BookPDF Available

Water quality: Indicators, human impact and environmental health

February 2013

Authors:

The University of Q

Water quality is fundamental for our health and affects the environment we share with other animals including marine, freshwater and terrestrial species. Water quality is often managed based on indicators for levels of bacteria and other chemical/physical contents. To assist in better management and monitoring of water quality, this book provides an overview of state of the art assessments of water quality; with an understanding how water quality is affected, and improving water quality for irrigation, drinking and recreation activities.

Content uploaded by You-Gan Wang

Content may be subject to copyright.

WATER RESOURCE PLANNING, DEVELOPMENT AND MANAGEMENT

WATER QUALITY

INDICATORS, HUMAN IMPACT

AND ENVIRONMENTAL HEALTH

Nova Science Publishers, Inc.

WATER RESOURCE PLANNING,

DEVELOPMENT AND MANAGEMENT

Additional books in this series can be found on Nova’s website

under the Series tab.

Additional E-books in this series can be found on Nova’s website

under the E-book tab.

ENVIRONMENTAL HEALTH - PHYSICAL,

CHEMICAL AND BIOLOGICAL FACTORS

Additional books in this series can be found on Nova’s website

under the Series tab.

Additional E-books in this series can be found on Nova’s website

under the E-book tab.

Nova Science Publishers, Inc.

WATER RESOURCE PLANNING, DEVELOPMENT AND MANAGEMENT

WATER QUALITY

INDICATORS, HUMAN IMPACT

AND ENVIRONMENTAL HEALTH

YOU-GAN WANG

EDITOR

New York

Nova Science Publishers, Inc.

transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical

photocopying, recording or otherwise without the written permission of the Publisher.

For permission to use material from this book please contact us:

Telephone 631-231-7269; Fax 631-231-8175

Web Site: http://www.novapublishers.com

NOTICE TO THE READER

The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or

implied warranty of any kind and assumes no responsibility for any errors or omissions. No

liability is assumed for incidental or consequential damages in connection with or arising out of

information contained in this book. The Publisher shall not be liable for any special,

consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or

reliance upon, this material. Any parts of this book based on government reports are so indicated

and copyright is claimed for those parts to the extent applicable to compilations of such works.

Independent verification should be sought for any data, advice or recommendations contained in

this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage

to persons or property arising from any methods, products, instructions, ideas or otherwise

contained in this publication.

This publication is designed to provide accurate and authoritative information with regard to the

subject matter covered herein. It is sold with the clear understanding that the Publisher is not

engaged in rendering legal or any other professional services. If legal or any other expert

assistance is required, the services of a competent person should be sought. FROM A

DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE

AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS.

Additional color graphics may be available in the e-book version of this book.

Library of Congress Cataloging-in-Publication Data

ISBN: 978-1-62417-111-6

Published by Nova Science Publishers, Inc.

New York

Nova Science Publishers, Inc.

CONTENTS

Preface vii

List of Contributors ix

Chapter 1 Water Quality Indices from Unbalanced

Spatio-Temporal Monitoring Designs 1

Sarah M. Raican, You-Gan Wang

and Bronwyn Harch

Chapter 2 Estimates of Likelihood and Risk Associated with Sydney

Drinking Water Supply from Reservoirs,

Local Dams and Feed Rivers 31

Ross Sparks, Gordon J. Sutton, Peter Toscas

and Rod Mc Innes

Chapter 3 Three-Dimensional Numerical Modeling of

Water Quality and Sediment-Associated

Processes in Natural Lakes 63

Xiaobo Chao and Yafei Jia

Chapter 4 Integrating Major Ion Chemistry with Statistical Analysis

for Geochemical Assessment of Groundwater

Quality in Coastal Aquifer of Saijo Plain,

Ehime Prefecture, Japan 99

Pankaj Kumar and Ram Avtar

Chapter 5 Suitability of Groundwater of Zeuss-Koutine Aquifer

(Southern of Tunisia) for Domestic and Agricultural Use 109

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli,

Rachida Bouhlila and Moncef Gueddari

Chapter 6 Application of Water Quality Indices (WQI) and Stable

Isotopes ( 18O and 2H) for Groundwater Quality

Assessment of the Densu River Basin of Ghana 131

Abass Gibrilla, Edward Bam, Dickson Adomako,

Samuel Ganyaglo and Hadisu Alhassan

Nova Science Publishers, Inc.

Contents

Chapter 7 Evaluation of Community Water Quality Monitoring

and Management Practices, and Conceptualization

of a Community Empowerment Model:

A Case Study of Luvuvhu Catchment, South Africa 161

L. Nare and J. O. Odiyo

Chapter 8 The Fate and Persistence of the Antimicrobial Compound

Triclosan and Its Influence on Water Quality 209

Teresa Qiu, Christopher P. Saint and Mary D. Barton

Chapter 9 Water Quality Assessment Methods: The Comparative Analysis 239

Tatyana I. Moiseenko, Alexandr G. Selukov

and Dmitry N. Kyrov

Chapter 10 Water Quality Impacts on Human Population Health in

Mining-and-Metallurgical Industry Regions, Russia 259

T. I. Moiseenko, N. A. Gashkina, V. V. Megorskii,

L. P. Kudryavtseva, D. N. Kyrov

and S. V. Sokolkova

Chapter 11 Chitosan Biopolymer for Water Quality Improvement:

Application and Mechanisms 273

Xinchao Wei and F. Andrew Wolfe

Index 295

Nova Science Publishers, Inc.

PREFACE

Water quality is fundamental for our health and affects the environment we share with

other animals including marine, freshwater and terrestrial species. Water quality is often

managed based on indicators for levels of bacteria and other chemical/physical contents.

To assist in better management and monitoring of water quality, we are proud to present

his book collecting 11 chapters providing the state of the art in assessment of water quality,

understanding how water quality is affected, and improving water quality for irrigation,

drinking and recreation activities.

Assessment of water quality is challenging due to spatial and temporal variability and its

multivariate nature arising from a number of parameters representing physical and biological

measures. Novel statistical approaches to establish representative annual indices are proposed

and illustrated by Raincan et al. (Ch. 1). These issues are also dealt with by Sparks et al. (Ch.

2) who focused on designing new monitoring programs.

The chemical process affecting water quality is investigated using a sediment-associated

process in lakes by Chao and Jia (Ch. 3). This is followed by a number of studies on

Groundwater quality. This is studied via ion chemistry by Kumar and Avtar (Ch. 4) in the

context of drinking consumption in Japan. Hamzaoui et al. (Ch. 5) also investigate

underground water quality based on a number of physical and chemical factors mainly for

irrigation purposes which will indirectly impact our health. Gibrilla et al. (Ch. 6) consider a

stable isotopes approach and show us how to establish quality indices for drinking and

agricultural purposes.

Water monitoring and management issues are well presented by Nare and Odiyo (Ch. 7).

Their chapter presented a very interesting study in South Africa, where the community takes

the major responsibility in monitoring and management.

Recycled or reclaimed water is studied by Qiu, Saint and Barton (Ch. 8) who provide a

comprehensive review on removal of triclosan. Water quality can also be assessed by

evaluating the physiological condition of fish, which provides a practically effective approach

for water monitoring (Moiseenko, Ch. 9). Water quality indices were also found correlated

with heavy metal accumulation in human bodies (Moiseenko, Ch. 10).

Finally, Wei and Wolfe (Ch. 11) present state-of-the-art technology in applying chitosan

for improving water quality by absorbing heavy metals.

Nova Science Publishers, Inc.

You-Gan Wang

viii

I sincerely believe that these 11 book chapters will be helpful for our researchers,

managers and other decision makers in understanding the current research topics and

important issues in making effective use of and managing our water resources.

You-Gan Wang, D.Phil. (Oxon)

Professor of Applied Statistics

Director, Centre for Applications in Natural Resource Mathematics (CARM)

School of Mathematics and Physics

The University of Queensland

Queensland 4072, Australia

Nova Science Publishers, Inc.

LIST OF CONTRIBUTORS

You-Gan Wang, Water Quality Indices from Unbalanced Spatio-temporal Monitoring

Designs. E-mail: you-gan.wang@uq.edu.au.

Ross Sparks, Estimates of Likelihood and Risk Associated with Sydney Drinking Water

Supply from Reservoirs, Local Dams and Feed Rivers. E-mail: Ross.Sparks@csiro.au.

Xiaobo Chao, Three-dimensional Numerical Modeling of Water Quality and Sediment-

Associated Processes in Natural Lakes. E-mail: chao@ncche.olemiss.edu.

Pankaj Kumar, Integrating Major Ion Chemistry with Statistical Analysis for

Geochemical Assessment of Groundwater Quality in Coastal Aquifer of Saijo plain, Ehime

prefecture, Japan. E-mail: pankajenvsci@gmail.com.

F. Hamzaoui, Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia)

for Domestic and Agricultural Use. E-mail: fadoua_fst@yahoo.fr.

Abass Gibrilla, Application of Water Quality Indices (wqi) and Stable Isotopes (δ18O and

δ2H) for Groundwater Quality Assessment of the Densu River Basin of Ghana.

E-mail: gibrilla2abass@yahoo.co.uk.

L. Nare, Evaluation of Community Water Quality Monitoring and Management

Practices, and Conceptualization of a Community Empowerment Model: A Case Study of

Luvuvhu Catchment, South Africa. E-mail: leratonare@yahoo.com.

Chris Saint, The Fate and Persistence of the Antimicrobial Compound Triclosan and its

Influence on Water Quality. E-mail: Christopher.Saint@ unisa.edu.au.

T. I. Moiseenko, Water Quality Assessment Methods: the comparative analysis.

E-mail: moiseenko.ti@gmail.com.

T. I. Moiseenko, Water Quality Impacts on Human Population Health in Mining-and-

metallurgical Industry Regions, Russia. E-mail: moiseenko.ti@gmail.com.

Xinchao Wei, Chitosan Biopolymer for Water Quality Improvement: Application and

Mechanisms. E-mail: weix@sunyit.edu.

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 1

WATER QUALITY INDICES FROM UNBALANCED

SPATIO-TEMPORAL MONITORING DESIGNS

Sarah M. Raican1, You-Gan Wang1,2,

∗

and Bronwyn Harch1,3

1CSIRO Mathematics, Informatics and Statistics, Australia

2Centre for Applications in Natural Resource Mathematics (CARM), School of

Mathematics and Physics, The University of Queensland, Australia

3CSIRO Sustainable Agriculture Flagship, Australia

ABSTRACT

This chapter investigates a variety of water quality assessment tools for reservoirs with

balanced/unbalanced monitoring designs and focuses on providing informative water quality

assessments to ensure decision-makers are able to make risk-informed management decisions

about reservoir health.

In particular, two water quality assessment methods are described: non-compliance

(probability of the number of times the indicator exceeds the recommended guideline) and

amplitude (degree of departure from the guideline). Strengths and weaknesses of current and

alternative water quality methods will be discussed. The proposed methodology is particularly

applicable to unbalanced designs with/without missing values and reflects the general

conditions and is not swayed too heavily by the occasional extreme value (very high or very

low quality).

To investigate the issues in greater detail, we use as a case study, a reservoir within

South-East Queensland (SEQ), Australia. The purpose here is to obtain an annual score that

reflected the overall water quality, temporally, spatially and across water quality indicators

for each reservoir.

∗ Corresponding author.

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

1. INTRODUCTION

Water quality monitoring is becoming increasingly important in order to meet public

health and safety regulations, for the protection and sustainability of our natural resources, as

well as to meet local and state government guidelines and regulations. Appropriate

assessment tools are required to summarize the annual health of the reservoir, accommodating

seasonal effects, health/environmental guidelines and the often unbalanced nature of the

monitoring design.

Water quality assessment tools that summarize measurements taken as part of a

monitoring program are a simple and concise way of providing information about the health

of the resource. There are a number of general issues that need to be considered when

undertaking water quality assessments including:

The question/reason for the water quality assessment;

Desired timescale for reporting water quality;

Seasonal differences;

Sampling frequency/timescale within the period of interest;

Balanced/unbalanced designs;

The choice of water quality indicators (variables which reflect the health of the

reservoir);

Selection of sites;

Data quality; and

Impact of missing values.

However, evaluating overall water quality status from a large number of variables and

samples is often difficult [Kannel et. al, 2007]. The type of water quality assessment

technique used depends on the objective of the monitoring program, the specific questions

being asked and the implemented monitoring/sampling design. Often, the design is required

to fit within financial budgetary constraints or relies on using existing data and monitoring

designs.

Extra consideration therefore needs to be given to the impact on water quality assessment

when dealing with unbalanced designs (i.e. increased sampling during certain periods/months

and/or an inappropriate sampling frequency that does not adequately capture the temporal

variation) and/or missing values.

The type of water quality assessment technique used depends on, the objective of the

monitoring program, the specific questions being asked and the implemented

monitoring/sampling design. Often, the design is required to fit within financial budgetary

constraints or relies on using existing data and monitoring designs. Extra consideration

therefore needs to be given to the impact on water quality assessment when dealing with

unbalanced designs (i.e. increased sampling during certain periods/months and/or an

inappropriate sampling frequency that does not adequately capture the temporal variation)

and/or missing values.

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 3

1.1. Current Approaches to Water Quality Assessment

A water quality index (WQI) is one assessment technique that enables a complex system

of water quality measurements to be summarized by a single number. This simple tool aims to

inform managers and decision makers about the overall water quality status, improves general

understanding of the water quality issues and is therefore likely to result in investment to

effectively manage our water resources [Cude, 2001]. These indices can be used to assess

changes and trends in overall water quality [Cude, 2001]. Sarkar and Abbasi [2006] identify

four main steps in the development of a water quality index:

1. Selection of an optimum set of parameters (indicators)

2. Transformation of parameters onto a common scale

3. Assignment of parameter weights

4. Aggregation of scores to form a final, single score.

A water quality index has been developed within an Ecosystem Health Monitoring

Program (EHMP) in South-East Queensland (SEQ). The EHMP [2007] was created as part of

the South-East Queensland Regional Water Quality Management Strategy [SEQRWQMS;

Smith and Storey, 2001] to monitor and report on the ecosystem health of freshwater systems.

A report card score (grade) is given to a number of reporting areas and is based on a

corresponding water quality index for each area. Indicators of ecosystem health were chosen

that responded to known disturbance gradients and a set of ecosystem health guidelines

developed using a set of minimally disturbed reference sites [Smith and Storey, 2001]. The

report card is calculated each year using samples taken during spring and the following

autumn and standardises for both spatial variability and measurement scale. The

standardisation of parameters occurs by calculating the ratio of the distance between the

measurement and the guideline to the distance between the worst-case scenario (WCS) and

the guideline [EHMP, 2007].

The WCS is calculated to be a theoretical limit or the 10th/90th percentile of data for all

sites and assessment periods for a given stream type. Calculation of the guideline and WCS

values for each stream group with similar physical conditions accommodates spatial

variability. A final water quality index is calculated by taking a series of arithmetic means

across various groupings.

Calculating the final score in this way means that each indicator, index and season are

given equal weighting in the final score regardless of how many indicators are within each

index. This approach is appropriate if few missing values occur, the sampling design is

balanced and there is likely to be minimal differences in the WCS levels between seasons or

data is collected from only one season. If this is not the case, then alternative scoring

procedures may be needed.

Instead of considering just one aspect of water quality, the Canadian Council of Ministers

of the Environment (CCME) develop a WQI taking account of: scope, frequency and

amplitude [CCME, 2001], with each factor ranging between 0 and 100. The use of more than

one aspect allows a more representative view of the water quality. Scope measures the

percentage of water quality variables for which the guideline is not met in at least 1 sample

during the period of interest.

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

Frequency measures the percentage of measurements in which the guideline was

exceeded and amplitude measures the departure from the guideline for measurements that did

not meet the guideline. The amplitude is computed by firstly calculating the number of times

the measurements do not meet the guideline (excursion). The normalized sum of excursions is

then calculated by dividing the excursion by the total number of observations. Finally an

amplitude value is computed by using an asymptotic function that scales the normalized sum

of excursions to range between 0 and 100.

The final water quality index is then the Euclidean distance of the three scores scaled by a

constant so the final score also ranges between 0 and 100. The scope, frequency and

amplitude all rely on information about the objective or guideline value. All

indicators/indicator values are combined in the calculation of the scope, frequency and

amplitude so there is no need to use averaging, weighted sum or the like to combine the

scores across indicators or water quality categories.

Many indices are however developed for each indicator (or sub-index) and therefore

require aggregation to form a single final score. Sarkar and Abbasi [2006] note that many

different variants of aggregating scores for indicators (or sub-indices) are used including

weighted sums, weighted geometric means, weighted products, an unweighted harmonic

mean square formula and modified additive aggregation functions. Aggregation across

multiple indicators and/or index categories and the combining of multiple water quality

aspects (i.e. scope, frequency and amplitude etc.) requires the indices/scores to be on a

common scale.

Scores are often normalised or scaled using a linear/non-linear function. Sargaonkar and

Deshpande [2003] develop non-linear relationships for each indicator to transform the

indicator values to a common scale ranging between 0 and 16. Each curve is constructed

using predetermined indicator limits for 5 water quality categories (excellent, acceptable,

slightly polluted, polluted and heavily polluted). More commonly the scores are scaled to be

on a range of 0 to 1 or 0 to 100. EHMP [2007] transform their scores on a 0 to 1 scale while

the CCME [2001] transform the scores onto a 0 to 100 scale. Swamee and Tyagi [2000] look

at different aggregation procedures and find that as result of the type of aggregation some

indices may not reflect an overall poor water quality when a subindex shows poor water

quality. Other aggregation indices may show the aggregated index to be unacceptable even

though the water quality is in fact acceptable, while others fail to give a complete picture of

the water quality. The choice of aggregation should depend on the reason for the summary

measure, which is related to the objectives of the monitoring program and the specific

questions being answered. Water quality indices can however be calculated without the need

for computation and aggregation of sub-indices. Swamee and Tyagi [2000] list a number of

approaches taken by various researchers including: factor analysis, principal component

analysis, fuzzy clustering analysis, multivariate ranking procedure and uniformity indexing

method. Said et al. [2004] also develop a water quality index not requiring aggregation

techniques, sub-indices or standardization procedures. Instead the index is a simple numerical

equation that is a function of the indicators with appropriate weights determined by level of

significance and effect of indicator on the water quality. Such approaches avoid the need for

and difficulties associated with aggregating across sub-indices but cannot reflect short-term

changes and localised changes may not be immediately reflected in the score [Said et al.,

2004].

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 5

1.2. Limitations and Strengths

Much thought has been given to index calculation, aggregation across sub-indices and

transformation of indices to a common scale, however there seems to be little research on the

development of indices for unbalanced designs or designs with missing values. There seems

to be even less research on standardising scores through time. This is particularly important

when dealing with multiple samples taken across the year in an annual water quality

assessment.

Many water quality variables have seasonal cycles with peaks occurring at particular

times of the year. Although an observed value may not be extreme during peak times, if

observed during a period when conditions are normally quite low and stable, it could indicate

potential problems with the water quality. When dealing with multiple samples in a given

period we therefore believe that seasonality should be taken into account and the water

quality index standardised through time.

Although standardised in both space and scale, it appears that the EHMP freshwater score

has not been standardised through time. If significant changes in water quality could be

expected between the spring and autumn samples then perhaps the score should also be

standardised across seasons.

According to Swamee and Tyagi [2000] this perhaps reduces the possibility that a poor

sub-index will be reflected in a poor overall water quality index. However aggregating in this

way enables the user to easily trace the reason for a poor/excellent water quality score

through the various stages of aggregation. The strength of the CCME WQI is its ability to

measure not one but three different aspects of water quality and little aggregation is required

as the sub-indices are calculated across all water quality indicators therefore removing the

need for aggregation.

However, no averaging makes it difficult to trace which water quality indicator/(s) has

(have) the greatest impact on the final index, which may impact on a decision makers ability

to specifically target management actions. The WQI calculations also allow regular water

quality monitoring data to be used where multiple samples may be recorded during the period

of interest (i.e. once per month for an annual index). However there appears to be little

consideration of the impact when the design is unbalanced and missing values occur or when

a particular season or period may be measured more frequently than another within the period

of interest. Seasonality is another aspect which does not appear to be well addressed within

this or many other articles. The need for seasonality depends on the reason for the water

quality assessment and the question being answered.

1.3. New Approaches to Water Quality Assessment

This Chapter investigates a variety of water quality assessment tools for reservoirs with

balanced/unbalanced monitoring designs and focuses on providing informative water quality

assessments to ensure decision-makers are able to make risk-informed management decisions

about reservoir health. In particular, two water quality assessment methods are described:

non-compliance (probability of the number of times the indicator exceeds the recommended

guideline) and amplitude (degree of departure from the guideline). Strengths and weaknesses

of current and alternative water quality methods will be discussed and a case study will be

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

provided from a local reservoir within South-East Queensland, Australia. The Queensland

Bulk Water Supply Authority (trading as Seqwater) manages and supplies water within SEQ.

In total there are 14 sites that Seqwater sample as part of their campaign/drought monitoring

program on this particular reservoir. A large number of water quality variables are monitored

with some monitored since 1997.

Water quality samples are usually taken from the reservoir once or twice per month;

however this does vary and can occur up to four times per month (or more on some

occasions). The reasons for the increased samples are varied but may include increased

sampling efforts during an event or increased sampling effort in periods of known high

variability. Samples may also be missing at some sites for some months therefore also

contributing to the unbalanced design problems.

The purpose for developing a new approach to water quality assessment was to calculate

an annual score that reflected the overall water quality, temporally, spatially and across water

quality indicators for each reservoir. The proposed methodology is applicable to unbalanced

designs with/without missing values and reflects the general conditions and is not swayed too

heavily by the occasional extreme value (very high or very low quality).

2. WATER QUALITY ASSESSMENT METHODS

We investigate a number of water quality assessment methods for the purpose of

quantifying the health of the reservoir and investigate temporal and spatial components of the

report card scoring procedures. Our aim is to provide an annual water quality assessment tool

for decision makers to enable them to make risk-informed management decisions. We believe

a two-step process is necessary to achieve this aim:

1. Calculate the frequency of samples exceeding the guideline (compliance);

2. Calculate the distance of exceedance from the guideline (amplitude).

Compliance is a common water quality assessment tool and is usually assessed using

known public health or environmental guidelines. Non-compliance is a measure of how often

water samples exceed these guidelines. Compliance/non-compliance is an important

management tool, however to make targeted management decisions we believe that a measure

of non-compliance is not sufficient on its own. We believe it is equally important to

understand the magnitude of the non-compliance, that is, the degree of departure from

compliance (amplitude). Different management actions may be required depending on the

non-compliance and amplitude scores. It is possible for two different years to have the same

non-compliance but have very different amplitudes (see Figure 1). The combination of both

non-compliance and amplitude measures therefore provides greater insight into the water

quality and gives deeper insight for the decision-makers. A non-compliance measure indicates

whether the guidelines are exceeded frequently, regularly or occasionally. The amplitude

informs on the extremity of the values that exceed the guideline.

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 7

Figure 1. Boxplots showing an example distribution of values for two separate years. Both distributions

have approximately the same non-compliance (a) 50% of samples > guideline but have very different

amplitudes (b) quantile95(year 1) > quantile95(year 2). The median of the distribution is indicated by

the solid black line and the guideline by the dashed line.

2.1. Non-Compliance Scores

We investigate four different methods for calculating non-compliance and detail the

strengths/weaknesses of each method and their most appropriate use.

The first and most simple non-compliance method (NC1) can be calculated as:

1above

total

NC N

, (1)

where Ntotal is the number of samples taken for a particular water quality indicator during the

period of interest (annual) and Nabove is the number of these samples that exceed the

recommended guideline. If there is more than one site being monitored, non-compliance is

calculated separately for each site. Figure 2 shows the number of samples above and below

the Queensland Water Quality guideline [Queensland, 2006] for Filtered Reactive

Phosphorus.

This scoring procedure can be used with a small or large number of samples and could

still be calculated even when only one sample is collected throughout the year or period of

interest. Repeat samples or periods of increased monitoring (typically likely to occur during

periods of increased variability) can bias the score towards the scores during the

months/periods of increased monitoring. For example, if a particular indicator is sampled

more during the summer period which is known to have increased variability and more

extreme levels/concentrations than the bias may tend towards higher non-compliance scores.

A score derived directly on this type of data is therefore likely to reflect the conditions during

those months/periods that are highly monitored and not necessarily reflect the overall

condition of the storage throughout the entire year. Bias could also result if there are missing

samples throughout the year and the range of temporal conditions is not adequately captured.

year 1 year 2

Indicator values

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

Missing data from periods prone to be more non-compliant may be biased towards smaller

non-compliance scores. This method is therefore probably more appropriate to continuous,

very regular sampling that is evenly spaced throughout the year to capture the range of

temporal conditions.

Figure 2. Data taken from 11 sites on a local reservoir for the 2006/2007 year for Filtered Reactive

Phosphorus taken at the surface using a 3m integrated tube. A non-compliance score (NC1) is

calculated separately for each site.

In order to remove/reduce bias resulting from over-sampling, under-sampling or missing

values an average of the proportion of values above the guideline each month (or period)

could be calculated (NC2):

NC p=, (2)

where pm is the proportion of samples that exceed the guideline during month, m. Every

month within the year has equal weighting and is therefore not biased by periods of increased

monitoring (see Figure 3). When each month is sampled an equal number of times then NC2

is simply equal to NC1. When the monitoring/sampling design is unbalanced then the NC2

score appears to be a more appropriate method for calculating non-compliance. As each

month is equally weighted in the final non-compliance score, the overall water quality

throughout the year is much better represented. This method is suited to both unbalanced and

balanced sampling designs and does not require samples to be taken at equal time intervals

throughout the year. The weakness is that the year (or period of interest) must be divided into

categories that represent the range of temporal conditions seen throughout the year and

averaged across categories. Too many missing months may also still create bias in the non-

compliance scores.

0.005 0.010 0.015 0.020

Filtered Reactive Phosphorus (mg/L)

18Jul2006 26Oct2006 3Feb2007 14May2007

Si te 1

Si te 2

Si te 3

Si te 4

Si te 5

Si te 6

Si te 7

Si te 8

Si te 9

Si te 10

Si te 11

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 9

Figure 3. Calculation of NC2 occurs by calculating the proportion of values exceeding the guideline for

each month (bounded by the gray lines) at each site. The months are then averaged within each site to

derive a non-compliance score for each site.

The last two non-compliance scores (NC3 and NC4) aim to overcome difficulties caused

by both unbalanced sampling designs and missing values. The first of these two approaches

uses linear interpolation between sampling points. This method assumes that the values

between samples can be approximated by a linear function of time. Instead of using the raw

samples we then use the interpolated values to calculate the non-compliance score to be:

3above

total

NC L

(3)

where Ltotal is the total length of time (i.e. 365 days for an annual score) and Labove is the

length of time the interpolated function is above the guideline (see Figure 4). This is simply a

modified version of Equation 1 that uses interpolated values rather than the collected samples.

This method represents the range of temporal variation without the need for categorisation

into smaller sub-periods and does not require the sampling to be equally spaced throughout

time. Multiple or repeat samples (samples taken at the same time) are simply averaged to

create a single value for each sampling time. This method can be used with a large or small

sample size throughout a given year however the more samples collected the greater the

accuracy in interpolation. This method can be used for both balanced and unbalanced

sampling designs. Bias may still result when a large proportion of the year is not adequately

sampled, however smaller proportions of the year can be linearly interpolated using the next

available adjacent samples. Bias may also result when samples are missing from key times

within the period of interest (i.e. times of significant decrease/increase in

concentrations/levels). For this reason we still recommend that sampling is done as regularly

and consistently as possible.

In the case where we may have upper and lower guideline limits (e.g. % saturation of

Dissolved Oxygen: 90 – 110% recommended range) we make the additional assumption that

the direction from which the guideline is exceeded is not important (being too high or too low

is of equal importance). Before calculating NC3 (Equation 3) we calculate

0.005 0.010 0.015 0.020

Filtered Reactive Phosphorus (mg/L)

18Jul2006 26Oct2006 3Feb2007 14May2007

Site 1

Site 2

Site 3

Site 4

Site 5

Site 6

Site 7

Site 8

Site 9

Site 10

Site 11

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

*()/2

VVul=−+ where u and l are the upper and lower guidelines respectively and

redefine the guideline according to the transformed data ()/2Gu ul

−+ . Calculation of

Labove and Ltotal is then based on the transformed values (Vi

*) and redefined guideline. A more

elegant approach for calculating non-compliance is to use a smooth function of time rather

than a linear interpolation between samples. We investigate fitted smooths using the family of

Generalized Additive Models [GAMs; Wood, 2006] where the response distribution belongs

to the family of exponential dispersion models (EDM). If samples are only taken at a single

site than a smooth could be fit to just one site (provided there is sufficient data). Alternatively

the following model could be fit:

() ()

gu st

=++ (4)

where s(t) is a spline-based smooth to represent the trend throughout the year, βj is a

categorical variable representing the shift in mean at site j and g(.) is a link function. It seems

reasonable to initially assume that sites within the same lake may have similar temporal

trends with a shift in site mean. In fitting the smooth (Equation 4), strength can be borrowed

from other sites when sites are missing samples. This helps with the infilling and interpolation

of values required for calculating the non-compliance score. Bias is reduced as missing values

are replaced by interpolated or infilled values and equal weights are then applied throughout

the year.

Figure 4. Linear interpolations for (a) all sites and (b) site 8. Non-compliance can be calculated as the

proportion of time the interpolations are above the guideline (length of gray arrow) in reference to the

total length of time (length of black and gray arrows).

0.005 0.010 0.015 0.020

Fil tered Reactive Phosphorus (mg/L)

18Jul2006 2 6Oct2006 3Feb2007 14 May2007

Site 1

Site 2

Site 3

Site 4

Site 5

Site 6

Site 7

Site 8

Site 9

Site 10

Site 11

0.005 0.010 0.015 0.020

Filt ered Reactive Phosphorus (mg/L)

18Jul2006 2 6Oct2006 3Feb2007 14 May2007

Site 8

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 11

Figure 5 however shows, there are some limitations with this. Some curves are

overestimated and some are underestimated resulting in non-compliance scores for some sites

greater than non-compliance based on the raw samples and some non-compliance scores less

than expected based on raw samples. An alternative would be to fit a separate smooth for

each site; however this may require a substantial amount of additional samples to enable the

model to be appropriately fitted. This underestimation at sites and overestimation at others

may be may be averaged out when the scores are combined across all sites. It is also possible

to adjust the smoothness parameter, however this would require user interaction and for all 16

variables could be very time consuming. As we also intend to automate the generation of

report card scores in the future we want as little user interaction as possible.

The non-compliance score (NC4) can be calculated as:

4above

total

NC S

(5)

where Sabove is the length of time the smooth exceeds the recommended guidelines and Stotal is

the length of the period of interest (i.e. 365 days for annual scores).

Figure 5. Spline interpolations (calculated by fitting a GAM) for (a) all sites and (b) site 8. Non-

compliance can be calculated as the proportion of time the spline curves are above the guideline (length

of gray arrow) in reference to the total length of time (length of black and gray arrows).

0.005 0.010 0.015 0.020

Date

Filtered Reactive Phosphorus (mg/L)

18Jul2006 26Oct2006 3F eb2007 14May2007

Site 1

Site 2

Site 3

Site 4

Site 5

Site 6

Site 7

Site 8

Site 9

Site 10

Site 11

0.005 0.010 0.015 0.020

Filt Reac Phosphorus (mg/L)

18Jul2006 26Oct2006 3F eb2007 14May2007

Site 8

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

Unfortunately, this model requires a reasonable sample size to fit an adequate model and

this method is more complex and computationally intensive. Difficulties may also be

encountered when a few sites experience unusually high concentrations or levels in

comparison to the other sites and there aren’t sufficient samples to fit a separate smooth for

each site. The fitted/interpolated values may be smoothed out by the values from other sites

so these extreme values may not be adequately captured. This is simply the nature of the

model and this does not invalidate the methodology, however making the procedure robust is

desirable.

Where upper and lower guideline limits are given, the data is transformed (Vi

*) and Sabove

and Stotal calculated using the transformed values (as explained for NC3) and adjusted

guideline value.

Table 1. Table summarises strengths, weaknesses and when the conditions under which

the non-compliance methods are recommended for use

Non-

Compliance

Method

Strengths Weaknesses Recommended for use when …

NC1 Simple and easy to calculate.

Can be calculated with small or

large sample sizes.

Easily biased by repeat samples

or periods of increased

monitoring.

May not reflect the overall

conditions throughout the

period of interest.

Missing values or under-

sampling may mean temporal

conditions throughout period

are not adequately captured.

The sampling is balanced,

continuous, very regular and

evenly spaced throughout the

period of interest with few

missing values.

NC2 Reduces bias due to over-

sampling, under-sampling and

a few missing values.

Every month throughout the

year has equal weighting.

Accounts for unbalanced

sampling designs.

The year (or period of interest)

must be divided into discrete

categories that represent the

range of temporal conditions.

Too many missing values can

be problematic.

The sampling design is

balanced or unbalanced, and

unevenly spaced throughout the

year with few missing values.

NC3 Represents range of temporal

variation.

No need to categorise into sub-

periods.

Multiple or repeat samples do

not bias results.

Biased is reduced when

missing values are present.

Can be used with small/large

sample sizes.

Appropriate for

balanced/unbalanced designs.

Assumes values between

samples are a linear function of

time.

Biased when a large proportion

of the year is not sampled.

Bias may result when samples

are missing from key times

within period of interest.

The sampling design is

unbalanced, irregular and

unevenly spaced throughout the

year with missing values.

Sampling at key times

(highs/lows) throughout period

will improve model

NC4 Represents range of temporal

variations.

Smooth functions can be used

rather than linear interpolations

between samples.

In fitting the smooth, strength

can be borrowed from other

sites when sites are missing

Sites are assumed to have

similar temporal trends or there

has to be a large number of

samples at each site to fit

separate curves for each site.

Complex and computationally

intensive.

Smooths will probably

The sampling design is

unbalanced, irregular and

unevenly spaced, there are

missing values and a large

number of samples at each site.

Sampling at key times

(highs/lows) throughout the

period of interest will improve

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 13

Non-

Compliance

Method

Strengths Weaknesses Recommended for use when …

samples.

No need to categorise into sub-

periods.

Multiple or repeat samples do

not bias results.

Bias is reduced when missing

values are present.

Appropriate for

balanced/unbalanced designs.

underestimate non-compliant

extremes.

Smoothness of curves

determined by choice of a

smoothness parameter.

the model.

2.2. Amplitude Scores

The second component of this water quality assessment involves calculating amplitude

scores (magnitude of non-compliance). We investigate six alterative amplitude scoring

methods. As the non-compliance scores by definition range from 0 to 1 where 0 indicates all

values were within the recommended guidelines and 1 indicates 100% non-compliance then

the amplitude scores will be developed on a similar scale of 0 to 1.

The first and simplest method for calculating amplitude (D1) is to calculate the observed

distance of each value from the guideline:

dVG=−

or expressed as the number of guidelines from the recommended guideline:

−

Both of these scores need further scaling so the amplitude ranges between 0 and 1 where

1 indicates most extreme distance from the guideline. To scale this we need information about

the limits or maximum values for each water quality indicator being assessed (akin to worst

case scenario (WCS) and guideline values used by the EHMP freshwater scoring). These

limits, dWCS and gWCS can be calculated as the 10th/90th percentile of di and gi respectively. The

10th percentile value is calculated if lower levels/concentrations indicate poor water quality.

Alternatively, the 90th percentile is used if higher concentrations are indicative of poor water

quality. For the occasional times when the 10/90th percentile is exceeded the score is scaled

back to the maximum score of 1.

The scaled amplitude scores therefore are:

1ii

WCS WCS

D median median

⎛⎞ ⎛⎞

⎜⎟ ⎜⎟

⎝⎠ ⎝⎠

(6)

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

The limits could be based on a single limit that is representative of the limit throughout

the year or could be a seasonally changing limit. Given the strong presence of seasonality

seen within many of the water quality indicators, an amplitude measure that calibrates

extremes for each season would seem to be more appropriate. An extreme value during the

summer period may be quite different to what would be classed as extreme during the winter

period. Extremes are therefore relative to the natural seasonal cycles/peaks that occur

throughout the year. One disadvantage of deriving a limit for each season/month however is

that there must be sufficient samples during each season/month in order to calculate the

limits. An alternative approach is to model the limit as a continuous function through time

using a quantile regression technique on all available data from all sites. The occasional

season/month with small or no samples will be interpolated with greater weight on

neighbouring seasons. This method will be discussed in more detail later.

Like the non-compliance method NC1 this amplitude measure suffers bias caused by

missing values, unbalanced designs and increased monitoring during particular periods of

time. If greater emphasis should be placed on those months with increased monitoring then

this bias is not problematic. However, our aim is to assess the overall condition of the lake

throughout the year and so alternative methods to reduce this bias would be more appropriate.

The second amplitude measure avoids the need to determine limits for each indicator.

Instead of a limit we simply determine the proportion of values above the guideline (non-

compliance) and the proportion of values above a second level which is highly undesirable to

exceed without severe ramifications. The second amplitude measure is calculated to be:

(1 2)

21 2

Dp +

=×

(7)

where p1 is the proportion of samples that are above the guideline and p2 is the proportion

above a second level which indicators should not exceed (see Figure 6). A score of 0 would

indicate all samples were below the recommended guideline (p1) while a score of 1 would

indicate all values exceed the second level (p2). This categorical approach already contains a

combination of non-compliance and amplitude in the one score. A second level could be

chosen to be a higher warning level or guideline or could be subjectively chosen for each

indicator using expert knowledge. For examples within this Chapter we have chosen the

second level to be twice the recommended guideline. As with most categorical methods, there

is some information loss, in this case we do not use the actual distances from the guidelines

just their location in respect to the guideline and a second guideline. A significant change in

scores may also result from a change in the value of the second guideline.

Another alternative to amplitude measures requiring calculation of indicator limits is the

probability associated with the location of the median distance in a reference distribution.

This reference distribution could be an empirical distribution based on historical data from a

number of years and sites or could be a known theoretical distribution. The third amplitude

score (D3) can be calculated as:

3Pr( )

median i

dist d=≤

, (8)

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 15

where di = Vi – G for Vi > G (see Figure 7).

This method is appropriate for ascertaining the typical conditions for the water body. The

reference distribution may need to be altered if we want to represent what is happening

throughout the entire period of interest. Alternatively, if we want to assess how far the

extreme values are from the guideline perhaps the 90th percentile distance should be used

instead of the median distance.

Figure 6. Calculation of an amplitude score using method 2. Two solid lines represent the first and

second guidelines, G1 and G2 respectively.

Figure 7. The empirical cumulative distribution (reference distribution) of distance values above the

guideline based on all years across the 3 dams for FRP (bottom). The bold line indicates the median

distance above the guideline (D3) for site 8 and shows the corresponding cumulative probability of D3

when compared to the reference distribution.

An alternative method uses the amplitude score from D4 to derive a corresponding

probability from a beta distribution with mean 0.5. The beta distribution is more appropriate

then D3 as we are able to incorporate the uncertainty associated with the original proportion

(D3) in the calculation of the score. The added benefit is that the transformation also results in

a proportion between 0 and 1. No scaling is necessary to create the index score and the beta

distribution is very flexible in shape. The fixed points in the beta distribution are 0, 0.5 and 1

0.005 0.010 0.015 0.020

FRP (mg/L)

18Jul2006 26Oct2006 3Feb2007 14May2007

Site 1

Site 2

Site 3

Site 4

Site 5

Site 6

Site 7

Site 8

Site 9

Site 10

Site 11

0.00 0.01 0.02 0.03 0.04

0.0

0.2

0.4

0.6

0.8

1.0

Distance above guideline (mg/L)

Cumulative probability

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

such that a proportion of D3=0 will map to D4=0, D3=0.5 will map to D4=0.5 and D3=1 will

map to D4=1. The cumulative distribution function for the beta distribution is:

(,)

(; , ) (, )

(,)

Bab

Fxab I ab

Bab

(9)

where B(a,b) is the beta function, Bx(a,b) is the incomplete beta function and Ix(a,b) is the

regularized incomplete beta function. The fourth amplitude score can be calculated as:

4(3,,)DFDaa= (10)

where a can be chosen as (n+1)/2 or other informative values to reflect the variability in n,

where n is the number of values used to calculate distmedian (see Figure 8).

Although both D3 and D4 do not require upper limits (or worst case scenarios), both are

prone to bias as a result of increased monitoring during certain periods (unbalanced design)

and missing values. Alternative methods that give equal weight throughout the entire period

of interest and methods that are more robust against missing values and unbalanced designs

would be more appropriate.

Figure 8. Probability from Beta distribution (mean = 0.5) used as an amplitude score D4. Bold line

indicates D3 score at site 8 and used to calculate the corresponding p-value from a beta distribution.

Two methods for calculating non-compliance (NC3 and NC4) are developed to handle

unbalanced designs and it is these approaches that are adapted to also provide our final two

amplitude scores (D5 and D6). The unbalanced design could be a result of missing values or

additional sampling. We again use the interpolated/infilled data instead of the collected

samples. However limits will still be required for each indicator to scale the amplitude

measurements to range between 0 and 1 inclusive.

For similar reasons as discussed previously, we would like the limits to reflect the

seasonal changes throughout the year. We could simply use the 10th/90th percentile of values

that exceed the guideline each month. As the amplitude scores use continuous functions it

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.5

1.0

1.5

probability (D4)

Beta distribution function

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 17

would be more appropriate to develop smooth functions for the limits. One such option

involves using quantile regression techniques [Koenker and D’Orey, 1987; Koenker et. al.,

1994].

Quantile regression techniques (parametric and non-parametric) have wide use in

environmental modelling. Modelling of droughts and floods (rainfall extremes) requires

knowledge of extreme quantiles to assist in the design of reservoirs, flood drains and run-offs

[Yu et al, 2003]. Morgan II et. al. [2006] uses the 50th quantile regression for modelling the

benthic index of biotic integrity and fish index of biotic integrity as functions of other water

quality parameters. As many water quality indicators do not have a Gaussian distribution and

transformations may be inadequate a non-parametric technique would be more suitable. Little

work has been done on non-parametric quantile regression techniques and software for such

techniques is very limited. Pin and Maechler [2008] have developed a non-parametric

quantile regression function for R [see http://www.r-project.org/]. This package computes

constrained B-Splines (COBS) non-parametric regression quantiles using linear or quadratic

splines [Bartels and Conn, 1980; Ng, 1996; Koenker and Ng, 1996; He and Ng, 1999;

Koenker and Ng, 2005]. As this method is non-parametric no distributional assumptions on

the errors are needed. The non-parametric quantile regression technique also allows us to

perform analyses without having to categorise the data into seasons/months. By using this

quantile regression technique we avoid the need for categorisation which often results in

information loss while at the same time still accounting for seasonality. Amplitude scores D5

and D6 will therefore use the 95th non-parametric quantile regression based on historical data

for indicator limits.

An equivalent amplitude score based on the linear interpolation technique developed for

NC3 can be computed as:

{

}

{}

()

5max ( ) ,0

kftG

tG dt

−

=−

∑∑

∫, (11)

where fk(t) is the linear function connecting the kth and (k+1)th sample points (Vi) sorted in

time, f95(t) is the 95th non-parametric quantile regression and G is the guideline value (see

Figure 9). Approximations to the sums/integrals are calculated by summing the distance of

the fitted values from the guideline at a smaller timescale (e.g. days if creating an annual

report).

If upper and lower guideline limits exist then the transformed data (Vi

*) is used in

calculating fk(t) and f95(t) rather than the observed values (Vi). The guideline should also be

adjusted to this transformed scale (see NC3 and NC4).

Although this method may be more computationally intensive then the previous methods

for amplitude, there will be reduced bias when the design is unbalanced and there are missing

values for some samples. The method can therefore be used for both balanced and unbalanced

designs so long as there are enough samples to detect the general trend throughout the period

of interest. Alternatively a smoothing approach (similar to that used for calculating NC4)

could be used when non compliance score NC4 is used. Like the linear interpolation

methodology, we simply calculate the ratio of areas underneath the limit and smooth based on

samples for the year/period of interest. The sixth amplitude score can be calculated as:

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

(

)

()

max ( ) , 0

6max ( ) , 0

tG dt

−

=−

∫

∫, (12)

where f(t) is the spline curve fitted to the observed samples as a function of time, t, in a GAM

framework and f95(t) is the non-parametric 95th percentile regression and G is the

recommended guideline. Approximations can be used to calculate the area underneath each of

the curves (see amplitude method D5). The smoothness is controlled and set as a default

parameter for all variables. This may seem too simple, however as the procedure is possibly

to be automated we don’t want to implement procedures that require intensive user input.

Transformation of the data (see D5) will again be necessary when we have a range for the

recommended guidelines. The functions f(t) and f95(t) are calculated on the transformed data

(Vi

*) rather than observed values (Vi) as was done for amplitude method D5.

Figure 9. Amplitude measure calculation (D5) for FRP at site 8 on a local reservoir. The distance value

is the ratio of the dark shaded are to the light shaded area.

Table 2. Table summarises strengths, weaknesses and when the conditions under which

the amplitude methods are recommended for use

Amplitude

Method

Strengths Weaknesses Recommended for use

when

D1 Simple

Can be calculated with small or large

sample designs.

Bias caused by

unbalanced designs,

missing values and

increased monitoring

during particular sub-

periods.

Need to determine limits

for each indicator.

The sampling design is

balanced, contains few

missing values is

regularly and evenly

sampled throughout the

period of interest.

0 100 200 300

0.01 0.02 0.03 0.04

Day during 2006/2007 year

FRP (mg/L)

(

)

(

)

guideline

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 19

Amplitude

Method

Strengths Weaknesses Recommended for use

when

D2 Avoids need to calculate indicator

limits.

Simple to calculate.

Categorical approach

(above/below levels)

results in some

information loss.

2nd level may need to be

chosen based on expert

knowledge unless

multiple guideline levels

exist.

A significant change in

scores may result from a

change in the second

level.

Missing values, repeat

samples or sub-periods

with more intensive

monitoring may bias the

results.

The sampling design is

balanced, there are very

few missing values and

the sampling is regular

and evenly spaced

throughout the period of

interest and multiple

guidelines exist.

D3 Can be modified to use other statistics

such as the 90th percentile instead of

median distance.

Relatively simple to calculate.

Requires sufficient

historical data to generate

empirical distribution if

theoretical distribution of

indicator values is

unknown.

Missing values, repeat

samples or sub-periods

with more intensive

monitoring may bias the

results.

Appropriate for a

balanced sampling design.

Assessing the ‘typical’

conditions for the

reservoir or water body.

It requires a regular and

relatively even-spaced

sampling throughout

time to capture temporal

variation and very few

missing values.

D4 Able to incorporate uncertainty

associated with original proportion D3

into the score.

Beta distribution is flexible in shape.

Transformation of D3 occurs on a 0 to

1 scale so no scaling is necessary.

Values of D3= 0, 0.5 and 1 are

transformed to the same values for D4.

Requires sufficient

historical data to generate

empirical distribution if

theoretical distribution of

indicator values is

unknown.

Missing values, repeat

samples or sub-periods

with more intensive

monitoring may bias the

results.

Sampling design must be

regular and relatively

even-spaced to capture

temporal variation.

Assessing the ‘typical’

conditions for the

reservoir or water body.

It requires a regular and

relatively even-spaced

sampling throughout

time to capture the

temporal variation and

very few missing values

to reduce potential bias.

D5 Reduced bias as a result of missing

values, unbalanced designs and sub-

periods of intense monitoring.

Limits change temporally to reflect

seasonal changes in the definition of an

‘extreme’ event.

Can be used with small/large sample

sizes.

Appropriate for balanced/unbalanced

designs.

Sufficient historical data

required for modelling

indicator limits.

Greater complexity and

slightly longer

computation time.

Assumes values between

samples are a linear

function of time.

The sampling design is

unbalanced, irregular

and unevenly spaced

throughout the year with

missing values.

Sufficient historical data

is needed to generate

indicator limits.

Sampling at key times

(highs/lows) throughout

the period of interest

will improve the model.

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

Table 2. (Continued)

Amplitude

Method

Strengths Weaknesses Recommended for use

when

Biased when a large

proportion of the year is

not sampled.

Bias may result when

samples are missing from

key times within period of

interest.

D6 Reduced bias as a result of missing

values, unbalanced designs and sub-

periods of intense monitoring.

Limits change temporally to reflect

seasonal changes in the definition of an

‘extreme’ event.

Smooth functions can be used rather

than linear interpolations between

samples.

In fitting the smooth, strength can be

borrowed from other sites when sites

are missing samples.

Multiple or repeat samples do not bias

results.

Appropriate for balanced/unbalanced

designs.

Sufficient historical data

required for modelling

indicator limits.

Most complex method and

longer computation times.

Sites are assumed to have

similar temporal trends or

there has to be a large

number of samples at each

site.

Smooths not always able

to model the occasional

extreme event.

The sampling design is

unbalanced, irregular

and unevenly spaced,

there are missing values

and a large number of

samples at each site.

Sufficient historical data

is needed to generate

indicator limits.

Sampling at key times

(highs/lows) throughout

the period of interest

will improve the model.

This method uses the same GAM fit used to calculate non-compliance (NC4) and so

suffers from the same under and over-estimation of the trend at sites. When averaging across

sites it is possible that this over/under-estimation may be averaged out across sites resulting in

a reasonable distance score representing the ‘average’ conditions on the lake. Providing a

sufficient number of samples have been collected at each site, model fits could be improved

by allowing separate curves to be fitted to each site.

Methods D5 and D6 are the preferred options for calculating amplitude however, due to

the limitations in the GAM approach to fitting a smooth curve to each site, we recommend D5

as the best approach when dealing with unbalanced monitoring designs and missing values,

so long as a sufficient number of samples have been collected to adequately capture the

general temporal changes over the period of interest.

2.3. Forming Overall Water Quality Indices

A water quality index can be developed using a combination of non-compliance and

amplitude scores across the site/(s) for a number of indicators of interest. The combining of

scores depends on the structure of the water quality index and how many sites and indicators

are used. We assume that data is collected for multiple indicators at a number of sites and the

indicators are grouped into at least two water quality categories.

Two initial methods for combining the non-compliance and amplitude scores for site i,

indicator j and water quality category k are:

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 21

(,)

,, ,,

()

nijk ijk

jk i

NC D

=×

∑ (13)

(,)

,,,,,

kiijkijk

SwNCD

=× ×

∑ (14)

where wi is a weighting for site i where the sum of the weights is equal to 1. If each site is

equally important than wi can be replaced by 1/n(j,k). If either D5 or D6 is used as the

amplitude score then due to these amplitude scores being conditional on exceeding the

guideline (non-compliance) and the implied multiplicative nature a geometric mean of the

non-compliance and amplitude scores would be more appropriate.

A squared transformation of Equation 13 results in:

(,)

,,,,,

kiijkijk

SwNCD

=× ×

∑. (15)

The combining of scores this way results in a score that is less than or equivalent to the

minimum of the two scores. It appears then that an extreme score in NC or D may not be well

reflected by the water quality index of Equation 14.

When D contains a non-compliance and amplitude component within the score (such as

D2 or even maybe D5 and D6) then the score could be simply calculated as:

(,)

,,,

kiijk

SwD

=×

∑. (16)

Some slight modifications to Equation 13 give alternative equations for calculating a

water quality index:

,, ,,

ijk ijk

NC D

Snn

=×

∑∑

(17)

,, ,,

nijk ijk

NC D

=∑

(18)

where nj is the number of sites within indicator j. The choice between Equation 13, 16 or 17

should relate to simplicity, understanding and usefulness of the score. Once a single score has

been calculated it may be useful to be able for managers/researchers interested in

understanding the scores to be able to trace and justify why the result was obtained. It makes

sense therefore that this occurs from the highest summary down in the order of: final score,

indicator, site and then NC or D. For this reason Equations 13 or 17 seem to be more

appropriate. We choose to use Equation 17 for the case study in next section.

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

To combine the scores across indicators and water quality categories the final score is

calculated as:

kkjjk

Cscore w w S

=∑∑

(19)

where k is the number of water quality categories, ,

1 and 1.

kjk

∑

If each indicator and each water quality category are of equal importance then:

where nk is the number of indicators in water quality category k.

3. AN APPLICATION

Many of the reservoirs within South-East Queensland (SE Qld), Australia are used for

drinking water, recreation, industry and agriculture. Understanding the conditions on the lake

is vital to the management of the reservoirs to ensure healthy water ecosystems for future

generations.

The methods for calculating water quality indices were developed for an annual

assessment of the reservoirs within SE Qld. They allow the reservoirs to be closely monitored

throughout the year and provide necessary information for the ongoing management of the

reservoirs.

The monitoring data collected by the Queensland Bulk Water Supply Authority (trading

as Seqwater) is often unbalanced in design. Monitoring occurs for many water quality

indicators on a monthly/fortnightly basis, however extra samples may be taken for event

monitoring, for calibration/validation or checking of results and for specific projects/studies

on the reservoir. As major events (algal blooms, rainfall and substantial runoff events) are

more likely to occur during summer this may mean increases in the number of samples over

this time thereby creating an unbalanced design. The number of sites that monitor each

indicator may vary over time and may also differ between indicators. The sites used for

calculating the water quality index should be representative of the dam and should remain

fairly consistent throughout time in order for comparisons across years to be valid.

Increased variability in many water quality indicators occurs during the Australian spring

(September - November) and summer (December - February) with peaks in concentrations

and levels also often occurring during summer. For this reason we define our reporting year to

be July to June so the summer period is not split between reporting years.

Preliminary analyses identified 16 water quality indicators within three water quality

categories that will be used in developing a water quality index assessment for the health of

the reservoirs. The first category (Water Quality) includes: surface and lake floor

and

jk k

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 23

concentrations of dissolved oxygen (DO), surface pH, surface turbidity, surface filtered

reactive phosphorus concentrations (surface FRP), surface total phosphorus (TP), surface

dissolved nitrogen (DiN) and surface total nitrogen (TN). The second category (Toxicants and

Pathogens; TandP) contain the surface algal ratio (total cyanophytes to total algal counts),

surface toxic species ratio (proportion of anabaena circinalis, microcystis aeruginosa and

cylindrospermopsis raciborskii to total cyanophyte counts), total toxic species at the surface

and e-coli surface concentrations. The Biological category includes surface chlorophyll-a, and

dissolved manganese (DiMang), ammonia nitrogen (NH3-N) and FRP lake floor

concentrations. In total there are 16 indicators split between three water quality categories.

These 16 indicators are not all monitoring at the same number of sites (see Table 3).

However, we believe it is more important to keep the number of sites consistent between

consecutive reporting years to enable appropriate comparisons across years. The removal or

addition of sites in some years may make it difficult to determine if changes are due to the

addition/removal of sites or changes in the overall water quality on the dam.

Table 3. The number of sites monitoring each of the 16 water quality indicators used for

the water quality index calculations

Indicator Number of sites

DO (surface) 14

DO (floor) 12

pH (surface) 14

Turbidity (surface) 14

FRP (surface) 13

TP (surface) 13

DiN (surface) 13

TN (surface) 13

Algal ratio (surface) 4

Toxic ratio (surface) 14

Total toxic species (surface) 14

E-coli (surface) 14

Chlorophyll-a (surface) 4

DiMang (floor) 11

NH3-N (floor) 11

FRP (floor) 11

3.1. Non-Compliance

The non-compliance scores have been calculated for FRP (at the lake floor) during the

2006/07 reporting year on one of the local reservoirs in SE Qld to compare the computational

differences between methods (see Table 4).

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

Table 4. Calculation of non-compliance scores for Filtered Reactive Phosphorus samples

taken at the site floor for numerous sites across a local SE Qld reservoir during

2006/2007. The scores are based on calculations using non-compliance methods 1 to 4

(NC1, NC2, NC3 and NC4)

Site Number of

samples

above the

guideline

Number of

samples

taken

Non-compliance

score using

Method 1 (NC1)

Average

monthly

proportion

Method 2

(NC2)

Non-compliance

using Method 3

(NC3)

Non-compliance

score using

Method 4 (NC4)

1 4 13 0.3077 0.3333 0.3781 0.4274

2 1 12 0.0833 0.0833 0.1068 0.2164

3 4 12 0.3333 0.3333 0.3507 0.3425

4 1 12 0.0833 0.0833 0.1370 0.2712

5 0 12 0.0000 0.0000 0.0000 0.1589

6 5 12 0.4167 0.4167 0.4301 0.3863

7 0 10 0.0000 0.0000 0.0000 0.0000

8 4 12 0.3333 0.3333 0.3370 0.4356

9 0 12 0.0000 0.0000 0.0000 0.1589

10 0 6 0.0000 0.0000 0.0000 0.1644

11 2 12 0.1667 0.1667 0.1699 0.2110

The first three non-compliance methods (NC1, NC2 and NC3) have very similar scores.

This is because the sampling design for this particular water quality indicator is not greatly

unbalanced. However, the fourth non-compliance scores are higher than the other non-

compliance scores for each site. Values which observe a 0% non-compliance (score = 0)

obtain scores greater than 0 for NC4. This occurs because we fit a single smooth function

across sites with a mean-shift for each site. The averaging of the function across sites means

some sites may have fits that indicate worse/better quality then what may have been observed.

To fit a separate curve for each site would require a greater number of samples to be taken. It

is for this reason that NC3 was chosen as the preferred method for calculating non-

compliance over NC4. Scores NC1 and NC2 are not suited to unbalanced sampling designs,

so NC3 also has the preference over these two scores. Non-compliance is therefore calculated

using method NC3.

3.2. Amplitude

Greater differences exist between the different amplitude scores than the non-compliance

scores (see Table 5). For example site 8 calculates a very large score using method D1 but

very small scores using methods D2, D5 and D6. This particular site has a single sample that

exceeded the recommended guideline. This particular value was close to the limit for that

month and so received a large amplitude score. Scores D2, D5 and D6 do not give as much

weight to this single value and so the amplitude score is very small. Amplitude scores D5 and

D6 are very similar in terms of the ranking of amplitude scores from each site. This makes

sense as the methodology for these two scoring procedures are similar. Amplitude score D4 is

very similarly ranked to both D5 and D6 even though the amplitude scores themselves are

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 25

very different. The other three scoring procedures give very different ranks and differing

amplitude scores. Since the linear interpolation method is chosen for the non-compliance

component (NC3) we also choose the linear interpolation method (D5) for the amplitude

component.

Table 5. Calculation of amplitude scores using Methods 1 to 7 for Filtered Reactive

Phosphorus at the site floor of 11 sites on a local SE Qld reservoir during 2006/2007

Site

No. of

values

Distance

Score 1

(D1)

Distance

Score 2

(D2)

Distance

Score 3

(D3)

Distance

Score 4

(D4)

Distance

Score 5

(D5)

Distance

Score 6

(D6)

1 4 0.6906 0.3077 0.8557 0.9635 0.3867 0.2296

2 1 0.8571 0.0833 0.4925 0.4925 0.0307 0.0316

3 4 0.3820 0.3118 0.5821 0.6375 0.1970 0.1339

4 1 0.6316 0.0833 0.7164 0.7164 0.0678 0.0666

5 0 0 0 0 0 0 0.0124

6 5 0.2258 0.3727 0.5323 0.5605 0.2661 0.1871

7 0 0 0 0 0 0 0

8 4 0.6892 0.3118 0.8010 0.9237 0.3276 0.2377

9 0 0 0 0 0 0 0.0124

10 0 0 0 0 0 0 0.0131

11 2 0.2218 0.1179 0.2587 0.2052 0.0188 0.0290

3.3. Final Water Quality Index

As the amplitude score is conditional on non-compliance (conditional probability) it

makes sense that the two scores are combined using a multiplicative operation. For this

reason we calculate the average of the two scores using the geometric mean or transformation

of a geometric mean rather than the arithmetic mean. We use Equation 17 as the means of

calculating a score for each indicator. However we also need to combine indicator scores if

we want to obtain a single score for the reservoir. We follow Equation 18 and assume that

each water quality category is of equal importance and that the water quality indicators within

each category are of equal importance:

kj k

Cscore S

=∑∑

. (20)

This means that greater emphasis is not placed on this particular category even though there

are twice as many indicators in this category than in each of the other two categories.

The average non-compliance and amplitude scores, category scores and final water

quality index for the reservoir in SE Qld are shown in Table 6. The water quality index

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

assigned for the year is 0.3825 where a score of 0 implies that all indicators were fully

compliant and a score of 1 implies all indicators were fully non-compliant. The

Toxicant/Pathogen category is the category of greater water quality. Dissolved oxygen

(bottom), surface pH, surface total phosphorus and surface total nitrogen are the indicators in

the water quality category that indicate very poor water quality. For the Biological category

chlorophyll-a surface concentrations indicate poor water quality followed closely by ammonia

nitrogen levels near the lake floor. For most of these indicators this appears to be due to a

high non-compliance component. Many sites must have observations that exceed the

recommended guideline levels.

Table 6. Report Card Scores for a local SE Qld reservoir during 2006/2007

Indicator No. of

sites NC D ,

Water Quality

Index Score

Report

Card

Score

Water Quality

DO Surface 14 0.5105 0.1286 0.2760 0.4604 0.3825

DO Bottom 12 0.8832 0.5318 0.6943

pH Surface 14 0.9303 0.5547 0.7273

Turbidity Surface 14 0.0261 0.1940 0.1488

FRP Surface 13 0.0264 0.0059 0.0325

Total Phosphorus

Surface

13 1.0000 0.5051 0.7107

DiN Surface 13 0.4249 0.1455 0.3397

Total Nitrogen

Surface

13 1.0000 0.5678 0.7535

Toxicants /

Pathogens

Algal ratio Surface 4 0.7753 0.3919 0.5565 0.2203

Toxic Species ratio

Surface

14 0.7826 0.0749 0.2429

Total Toxic Species

Surface

14 0.0000 0.0000 0.0000

E. Coli Surface 14 0.0325 0.0343 0.0816

Biological

Chlorophyll a

Surface

4 0.9438 0.4451 0.6503 0.4668

Dissolved

Manganese Bottom

11 0.4699 0.3331 0.4502

Ammonia Nitrogen

Bottom

11 0.8048 0.3482 0.5632

FRP Bottom 11 0.1736 0.1177 0.2036

CONCLUSION

This Chapter details the development of water quality indices using the combination of

two components: non-compliance and amplitude and within the context of monitoring designs

being balanced or unbalanced.

Non-compliance measures how often values exceed a recommended guideline during a

given period and amplitude measures departure from the guideline. A number of calculation

methods were developed for each of these components, with recommendations in here based

on the spatio-temporal monitoring context of the monitoring undertaken. The purpose of the

water quality index and the sampling design of the monitored indicators will determine which

method is most appropriate.

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 27

Many water quality indices assume regular and balanced sampling designs. In the case

scenario for a local reservoir within SE Qld the design is often unbalanced or missing data

and many of the documented water quality indices are inappropriate and would give biased

results. We have proposed methods that could be used when the sampling design is

unbalanced and missing values may occur. Some of these methods reduce bias that is present

when certain sub-periods are over-represented or missing values.

The impact of this bias has been investigated for the 2006/2007 year (Table 7). Three

scoring methods are compared. The first scoring method uses NC1 and D1, the second uses

NC3 and D5 while the third uses NC4 and D6. The differences between the scoring methods 1

and 3 are compared to the adopted scoring method (method 2: NC3 and D5). Results for the

non-compliance scores for each site and indicator combination differ by up to 0.25 for scoring

method 1 and 2 but up to 0.44 between scoring methods 3 and 2. Much larger differences

exist between the amplitude methods with differences in score of at most 0.7331 for each site

and indicator combination. By the time the scores are multiplied and averaged across sites to

derive an indicator score the difference in scores is at most 0.1271. The WQI scores differ by

only 0.0256 and the final score differs by 0.03 at most. This reduction in differences between

methods may be due in part to the averaging across sites, indicators and water quality

groupings. This averaging is necessary if we are to adequately represent the overall condition

of the lake. To avoid the overall water quality being influenced by a single site or indicator

with poor water quality we take an average across a number of sites and indicators. This

provides a more complete picture of the overall water quality. A poor score then indicates

poor water quality at a number of sites and indicators and the reverse is also true.

The water quality indices developed within this Chapter are designed for data containing

spatial and temporal information (i.e. when there are multiple samples taken during a given

period of interest at a number of different sites). The presence of seasonality is an important

consideration in the derivation of indicator limits. Many water quality variables show strong

peaks or increased levels during the Australian summer. An extreme in summer may be much

higher than a corresponding extreme in winter. The limits should therefore reflect changes

with respect to season and so we model the seasonal limits using a non-parametric quantile

regression.

Table 7. Comparison of scoring differences using Score1: NC1 and D1, Score 2: NC3

and D5 and Score 3: NC4 and D6

Scale NC for each

site and

indicator

D for each

site and

indicator

NC * D for

each site and

indicator

Indicator

score

WQI

score

Final

Score

1 –

Score

Minimum -0.2541 -0.7331 -0.2432 -0.0673 0.0255

Q1 -0.0679 0.0000 0.0000 -0.0047

Median 0.0000 0.0313 0.0039 0.0416 0.0256 0.0325

Q3 0.0000 0.1280 0.0583 0.0708

Maximum 0.1021 0.8264 0.3947 0.1271 0.0464

Score

3 –

Score

Minimum -0.1984 -0.7065 -0.1006 -0.0509 -0.0182

Q1 0.0000 -0.0165 -0.0061 -0.0120

Median 0.0000 0.0000 0.0000 0.0000 -0.0015 -0.0028

Q3 0.0904 0.0009 0.0131 0.0130

Maximum 0.4432 0.1232 0.1726 0.0298 0.0114

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

The non-compliance component of the water quality index is of at least equal if not

greater importance than the amplitude component. The recommended guidelines used for the

non-compliance scores are often safe levels beyond which serious public health or

environmental problems may arise. If guidelines continue to remain non-compliant for large

proportions of the year then management and preventative action may need to be

implemented so reoccurrence doesn’t occur in following years. Amplitudes D5 and D6

already incorporate aspects of non-compliance into the score and so do not need to be

combined with a non-compliance score. However, we emphasize the importance of the non-

compliance score by combining the amplitude score (D5) with the corresponding non-

compliance score (NC3). The combination of these two scores provides a water quality index

that captures the non-compliance and amplitude while incorporating spatial, temporal and

seasonal information in the scores. The methods NC3 and D5 reduce bias that may result from

using an unbalanced sampling design or from missing values within the data. There is also

room for further research into the likelihood of scores for each water quality index when they

contain a varied number of indicators. As the number of indicators in each index increases the

scores are more likely to be closer to the mean 0.5, with much less chance of obtaining scores

close to 0 or 1 (Figure 10). The indicator scores were randomly sampled from a uniform

distribution with range of 0 to 1 to reflect the possible range of scores under an independence

assumption. This follows the central limit theorem but creates some difficulties in the

interpretation of scores or converting the scores to report card grades (A, B, C, D, Fail). We

avoid these difficulties by simulating some worst case scenarios for each of the grades based

on the number of indicators in each of the categories. These simulated cut-offs are then used

to determine the range of scores for a particular grade. In this way we account for the effect of

the number of indicators in each category within our calculations.

Figure 10. Distributions of final report card scores based on 3 categories with (a) 1 indicator in each

category (b) 2 indicators in each category, (c) 4 indicators in each category, (d) 8 indicators in one

category and 4 indicators in each of the other two categories (e) 8 indicators in each category and (f) 10

indicators in each category. Each indicator score was randomly sampled from a uniform distribution

with range 0 to 1. The scores were combined using Equation 19 and the process simulated 10000 times

to create the distributions above.

1,1,1

score values

Frequenc y

0.00.20.40.60.81.0

0 400 1000

2,2,2

scor e values

Frequenc y

0.0 0.2 0.4 0.6 0.8 1.0

0 500 1500

4,4,4

score values

Frequenc y

0.0 0.2 0.4 0.6 0.8 1.0

01000

8,4,4

score values

Frequenc y

0.00.20.40.60.81.0

0 1000 2500

8,8,8

scor e values

Frequenc y

0.0 0.2 0.4 0.6 0.8 1.0

0 1000 2500

10,10,10

score values

Frequenc y

0.0 0.2 0.4 0.6 0.8 1.0

0 500 1500

Nova Science Publishers, Inc.

Water Quality Indices from Unbalanced Spatio-Temporal Monitoring Designs 29

ACKNOWLEDGMENTS

This work is based on Lennox and Wang [2008] which can be consulted for further

information. The work has been funded by the Queensland Bulk Water Supply Authority

(trading as Seqwater) and the Commonwealth Scientific and Industrial Research

Organisation. We also wish to thank Seqwater staff, Noel Cressie and Abdel El-Sharaawi for

their valuable comments/suggestions on this work.

REFERENCES

Bartels, R. and Conn, A. (1980). Linearly Constrained Discrete L_1 Problems, ACM

Transaction on Mathematical Software 6, 594–608.

Canadian Council of Ministers of the Environment. 2001. Canadian water quality guidelines

for the protection of aquatic life: CCME Water Quality Index 1.0, Technical report. In

Canadian environmental quality guidelines, 1999; Canadian Council of Ministers of the

Environment; Winnipeg.

Cude, C. G. (2001). Oregon water quality index: a tool for evaluating water quality

management effectiveness. Journal of the American Water Resources Association 37(1),

125-137.

EHMP. (2007). Ecosystem Health Monitoring Program 2005-06 Annual Technical Report.

South East Queensland Healthy Waterways Partnership, Brisbane.

He, X., and Ng, P. (1999). COBS: Qualitatively Constrained Smoothing via Linear

Programming; Computational Statistics 14, 315–337.

Kannel, P. R., Lee, S., Lee, Y. S., Kanel S. R., and Khan, S. P. (2007). Application of water

quality indices and dissolved oxygen as indicators for river water classification and urban

impact assessment. Environmental Monitoring Assessment 132, 93-110.

Koenker, R. W., and D’Orey, V. (1987). Algorithm AS 299: Computing regression quantiles.

Applied Statistics 36(3), 383-393.

Koenker, R. W., Ng, P., and Portnoy, S. (1994). Quantile Smoothing Splines. Biometrika

81(4), 673-680.

Koenker, R., and Ng, P. (1996). A Remark on Bartels and Conn's Linearly Constrained

L1Algorithm, ACM Transaction on Mathematical Software, 22, 493–495.

Koenker, R., and Ng, P. (2005). Inequality Constrained Quantile Regression, Sankhya, The

Indian Journal of Statistics 67, 418–440.

Lennox, S. M., and Wang, Y-G. (2008). Report Card Tool for Water Quality Monitoring at

Wivenhoe, Somerset and North-Pine Storages. CSIRO Mathematical and Information

Sciences Report 08/108.

Morgan , R. P., Kline, K. M, and Cushman, S. F. (2006). Relationships among nutrients,

chloride and biological indices in urban Maryland streams. Urban Ecosystems, 10, 153-

166.

Ng, P. (1996). An Algorithm for Quantile Smoothing Splines, Computational Statistics and

Data Analysis, 22, 99–118.

Ng, P., and Maechler, M. (2008). COBS – Constrained B-splines (Sparse matrix based). R

package version 1.1-5. http://wiki.r-roject.org/rwiki/doku.php?id=packages:cran:cobs

Nova Science Publishers, Inc.

Sarah M. Raican, You-Gan Wang and Bronwyn Harch

Said, A., Stevens, D. K., and Sehlke, G. (2004). An Innovative Index for Evaluating Water

Quality in Streams. Environmental Assessment, 34(3), 405-414.

Sargaonkar, A., and Deshpande, V. (2003). Development of an Overall Index of Pollution for

Surface Water Based on a General Classification Scheme in Indian Context.

Environmental Monitoring and Assessment, 89, 43-67.

Sarkar, C. andAbbasi, A. (2006). Qualidex – A New Software for Generating Water Quality

Indice. Environmental Monitoring and Assessment, 119, 201-231.

Smith, M. J., and Storey, A. W. (2001). Design and Implementation of Baseline Monitoring

(DIBM3): Developing an Ecosystem Health Monitoring Program for Rivers and Streams

in Southeast Queensland. Report to the South East Queensland Regional Water Quality

Management Strategy, Brisbane.

Swamee, P. K., and Tyagi, A. (2000). Describing Water Quality with Aggregate Index.

Journal of Environmental Engineering, 126(5, 451-455.

Queensland. (2006). Queensland water quality guidelines 2006. Environmental Protection

Agency, Brisbane (accessed online 24th September 2008: http://www.epa.qld.gov.au/

publications?id=1414).

Wood, S. N. (2006). Generalized Additive Models: An introduction with R; Chapman and

Hall/CRC Press.

Yu, K., Lu, Z., and Stander, J. (2003). Quantile regression: applications and current research

areas. The Statistician, 52, 331-350.

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 2

ESTIMATES OF LIKELIHOOD AND RISK ASSOCIATED

WITH SYDNEY DRINKING WATER SUPPLY FROM

RESERVOIRS, LOCAL DAMS AND FEED RIVERS

Ross Sparks1, Gordon J. Sutton1,2, Peter Toscas1

and Rod Mc Innes3

1CSIRO Mathematical, Informatics and Statistics, Australia

2School of Chemistry, University of New South Wales, Australia

3Sydney Catchment Authority, Australia

ABSTRACT

As the main supplier of potable water to Sydney, Sydney Catchment Authority aims

to supply water that is in accordance with Australian drinking water quality standards.

The standard specifies thresholds for acceptable ranges for over 80 analytes and

attributes. When the quality thresholds are exceeded, the water needs to be treated, so that

the treated water complies with the standard. The total expected cost of treating the water

is the risk to be calculated.

The main objective of the chapter is to describe a water quality risk assessment

process that is free from sampling biases. The risk assessment process involves modelling

each analyte based on historical data so that estimates can be obtained for the likelihood

of exceeding the water quality thresholds. These likelihood estimates are then combined

with associated estimated water treatment costs to produce an annual water treatment cost

estimate, or equivalently, an annual risk estimate.

Costs of exceeding thresholds are in their early stages of development and, at best,

costs have been developed for each threshold exceedance separately. Some analytes have

one-sided minor and major thresholds, while others have two-sided thresholds. Risk

values (equal to the integral of exceedance likelihood estimates times the treatment cost

associated with such exceedances) have been explicitly calculated for 2006.

A future aim is to repeat this process for all past years as a way of monitoring the

year-to-year variation in risk for each analyte. Since the process is to be to be repeated for

each year - the chapter works towards providing a robust approach to estimating the risk,

which is:

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

1. Flexible in distributional assumptions; and

2. Capable of selecting variables that are important for estimating risk.

The application datasets come from the Sydney Catchment Authority and comprise

approximately 80 analytes sampled at 35 different locations. Given the number of

analyte-site combinations and the aim of yearly model fitting, the modelling process

needs to be repeatable, so the process was fully automated using R scripts. Although the

real-risk assessment problem is multivariate and spatio-temporal in nature, the associated

cost profiles were not available in this form. The chapter includes an outline of a process

that could be followed in the future if such cost profiles become routinely available. The

cost profiles could be used to estimate the annual risk of all analyte exceedances, taking

into account interactions. This requires a completely different way of thinking about the

problem and requires a detailed understanding of variation and associated costs.

INTRODUCTION

The importance of water quality and its impact on health is well documented in the

literature (e.g., see Abera et.al. [2011], Fawell and Nieuwenhuijsen [2008], Krewski et al.

[2004], Hales et al. [2003], Urbansky and Schock [1999], Yang [1998]). There is evidence

that poor water quality is linked to increased risk of cancer (e.g., see Morris [1995]). In

addition, the concerns about the stringency of standard safe water thresholds are documented

in Grinsven et al. [2006], Cross et al. [2004] and Pavlov et al. [2004] and therefore the risks

are not well understood. This indicates the importance of collecting the data, controlling the

variation in analytes in the drinking water system, and improving the understanding and

management of the risks associated with water quality. The key purpose of the water quality

risk assessment in this paper is to provide risk profiles for each water quality hazard group.

This will assist in managing hazards by applying risk controls to hazard events in the supply

systems with the highest priority (e.g., see Rosen et al. [2010]). There are several approaches

to risk management and some of these are compared in Chowdhury et al. [2009]. In this

chapter, we take a traditional approach to risk assessment and estimate the likelihood and

consequences in terms of operational costs to calculate risk. In addition, the risk assessment

process will also be used to develop a water quality management framework. Monitoring for

deterioration in catchment water quality after adjusting for changes in climate is briefly

discussed in the Conclusion Section of the Chapter. The purpose of the statistical analysis in

this chapter is to apply statistically robust data analysis methods for the calculation of the risk

assessment. The following general steps are carried out on each analyte to achieve this

objective:

1. Step 1: Develop model

Use all the available data to develop a model that can predict the analyte values

throughout the year without sampling bias. This is done as follows:

a. Clean the data – making certain all measures are recorded using the same unit of

measurements – discarding all entries that are obvious errors.

b. Decide on a set of parametric distributions to select from, with the aim of

describing how measurements vary in distribution over time.

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 33

c. Decide on explanatory variables to consider in the models.

d. Fit models for each distribution, based on the explanatory variables that can

predict temporal changes in the parameters of the distribution and that can

interpolate the daily trend in the distribution of values for the target year.

e. Decide which model, and hence distribution, best describes how measurements

vary about the trend.

2 Step 2: Estimate Likelihood

Use the model for the selected distribution to predict the trend in the analyte’s values

throughout the target year. Estimate the probability of exceeding the analyte’s

threshold for each day using the predicted distributions. Find the sum of the daily

probabilities exceeding the threshold for a year. This estimates the expected number

of days per annum that the analyte exceeds the threshold.

3 Step 3: Estimate Risk

Estimate the risk (expected value of likelihood multiplied by consequence) of poor

water quality for the particular analyte as follows:

a. Establish (potentially multiple) poor quality thresholds that define different

consequence levels and establish the associated consequences (treatment and

operating costs to SCA (Sydney Catrchment Authority)).

b. Estimate the daily probabilities of exceeding each poor water quality threshold

using the model for the selected distribution, as in Step 2.

c. Estimate the annual risk caused by variability in the analyte (sum over all days

within a year the daily likelihoods for each consequence level multiplied by the

respective consequences). This estimates the total expected consequence, or cost

above normal operation, for the target year.

4 Step 4: Estimate uncertainties

Estimate the uncertainty in the estimates of likelihood and risk.

In more detail, empirical models are used to predict analyte values, particularly on days

when no analyte value is measured. These models build equations linking the analyte value to

a set of explanatory variables in a way that the equations can be used to reasonably accurately

predict the analyte value given the explanatory variable values. More specifically, the

equations define the distribution the analyte is expected to adhere to, as functions of the set of

explanatory variables. The models have parameters assigned to each explanatory variable that

describe the impact the explanatory variable has on predictions in the model. Methods

designed to minimise the model error are used to estimate parameters. Model selection

methods are used to decide on the most suitable distribution for the error structure in the

model. The fitted models estimate the trend in mean values over the target year and the spread

of values around that trend. These evolving mean values and volatility measures are used to:

• Estimate the likelihood – as the estimated number of days the water quality threshold

will be exceeded for the target year;

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

• Estimate the risk – as the total estimated cost to SCA due to variable analyte values

(for each analyte and supply site combination); and

• Estimate the uncertainty in the risk and likelihood estimates.

We start by looking at estimating the likelihood.

ESTIMATING THE LIKELIHOOD USING THE FITTED MODEL

In this section we develop estimates for the likelihood of exceeding water quality

thresholds. We assume that a model has been fitted for the analyte of interest and that the

model defines the probability density function of the analyte at each time point. In the SCA

case study, gamlss models (Rigby and Stasinopoulos [2005], Stasinopoulos, Rigby and

Akantziliotou [2006], and Stasinopoulos Rigby [2007]) were developed for each analyte, with

models selected from the gamlss family of distributions: normal, log-normal, Inverse

Gaussian and Zero Adjusted Inverse Gaussian. Discussion of the modelling for the likelihood

is left to the model formulation section. For the sake of illustration, in this section we use a

time-varying Gaussian distribution, having time-varying mean and standard deviation

parameters, rather than a generic distribution. The method is applicable to any defined and

calculable probability distribution.

Let the density for the analyte be denoted by ),/( ttt

. The average likelihood (or

average probability) of exceeding either the upper or lower threshold is then estimated as:

where the time period (t ranging from 0 to 1) represents one year, and

is the probability of the analyte being below threshold c at time t.

If there is only an upper threshold then this average is calculated as:

The above integral is estimated by breaking time into daily intervals and summing over

days. The average likelihood is estimated by first deciding on a time of day at which to

calculate the likelihood. Estimates of the mean and standard deviation (denoted t

ˆ and t

ˆ,



dttcFtcFp ttLttu )),/)((),/)((1(

σμσμ

+−= ∫



dxxfcF

tttt ∫

),/(),/(

σμσμ



.)),/(1(

dtcFp ttuu

σμ

∫−=

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 35

respectively) then give an estimate of the likelihood of exceeding a threshold on the day

represented by time t of the target year, at the decided time of the day, as:

is then estimated by averaging these estimated daily likelihoods over each day of the year.

Similarly, an estimate of

can be found when there is only an upper threshold.

The number of days in the year that will exceed one of the thresholds is estimated by

365.25 days times the estimate of the daily likelihood value for either both thresholds (upper

and lower) or the upper threshold only.

Figure 1. Simple illustration of the relationship between the fitted models and estimating probability of

exceeding the threshold.



)

/)(()

/)((1

ˆttLttut tcFtcFp

−=



)

/)((1

ˆttuut tcFp

−=

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

A GRAPHICAL REPRESENTATION OF THE LIKELIHOOD ESTIMATION

PROCESS

Figure 1 is used to demonstrate how the daily likelihood values are estimated for two

days represented by t=0.6 and t=0.8 during the target year. A simple fitted model has been

used comprising a linear trend term and a constant standard deviation for the volatility,

giving:

where t is the fraction of the target year.

The distribution of analyte values about the mean, t

, is assumed to be normally

distributed and the upper threshold has been set as 17. The density functions used to estimate

the probability of exceeding the threshold at times t=0.6 and t=0.8 of the year are the bell

shape curves plotted in Figure 1.

Figure 2. Examples of daily likelihood values – probability of exceeding an upper and lower threshold.



12t+3



Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 37

The probabilities p0,6=P(Yt>17|t=0.6) and p0,8=P(Yt>17|t=0.8) are equal to the areas

shaded between the density estimates (bell shape curves) and the vertical line drawn at t=0.6

and t=0.8, respectively. The area is found for each time t using the normal distribution

function, with mean t

and standard deviation t

, giving the probability that the random

analyte value taken from the density at time t is greater than the threshold 17.

The probability of the analyte value exceeding the threshold for each day of the year is

summed over all days in a year.

See Figure 2 for a graphical representation of estimating daily probabilities of exceeding

either an upper or a lower threshold. The two examples of the probabilities of exceeding

either of the thresholds illustrate that at times both tails of the distribution need to be

considered for calculating daily likelihood values.

ESTIMATING THE UNCERTAINTY IN THE LIKELIHOOD ESTIMATES

The approach to estimating uncertainty in the likelihood estimates is developed for the

situation where there is an upper and a lower threshold, and the upper threshold problem

follows as a simplification of the results.

The likelihood is taken as the average of

over days t = 1, 2, .., 365 (or 366) in the target year, where t

and t

are estimates of the

mean and standard deviation for day t, estimated from the selected gamlss regression model.

We note that in the preceding section that introduces the likelihood, t is instead a continuous

variable on [0,1], however here it is treated as discrete. This section develops a process for

estimating how inaccurate ˆ

is relative to the true value, p

, where

Moreover, we are interested in establishing the uncertainty in the likelihood estimate for

the sum of ˆ

over all days within the target year.

The parameters of the time varying distributions,

and

, are estimated using

regression models with time varying covariates, such as flows. Uncertainties in risk estimates

are generated by the uncertainty in the fitted model, i.e., the uncertainty in the estimated

regression parameters in the gamlss model (Rigby and Stasinopoulos [2005], Stasinopoulos,

Rigby and Akantziliotou [2006], and Stasinopoulos Rigby [2007]).

Both the sequential estimates of the parameters ( ˆ

and ˆ

) and the covariates are highly

autocorrelated across days, and this autocorrelation influences the uncertainty in likelihood

estimates. Perhaps the best way to cope with such situations is to develop a hierarchical

Bayesian approach to estimating the uncertainty by establishing the posterior distribution of



)

/)(()

/)((1

ˆttLttut tcFtcFp

−=



),/)((),/)((1 ttLttut tcFtcFp

−=

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

the likelihood estimate. An Markov Chain Monte Carlo (MCMC) approach could be used to

estimate this posterior probability.

We did not follow this approach because we anticipated that it would take too long to

apply to all analyte-site combinations. The alternative followed in this chapter is to establish

reasonable estimates of the standard errors (standard deviations) of the estimated likelihood

values.

At times it was necessary to make some simplifying assumptions. The first important

assumption is that we assume all covariates are measured without error or with negligible

error (i.e., near zero uncertainty in the measurement process).

The estimates t

ˆ are established using the same fitted models. However, flow and

storage variables are highly autocorrelated, and therefore t

ˆ values are highly autocorrelated.

Consequently, independence of consecutive daily estimates t

ˆ cannot be assumed. We find

the uncertainty in the sum of t

ˆ under the assumption that t

ˆ is an unbiased estimate of t

for all t, i.e. tt ppE =]

[. This is a relatively strong assumption in practice because the

thresholds that define t

p are often in the tails of the fitted distributions and the tails are less

accurately estimated than the high density regions. Based on this assumption, the variance in

the estimated sum probability of exceeding a threshold is given by:

We now assume that t

ˆ is estimated from a gamlss model. From Cogley [1999] and

Thiebaux and Zwiers [1984], the effective sample size used to fit the model, e

n, is given by:

where m is the number of observations used to fit the model and ˆ

(

) is the estimated

autocorrelation in the model residuals. The R function effectiveSize, in the R library coda, is

used to estimate ne. The autocovariance of ˆ

can then be estimated as:

where )(

−t is the autocorrelation between the estimates of the probability of exceeding

the threshold on days t and

, so that



[]

⎟

⎠

⎞

⎜

⎝

⎛−= ∑∑ ==

365

1])

[

()

ttpEpEpVar



[

]

(

)

()

(365

365

ttt

ttppppE −−= ∑∑ ==



].)

)(

[(

365

∑∑

== −−= ttt ppppE

ττ



,))(

)/1(21/( 1

∑−

=Δ

−+= m

jte jmjmn



,/)()1()1()]

)(

[( etttt ntppppppppE

τρ

ττττ

−−−≈−−

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 39

What is missing in this equation is some goodness of fit measure for the gamlss model,

and this could be estimated using the bootstrap approach or MCMC. Bootstrap and MCMC

are both computationally intensive, so for the sake of simplicity the model uncertainty part of

the variance is ignored.

Therefore, an estimate of

[]

)

)(

(

ττ

ppppE tt

−

is given by:

where )(

−t is approximated by the sample autocorrelation for variable

)

1/(

)

ˆttttt ppppp −=− . This then gives the estimated variance of the annual

likelihood estimate as:

Simulation studies found this estimated variance to be fairly volatile, but the alternative

MCMC approach was considered untenable given the computational cost of the model

selection and retrospective surveillance aspects of the work.

Very roughly, we expect approximately 70% of the total probabilities to be within the

bounds

ESTIMATING THE COST RISK ASSOCIATED

WITH EXCEEDING UPPER THRESHOLDS

Using a step-function: For each analyte-site combination, let the increase in drinking

water supply and operating costs be m

ccc

...

21 when analyte measurement boundaries

bbb <<< ...

21 are exceeded. In addition, assume that the probability of the analyte at a

site on day t exceeding j

b is given by:

For day t the expected increase in supply and operating costs is given by:



∑∑∑===−−−≈ 365

365

1./)()1()1()

(tett

ttntpppppVar

τττ τρ



,/)(

)

ˆett ntpppp

τρ

ττ

−−−



./)

)

)(

)

(

ˆ365

365

ttt

ttnpppptparV ∑∑∑===−−−=

τττ

τρ



(

ˆ365

365

1∑∑ == ±tt

ttparVp



,),/(1 ttt

ttj dxxfp

σμ

∫

−=

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

These can be thought of as costs above the normal running costs associated with adverse

variation for the respective analyte being considered. The annual expected cost due to

variable analyte values is given by:

The individual expected daily costs are estimated by:

and the expected annual cost is estimated by:

ˆis taken as the increase to the cost risk associated with analyte variability over

the year.

USING A DOUBLE-SIDED TWO STAGED (MINOR OR MAJOR)

STEP-PROFILE COST FUNCTION WITH AN ESCALATING

COST FOR CONSECUTIVE MAJOR EXCEEDANCES

For each analyte-site combination, define the analyte measurement boundaries

∞=<<<<= +110 ...0 mm bbbb . Now define analyte measurement levels L0, L1,…, Lm so

that when an analyte measurement lies between measurement boundaries j

b and 1+j

b it is in

level Lj.

This gives the probability of an analyte measurement being in level Lj at a site on day t as

−tjtj pp , where tj

p is defined, as before, as:

We now extend the cost structure to allow both lower and upper quality thresholds and

also consecutive events. Denote the analyte measurement at the site of interest on day t as t



....)()()1(0 3222111 tmmtttttt pcppcppcpEC ×

−

×+−×=



365

EC EC

=∑



...)

ˆˆ

()

ˆˆ

()

1(0

ˆ3222111 tmmtttttt pcppcppcpCE ×++−×+−×+−×=



365

ˆˆ

EC EC

=∑



ttt

ttj dxxfp

),/(1

σμ

∫

−=

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 41

Let L0 and Lm represent major events, then define a new cost level for day t that applies

when both t

y and 1−t

y are major events (of the same sort).

So, define the increases in drinking water supply and operating costs on day t as:

cc0, for y

and y

−1 lying in level L0 (consecutive major events);

c0, for y

lying in level L0 and y

−

1 not (major event);

c1...cm−1, for y

lying in levels L1, L2,…,Lm-1 .

, for y

lying in level Lm and y

−

1 not (major event); and

ccm, for y

and y

−1 lying in level Lm (consecutive major events).

Note that Lj is the region where bj < y < bj+1 , j=0,…,m.

Figure 3 provides an example with m=4. Although this figure considers the application

that excludes changing costs for consecutive events, it does plant the seed for considering

consecutive events.

Figure 3. Demonstrating the minor and major exceedance thresholds for day t.

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

The following table gives the cost incurred on day t, as a function of the analyte levels on

day t and day (t-1), for the general case that there are thresholds on both the low and the high

sides and the second of two consecutive major events incurs an extra cost (i.e., 0c

c and 4c

c).

For illustration the number of boundaries is restricted to 4, 01 5

,bb b

<<… where 00b=

and 5

b is infinite.

Level on day (t-1) (Yesterday)

L0 L1 L2 L3 L4

Level on

day t

(Today)

L0 0c

L3 3

c4c

The next table gives the costs incurred on day t, for the same case as in the previous table,

along with the corresponding probabilities of occurrence, assuming the model residuals are

uncorrelated.

Cost Probability Interpretation

y and 1−t

y in L0 0c

c )1)(1( 1)1(1 −

−

tt pp consecutive major events

y in L0, 1−t

y not L0 0

c ))(1( 1)1(1 −

−

tt pp major event

y in L1 1

c 21 tt pp

−

minor event

y in L2 2

c 32 tt pp

−

in control

y in L3 3

c 43 tt pp

−

minor event

y in L4, 1−t

y not L4 4

c )1( 4)1(4 −

−

tt pp major event

y and 1−t

y in L4 4c

c 4)1(4 −tt pp consecutive major events

By setting the appropriate costs to zero and appropriate consecutive analyte measurement

boundaries equal, this structure can accommodate cases with just lower thresholds or just

upper thresholds as well as the double sided case. Similarly, the cost of two consecutive days

with major events need not be implemented.

The following formula gives the expected cost incurred on day t for the double sided

case, including consecutive major events:

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 43

The expected cost, or risk, for the year 2006 is then given as:

where t = 1 represents January 1, 2006. Note that to calculate 2006∈t

EC , the expected cost

over 2006, 1t

and 4t

p will be needed for 31 Dec 2005.

ESTIMATING THE UNCERTAINTY IN THE RISK ESTIMATOR,

EXCLUDING ALTERED COSTS FOR CONSECUTIVE EXCEEDANCES

The challenge in this section is to come up with a relatively simple way of estimating the

uncertainty for the annual risk estimator defined as the total cost for the year caused by

variable quality in analyte values at a supply site.

Assume there are m thresholds. Define ti

p as before and set 1

p and .0

+tm

Then, write the estimated daily cost as:



̂ 

̂

̂ 

̂

̂ 

̂

 

 



 





 



where cT=c1c2c3... cm

[]

; ˆ

T=ˆ

Tˆ

T... ... ˆ

[

]; and ˆ

ti =ˆ

ti −ˆ

ti+1

Then the estimated total annual cost is given by:

where

[

]

TTTTT qqqqq 365321 ˆ

...

ˆˆˆˆ = and

⊗

is the Kronecker product.

We note that t

y is being represented as a multinomial with levels L1, …, Lm and with

probabilities tmt qq ,,

1…, such that ,1

∑

iti

q for all t, and 0][

tjti qqE for .

i≠

The variance of the estimated daily cost is given by:



)ity)(probabil(cost tii

∑



+−

−

+−= −− )()())(1())(1( 3222111)1(101)1(10 ttttttttc ppcppcppcppc



4)1(444)1(44433 )1()( −−

−+− ttctttt ppcppcppc



365

2006 ∑

∈=

tt ECEC



)1(

ˆ365

365

1qcqcCE TT

tt⊗== ∑∑ ==

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

where t

Σ is the covariance matrix of t

ˆ, given by:

The matrix t

is estimated by replacing tj

qby tj

ˆin the equation for t

Σ above. This

gives an estimate for the variance of daily costs as:

Note that this only estimates the uncertainty for the cost on day t, not for the total cost for

the year.

To obtain an estimate for the annual expected cost, it is assumed that there is

autocorrelation between consecutive estimates of the likelihoods, i.e.

where )(

−t is the autocorrelation between estimate ˆ

t/(1 −ˆ

t) and ˆ

/(1 −ˆ

Therefore estimating the standard deviation of the estimated annual expected cost is much

more complicated than estimating the standard deviation for daily risk estimates and a few

assumptions are needed to simplify the process. Two methods are now presented for

estimating this standard deviation, each based on different assumptions about the

autocorrelations.

Method 1: We assume that the distribution of t

p is invariant of time t, such that tj

ˆ are

roughly constant over time. In addition, we assume that we can approximate the covariance

matrix of q

ˆby mm

R×

Σ⊗ , where R is the 365x365 matrix with ones down the main

diagonal and the time-invariant autocorrelations of tj

ˆ on the jth off diagonal for all j; and

mm×

Σis the time-invariant covariance matrix for all .

As such the variance of the total estimated cost for the year is given by:



var( ccCE t

tΣ=



)1(..

....

..)1(

22221

12111

tmtmtmttmt

tmttttt

qqqqqq

⎥

⎦

⎤

⎢

⎣

⎡

−−−

=Σ



./)

ˆˆ

()

ar(v

ˆ1

111

jk tktjkj

jtjtjjt nqqccqqcCE

∑

∑−

=+== −−×=



,/)1()1()()

cov( ekktktkktk nqqqqtqq

τττ

τρ

−−−=



,11)1)()(1()

var( 365

1ccRcRcCE mm

tt××

=Σ×=⊗Σ⊗⊗=

∑

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 45

where T

1is the 1x365 matrix of 1’s.

The time-invariant autocorrelations in

are approximated by averaging the estimated

autocorrelations of tj

qover j and mm ×

is approximated by the time-average sample-

estimated covariance matrix of .

qTreating each t

q as comprising the probabilities of a

multinomial distribution, mm×

Σ is approximated by:

Method 2: We now allow the distribution of q

to vary with time and assume that the

covariance matrix of qcan be written:

(1)

where t

Σ on the diagonal are defined previously; and )|))(|

−

on the off-diagonals

are the covariances between t

ˆ and

ˆ. The ( i

th , jth ) element of )|))(|

tΣ− is

assumed to be given by: })1()1(|)(|{ jjtitijti ppppppt

τττ

τρ

−−−− when ij ≠ and

})1()1(|)(|{ iititi ppppt

ττ

τρ

−−−− when .ij

An estimate for T

Σ, denoted T

ˆ, is obtained by replacing each variable by its estimate.

That is, tj

pby tj

ˆand )(

−tby )(

−

t. )(

−

t is approximated as the average over

all j of the sample autocorrelation of tj

ˆat time difference )(

−

t days.

The variance estimate of the annual cost estimate is then given by:



)

ˆˆˆˆ

....

ˆˆ

..)

ˆˆˆ

ˆˆ

)

365

22221

12111

∑

⎥

⎦

⎤

⎢

⎣

⎡

−−−

=Σ

tmtmtmttmt

tmttttt

mm n

qqqqqq



)362()363()364(

)362()1()2(

)363()1()1(

)364()2()1(

3653,3652,3651,365

365,333232

365,223221

365,113121

⎥

⎦

⎤

⎢

⎣

⎡

ΣΣΣΣ

=Σ

ρρρ



).1(

)1()

ar(v

ˆ365

1ccCE T

tt⊗Σ⊗=

∑=

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

The uncertainty in the total costs estimate

365

ˆt

∑

is given by the estimated standard

error:

A similar approach can be used for a two stage (minor or major) additional cost step

function.

ASSESSING TRENDS IN THE LIKELIHOOD AND RISK

We decided to calculate the average likelihood and risk for each annual period, rather

than averaging over the whole history, and then presenting the trend in these annual

likelihoods and risks over the whole history. The advantage of this approach is it allows SCA

to monitor trends in the annual risk over time in a way that is linked to their associated

budgeted costs. In addition, annual averages remove the seasonal influence on the likelihood

and risk scores. Understanding the variation in annual likelihoods and risks across years is

essential for managing the increased costs associated with these risks.

Future analyses should look at evaluating whether SCA’s risk management strategy is

improving from year to year. This calls for a different monitoring strategy to the one referred

to in this chapter. This is much more difficult because the variation in risk that is beyond the

control of SCA, such as weather changes, need to be removed before the management

performance can be assessed for those aspects (hazard events) that are within the control of

SCA.

MODELS FOR ESTIMATING THE LIKELIHOOD

Models were developed for estimating the likelihood of analyte values exceeding upper

or lower thresholds for all days within a year. These models needed to account for seasonal

variation, flow characteristics and other explanatory variables that were measured daily.

These models were used to predict analyte values during periods where measurements are not

made.

Having daily measurements for each analyte would have been ideal for modelling.

However, only a few analytes were measured daily (e.g., pH). Another, less ideal, scenario

would have been having the sampled days sufficiently frequent. If the sampled days were

representative of the period of study it would provide enough accuracy to the likelihood of

threshold exceedance estimates. However, this too was seldom true, which made assessing

the annual risk of exceeding thresholds at a site problematic. Therefore, the purpose of

building a model was to have the ability to interpolate analyte values even for days when the

analyte had not been measured.



)

ar(v

ˆ365

∑=tt

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 47

Potential explanatory variables: Operating models where established by finding the set

of explanatory variables that could best predict the analyte value on any day. Water

temperature was a potential explanatory variable for many analytes, but was not used because

of missing data problems. Therefore harmonics for within a day and across the year was used

as a surrogate for temperature. The following explanatory variables were considered:

• Daily and annual harmonics

• Flow variables – mean flow, maximum flow and flow velocity at representative sites

for each water catchment

• Storage level at the specified sites

• Depth

• Depth interactions with harmonics and/or flows.

Model selection: The model selection process was based on the generalized additive

modelling approach using the gamlss library of functions in R-code (Stasinopoulos, Rigby

and Akantziliotou [2006]). The gamlss library is a collection of functions to fit Generalized

Additive Models for Location, Scale and Shape using the R package. Generalized Additive

Models for Location, Scale and Shape (gamlss) were introduced as a way of overcoming

some of the limitations associated with Generalized Linear Models (GLM) and Generalized

Additive Models (GAM) (e.g., see Hastie and Tibshirani [1990]). In gamlss, the exponential

family distribution assumption for the response variable (y) is relaxed and replaced by a

general distribution family, including highly skewed and/or kurtotic distributions. The

following subset of the available gamlss distributions was used for model selection and was

considered sufficiently versatile to model the analytes:

1. Normal

2. Log-normal

3. Inverse Gaussian

4. Zero adjusted Inverse Gaussian where the low detects or zeros are fitted using a

logistic regression model and the remaining values are fitted using the Inverse

Gaussian distribution.

The models were fitted using maximum (penalised) likelihood estimation. This is

implemented in the gamlss package in R; see Stasinopoulos et al. [2006].

The EM algorithm and/or MCMC algorithm (e.g., see Sparks et al. [2011]) was used to

deal with partial missing information where analyte values have exceeded their measurement

limits (e.g., the measured values are only known to be below 10).

Variable selection process: For each model, a forward selection approach was used to

include variables that were significant in a step-wise fashion, one explanatory variable at a

time. A first explanatory variable was selected that best explained the analyte’s behaviour on

its own. Subsequent explanatory variables were included if and when they:

a) reduced the model’s AIC (Akaike Information Criterion) when included; and

b) reduced the model’s AIC more than the other remaining explanatory variables.

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

Model selection (variables and distribution) of every analyte-site combination proved too

onerous a task to be carried out independently in a hands-on manual fashion, because of the

number of combinations of site and analytes. Therefore the process was automated so that it

was repeatable for different time periods. Several rules needed to be developed (such as small

observation sizes limit the complexity of the model selected) so that reasonable models would

be selected in the automatic model selection process.

Figure 4. Time series plot of pH at Wollondilly River for the full period from 1990 to the end of 2006.

An example of model fit is presented in Figure 4 for pH for 1991 to 2006. The black

circles are the measured values, the blue points are the model fitted values and the green lines

are the prediction bounds. The red lines are the quality thresholds for pH. Notice that Sydney

Catchment Authority was able to deliver within the threshold limits by the end of 2006, while

at the end of 1990 it was not. That is, there has been a steady reduction from high alkalinity in

1990 to a more acceptable range in 2006. In addition, the pH values are fairly predictable

meaning that they are largely in-control. Notice the model does on occasions produce strange

predictions when extrapolating beyond the range of flows experienced in the past dataset.

These can be seen by the prediction intervals dramatically dropping in values in Figure 4. In

the future, the historical data used to fit the models will include more extreme events and thus

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 49

extrapolation will become less likely. Therefore, having a long historical dataset is essential

to avoiding this issue in estimating the risks of exceeding thresholds.

CONSEQUENCE - COST ESTIMATES

The cost of a water quality exceedance is viewed from a water consumer perspective –

that is, the consequence is either a loss of consumer welfare, e.g., health or aesthetic (taste,

odour), or alternatively, the opportunity cost of mitigation actions at multiple barrier

exceedances in the water supply network to reduce the risk of this consequence to the

consumer occurring. Ideally, sufficient analysis would be undertaken to ensure that the

selection between and within these alternatives is optimised so that the least cost consequence

is selected in all cases. In this application, the available information has not allowed more

than a rudimentary optimisation. Further development of this technique should focus on this

aspect. However, with this caveat, the consequence of a given breach in a particular analyte’s

thresholds (usually three thresholds have been set for each analyte - minor, major and

emergency) would vary according to:

• The location of the testing station – the further upstream the greater the probability

that this threshold will not be breached at downstream locations, and that therefore

there will be no consumer losses or mitigation action required to avoid them.

• The size and location of the community being supplied with water from a particular

source (the representative monitoring station).

o There are strong economies of scale for water supply for lower level breaches of

water quality thresholds – water supplies to large communities can have typical

contaminants treated for lower unit cost, thus significantly reducing the costs of

addressing a given low level water quality event.

o Simultaneously, there are strong diseconomies of scale for water supply for

higher level breaches, particularly when breaches are uncorrelated between water

sources. To explain, for breaches in small water supply systems (e.g. less than

100 connections), when other water sources are not in breach, alternative

supplies can be trucked in at a cost of around $7 per KL, or less than $5,000 per

day total. Whereas, for breaches in large water quality systems, logistic and

congestion costs may make alternative supplies uneconomic. In these situations,

least cost short term solutions such as boil water alerts have been estimated

(Sloane Cook and King [2004] and Jaguar Consulting Pty Ltd [2004]) to cost

from $78M to $350M in 2004 dollars, or roughly an increase of $0.90 to $3 Per

KL supplied, depending on the extent of consumer costs (time lost, income,

health etc.) included. Some of this high cost is caused by the mitigation not being

able to address this risk in full, and so costly consequences to health can ensue.

Harrington et al. [1991] details how pathogen contamination can lead to such

high consumer costs. Note that it is the volume of contaminated water and the

subsequent impact on consumers that drives total cost, and that therefore creates

the diseconomy. (For such systems, longer term solutions such as diversifying

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

supplies may be required, which would have minimum costs of the order of

$1/KL (transfers) to $3 per KL (desalination). However, long term capital

solutions to address infrequent water quality events can result in poorly utilised

capital, with sometimes exorbitantly high unit costs.

• The type of response by consumers.

o For example, even when a boil water alert is in place, many consumers will not

deem the risk great enough to comply, while others will be ignorant of the alert.

In a situation where there are no known clinical consequences (e.g. Sydney in

1998), these people will not bear any costs. However, where there are, and

people may suffer chronic or acute ill health, they will bear significant loss of

income and direct or indirect health care costs (e.g., increase in insurance

premiums).

o Alternatively, water quality perceptions can be changed by a well publicised

breach, even where there is no health or aesthetic consequence. Consumers will

increase averting behaviours such as purchasing bottled water, which can be very

costly (e.g., thousand fold increase from $2 per kL to greater than $2 per L),

even though the total quantity drunk is small.

• The nature of the water supply

o The correlation between water quality in different parts of the network is a major

driver of cost. All things being equal, spatially uncorrelated water quality events

in the bulk water supply system are lower cost, with fewer consumers impacted,

if the system allows water to be supplied from uncontaminated sources.

o The frequency of the breach – where breaches are unusual, there will be one-off

incident costs (incident management, ad hoc testing, overtime, etc.). These costs

can be in the order of $50,000 to $750,000 per event even where it is later

established a breach did not cause impact to consumers (SCA [2008]). Where

breaches are frequent, lower unit-cost-per-incident responses can be put in place

(e.g., putting increased monitoring on routine programs rather than ad hoc

delivery). See Miller et al. [2004], Pg. 44ff for a discussion of consequence

scoring for raw water sources.

There are a wide range of considerations that apply in modelling costs, and these become

more complex the greater the level of accuracy that is required. It has not been possible to

cover all issues, but the following pragmatic approach has been applied.

Costing Approach

The approach taken is to give all analytes a cost, even when this cost cannot be accurately

estimated. This approach is analogous to a Bayesian prior, and can be refined in subsequent

development.

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 51

Analytes have been grouped according to the types of mitigation cost that they would

normally incur, and an approximate cost allocated to that group. All costs are divided into

three groupings: the cost of exceeding a minor threshold (labelled 1st), the cost of a single day

exceedance of a major threshold (2nd) and the cost when consecutive days exceed the major

threshold (3rd).

Note that all costs are on an incremental basis. That is, the cost does not include any

baseline or routine monitoring or treatment, only the costs incremental (additional) to this

baseline cost.

The key groups were:

Analytes with pathogenic impact but difficult to treat:

• Giardia and Cryptosporidium per event at each threshold 1st $10,000, 2nd $100,000,

3rd $4,680,000. The huge cost increase at the higher thresholds is based on boil water

costs to community and disease burden (see Sloane Cook and King [2004] and

Jaguar Consulting [2004]). As noted above, these studies estimated total incident cost

for these kinds of events in the tens and hundreds of millions, which gives very high

daily and consecutive day costs for incidents lasting for a few months at the

maximum.

Analytes treatable by disinfection:

• E. Coli as indicator of treatable pathogen presence. Cost per event at each threshold

1st $1,000, 2nd $15,000, 3rd $10,000. This cost curve assumes incremental disinfectant

and monitoring costs. For pathogens, the fact that the routine operational monitoring

is frequent, means that the incremental monitoring cost for an incident is low).

• For other treatable pathogens, the incremental cost is assumed to be $1,000 at first

threshold only reflecting treatment, as monitoring of indicator pathogens would be

covered by the other item.

Analytes not readily1 treatable such as pesticides:

• Cost per event at each threshold 1st - $5,000, 2nd $10,000, 3rd $30,000 which includes

intensive monitoring and incident management to determine appropriate responses

(e.g., Water source management, or changing farm or forest pesticide practices).

These costs are based on the lower bound historical incident management cost cited

above $50,000. This is incremental monitoring only, but non-routine (cf. e.g., E.coli).

• True colour – No costs for any thresholds. Penalties related to this item were

removed from the Bulk Water Supply Agreement with SCA’s major customers in

2003.

1 Some pesticides may be removed in small systems (cf. cyanobacteria) using activated carbon, e.g. US EPA.

[2000]. However, in Australian coastal urban water systems, these treatments are not typically available in

plant infrastructure because influent is normally high quality. Thus, activated carbon would have high

incremental cost, .

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

Figure 5. The estimated costs for each analyte at the major drinking water supply point of Lake

Burragorang Dam Wall (OM denotes Organic matter and Sed denotes sediment (turbidity)).

Analytes with biological impact:

• Microcystin LR equiv, Toxic Cyanobacterial Count (cells/mL), Toxic Cyanobacterial

biovolume (mm3/L) 1st $100,000, 2nd $200,000 and 3rd $1,000,000 involving

treatment costs.

• This costing is based on applying activated carbon treatment if available, or where

not, as in a major incident response, on the opportunity cost of loss of water use to

consumers. No health impact has been assumed.

• Areal Standard Unit (algae) 1st $50,000, 2nd $150,000, 3rd $200,000, where the first

threshold cost only includes extra monitoring and management costs. These relate to

incident costs. The second and third thresholds include incremental treatment costs

for algae, but do not assume any loss of supply to consumers (cf. toxic algae above).

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 53

• Chlorophyll–a – 1st $50,000. These relate to incident costs only, as explained above

for the Areal Standard unit measure.

Analytes with chemical impact:

• This includes various forms of nitrogen, e.g., nitrate or ammonia. For these analytes,

a single 1st of $10,000 was set, based on incident costings.

Figure 6. The estimated costs for each analyte at other places in the drinking water supply chain.

RESULTS

The results of applying the above theory are now presented. The annual risk estimates are

provided for 2006, with likelihoods and consequence estimates being based on the

interpolated values from each analyte’s gamlss model, which were fitted using known flows

for 2006, annual and daily harmonics and time as explanatory variables.

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

Figures 5 to 7 present the annual risk estimates for the SCA supply sites as barplots, with

estimates truncated at 5 million dollars. The vertical dashed lines separate the analytes into

their higher-level groupings, such as Microbial pathogens and Algae, as labelled. There are

too many analytes to label the bars at the individual analyte level.

Instead, the analytes with high estimated annual risk are summarised in Table 1. Note that

not all analytes are measured at each site.

Table 1. Summary of analytes with high estimated annual risk, by location

Water

Quality

Hazard

Analyte

Burragorang

Wingercarri

Lake

Yarrunga

Woronora

Lake

Illawara

epean

mircrobial

pathogens

Cryptosporidim x

Giardia x

algae areal standard

unit

x x x x x x x

toxic

cyanobacterial

x x x

heavy

metals

barium x x

boron x x

Organic Total organic

carbon

Nutrients nitrogen

oxidised

x x

sediments turbidity x x

Non-

metals

sulphate x

chlorine free x

physical alkalinity x x x

total hardness x x x x x

dissolved

oxygen

x x x x

temperature x x

The sites investigated are Burragorang, Wingecarribee, Lake Yarrunga, Woronora, Lake

Prospect, Illawarra and Lake Nepean. The full list of the analytes is recorded in Table 2 in

Appendix A together with their threshold values. The source of the threshold values are given

in the last column of Table 2 in Appendix A.

Figure 5 presents the results for the supply site: Lake Burragorang @ 500m u/s Dam Wall

(DWA2). The dam wall, which is 500 metres downstream from the Lake Burragorang

measurement site, is the major source of drinking water supplied to Sydney.

Figures 6 and 7 present the results for upstream supply sites. These annual risk (cost)

estimates are not assumed accurate in an absolute sense. Instead, they should be interpreted in

a relative sense, with relatively high values being considered indicative of high-risk areas.

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 55

Figure 7. The estimated costs for each analyte at other places in the drinking water supply chain.

Table 1 presents the analytes posing high estimated annual risk for each supply site.

There is a high risk from algae, especially areal standard unit algae, across all supply sites.

This is in part due to the high cost of occasional blooms, but this analysis highlights a need to

investigate the causes of the high algae levels and highlights that there are potentially large

benefits to be gained from investing in preventative measures, as part of the water catchment

management.

Exact interpretation and consideration of the biology is beyond the scope of this chapter

and is left to experts in water ecology.

The other area of common high risk is the physical higher-level water quality hazard

grouping. The analysis indicates that alkalinity and hardness both tend to have a high

estimated risk (cost) and that it would be prudent to investigate possible management

methods. These may be difficult of manage, considering the geology of the catchment area.

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

FUTURE RISK ASSESSMENT CHALLENGES

Estimating the costs associated with past events and using these to estimate future costs is

made difficult by the multivariate nature of these events. Generally, analyte values are

correlated with time, and therefore often several thresholds are exceeded simultaneously. In

this chapter, cost estimates have been derived univariately, which is largely because the cost

thresholds were assumed univariate. If water treatment costs are triggered by univariate

threshold exceedances, then these costs may reflect accurately the nature of the operational

costs. However, the true cost to society is unlikely to be reflected in univariate rectangular

cost boundaries.

The challenge of defining thresholds that account for the multivariate nature of the

problem is enormous. However, if true society risk estimates are required then working

towards this objective is necessary. An initial approach to this end would be to approximate

the cost boundaries by simple hyper-ellipsoidal lines and estimate the likelihoods under the

assumption that analyte values are normally distributed, so that distributions of quadratic

forms could be used to estimate the likelihood of such exceedances.

The current approach, at best, provides users with the analyte grouping that are more

likely to exceed univariate quality thresholds. This is useful in that it helps decide which

analytes need to be more closely monitored, however, the risk estimates derived in this

chapter are not an accurate reflection of the operational costs resulting from poor water

quality, and are at best interpreted in a relative sense.

Most catchment managers are interested in whether they are improving the condition of

their catchment. However, this can only be efficiently achieved after correcting for changes in

climate and environmental conditions from year to year. Risk adjusted control charts are used

in health surveillance (e.g., Steiner et al. [2000]) and provide a methodology for achieving

this aim. These charts would first correct for the changes in climate and the environment that

are beyond the managers’ control by removing their influence on the analyte trends.

Examples for climate might be temperature, rainfall, flows, etc, and for the environment,

might be the number of wildlife in the catchment, the number of livestock, percentage trees,

percentage of land under agriculture, etc. Adjusted scores, derived from the adjusted trends,

could then be interpreted as measures of improvement in the catchment that are attributable to

the managers’ actions.

CONCLUSION

This chapter examines the Sydney Catchment Authority’s operating risk relating to

treatment costs and costs of managing poor drinking water outbreaks for individual analytes.

The risk for each analyte at each site was calculated by combining a fitted time-dependent

model for the analyte that gave the distribution of analyte values for each day, with a cost

profile for the analyte exceeding certain thresholds.

The analysis was performed on previously collected analyte measurements that were

made irregularly and often infrequently, which posed modelling challenges. The modelling

was performed using gamlss models. Estimates of the uncertainties in the likelihood and

annual risk estimates were derived that accounted for autocorrelation in the analytes.

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 57

Some conceptual changes could improve the cost estimates. One improvement would be

to consider consequences outside the Australian Drinking Water Guidelines, such as the

health and other consequences for the customers of drinking the water or to consider the

aquatic environmental concerns. Another possible improvement would be to treat risk as

multivariate by attempting to estimate the true cost of multivariate hazard events rather than

approximate it as the sum of multiple univariate hazard events. Both these objectives offer

potential future research opportunities.

REFERENCES

Abera, S., Zeyinudin, A., Kebede, B., Deribew, A., Ali, S. and Zemene, E. (2011).

Bacteriological analysis of drinking water sources. African Journal of Mibrobiology

Research. 5: 2638-2641.

Chowdhury, S., Champagne, P. and McLellan, P.J. (2009). Uncertainty characterization

approaches for risk assessment of DPBs in drinking water: a review. J. Environ. Manage.

90:1680-1691.

Cogley, JG (1999). Effective sample size for glacier mass balance. Geografiska Annaler. 81;

497-507.

Cross, A., Cantor, K.P., Reif, J.S., Lynch, C.F., Ward, M.H. (2004). Pancreatic cancer and

drinking water and dietary sources of nitrate and nitrite. American Journal of

Epidemiology. 159: 693-701.

Hales, S., Black, W., Skelly, C., Salmond, C., and Weinstein, P. (2003). Social deprivation

and public health risks of community drinking water supplies in New Zealand. J.

Epidemiol. Community Health. 2003;57:581-583 doi:10.1136/jech.57.8.581.

Harrington, W. Krupnick, A. J. Spofford W. O. (1991) Economics and episodic disease: the

benefits of preventing a giardiasis outbreak Resources for the Future. Quality of the

Environment Division.

Fawell, J and Nieuwenhuijsen, M.J. (2008), Contaminants in drinking water: Environmental

pollution and health. British Medical Bulletin. 68 (1): 199-208.

Jaguar Consulting Pty Ltd (2004) Drinking Water Quality Regulatory Framework For

Victoria Regulatory Impact Statement For The Safe Drinking Water Regulations, for

Department of Human Services, Victoria, September.

Krewski, D., Balbus, J., Butler-Jones, D., Haas, C., Isaac-Renton, J., Roberts, K. and Sinclair,

M. (2004). Managing the mirobiological risks of drinking water. Journal of Toxicology

and Environmental Health Part A. 67: 1591-1617.

Miller, R, Guice, J and Deere, D. (2004) Risk Assessment for Drinking Water Sources CRC

for Water Quality and Treatment, Research Report No 78 CRC for Water Quality and

Treatment

Morris, R.D. (1995). Drinking water and cancer. Environ. Health Perspect. 103: 225-231.

Pavlov, D., de Wet, C.M.E., Grabow, W.O.K., Ehlers, M.M. (2004). Potentially pathogenic

features of hterotrophic plate countbecteria isolated from treated and untreated drinking

water. International Journal of Food Microbiology. 92: 275-287.

Rigby, R. A. and Stasinopoulos D. M. (2005). Generalized additive models for location, scale

and shape,(with discussion), Appl. Statist., 54, part 3, pp 507-554.

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

Rosen, L., Lindhe, A., Bergstedt, O., Norberg, T. and Pettersson, T.J.R. (2010). Comparing

risk-reduction measures to reach water safety targets using an integrated fault tree model.

Water Science and Technology Water Supply. 10: 428-436.

Sloane Cook and King Pty Ltd (2004) Prospect Raw Water Pumping Station Economic

Appraisal Final Draft, NSW Department Of Commerce, May.

Sparks, RS, Sutton, G. and Toscas, P. (2011). “Modeling inverse Gaussian random variables

when there are missing data and low detection limits. Advances in Decision Sciences.

doi:10.1155/2011/571768.

Stasinopoulos D. M., Rigby R.A. and Akantziliotou C. (2006) Instructions on how to use the

GAMLSS package in R. Accompanying documentation in the current GAMLSS help

files, (see also http://www.gamlss.org/).

Stasinopoulos D. M. Rigby R.A. (2007) Generalized additive models for location scale and

shape (GAMLSS) in R.Journal of Statistical Software, Vol. 23, Issue 7, Dec 2007,

http://www.jstatsoft.org/v23/i07.

Steiner, S. H., Cook, R.J., Farewell, V.T., Treasure, T. (2000). Monitoring surgical

performance using risk-adjusted cumulative sum charts. Biostatistics. 1: 441-452.

Thiebaux, HJ and Talkner, P (1984). On the interpretation and estimation of effective sample

size. J. Climate Appl. Meteorol., 23; 800-811.

Urbansky, E.T. and Schock, M.R. (1999). Issues in managing the risks associated with

perchlorate in drinking water. Journal of Environment Management. 56: 79-95.

US EPA (2000). Summary of Pesticide Removal/Transformation Efficiencies from Various

Drinking Water Treatment Processes - Prepared for the Committee to Advise on

Reassessment And Transition (CARAT) October 3

van Grinsven, H.J., Ward, M.H., Benjamin, N., and de Kok , T.M. (2006). Does the evidence

about health risks associated with nitrate ingestion warrant an increase of nitrate standard

for drinking water? Environmental Health. 5:26. DOI: 10.1186/1476-069X-5-26.

Yang, C. (1998). Calcium and magnesium in drinking water and risk of death from

cerebrovascular disease. Stroke. 29: 411-414.

ATTACHMENT A– ANALYTES FOR ASSESSMENT, FOLLOWED BY AN

EXAMPLE DATA SET

Table 2. Analytes to be assessed under each water quality hazard group

Water Quality

Hazard Group

Analyte Concentration

Limit for

Assessment

Guideline

Microbial

Pathogens

E. coli (cfu/100mL) 500 BWSA

Coliforms Total (cfu/100mL) 1000 BWQIRP

Enterococci (cfu/100mL) 35 ANZECC

Cryptosporidium (oocysts/100L) 1 BWQIRP

Giardia (cysts/100L) 1 BWQIRP

Adenovirus 1 ADWG

Enterovirus 1 ADWG

F-RNA Phage 1 ADWG

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 59

Water Quality

Hazard Group

Analyte Concentration

Limit for

Assessment

Guideline

Reovirus (Reoviridae) 1 ADWG

Algae

Areal Standard Unit (algae) 500 BWSA

Chlorophyll a (μg/L) 5 ANZECC

Microcystin LR equiv (μg/L) 1.3 BWQIRP

Toxic Cyanobacterial Biovolume (mm3/L) 0.4 BWQIRP

Toxicigenic Cynobacterial Count

(cells/mL)

5000 BWQIRP

Metals

Aluminium (acid soluble) (mg/L) 0.2 ADWG

Aluminium Total (mg/L) 1.0 BWQIRP

Iron Total (mg/L) 0.3 ADWG

Manganese Total (mg/L) 0.1 ADWG

Sodium Total (mg/L) 180 ADWG

Heavy Metals

Total Antimony (mg/L) 0.003 ADWG

Total Arsenic (mg/L) 0.007 ADWG

Total Barium (mg/L) 0.7 ADWG

Total Boron (mg/L) 4 ADWG

Total Cadmium(mg/L) 0.002 ADWG

Total Chromium (mg/L) 0.05 ADWG

Total Copper (mg/L) 2.0 ADWG

Total Lead (mg/L) 0.01 ADWG

Total Mercury (mg/L) 0.001 ADWG

Total Molybdenum (mg/L) 0.05 ADWG

Total Nickel (mg/L) 0.02 ADWG

Total Selenium (mg/L) 0.01 ADWG

Total Silver (mg/L) 0.1 ADWG

Total Uranium (mg/L) 0.02 ADWG

Total Zinc (μg/L) 8 ANZECC

Pesticides

Aldrin (μg/L) 0.01 ADWG

Amitrole (μg/L) 1 ADWG

Atrazine (μg/L) 0.1 ADWG

Chlordane (μg/L) 0.01 ADWG

Chlorpyrifos (μg/L) 0.1 ADWG

Clopyralid (μg/L) 1000 ADWG

2,4 Dichlorophenooxyacetic acid (μg/L) 0.1 ADWG

DDT (μg/L) 0.06 ADWG

Dieldrin (μg/L) 0.01 ADWG

Diquat (μg/L) 0.5 ADWG

Diuron (μg/L) 30 ADWG

Total Endosulfan (μg/L) 0.05 ADWG

Heptachlor (μg/L) 0.05 ADWG

Nova Science Publishers, Inc.

Ross Sparks, Gordon J. Sutton, Peter Toscas et al.

Table 2. (Continued)

Water Quality

Hazard Group

Analyte Concentration

Limit for

Assessment

Guideline

Hexazinone (μg/L) 2 ADWG

Lindane (μg/L) 0.05 ADWG

Molinate (μg/L) 0.5 ADWG

Paraquat (μg/L) 1 ADWG

Picloram (μg/L) 300 ADWG

Propiconazole (μg/L) 0.1 ADWG

Temephos (μg/L) 300 ADWG

Triclopyr (μg/L) 10 ADWG

2,4,5-T 0.05 ADWG

Organic

Matter

Total Organic Carbon (mg/L) 10 SCA Science

and Research

Nutrients

Ammonia (as NH3) (µS/cm) 0.5 ADWG

Ammonium (mg/L) 0.01 ANZECC

Nitrate as N (mg/L) 50 ANZECC

Nitrogen Oxidised (mg/L) (NOx) 0.01 ANZECC

Nitrogen Total (mg/L) 0.35 ANZECC

Phosphorus Filterable (mg/L) 0.005 ANZECC

Phosphorus Total (mg/L) 0.01 ANZECC

Sediment Turbidity Field (NTU) 40 BWQIRP

Total Dissolved Solids (mg/L) 1000 ANZECC

Non-metals Chloride (mg/L) 250 ADWG

Chlorine Free (mg/L) 0.6 ADWG

Cyanide (mg/L) 0.08 ADWG

Iodide (mg/L) 0.001 BWQIRP

Sulphate (mg/L) 250 ADWG

Non-metals

Chloride (mg/L) 250 ADWG

Chlorine Free (mg/L) 0.6 ADWG

Cyanide (mg/L) 0.08 ADWG

Iodide (mg/L) 0.001 BWQIRP

Sulphate (mg/L) 250 ADWG

Physical

Alkalinity (mg as CaCO3/L) 15 - 60 BWSA

Conductivity Field (µS/cm) 350 ANZECC

Dissolved Oxygen (%Sat) 90-110 ANZECC

pH (Field) 5.5 – 8.5 BWQIRP

Temperature (Deg C) 5.0-25 BWSA

Total Hardness (mgCaCO3/L) 25 – 70 BWSA

Nova Science Publishers, Inc.

Estimates of Likelihood and Risk Associated with Sydney Drinking Water Supply … 61

Water Quality

Hazard Group

Analyte Concentration

Limit for

Assessment

Guideline

True Colour at 400nm 40 BWQIRP

Radiological

Gross alpha emitters (Bq/L) 0.5 ADWG

Gross beta emitters (Bq/L) 0.5 ADWG

BWSA = Bulk Water Supply Agreement; ADWG = Australian Drinking Water Guidelines; ANZECC

= ANZECC Water Quality Guidelines; BWQIRP = Bulk Water Quality Incident Response Plan.

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 3

THREE-DIMENSIONAL NUMERICAL MODELING OF

WATER QUALITY AND SEDIMENT-ASSOCIATED

PROCESSES IN NATURAL LAKES

Xiaobo Chao* and Yafei Jia

National Center for Computational Hydroscience and Engineering,

The University of Mississippi, US

ABSTRACT

This chapter presents the development and application of a three-dimensional water

quality model for predicting the distributions of nutrients, phytoplankton, dissolved

oxygen, etc., in natural lakes. In this model, the computational domain was divided into

two parts: the water column and the bed sediment layer, and the water quality processes

in these two domains were considered. Three major sediment-associated water quality

processes were simulated, including the effect of sediment on the light intensity for the

growth of phytoplankton, the adsorption-desorption of nutrients by sediment and the

release of nutrients from bed sediment layer. This model was first verified using

analytical solutions for the transport of non-conservative substances in open channel

flow, and then calibrated and validated by the field measurements conducted in a natural

oxbow lake in Mississippi. The simulated concentrations of water quality constituents

were generally in good agreement with field observations. This study shows that there are

strong interactions between sediment and water quality constituents.

INTRODUCTION

Sediment is a major nonpoint source pollutant. It may be transported into surface water

bodies from agricultural lands and watersheds through runoff. These sediments could be

associated with nutrients, pesticides, and other pollutants, and greatly affect the surface water

qualities. Therefore, sediment has been listed as the most common pollutant in rivers,

* E-mail address: chao@ncche.olemiss.edu.

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

streams, lakes, and reservoirs by the US Environmental Protection Agency (USEPA).

Sediment creates turbidity in water bodies, reducing the light intensity in water columns,

which is one of the most important factors affecting the phytoplankton growth. In water

bodies, nutrients can interact with sediment particles through the processes of adsorption and

desorption. In addition, nutrients in bed sediments may be released into the water column.

Those sediment-associated processes play important roles in water quality interaction

systems.

Numerical modeling is a very effective approach to study water quality constituents in

surface water bodies. In recent years, some well-established three dimensional models, such

as WASP6[1], CE-QUAL-ICM[2], Delft3D-WAQ[3], MIKE3_WQ[4], MOHID[5], EFDC

[6], etc., have been used to simulate water quality constituents in rivers, lakes, and coastal

waters. These models generally cover basic physical, chemical and biological processes of

aquatic ecosystems. However, only a few are capable of simulating the effects of sediment on

the water quality. In WASP6, CE-QUAL-ICM, and MIKE3-WQ, the effect of suspended

sediment (SS) concentration on the growth of phytoplankton was not taken into account. The

processes of adsorption-desorption were not simulated in MIKE3-WQ and MOHID. In

WASP6, CE-QUAL-ICM and EFDC, the adsorption-desorption of nutrients by sediment

were described using a simple linear isotherm. In the above mentioned models, the release

rate of nutrients from the bed sediment was determined based on the concentration gradient

across the water-sediment interface; however, the effects of pH and dissolved oxygen

concentration on the release rate were not considered. In view of the limited understanding of

the sediment-associated water quality processes, the accuracy of simulations using the

aforementioned models has room for improvement.

The development and application of a three-dimensional model for simulating the water

quality constituents in natural lakes were presented. The computational domain was divided

into two parts: the water column and the bed sediment layer. This model was decoupled with

a three-dimensional free surface hydrodynamics model CCHE3D [7], and the major

sediment-associated processes were simulated, including the effect of sediment on light

penetration, the adsorption-desorption of nutrients by sediment and the release of nutrients

from bed sediment.

This model was first verified by a mathematic solution consisting of the movement of a

non-conservative tracer in a prismatic channel with uniform flow, and the numerical results

agreed well with the analytical solutions. Then it was applied to a real case to validate its

capability for simulating the concentrations of phytoplankton and nutrients in Deep Hollow

Lake, a small oxbow lake in the Mississippi Delta. The concentrations of water quality

constituents obtained from the numerical model were generally in good agreement with

observations.

This chapter presents detailed technical information on the modeling of water quality and

sediment-associated processes in natural lakes. The general water quality processes

considered in the numerical model, including the phytoplankton kinetics, nitrogen cycle,

phosphorus cycle, dissolved oxygen balance, and processes in benthic sediment layer are

described first, and then the sediment-associated water quality processes are presented. The

detailed information on the development, verification, and application of a three-dimensional

water quality model are given in the following three Sections. The discussion and conclusion

are described in the last two Sections.

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 65

WATER QUALITY PROCESSES

Water quality is scaled by a combination of many chemical variables, which interact

based on laws of chemistry and bio-chemistry under natural conditions. The interactions of

water quality constituents in water column and sediment layer are shown in Figure 1.

Figure 1. Interactions of water quality constituents in water column and sediment layer.

In the water column, four biochemical processes were considered: the phytoplankton

kinetics, nitrogen cycle, phosphorus cycle, and dissolved oxygen (DO) balance.

The conceptual framework for the eutrophication kinetics was mainly based on the

WASP6 model [1]. Eight state variables were involved in the interacting systems: ammonia

nitrogen (NH3), nitrate nitrogen (NO3), phosphate (PO4), phytoplankton (PHYTO),

carbonaceous biochemical oxygen demand (CBOD), DO, organic nitrogen (ON), and organic

phosphorus (OP).

In bed sediment layer, the decomposition of organic material releases nutrients to the

sediment porous water and also results in the exertion of oxygen demand at the sediment-

water interface. Similar to the water column, eight state variables are involved, including

NH3, NO3, PO4, PHYTO, CBOD, DO, ON, and OP.

Phytoplankton Kinetics

Phytoplankton are microscopic plant-like organisms that live in water environments, play

a central role in aquatic ecosystem. Via photosynthesis, phytoplankton produce organic

compounds from inorganic nutrients. They consume nutrients, including inorganic nitrogen,

phosphate, silica, and carbon dioxide from the water and release oxygen as a by-product to

the water. The source term Sm due to phytoplankton growth and reduction is calculated by

Light, Temperature

Water

Column

CBOD

C: N: P

DO PHYTO

Reaeration

Sediment

Layer CBOD DO PHYTO OP

ONNH

Denitrification

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

()

mppset

SGDPM=−− (1)

where M is the phytoplankton biomass; Gp is the growth rate of phytoplankton (day-1); Dp is

the reduction rate of phytoplankton (day-1); and Pset is the effective phytoplankton settling

rate(day-1). In the model, the total chlorophyll is used as a simple measure of phytoplankton

biomass. The relationship between the chlorophyll and phytoplankton biomass can be

expressed as

chl

= (2)

in which Cchl is the chlorophyll concentration (mg/l);

is the carbon to chlorophyll ratio.

The growth rate of phytoplankton is determined by the availability of nutrients, the

intensity of light, and by the ambient temperature. The effects of each factor are considered to

be multiplicative:

TINmxp fffPG = (3)

in which fN, fI and fT are the limitation factors due to nutrient availability, light intensity, and

temperature, respectively; Pmx is the maximum phytoplankton growth rate (day-1).

The limitation factor fN is calculated based on Michaelis-Menten Equation and Liebig’s

law of the minimum [1]:

⎟

⎠

⎞

⎜

⎝

⎛

+++

mPPO

mNNONH

NONH

NKC

KCC

33 ,min (4)

in which CNH3, CNO3 and CPO4 are the concentrations of NH3, NO3 and PO4, respectively

(mg/l); KmN and KmP are the half-saturation constants for nitrogen and phosphorus uptake,

respectively(mg/l).

The limitation factor fI is calculated by integrating the Steele equation over water depth

and time [8]:

⎥

⎦

⎤

⎢

⎣

⎡

⎟

⎠

⎞

⎜

⎝

⎛−−

⎟

⎠

⎞

⎜

⎝

⎛−

=⋅−Δ+− d

)(

0expexp

72.2 (5)

where fd is the photoperiod;

z is the vertical thickness of a computational element (m); zd is

the distance from the water surface to the top of a computational element in the water column

(m); Ke is the total light attenuation coefficient (m-1), it is determined by the pure water and

the concentrations of chlorophyll and suspended sediment in the water column; I0 is the daily

light intensity at the water surface (ly/day); Im is the saturation light intensity of

phytoplankton (ly/day).

The field observations [2,6] show that there is an optimum temperature Tm (oC) for the

growth of phytoplankton. When the temperature T (

oC) is below Tm, the growth rate of

phytoplankton increases as a function of T; while the growth rate decreases when the

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 67

temperature T is above Tm. The limitation factor fT can be calculated using formulas proposed

by Cerco and Cole [2]:

()

[

]

exp mT TTKTgf −−= when m

TT ≤ (6)

()

[

]

exp TTKTgf mT −−= when m

TT > (7)

in which KTg1 and KTg2 are coefficients representing the effects of temperature on the

phytoplankton growth below and above Tm, respectively.

Phytoplankton losses mainly include endogenous respiration, death and grazing by

zooplankton. The reduction rate of phytoplankton is given as follows:

2020 −− ++= T

pzgzoopzgpd

prprp CkkkD

θθ

(8)

where kpr and kpd are the rates of endogenous respiration and death, respectively (day-1); kpzg is

the zooplankton grazing rate (l/mgC/day); Czoo is the concentration of zooplankton(mgC/l);

pr and

pzg are the temperature coefficients.

The effective phytoplankton settling rate Pset can be given as:

set D

= (9)

where

s4 is the settling velocity of phytoplankton (m/day) and Dj is the depth of

element j (m).

Nitrogen Cycle

The major components of nitrogen cycle in aquatic environments include: ammonia

nitrogen (NH3), nitrate nitrogen (NO3), phytoplankton (PHYTO) and organic nitrogen (ON).

NH3 and NO3 are consumed by phytoplankton for its growth. Due to physiological

reasons, NH3 is the preferred form of inorganic nitrogen for phytoplankton. Nitrogen is

returned from the phytoplankton biomass pool to particulate and dissolved organic nitrogen

pools as a result of phytoplankton death, endogenous respiration and zooplankton grazing.

ON is converted to NH3 at a temperature-dependent mineralization rate, and NH3 is then

converted to NO3 at a temperature- and oxygen-dependent nitrification rate. In the absence of

oxygen, NO3 can be converted to nitrogen gas (denitrification) at a temperature-dependent

rate. Nitrogen may interact with sediment through the processes of adsorption and desorption.

Dissolved inorganic and organic nitrogen at the bed sediment layer may also be released to

the water column under certain conditions.

In the numerical model, the kinetic source terms for NH3, NO3 and ON were calculated

using formulas presented by Wool et al. [1].

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

Phosphorus Cycle

The major components of phosphorus cycle in aquatic environments includes: phosphate

(PO4), phytoplankton (PHYTO) and organic phosphorus (OP).

Phosphorus kinetics are basically similar to the nitrogen kinetics except there is no

process analogous to denitrification. PO4 is utilized by phytoplankton for its growth and is

incorporated into phytoplankton biomass. Phosphorus is returned to the water column from

dead or decaying phytoplankton biomass in the bed sediment. The various forms of OP

undergo settling, hydrolysis and mineralization, and are converted to inorganic phosphorus at

temperature-dependent rates. In addition, phosphorus may interact with sediment through the

processes of adsorption and desorption. Dissolved inorganic and organic phosphorus at the

bed sediment layer may also be released to the water column under certain conditions.

In the numerical model, the kinetic source terms for PO4 and OP were calculated using

formulas summarized by Wool et al. [1].

Dissolved Oxygen Balance

Dissolved oxygen (DO) is one of the most important parameters in water quality analysis

and is used to measure the amount of oxygen available for biochemical activity in waters.

Five water quality constituents, including ammonia nitrogen (NH3), nitrate nitrogen (NO3),

phytoplankton (PHYTO), carbonaceous biochemical oxygen demand (CBOD) and DO, are

involved in the DO processes. The decomposition of organic material in benthic sediment can

greatly reduce the concentration of DO in the water column. In the model, sediment oxygen

demand (SOD) is used to calculate the sink of oxygen due to this process.

The level of DO is increased due to the atmospheric reaeration and photosynthesis of

aquatic plants; and decreased due to phytoplankton respiration, oxidation of CBOD,

nitrification and sediment oxygen demand (SOD).

CBOD is a measure of oxygen demand consumed by bacteria in the oxidation of organic

(carbonaceous) matters present in water. The major source of CBOD is the result of

phytoplankton death. The loss of CBOD usually results from the settling of the particulate

CBOD, oxidation of the carbonaceous materials and the denitrification reaction under low

DO conditions.

The kinetic source terms for DO and CBOD can be calculated using formulas presented

by Wool et al. [1] and Chapra [8].

Processes in Bed Sediment Layer

For simplicity, the bed sediment layer is represented using one layer in the model. Similar

to the water column, eight state variables including NH3, NO3, PO4, PHYTO, CBOD, DO,

ON, and OP are involved in the bed sediment layer.

The decomposition of phytoplankton releases organic nitrogen and organic phosphorus

based on the ratios of nitrogen/carbon and phosphorus/carbon, respectively. The ON and OP

are converted to NH3 and PO4 at temperature-dependent decomposition rates. NH3 and PO4

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 69

can also be generated by the anaerobic decomposition of phytoplankton. In addition NO3 can

be converted to nitrogen gas at a temperature-dependent rate due to denitrification.

The anaerobic decomposition of phytoplankton releases CBOD based on the ratio of

oxygen/carbon. Under the anoxic condition, the denitrification reaction provides a sink for

CBOD. The decomposition reaction of benthic organic carbon causes reductions of CBOD

and DO.

The water quality constituents in this layer may be exchanged with overlying water

column due to the processes of diffusion, resuspension and settling.

SEDIMENT-ASSOCIATED WATER QUALITY PROCESSES

The field observations conducted by Stefan et al.[9] show that suspended sediment

reduces light penetration in water column. Therefore it reduces the growth of phytoplankton

by limiting the amount of solar energy available for photosynthesis. The processes of

adsorption-desorption of nutrients by suspended sediment are also important for the fate and

transport of nutrients. In addition, nutrients may exchange between bed sediment layer and

water column due to diffusion. To study these complex processes, the effects of sediment on

the phytoplankton and nutrients were considered, and three major sediment-associated water

quality processes, including the effect of sediment on the growth of phytoplankton, the

adsorption-desorption of nutrients by sediment and the release of nutrients from bed sediment

layer, were taken into account.

Effect of Sediment on the Growth of Phytoplankton

The light attenuation coefficient K

e in Eq.(5) is determined by the effects of water,

chlorophyll and SS, and can be expressed by the formula proposed by Stefan et al. [9]:

sschle KKKK ++= 0 (10)

where K0 is the light attenuation by pure water (m-1); Kchl is the attenuation by chlorophyll (m-

1); Kss is the attenuation by SS (m-1). Wool et al. [1] proposed a formula to calculate Ke based

on the light attenuation due to pure water and chlorophyll without considering the effect

of SS:

67.0

0054.00088.0 chlchle CCKK ++= (11)

in which Cchl is the concentration of chlorophyll (mg/l).

In fact, SS increases both water surface reflectivity and light attenuation. Field

measurements by Stefan et al. [9] showed that the attenuation by SS can be given by:

sK ss

= (12)

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

where s is the concentration of suspended sediment;

is the attenuation parameter due to

suspended sediment. Combining Eqs. (11) and (12), Eq. (10) can be written as:

sCCKK chlchle

+++= 67.0

0054.00088.0 (13)

The parameter K0 and

can be obtained based on field measurements.

Processes of Adsorption-Desorption of Nutrients by Sediment

Mathematical Descriptions

Adsorption and desorption are important processes between dissolved nutrients and

suspended sediment in the water column. Adsorption is a process in which dissolved nutrients

become associated with suspended particles. Desorption is the reverse process and thus refers

to the release of adsorbed nutrients from suspended sediment.

Therefore, nutrients can be directly transported in the flow, they can also be transferred

from the dissolved phase to the particulate phase associated with sediment and then

transported with sediment in the flow.

In water quality processes, the reaction rates for adsorption-desorption are much faster

than that for the biological kinetics, so an equilibrium assumption can be made [1,2]. It is

assumed that the interaction of the dissolved and particulate phases reach equilibrium

instantaneously in response to nutrient and sediment inputs so as to redistribute nutrients

between dissolved and solid-phase compartments.

Thus, the processes of adsorption-desorption are assumed to reach equilibrium at each

time step in the numerical simulation.

In some models, the adsorption-desorption of nutrients by sediment is described by a

linear isotherm, and the ratio of particulate and dissolved nutrient concentration is assumed as

a constant [1, 10]. However, most experimental results show that the Langmuir equilibrium

isotherm is a better representation of the relations between the dissolved and particulate

nutrient concentrations [11,12,13].

It is assumed that the volume of the nutrient/water/sediment mixture solution is V0, which

is a constant before and after adsorption. C0 is the total initial nutrient concentration at each

time step, and s is the sediment concentration. The concentrations of dissolved and particulate

nutrients are defined as:

d= (14)

p= (15)

where Cd and Cp are the concentrations of dissolved and particulate nutrients when the

adsorption reaches equilibrium; Md and Mp are the masses of dissolved and particulate

nutrients, respectively.

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 71

The total initial amount of nutrients in the solution is same as when the adsorption

reaches equilibrium, so the following equation can be obtained:

0000 VCVCVC pd += (16)

pd CCC +=

0 (17)

The equilibrium adsorption content (Q) is defined as:

Qpp

p===

(18)

where M

s is the mass of sediment in water column. In the proposed model, the Langmuir

equation was adopted to calculate the adsorption and desorption rate [8,13]. The equilibrium

adsorption content (Q) can be expressed as:

KCQ

Q+

=1 (19)

where Qm is the maximum adsorption capacity; and K is the ratio of adsorption and

desorption rate coefficients. Based on Eqs. (17) and (18), the following equations can be

obtained:

sQC p= (20)

sQCCCC pd −=−= 00 (21)

By substituting Eqs.(18) and (21) into Eq.(19) and simplifying, a quadratic equation is

obtained

2=+

⎟

⎠

⎞

⎜

⎝

⎛++− mpmp sQCCsQ

CC (22)

By using the quadratic formula, Cp can be solved as:

⎥

⎦

⎤

⎢

⎣

⎡−

⎟

⎠

⎞

⎜

⎝

⎛++±

⎟

⎠

⎞

⎜

⎝

⎛++= mmmp sQCsQ

CsQ

CC 0

00 4

1 (23)

In order to determine the “+” or “–” sign to be used in front of the radical in Eq (23), this

equation was rewritten as:

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

⎥

⎦

⎤

⎢

⎣

⎡+

⎟

⎠

⎞

⎜

⎝

⎛−+±

⎟

⎠

⎞

⎜

⎝

⎛++= K

CsQ

CC m

mmp

00 (24)

In Eq. (24), ⎟

⎠

⎞

⎜

⎝

⎛−+>+

⎟

⎠

⎞

⎜

⎝

⎛−+ m

msQ

0. If the “+” is adopted, then

CC p>.

Based on Eq. (21), it is known that Cp has to be less than C0. So in Eq. (24), only the “–”

sign is possible, and it becomes

⎥

⎦

⎤

⎢

⎣

⎡+

⎟

⎠

⎞

⎜

⎝

⎛−+−

⎟

⎠

⎞

⎜

⎝

⎛++= K

CsQ

CC m

mmp

00 (25)

Based on Eq (21), Cd can be expressed as:

⎥

⎦

⎤

⎢

⎣

⎡+

⎟

⎠

⎞

⎜

⎝

⎛−++

⎟

⎠

⎞

⎜

⎝

⎛−−= K

CsQ

CC m

mmd

00 (26)

Eqs. (25) and (26) are used to calculate the concentration of particulate and dissolved

nutrients when the adsorption reaches equilibrium. The two equations show the

concentrations of particulate and dissolved nutrients due to adsorption-desorption are

determined by the total initial concentration of nutrients C0, the adsorption constants K and

Qm, and the suspended sediment concentration s. As C0 and s may vary with time, the ratio of

Cp and Cd is not a constant.

Comparison with Experimental Measurements

A laboratory experiment was conducted by Bubba et al. [13] to study adsorption

processes of ortho-phosphorus by sediment. Figure 2 shows the measured concentrations of

dissolved and particulate phosphate at equilibrium. The results obtained based on the

Langmuir equation (Eq.19) and linear isotherms with ratios for the particulate to dissolved

phosphate ranges from 0.01 to 0.5 suggested by some researchers [1,2] were also plotted for

comparison. It can be observed that the Langmuir equation is generally in good agreement

with the full range of the data, but the linear isotherm appears to be valid only when the

concentration of dissolved phosphate is very low, producing large errors outside this range.

Eqs. (25) and (26) were tested using this experimental measurement case. Based on the

measured data, the maximum adsorption capacity Qm and the Langmuir adsorption constant K

were 0.7 l/mg and 5.1×10-3 mg P/mg, respectively.

At equilibrium adsorption conditions, the concentrations of SS, the particulate and

dissolved phosphate under different initial concentrations of 2.5, 5, 10, 20, and 40 mg/l were

measured.

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 73

Two Stations’ data (Birkesig and Vestergard) were compared with computation results

obtained by Eqs. (25) and (26). Figures 3a and 3b show that the computed concentrations of

particulate and dissolved phosphate are generally in good agreement with measurements.

Figure 2. Relations between dissolved and particulate phosphate concnetrations obtained by Langmuir

equation and linear assumption.

Figure 3. Concnetration of phosphate versus different initial concnetration at equilibrium.

Release of Nutrients from Bed Sediment

Mathematical Descriptions

The process of the decomposition of organic material in bed sediment can release

nutrients to the sediment interstitial waters and remove oxygen from the overlying water. As a

result, the bed release is important sources of nutrients in the water column.

In many numerical models, the release rate of nutrients from bed sediment is determined

based on the concentration gradient across the water-sediment interface [1,2,4,5]. In fact, the

bed release rate is also affected by pH, temperature and dissolved oxygen concentration

[14,15]. Based on Romero [16], the bed release rate can be expressed as:

Dissolved concentration (mg/l)

Particulate concentration (mg/l)

0 50 100 150 200

Langmuir Eq.19

linear with ratio=0.01

linear with ratio=0.50

measurements

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

20 7

Tdos

diff sed c

dos pHs

KDOK pH

−⎛⎞

−

⎜⎟

++−

⎝⎠

(27)

where Sdiff is the bed release rate (mg/m2day); Sc is the diffusive flux of nutrients (mg/m2day);

Kdos (mg/l)and KpHs are the values that regulate the release of nutrient according to the

dissolved oxygen (DO) and pH in the bottom layer of the water column of depth Δzb (m);

sed

is the temperature coefficient. The diffusive flux Sc can be calculated using Fick’s first law

which expresses that the flux is directly proportional to the concentration gradient and the

porosity of sediment [17,18]:

)()( wbwb

mc CCkCC

DS −=−

≈−=

(28)

where Dm is the molecular diffusivity (m2/day);

is the porosity of sediment; Δzb is the

diffusive sub-layer thickness near the bed (m); k is the diffusive exchange coefficient at

water-sediment interface (m/day); Cw and Cb are the concentration of nutrients in water and

water-sediment interface, respectively.

Steinberger and Hondzo [19] investigated the factors affecting k and established an

empirical relation for diffusional transfer of dissolved oxygen across the bed surface. Their

studies show that k is governed by the Reynolds and Schmidt numbers. Based on their

studies, k is expressed by

⎪

⎩

⎪

⎨

⎧

⎟

⎠

⎞

⎜

⎝

⎛

⎟

⎠

⎞

⎜

⎝

⎛Δ

33.089.0

012.0

≠

(29)

where U = depth averaged velocity; and

= kinematic viscosity. This formula takes the flow

conditions into account for estimating the exchange coefficient k.

Comparison with Experimental Data

A laboratory experiment was conducted by Kim et al.[14] to study the effects of pH and

DO on the phosphorus release rate at Jamsil Submerged Dam (JSD) Station in Han River.

Eq.(27) was tested using the experiment measurements. Figures 4 and 5 show the

comparisons between the computational results by Eq. (27) and experimental measurements.

In general, the effects of pH and DO on the phosphorus release rate are reasonably

predicted by Eq. (27). The experimental and computational results show the release rate of

phosphorus decreases with the increase of DO. Under acidic conditions, the release rate

decreases as pH increases, while under basic conditions, the release rate increases as pH

increases. These trends are similar to other experimental measurements [10,20,21].

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 75

Figure 4. The effect of DO on the release rate of phosphorus at JSD Station in Han River.

(a) under acidic condition (b) under basic condition

Figure 5. The effect of pH on the release rate of phosphorus at JSD Station in Han River.

UMERICAL

ODEL

EVELOPMENT

The proposed water quality model (CCHE3D_WQ) was developed based on CCHE3D

hydrodynamic model, a three dimensional model developed at the National Center of

Hydroscience and Engineering, the University of Mississippi [7]. CCHE3D is a three-

dimensional model that can be used to simulate unsteady turbulent flows with irregular

boundaries and free surfaces. It is a finite element model utilizing a special method based on

the collocation approach called the efficient element method. This model is based on the 3D

Reynolds-averaged Navier-Stokes equations. By applying the Boussinesq Approximation, the

turbulent stress can be simulated by the turbulent viscosity and time-averaged velocity. There

are several turbulence closure schemes available within the model for different purposes,

including the parabolic eddy viscosity, mixing length, k–ε and nonlinear k–ε models. This

model has been successfully applied to analyze wind-driven flow, turbulent buoyant flow,

turbulent flow fields in scour holes and around a submerged training structure in a meander

bend [22, 23,24].

02468

Release rate (mg P/ m

day)

DO (mg/l)

measurements

computation

pH=8

T=20

02468

Release rate (mg P/ m

day)

pH value

measurement s

computati on

DO=0.5mg/l

T=20

6 8 10 12 14

Release rate (mg P/ m

day)

pH value

measurement s

computat ion

DO=0.5mg/l

T=20

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

Governing Equations

The governing equations of continuity and momentum of the three-dimensional unsteady

hydrodynamic model can be written as follows:

∂

u (30)

iji

jij

ifuu

⎟

⎠

⎞

⎜

⎝

⎛−

∂

−=

∂

∂''

(31)

where ui (i=1,2,3) = Reynolds-averaged flow velocities (u, v, w) in Cartesian coordinate

system (x, y, z); t = time;

= water density; p = pressure;

= fluid kinematic viscosity; ''

jiuu−

=Reynolds stress; and fi = body force terms.

The free surface elevation (

s) is computed using the following equation:

0=−

∂

ηηη

(32)

where us, vs and ws = surface velocities in x, y and z directions;

s = water surface elevation.

In water column, each one of the water quality constituents can be expressed by the

following mass transport equation:

∑

+++=+++ i

iiii S

C)()()(

)()()(

∂

(33)

in which u, v, w are the water velocity components in x, y and z directions, respectively; C i is

the concentration of the ith water quality constituent; Dx, Dy and Dz are the diffusion

coefficients in x, y and z directions, respectively; i

SΣis the effective source term for the ith

water quality constituent.

In bed sediment layer, the decomposition of organic material can have profound effects

on the concentrations of oxygen and nutrients in the overlying waters. In this layer, the pore

water may infiltrate in and out and thus induce additional nutrients transfer.

To simulate the water quality processes in this layer, the exchange between the sediment-

water interface and the kinetic transformation of water quality constituent are considered. To

simplify the model, the bed sediment layer is represented using one layer. In general, the

depth of this layer is about 0.1 to 0.3 m. In the current version, only diffusive flux of water

quality constituents is considered in this layer:

∑

∂

(34)

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 77

where Cj =concentration of water quality constituents in sediment layer; i

SΣ is the effective

source term, which includes biochemical kinetic transformation rate, bed release rate, and the

settling and re-suspension terms.

Wind_Induced Eddy Viscosity

In a natural lake, wind stress is the important driving force for lake flow circulation. For

simulating the wind driven flow, the distribution of vertical eddy viscosity is required. A

parabolic eddy viscosity distribution was proposed by Tsanis [25] based on the assumption of

a double logarithmic velocity profile. The vertical eddy viscosity was expressed as:

))((

*zHzzz

t−++=

(35)

in which λ= numerical parameter; zb and zs = characteristic lengths determined at bottom and

surface, respectively; u*s = surface shear velocity; H = water depth. To use this formula, three

parameters, λ, zb and zs have to be determined. For some real cases with very small water

depths, using this formula to calculate eddy viscosity may cause some problems.

Koutitas and O’Connor [26] proposed two formulas to calculate the eddy viscosity based

on a one-equation turbulence model. Their formulas were:

)2(

max, zztt

ηηνν

−= ( 5.00 ≤≤ z

) (36)

)15)(1(

max, −−= zztt

ηηνν

( 15.0 ≤< z

) (37)

Where

HuHu sst *

25.0

*max, 142.03.0105.0

λλν

== − (38)

in which,

= numerical parameter;

z = non-dimensional elevation, Hz

z/=

In this model, a new formula was proposed based on experimental measurements

conducted in a laboratory flume with steady-state wind driven flow reported by Koutitas and

O’Connor [26]. The form of eddy viscosity was similar to Koutitas and O’Connor’s

assumption (Eq. 36 and 37) and expressed as:

)(

max, ztt f

ηνν

= (39)

Based on measured data, a formula is obtained to express the vertical eddy viscosity:

)62.078.224.3( 2

max, ++−= zzztt

ηηηνν

(40)

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

A figure was plotted to compare the vertical distributions of eddy viscosity obtained from

formulas provided by Tsanis, Koutitas, and the proposed model (Figure 7, next Section).

Boundary Conditions

Wind stress is generally the dominant driving force for flow currents in natural lakes. The

wind shear stresses at the free surface are expressed by

windwindwinddawx VUUC +=

ρτ

(41)

windwindwinddawy VUVC +=

ρτ

(42)

where a

= air density; Uwind and Vwind = wind velocity components at 10 m elevation in x and

y directions, respectively. Although the drag coefficient Cd may vary with wind speed

[26,27], for simplicity, many researchers assumed the drag coefficient was a constant on the

order of 10-3 [28, 29]. In this model, Cd was set to 3

100.1 −

×, and this value is applicable for

simulating the wind driven flow in Deep Hollow Lake [30].

On the free surface, the normal gradient of mass concentration is set to be equal to zero.

In the vicinity of the solid walls and bed, the gradients of flow properties are steep due to wall

effects. The normal gradient of concentration on the wall is set to be zero.

At the inlet boundary, flow discharge and mass concentration are required; at the outlet,

the water surface elevation is set as boundary condition.

Numerical Solution

The numerical model was developed based on the finite element method. Each element is

a hexahedral with three levels of nine-node quadrilaterals, and the governing equations are

discretized using a 27-node hexahedral (Figure 6).

Staggered grid is adopted in the model. The grid system in the horizontal plane is a

structured conformal mesh generated on the boundary of the computational domain. In

vertical direction, either uniform or non-uniform mesh lines are employed. In order to get

more accurate results, the mesh lines are placed with finer resolution near the wall, bed and

free surface.

The unsteady equations are solved by using the time marching scheme. A second-order

upwinding scheme is adopted to eliminate oscillations due to advection.

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 79

Figure 6. Coordinate system and computational element. (In Figure6, the x, y ,and z are axes of the

Cartesian coordinate system and ξ , η and ζ are axes of a local system).

In this model, a convective interpolation function is used for this purpose. This function

is obtained by solving a linear and steady convection-diffusion equation analytically over a

one-dimensional local element shown in Figure 6.

Although there are several other upwinding schemes, such as the first order upwinding,

the second order upwind and Quick scheme, the convective interpolation function is selected

in this model due to its simplicity for the implicit time marching scheme [7].

The velocity correction method is applied to solve the dynamic pressure and enforce

mass conservation. Provisional velocities is solved first without the pressure term, and the

final solution of the velocity is obtained by correcting the provisional velocities with the

pressure solution [7, 22]. The system of the algebraic equations is solved using the Strongly

Implicit Procedure (SIP) method. Flow fields, including water elevation, horizontal and

vertical velocity components, and eddy viscosity parameters were computed by CCHE3D and

set as input data. After getting the effective source terms i

SΣ, the concentration distribution of

water quality constituents can be obtained by solving mass transport equations (33)

numerically. The effective source term i

SΣincludes the kinetic transformation rate, external

loads and sinks for water quality constituents. The kinetic transformation rate can be obtained

by analyzing the complex processes of water quality constituents. Normally, the external

loads and sinks come from the boundary, including inlet, outlet, benthic, water surface and

atmospheric. In sediment layer, the concentration distribution of water quality constituents

can be obtained by solving equations (34) numerically after obtaining the effective source

terms j

SΣ.

MODEL VALIDATION AND VERIFICATION

Model Validation for Wind-Driven Flow

A laboratory test case was carried out in a wind-wave flume by Koutitas and O’Connor

[26] to study the wind-driven flow.

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

The length and width of the measureing flume were 5 m and 0.2m, and the water depth

was 0.31m. The vertical current profile near the middle section was measured. The detailed

information of the experiment was described by Koutitas and O’Connor [26]. Figure 7 shows

the vertical distributions of eddy viscosity obtained from experimental measurements and

formulas provided by Tsanis (Eq.35), Koutitas (Eq. 36 and 37), and the proposed formula

(Eq.40). This proposed formula can be used to calculate the eddy viscosity over the full range

of water depth. At water surface ( 1/

Hz ), Eqs. (35) and (37) show the eddy viscosity is

zero or very closed to zero.

Figure 7. Comparison of vertical eddy viscosity formulas and experimental data.

The newly proposed formula shows the eddy viscosity at the water surface is about 16%

of the maximum eddy viscosity max,t

The developed numerical model was applied to simulate the velocity profile of the wind

driven flow for the experimental case. A non-uniform grid with 21 vertical nodes with fine

resolution near surface and bed were used for model validation. This non-uniform grid was

generated using a flexible and powerful two-direction stretching function EDS [31].

The parameters in the stretching function were set to E=1, D=0.5, and S=5.2. In the

numerical modeling, both Eq. (35), and (40) were used for calculating the eddy viscosity, and

the simulated vertical velocity profiles were compared with experimental measurements

(Figure 8). Numerical results are generally in good agreement with experimental

measurements. Near the water surface, the model overestimated the velocity using Tsanis’s

formula (Eq. 35) to calculate the eddy viscosity; while it underestimated the velocity by using

the proposed formula (Eq. 40). Near the bottom, the simulation obtained by using the

proposed formula gave better predictions.

νt/(u*sH)

z/H

00.05 0.1 0.15

0.2

0.4

0.6

0.8

measurements

this study

Koutitas (1980)

Tsanis (1989)

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 81

Figure 8. Normalized vertical velocity profile (us= surface velocity).

Model Verification for the Mass Transport Simulation

The proposed water quality model was tested against an analytical solution for predicting

the concentrations of a non-conservative substance in a hypothetical one-dimensional river

flow with constant depth and velocity.

A continuous source of a non-conservative substance was placed at the upstream end of a

straight channel for a finite period of time, τ (Figure 9). Under the unsteady condition, the

concentration of the substance throughout the river can be expressed as:

ss CK

C−

∂

(43)

where U = velocity; Cs = concentration of substance; Dx = mixing coefficient; and Kd = decay

rate. An analytical solution given by Chapra (1997) is:

⎥

⎦

⎤

⎢

⎣

⎡

⎟

⎠

⎞

⎜

⎝

⎛Γ+

⎟

⎠

⎞

⎜

⎝

⎛Γ++

⎟

⎠

⎞

⎜

⎝

⎛Γ−

⎟

⎠

⎞

⎜

⎝

⎛Γ−= tD

Utx

erfc

Utx

erfc

UxC

txC

)1(

exp

)1(

exp

),( 0 (t<τ) (44)

⎥

⎦

⎤

⎢

⎣

⎡

⎟

⎠

⎞

⎜

⎝

⎛

−

Γ−−

−

⎟

⎠

⎞

⎜

⎝

⎛Γ−

⎪

⎩

⎪

⎨

⎧

⎟

⎠

⎞

⎜

⎝

⎛Γ−= )(2

)(

)1(

exp

),( 0

tUx

erfc

Utx

erfc

UxC

txC

⎪

⎭

⎪

⎬

⎫

⎥

⎦

⎤

⎢

⎣

⎡

⎟

⎠

⎞

⎜

⎝

⎛

−

Γ−+

−

⎟

⎠

⎞

⎜

⎝

⎛Γ+

⎟

⎠

⎞

⎜

⎝

⎛Γ++ )(2

)(

)1(

exp

tUx

erfc

Utx

erfc

(t > τ) (45)

u/us

z/H

-0.5 00.5 11.5

0.2

0.4

0.6

0.8

measurements

this study

Tsanis

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

where f41 +=Γ , and 2

fxd

=. For the river conditions shown in Figure 9, with a depth

of 10 m, U = 0.03m/s, Dx = 30 m2/s,

= 6 hr, and the values of Kd = 0, 1.0/day and 2.0/day,

respectively. Figure 10 shows the time series of concentration at the section x = 2000 m

obtained by the numerical model and analytical solution. The maximum error is less than 2%.

Figure 9. Test river for verification case.

Figure 10. The time series of concentration at the section x = 2000m obtained from the numerical

model and analytical solution.

MODEL APPLICATION TO DEEP HOLLOW LAKE

Study Area

Mississippi Delta is one of the most intensively farmed agricultural areas of the United

States. The soils in this region are highly erodible, resulting in a large amount of sediment

discharged into the water bodies.

Sediments are normally associated with many pollutants and greatly affect water quality

and aquatic lives. The Mississippi Delta Management System Evaluation Area (MDMSEA)

project is part of a national program designed to evaluate the impact of agricultural

production on water quality and to develop best management practices (BMPs) to minimize

adverse effects. Deep Hollow Lake, a small oxbow lake located in Leflore County,

Mississippi, was selected as a study site for MDMSEA.

Time (hour)

Concentration of mass(mg/l)

0 12 24 36 48 60 72 84 96 108 120 132 144

analytical solution (Kd=0)

analytical solution (Kd=1/day)

analytical solution (Kd=2/day)

simulation(Kd=0)

simulation(Kd=1/day)

simulation(Kd=2/day)

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 83

Figure 11 shows the study area of Deep Hollow Lake. It has a morphology typical of an

oxbow lake, with a length of about 1 km and a width of about 100 m. Lake water depth

ranges from 0.5 m to 2.6 m, with greatest depth in the middle. The lake receives runoff from a

two square kilometer watershed that is heavily cultivated.

Figure 11. The study area of Deep Hollow Lake.

Weekly or biweekly samples of suspended sediment, nutrients, chlorophyll, bacteria, and

other selected water quality variables were collected at Stations DH1, DH2 and DH3. Two of

the major inflows, located at the Stations UL1 and UL2, were monitored for flow and water

quality by the U.S. Geological Survey. The nutrient concentrations in Deep Hollow Lake are

mainly dependent on the fertilizer loadings in the surrounding farmland and the quantity of

runoff. Field measurements show that the concentrations of nitrate and ammonia in the lake

are very low, while the concentration of phosphorus is relatively high in comparison with

other areas of the United States. Suspended sediment concentrations are relatively high,

exceeding published levels known to adversely impact fish growth and health [32].

The water quality of the lake is sensitive to suspended sediment concentrations because

photosynthetic activity is limited by elevated turbidity levels following runoff events.

Researches in lakes similar to Deep Hollow have shown that the nutrients can be released

from suspended and bed sediments [15]. So the sediment-associated water quality processes

need to be considered in the water quality modeling.

Based on bathymetric data, the computational domain was discretized into a structured

finite element mesh using the CCHE Mesh Generator[33].

In the horizontal plane, the irregular computational domain was represented by a mesh

with 95×20 nodes. In the vertical direction, the domain was divided into 8 layers with finer

spacing near the bed. The bed layer was represented by a 0.1 m single layer.

Light Attenuation Coefficient in Deep Hollow Lake

Light intensity is one of important factors for the growth of phytoplankton. In the

numerical model, the effect of light intensity on the growth of phytoplankton is determined by

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

the light limitation factor, which can be calculated based on the light attenuation coefficient,

light intensity and water depth (Eq. 5).

As shown in Eq.(10), the light attenuation coefficient K

e is affected by water,

concentrations of chlorophyll and SS. It has been observed that the light intensity in the water

column decreases with the water depth, and the attenuation of light is proportional to the light

intensity in the water:

−= (46)

in which I is the light intensity in the water column; zd is the distance to water surface.

On the water surface, zd = 0 and 0

II =, so

)(

eII −

= (47)

Eq.(47) can also be written as:

de zK

I−=

⎟

⎠

⎞

⎜

⎝

⎛

ln (48)

So Ke can be obtained by linear fitting based on the measured I, I0 and zd.

A field measurement was conducted to measure the light intensity on the water surface

and in the water column, and concentrations of suspended sediment and chlorophyll in Deep

Hollow Lake. Based on the measured data, the light attenuation coefficient Ke, and the

parameters K0 and

in Eq. (13) can be obtained. Light intensities were measured using a

LICOR LI-250 light meter and a spherical quantum radiation sensor which measures photon

flux from all directions underwater. This measurement is called Photosynthetic Photon Flux

Fluence Rate (PPFFR). Units are in micromole per second per square meter. Data was

collected at Deep Hollow Lake approximately every two weeks from January to March, 2004

(weather permitted). Light radiation was measured during conditions of unobstructed sunlight

(no clouds) at three sites: DH1, DH2, and DH3 Stations (shown in Figure 11) between the

hours of 10:00 AM and 2:00 PM. At each station, light intensities were measured in air, at

water surface and approximately every 10 cm in the water body until the measurement unit

touched the lake bottom. In addition, the concentrations of SS and chlorophyll at the same

sites were also measured. Total 20 sets of data were obtained.

Figure 12 shows the light intensity of three stations (DH1, DH2 and DH3) on water

surface and in the water column measured on January 20, 2004. It can be observed that light

intensity decreases quickly with the increase of depth in the water column due to the effects

of water, sediment and phytoplankton. The light can penetrate about 1/3 of the total water

depth. Figures 13a, 13b and 13c show the relations between )/ln( 0

II and distance to the

water surface zd at DH1, DH2 and DH3 stations on Jan 20, 2004. It is obviously there is a

linear relation between )/ln( 0

II and zd, and the slope of the line is the measured light

attenuation coefficient.

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 85

Figure 12. Measured light intensity in water at DH1, DH2 and DH3 Stations (Jan. 20, 2004).

Figure 13a. Relations between )/ln( 0

II and distance to the water surface z at DH1 Station.

Figure 13b. Relations between )/ln( 0

II and distance to the water surface z at DH2 Station.

Light intensity (umol/s/m2)

Distance to water surface (m)

0 500 1000 1500 2000

0.2

0.4

0.6

0.8

1.2

DH1

DH2

DH3

Distance to water surface (m)

ln (I/Io)

0 0.2 0.4 0.6 0.8 1 1.2

-14

-12

-10

-8

-6

-4

-2

measurements

fitted line

ln(I/Io)=-10.91zd

Distance to water surface (m)

ln (I/Io)

0 0.2 0.4 0.6 0.8 1 1.2

-14

-12

-10

-8

-6

-4

-2

measurements

fitted line

ln(I/Io)=-11.34 zd

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

Figure 13c. Relations between )/ln( 0

II and distance to the water surface z at DH3 Station.

The Eq. (13) can also be written as

67.0 )054.00088.0( KsCCK chlchle +=+−

(49)

Set )054.00088.0()( 67.0

chlchlee CCKKf +−= ,so

67.0 )054.00088.0()( KsCCKKf chlchlee +=+−=

(50)

Based on all sets of field measured data, a regression line, with the slope

and

intersection K0 equal to 0.0452 and 1.2, respectively, was obtained (Figure14). So Eq. (13)

can be expressed as

sCCK chlchle 0452.0054.00088.02.1 67.0 +++= (51)

Figure 14. )( e

Kf versus suspended sediment concentration s.

Distance to water surface (m)

ln (I/Io)

0 0.2 0.4 0.6 0.8 1 1.2

-14

-12

-10

-8

-6

-4

-2

measurements

fitted line

ln(I/Io)=-11.52 zd

f (Ke) = 0. 0452s + 1.2

0 50 100 150 200 250 300

f(ke)

Sed iment co ncentratio n (mg/l)

measurements

fitted line

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 87

Based on the field measurements in Deep Hollow Lake, the value of

was equal to

0.0452. This value was close to the value of 0.043 obtained by Stefan et al. [9] for a nearby

lake. Eq. (51) was adopted to calculate the light attenuation coefficient in Deep Hollow Lake.

Figure 15 compares the light attenuation coefficient obtained from field measurements and

calculations from Eq. (51). Good agreement was obtained with an r2 value of 0.86.

Figure 15.Comparison of the light attenuation coefficient obtained from field measurements and Eq.

(51).

Model Application

In Deep Hollow Lake, wind stress is the most important driving force for lake flow

circulations. The CCHE3D hydrodynamic model was first calibrated using field

measurements obtained from Deep Hollow Lake using an Acoustic Doppler Current Profiler

(ADCP), and then it was applied to simulate the flow fields during the selected simulation

period.

Figure16 shows the comparison of simulated flow velocities with field measurements

conducted on November 12, 2003. In general, numerical model provides good predictions for

flow fields.

The proposed water quality model CCHE3D_WQ was calibrated using weekly/biweekly

field measured data obtained in the lake between April to June, 1999. In this period, the flow

fields including water surface elevation, velocity, eddy viscosity, etc., were obtained from the

CCHE3D hydrodynamic model. After obtaining the flow fields, the concentrations of water

quality constituents can be simulated using the proposed CCHE3D_WQ model.

Figure 17 shows the measured suspended sediment at DH1 Station in 1999. The SS

concentration is relatively high in spring, fall, and winter, so the sediment-associated water

quality processes play important roles in the lake aquatic ecosystem. The above mentioned

three major sediment-associated processes were integrated into the CCHE3D_WQ model to

simulate the water quality constituents in the lake.

0 5 10 15 20

measured Ke

calculated Ke

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

Figure 16. Observed and simulated velocities at Station DH1 (Nov. 12, 2003).

The light attenuation coefficient Ke was calculated using Eq.(51) by considering the

effects of water, chlorophyll and suspended sediment. The adsorption and desorption of

nutrients by sediment were described using Langmuir equation, and the concentrations of

dissolved and particulate nutrient at equilibrium were calculated using Eqs. (25) and (26). In

Deep Hollow Lake, the concentrations of ammonium and nitrate were very low, so the

adsorption and desorption of ammonium and nitrate from sediment were expected to be

insignificant and were not incorporated.

Figure 17. Measured SS concentration at Station DH1.

Dissolved nutrients may be released to the water column from bed sediments. In the

numerical simulation, the bed release rate of ammonium, nitrate, organic nitrogen, phosphate,

and organic phosphorus were calculated using Eq. (27). For calibration runs, parameters in

the water quality model were adjusted repeatedly to obtain a reasonable reproduction of the

field data. The model parameters were obtained either directly from experiments and field

U(cm/s)

z(m)

-10 -5 0 5 10 15

-0.2

0.2

0.4

0.6

0.8

1.2

1.4

measurement

computation

V(cm/s)

z(m)

-10 -5 0 5 10 15

-0.2

0.2

0.4

0.6

0.8

1.2

1.4

measurement

computation

100

150

200

250

300

350

0 30 60 90 120 150 180 210 240 270 300 330 360

SSconce ntr ation(mg/l)

Julianday,1999

measurements

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 89

measurements, or by others [1,12,13,34,35,36]. Table 1 shows some calibrated parameters

used in the model to simulate the water quality constituents in Deep Hollow Lake. All

adopted parameter values are in the range reported in the literatures [1,2]. In the water quality

model, the measured water temperature, surface light intensity, wind field, SS concentration,

and pH were set as input data. The measured depth averaged concentrations of nutrients and

chlorophyll were used for model calibration.

Figure 18 shows the comparison of the simulated chlorophyll concentrations with field

measurements. In this figure, the simulated results at different layers were plotted for

comparison. Figure 19 and 20 show the model calibration results of nitrogen and phosphorus

at Station DH1. In general, the model provided reasonable reproduction of patterns and

acceptable magnitudes for water quality constituents. The mean values of the model results

are generally in good agreement with the field observations. However, the ability of the

model to reproduce temporal variations in field measurements was not as good as the mean

values. Those differences may arise due to the fact that measurements occurred weekly while

the time step for the simulation was 1 hour.

Table 1. Calibrated values of parameters for the water quality model applied to Deep

Hollow Lake

Parameter definition Symbol Value Units

Maximum phytoplankton growth rate Pmx 2.0 day-1

Background light attenuation coefficient K0 1.2 m-1

Light attenuation parameter related to SS

0.0452 l/mg/m

Saturation light intensity of phytoplankton Im 300 ly/day

Half-saturation constant for nitrogen in

phytoplankton growth

KmN 0.01 mg/l

Half-saturation constant for phosphorus in

phytoplankton growth

KmP 0.001 mg/l

Effect coefficient of temperature below optimal

temperature on growth

KTg1 0.006 none

Effect coefficient of temperature above optimal

temperature on growth

KTg2 0.008 none

Phytoplankton endogenous respiration rate kpr 0.125 day-1

Phytoplankton mortality rate kpd 0.02 day-1

Temperature correction coefficient

pr 1.068

Settling velocity of phytoplankton

s4 0.01 m/day

Diffusive exchange coefficient of nitrogen at water-

sediment interface

kn 0.01 m/day

Control value of nitrogen release via DO Kndos 0.5 mg/l

Ratio of adsorption and desorption rate coefficients

for phosphorus

K 0.7 l/mg

Maximum adsorption capacity of phosphorus Qm 0.0051 mg P/mgSS

Diffusive exchange coefficient of phosphorus at

water-sediment interface

kp 0.01 mg/m2day

Temperature coefficient

sed 1.05 none

Control value of phosphorus release via DO Kpdos 0.5 mg/l

Control value of phosphorus release via pH KpH 18 none

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

Figure 18. The concentration of chlorophyll at Station DH1 for calibration.

a. Ammonia b. Nitrite

Figure 19. The concentration of nitrogen at Station DH1for calibration.

a. Phosphate b. Organic phosphorus

Figure 20. The concentration of phosphorus at Station DH1for calibration.

In addition, the water quality processes in the lake system are likely more complicated

than those processes considered in the numerical model.

The period from September to December 1999 was chosen for model validation. After

obtaining the flow fields from CCHE3D hydrodynamic model, the measured boundary

conditions, weather data and SS concentrations, CCHE3D_WQ model was used to simulate

the concentrations of water quality constituents. Parameter values in the water quality model

were same as those calibrated values. Figures 21 and 22 show the simulated and measured

Julian day, 1999

Concentration of chlorophyll(mg/l)

100 120 140 160 180 200

0.1

0.2

0.3

surface

layer3

layer5

bottom

measurement

Julian day, 1999

Concentration of NH3(mg/l)

100 120 140 160 180 200

0.1

0.2

0.3

simulation

measurements

Julian day, 1999

Concentration of NO3(mg/l)

100 120 140 160 180 200

0.2

0.4

0.6

simulation

measurements

Julian day, 1999

Concentration of PO4(mg/l)

100 120 140 160 180 200

0.1

0.2

0.3

simulation

measurements

Julian day, 1999

Concentration of OP(mg/l)

100 120 140 160 180 200

0.2

0.4

0.6

simulation

measurements

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 91

concentrations of chlorophyll and nitrogen. For the validation run, trends and quantities of

concentration of chlorophyll and nitrogen obtained from the numerical model are generally in

agreement with the observations. Figure 23 shows the simulated and observed concentrations

of ortho-phosphorus and total organic phosphorus, respectively. Without considering the

sediment-associated processes of adsorption, desorption and bed release, the model

overestimated ortho-phosphorus concentration and underestimated organic phosphorus. After

considering those processes, the root mean square error (RMSE) of ortho-phosphorus

concentration was reduced from 0.029 to 0.016 mg/l, or reduced 45%; for organic phosphorus

RMSE was reduced from 0.051 to 0.037 mg/l, or reduced 28%.

Figure 21. The concentration of chlorophyll at Station DH1 for validation.

a. Ammonia b. Nitrite

Figure 22. The concentration of nitrogen at Station DH1for validation.

a. Phosphate b. Organic phosphorus

Figure 23. The concentration of phosphorus at Station DH1for calibration.

Julian day, 1999

Concentration of chlorophyll(mg/l)

240 260 280 300 320 340 360

0.1

0.2

0.3

surface

layer3

layer5

bottom

measurement

Julian day, 1999

Concentration of NH3(mg/l)

240 260 280 300 320 340 360

0.2

0.4

0.6

simulation

measurements

Julian day, 1999

Concentration of NO3(mg/l)

240 260 280 300 320 340 360

0.2

0.4

0.6

simulation

measurements

Julian day, 1999

Concentration of PO4 (mg/l)

240 260 280 300 320 340 360

0.05

0.1

0.15

0.2

with sediment-associated processes

without sediment-associated processes

measurements

Julian day, 1999

Concentration of OP (mg/l)

240 260 280 300 320 340 360

0.1

0.2

0.3

0.4

0.5

with sediment-associated processes

without sediment-associated processes

measurements

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

CCHE3D_WQ was also applied to simulate the concentration of phosphorus in bottom

sediment layer.

In this layer, the concentration is affected by the processes of biochemical kinetic, bed

release, settling and re-suspension (Eq. 34).

Due to lack of field data, a hypothetical case was assumed with the initial concentrations

of phosphate and organic phosphorus in sediment layer of 0.04 mg/l and 0.2 mg/l,

respectively. Figure 24 shows the simulated concentrations of ortho-phosphorus and total

organic phosphorus.

a. Phosphate b. Organic phosphorus

Figure 24. The concentration of phosphorus in bed sediment layer and lake water.

DISCUSSION

Comparison of Langmuir Equation and Linear Approach for Modeling the

Adsorption-Desorption

In many water quality models, both Langmuir equation and linear approach have been

applied to simulate the processes of adsorption-desorption [1, 2, 3, 37, 38]. For linear

approach, the adsorption-desorption of nutrients by sediment is described by a linear

isotherm, and the equilibrium adsorption concentration Q (mg /mg SS) can be expressed as

dp CKQ = (52)

in which Kp is the partition coefficient. Based on Eqs.(17), (20) and (21), Cp and Cd can be

calculated by

= (53)

= (54)

Julian day, 1999

Concentration of OP(mg/l)

240 260 280 300 320 340 360

0.2

0.4

0.6

sediment layer

lake water

Julian day, 1999

Concentration of PO4(mg/l)

240 260 280 300 320 340 360

0.1

0.2

0.3

sediment layer

lake water

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 93

Eqs.(53) and (54) were adopted by some water quality models for calculating the

particulate and dissolved nutrient concentrations at each time step due to adsorption-

desorption. In fact, this linear approach is satisfactory only for the low nutrient concentration

cases [8]. If the nutrient concentration is very low, the term )1( d

KC+in Langmuir equation

(Eq.19) approximately equals to 1, so Eq. (19) can be simplified as:

dm KCQQ = (55)

if the partition coefficient Kp is set as

KQK mp = (56)

then Eq.(55) is same as the linear approach equation (52). So Eq.(52) is the linear portion of

the Langmuir equation.

In order to find the value of “low nutrient concentration” for adopting the linear

approach, Eqs (53), (54) and (25), (26) were used to calculate the particulate and dissolved

concentrations of nutrients in Deep Hollow Lake due to adsorption-desorption. For this case

Langmuir adsorption constant K and the maximum adsorption capacity Qm were taken as 0.7

l/mg and 0.0051 mg /mg SS, respectively. If the linear approach was adopted, based on

Eq.(56), the partition coefficient Kp is 0.00357 l/mg.

Figures 25a and 25b show the concentrations of particulate and dissolved phosphate

calculated based on the linear approach (Eqs. 53 and 54) and Langmuir equation ( Eqs. 25 and

26) under different total initial concentrations.

a. Particulate concentration b. Dissolved concentration

Figure 25. Comparison of particulate and dissolved phosphate concentrations obtained based on the

Linear approach and Langmuir equation.

It can be observed that the linear approach and Langmuir equation have same results if

the initial concentration is less than 0.1 mg/l. In Deep Hollow Lake, the phosphate

concentration can be greater than 0.3 mg/l, so Langmuir equation should give more accurate

predictions.

Initial concention (mg/l)

Particulate concention (mg/l)

0 0.1 0.2 0.3 0.4 0.5

0.02

0.04

0.06

0.08

0.1

Eq.25

Linear Approach

Initial concention (mg/l)

Dissolved concention (mg/l)

0 0.1 0.2 0.3 0.4 0.5

0.1

0.2

0.3

0.4

Eq. 26

Linear Approach

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

Sensitivity of Chlorophyll Concentration to SS

Field data show that major water quality problems in Deep Hollow Lake in the late 1990s

were caused by excessive sediment loads carried by runoff from surrounding cultivated lands.

In order to improve the water quality of the lake, various best management practices (BMPs),

such as edge-of field BMPs and agronomic BMPs were employed to reduce sediment loads

on the farm lands. After the reduction of lake sediment concentration, there were significant

increases in Secchi depth and chlorophyll concentration, and fish populations responded

positively [32].

To show the sensitivity of chlorophyll concentration to SS, a series of hypothetical lake

SS concentrations were input to the model, while the flow conditions and nutrient loadings at

inlet boundaries were kept the same. The interactions between SS and nutrients in the lake

were considered in the model simulation. The calibration and validation periods were selected

for sensitivity study.

It was assumed the SS levels were varied from 10% to 300% of the current SS condition

(base condition), and the model was applied to simulate the response of chlorophyll

concentration in the lake. As expected, the concentration of chlorophyll is inversely related to

the sediment concentration. Figure 26 shows the sensitivity of temporal mean chlorophyll

concentration to SS for the entire simulation period at DH1 Station. In this figure, the square

represents the base condition (SS=100%, Chlorophyll=100%). When lake SS was reduced by

50%, simulated mean chlorophyll concentration increased about 40%. When lake SS was

doubled, the chlorophyll concentration fell to 37% of the base condition. Field observations

of SS and chlorophyll in the years 1999, 2000 and 2001 were used to evaluate the model

sensitivity analysis. Measured mean concentrations of SS and chlorophyll under the base

condition (Year 1999) were 82 mg/l and 40 mg/l, respectively.

Mean concentrations of SS in 2000 and 2001 were reduced to 56% and 62% of base

condition, and the chlorophyll concentrations increased to 142% and 137% of the base

condition, respectively (Figure 26). This tendency agrees with the numerical predictions

shown in Figure 26.

Figure 26. Sensitivity of temporal mean chlorophyll concentration to temporal mean SS.

20%

40%

60%

80%

100%

120%

140%

160%

180%

200%

50%

100%

150%

200%

250%

300%

Percentage of current chlorophyll

concentration

Percenta

e of current SS concentration

model prediction

ob servatio n-2000

ob servatio n-2001

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 95

Sensitivity of Chlorophyll Concentration to Nutrient Loadings

To examine the effects of changing nutrient loadings on chlorophyll concentration,

scenarios were generated by reducing and increasing the observed concentrations of dissolved

inorganic nitrogen and phosphorus. The calibration and validation periods were selected for

sensitivity study. It was assumed the nutrient loadings were varied from 10% to 300% of the

current nutrient conditions (base conditions), and the model was applied to simulate the

chlorophyll concentration in the lake. Reducing nutrient loads produced lower chlorophyll

concentrations, as expected. Figure 27a and Figure 27b show the sensitivities of temporal

mean concentration of chlorophylls under the base conditions and reduction/increasing loads

for inorganic nitrogen and phosphorus, respectively. In these figures, the squares represent the

base condition (nutrient=100%, Chlorophyll=100%). The temporal mean chlorophyll

concentration under base nutrient load was about 0.059 mg/l.

According to model simulations, reducing the lake inorganic nitrogen concentration by

50% would reduce the average chlorophyll concentration by approximately 0.0145mg/l, or

about 24%. When the concentration of inorganic nitrogen was doubled, the chlorophyll

concentration increased about 16.5%.

For inorganic phosphorus, reducing the concentration by 50% in the lake would reduce

average chlorophyll concentration by approximately 2%. When the concentration of inorganic

phosphorus was doubled, the average concentration of chlorophyll increased less than 0.5%.

This analysis indicates that the concentration of chlorophyll is much more sensitive to the

inorganic nitrogen than that to inorganic phosphorus in Deep Hollow Lake.

a. Inorganic nitrogen b. Inorganic phosphorus

Figure 27. The effect of reduction/increasing of nutrient loadings on the chlorophyll concentration.

CONCLUSION

A three-dimensional numerical model (CCHE3D_WQ) was developed to simulate the

concentration of water quality constituents in shallow natural lakes where sediment-associate

processes are important. Four biochemical cycles were simulated, and eight state variables

were involved in the interacting systems. Sediment-associated water quality processes were

studied. A formula was obtained based on field measurements to calculate the light

attenuation coefficient by considering the effects of suspended sediment and chlorophyll. The

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

concentrations of particulate and dissolved nutrients due to adsorption-desorption were

calculated using two formulas derived based on the Langmuir Equation. The release rates of

nutrients from the bed sediment were calculated by considering the effects of the

concentration gradient across the water-sediment interface, pH, temperature and dissolved

oxygen concentration. Those formulas were tested using field measured data.

The sediment-associated processes were integrated into CCHE3D_WQ. This model was

first verified using analytical solutions of pollutant transport in open channel flow, and then it

was applied to the study of the concentration of water quality constituents in Deep Hollow

Lake. Realistic trends and magnitudes of nutrient and phytoplankton concentrations obtained

from the numerical model generally agreed with field observations. The effects of sediment-

associated processes are quite important in water quality processes. Without considering the

sediment-associated processes, the model overestimated ortho-phosphorus concentration and

underestimated organic phosphorus. After considering those processes, the numerical results

produced better agreements with field observations.

The model was also used to conduct analyses of the sensitivities of lake chlorophyll

concentration to SS concentration and nutrient loadings. Lake primary productivity is mainly

limited by suspended sediment concentration, which limits light penetration. After the

reduction of SS in the lake, there is significant increase in chlorophyll concentration. The

numerical results also show that the concentration of chlorophyll is much more sensitive to

the inorganic nitrogen than that to the inorganic phosphorus in Deep Hollow Lake.

This model provides a useful tool for predicting water quality constituents in natural

lakes. It helps us understand the interaction processes between water quality and sediment in

natural lakes. It is also useful for the decision makers to evaluate BMPs established in the

lake watersheds.

REFERENCES

[1] T. M. Wool, et al. Water Quality Analysis Simulation Program (WASP) version 6

User’s Manual, US Environmental Protection Agency, Atlanta, GA (2001).

[2] C. F. Cerco and T. Cole, User’s Guide to the CE-QUAL-ICM: Three-dimensional

Eutrophication Model, Technical Report EL-95-1 5, U.S. Army Corps of Engineers,

Vicksburg, MS (1995).

[3] Delft Hydraulics, Delft3D-WAQ: Technical Reference Manual, Delft Hydraulics, The

Netherlands (2003).

[4] Danish Hydraulic Institute, Coastal and Inland Waters in 3D, <http://www.dhi.dk>.

[5] A. R. Trancoso, S. Saraiva, L. Fernandez, P. Pina, P. Leitao and R. Neves, Ecological

Modeling, 187, 232–246 (2005).

[6] Z.G.Ji, Hydrodynamics and Water Quality: Modeling Rivers, Lakes, and Estuaries,

John Wiley, New Jersey, USA (2008).

[7] Y.Jia, S. Scott, Y. Xu, S. Huang and S.S.Y. Wang, Journal of Hydraulic Engineering,

131(8), 682–693(2005).

[8] S. C. Chapra, Surface Water Quality Modeling, The Mcgraw-Hill Companies, Inc, New

York (1997).

[9] H. G. Stefan, et al., Water Resources Research, 19, No. 1, 109-120 (1983).

Nova Science Publishers, Inc.

Three-Dimensional Numerical Modeling of Water Quality … 97

[10] M. Ishikawa and H. Nishmura, Water Research, 23, 351-359 (1989).

[11] I. Fox, M. A. Malati and R. Perry, Water Research, 23, 725-732 (1989).

[12] A. Appan and H. Wang, Journal of Environmental Engineering, 126, 993-998 (2000).

[13] M. D. Bubba, C. A. Arias and H. Brix, Water Research, 37, 3390–3400 (2003).

[14] L. H. Kim, E. Choi and M. K. Stenstrom, Chemosphere, 50 (1), 53–61 (2003).

[15] L. H. Fisher and T. M. Wood, Effect of Water-column pH on Sediment-phosphorus

Release Rate in Upper Klamath Lake, Oregon, 2001, USGS Water Resources

Investigation Report 03-4271 (2004).

[16] J. R. Romero, et al., Computational Aquatic Ecosystem Dynamics Model: CAEDYM

v2, Science Manual, University of Western Australia (2003).

[17] M. R. Loeff, et al., Limnology and Oceanography, 29, 675-686 (1984).

[18] P. A. Moore, K. R. Reddy and M. M. Fisher, Journal of Environmental Quality, 27,

1428-1439 (1998).

[19] N. Steinberger and M. Hondzo Journal of Environmental Engineering, 125, 192-200

(1999).

[20] P. C. M. Boers, Water Research, 25, 309-311 (1991).

[21] A. Kleeberg and G. Schlungbaum, Hydrobiologia, 253, 263-274 (1993).

[22] Y. Jia, T. Kitamura and S. S. Y. Wang, Journal of Hydraulic Engineering, 127, 219-

229 (2001).

[23] X. Chao, Y. Jia and S. S. Y. Wang, Journal of Hydraulic Engineering, 135, 554-563

(2009).

[24] X. Chao and Y. Jia, Three-dimensional Numerical Simulation of Cohesive Sediment

Transport in Natural Lakes, in Book "Sediment Transport", INTECH Publisher, 137-

160 (2011).

[25] I. K. Tsanis, Journal of Hydraulic Engineering, 115 (8), 1113-1134 (1989).

[26] C. Koutitas and B. O’Connor, Journal of Hydraulic Engineering, 106 (11), 1843-1865

(1980).

[27] K. R. Jin, J. H. Hamrick and T. Tisdale, Journal of Hydraulic Engineering, 126(10),

758–771 (2000).

[28] W. Huang and M. Spaulding, Journal of Hydraulic Engineering, 121(4), 300-

311(1995).

[29] F. J. Rueda and S. G. Schladow, Journal of Hydraulic Engineering, 129(2), 92-101

(2003).

[30] X. Chao, Y, Jia and D. Shields, “Three Dimensional Numerical Simulation of Flow and

Mass Transport in a Shallow Oxbow Lake”, World Water and Environmental

Resources Congress 2004, ASCE, Salt Lake City, USA, June 27-July 1 (CD-ROM).

[31] Y. Zhang, Y. Jia and S. S. Y. Wang , “Techniques for mesh density control”, The 7th

International Conference on Hydroscience and Engineering, Philadelphia, USA, Sep. 10

– Sep. 13.

[32] R. A. Rebich and S. S. Knight. The Mississippi Delta Management Systems Evaluation

Area Project, 1995-99, Mississippi Agriculture and Forestry Experiment Station,

Information Bulletin 377, Division of Agriculture, Forestry and Veterinary Medicine,

Mississippi State University (2001).

[33] Y. Zhang, CCHE Mesh Generator User Manual, Technical Report, The University of

Mississippi (2002).

[34] R. Portielje and L. Lijklema, Hydrobiologia, 253, 249-261 (1993).

Nova Science Publishers, Inc.

Xiaobo Chao and Yafei Jia

[35] D. M. DiToro, Sediment Flux Modeling, Wiley-Interscience, New York (2001).

[36] M. R. Hipsey, Computational Aquatic Ecosystem Dynamics Model: CAEDYM v2,

User Manual, University of Western Australia, Perth, Australia (2003).

[37] J. G. C. Smits and D. T. Molen, Hydrobiologia 253: 281-300 (1993).

[38] X. Chao, Y. Jia, C. Cooper, D. Shields and S.S.Y. Wang, Journal of Environmental

Engineering, 132, No. 11, 1498-1507 (2006).

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 4

INTEGRATING MAJOR ION CHEMISTRY

WITH STATISTICAL ANALYSIS FOR GEOCHEMICAL

ASSESSMENT OF GROUNDWATER QUALITY

IN COASTAL AQUIFER OF SAIJO PLAIN,

EHIME PREFECTURE, JAPAN

Pankaj Kumar1 and Ram Avtar2

1Graduate School of Life and Environmental Sciences,

University of Tsukuba, Tsukuba, Japan

2Institute of Industrial Science, The University of Tokyo,

Komaba Meguro-Ku, Tokyo, Japan

ABSTRACT

A comprehensive study of major ions, silica and isotopes was carried out to

understand the geochemical processes controlling groundwater quality in coastal aquifer

of Saijo plain, Western Japan. Various graphs were plotted using chemical data to enable

hydrochemical evaluation of the aquifer system based on the ionic constituents, water

types, hydrochemical facies, and factors controlling groundwater quality.

Carbonate weathering and atmospheric precipitation are strong factors controlling

the chemistry of major ions. From stable isotopic results, it was found that most of

sample points plotted near the local meteoric water line (LMWL) i.e. origin of ground

water is meteoric in principle; however point away from the LMWL favors exchange

with rock minerals mainly salinization process. This study is crucial considering that

Saijo city is known as one of the water capital of Japan and groundwater is the exclusive

source of drinking water in this region.

Keywords: Groundwater, salinization, weathering

Nova Science Publishers, Inc.

Pankaj Kumar and Ram Avtar

100

1. INTRODUCTION

Coastal aquifers are commonly stressed with enhanced pumping for water supply,

ultimately results in lowering of water table, increase of land subsidence and intrusion of

saline water into freshwater aquifers (Capaccioni et al. 2005; Lee et al. 2007; Trabelsi et al.

2007; Gattacceca et al. 2009). Because of dynamic nature of sea water- fresh water interface,

lack of regular monitoring network for groundwater from the coastal aquifer, can’t ensure its

diligent sustainable management (Steyl 2010). Globally a huge number of studies have been

done to understand the hydrodynamics of the coastal aquifers to assess safe yield for

groundwater withdrawal in order to implement appropriate management policies (Voudouris

2006). Saltwater intrusion is the migration of saltwater into freshwater aquifers under the

influence of groundwater development (Freeze and Cherry 1979). Integrated approach of

geochemical and isotopic analysis proved as an efficient tool to understand interaction

between groundwater and its surrounding environment which contribute to its better

management (Adams et al. 2001; Schiavo et al. 2009). Combined study of major ions (HCO3-

, Cl- , SO42- , NO3- , Ca2+ , Na+, Mg2+ and K+ ) and environmental isotopes (δD, δ18O and

87Sr/86Sr) is an proficient investigative means to know the aquifer matrix chemistry and

water–rock interaction along the flow path in the aquifer to trace groundwater evolution and

mechanism of salt water – fresh water mixing (Kim et al. 2003; Vengosh et al. 2005; Chen et

al. 2007; Jorgensen et al. 2008; Schiavo et al. 2009; Langman et al. 2010). Piper and Durov

diagrams (in original and modified versions) have been consistently applied to study the

cation exchange and reverse cations reaction during dynamics of sea water intrusion or

freshening occurring in alluvial coastal aquifer (Ray et al. 2008; Petalas et al. 2009; Forcada

2010). Kurihara (1972) described that geological structure in Saijo plain is very complex

because of subsidence took place in alluvial fan of Kamo River. Nakano et al. (2008) reported

that groundwater in the Saijo city shows a different quality between the eastern Saijo plain

and western Shuso one, indicating that each groundwater constitutes an independent aquifer

system. Saijo city office (2008) also reported that there is a decline in groundwater level in

northern coastal area of the Saijo plain, which may trigger sea water intrusion. Saijo plain is

known for its plenty of good quality groundwater in western Japan but despite its importance,

little is known about the natural phenomena that govern the chemical composition of

groundwater. The prime objective of this study is to elucidate the mechanism or chemical

processes that control water chemistry with special attention on sea water- fresh water

interaction pattern on this small coastal aquifer of Saijo plain.

2. STUDY AREA

Saijo plain is located in Saijo City, Ehime Prefecture, and the north-western part of the

Shikoku Island in the western Japan (Figure 1). The Saijo plain is bounded on the south by

the range of Ishizuchi mountains (1982 m.a.m.s.l.), the highest peak in the western Japan, on

the north by the Seto Inland Sea, on the west by the Shuso plain and on the east by Toyo

region. From the geological point of view, mountains mainly consist of the Sambagawa

metamorphic rocks, whereas the alluvial plain is composed of the Pliocene-Quaternary

sediment (with approximately 1 km thickness).

Nova Science Publishers, Inc.

Integrating Major Ion Chemistry with Statistical Analysis … 101

Figure 1. Geological map of Saijo pain, Western Japan.

There are several active faults with right lateral motion. The Okamura fault, an active

fault segment of the Median Tectonic Line (MTL), is the boundary between the mountainous

area and the plain.

The Komatsu fault is the main fault in the plain, location of which is partly unknown

because of the overlying sediments. In Saijo city, drainage system mainly consists of four

rivers namely Kamo, Uzui, Nakayama and Daimyojin Rivers flow from the mountain to the

coast. The Kamo and the Uzui Rivers are the major component recharging the unconfined

groundwater of the alluvial fan in the Saijo plain. The Saijo plain belongs to the Setouchi

climate, a temperate climate with a relatively small amount of mean annual precipitation

(1433 mm, Saijo city, 2008). The mean annual temperature is approximately 15°C.

The spatial distribution of groundwater under the Saijo plain is divided into three zones

from southern mountains to northern coast: the spring, the artesian and the coastal

groundwater zones. The spring area is mainly featured as a recharge area despite some local

upwelling.

3. METHODOLOGY

Sampling wells were selected in such a way that they represent different geological

formations and land use patterns at varying topography of the coastal area of Saijo plain. A

total of twenty two water samples (eleven from each confined and unconfined aquifer) were

Nova Science Publishers, Inc.

Pankaj Kumar and Ram Avtar

102

collected during July, 2010. In-situ measurements for EC, pH, ORP and temperature were

done using respective probes. The water samples were collected in pre-rinsed clean

polyethylene bottles. The concentration of HCO3− was analyzed by acid titration (using

Metrohm Multi-Dosimat) while other anions Cl−, NO3−, SO42−, Br- and PO43− were analyzed

by DIONEX ICS-90 ion chromatograph. Major cations and trace elements were determined

with inductively coupled plasma-mass spectrometry (ICP-MS). Oxygen and hydrogen

isotopes were analyzed by mass spectrometer (MODEL MAT 252, Thermo Finnigan Inc.) in

university of Tsukuba. The results for both isotopes are expressed through deviation from the

VSMOW (Vienna Standard Mean Ocean Water) standard using the δ-scale according to the

equation and unit is per mil.

where, R is the isotopic ratio (i.e. 2H/1H and 18O/16O) for the sample and standard. Analytical

precisions of stable isotopes were better than 0.1‰ for δ18O and 1.0‰ for δ2H. For major

ions, analytical precision was checked by normalized inorganic charge balance (NICB)

(Kumar et al. 2010). This is defined as [(Tz+ - Tz-)/(Tz+ + Tz-)] and represents the fractional

difference between total cations and anions. The observed charge balance supports the quality

of the data points, which is better than ±5% and generally this charge imbalance came in

favor of positive charge.

Factor analysis is used here for the classification and assessment of groundwater quality

using the Statistical Package for Social Sciences (SPSS) software package (SPSS Inc., USA,

version 12).

Factor analysis was applied on experimental data standardizing through z-scale

transformation to avoid misclassification of samples due to the wide differences in data

dimensionality (Singh et al., 2011). It gave information about the variables responsible for the

spatial variation in groundwater quality.

4. RESULTS AND DISCUSSIONS

4.1. General Water Chemistry

Statistical summary (i.e. minimum, maximum, average and the standard deviation) of the

analytical results for each water-quality characteristic analyzed is given in Table 1. Average

value trend for cations and anions in groundwater samples was found as Ca2+ > Na+ > Mg2+ >

K+ and HCO3-> Cl- > SO42- > NO3- respectively.

All the physico-chemical parameters within the highest desirable permissible limit

recommended by WHO (World Health Organization) (2004), except Cl- at some sampling

points. High HCO3- represents the major source of alkalinity caused by the presence of

carbonaceous sandstones in the aquifers and weathering of carbonate minerals related to the

flushing of carbonate enrich water from unsaturated zone, where it is formed by

decomposition of organic matter.

×1 0 0 0

Sample VSMOW

VSMOW

−

⎡⎤

=⎢⎥

⎣⎦

‰

Nova Science Publishers, Inc.

Integrating Major Ion Chemistry with Statistical Analysis … 103

Table 1. Statistical summary of hydrogeochemical parameters of groundwater samples

of Saijo plain, Japan

Parameters Minimum Maximum Average Std. Dev.

Temp.( ºC) 14.0 23.2 17.8 2.6

Ph 5.9 7.6 6.9 0.5

EC (µs/cm) 87.0 1594.0 203.6 316.3

ORP (mv) 34.0 280.0 175.1 64.6

Na+ (mg/L) 1.1 97.1 11.3 23.7

K+!(mg/L) 0.7 6.0 2.2 1.3

Ca2+!(mg/L) 1.1 92.4 18.1 17.2

Mg2+!(mg/L) 0.9 76.4 6.3 15.7

Cl-!(mg/L) 1.9 502.7 29.1 106.3

SO42-!(mg/L) 6.5 56.6 14.2 11.3

HCO3- (mg/L) 25.3 57.1 41.0 7.6

NO3- (mg/L) 0.2 37.8 6.2 8.6

SiO2 (mg/L) 5.9 33.2 11.9 5.5

δ18O (‰) -9.4 -7.8 -8.5 0.4

δD (‰) -61.6 -53.8 -57.9 2.5

Figure 2. Scatter diagram for groundwater showing relationship between (a) Ca2+Mg2+ and Tz+ and (b)

Na++K+ and Tz+ where Ca2+Mg2+ accounts for most of cations in the groundwater.

Nova Science Publishers, Inc.

Pankaj Kumar and Ram Avtar

104

Scatter plot between (Ca2+ + Mg2+) versus Tz+ for the groundwater yields a strong linear

trend indicating that they accounts for most of the cations (Figure 2a) whereas, the trend

between (Na+ + K+) versus Tz+ was weak (Figure 2b).

These results suggest that the carbonate weathering is the dominant process and

contribution of cations via alumino-silicate weathering is low in comparison to carbonate

weathering. Low average value of silica also supports the above mentioned fact. Stiff diagram

is used to make a quick visual comparison between waters from different sources and to gain

better insight into the hydro-geochemical processes operating in the coastal aquifer of Saijo

plain which resulted in the spatio-temporal variation in hydrochemistry (Figure 3). From this

diagram, it was found that water samples were mainly dominated by Ca-HCO3 type which is

very consistent with the geological feature of the alluvial plain. On contrast, on the west side

of the plain there is preliminary signature of sea water intrusion which was characterized by

Ca-Cl and Na-Cl type of water samples taken from bore wells with screen depth more than 25

meter. This shows that average high value of Cl- might be because of encounter of screen

depth to the sea water – fresh water interface. On the east side of plain, water is enriched with

anions SO42- and NO3- indicating the dominance of anthropogenic inputs like agricultural

activities and sewage effluents.

Figure 3. Hexadiagram showing groundwater qulity at spatial scale.

4.2. Isotopic Signature of Groundwater

Value for δ18O ranges from (-9.4 to -7.8 ‰) while corresponding value for δ2H ranges

from (-61.6 to -53.8‰) (Table 1). The relationship between δ18O and δ2H values for confined

Nova Science Publishers, Inc.

Integrating Major Ion Chemistry with Statistical Analysis … 105

and unconfined water is shown in Figure 4. Except few points, most of the water samples

were clustered on or near the meteoric water line. Thus, these results are not affected

according to the deviations isotopic compositions away from the meteoric water line,

including evaporation from open surface and exchange with rock minerals. It was found that

confined groundwater samples tend to have lower value than unconfined water which

suggested that recharge elevation of confined groundwater is higher than one of unconfined

water. Confined groundwater samples away from the LMWL might be results of evaporation

and exchange with rock minerals. To confirm the process behind it, scatter plot was drawn to

show relationship between δ18O and chloride concentrations of the groundwater in the study

region (Figure 5). The dash or broken line in this figure denotes the precipitation-seawater

mixing line.

Figure 4. Scatter plot showing relation between δ180 and δ2H for groundwater. (Here LMWL –Local

meteoric water line).

Figure 5. Scatter plot showing relation between δ180 and Cl for groundwater and its trend on mixing

line.

Groundwater samples were plotted along the δ18O axis is due to the verylow chloride

concentration of groundwater samples compared to their wide range of δ18Ovalues. However,

two groundwater samples are distinguishable from other samples indicating relatively high

Nova Science Publishers, Inc.

Pankaj Kumar and Ram Avtar

106

chloride concentrations with value of 52.0 and 502.7 mg/l, respectively (also shown in west

side in Figure 3), and they converge toward the composition of sea water along the mixing

line. This suggests that salinization is caused solely by mixing with coastal marine water.

4.3. Factor Analysis

The factor analysis (shown in Table 2) resulted in three principal components

representing three main sources of variation in the hydrochemistry at the regional level. These

three components accounts for 76.38 % of the variance in the hydro-chemical data. Factor 1

has higher loadings for EC, Ca2+, Mg2+ and HCO3

- which Indicating weathering of

carbonaceous materials in the area. Factor 2 has higher loading of Cl-, δ18O and δ2H which

suggests about groundwater salinization. Factor 3 shows higher loading for SO42- and NO3-

indicating anthropogenic activities determining water quality at some points.

Table 2. Multivariate factor analysis score for groundwater samples of Saijo plain,

Japan

Variables F1 F2 F3

Temp. 0.35 -0.42 0.867

pH 0.22 0.37 0.13

EC 0.98 0.34 0.65

ORP -0.49 -0.24 0.64

Na+ 0.55 0.72 -0.18

K+ 0.25 0.80 0.28

Mg2+ 0.73 0.13 0.16

Ca2+ 0.94 0.61 0.39

Cl- 0.47 0.88 0.51

NO3- 0.12 0.56 0.72

SO42 - 0.38 0.27 0.86

HCO3- 0.92 -0.71 0.17

SiO2 0.46 -0.28 -0.02

δ18O -0.03 0.77 -0.70

H -0.32 0.51 -0.43

Eigen Value 6.86 2.67 1.92

Percentage of Variance 45.74 17.82 12.82

Cumulative percentage 45.74 63.56 76.38

CONCLUSION AND RECOMMENDATIONS

Hydrochemical analysis suggests that most of the water samples in Saijo plain were

showing good quality i.e. all the chemical parameters were within the highest desirable limit

set by WHO, while two samples shown high chloride concentration i.e. state of salinization

which needs immediate attention. From isotopic study, it was found that most of the water

Nova Science Publishers, Inc.

Integrating Major Ion Chemistry with Statistical Analysis … 107

samples are of meteoric origin. Relation between δ18O and chloride firmly supported local

SW-FW mixing reason behind elevated chloride concentration in case of two water samples.

Factor analysis also firmly supported the results from graphical representation. From above

work it was found that for samples taken from bore well with screen depth >25 meter has a

problem of salinization so it will be recommended to have an alternate tube/bore well with

screen depth <15 meter in order to prevent the encounter of SW–FW interface as an

alternative. To validate qualitative aspects of water quality, study for present and future status

of SW intrusion through numerical simulation can be a future perspective.

ACKNOWLEDGMENTS

The authors are highly thankful to the Monbukagakusho (MEXT) Japanese Government

fellowship which helped to pursue research. Authors also want to put on record contribution

of Graduate School of Life and Environmental Science, University of Tsukuba for facilitating

data analysis in its labs.

REFERENCES

Adams, S., Titus, R., Pietersen, K., and Tredoux, G. (2001). Hydrochemical characteristics of

aquifers near Sutherland in the Western Karoo, South Africa. Journal of Hydrology, 241,

91-93.

Capaccioni, B., Didero, M., Paletta, C., and Didero, L., (2005). Saline intrusion and

refreshening in a multilayer coastal aquifer in the Catania Plain (Sicily, Southern Italy):

dynamics of degradation processes according to the hydrochemical characteristics of

groundwaters. Journal of Hydrology, 307 (1– 4), 1–16.

Chen, K.P., and Jiao, J.J. (2007). Seawater intrusion and aquifer freshening near reclaimed

coastal area of Shenzhen. Water Science and Technology, 7,137-145.

Forcada, E. G. (2010). Dynamics of Sea water interface using hydrochemical facies evolution

diagram. Ground Water, 48, 2, 212-216.

Freeze, R.A., and Cherry, J.A., 1979. Groundwater, Prentice-Hall.

Gattacceca, J.C.,Coulomb, C.V., Mayer, A., Claude, C., Radakovitch, O., Conchetto, E., and

Hamelin, B. (2009). Isotopic and geochemical characterization of salinization in the

shallow aquifers of a reclaimed subsiding zone: The southern Venice Lagoon coastland.

Journal of Hydrology, 378, 46–61.

Jorgensen, N.O., Andersen, M.S., and Engesgaard, P. (2008). Investigation of a dynamic

seawater intrusion event using strontium isotopes (87Sr/86Sr). Journal of Hydrology,

348, 257– 269.

Kim,Y., Lee, K.S., Koh, D.C., Lee, D.H., Lee, S.G., Park, W.B., Koh, G.W., and Woo, N.C.

(2003). Hydrogeochemical and isotopic evidence of groundwater salinization in a coastal

aquifer: a case study in Jeju volcanic island, Korea. Journal of Hydrology, 270, 282-294.

Kumar, P., Kumar, M., Ramanathan, A.L., and Tsujimura, M. (2010). Tracing the factors

responsible for arsenic enrichment in groundwater of the middle Gangetic Plain, India: a

source identification perspective. Environmental Geochemistry and Health, 32,129–146.

Nova Science Publishers, Inc.

Pankaj Kumar and Ram Avtar

108

Kurihara, G. (1972). Geology of the alluvial plains of the southern coastal area of Setouchi.

Tohoku Univ. Inst. Geol. Pal. Center, 73, 31-65 (in Japanese with English abstract).

Langman, J. B., and Ellis, A. S. (2010). A multi-isotope (δD, δ18O, 87Sr/86Sr, and δ11B)

approach for identifying saltwater intrusion and resolving groundwater evolution along

the Western Caprock Escarpment of the Southern High Plains, New Mexico. Applied

Geochemistry, 25, 159–174.

Lee, J.Y., and Song, A.S.H. (2007). Evaluation of groundwater quality in coastal areas:

implications for sustainable agriculture. Environmental Geology, 52, 1231–1242.

Nakano, T., Saitoh, Y., and Tokumasu, M. (2008). Geological and human impacts on the

aquifer system of the Saijo basin, western Japan. Proceedings of 36th IAH Congress.

Petalas, C., Pisinaras, V., Gemitzi, A., Tsihrintzis, V.A., and Ouzounis, K. (2009). Current

conditions of saltwater intrusion in the coastal Rhodope aquifer system, northeastern

Greece, Desalination, 237, 22-41.

Ray, R.K., and Mukherjee, R. (2008). Reproducing the Piper Trilinear diagram in

Rectangular Coordinates. Ground Water, 46 (6), 893-896.

Saijo City Office. (2008). Annual report of ground water for 2008. Saijo city, life

environment section, 88pp (in Japanese).

Schiavo, M.A., Hauser, S., and Povinec, P.P. (2009). Stable isotopes of water as a tool to

study groundwater–seawater interactions in coastal south-eastern Sicily. Journal of

Hydrology, 364, 40– 49.

Singh, C.K., Shashtri, S., and Mukherjee, S. (2011). Integrating multivariate statistical

analysis with GIS for geochemical assessment of groundwater quality in Shiwaliks of

Punjab, India. Environmental Erath Science, 62, 1387-1405.

Steyl, G., and Dennis, I. (2010). Review of coastal-area aquifers in Africa. Hydrogeology

Journal, 18, 217–225.

Trabelsi, R., Zairi, M., and Dhia, H. (2007). Groundwater salinization of the Sfax superficial

aquifer, Tunisia. Hydrogeology Journal, 15 (7), 1341–1355.

Vengosh, A., Kloppmann, W., Marei, A., Livshitz, Y., Gutierrez, A.,Banna, M., Guerrot, C.,

Pankratov, I., and Raanan, H. (2005). Sources of salinity and boron in the Gaza strip:

natural contaminant flow in the southern Mediterranean coastal aquifer. Water Resource

Research, 41, W01013.

Voudouris, K.S. (2006). Groundwater Balance and Safe Yield of the coastal aquifer system in

NEastern Korinthia, Greece. Applied Geography, 26, 291–311.

WHO. (2004). Guidelines for drinking water quality-II pp 333. Geneva, Environmental

Health Criteria, 5.

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 5

SUITABILITY OF GROUNDWATER OF ZEUSS-KOUTINE

AQUIFER (SOUTHERN OF TUNISIA) FOR DOMESTIC

AND AGRICULTURAL USE

Fadoua Hamzaoui-Azaza

∗

1, Besma Tlili-Zrelli1,

Rachida Bouhlila2 and Moncef Gueddari1

1Laboratory of Geochemistry and Environmental Geology,

Department of Geology, Faculty of Mathematical,

Physical and Natural Sciences, University Campus,Tunis, Tunisia

2Modeling in Hydraulic and Environment Laboratory,

National Engineers School of Tunis, Tunisia

ABSTRACT

In arid and semi-arid regions from North African countries, ground water forms the

major source of water supply for socio-economic growth and sustainability of the

environment.

Hydrogeochemical characteristics of deep groundwater in southeast Tunisia have

been assessed to identify its suitability for drinking and irrigation applications

In order to evaluate the quality of groundwater in Zeuss-Koutine, which represents

the principal resource of water for Medenine, the hydrochemical data have been analyzed

using geochemical methods and multivariate statistical techniques such as Principal

component analyses and Cluster analyses.

In this study 14 wells were sampled in summer (July) and their water was analyzed

for various variables, including temperature, pH, Total Dissolved Solids (TDS), Na+, Cl-,

Ca2+, Mg2+, SO42-, K+, HCO3-, Fe3+, Mn2+, Zn2+, Al3+, Pb2+, Cr3+, Cu2+ and F-

The results showed that the majority of ions are above the maximum desirable limit

although trace metals are within the maximum permissible limit for drinking water.

Besides and on the basis on the water quality index WQI, groundwater has been

∗ Corresponding autors: Fadoua Hamzaoui-Azaza, 1Laboratory of Geochemistry and Environmental Geology,

Department of Geology, Faculty of Mathematical, Physical and Natural Sciences, University Campus.Tunis.

Tunisia; fadoua_fst@yahoo.fr.

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

110

classified into “unsuitable for drinking purposes”, “Very poor water “and “Poor water”

and can’t be used for drinking purposes without any treatment. The calculated values for

sodium adsorption ratio (SAR) indicate well to permissible use of water for irrigation

signifying that sodicity is very low. According to the USSL diagram, the most dominant

classes of water sampling are C4a-S1 and C4b-S1 indicating very high and extremely

high salinity waters which are unsuitable for irrigation with a restrict drainage.

In the whole domain, a significant increase in the degree of water mineralization was

observed from southwest to northeast, following the regional flow direction.

Interpretation of hydrochemical data reveals that Water quality is mainly dominated

by dissolution of evaporates minerals. Results obtained from principal component

analyses (PCA) showed that 3 components explaine more than 75 % of the total variance

in the groundwater quality and demonstrated that the variable responsible for water

quality are largely related to soluble salts species (Na+, Cl-, Ca2+, Mg2+, SO42- and K+).

Cluster analysis showed that sites sampling can be grouped in tow clusters.

INTRODUCTION

Groundwater is an important source of water supply throughout the world. It is estimated

that approximately one-third of the world’s population use groundwater for drinking (Nickson

et al. 2005). In arid zones, water is a rare and precious resource. The exploitation of water

resources is a complex problem in the framework of sustainable agricultural development in

these regions (Hamzaoui-Azaza et al., 2009).

The current situation of the water resources and their uses in Tunisia, which is the most

drought-stressed countries in the middle East and North Africa region, present a common

stakes with many areas of the Mediterranean basin: limited and already largely exploited

resources to answer the growth of the needs, the increasing use of so-called non conventional

resources, the overdependence on groundwater to meet increasing demands of domestic,

agriculture, and industry sectors, an increasing merchandising of the resources, constraining

climatic conditions which come to reinforce the tensions around water. Under such

conditions, Tunisia as all other countries subject to these same constraints, have no choice but

to increase their vigilance around the issues related to rational management and conservation

of water resources. While the demography is expanding in South Tunisia, the water needs of

the three main sectors, irrigation, drinking water supply and industry, are expected to increase

in order to provide the population with employment and life conditions enabling people to

settle in their homeland. In fact in this region of Tunisia, the groundwater is currently the only

available water resources upon which depend all the socioeconomic sectors. In the medium

term, desalination of sea water, preferably using solar energy available in these latitudes,

could be one of the best alternatives to the supply of freshwater.

In terms of quality, different classifications for water use are commonly cited and

geochemical and statistical methods enable us to deduce the sources of water mineralization.

The chemical composition of groundwater is controlled by many factors that include the

climate conditions, the geological structure and mineralogy of the watershed and the aquifer,

and the geological processes within the aquifer (Rosen and Jones 1998). The interaction of all

these factors leads, for a specific aquifer, to a particular type of water whose chemical

composition determines its uses.

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 111

Geochemical methods and statistical techniques including Hierarchical cluster analysis

(HCA) and Principal component analysis (PCA) are respectively used to identify factors and

the dominant mechanisms controlling groundwater chemistry.

A water quality index (WQI), which can serve as a useful tool for correct evaluating the

quality of groundwater and surface water (Abassi 1999, Adak et al., 2001), was generated for

the entire study area to define the suitability of water for drinking purpose and to classify

groundwater in the area into spatial water quality types. Generally, indices have been

developed to summarize water quality data in an easily and understandable format (Saeedi

2010).

The high salinities of groundwater from some of the wells in the area and the large

dependence of the communities on groundwater for irrigation require a global evaluation of

the quality of the resource in the area for irrigation purposes.

Numerous parameters such as Sodium adsorption ratio (SAR), Percent Sodium (% Na),

Residual Sodium Carbonate (RSC) and Percent Magnesium (% Mg) have been used to

evaluate the suitability of groundwater for irrigation purposes (Paliwal 1972, Domenico and

Schwartz 1990, Tijani 1994, Shaki 2006).

In the current study, a first attempt was made to assess the groundwater quality of Zeuss

Koutine aquifer and its suitability for drinking and agricultural uses, thus helping to the

effective and sustainable management of groundwater resources in this area. Sustainable

development of water sources must go hand in hand with improved sanitation and hygiene

practice.

STUDY AREA

The study site, Zeuss-Koutine aquifer, is described in details by Hamzaoui Azaza et al.,

2011. Tt belongs to the region of south eastern Tunisia. It is situated in the north of the city of

Médenine between the Jeffara plain and Matmata mountains, actually ending in the saline

depression (Sebkha) of Oum Zessar before ending into the Gulf of Gabès in the

Mediterranean sea. It is bordered, at the north, by the Segai plain and the town of Mareth, in

the south by the Trias Sandstone aquifer and in the south east by the Jorf Miocene aquifer

(Figure 1). The study site is characterized by steppe vegetation. It is characterized by an arid

to semi-arid climate with a low and highly irregular total rainfall (150-240 mm). Temperature

differences are extreme between the seasons ranging from-3°C (winter) to as high as 48°C

(summer) (Ouassar 2007).The hydrogeographic network is quite dense.

It is constituted primarily by three big Wadiis named Zeuss Wadi, Koutine-Om Ezzassar

Wadi and Zigzaou Wadi. The geological formations consist of alternating continental and

marine origin. The oldest submerging layers are represented by a marine, superior Permian,

and the most recent ones are of the recent Quaternary (Morhange and Pirazzoli 2005).

Between the Permian and Quaternary formations we find confined strata of different

ages, which are generally declining in northward direction (Gaubbi 1988). Zeuss-Koutine

region is bordered by three main structures that define the South-Eastern Tunisia: the Dahar

monocline, West and North-West, the Medenine Tebaga monocline, South, and the plain of

the Djeffara, East and North (Hamzaoui Azaza et al., 2011).

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

112

Figure 1. Location map of the study area and Spatial distributions of groundwater level in 2004.

The Zeuss-Koutine aquifer unit constitutes an important water source used for various

purposes at an average abstraction estimated at a rate of 350 l/s. The extraction of water from

this reservoir started in 1962. The continuous monitoring has shown that the withdrawal rate

increased from 102 l/s in 1974 to 440 l/s in 2005 resulting in a decline of the mean

piezometric level of 0.33 m in 1982 and 1.02 m in 2004 (DGRE, 2004).

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 113

Zeuss-Koutine aquifer is constituted by the following litho-stratigraphic units: Jurassic

limestone and dolomite; Albo-Aptian calcitic dolomites; Turonian dolomites and dolomitic

limestone and Lower Senonian limestone. However, in Medenine region, the Jurassic

limestone and dolomite constitute the main aquifer material because of its considerable

thickness reaching up to 120 m. The thickness of this aquifer is between 30 and 200 m; and

its depth varies between 70 and 250 m (Hamzaoui-Azaza 2011).

Results of aquifer tests indicate that transmissivities of the aquifer varies between 0.055

and 0.2 m2 s-1 (OSS 2005).

This aquifer unit is recharged by water ﬂowing from Matmata Mountains, where

important outcrops are made up of Jurassic limestone, as well as by local infiltration from

several rivers. Besides, The Zeuss-Koutine aquifer unit is fed by the underlying Triassic

Sandstone aquifer. The general groundwater flow in the horizontal plane s takes part from

South-West to North-East (OSS 2005).

Water of the Zeuss-Koutine aquifer is used unevenly by different economic sectors.

However, drinking water supply remains the primary use. Anthropogenic activities in the

study area rely mainly on agricultural.

SAMPLE COLLECTION AND ANALYTICAL TECHNIQUES

In order to study the groundwater quality in the Zeuss-Koutine aquifer 14 groundwater

samples, taken from wells used for drinking, industrial, irrigation, and other domestic

purposes, were collected in July 2005 from boreholes ranging in depth from 91 to 577 m after

10 min of pumping. The geographical location of the sampling sites is shown in Figure 4.

Groundwater samples were collected in sterilized polythene bottles through 0.45-μm in-line

ﬁlters. All samples were transported to the laboratory to analysis and kept in a refrigerator

below 4◦ C. Each of the groundwater samples was analyzed for 17 parameters such as

physico-chemical parameters, majors and trace elements.

Unstable parameters temperature, pH and salinity were measured in the ﬁeld using well-

calibrated digital sensors. The pH electrode was calibrated against pH 4, 7, and/or 10 buffers.

Samples for the analysis of cations were acidiﬁed to pH 2 by adding several drops of ultra-

pure nitric acid. The analyses for various chemical parameters to assess the groundwater

quality were carried out using standard procedures (Rodier, 1996). The water samples were

analyzed at an approved chemical laboratory in Tunisia. Calcium and magnesium

concentration were determined by complexometric titration method using Ethylene Diamine

Tetra-acetic Acid (EDTA). Sodium and potassium concentrations were determined using

Flame Photometer. Bicarbonate concentration was determined by acidimetric titration method

using methyl orange as indicator. Chloride was determined by using 0.1N AgNO3 solution.

Sulphate was determined by gravimetric method.

Metal ion concentrations were determined by atomic absorption spectrometry.

Quantiﬁcation of metals by means of a calibration curves of aqueous standard solutions of

respective metals. These calibration curves were determined several times during the period

of analysis. The quality of the chemical analyses was carefully inspected by checking ion

balances. The ion balance errors for the analyses were within ± 5%.

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

114

Visually communicating iso-concentration/contour maps were constructed using Surfur

7.0 software to delineate spatial variation of physico-chemical characteristics of groundwater

samples.

RESULTS AND INTERPRETATION

Understanding the groundwater chemistry is important as it is the main factor

determining its suitability for drinking, agricultural and industrial purposes (Subramani et al.

2005). The chemical composition of groundwater is generally controlled by several factors

that include Climate, soil characteristics, geological structure and mineralogy of the

watersheds and aquifers and geochemical processes within the aquifer (Jallali, 2010). The

mixing/non-mixing of different types of groundwater may also play important roles in

determining the quality of the groundwater (Reghunath et al., 2002). The interaction and the

combination of all factors leads to different water types. Physical and chemical parameters

including statistical measures such as minimum, maximum, average, and median are given in

Table 1. The physical and chemical parameters of the analytical results of groundwater were

compared with the standard guideline values recommended by the World Health Organisation

(WHO 2004) for drinking water.

Table 1. Summary statistics of the analytical data such as minimum, maximum,

average, and median

Parameters Average Mean Min Max Variance

Standard

deviation

WHO

2004

T°C 27.80 27.10 22.00 36.10 11.29 3.36

pH 7.56 7.60 7.00 7.90 0.05 0.22

Optimum 6.5-

9.5

Na(mg/l) 641.66 740.00 270.00 1327.56 95845.22 309.59 200 (mg/l)

K (mg/l) 18.02 20.00 8.60 33.24 47.53 6.89 30 (mg/l)

Ca (mg/l) 273.77 302.00 107.00 445.00 11319.35 106.39 200 (mg/l)

Mg (mg/l) 121.71 133.00 73.00 194.08 1208.14 34.76 200 (mg/l)

SO4 (mg/l) 945.62 930.00 534.05 1484.64 95057.76 308.31 400 (mg/l)

HCO3 (mg/l) 184.08 191.64 106.00 240.34 1533.39 39.16 380 (mg/l)

Cl (mg/l) 1001.65 1135.97 338.00 2159.11 293109.19 541.40 250 (mg/l)

TDS (mg/l) 3199.93 3480.00 1520.00 5400.00 1486496.07 1219.22 1000 (mg/l)

Fe 0.18 0.01 0.00 1.67 0.22 0.47

Mn (ug/l) 0.01 0.00 0.00 0.06 0.00 0.02 0.4 (mg/l)

Cu (ug/l) 0.02 0.00 0.00 0.05 0.00 0.02 2 (mg/l)

Al (ug/l) 0.20 0.06 0.00 0.79 0.06 0.25 0.2 (mg/l)

Si (ug/l) 7.54 5.67 2.81 13.64 17.66 4.20

Zn (ug/l) 0.06 0.03 0.00 0.20 0.00 0.06 3 (mg/l)

Pb (ug/l) 2.38 0.00 0.00 6.30 4.50 2.12 0.01 (mg/l)

Cr (ug/l) 0.37 0.03 0.00 1.40 0.24 0.49 0.05 (mg/l)

F (mg/l) 2.18 2.20 1.64 2.80 0.10 0.31 1.5 (mg/l)

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 115

Water chemical data has been first approached by a description of the spatial variation of

some ions, than has been used for computation irrigation quality parameters Sodium

adsorption ratio (SAR), Percent Sodium (% Na), Residual Sodium Carbonate (RSC) and

Percent Magnesium (% Mg), besides, tow statistical analyses methods were applied :

Principal Component Analysis (PCA) and Cluster Analyses.

Physico-Chemical Parameters

Temperature and pH

The temperature variation of the groundwater in the study area ranges from 22◦ C to

30.6◦C. The pH of the groundwater in the study area ranges from 7.0 to 7.9 indicating that the

waters are generally neutral to slightly alkaline. The pH values of groundwater samples are

within the permissible limit suggested by WHO (WHO, 2004).

Salinity

In natural waters, dissolved solids consists mainly of inorganic salts such as bicarbonates,

sulfates, chlorides, calcium, phosphates, and magnesium, potassium, sodium, etc., and a small

amount of organic matter and dissolved gases. The salinity value ranging from 1520 mg/l to

5400 mg/l with an average value of 3199 mg/l. The highest values of salinity are generally

registered when the movement of groundwater is at its least, hence the salinity is influenced

with depth/time and recharge/discharge area relationships. Water from recharge areas is

usually diluted in contrast in the discharge areas, it is often relatively saline (Chilton 1992).

Consequently, the increase in the groundwater salinity of the aquifer from southwest

towards the north and northeast may be attributed to the farthest distance from the naturally

recharged area and the increased saturated thickness of the aquifer in this direction.

Spatial distribution maps, showing an increase towards groundwater flow (Figure 2). The

map shows also that P9, P10, P11 and P12 wells that are located in the upstream section of

the aquifer (recharge zone) have the lowest salinity levels. Relatively low TDS values

characterize wells located near the River and reveal the dilution of groundwater by the water

from Triassic aquifer. About 100% of the samples analyzed were found above the desirable

limit of 1000 mg/L (WHO 2004) This spatial variations reﬂects the variation of Na, Cl, Ca,

Mg, K and SO4 concentrations, with the linear correlations (R2= 0.98, 0.96, 0.97, 0.79, 0.91

and 0.77) between salinity and the concentrations of these ions, suggesting that the increase of

these elements et especially Na and Cl concentrations contributes to the salinity increases.

Major Ions

Chlorides and Sodium

Chloride is a usually distributed element in nature in one or in combination with other

elements. It has a high affinity towards sodium. Therefore, its concentration is high in ground

waters, where the temperature is high and rainfall is less (Geetha et al, 2008; Ramakrishnaiah

et al., 2009). Chloride Concentrations vary between 338 and 2159 mg/l with an average value

of 1002 mg/l. Those of the sodium are ranging between 270 and 1328 mg/l with an average

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

116

value of 740 mg/l. Sodium is the dominant cation in Zeuss-Koutine aquifer, whereas Chloride

is the dominant anion. Excess in Chloride and sodium concentration is found in wells.

The possible source of sodium and chloride concentration in groundwater would be due

to the dissolution of rock salts and weathering of sodium-bearing minerals (Krishna Kumar et

al., 2009).

If the halite dissolution process is responsible for the sodium, Na/Cl ratio should be

approximately 1, while the Na/Cl ratio grather than 1 reflect generally a release of sodium

from silicate weathering (Meyback, 1987), if Na/Cl ratio is less than 1, the reduction of Na

concentration may be due to ion exchange process (Jeevanandam et al., 2007). In this present

study, Na/Cl ratio is greater than 1 in the predominant groundwater samples (9 samples).

The chloride and sodium ions concentration in groundwater of the study area exceeds the

maximum allowable limit of 600 mg/l in respectively nine and ten locations (WHO 2004).

Figure 2. Spatial distribution of Salinity and location of the sampling wells.

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 117

Calcium and Magnesium

The origin of Calcium and magnesium ions present in groundwater is generally the

weathering of dolomites, limestone, gypsum and anhydrites, whereas the cation exchange

process may explain the contribution of calcium in aquifer (Subramani et al., 2010).

The calcium concentrations oscillate between 107 and 445 mg/l, 64 % of the samples

exceed the desirable limit of 200 mg/l as per WHO standard (WHO 2004). Those of

Magnesium range between 73 and 194 mg/l. The recommended limit for Magnesium in

natural water is 150 mg/l (WHO 2004). 80% of the groundwater samples fall below this limit.

Sulfates

Natural concentrations of sulphates may be enriched by weathering of sulfate minerals,

such as gypsum and anhydrite. Sulfate concentrations varied from 534 to 1485 mg/l with an

average value of 945 mg/l. The spatial distribution maps of sulfate concentration show that

high concentrations are located in North-Est, in P4 and P5 well and in P13 and P3 well at the

South-East. All groundwater samples exceed the desirable limit of 250 mg/l as per WHO

standard (WHO 2004).

Potassium

Potassium concentrations in the Zeuss-Koutine aquifer were nearly homogeneous within

each sample site. The values range from 9 to 33 mg/L, Potassium concentration in

groundwater of the study area is within the maximum allowable limit in all the sample

locations (WHO 2004). This low level of potassium is explained by the tendency of this

element to be fixed by clay minerals and to participate in the formation of secondary minerals

(Matheis 1982, Nwankwoala and Udom 2011).

Alkalinity

Water alkalinity of the Zeuss-Koutine aquifer is the only product of bicarbonate ions,

given that pH values are almost near to neutral. The bicarbonate ion concentrations ranged

from 106 to 240 mg/l with an average value of 184 mg/l. The bicarbonate concentration in

groundwater is derived from carbonate weathering:

CaCO3 + CO2 + H2O → Ca2+ +2 HCO3- and CO2 + H2O → H++HCO−3

The spatial distribution of alkalinity is rather different from the maps of the main ions. In

fact, wells where water shows the lowest concentration major ion are characterized by the

highest bicarbonate concentration levels. The increase of bicarbonate concentration in the

groundwater may be due to the availability of carbonate minerals in the recharge areas and

silicate weathering (Elango et al., 2003, Krishna Kumar et al., 2009). The alkalinity values

also exceed the desirable limit of 200 mg/l in about 30% of the samples.

Trace Elements

Minor elements are less abundant and are often concentrated under certain conditions. In

fact, many physical and chemical processes may control amounts of these elements in

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

118

groundwater such as dispersion, complexation, acid-base reactions, oxidation-reduction,

precipitation-dissolution…(Fetter 2001).

The trace elements concentrations have a few extreme values. The concentrations of Fe3+,

Mn2+, Zn2+, Al3+, Pb2+, Cu2+ and Cr3+ were lower than the maximum permissible level

prescribed by WHO standards set for drinking water. The observed variations are not

explained by concurrent variations in TDS. The comparison of the hydrochemical data with

the WHO standards shows that 100 % of the samples exceeded the guide value for fluoride

ion (1.5 mg/l), which explains the existence of many cases of dental fluorosis in the south of

Tunisia. In fact, research on the relationship between fluorine concentration in drinking water

and endemic fluorosis has been conducted in many parts of the world (Guo et al., 2007).

In general, the main fluoride source in groundwater is related to the mineralogy of the

bedrock and their concentration dispersion values are due to different bedrock types.

Dissolution of Fluorite (CaF2) is a plausible source of fluoride ion in groundwater (Abu

Rukaha and Alsokhny 2004).

Hydrochemical Facies

Based on the major cation and anion, The chemical groundwater character of the study

area were represented by drawing piper trilinear diagram to identify and compare water types

(Piper, 1944). Calcium and sodium are mainly the dominant cations in Zeuss-Koutine

groundwater and among the anions, chloride and sulfate are dominant.

Two major groundwater groups can be distinguished in the piper diagram (Figure 3): the

Na–Cl group, which is characterising discharge area and the Ca–SO4–Cl group characterising

recharge area.

Figure 3. Piper trilinear diagram for the various cations and anions composition of the water samples.

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 119

Suitability for Drinking Purposes Using Water Quality Index (WQI)

Water Quality Index is a useful and efficient method for assessing the quality of water

and its suitability for drinking purposes. WQI is a mathematical instrument used to transform

large quantities for water quality data into a single number which represents the water quality

level (Saedi 2010, Chander 2011). The adopted methodology to develop a WQI involves

three steps. The first one is to assign weight (wi) to the sampling points based on the

concentration of the physico-chemical parameters and/or biological constituents of the water

and their relative importance in the quality of water for drinking purposes (Table 2) (Yiadana

2010).

Table 2. Weight and Relative weight of chemical parameters used for WQI computation

Chemical parameter (mg/l) WHO (2004) Weight (wi) Relative weight Wi

TDS 1000 5 0.172

Na+ 200 3 0.103

Ca2+ 200 3 0.103

Mg2+ 200 3 0.103

K+ 30 3 0.103

F- 1.5 5 0.172

Cl- 250 3 0.103

SO42- 400 3 0.103

HCO3- 380 1 0.034

The second step consists on the computation of the relative weight (Wi) which is

calculated from (Eq.1)

Wi = wi /∑ n i=1 wi (Eq.1)

where wi is the weight of each parameter,

n is the number of parameters

In the third step, a quality rating scale (qi) from each parameter is computed from (Eq.2)

qi = (Ci/Si) *100 (Eq.2)

where Ci is the concentration of each chemical parameter in each water sample in mg/l Si is a

world health organization’s standard for each of the major parameters in drinking water

(WHO), 2004.

The sub index SIi of ith parameter and the water quality index WQI were respectively

computed from Eqs. (3) and (4).

SIi = Wi * qi (Eq.3)

WQI = ∑ SIi

(Eq.4)

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

120

Depending on WQI values, water quality was than categorized into five consecutive

classifications (Table 3) (Sahu and Sikdar, 2008). In the current study, a total of nine

chemical parameters (Total Dissolved Solids (TDS), Na+, Cl-, Ca2+, Mg2+, SO42-, K+, HCO3-

and F-) of 14 water samples were used to calculate WQI, in order to assess a groundwater

suitability for domestic (drinking) use of the Zeuss Koutine aquifer.

Table 3. Water quality index (WQI) ranges

WQI range Water quality description

WQI<50 Excellent water

50 < WQI < 100.1 Good water

100 <WQI< 200.1 Poor water

100 <WQI< 200.1 Very poor water

100 <WQI<2 00.1 Water unsuitable for drinking purposes

Table 4. WQI and water description for location samples

WQI Water quality description

P1 303.41 Water unsuitable for drinking purposes

P2 446.52 Water unsuitable for drinking purposes

P3 358.80 Water unsuitable for drinking purposes

P4 298.82 Very poor water

P5 347.70 Water unsuitable for drinking purposes

P6 287.27 Very poor water

P7 287.54 Very poor water

P8 262.26 Very poor water

P9 148.61 Poor water

P10 139.21 Poor water

P11 139.24 Poor water

P12 146.08 Poor water

P13 324.63 Water unsuitable for drinking purposes

P14 206.73 Very poor water

The weights assigned to the parameters ranged between 1 and 5 (Table4), and were based

on the extent of health effects of those parameters. The highest weight (5) were assigned to

TDS and F- due to their major importance in water quality assessment and their health

implications when they have high concentration in water (Serinivasamoorthy et al., 2008;

Yiadana et al., 2010). Furthermore, Fluoride and TDS are the most sensitive water quality

parameters in the study area, and hence they need to be monitored regularly with higher

accuracy.

Based on the water quality index, in the analyzed samples of the study area, none fall

within the “Excellent” and “good” categories. 35.7%, 35.7% and 28.6% fall in “unsuitable for

drinking purposes”, “Very poor water “and “Poor water” respectively. Poor and very poor

water can’t be used for drinking water without any treatment and conventional disinfection,

whereas water “unsuitable for drinking purposes” could only be used for aquaculture,

irrigation and industrial purposes (Jindal and Sharma 2011).

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 121

Suitability for Irrigation

The water quality for irrigation purposes is assessed on the basis of its salinity level and

its sodicity (Kelly 1951). The longtime effects of irrigation water on soil physical properties

and crop productivity depend on the total salt, sodium, bicarbonate and carbonate

concentrations of the irrigation water, and also the soil’s initial physical properties (Yidana

2010).

In the study area, irrigation suitability of groundwater was evaluated based on sodium

percentage % Na, Sodium adsorption ratio SAR, Residual sodium carbonate RSC,

Magnesium percentage % Mg and permeability index PI.

SAR Sodium Adsorption Ratio

The sodicity hazard of water is generally described by the sodium adsorption ratio.

Indeed, there’s a significant relationship between SAR values of irrigation water and the

extent to which sodium is adsorbed by the soils (Subba 2006). If water used for irrigation has

high sodium concentrations and low calcium concentrations, the cation exchange complexe

may become saturated with sodium which can destroy the soil structure due to the dispersion

of clay particles (Todd, 1980).

SAR (Sodium Adsorption Ratio) is computed from Eq. (5) where concentrations are

reported in meq/l.

SAR = Na+/√ ((Ca2+ + Mg2+)/2) Eq. (5)

Excessive sodium concentrations in irrigation water can also result in sodium hazard,

particularly in dry climates such as south of Tunisia, causing reduced permeability. Indeed, in

arid climates, where plants tend to uptake more water than in cooler climates, more dissolved

solids are concentrated in the root zone as the soils dry up, resulting in salinity hazard (Shaki

2006). The SAR values in the study area range between 4.08 and 13.9 with mean of 7.87.

78.57% of groundwater samples have SAR<10 and 21.42% have SAR >10.

To ameliorate a water classification for irrigation suitability, the SAR has been plotted

with the salinity measurement on USSL diagram (Richards, 1954) (Figure 4).

As seen in the figure, 42.85% of water samples analyzed fall in C4-bS1 field indicating

extremely high salinity with low SAR, 35.71% in C4a-S1 (very high salinity with low SAR)

and 21.43% in C4b-S2 field (extremely high salinity with medium SAR).

Very high and extremely high salinity waters are unsuitable for irrigation with a restrict

drainage. An adequate drainage with low salinity waters and plants having good salt tolerance

should be selected.

Percent Sodium % Na

Sodium concentration plays an important role in evaluating the groundwater quality for

irrigation because sodium causes an increase in the hardness of soil as well as a reduction in

its permeability (Tijani 1994).

Excessive Na+ causes a dispersion of soil mineral particles and a decrease of water

penetration (Jalali 2007). Sodium percentage is calculated using the formula given below

(Eq.6), where all the ionic concentrations are expressed in milliequivalents per litre (meq/l).

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

122

% Na= (Na+ / Na+ + Ca++ + Mg++ + K+) *100 Eq. (6)

% Na of all samples ranged between 40 and 60 indicates that the groundwater is

permissible for irrigation purposes.

Figure 4. USSL diagram for the study area.

Residual Sodium Carbonate RSC

In water with high bicarbonates concentration, there is tendency of calcium and

magnesium to precipitate as carbonates and, consequently the water of the soils becomes

more concentrated. As a result, the cumulative concentration of carbonates and bicarbonates

react with sodium as sodium bicarbonate which influences the suitability of groundwater for

irrigation (Shaki 2006, Kumar et al., 2007).This is denoted as residual sodium carbonate

RSC, which is suggested by Eaton, 1950 and calculated as

RSC = (CO3-- + HCO3-) – (Ca++ + Mg++) Eq. (7)

Thus, estimation of residual sodium carbonate RSC is another way to examine irrigation

water. Computed RSC indicate that all samples have negative RSC showing that the

cumulative concentration of CO3-- and HCO3- is lower than the combined Ca++ and Mg++

concentrations which involve that there is no residual carbonate to react with sodium in the

soil. Based on RSC, all groundwater of the study area are suitable for irrigation.

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 123

Percent Magnesium % Mg

The cation exchange behavior of magnesium is similar to that of calcium. Both ions are

strongly adsorbed by clay minerals and other surfaces having exchange site (Hem, 1985).

Excess of magnesium in water will adversely affect the soil quality rendering it alkaline,

resulting in decreased and adversely affected crop yields (Ravikumar et al., 2011). An index

for calculating the magnesium ratio was developed by Paliwal (1972) and it’s computed by

the following equation (Eq.8).

% Mg = (Mg/ (Ca + Mg)) * 100 Eq. (8)

where all the ionic concentrations are expressed in milliequivalents per litre (meq/l).

Percent of magnesium % Mg of samples analyzed range between 29.8% and 58.43%. In

85.7% of samples, % Mg is less than 50% suggesting their suitability for irrigation, while

only 14.3% fall in the unsuitable class with % Mg more than 50% indicating their adverse

effect on crop yield.

Permeability Index

The permeability index is an important factor which influences quality of irrigation water

in relation to soil for development in agriculture (Srinivasamoorthy et al., 2011). The

permeability index PI of a water samples is computed from Eq. (9) where the concentration of

all ions is in meq/l.

PI = ((Na+ √HCO3-)/ (Na+ + Ca++ + Mg++)) *100 Eq. (9)

Permeability indices were plotted with the total ionic content of the groundwater samples

on a Doneen’s chart (Domenico and Schwartz 1990), which represent three different classes:

CI with best water type for irrigation, CII water generally acceptable and CIII waters

unacceptable. In the study area, the PI ranged from 52.44 to 64.09 with mean of 56.13 (Figure

5) indicating that all the samples are within CI best water type for irrigation purposes.

Finally; table 5 has been established to summarize several parameters throughout the

study period to evaluate suitability of Groundwater for irrigation purposes.

Multivariate Data Analysis

To carry out multivariate statistical techniques such as cluster analysis and principal

components analysis, the computer program ANDAD 6.00, performed by the Geo-Systems

Center of Instituto Superior Tecnico Portugal (CVRM, 2000), was used. Many

hydrogeochemical studies applied these techniques in an attempt to analyse and understand

chemical quality of groundwater data set. The multivariate statistical analysis is a quantitative

and independent technique of groundwater classification permitting the ranging of

groundwater samples and examining relationships between chemical parameters and

groundwater at the same time (Cloutier et al., 2008). For the multivariate analysis the data for

each parameter were standardized to a range of –1 to 1.

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

124

Figure 5. A Doneen’s chart for the study area.

Table 5. Suitability of Groundwater for irrigation based on various classifications

Parameters Range Class N° of samples Percentage of

samples

SAR <10 Low 11 78.57%

10-18 Medium 3 21.42%

18-26 High - -

> 26 Very high - -

% Na <20 Excellent

20-40 Good

40-60 Permissible 1-2-3-4-5-6-7-8-9-10-

11-12-13-14 100%

60 -80 Doubtful

> 80 Unsuitable

RSC < 1.25 Good 1-2-3-4-5-6-7-8-9-10-

11-12-13-14 100%

1.25 – 2.5 Doubtful

> 2.5 Unsuitable

% Mg <50 Suitable 1-2-3-4-5-6-7-8-9 85.7%

>50 Unsuitable 10-11 14.3%

<75 Soft

PI (%) 0 - 75 CI best water type for

irrigation

1-2-3-4-5-6-7-8-9-10-

11-12-13-14 100%

75 - 80 CII water generally

acceptable

80 - 120 CIII waters unacceptable.

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 125

Principal Component Analysis (PCA)

The PCA was carried out for data reduction in order to systematize the interpretation of

large sets of data and to characterize the linear correlations and loadings of the water quality

parameters. For application of PCA, we considered the data of the 14 water points. PCA was

applied to a matrix of 14 rows (wells) by 8 columns (variables). Table 6 presents the

eigenvalues and the cumulative eigenvalue of variance.

PCA results reveals show that the 2 first axes explain approximately more than ¾ of the

total variance in the data set, the first component accounted for about 66.55 % and the second

component accounted for about 14.70 %. These tow components together accounted for about

81.26 % of the total variance and the rest of the components only accounted for about 18.74

%. Thus, only the 2 factorial plans were retained for interpretation. Most signiﬁcant variables

in the components, represented by loadings higher than 0.6, are taken into consideration for

interpretation (Figure 6). The 1st factorial axis (F1) is interpreted by the relative projection of

the major ions in a specific group on the positive side, as a mineralization axis. The second

factor is best represented by HCO3-. We can note that this variable is not associated with any

other elements, showing an independent behavior regarding the other groups of variables.

Table 6. Results from the principal component analysis: eigenvalues, %total variance

and % cumulative variance (principal vectors are in bold)

Factors Eigenvalues %total variance %cumulative variance

1 6.655649 66.556488 66.556488

2 1.470906 14.70905 81.265541

3 1.050427 10.50427 91.769814

4 0.493581 4.935808 96.70562

5 0.18061 1.806096 98.511719

6 0.099812 0.998121 99.509842

7 0.032324 0.323242 99.833076

Cluster Analysis (CA)

Cluster analysis is a commonly used method for identifying and selecting the

homogeneous groups from the hydrogeochemical datasets. There are two types of cluster

analysis: R and Q-modes (Cruz and França 2006). The approach used in this study is based on

Q-mode so that similarities between different parameters could be revealed rather than

similarities between variables.

In our study, and in order to perform CA, an agglomerative hierarchical clustering was

developed using a combination of the Ward’s linkage method as a clustering algorithm and

Euclidean distances as a measure of similarity. The monitoring result of such analyses is a

graph, called dendrogram showing the degree of similarity between parameters (Hamzaoui et

al., 2011).

On the basis of chemico-physicals parameters and major ions concentrations, all 14

sampling sites were clustered into two homogenized water quality groups (Figure 7):

Cluster 1 (wells P1, P3, P 4, P13, P2, P7, P5 and P8) located in the downstream region.

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

126

Cluster 2 (Wells P6, P7, P9, P10, P11, P12, P13 and P14), these wells are nearer to the

upstream zone.

Figure 6. Representation of the parameters in the first factorial plane (axis 1 and 2).

Figure 7. Dendrogram of the Q-mode cluster analysis.

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 127

CONCLUSION

Zeuss–Koutine aquifer, located in southeastern Tunisia, is used intensively as a unique

water resource to meet the growing needs of the various sectors. The groundwater quality

monitoring in this region is significant. In this work and thanks to collected and measured

data, the groundwater quality in Zeuss–Koutine aquifer has been assessed for its use for

domestic and irrigation purposes. The salinity value ranges from 1520 mg/l to 5400 mg/l with

an average value of 3199 mg/l and obviously, the fresh water is rather encountered near the

recharge zones while downstream, water in saline. .

The assessment of water samples according to the permissible limits prescribed by WHO

for drinking purposes indicated that the values of the majority of parameters (Salinity, major

elements except potassium concentrations, Temperature, pH and Fluorine concentration)

exceed the WHO limits. Furthermore, the suitability of water for drinking purposes was

examined using WQI indicating that groundwater in this region is classified into“unsuitable

for drinking purposes”, “Very poor water “and “Poor water” and can’t be used for drinking

purposes without any treatment .

Due to very high to extremely high salinity hazard, all water samples of Zeuss Koutine

aquifer are unsuitable for irrigation with a restricted drainage and need rather an adequate

drainage, with low salinity waters and selected plants having a good salt tolerance..

Nevertheless, according to Doneen’s chart, all samples fall in within CI of best water type for

irrigation purposes.

According to the order of the cations and anions dominance, Zeuss-Koutine groundwater

is divided into two facies: a Na–Cl facies, which characterises discharge area and Ca–SO4–Cl

facies in recharge area.

Application of multivariate statistical analysis including Cluster Analysis and Principal

Components Analysis, on the 14 samples collected from Zeuss-Koutine aquifer confirms

results obtained by conventional geochemical methods.

ACKNOWLEDGMENTS

The authors gratefully thank the National Society of Drinking Water in Tunisia

(SONEDE), the Resources Water Direction of Tunis (DGRE) and the Regional Direction of

Agriculture and Water Resources of Medenine (Southeaster Tunisia). We would like to

acknowledge Mr Taher Atoui for his gently help during the field visits.

REFERENCES

Abbasi, SA. Environmental pollution and its control. Cogent. International. Philadelphia and

Pondicherry, 1999, 442 p.

Abu Rukaha, Y.; Alsokhny, K. Geochemical assessment of groundwater contamination with

special emphasis on fluoride concentration, North Jordan. Chemie der Erde. 2004. 64,

171-181.

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

128

Chilton ,J. Water Quality Assessments - A Guide to Use of Biota, Sediments and Water in

Environmental Monitoring - Second Edition. Edited by Deborah Chapman. Great Britain.

1992, 88 p.

Cloutier, V; Lefebvre, R; Therrien, R; and Savard, MM. Multivariate statistical analysisof

geochemicaldata as indicative of the hydro-geochemical evolution of groundwater in a

sedimentary rock aquifer system. Journal of Hydrology. 2008, 353, 294–313.

Cruz, JV.; França, Z. Hydrogeochemistry of thermal and mineral water springs of the Azores

archipelago (Portugal). Journal of Volcanology and Geothermal Research. 2006, 151(4),

382-398.

CVRM. Programa ANDAD. Manual do Utilizador: CVRM-Centro de Geosistemas, Instituto

Superior Tecnico, Lisbon, Portugal. 2000.

DGRE., a. Annuaires de l exploitation des nappes profondes en Tunisie. 2004, 200 p. DGRE.,

b. Annuaires piézométriques en Tunisie. 2004, 150 p.

Domenico, PA; and Schwartz FW. Physical and chemical hydrogeology.Wiley, New York,

1990, 410–420.

Eaton, AD.; Clesceri, LS.; and Greenberg, AE. Standard methods for the examination of

water and wastewater (19th ed.). Washington DC: American Public Health Association.

1995. 1325 p.

Elango, L; Kannan, R. and Kumar, S. Major ion chemistry and identification of

hydrogeochemical processes of groundwater in a part of Kancheepuram District, Tamil

Nadu, India. Journal Environmental Geosciences, 2003, 10(4), 157-166.

Fetter, CW. Applied Hydrogeology (4th ed.), Prentice-Hall, Upper Saddle River, New Jersey.

2001, 598p.

Gaubbi, E. Evolution de la piézométrie et de lagéochimie de la nappe de Zeuss–Koutine.

Master thesis. Tunis, Tunisia: University El Manar. 1988, p 63.

Geetha, A; Palanisamy PN; SIVAKUMAR, P; Ganesh kumar, P. and Sujatha M. Assessment

of Underground Water Contamination and Effect of Textile Effluents on Noyyal River

Basin In and Around Tiruppur Town, Tamilnadu. E-Journal of Chemistry, 2008, 5, 4,

696-705.

Guo, H., Wang, Y. Geochemical characteristics of shallow groundwater in Datong basin,

northwestern China. Journal of Geochemical Exploration. 2005, 87, 109–120.

Hamzaoui-Azaza, F.; Bouhlila, R.; and Gueddari, M. Geochemistry of fluoride and major ion

in the groundwater samples of Triassic aquifer (south eastern Tunisia), through

multivariate and hydrochemical techniques: Journal Applied Sciences Research. 2009, 5,

1941–1951.

Hamzaoui-Azaza, F., 2011, Géochimie et Modélisation des Nappes de Zeuss-Koutine, des

Gre `s du Trias et du Mioce `ne de Jorf-Jerba-Zarzis: Unpublished Thesis, Department of

Geology, University of Tunis El-Manar, Faculty of Science of Tunis; 2011262 p.

Hamzaoui Azaza, F ; Ketata, M; Bouhlila, R. ; Gueddari,M.; and Riberio, L.

Hydrogeochemical characteristics and evaluation of drinking water quality in Zeuss-

Koutine aquifer, south-eastern Tunisia. Environmental Monitoring and Assessment, 2011,

174, -283-298.

Hem, JD. Study and interpretation of the chemical characteristics of natural water: United

States; Geological Survery Supply paper. 1985, 263 p.

Jalali, M. Assessment of the chemical components of Famenin groundwater, western Iran:

Environmental Geochemal Health. 2007, 29, 357–374. no access.

Nova Science Publishers, Inc.

Suitability of Groundwater of Zeuss-Koutine Aquifer (Southern of Tunisia) … 129

Jalali, M. Groundwater geochemistry in the Alisadr, Hamadan, western Iran. Environmental

Monitoring and Assessment. 2010, 166, (1-4), 359-369.

Jeevanandam, M.; Kannan, R.; Srinivasalu, S; and Rammohan, V. Hydrogeochemistry and

Groundwater Quality Assessment of Lower Part of the Ponnaiyar River Basin, Cuddalore

District, South India. Environmental Monitoring and Assessment. 2007, 132 (1-3),

263-274.

Krishna Kumar, S.; Rammohan, V.; Dajkumar Sahayam J.; and Jeevanandam M. Assessment

of groundwater quality and hydrogeochemistry of Manimuktha River basin, Tamil Nadu,

India. Environmental Monitoring and Assessment. 2009,159 (1-4), 341-351.

Kumar, M.; Kumari, K.; Ramanathan, AL.; and Saxena R. A comparative evaluation of

groundwater suitability for irrigation and drinking purposes in two intensively cultivated

districts of Punjab, India. Environmental Geology. 2007, 53,553–574.

Mathess, G. The properties of groundwater. Wiley, New York; 1982.

Meybeck, M. Global chemical weathering of surfıcial rocks estimated from river dissolved

loads. American Journal of Science. 1987, 287, 401–428.

Morhange, C.; and Pirazzoli, PA. Mid-Holocene emergence of Southern Tunisian coasts.

Marine Geology. 2005, 220, 205–213.

Nwankwoala, HO.;and Udom, GJ. Hydrochemical Facies and Ionic Ratios of Groundwater in

Port Harcourt, Southern Nigeria. Research Journal of Chemical Sciences. 2011, 1(3), 87-

101.

OSS : Observatoire du Sahara et du Sahel.Etude Hydrogéologique du Système Aquifère de la

Djeffara Tuniso-Libyenne: Rapport de synthèse, Tunisia. 2005, 209 p.

Ouessar, M. Hydrological Impacts of Rainwater Harvesting in Wadi Oum Zessar Watershed

(Southern Tunisia): Unpublished Thesis, Faculty of Bioscience Engineering, Ghent

University, Ghent, Belgium. 2007, 154.

Paliwal, KV. Irrigation with saline water. Monogram, New Delhi: IARI. 1972, 198p.

Piper, AM. A graphic procedure in the geochemical interpretation of water analysis.

American Geophysical Union, Tranc. 1944, 914-923.

Ramakrishnaiah C R; Sadashivaiah C. and Ranganna G. Assessment of Water Quality Index

for the Groundwater in Tumkur Taluk, Karnataka State, India. E-Journal of Chemistry,

2009, 6(2), 523-530.

Ravikumar, P.; Somashekar, RK.; and Angami, M. Hydrochemistry and evaluation of

groundwater suitability for irrigation and drinking purposes in the Markandeya River

basin, Belgaum District, Karnataka State, India. Environmental Monitoring Assessement .

2011, 173, 459–487.

Reghunath, R; Sreedhara Murthy, TR. and Raghavan, BR. The utility of multivariate

statistical techniques in hydrogeochemical studies:an example from Karnataka, India.

Water Research. 2002, 36, 2437–2442.

Rodier, J. L’Analyse de l’Eau: Eaux Naturelles, Eaux Résiduaires, Eau de Mer, 8th ed.:

DUNOD, Paris. 1996,1384 p.

Rosen, MR. and Jones, S. Controls on the groundwater composition of the Wanaka and

Wakatipu basins, Central Otago, New Zealand. Hydrogeology Journal. 1998, 6, 264-281.

Saeedi, M; Sharifi, OA; and Meraji, H. Development of groundwater quality index.

Environmental Monitoring Assessement. 2010, 163, 327–335.

Sahu, P.; and Sikdar, PK. Hydrochemical framework of the aquifer in and around East

Kolkata wetlands, West Bengal India. Environmenetal Geology. 2008, 55, 823–835.

Nova Science Publishers, Inc.

Fadoua Hamzaoui-Azaza, Besma Tlili-Zrelli, Rachida Bouhlila et al.

130

Service Quality. 2005, 15(2), 195-208.

Shaki, AA; and Adeloye, A.J. Evaluation of quantity and quality of irrigation water at

Gadowa irrigation project in Murzuq basin, southwest Libya. Agricultural water

management. 2006, 84, 193 – 201.

Srinivasamoorthy, K.; Chidambaram, M.; Prasanna, M.V.; Vasanthavigar, M.; John Peter, A.;

and Anandhan, P. Identification of major sources controlling Groundwater Chemistry

from a hard rock terrain A case study from Mettur taluk, Salem district, Tamilnadu, India.

Journal of Earth System Sciences. 2008, 117(1), 49–58.

Subramani, T.; Elango, L.; and Damodarasamy, SR. Groundwater quality and its suitability

for drinking and agricultural use in Chithar River Basin, Tamil Nadu, India.

Environmental Geology. 2005, 47, 1099-1110.

Tijani, MN. Hydrochemical assessment of groundwater in Moro area, Kwara State, Nigeria.

Environmental Geology. 1994, l24, 194–202.

Venugopal, T; Giridharan, L. and Jayaprakash, M. Groundwater Quality Assessment Using

Chemometric Analysis in the Adyar River, South India. Arch Environ. Contam

Toxicol.2008, 55, 180–190.

World Health Organization (WHO). World Health Organization Guidelines for Drinking

Water Quality; Volume 1: Recommendations, 3rd ed.: World Health Organization,

Geneva, Switzerland. 2004, 188 p.

Yidana, SM; Bruce Banoeng, Y.; and Akabzaa, TM. Analysis of groundwater quality using

multivariate and spatial analyses in the Keta basin, Ghana. Journal of African Earth

Sciences. 2010, 58, 220–234.

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 6

APPLICATION OF WATER QUALITY

INDICES (WQI) AND STABLE ISOTOPES

(18O AND 2H) FOR GROUNDWATER

QUALITY ASSESSMENT OF THE DENSU

RIVER BASIN OF GHANA

Abass Gibrilla1, Edward Bam1, Dickson Adomako1,

Samuel Ganyaglo1 and Hadisu Alhassan2

1Nuclear chemistry and Environmental Research Centre, National Nuclear Research

Institute, Ghana Atomic Energy Commission, Kwabenya-Accra, Ghana

2Ghana Urban Water Company Ltd, Weija

ABSTRACT

Groundwater and surface water (Densu River) were collected for physical, chemical

and stable isotope analysis to determine their suitability for drinking and agricultural

purposes. Rain water was also sampled on event basis at Koforidua for stable isotope

analysis. The results showed that, groundwater in the study area are generally fresh and

slightly acidic to neutral while the surface water is slightly alkaline. The WQI values

were found ranging from 0-50 belonging to “excellent” and “good” water quality. An

integrated approach of heavy metal evaluation indices using Contamination index (Cd),

heavy metal pollution index (HPI) and heavy metal evaluation index (HEI) were used to

evaluate the extent of pollution and suitability of the samples for drinking with respect to

heavy metals. The three indices showed similar trends with strong correlations but with

different water quality classifications. Whereas the Cd showed that 95% of groundwater

and 100% of surface water were highly polluted, HPI and HEI indicated that, 4.76% and

0% of groundwater and 25% and 0% of the surface water were polluted. A modification

of Cd and HPI using multiple of mean criteria showed a comparable classification with

HEI. Cd, HPI and HEI showed that 4.8%, 0% and 0% of the groundwater were highly

polluted while 25%, 0% and 0% of the surface water were respectively classified as

highly polluted. Chemical indices like percentage of sodium, Sodium adsorption ratio,

residual sodium carbonate, and permeability index (PI) indicate that, the groundwaters in

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

132

the study area are suitable for irrigation. A comparison of the isotopic data of the rain

water, Local Meteoric Water Line (LMWL) and Global Meteoric Water Line ( GMWL)

indicates that the groundwater in the study area is mainly meteoric with few groundwater

and all the surface water showing an evidence of evaporation. The d-excess values show

that the groundwater has undergone dilution with the rainfall and this is observed from

the decrease of the d-excess of the groundwater with increase in Oxygen-18. This

observation also suggests a modern day recharge to the groundwater.

INTRODUCTION

In Ghana, groundwater is the major source of potable water for most rural communities.

Groundwater is a valuable natural resource; it occurs almost in all geological formations

beneath the earth surface not in a single widespread aquifer but in thousands of local aquifer

systems with similar characteristics [1]. The presence of dissolved minerals coupled with

some special characteristics of groundwater as compared to surface water makes it a preferred

choice for many purposes [2], [3]. In many rural communities in the Densu River basin,

groundwater is the major source of water for domestic and other uses. This is partly due to

pollution of the Densu River and its tributaries. In recent years, these communities are

experiencing rapid growth due to urbanization. The potential threat to groundwater resources

in the area due to over exploitation, agriculture and improper waste disposal practices are

envisaged [4]. Knowledge of the occurrence, quality and recovery of the groundwater

resources is, therefore, essential for proper implementation of integrated water resources

management programme [5]. The quality and suitability of groundwater for domestic,

industrial and agricultural purposes depends on the quality of recharge water, atmospheric

precipitation, in-land surface water, and on sub surface geochemistry. Because of the

potential of absorption and transportation of waste materials (domestic, industrial and

agriculture), river basins are highly vulnerable to pollution, hence the need for regular

monitoring and control of water quality in these areas [6].

Interest in chemical and trace metal pollution has generated a desire both on the national

and international scales, for integrating numerous parameters associated with water quality in

a specific index, hence, the development of several water quality indices and metal pollution

indices [7], [8], [9]. Some of these pollution indices had been successfully used by [5], [10],

[11], [12] to assess the quality of surface and groundwater with respect to chemical and heavy

metal. These pollution indices are intended to provide a useful and comprehensible guiding

tool for water quality executives, environmentalist, decision makers and potential users of a

given water system [11]. Stable isotopes of 18O and 2H have also been widely used in water

resources management. This is because, they allow conclusion to be drawn as regards the

recharge process, the location of recharge and discharge areas, aquifer continuity, sources of

ions in water and turnover time [13], [14]. Groundwater hydrochemistry in the Densu river

basin has been fairly studied by various authors [15], [16], [17], [18] and their properties are

well known to the extent that groundwater is tapped in commercial quantities to meet both

domestic and industrial needs of the people. All these authors generally appear to attribute the

groundwater hydrochemistry to rock-water interaction.

Later studies by [16],[19] using stable isotopes of 18O and 2H showed an evidence of

evaporated waters recharging the groundwater system in some areas; this implied that

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 133

anthropogenic activities on the surface and the unsaturated zone [4] may pose a serious

challenge to the groundwater quality in the near future. In this chapter, Water Quality Index

(WQI), percentage of sodium (Na%), Sodium Adsorption Ratio (SAR), Residual Sodium

Carbonate (RSC), permeability index (PI), contamination index (Cd), heavy metal pollution

index (HPI) and heavy metal evaluation index (HEW) which are regarded as one of the most

effective ways to communicate water quality [20] will be used. The objective of this chapter,

therefore, is to study the suitability of the groundwater and the surface water for drinking and

irrigation purposes using the data obtained through quantitative analysis and [21] water

quality standards. The study will also employ stable isotopes as complementary tool to study

the origin of groundwater in the study area.

METHOLOGY

Study Area

The Densu river basin lies between latitude 5o 30’ N to 6o 20’ N and longitude 0o 10’ W

to 0o 35’ W (Figure 1). The river shares its catchment boundary with the Odaw and Volta

basins to the east and north respectively, the Birim basin in the northwest and the Ayensu and

Okrudu in the west.

Figure 1. Geological map of the study area.

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

134

The Densu River takes its source from Atewa-Atwiredu mountain range near Kibi (East

Akyem District) in the Eastern Region of Ghana. The river is about 116 km long with a

catchment area of 2564 km2 covering nine administrative districts.

The main tributaries include rivers Adeiso, Nsakyi, Dobro, and Kuia (Figure 1). The

Densu River enters the Weija reservoir, one of the two main sources of water supply for the

city of Accra and finally discharges into Sakumo lagoon and Gulf of Guinea near Bortianor

west of Accra. Most communities upstream and midstream of the river depend on

groundwater and to a lesser extent the raw water without any form of treatment.

The year 2000 population and housing census estimated the total population of the Densu

River basin to be about 1.2 million people [22]. The economic activities in the catchment are

mainly cultivation of crops such as cocoa, maize, cassava, vegetables, pineapples and

cocoyam with few livestock and fish farming. Artisanal mining popularly called “galamsey”

in some of the major rivers is fast becoming a lucrative business in the study area.

Climate and Geology

The basin falls under two distinct climatic zones characterized by two rainfall regimes

with different intensities [23]. The major rainy season extends from April/May to July. The

minor season occurs between September and November.

The mean annual rainfall recorded for 10 years during the period 1993 to 2003 obtained

from the Ghana Meteorological Agency (GMA) varies from about 1200 mm at Nsawam to

about 1487 mm in the river source area at Kibi. The mean annual temperature is about 27 oC,

with March/April being the hottest (32oC) and August being the coldest month (23oC).

Maximum and Minimum monthly temperature, normal rainfall distributions and relative

humidity are shown in Figure 2.

Figure 2. Distribution of rainfall, temperature and relative humidity in the study area, adapted from

[25].

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 135

The middle to northern portion under study is mainly underlain by Precambrian

crystalline rocks, comprising of Birimian (upper and lower) and Cape Coast Granites while

the southern portion is underlain by the Togo series. Detailed description of the geology and

hydrogeology can be found in our previous work [19]. The dominant soils are ochrosols, with

patches of gleisols and lithosols.

Data and Field Work

The principal dataset presented in paper includes the physical, chemical, isotopic and

trace metals analyses collected by [24] and also from [15] and [25].

Rainwater samples were collected from the study area on an event basis (2006–2008) at

Kibi meteorological stations which is about 320 m a.s.l. for isotope analysis. The rain water

samples were collected by a 500ml vial through a 200mm diameter funnel with a pingpong

ball to avoid evaporation. The rain water samples were then collected into 60ml air-tight

polyethylene vials soon after the rain event for the analysis.

A total of twenty one boreholes, one hand-dug well and four Surface water points (River

Densu) were sampled at various locations. Figure 1. All the water samples were collected in

500ml pre-conditioned high density polyethylene bottles.

They were first conditioned by washing with five (5%) percent nitric acid, and then

rinsing several times with distilled water. This was carried out to ensure that the sampling

bottles were free from contaminants. Samples for isotopes analysis were collected in 60ml

glass bottle filled to the brim and securely capped.

At the sampling points, the boreholes were pumped to purge the aquifer of stagnant water

to acquire fresh samples for analysis. Most of the wells were being used for domestic water

supply during the sampling period; therefore, purging lasted for 5-10 minutes.

pH, temperature, electrical conductivity (EC) and total dissolve solids (TDS)

measurements were conducted in situ in the field using a pre calibrated HACH pH meter and

HACH conductivity meter. Alkalinity titration was done at the wellhead using a HACH

digital titrator. For chemical and trace metal analysis, the samples were filtered on site

through 0.45µm cellulose filters with the aid of a hand operated vacuum pump and collected

in the 500ml bottle. The sample for metal analysis was preserved by adjusting to pH<2 with

6N ultrapure nitric acid [26]. The bottles and caps meant for collecting the samples were

rinsed three times with the filtered water after which they were filled to the brim and caped.

All the samples were then kept in an ice chest containing ice bricks and transported to the

laboratory.

Laboratory Analysis

Na+ and K+ were analyzed using flame photometer, magnesium (Mg2+) and calcium

(Ca2+) using Fast Sequential Atomic Absorption Spectroscopy (Varian AA240FS). Chloride

(Cl-), Sulphate (SO42-) and Nitrate (NO3-) were analyzed using ion chromatography system

(Dionex ICS-90) at Nuclear Chemistry and Environmental Research Centre, NNRI, GAEC.

The ionic-balance-error was computed, taking the relationship between the total cation (Ca2+,

Mg2+, Na+ and K+) and the total anions (HCO3-, Cl-, SO42-) for each set of complete analysis

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

136

of water samples. Only samples which fall within ±5% were reported in this work. The

acidified samples were used to determine the trace metals concentrations using Instrumental

Neutron Activation Analysis (INAA) using Ghana research reactor 1 (GHARR 1) at Ghana

atomic Energy Commission (GAEC). The stable isotope analysis of the samples was carried

out at HMGU, Institute of Groundwater Ecology (Neuherberg/Germany) using Isotope mass

spectrometry.

The variation in isotope ratio D/H and 18O/16O in water samples are expressed in terms of

per mille deviation (‰) relative to internal standards that were calibrated using the Vienna-

Standard Mean Ocean Water (V-SMOW). The data was then normalized following [26] as

follows

‰  1000*1⎥

⎦

⎤

⎢

⎣

⎡−

−SMOWV

sample

where

R sample represent either the 18O/16O or the D/H ratio of the samples.

RV-SMOW represents either the 18O/16O or the D/H ratio of the V-SMOW.

The analytical reproducibility was 0.1‰ for oxygen and 1‰ for deuterium.

Sample Preparation

0.50ml of each water sample (weighing 0.50g) was pipette using a calibrated Eppendorf

tip ejector pipette (Brinkmann Ins., Inc., Westbury, NY) into clean pre-weighed 1.5ml

polyethylene vials, and fitted with polyethylene snap caps and heat-sealed. Four of these

sample vials were placed into a 7.0ml volume polyethylene vial and heat-sealed (for medium

lived radionuclide) for irradiations. Two replicates were prepared for each sample. However,

for short lived radionuclide, only one sample was put into the 7.0ml vial. The elemental

comparator standard used in this work was synthetic standards prepared by pipeting aliquots

of multielement and single elements NIST standard solution and validated against IAEA

Standard Reference Material (SRM) 1547 Peach leaves prepared as the samples.

Sample Irradiation, Counting and Analysis

All the samples, synthetic standards prepared by pipeting aliquots of multi-element and

single elements standard solution and standard reference materials were irradiated in the inner

pneumatic irradiation sites of the Ghana Research Reactor-1 (GHARR-1) facility operating at

half full power of 15 kW with corresponding thermal neutron flux of 5.0 x 1011 n/cm2s1. The

scheme for irradiation and counting was chosen according to the half lives of the elements of

interest Table 1. The detector used in this work was an n-type high purity germanium (HPGe)

detector Model GR 2518 (Canberra Industries Inc.) with a resolution of 0.85 keV (FWHM)

and 1.8 keV (FWHM) for 60Co gamma-ray energies of 1332 keV.

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 137

Table 1. Nuclear data and irradiation scheme of elements determined in this work,

adapted from [25]

Element Target

isotopes

Formed

isotope

Half-life Gamma ray

energy Kev

Irradiation

time (ti)

Cooling

time (td)

Counting

time (tc)

Cu 65Cu 66Cu 5.10 min 1039.2 120 sec 60-300 sec 600 sec

Al 27Al 28Al 2.24 min 1778.9 120 sec 60-300 sec 600 sec

Mn 55Mn 56Mn 2.58 h 846.8 3600 sec 60-300 sec 600 sec

As 75As 76As 26.32 h 559.1 14400 sec 2 days 600 sec

Zn 64Zn 65Zn 243.9 days 1115.6 14400 sec 3-30 days 12-24 h

Fe 58Fe 59Fe 44.49 days 1099.2 14400 sec 3-30 days 12-24 h

Cr 50Cr 51Cr 27.7 days 320.0 14400 sec 3-30 days 12-24 h

Table 2. Validation result for elemental concentration of NIST-1547 Peach Leaves

Certified Reference Material (CRM: mg/kg),

adapted from [25]

Element NIST 1547 Peach leaves

This work Recommended value %RSD Z-Score

Al 252±6 249±8 2.38 0.38

Cu 3.9±0.2 3.7±0.4 5.13 0.50

Cr 1.01±0.01 1.02±0.03 0.99 -0.03

Zn 17.7±0.3 17.9±0.4 1.69 -0.50

Fe 210±10 218±14 4.76 0.57

Mn 97.6±7 98±8 7.17 -0.05

As 0.06±0.01 0.05±0.01 16.67 1.00

Concentrations are reported as mean ± SD.

100*)/(% xRSD

()

−

=− x

scoreZ ,

where

x= measured mean value.

µ= recommended value.

σ= standard deviation of the recommended value.

The detector operated on a bias voltage of (-ve) 3000 V with relative efficiency of 25% to

NaI detector. A Microsoft Soft window based software MAESTRO was used for the spectra

analysis, employing the relative standardizing (comparator) method.

The analytical results of the present studies were validated using NIST-1547 Peach

Leaves Certified Reference Material as shown in Table 2 where mean elemental

concentrations and standard deviation, certified/ recommended value and standard deviation,

percentage RSD and Z-score were tabulated. The high precision in the data were suggested by

the low RSDs (in %) which were < 10% for all the elements except As where RSDs is

16.67%. All the Z-score values were below 3, suggesting that the data are within 95%

confidence limit. Hence, the elemental concentration data for samples analyzed in this study

are reliable within ±10%. The detection limits of the technique for the determined elements

were calculated under identical experimental conditions and are: Al=1.5mg/l, Cr=0.001mg/l,

Cu=0.5mg/l and Zn=2mg/l for groundwater and surface water.

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

138

Estimation of the Water Quality Index (WQI)

Water Quality Index (WQI) is a very reliable, useful and efficient method for assessing

and communicating the information on the overall quality of water [28], [29]. The

determination of WQI helps in deciding the suitability of groundwater sources for its intended

purpose. From the early 1960s, different WQI have been developed [30], [31]. This work will

employ the use of WQI proposed by [9] in assessing the suitability of the water in the study

area for drinking.

[

]

∑=

nqWAntiWQI 101 loglog

where

Wn = Weightage factor and calculated from the following equations

)( −

=in SKW

K proportionality constant derived from

()

−

−⎥

⎦

⎤

⎢

⎣

⎡

=∑

Si are the [21], [32] standards values of the water quality parameter. The calculated

Weightage factors of each parameter are given in Table 3.

Table 3. Water Quality Parameters, their standard values, their ideal values and the

assigned weightage factors, adapted from [25]

Parameter Standard Value, Si Ideal value, Cid 1/Si Assigned Weightage factor, Wi

pH 8.5 7 0.1176 0.00396

Total dissolve solids 500 0 0.0020 6.73E-05

Alkalinity 120 0 0.0083 0.000281

Electrical conductivity 1400 0 0.0007 2.4E-05

Chloride 250 0 0.0040 0.000135

Sulphate 250 0 0.0040 0.000135

Sodium 200 0 0.0050 0.000169

Nitrate 50 0 0.0200 0.000674

Calcium 75 0 0.0133 0.000448

Magnesium 30 0 0.0333 0.001122

Mn 0.5 0 2.0000 0.067400

Fe 0.3 0 3.3333 0.112202

Cu 2.0 0 0.5000 0.01683

Cr 0.05 0 20.000 0.673211

Zn 3.0 0 0.3333 0.01122

Al 0.3 0 3.3333 0.112202

∑ =29.71 ∑= 1

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 139

Table 4. Water Quality Index Scale, adapted from [25]

Water quality Description

0-25 Excellent

26-50 Good

51-75 Poor

76-100 Very poor

>100 Unfit for drinking(UFD)

Quality rating (qn) was calculated using the formula

()

100*

tan

⎥

⎦

⎤

⎢

⎣

⎡

−

idealdards

idealactual

nVV

where,

qn = Quality rating of ith parameter for a total of n water quality parameters

Vactual = Value of the water quality parameter obtained from laboratory analysis

Videal = Value of that water quality parameter can be obtained from the standard tables,

Videal for pH = 7 and for other parameters it is equivalent to zero.

Vstandard = WHO, 2004 / ISI, 1993 standard of the water quality parameter

The calculated WQI values are then used rate the ground water quality as excellent, good,

poor, very poor and unfit for human consumption (Table 4).

Heavy Metals Indexing Approach

Three standard methods for heavy metal contamination evaluated in this study are

Contamination index (Cd) developed by [33], Heavy Metal pollution Index (HPI) proposed by

[34] and Heavy Metal evaluation Index (HEI) proposed by [35].

Contamination Index (Cd)

This method evaluates the quality of water by determining the degree of contamination.

The Cd is the sum of the contamination factors of individual components exceeding the upper

permissible limit for each water sample analysed. For this reason, the Cd can be used to give a

summary of the combined effects of heavy metals in drinking water. The Cd is calculated

using the relation

∑

fid CC

where

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

140

1−

⎥

⎦

⎤

⎢

⎣

⎡

fi C

Cfi = contamination factor for the i-th components

CAi = measured value for the i-th components

CNi = upper permissible limit of the i-th component (N denoting the ‘normative value’).

Even though, the original authors did not consider the measured values below the upper

permissible concentration value, this study will use all the measured parameters irrespective

of their value for the sake comparison. The upper permissible concentration values CNi was

taken as the maximum admissible concentration (MAC) and was obtained from [21]

guidelines for drinking water quality.

The waters were then classified based on their Cd values into three groups as follows: Cd

<1 as low, Cd =1-3 as medium and Cd >3 as high.

Heavy Metal Pollution Index (HPI)

HPI signifies the overall quality of water with respect to heavy metals. The HPI is based

on weighted arithmetic mean quality method following two steps. The first step involves

establishing a rating scale to weigh each selected parameter. The Second step involves the

selection of the pollution parameters on which the index is to be based. The rating system is

an arbitrarily value between zero to one. This is defined by the perceived importance of the

individual quality parameter under considerations in a comparative way or it can be assessed

by making values inversely proportional to the recommended standard for the corresponding

parameter [31], [34].

In this study, the weightage factor (Wi) for each parameter was defined as the inverse of

the recommended standards denoted as (Si) as suggested by [36]. The concentration limits

(highest permissible value for drinking water Si and maximum desirable value Ii) for each

parameter were taken from [21] and [32]. The Si refers to maximum allowable concentration

in groundwater in the absence of alternative source. The HPI was calculated using the model

[34] and is given by

∑

HPI

where

Qis the sub-index of the i-th parameter.

W is the unit weightage of the i-th parameter

n is the number of parameters measured

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 141

The sub-index ( i

Q) of the parameters was calculated by

100*

∑

=−

−

iii

iIS

where

Mis the measured value of the heavy metal of the i-th parameter

I is the ideal value of the i-th parameter

S is the standard value of the ith parameter

Generally, pollution indices are determined to assess the use of water for specific use; the

calculated indices will be used for the purpose of drinking water. The critical HPI value for

drinking water is 100, above which the water is unsuitable for drinking.

Heavy Metal Evaluation Index (HEI)

The HEI is also a useful method that gives an overall quality of water with respect to

heavy metal.

∑=

mac

HEI 1

where

His the measured value of the i-th parameter

mac

His the maximum admissible concentration of the i-th parameter.

RESULTS AND DISCUSSIONS

The statistical summary of the physico-chemical parameters and trace metals measured in

the groundwater and surface waters are presented in Table 5. The pH of the groundwater

ranges from 5.58-7.09. The lowest pH occurred at ANY while the highest pH was found at

DPOT. The mean pH was 6.47. The very acidic groundwater were found in the Birimian

ANY, AK, NK, POT (5.58-6.66) while in the granite where most of the samples were located,

the pH are slightly acidic to neutral (5.98-7.09). This deviation can be attributed to the

activities occurring in the unsaturated zone which might have effect in the groundwater

before recharge since these areas are characterized by intensive agricultural activities.

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

142

Table 5. Statistical summaries of the physical parameters and trace metal

in the study area

Groundwater Surface water

Parameter Min Max Mean SD Min Max Mean SD

pH 5.5800 7.0900 6.4710 0.3920 7.3800 7.4800 7.4450 0.0451

Temp 26.1000 28.6000 26.9333 0.6843 26.1000 26.7000 26.3750 0.2500

TDS 49.5000 361.0000 168.9905 85.3246 50.9000 96.0000 68.7750 19.9075

COND 98.8000 722.0000 337.9619 171.2090 101.9000 191.2000 137.3500 39.4394

Sal 0.0000 0.4000 0.1524 0.0981 0.0000 0.1000 0.0500 0.0577

Alkalinity 11.2600 193.2100 85.2390 53.6163 19.4400 85.2900 53.1150 31.2376

Na+ 30.8000 226.7000 100.4238 65.8449 24.9000 66.5000 39.4800 19.0500

K+ 0.8000 52.9000 10.1429 11.4894 1.9000 9.9000 5.6300 4.1700

Cl- 3.4000 124.5000 36.9619 31.1747 4.9000 11.3000 9.1500 2.9500

HCO3- 12.4000 233.6000 103.6862 65.4912 23.4600 102.8000 63.9700 37.2100

Mg2+ 0.3700 68.0000 7.7814 14.1066 2.0500 3.7600 2.8900 0.7000

Ca2+ 2.8400 99.6000 19.9014 31.6991 2.1800 4.5400 3.4400 0.9700

SO42- 4.9500 26.6800 17.8181 5.4075 9.3800 14.3400 11.9300 2.0400

NO3- 0.1790 64.4600 23.3362 19.1288 2.8900 7.1600 4.4800 1.9100

PO43- 0.0200 26.1300 1.3871 5.6700 0.0200 0.1700 0.1000 0.0600

Cu 0.0100 9.5700 4.1340 3.2579 3.6200 9.7800 7.0150 2.6406

Al 0.9320 2.6680 1.5854 0.4840 1.1680 3.2040 2.0115 0.9893

Mn 0.0025 2.2775 1.1496 0.7039 0.1950 0.9850 0.4463 0.3650

Cr 0.0010 0.0030 0.0013 0.0006 0.0010 0.0020 0.0014 0.0005

As 0.0010 0.0200 0.0099 0.0076 0.0010 0.0200 0.0103 0.0078

Fe 0.1870 0.6860 0.3206 0.1230 0.2890 0.4480 0.3403 0.0729

Zn 2.0000 3.2000 2.1619 0.3584 2.0000 9.1000 3.9000 3.4708

All units are in mg/l except pH in pH units, COND in µs/cm, Temp in oC, Sal in ppt.

All the surface waters D POT, D AKD, D MAN and NS were found to have pH values

very close (7.38-7.48). The WHO recommended limit for potable water is 6.5-8.5. This

implies that about 42.30% of the samples fall outside the recommended range while 58.7%

fall within the range.

Electrical conductivity (EC) values are generally low. Minimum and maximum values

are 98.8µs/cm-1 and 722µs/cm-1 respectively, with the mean value of 337.96. Total dissolve

solid (TDS) ranged from 49.5-361mg/l. According to TDS classification by [37], all the

groundwater’s are fresh (TDS<1000). The surface water EC and TDS range from 101.9 to

191.2 and 50.9 to 96 mg/l with mean values of 137.35 and 68.78 respectively. Major cations

(Ca2+, Mg2+, Na+ and K+) in both groundwater and surface water were also generally low with

Na+ being the most dominant cation. HCO3- is also the most dominant anion with values

ranging from 12.4 to 233.6 mg/l for groundwater and 23.46 to 102.8 for the surface water.

The nitrate in the groundwater varied from 0.18 - 64.46mg/l with an average of 23.34

mg/l. Eventhough it has been observed that igneous rocks contain small amounts of nitrate

[38], most nitrate in water comes from fertilizers, nitrification by leguminous plants and

animal excreta. Industrial and domestic sources can also contribute to higher elevation of

nitrate in water. Nitrogen is an essential component of protein hence occurs in all living

organisms. When these materials decay through microbial activities, the complex protein

changes through amino acid to ammonia, nitrite and finally nitrate. The nitrate produce may

then leach to the groundwater.

Fears have been expressed that nitrate contaminated water supplies carries the risk of

methaemoglobinaemia (blue-baby syndrome) and stomach cancer. The main pollution risk for

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 143

the aquifers is vertical infiltration of precipitation and flushing of pollutants from the soil.

About 57.7% of the samples have nitrate concentrations above the recommended value of

10mg/l by WHO.

There was no general trend in the different geological formations; nitrate from the rocks

is therefore not likely to be the source of the nitrate, but since the area is a forest zone

dominated by agriculture activity and also heavily populated, higher nitrate concentrations

can be attributed to decay of organic matter, nitrogen fixation, fertilizer applications and

sewage.The surface water nitrate was also observed to be almost uniform with little variation

from the upper to the lower portion of the study area. This can also be attributed to run-offs,

sewage and industrial effluents being discharge into the river.

SO42- and PO43- were observed to have generally low concentration in both the surface

and the groundwater in the study area.

Water Quality Index (WQI)

The chemistry of groundwater has been utilized as a measure to outlook the quality of

water for drinking and other purposes [5], [12]. Tiwari and Mishra [9] specifically used WQI

to determine the suitability of groundwater for drinking purpose. A location wise calculated

WQI values for the different geological formations and the surface water were presented in

Table 6.

Table 6. Results of the calculated WQI of the sampling points, adapted from [25]

Location CODE Geological type WQI Quality

Nkroso NK Birimian 27.94 Good

Potroase1 POT 1 Birimian 11.14 Excellent

Potroase Yawofori POT YOF Birimian 17.48 Excellent

Anyinase ANY Birimian 15.00 Excellent

Akooko AK Birimian 19.20 Excellent

Potroase2 POT 2 Birimian 16.83 Excellent

Tinkong1 TK 1 Cape Coast Granitoid 10.90 Excellent

Tinkong2 TK 2 Cape Coast Granitoid 10.26 Excellent

Crig CR Cape Coast Granitoid 10.83 Excellent

Maase1 MS 1 Cape Coast Granitoid 11.77 Excellent

Maase2 MS 2 Cape Coast Granitoid 16.24 Excellent

Maase T 1 MST 1 Cape Coast Granitoid 13.89 Excellent

Maase T 2 MST 2 Cape Coast Granitoid 14.08 Excellent

Adowkwanta ADK Cape Coast Granitoid 15.95 Excellent

Omenako OM Cape Coast Granitoid 7.46 Excellent

Kukua KK Cape Coast Granitoid 8.76 Excellent

Metemano MT Cape Coast Granitoid 6.43 Excellent

Teacher Mante TM Cape Coast Granitoid 14.27 Excellent

Asosotwene AST Cape Coast Granitoid 11.20 Excellent

Afabeng Borehole AF B Cape Coast Granitoid 25.09 Good

Afabeng Hand dug well AF H Cape Coast Granitoid 28.83 Good

Densu Potroase D POT Surface water 15.25 Excellent

Densu mangoase D MAN Surface water 11.93 Excellent

Densu Akyem Odumase D AKD Surface water 15.84 Excellent

Densu Nsawam D NS Surface water 27.92 Good

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

144

The results of the computed WQI values of the Birimian ranges from 11.14 to 27.94,

while the Cape Coast granitoid and the surface water ranged from 7.46 to 28.83 and 11.93 to

27.92 respectively. All the two geological formations showed ‘excellent water’ quality in

most locations with the Birimian having 83.33%, Cape Coast granitoid 81.25% and the

surface water 75%. Few locations NK in the Birimian, AF B and AF H in the Cape-Coast

granitoid and D NS for the surface water, however, showed ‘good’ water quality.

It is evident that, though the geological materials contribute to the presence of dissolved

ions in the water, these areas of ‘good’ water quality might be affected by leaching from point

source pollutants from nearby effluents, domestic disposal site or agricultural wastes (agro-

chemicals, fertilizers etc.) A comparison of the WQI values of the samples reveals that,

groundwater in all the geological formations and the surface water based on the parameters

measured are suitable for drinking. However, Nkroso in the Birimian, Afabeng borehole and

hand dug wells showed evidence of gradual contamination.

Groundwater and Surface Water Classification

The groundwater and the surface water were classified using [39] modification of [40]

method. When pyrites and other sulphide minerals in the aquifer material undergo oxidation,

acid solutions are produced in which heavy metals become highly mobile.

The relationship between pH and total metal content (mg/l) for the analysed samples are

presented in Figure 3. The total metal content was calculated as sum of all the measured metal

(Zn, Fe, Mn, As, Cr, Al and Cu). All the surface water and about 85.71% of the groundwater

samples plot in the field of near neutral-high metal while 14.29% of the groundwater plot in

the region of acid-high metal. In all the cases, the samples plotted in the region of high metal.

Figure 3. Classification of water based metal load and pH.

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 145

Even though, the metal levels were below the WHO guidelines for drinking water, this is

a source of worry because all the groundwater and the entire stretch of the surface water

(Densu River) are used for drinking and irrigation purposes [11]. The high metal

concentration in most of the groundwater may originate from the geology and to a lesser

extent corrosion of pipes or pipe fittings. Most of these pumps have not seen any major

rehabilitation with some functioning with difficulty due to corrosion.

Heavy Metal Pollution Indices

The results of the statistical and location wise Cd, HPI and HEI of the groundwater and

surface water are presented in Table 7 and Table 8 respectively.

Contamination Index (Cd)

Cd values ranged from -0.64 to 20.19 with mean value of 8.44. The computed Cd values

showed that, with the exception of MT (Cd = -0.64) which belongs to the low category, all the

groundwater belong to high contamination level (Cd >3). The surface water Cd ranged from

3.35 to 20.08 with mean value of 10.16 representing 100% contamination.

Table 7. Heavy metal pollution indices, mean deviation (MD) and percentage deviation

(%D) for groundwater in the study area

Location Code Cd MD % D HPI MD % D HEI MD % D

Tinkong1 TK1 3.74 -4.70 -55.73 53.63 -10.61 -16.52 10.74 -4.70 -30.46

Tinkong2 TK2 3.63 -4.81 -57.02 53.25 -10.99 -17.11 10.63 -4.81 -31.17

Crig CR 3.66 -4.78 -56.64 51.81 -12.43 -19.35 10.66 -4.78 -30.96

Maase1 MS1 4.80 -3.63 -43.07 55.36 -8.88 -13.83 11.80 -3.63 -23.54

Maase2 MS2 10.06 1.62 19.25 72.88 8.64 13.45 17.06 1.62 10.52

Maase T 1 MST1 9.21 0.77 9.14 69.84 5.60 8.71 16.21 0.77 4.99

Maase T 2 MST2 9.61 1.17 13.86 62.87 -1.37 -2.14 16.61 1.17 7.57

Adowkwanta ADK 9.74 1.31 15.47 77.11 12.87 20.03 16.74 1.31 8.46

Nkroso NK 16.96 8.52 101.00 87.03 22.79 35.47 23.96 8.52 55.20

Potroase1 POT1 4.42 -4.02 -47.62 47.16 -17.08 -26.59 11.42 -4.02 -26.03

Potroase2 POT2 11.31 2.87 34.05 72.82 8.58 13.35 18.31 2.87 18.61

Potroase

Yawofori POT YOF 10.72 2.28 27.07 78.28 14.04 21.85 17.72 2.28 14.79

Omenako OM 2.50 -5.93 -70.33 40.32 -23.92 -37.24 9.50 -5.94 -38.44

Kukua KK 2.06 -6.38 -75.58 41.04 -23.20 -36.12 9.06 -6.38 -41.31

Metemano MT -0.64 -9.08 -107.55 29.34 -34.90 -54.33 6.36 -9.08 -58.79

Teacher Mante TM 10.46 2.02 23.98 62.68 -1.56 -2.43 17.46 2.02 13.11

Asosotwene AST 6.95 -1.49 -17.64 51.54 -12.70 -19.77 13.95 -1.49 -9.64

Afabeng Borehole AF B 20.19 11.75 139.24 109.22 44.98 70.01 27.19 11.75 76.10

Afabeng Hand

dug well AF H 15.01 6.57 77.86 92.23 27.99 43.57 22.01 6.57 42.55

Anyinase ANY 3.81 -4.63 -54.85 51.12 -13.12 -20.43 10.81 -4.63 -29.98

Akooko AK 16.32 7.88 93.43 79.47 15.23 23.70 23.32 7.88 51.06

Min -0.64 29.34 6.36

Max 20.19 109.22 27.19

Mean 8.44 64.24 15.44

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

146

Table 8. Heavy metal pollution indices for surface water in the study area

Location

Mean

Deviation

Deviation HPI

Mean

Deviation

Deviation HEI

Mean

Deviation

D POT 3.35 -6.81 -67.06 44.34 -21.12 -32.26 10.35 -6.81 -39.71

D MAN 6.5 -3.67 -36.07 50.62 -14.84 -22.67 13.5 -3.67 -21.36

D AKD 10.73 0.57 5.58 66.28 0.82 1.26 17.73 0.57 3.3

D NS 20.08 9.91 97.55 100.59 35.13 53.67 27.08 9.91 57.76

Min 3.35 44.34 10.35

Max 20.08 100.59 27.08

Mean 10.16 65.45 17.16

Heavy Metal Pollution Index (HPI)

The groundwater showed a wide variation in HPI ranging from 29.34 to 109.22 with

mean value of 64.24. The HPI showed that 95.24% of the groundwater samples were below

the critical value of 100 with only AF B slightly exceeding the limit representing 4.76% of

the samples in the study area.

Table 9. PHI calculation for groundwater in the Densu River Basin based on WHO,

2004 and Indian Standard 1991, 10500

Table 10. HPI calculation for surface water in the Densu River Basin based on WHO,

2004 and Indian Standard 1991, 10500

Metals

Mean

concentration

(Mi) (ppb)

Highest permitted

value for drinking

water (Si) (ppb)

Desirable

maximum value

(Ii)( ppb)

Unit

weighting

factor (Wi)

Sub-index

(Qi)

Wi*Qi

HPI

Cu 7015 1500 50 0.00067 480.345 0.32023 65

Al 2011.5 200 30 0.00500 1165.59 5.827941

Mn 446.25 300 100 0.00333 173.125 0.577083

Cr 1.35 10 0 0.10000 13.5 1.35

As 10.25 50 0 0.02000 20.5 0.41

Fe 340.25 1000 100 0.00100 26.6944 0.026694

Zn 3900 15000 5000 0.00007 11 0.000733

∑Wi = 0.13007, ∑Wi*Qi = 8.5126.

Metals

Mean

concentration

) (ppb)

Highest permitted

value for drinking

water (S

) (ppb)

Desirable

maximum value

)( ppb)

Unit

weighting

factor (W

)

Sub-index

)

HPI

Cu 4134 1500 50 0.00067 281.655 0.18777 64

Al 1585.38 200 30 0.00500 914.93 4.57465

Mn 1149.64 300 100 0.00333 524.821 1.749405

Cr 1.34286 10 0 0.10000 13.4286 1.342857

As 9.85714 50 0 0.02000 19.7143 0.394286

Fe 320.571 1000 100 0.00100 24.5079 0.024508

Zn 2161.9 15000 5000 0.00007 28.381 0.001892

∑W

= 0.13007, ∑W

= 8.2754

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 147

The surface water HPI ranged from 44.34 to 100.59 with a mean value of 65.46. This

represent 75% low contamination. The detailed average calculation of the HPI with unit

weightage (Wi) and standard permissible values (Si) are presented for groundwater and

surface water in Table 9 and Table 10 respectively.

Heavy Metal Evaluation Index (HEI)

The HEI developed by [35] also gave a meaningful insight into the extent of the heavy

metal pollution in the study area. The groundwater HEI ranged from 6.36 to 27.19 with a

mean value of 15.44.

The surface water values also ranged from 10.34 to 27.08 with a mean value of 17.16.

The HEI values were grouped based on the mean values using multiple of mean into three

signifying different levels of contaminations.

The proposed grouping criteria are low (HEI<15), medium (HEI=15-30) and high

(HEI>30). Using this criteria, 47.6%, 52.4% and 0% of the groundwater samples show low,

medium and high contamination, respectively, whereas, 50%, 50% and 0% of the surface

water low, medium and highly contaminated.

Comparison of the Three Indices

A comparative study of the three indices (Cd, HPI and HEI) showed a similar trend in all

the sampling points Figure 4. A steady rise in all the indices as the river flows from upstream

to downstream is an indication of effluent, domestic or industrial waste entering the river.

Figure 4. Comparison and spatial distribution of the three pollution indices.

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

148

Table 11. Classification of groundwater and surface water quality based on modified

categories of the pollution indices in the study area

It is interesting to also note that, despite, the similar trends in the indices; they presented

somehow different water quality information. For instant, whereas Cd suggest 95% of

groundwater and 100% of surface water as highly polluted, HPI and HEI indicated that,

4.76% and 0% of groundwater and 25% and 0% of the surface water were polluted.

For the purpose of comparison with HEI, the classification Cd and HPI were slightly

modified using the multiple of mean criteria as HEI. The reclassification scheme showed

comparable results and water quality information, thus for Cd, 61.9% of groundwater and

50% of surface water were classified as low, 33.3% and 25% of groundwater and surface

water were classified as medium, while 4.8% and 25% of the groundwater and surface were

highly contaminated respectively.

The HPI also classified 47.6% of groundwater and 50% of surface water as low, 52.6% of

groundwater and 50% surface water classified as moderately polluted with 0% high pollution

in both cases Table 11.

The relationship between the indices and the metals responsible for the calculated indices

were also examined using Pearson correlation matrix (Table 12). Cd , HPI and HEI showed

strong correlation (Cd vrs HPI =0.751, Cd vrs HEI =0.820 and HPI vrs HEI =0.993). A similar

strong correlation was observed for Cd, HPI and HEI with Cu, Al and Mn suggesting that,

these metals are the dominant or controlling parameters of these indices.

In summary, it can be said that, despite the comparable trends shown by the three indices,

the HPI and HEI are more suitable for water quality index determination in the study area.

Furthermore, the simplicity of HEI as compared to HPI and Cd methods makes it a more

preferable choice. This study corroborates well with the findings of [11] and [35].

Index

method

Class

Degree of

pollution

No. of

location

Percentage

(%)

Groundwater

<10 Low 13 61.9

10-20 Medium 7 33.3

>20 High 1 4.8

HPI <60 Low 10 47.6

60-120 Medium 11 52.4

>120 High 0 0

HEI <15 Low 10 47.6

15-30 Medium 11 52.4

>30 High 0 0

Surface water

<10 Low 2 50.0

20-40 Medium 1 25.0

>20 High 1 25.0

HPI <60 Low 2 50.0

60-120 Medium 2 50.0

>120 High 0 0

HEI <15 Low 2 50.0

15-30 Medium 2 50.0

>30 High 0 0

Nova Science Publishers, Inc.

Table 12. Correlation matrix among metals, pollution indices and the physical parameters in the study area

pH T oC TDS EC Sal Alk Cu Al Mn Cr As Fe Zn Cd HPI

T oC -0.343

TDS 0.548* -0.351

EC 0.549** -0.350 1.000**

Sal 0.511* -0.444* 0.880** 0.880**

Alk 0.813** -0.392 0.633** 0.632** 0.472*

Cu 0.287 -0.011 0.275 0.278 0.303 0.132

Al 0.176 -0.113 0.263 0.263 0.366 0.101 0.503*

Mn -0.273 0.235 -0.074 -0.077 -0.010 -0.167 0.328 0.574**

Cr 0.042 0.097 .0469* 0.468* 0.404 -0.068 0.327 0.181 0.293

As 0.172 -0.105 -0.084 -0.083 -0.156 0.192 0.032 -0.298 -0.161 -0.363

Fe 0.277 -0.346 0.233 0.234 0.225 0.345 -0.105 0.000 -0.432 -0.256 -0.284

Zn 0.385 0.193 0.329 0.332 0.259 0.402 0.350 0.302 0.086 0.139 0.250 -0.164

Cd 0.252 0.025 0.281 0.283 0.321 0.139 0.969** 0.663** 0.510* 0.346 -0.023 -0.147 0.425

HPI 0.020 0.047 0.236 0.235 0.302 -0.011 0.582** 0.863** 0.850** 0.466* -0.208 -0.301 0.292 0.751**

HEI 0.059

0.052

0.249

0.312

0.018

0.668**

0.858**

0.827**

0.458*

-0.180

-0.287

0.335

0.820**

0.993**

* Correlation is significant at the 0.05 level (2-tailed).

** Correlation is significant at the 0.01 level (2-tailed).

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

150

Water for Irrigation Purpose

The quality of water used for irrigation is vital for crop yield, maintenance of soil

productivity and protection of the environment [41]. At the same time, the quality of

irrigation water is very much influenced by the land constituents of the water source. For the

purpose of this work, Sodium Adsorption Ratio (SAR), Sodium Percentage (Na%), Residual

Sodium Carbonate (RSC) and Permeability Index (PI) were used to determine the suitability

of the groundwaters for irrigation purposes.

Sodium Absorption Ratio (SAR)

High Sodium concentration leads to development of an alkaline soil, these results in an

excess Na+ in water producing the undesirable effects of changing the soil properties

(formation of crust, water-logging, reduced soil aeration, reduced infiltration rate and reduced

soil permeability [42]. Therefore, in assessing the suitability of groundwater for irrigation,

Na+ concentration is essential. The degree to which irrigation water enters into cation

exchange reactions in soil can be indicated by SAR [43]. The Na+ replacing adsorbed Ca2+

and Mg2+ is a hazard as it causes damage to the soil structure, making it compact and

impervious. SAR is defined as

22 ++

MgCa

SAR

where, the concentrations are reported in meq/L.

Applying this index to the samples indicates that, 100% of the Birimian and surface water

belong to excellent category with SAR ranging from 2.62 to 6.7 and 2.62 to 5.56 respectively.

The Cape Coast granitoid however, yielded only 73% in the excellent category with 20% and

7% falling under Good and doubtful category.

Sodium Percentage (%Na)

Soils containing a large proportion of sodium with carbonate as the predominant anion

are termed alkali soils; those with chloride or sulphate as the predominant anion are known as

saline soils [44]. Na+ concentration in water is widely use in assessing the suitability of water

for irrigation purposes [45] and plays a vital role in the classification of river water for

irrigation. This is due to the fact that, sodium reacts with soil resulting in clogging of

particles, thereby reducing the permeability [44]; [46] and [47]. It is usually expressed in

terms of percent sodium (Na%) can be calculated by the following relation

100*% 22 ⎟

⎟

⎠

⎞

⎜

⎝

⎛

+++

=++++

NaKMgCa

KNa

where all ionic concentrations are in meq/l

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 151

Figure 5. Rating of groundwater samples on the basis of electrical conductivity and percent sodium

(after Wilcox, 1955), adapted from [25].

The calculated Na% (meq/l) in the samples were plotted against Electrical conductivity in

µs/cm [45] diagram in Figure 5. The results revealed that 100% of the surface water and 83%

of the Birimian fell in the field of excellent to good while 16% of the Birimian samples fell

under permissible to doubtful.

The Cape-Coast granitoid however, showed a wide variation in the Wilcox diagram with

53% of the samples belonging to field of excellent to good, 33% belong to Permissible to

doubtful and 13% being doubtful to unsuitable.

Residual Sodium Carbonate

In addition to the SAR and Na%, the difference between the excess sum of carbonate and

bicarbonate in groundwater against the sum of calcium and magnesium also has a significant

effect on groundwater suitability for irrigation. When total carbonate levels exceed the total

amount of calcium and magnesium, the water quality may be deteriorated. This is because,

high excess carbonate concentration often known as “residual” combines with calcium and

magnesium to form a solid scale like material which settles out of the water [48]. This excess

is denoted by ‘residual sodium carbonate’ (RSC) and is determined as suggested by [49].

()()

++−−

+−+=

222

MgCaCOHCORSC

where all ionic concentrations are expressed in meq/l.

Irrigations water with high RSC have high pH values; hence, soils or land irrigated by

such waters becomes infertile and lowers crop yield owing to the deposition of sodium

carbonate responsible for the black colour of the soil [46] and [50].

Residual sodium carbonate (RSC) has been calculated to determine the suitability or

otherwise of the groundwater in the study area for agricultural purpose. The classification of

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

152

irrigation water was done according to the [51]. RSC < 1.25 meq/l is safe for irrigation,

values between 1.25 to 2.5 meq/l is of marginal quality and a value > 2.5 meq/l is unsuitable

for irrigation [46]. In this study, the surface water and the Birimian samples show RSC values

ranging from −0.03 to −1.14 meq/l and -0.38 to 1.13 respectively representing 100% safe for

irrigation.

The Cape Coast granitoid samples ranges from -0.01 to 3.37 representing 73% safe for

irrigation and 20% marginal quality for irrigation and 7% unsuitable for irrigation.

Furthermore, the negative RSC at all sampling sites indicates that there is no complete

precipitation of calcium and magnesium [52].

Permeability Index (PI)

The PI is also a useful tool which indicates whether water samples are suitable for

irrigation. Irrigated waters influenced by sodium, calcium, magnesium and bicarbonate

contents affect the permeability of soil after a long term use. [46]. Doneen, [53] classified

water based on PI which is given as

(

)

+++

−+

=22

3100*

MgCaNa

HCONa

According to the criteria water can be classified as Class I, II and III. Class I and II are

categorized as good for irrigation with 75% or more maximum permeability. Class III water

is unsuitable with 25% maximum permeability.

All the Birimian and surface waters Figure 6 fall in class I. 87.5% of the Cape Coast

granitoid belongs to Class I while the rest of the 12.5% belong to Class II group, hence, based

on Doneen’s chart all the samples are suitable for irrigation.

Figure 6. Classification of irrigation water for soils of medium permeability (Doneen chart).

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 153

Stable Isotope Analysis

The statistical summary of the isotopic (δ18O and δ2H) composition in rain water,

groundwater and surface water are shown in Table 13.

Rain Water Isotopic Composition

The majority of the rain water stable isotopes (40) were sampled at Koforidua covering a

two year period (2007-2008) with few samples from Accra (7) from March to July 2007.

The relationship between δ D and δ18O based on the 47 samples of the rain water are

shown in Figure 7. According to the measured δ D and δ18O in the 47 sets of rainfall, the data

define a linear trend that represents the local meteoric water line for the Densu river basin.

The slope of 7.3 and intercept of 8.6 are both slightly below the global meteoric water line

[54]. The δ18O values of the rainfall in the Densu river basin ranged from -6.01 to -1.80 with a

mean value of -4.25, the δ D ranged from -35.70 to -2.80 with a mean value of -22.59. The

seasonal variation of rainfall amount and δ D are shown in Figure 8. The lighter isotopic

values are observed in May to September in both years. These months represents the period of

heavy rainfall events, where the humidity is high in both years, hence, much of the

groundwater recharge may occur during these months.

Table 13. Statistical summary of the δ18O, δ 2H and d-excess of the rainwater,

groundwater and surface water in the study area

δ18O δ 2H d-excess

Min Max Mean SD Min Max Mean SD Min Max Mean SD

Groundwater -3.63 -1.43 -2.82 0.49 -16.46 -3.48 -11.24 3.04 7.01 14.08 11.35 1.45

Surface water -3.34 -2.21 -2.76 0.35 -16.1 -10.3 -12.54 1.88 5.58 11.22 9.50 1.88

Rain water -6.35 -1.92 -4.33 1.35 -37.8 -3.7 -23.75 9.81 4.28 14.12 10.90 2.49

Figure 7. Relationship between 18O and 2H in the rain water, groundwater and surface water.

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

154

Figure 8. Seasonal variation of rainfall amount and δ 2H in the study area.

Groundwater Isotopic Composition

Deuterium values of twenty one (21) groundwater ranges from -16.46 to -3.48‰ while

oxygen-18 values range from -3.63 to -1.43 ‰ VSMOW. The average values of δ 2

H and δ

18O were -11.24±1.62‰ and -2.82±0.37‰ respectively. The best fit regression line of the

groundwater was 2.583.5 182 +∂=∂ OH with a correlation coefficient of r2 =0.896. The

surface water had δ 2

H and δ 18O ranging from -6.1 to -10.3‰ and -3.34 to -2.21‰

respectively with an average of -12.54±1.64 for δ 2

H and -2.76±0.09 for δ18O. The best fit

regression line was 3.108.4 182 −∂=∂ OH with a correlation coefficient of r2=0.578. It was

observed that both the groundwater and the surface water have slopes lower than the rainfall.

The lower slope implies primary evaporation during precipitation as the moisture moves

allowing time for more contact and exchange with the atmosphere. The observed isotopic

variation in the rainfall might be due to different rainfall events.

Origin of Groundwater

The isotopes of oxygen 18O and hydrogen δ 2H are a sensitive tracer and widely used in

studying the natural water circulation and groundwater movements. Differences in the content

of δ 2

H and δ 18O in groundwater, surface water and rainfall collected at Koforidua were

exploited on a similar graph to show the extent of variation in the study area Figure 7. A local

meteoric water line (LMWL) obtained by [55] for the Accra plains (which forms part of the

study area) defined by the equation

61.1386.7 182 +∂=∂ OH

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 155

and the Global meteoric water line [54] defined by the equation

108 182 +∂=∂ OH

were also inserted in the graph.

The stable isotope composition relative to Global Meteoric Water Line (GMWL) reveals

important information on the groundwater recharge patterns, relationship between ground and

surface water. As evident from Figure 7, most groundwater samples plot in between the

GMWL and LMWL with few samples falling on the lines. This suggests a meteoric origin for

the groundwater. The groundwater in the study area appears to group in a narrow range,

signifying a well-mixed system with relatively constant isotopic composition. A few (15%) of

the groundwater samples plot slightly away from the meteoric water line showing an evidence

of small isotopic enrichment by evaporation on the surface or in the unsaturated zone before

recharge. This shows that, generally, the meteoric water recharging the groundwater system

in the area is homogeneous with evaporation playing an insignificant role on the infiltrating

water. Similar observations were made in the southern portion of the basin [56]. All the

surface waters plot relatively below the global meteoric water line indicating a degree of

isotopic enrichment. This observation can be attributed partly to the open flow of the river

and to some extent the isotopic enrichment could be reflecting the integration of isotopic

composition of the Densu river tributaries. The narrow range in the stable isotopic

composition of the surface water shows a homogeneous and well mixed system between the

Densu River and its tributaries.

It is interesting to note that few of the groundwater samples exhibit similar isotopic

composing to that that of the surface water (River Densu), this suggest a possible hydraulic

connection between the aquifers and the river water, some degree of fractionation both on

land surface and in the unsaturated zone and most probably mixing mechanisms by

anthropogenic activities such as irrigation, which might result in the groundwater being

recharge by enriched waters.

Deuterium Excess (D-excess)

The d-excess reflects the conditions that lead to kinetic isotope fractionation between

water and vapour during primary evaporation in the oceans [57]. This number also shows the

extent of deviation of a given sample from the meteoric water line. The calculated deuterium

excess of the rainfall was found to range from 7.01 to 14.08‰ with most samples plotting

together confirming a common moisture source for the rainfalls. The groundwater d-excess

values ranges from 5.58 to 11.22‰, while that of the surface water ranges between 4.28 to

14.12‰. As the δ18O increases (more enriched) the deuterium excess in all the samples

decreases gradually Figure 9.

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

156

Figure 9. d-excess vs δ18O plot in the study area.

It is also noticed that the groundwater samples plot together with the rainfall and the

surface water. This observation can be partly attributed to the dilution of the groundwater

with the rainfall which could increase the d-excess [58]. Furthermore, the tendency towards

the rainfall may imply a modern recharge to the groundwater systems.

CONCLUSION

The study investigated the suitability of groundwater and surface water in the Densu river

basin for drinking and agricultural purposes using a wide range of water quality indices. The

results showed that, the groundwater and surface in the basin are generally suitable for

drinking and irrigation. The stable isotopes showed that, the groundwater are generally of

meteoric origin and of modern day recharge.

ACKNOWLEDGMENT

The authors wish to thank the technicians of the isotope hydrology lab of Ghana Atomic

Energy Commission for their support in sample analysis. We also wish to thank Prof Amuesi

and Mr. Oware Kesse for providing transport to the field.

REFERENCES

[1] M. Vasanthavigar, K. Srinivasamoorthy, R. Rajiv Ganthi, K. Vijayaraghavan, V. S.

Sarma, Arabian Journal of Geosciences. 5, 245, (2012).

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 157

[2] P. N. Rajankar, S. R. Gulhane, D. H. Tambekar, D.S. Ramteke, S.R. Wate, E-Journal of

Chemistry. 6, 905 (2009).

[3] P. K. Goel, Water Pollution – Causes, Effects and Control, New age Int. (P) Ltd, New

Delhi (2000).

[4] K. P. Bam, Major and trace elements in soil profile of the unsaturated zone of the

Densu river basin, Ghana, University of Ghana, Unpublish Mphil thesis. (2009).

[5] M. Vasanthavigar, K. Srinivasamoorthy, K. Vijayaragavan, R.R. Ganthi, S.

Chidambaram, P. Anandhan, R. Manivannan, S. Vasudevan, Environmental Monitoring

and Assessment. 171, 595, (2010).

[6] P. Simeonova, V. Simeonov, G. Andreev, Central European of Chemistry. 1, 136,

(2003).

[7] S.V. Mohan, P. Nithila and S.J. Reddy, J. Environ. Sci. Health A. 31, 283, (1996).

[8] N. Nishidia, M. Miyai, F. Tada, S. Suzuki, Environ. Pollution. 4, 241, (1982).

[9] T. N. Tiwari and M. Mishra, Indian J. Environmental Protection, 5, 276, (1985).

[10] B. Prasad and K.K. Mondal, Mine Water Environ., 27, 40, (2008).

[11] M.A.H. Bhuiyan, M.A. Islam., S.B. Dampare, L. Parvez., S. Suzuki, Journal of

Hazardous Materials. 179, 1065, (2010).

[12] W. M. Edmunds, J. J. Carrillo-Rivera and A, Journal of Hydrology, 258, 1, (2002).

[13] J.Ch Fontes, Environmental Isotopes in groundwater hydrology In handbook of

Environmental Isotopes Geochemistry, (1980).

[14] IAEA, Guidebook on Nuclear Techniques in Hydrology, Technical Report series No.

91, IAEA, Vienna, (1983).

[15] A. Gibrilla, T.T. Akiti, S. Osae, D. Adomako, S.Y. Ganyaglo, E.P.K. Bam, A. Hadisu,

Journal of Water Resource and Protection, 2, 1071, (2010a).

[16] D. Adomako, S. Osae, T.T. Akiti, S. Faye, P. Maloszewski, Environ. Earth Sci., DOI

10.1007/s12665-010-0595-2, (2010).

[17] J. R. Fianko, S. Osae, D. Adomako, D.G. Achel, Environmental Monitoring

Assessment, 30, 145, (2008).

[18] WRC, Groundwater assessment: an element of integrated water resources management:

the case of Densu River Basin, Technical report, Water Resources Commission, Accra.

(2006).

[19] A. Gibrilla, T.T. Akiti, S. Osae, D. Adomako, S.Y. Ganyaglo, E.P.K. Bam, A. Hadisu,

Journal Water Resource and Protection, 2, 1010, (2010b).

[20] D. K. Sinha, A. K. Srivastava, Indian Journal of Environmental Protection, 14, 340,

(1994).

[21] WHO, World Health Organisation guidelines for drinking water quality, Third edition.

Geneva, ISBN 92 45 154638 7, (2004).

[22] Ghana Statistical Service, 2000 population and housing census, (2000).

[23] K. B. Dickson and G. Benneh, A New Geography of Ghana. Longman, London. (1980).

[24] A. Gibrilla, pollution, hydrochemical and isotopic studies of groundwater in the

northern Densu river of basin, Ghana. University of Ghana. Unpublish Mphil thesis.

[25] A. Gibrilla, E.K.P. Bam, D. Adomako , S. Ganyaglo, S. Osae, T.T. Akiti, S. Kebede, E.

Achoribo, E. Ahialey, G. Ayanu, E.K. Agyeman, Water Qual Expo Health. DOI

10.1007/s12403-011-0044-9. 3, 63, (2011).

[26] M. Radojevic,. and V.N. Bashkin, , Organic matter. In: Practical Environmental

Analysis. The Royal Society of Chemistry, Cambridge. 325-329. 1999.

Nova Science Publishers, Inc.

Abass Gibrilla, Edward Bam, Dickson Adomako et al.

158

[27] T. B. Coplen, Chemical Geology, 72, 293, (1988).

[28] S. S. Asadi, P. Vuppala, M.A. Reddy, International Journal of Environmental Research

and Public Health, 4, 45, (2007).

[29] S. K. Pradhan, D. Patnaik, S.P. Rout, Indian. Journal of Environmental Protection, 21,

355 (2001).

[30] R. D. Harkins, Journal Water Pollution Control Federation, 46, 588, (1974).

[31] R. K. Horton, Journal Water Pollution Control Federation, 37, 300, (1965).

[32] ISI, Indian standard specification for drinking water. New Delhi. ISI, 10500, (1993).

[33] B. Backman, D. Bodis, P. Lahermo, S. Rapant, and S Tarvainen, Environmental

Geology 36, 55, (1997).

[34] S.V. Mohan, P. Nithida and S.J. Reddy, Journal Environmental Science and Health,

A31, 238, (1996).

[35] A.E. Edet,. and O.E. Offiong, GeoJournal, 57; 295, (2002).

[36] S.J. Reddy, Encyclopedia of environmental pollution and control. Environmental

Media, Karlla, India, 1, 342, (1995).

[37] R. A. Freeze and J.A Cherry, Groundwater. New Jersey: Prentice Hall. (1979).

[38] S. N. Daviest and R.J.M. Dewiest, Hydrogeology, John Willey and sons, New York,

(1966).

[39] R. Caboi, R. Cidu, L. Fanfani, P. Lattanzi and P. Zuddas, Chron. Rech. Miniere 534,

21, (1999).

[40] W.H. Ficklin, G.S. Plumee, K.S. Smith and J.B. McHugh, Geochemical classification

of mine drainges and natural drainages in mineralised areas. In: Kharaka Y.K. and

Maest A.S. (eds), Water-rock interaction. Vol 7. Balkema, Rotterdam, pp 381 (1992).

[41] V. Singh, U.C. Singh, Indian Journal of Science and Technology, 1, 1 (2008).

[42] W.P. Kelly, Alkali soils – Their formation, Properties and Reclamation, Reinhold

Publication, New York (1951).

[43] T. M. Manjusree, S. Joseph, J. Thomas, Journal of the Geological Society of India, 74,

459 (2009).

[44] T. Suresh, N.M. Kottureshwara, Rasayan J. Chem., 2, 221 (2009).

[45] L.V. Wilcox, Classification and use of the irrigation waters, U.S. Department of

Agriculture Circular No. 969, Washington, District of Columbia, (1955).

[46] A. Nagaraju, S. Suresh, K. Killham, K, Hudson-edward, Turkish J. Eng. Env. Sci.

30,203 (2006).

[47] D. K. Todd, Ground water hydrology, New York: Wiley (1980).

[48] S. K. Sundaray, B. B. Nayak, D. Bhatta, Environmental Monitoring and Assessment,

155, 227 (2009).

[49] L.A. Richard, Diagnosis and improvement of saline and alkali soils, Agricultural

handbook 60, Washington, DC: USDA (1954).

[50] F.M. Eaton, Soil Sci, 69, 123 (1950).

[51] U S Salinity Laboratory staff et al, Diagnosis and improvement of saline and alkali

soils, U.S. Deptt. Agr. Handbook. (1954).

[52] T.N. Tiwari and A. Manzoor, Indian Journal of Environmental Protection, 8, 494

(1988).

[53] L.D. Doneen, Notes on water quality in Agriculture, Published in water Science and

Engineering, Paper 4001, Department of water Science and Engineering, University of

California. (1965).

Nova Science Publishers, Inc.

Application of Water Quality Indices (WQI) and Stable Isotopes ... 159

[54] H. Craig, Isotopic variation in meteoric water, Science, 133,1702 (1961).

[55] T. T. Akiti, Environmental isotope study of groundwater in crystalline rocks of the

Accra Plains, 4th Working Meeting on Isotopes in Nature, proceedings of an advisory

group meeting, IAEA, Vienna. (1986).

[56] Goucey et al. Application of Isotope Techniques for the Assessment of Groundwater

Resources: Densu river basin report. (2008).

[57] W. Dansgaard, Tellus, 16, 436 (1964).

[58] F. Yuan and S. Miyamoto, Characteristics of oxygen-18 and deuterium composition in

waters from Pecos River in American Southwest J. Chemical Geology. 255. pp 220-230

(2008).

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 7

EVA L U AT I O N O F COMMUNITY WATER QUALITY

MONITORING AND MANAGEMENT PRACTICES, AND

CONCEPTUALIZATION OF A COMMUNITY

EMPOWERMENT MODEL: A CASE STUDY OF

LUVUVHU CATCHMENT, SOUTH AFRICA

L. Nare

∗

and J. O. Odiyo

Department of Hydrology and Water Resources, University

of Venda, Thohoyandou, South Africa

ABSTRACT

South Africa has projected herself as a democratic society and one of the

cornerstones of democracy is community involvement and participation in development.

Therefore, a study to evaluate and conceptualise a model for community empowerment in

water quality monitoring and management was carried out in Luvuvhu Catchment of

South Africa. The first task was to prove that the communities in the catchment were

vulnerable to water quality problems. Potential sources of pollution and types of

pollutants were identified. Community vulnerability and risk caused by the pollutants was

confirmed by analysing results obtained from DWA surface and groundwater quality

monitoring stations dotted around the catchment. Water treatment efficiencies at three

water treatment plants supplying water to communities in the catchment were calculated

to confirm community vulnerability due to inadequate supplies and poor water quality.

The contemporary water quality monitoring and management practice was evaluated to

identify any gaps relating to community involvement and participation. The willingness

of the community to be involved and participate in water quality monitoring was

evaluated. Indigenous knowledge systems related to water quality management were

investigated to identify ways in which these could be incorporated into the national water

quality monitoring framework as a way of strengthening the community’s voice in water

quality monitoring and management. The results showed that communities in Luvuvhu

Catchment are vulnerable to water pollution problems. The contemporary water quality

∗ E-mail: leratonare@yahoo.com.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

162

monitoring and management practice does not promote community involvement and

participation despite the fact that the communities were willing to participate in water

quality monitoring and were even prepared to pay for the services. The study found a

wealth of indigenous knowledge related to water quality monitoring and management.

Based on the above results, a model for empowering communities to participate in water

quality monitoring and management was proposed and its acceptability and

reasonableness determined. The model is anchored on three frameworks; technological,

empowerment and communication frameworks. The model envisages a situation where

simple technologies based on indigenous knowledge are developed to monitor water

quality at the community level as a way of empowering communities to take

responsibility for managing their own resources including water.

Keywords: Community, participation, empowerment, model, Luvuvhu Catchment

1. INTRODUCTION

There are a number of factors that compel South Africa to adopt effective community

participation in water quality monitoring and management. The first factor relates to water

scarcity in the country. Although Luvuvhu catchment has abundant water resources for now,

the catchment will face problems of water availability in the near future. The surplus water is

provided by Nandoni Dam, but the dam was built to supply the growth in urban and industrial

demands in Thoyandou and Louis Trichardt (DWAF, 2006). According to the Thulamela

Municipality Integrated Development Plan (IDP) (2008), there was a backlog of 93 121

households without access to reticulated water in 2008. When these households are connected

to the municipal water supply system, the amount of water available in the catchment is likely

to be reduced. Luvuvhu Catchment also supplies other catchments with water. About 2.4

million m3/year of water is transferred from the Albasini dam on the Luvuvhu River to

Makhado Municipality which is in the Sand River Catchment. Vondo Regional Water

Scheme on Mutshundudi tributary of the Luvuvhu River supplies water to some villages

which fall under Letaba River Catchment (DWAF, 2005). The high population growth and

economic development in Makhado Municipality and other areas which obtain water from the

Luvuvhu Catchment will also worsen the situation as more water will be required to meet the

demand. The impacts of climate change are also expected to reduce the available water

resources (DWAF, 2005).

Pollution of water is becoming a significant problem in South Africa as the country

continues to industrialize. In many reservoirs, water quality has deteriorated as a result of

increasing salinity, nutrient enrichment and bacteriological pollution upstream (Silberbauer et

al., 2001). Major sources of pollution for surface waters are agricultural drainage and wash-

off (irrigation return flows, fertilizers, pesticides, and runoff from feedlots), urban wash-off

and effluent return flows (bacteriological contamination, salts and nutrients), industries

(chemical substances), mining (acids and salts) and areas with insufficient sanitation services

(microbial contamination). The worst examples of pollution in South Africa include the

Klipspruit stream case in Soweto where the stream was extremely polluted with acid and

toxic metals from old mine dumps and tailings (Moyo and Mtetwa, 2002). Sulphate

concentrations were more than 2000 times over the WHO standards. Van Zyl (1999) gives an

example of the Juskei River in Alexandria township of Johannesburg where sewage from an

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 163

over populated settlement, heavily contaminated the river, with E. coli counts averaging 3

million per 100ml. SANS 241 require that 95% of all samples taken from a point should show

no viable growth of E. coli for the water to be declared safe for human consumption.

In 2005 basic sanitation services were available to only 67% of the population in South

Africa (Anderson et al., 2008). This means that there were still over 30% of the people

without access to basic sanitation and thereby posing a threat to the quality of water resources

in the country.

The perception of water pollution as a community problem in South Africa is very low. A

study conducted by Marie Wentzel of Statistics South Africa in 2008 found that slightly less

than 11% of all households who participated viewed water pollution as a community problem.

This poses a problem because the same author argues that “awareness of environmental

problems and the willingness to deal with them are more likely to be present among

populations with more enlightened perception of environmental issues”.

In South Africa, 3.6 million people do not have access to safe drinking water, while 26

600 deaths annually are related to diarrhoea, in most cases preventable (Barwell, 2006).

Investigations show that an unacceptably high incidence of poor drinking water quality occurs

in non-metro South Africa (Hodgson and Manus, 2006).

In 2000, South Africa experienced one of the worst cholera epidemics in the country’s

recent history (Genthe and Steyn, 2006). By the end of the year, the cholera outbreak had

spread to eight of South Africa’s nine provinces, with a total of 1 0638 99 reported cases and

229 reported deaths. On 5 September 2005, a typhoid outbreak was confirmed in Delmas

resulting in five reported deaths and the hospitalisation of 17 people (Department of Health,

2005b). Delmas is in Mpumalanga, a province which is adjacent to Limpopo province where

Luvuvhu catchment lies. Thus the population in the study area is vulnerable to waterborne

diseases which could be prevented with good water quality monitoring and management

measures.

Therefore, management of South Africa’s water quality and availability is essential since

it is predicted that the demand for water will outstrip its supply by 2025 (Hirji et al., 2002).

South African water resources are expected to decline markedly in the years to come for the

reason that the ratio of runoff to rainfall is amongst the lowest of any populated region of the

world (Oberholster et al., 2007). Involvement and participation of communities in water

quality monitoring and management will increase the capacity of the nation to manage

properly the quantities and quality of the scarce water resources.

It is apparent that there is need to go beyond the usual framework for monitoring and

managing water quality. Involvement of indigenous communities in this case could be a

viable option. Tsiho (2007) states that communities all over the world have developed their

own knowledge and practices for observing, measuring, and predicting environmental quality

change, which are embedded in their indigenous languages and cultural beliefs. He argues

that "there is little doubt that people at the grassroots have knowledge of their environment

that transcends conventional social, economic and biological indicators." Therefore there is

need to create space for this indigenous knowledge to be incorporated into water quality

management strategies currently being used. Indigenous knowledge is not only at its best

when it is matched with contemporary science (Hens, 2006). Bell and Davies (2001) argued

that, community involvement in Integrated Water Resources Management (IWRM) or in

other environmental issues is based on “the need to use indigenous knowledge (IK) as well as

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

164

opinions that are vital for environmental protection, including proper water resource use and

management”.

The use of participatory approaches in the management of water resources is one of the

Principles of the Dublin Convention. Although many researchers agree on the importance of

local community involvement in IWRM, the level of involvement is still low in most

developing countries (GWP, 2000). Public participation in decision-making, especially

focusing on historically disadvantaged and marginalised communities, concerning water

resource protection is one of the basic principles of IWRM in South Africa (Department of

Health, 2005a). The participation of stakeholders in decision making processes and therefore

governance of water resources is a critical strategy in ensuring the sustainability of

watersheds in the provision of resources (Ong’or, 2005). The National Water Act of 1998

(Chapter 7) advocates for the promotion of community participation in protecting, using,

developing, conserving, managing and controlling of water resources (Karar, 2003).

Communities are the primary stakeholders in the watersheds where they live and therefore

they need to be empowered with knowledge and tools that will enable them to effectively

participate in the governance of water resources (DWAF, 2005).

Community participation involves holding discussions and open forums between

community members themselves and with government authorities or non governmental

organizations involved in advocacy so as to contribute ideas for inclusion in policy

development and change in operation strategy (DWAF, 2005). If given a chance,

communities can participate effectively in matters relating to water resources management. In

Kalomo (Zambia), the local community was mobilized to manage provision of water services,

whereby villagers protected a catchment area by building a fence around a borehole and

regularly cleaned the water point (Dungumaro and Madulu, 2002). Evidence from Gujarat

(India) demonstrates the linkages between local community involvement in water project

management and empowerment of stakeholders, especially imparting them with the capacity

to negotiate with other stakeholders at higher levels concerning issues that affect their

livelihood and lifestyle (Dungumaro and Madulu, 2002).

The WHO Protocol on Water and Health of 2006 recognizes that water “has social,

economic and environmental values, and should therefore be managed so as to realize the

most acceptable and sustainable combination of these values”. The Constitution of South

Africa (Chapter 2, Section 24), gives the citizens the right “to an environment that is not

harmful to their health and well-being” (Anderson et al., 2008). Despite improved access to

safe water and sanitation services and the creation of a legal framework to protect the

country's natural resource base, inequities and inequalities in both these areas remain

(Hemson and O’Donovan, 2005).

All the above arguments may not be realized unless the communities are adequately

empowered to deal effectively with issues that relate to the management of their natural

resources including water. They need to be armed with knowledge, skills and tools that will

enable them to sit and discuss with technocrats on a relatively equal footing. Kauzeni and

Madulu (2000) found that, though community participation is emphasized in developing land

use plans, in many cases local communities and their local knowledge are ignored by planners

in developing and managing land and water resources. This study therefore sought to identify

indigenous knowledge, attitudes and practices in Luvuvhu Catchment relating to water quality

monitoring and management that could be incorporated into the formal water quality

monitoring framework and enable communities to participate effectively in this regard.

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 165

Friis-Hansen (1999) emphasized the need for taking indigenous knowledge on board

when planning, developing, implementing and managing water resources. He argued that

although experiences and knowledge of local people lack scientific explanations, they are a

strong weapon in solving local problems. Research in local knowledge could ensure

community participation, and indigenous/local knowledge could be used to facilitate

development of water projects that are environmentally sustainable and meet national and

community development objectives (Adams et al., 1994). It has been shown that the

implementation of well considered, community accepted drinking water quality management

procedures can effectively change an unacceptable water quality to one that satisfies drinking

water specifications (Mackintosh and Colvin, 2003).

South Africa is a signatory to a number of international and regional agreements relating

to management of water resources such as the SADC Protocol on Shared Waters and

therefore the communities in the country need to be empowered to deal with pollution and

water quality issues so that the country does not find itself in breach of those agreements.

Developing a model which will enable communities to participate effectively in water quality

monitoring and management will most likely go a long way in avoiding a situation where the

country is found to be in breach of the agreements.

“Indigenous People’s Kyoto Water Declaration” at the Third World Water Forum in

Kyoto (Japan) in March 2003 was an attempt to bring to the forefront the plight of indigenous

communities from around the globe. This study therefore was a tangible contribution towards

allowing an indigenous community to have a voice over the management of their natural

resources including water. Indigenous knowledge (IK) concerning the environment, is in

many societies around the world, in danger of being lost, since western science has lately

been controlling the development of environmental management practices to such a large

extent (Wigrup, 2005).

It is therefore of the greatest importance to record and assess such IK before it becomes

extinct. Luvuvhu Catchment covers part of Makhado Local Municipality and most parts of

Thulamela Locality Municipality in Vhembe District. The two local municipalities are

dominated by two indigenous groups i.e. the Venda and the Tsonga. Therefore by pulling out

some aspects of their knowledge, attitudes and practices and integrating them into the national

water quality framework, the study will contribute towards the preservation of this

knowledge.

2. THE STUDY AREA

The Luvuvhu Catchment forms part of the Luvuvhu/Letaba Water Management Area

which lies between 29049’E and 31055’ E and 2401’S and 29049S. The Luvuvhu River and

some of its tributaries (including the Mutshindudi and Mutale Rivers) rise in the

Soutpansberg Mountains. It flows for about 200km through a diverse range of landscapes

before it joins the Limpopo River near Pafuri in the Kruger National Park. Except for the

Thohoyandou and Malamulele urban centres, the catchment mainly consists of rural

settlements, commercial agriculture and the Kruger National Park. The major uses of water in

the catchment are domestic water supplies followed by commercial agriculture and then

environmental use.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

166

Figure 1. Map of the portion of the Luvuvhu River Catchment that comprised the study area showing

treatment plants.

There are four water treatment plants in the catchment. These are Albasin, Malamulele,

Xikundu and Mhinga. Albasin is located in a commercial area and it supplies water to Louis

Trichardt which is in the Sand River Catchment and therefore the area around Albasin was

not included in the study. The study area effectively covered the areas that are supplied with

water from Malamulele, Xikundu and Mhinga treatment plants (Figure 1).

ETHODOLOGY

The first objective of the study was aimed at proving that indeed there were pollution

problems in Luvuvhu Catchment warranting investigation. The study had to prove

community vulnerability to pollution problems in the catchment and that they faced potential

risks to their health due to the pollution problems. The first action was to review any literature

and official documents that could prove that there was potential vulnerability and health risk

faced by communities due to pollution problems in the catchment. The second action was to

analyse water quality from Luvuvhu River to prove that indeed the water in the river was

polluted as suggested by the literature reviewed. Water quality results from four monitoring

stations along the Luvuvhu River the period 2005 – 2007 were obtained from the Department

of Water Affairs (DWA) and analysed to further confirm the vulnerability and risks faced by

communities in the catchment. The water quality monitoring program in Luvuvhu River

commenced in 2005 and at the time the data were obtained, complete data sets were only

available for four stations between 2005 and 2007.

The third action was to monitor water quality from one street tap from each of the 15

villages participating in the study for 8 months to further confirm and fortify the case for

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 167

community vulnerability and risks in the catchment. Microbiological analysis of water was

carried out using Total Coliforms and E. coli as indicators.

The fourth action was to analyse water treatment efficiencies from the treatment plants

that supplied water to the community in the catchment to find out if the quantities were

adequate and the quality good for the target population. Treatment efficiency for a plant in

relation to quantities of water processed was calculated from dividing the volume of treated

water per day by the corresponding volume of raw water that went in and then multiplying the

product by 100. The treatment efficiency in relation to quality was calculated from dividing

the magnitude of a particular water quality parameter in treated water by the magnitude of the

same parameter in raw water. Only Mhinga Water Treatment Plant had enough data to cover

a period of two years. Therefore treatment efficiencies could only be calculated for the

Mhinga Water Treatment Plant.

The fifth action was to collect and analyse data from health facilities in the catchment

relating to waterborne diseases such as diarrhoea to really prove that the communities in the

catchment were indeed at risk due to water pollution problems. The final or sixth action under

the first objective was to find out the feelings and perceptions of the communities themselves

relating to the quality and safety of their water supplies.

The second objective was to analyse and identify any gaps in the current water quality

monitoring and management frameworks. This involved an extensive review of the policy,

legal, institutional and organizational frameworks relating to water quality monitoring and

management in South Africa. It also involved conducting interviews with practitioners and

communities to solicit their views on the current frameworks.

The third objective sought to evaluate indigenous knowledge, attitudes, practices and

perceptions related to water quality monitoring and management in Luvuvhu Catchment. This

was achieved in two parts. The first part was achieved through interviews with around 8 000

members of the community from the catchment. Participatory tools and a structured

questionnaire were used to collect data from the community. The second part evaluated the

community’s perception of water safety using turbidity as the first part had indicated that

communities relied on the physical appearance of water to decide whether it was still suitable

for use or not. The second part of the study therefore sought to identify the point at which

members of the community would reject the use of water based on its physical appearance. A

template was used to collect data from the respondents.

The fourth objective sought to evaluate community participation in water quality

monitoring and management in Luvuvhu Catchment. This involved carrying out interviews

with practitioners and members of the community and review of any related literature and

official reports.

The fifth objective sought to evaluate the reasonableness and acceptability of the

proposed model for effective community participation in water quality monitoring and

management for the communities in Luvuvhu Catchment and other stakeholders. The

proposed model was circulated to various academics and government officials to solicit for

their inputs.

A survey was carried out among the community leadership in the 15 participating

villages using a questionnaire. This included the Chief of the area, Village Heads, Civic

Chairman and the Ward Committee Member from each village. The study sought to evaluate

the perception of the leadership in relation to the proposed model, finding out if it was

acceptable to them and capturing their contributions.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

168

3.1. Water Quality Monitoring

Secondary water quality data collected by Department of Water Affairs (DWA) from

their monitoring stations along the Luvuvhu River were used in the study. These were

monthly total water quality data covering a period of two years starting from July 2005 to

August 2007 to indicate the quality status. The quality of water from street taps in the villages

participating in the study was monitored for a period of eight months. The results for Total

Coliforms and E. coli were almost always negative and therefore the exercise was stopped

after eight months.

3.2. Community Surveys

Two major community surveys were carried out to evaluate indigenous knowledge,

attitudes, practices, perceptions relating to water quality monitoring and management. These

studies involved people obtaining their water for different uses from the Luvuvhu River

system. The communities mostly have access to reticulated water supplied by three treatment

plants namely Malamulele, Xikundu and Mhinga. For the first study the sample involved a

large number of people (over 8 000 out of 53 333 people) (15%) and the use of structured

interviews to collect data from individuals proved to be very expensive. Therefore

participatory tools were used to gather qualitative data which was then converted into

quantitative data. Using the formula;

n = [ (z2 * p * q ) + ME2 ] / [ ME2 + z2 * p * q / N ].

This normally used in social science studies to determine the sample size, the ideal

sample for the given study population would have been around 5000 respondents but the

sample had to be increased to 8000 to improve on the accuracy of results.

The exercise was divided into a number of sessions with each session having 45 people.

In a session the people were divided into three groups of 15 people each. Each group was

assigned some themes relating to water quality monitoring and management and asked to

work through them using relevant participatory tools and then write the findings on flip charts

following instructions. The groups would then present at a plenary session and the Research

Assistants (RA) would take down the responses from each group. The data were then

converted into quantitative data as the responses were presented in terms of how many of the

groups gave a particular response. This automatically converted the data into numbers which

could be analysed statistically instead of qualitative statements. Structured questionnaires

were used to interview government officials and other stakeholders at catchment level.

The second part of the study was based on the findings of the first one and aimed at

finding the point at which the community would reject the use of water for various purposes

based on the physical appearance of water. The perceptions of 1 000 people (2% of the total

population) relating to safety of water based on the degree of turbidity were evaluated.

Samples of water with known turbidity, chemical and microbiological values were shown to

each respondent.

The respondent was then asked whether he/she would be able to use the water from each

sample for various uses such as drinking, cooking, bathing and laundry. The responses were

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 169

recorded on a template. A profile of turbidity values which the communities considered to be

unsafe for each use was compiled. A table outlining the relationship between the level of

turbidity and possible health impacts as given in the South African Water Quality Guidelines

(1996) was used to interpret and make sense of the turbidity values at which respondents

rejected the water for different uses.

3.3. Sampling

3.3.1. Water Sampling Points

Total water quality data from the following DWA sampling points was collected and

analyzed;

• Downstream of Thohoyandou sewage works

• Downstream of Waterfall sewage works

• Downstream of Elim sewage works

• Vuwani Oxidation Ponds

3.3.2. Community Sampling

There are four water treatment plants in the part of the catchment where the study was

done namely Albasini, Malamulele, Xikundu and Mhinga. Albasini supplies water to Louis

Tritchadt which lies outside the catchment and therefore was not included in the study.

Fifteen percent (15%) of villages being supplied by each of the remaining treatment plants

were included in the study. The selection was done randomly. Each village was given a

number and the number was written on a small piece of paper which was put in a container

and the contents thoroughly mixed up before a paper was selected and the number recorded.

The paper was then thrown back into the container. The process was repeated until 15% of

the villages from each treatment plant were picked.

Multi - stage sampling was carried out to come up with the people who took part in the

study. Once the villages were selected, 15% of the households in each village were selected to

participate in the study. Random sampling was carried out to select the participating

households. Each household in a village was assigned a number, which was then written on a

small piece of paper. All the pieces of paper were then put in a container and shuffled

thoroughly before the chief of the village was asked to pick a paper. The number on the paper

was written down and the piece of paper thrown back into the container. The process was

repeated until 15% of the households were selected. Three members from each selected

household were then invited to take part in the study. If one of the required members was not

available in a particular household, a member from the next household was invited to

participate in the study.

3.4. Analysis of Data from Community Survey

Two sets of data were gathered from the community surveys. The first set concerned the

data relating to the knowledge, attitudes, practices and perceptions (KAPP) relating to water

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

170

quality management. The second set of data consisted of responses of community members to

the water samples containing the different degrees of physical appearance of water (turbidity).

In the first part of the KAPP study, communities indicated that they depended on the physical

appearance of water to determine whether a body of water was still fit for human

consumption. Turbidity was therefore used as an indicator of physical appearance of water in

this study.

The analysis of data from KAPP study started with the conversion of qualitative data

gathered through participatory approaches to quantitative data that is capable of being

analysed by statistical methods. The responses from the different groups were then entered

into EXCEL, a computer programme or spreadsheet before being transported into SPSS

which analyses statistical data. SPSS stands for Statistical Package for Social Sciences and it

organises quantitative research data to determine the relevance of variables associated with

the research topic. PASW (Predictive Analysis Software) brand was used for this analysis.

The information obtained from analysing the data from the KAPP study was subjected to

further statistical analysis to obtain the mean and standard deviation.

The data from the turbidity/colour study was analysed manually. The information

obtained was used to develop a profile of different degrees of turbidity/colour and the likely

responses from members of the community to the water.

4. RESULTS

4.1. Water Quality Monitoring and Contemporary Management Practice

Although South Africa has a comprehensive and sophisticated water quality monitoring

and management programme backed by very elaborate and extensive policy, legal,

institutional and technical frameworks, implementation at community level leaves a lot to be

desired. While the national constitution guarantees all citizens the right to access sufficient

water and a safe environment, the practical modalities on the ground make it impossible to

achieve this. The way in which the water quality monitoring programme is implemented

currently makes it hard for the country to guarantee the above constitutional provisions.

While the policy, legal and technical frameworks are comprehensive enough to guarantee the

above constitutional provisions, the structures of the institutional framework and lack of

resources make it impossible to do that.

The water quality monitoring programme is currently being transformed from a

centralised to a decentralised one. The transition phase also contributes to the weaknesses in

the system. While the main actors in the old system were the Departments of Water Affairs

and Health, the Water Services Act transfers the primary responsibility for managing the

quality of water for domestic use to the Water Services Authorities. This has implied the

reorganisation of the whole programme. DWA has had to shed the key responsibility of

managing water quality for domestic uses to the WSAs. This meant transferring staff and

resources to the WSAs and the transition is not yet completed and at the moment that is

affecting water quality monitoring on the ground. As indicated in Figure 6 DWA at regional

office level has three sections charged with water quality monitoring and management (Water

Services, Water Quality Management and Geohydrology). The Water Quality Management

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 171

section is headed by an Assistant Director. The Assistant Director in the case of Polokwane

Regional Office is incharge of two Water Management Areas i.e. Limpopo WMA and

Luvuvhu/Letaba WMA. The section is responsible for monitoring quality in surface waters

and there are only ten officers to cover the whole of Luvuvhu/Letaba WMA. Luvuvhu River

only, has 31 points to be monitored at least once a month.

The inadequate manpower has a serious impact on the quality of the service being

rendered. The officers are not able to collect data every month due to other work

commitments and therefore a lot of gaps exist in the data collected over the years. For

example, for the period between July 2005 and June 2007 only four stations out of thirty one

had complete sets of data for a year. Although the data was computerised and could be easily

accessed by making a request through internet, it was not analysed. This brings into question

the usefulness of the data collected as a tool for community interventions and decision

making.

The geohydrology section in DWA Polokwane Regional Office is responsible for

monitoring and management of groundwater quality. According to Mr. WH du Toit the

Information Manager in the Water Resource Directorate at the Polokwane Regional Office,

the National Groundwater Quality Monitoring Program has 55 monitoring stations in

Limpopo Province which includes Luvuvhu Catchment and the data is available on Water

Management System (WMS). The data obtained from DWA after request through internet

also suffered from lack of completeness. For example, there is hardly a point with a complete

set of data for a year on any given water quality monitoring parameter between 1995 and

2007. Some of the points were only monitored for only one month in a year. Consequently

the data available over the internet is not analysed meaning that only experts can make use of

it.

As already indicated the primary responsibility of monitoring quality of water for

domestic use lies with WSAs. The role of DWA through the Water Services section is to

support and regulate the WSAs. Therefore the district structure for DWA has remained with

skeletal support staff just to support the work of the WSAs. In Luvuvhu Catchment the WSAs

are Vhembe District Municipality which treats water and Thulamela and Makhado Local

Municipalities which distribute the water to consumers. The water quality function is

supposed to be carried out by the district municipality but Vhembe District Municipality has

not completed setting up the water services structures and therefore DWA is still responsible

for that function.

The water quality programme in Vhembe is carried out by staff based at four water

treatment plants in the catchment by taking samples and forwarding them to a laboratory

located in Sibasa. The samples are taken from the raw water supplying a treatment plant, from

the treatment plant itself and from a designated point just outside the treatment plant. Other

sampling points are situated along the distribution line for every 10 000 people using the

system. Water can get recontaminated along the distribution system (McGhee, 1991).

Therefore the above arrangement cannot guarantee that the consumers are receiving safe

water. The laboratory caters for other local municipalities under Vhembe district but under

different catchments such as Nzhelele and Sand. It has employed 8 technicians and according

to Mr. Nemaunzeni (Control Engineering Technician) the laboratory is coping with samples

from the whole district. This has made the sampling points to be very limited. If more

samples were to be randomly added from community level, the laboratory may fail to cope

with the workload.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

172

The Department of Health through its Environmental Health section is tasked with

monitoring water quality at the point of consumption and giving health and hygiene education

at community level from the Drinking Water Quality Management Guide for Water Services

Authorities (2005). The Environmental Health Services are currently in transition. They are

being moved from the Department of Health to Vhembe District Municipality. At the moment

an operational structure has not been worked out for the services but staff has just been

seconded to work in areas under the local municipalities.

There are seven Environmental Health Officers (EHO) deployed under Thulamela

Locality Municipality which covers most parts of Luvuvhu Catchment to serve an estimated

population of 800 000 people (Vhembe IDP, 2009). This translates to a ratio of 1 EHO to 114

286 people to be served. The World Health Organisation recommends a ratio of 1 EHO to 10

000 people but South Africa relaxed this to 1 EHO to 15 000 people (Agenbag and Gouws,

2004). Figure 2 gives a comparison of functional Environmental Health Officers per

population in South Africa between 2006 and 2007. Limpopo Province had 1 EHO for 32 000

people, which is more than the nationally recommended staffing levels. But the current

staffing situation for the areas under Thulamela Local Municipality is far much worse as

shown above. This has serious implications on the ability of the EHOs to carry out their

mandate of monitoring water quality at the point of consumption and giving health and

hygiene education to the community. The EHOs also have other duties in addition to water

quality monitoring including; malaria control, food control and sanitation.

The EHOs themselves on the ground do not see water quality monitoring as a priority.

One of the EHOs under Vhembe District Municipality Mr. Matshakatini said they only

prioritised water quality monitoring during outbreaks of waterborne diseases such as cholera.

He said when they were still under DoH they only monitored water quality when ordered to

do so by their superiors but currently there is confusion since they do not have a structure yet

to report to.

Figure 2. Comparison of functional Environmental Health Practitioners (EHP) per population in South

Africa (2006 – 2007) (Agenbag, 2008).

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 173

There is very limited contact between the EHOs and the rural population they serve. Five

of the seven EHOs under Thulamela Municipality operate from Thulamela Municipality

offices in Thohoyandou and the other two are based in Malamulele. There are no officers

based at clinics which are at community level. With such limited interface with communities,

their ability to influence behaviour change towards water quality management through

delivery of health and hygiene education is in doubt.

Although DWA has set up a water quality information system as required by the National

Water Act and the information is readily available on the internet or through request to the

department, the quality of the information and access by all the members of the public

remains a concern. As already indicated, a lot of the data sets are not complete and the

information is kept in the form of raw data which can only be useful to professionals who are

able to analyse and interpret it.

DWA has put into place an audit system to monitor the water quality monitoring function

of WSAs. Mr. Nemaunzeni who is in charge of the laboratory at Sibasa confirmed that once a

month a private laboratory takes samples from designated water sampling points and analyses

them for quality control purposes.

From the foregoing discussion, it is clear that although water quality monitoring and

management is well organised on paper in South Africa, on the ground the programme is not

likely to satisfy the provisions of the national constitution and ensure that every citizen in

Luvuvhu Catchment is protected from using unsafe water. This is why waterborne diseases

are still common. Therefore there is need for authorities to be innovative and bring other

stakeholders including local communities on board.

4.2. Indigenous Knowledge and Community Perceptions Relating to Water

Quality Monitoring

According to the Health Belief Model, for an individual to undertake recommended

preventive health action, he/she needs to perceive a threat to his/her health, be simultaneously

cued to action and his/her perceived benefits outweigh his/her perceived losses. Therefore for

one to successfully mobilize communities in Luvuvhu Catchment to participate in water

quality monitoring and management, one needs to understand their perceptions of water

quality related threats/risks to their health. The study sought to evaluate indigenous

knowledge and community perceptions in the catchment and utilized them as a basis for

mobilizing the communities to participate in water quality monitoring and management

projects.

There was a wealth of knowledge relating to water quality management among the

communities in the catchment as provided in Table 4. The communities knew how water got

polluted, with a high percentage of the groups (ranging from 91 – 99% mean = 93; Sample

standard deviation (S) = 2.4) saying water got polluted through the introduction of foreign

materials. The percentage of the groups who said they would recognize that water had been

polluted through its physical appearance ranged from 61 – 68% (mean = 60; S = 1.5). The

percentage of the groups that said they would employ a combination of conventional and

traditional methods to treat water ranged from 79 – 89% (mean = 85; S = 2.9). An example of

a conventional method mentioned was the use of chlorine products to disinfect water at

household level, while the traditional methods included putting tree branches with fresh

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

174

leaves in the water to be treated, allowing the water to settle for a while before using it,

filtering water through sand and boiling water before using it. The communities were aware

that poor home hygiene could contaminate water, with the percentage of the groups

participating in the study who confirmed this ranging from 83 – 89% (mean = 86; S = 2.3).

For the groups that said they would stop using a source and report to the local leadership if

they suspected that the source was polluted the percentages ranged from 69 – 77% (mean =

73; S = 2.4). The communities were also aware of the fact that consumption of polluted water

affected the gastro – intestinal system with the percentage of the participating groups saying

this ranging from 81- 88% (mean = 86; S = 2.4).

The communities generally said they would reject using water for various purposes based

on the aesthetic quality of the water more than any other reason. For drinking water, 64 –

71% (mean = 66; S = 2.3) said they would reject it for aesthetic reasons and 71 – 79% (mean

= 75; S = 1.9) said they would reject water for bathing for the same reason. For cooking, 61 –

67% (mean = 63; S = 2.3) said they would reject water for aesthetic reasons while 82 – 89%

(mean = 85; S = 1.8) said they would do the same for laundry purposes.

The community’s perception of water safety was evaluated using turbidity as an indicator

of poor quality in water. Although the same communities had said they would recognize that

water had been polluted through its physical appearance, their perception of unsafe water

using physical appearance (turbidity) as an indicator differed significantly from acceptable

scientific value. For example, about 35% of the participants said they would use water for

drinking with turbidity of 92 NTU. This is way above the maximum limit of 5 NTU

recommended in SANS 241. According to South African Water Quality Guidelines (Volume

1: Domestic Water Use) any water with a turbidity value above 10 NTU carries an associated

risk of disease due to infectious disease agents and chemicals adsorbed onto particulate

matter. This study demonstrated the point made in the Health Belief Model in that, now that it

is clear the communities in Luvuvhu are at risk because they tolerate water with very high

turbidity, the health and hygiene education by EHOs can be effective in changing this

behaviour.

Since the communities felt that the reticulated water supply was not safe in terms of

smell, taste, and physical appearance, they were advised to treat water at the point of use with

simple technologies such as using chlorine compounds or boiling water before use in the

short term. In the long term, the communities were advised to engage authorities at the

treatment plants so that the treatment processes can be improved upon.

4.3. Community Participation in Water Quality Monitoring and

Management

South Africa has very comprehensive policy, legal, technical and institutional

frameworks that are either meant or can be used to promote community participation in water

quality monitoring and management.

The various policies and pieces of legislation discussed in chapter two encourage and

facilitate participation of communities in development activities including water quality

monitoring and management. The National Constitution guarantees the citizens the right to be

consulted in any activities that affect them and since water quality affects public health it

becomes imperative that they be consulted on it.

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 175

South African Water Quality Guidelines (Volume 1: Domestic Water Use) give a

tabulated relationship between the concentration of a water quality constituent and the

possible risk to human health as illustrated for pH and turbidity in Tables 1 and 2

respectively. This tabulation makes it possible and easy to involve members of the public in

water quality monitoring and management. For example if members of the community were

trained in the use of simple equipment such as pH and turbidity meters and given Tables 1

and 2 they would be able to take samples, analyse them and interpret the results for

themselves.

Table 1. Effects of pH on Aesthetics and Human Health

pH Range (pH

units) Effects

<4.0 Severe danger of health effects due to dissolved toxic metal ions. Water tastes sour

4.0-6.0 Toxic effects associated with dissolved metals, including lead, are likely to occur at

a pH of less than 6. Water tastes slightly sour

Target Water

Quality Range

6.0-9.0

No significant effects on health due to toxicity of dissolved metal ions and

protonated species, or on taste are expected. Meta ions (except manganese) are

unlikely to dissolve readily unless complexing ions or agents are present. Slight

metal solubility may occur at the extremes of this range. Aluminium solubility

begins to increase at pH 6, and amphoteric oxides may begin to dissolve at a pH of

greater than 8.5. Very slight effects on taste may be noticed on occasion

9.0-11.0

Probability of toxic effects associated with deprotonated species (for example,

ammonium deprotonating to form ammonia) increases sharply. Water tastes bitter

at a pH of greater than 9

>11.0 Severe danger of health effects due to deprotonated species. Water tastes soapy at a

pH of greater than 11

Table 2. Effects of Turbidity on Aesthetics and Human Health

Turbbidity

Range (NTU) Effects

Target Water

Quality

Range 0-1

No turbidity visible

No adverse aesthetic effects regarding appearance, taste or odour and no

significant risks of associated transmission of infectious micro-organisms. No

adverse health effects due to suspended matter expected

1-5

No turbidity visible

A slight chance of adverse aesthetic effects and infectious disease transmission

exists

5-10

Turbidity is visible and may be objectionable to users ar levels above 5 NTU.

Some chance of transmission of disease by micro-organisms associated with

particulate matter, particularly for agents with a low infective dose such as

viruses and protozoan parasites

>10

Severe aesthetic effects (appearance, taste and odour).

Water carries an associated risk of disease due to infectious disease agents and

chemicals adsorbed onto particulate matter. A chance of disease transmission

at epidemic level exists at high turbidity

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

176

Table 3. Colour coded classification system (Quality of Domestic Water Supplies Series

1998, Volume 1: Assessment Guide)

Blue B

Class 0

Ideal water quality – suitable for lifetime use.

Green G Class I Good water quality – suitable for use, rare instances of negative

effects.

Yellow Y Class II Marginal water quality – conditionally acceptable. Negative

effects may occur in some sensitive groups

Red R Class III Poor water quality – unsuitable for use without treatment. Chronic

effects may occur

Purple P Class IV Dangerous water quality – totally unsuitable for use. Acute effects

may occur.

The Quality of Domestic Water Supplies (Volume 1: Assessment Guide) gives a colour

coded classification system as shown in Table 3, which is created after analysing the quality

of water at a particular locality for a period of time, for example a year or two. The use of

colour is innovative in that even illiterate members of the public can follow and understand

the quality of their water resources and the implication to their health. For example they are

taught that if the colour displayed is blue then it means they can use the water without any

precautions but if it is red then they need to treat the water first and if it is purple then they

should never use the water. This framework would empower communities to seek

explanations from experts and participate in any remedial action.

DWA also publishes Water Services Authority Awareness Pamphlet and a Consumer

Awareness Booklet on water quality issues. These if properly distributed go a long way in

keeping members of the public informed and allow them to contribute to water quality

monitoring and management. The problem is that all these opportunities offered by the

different frameworks are not being utilised at the moment. The data generated from water

quality monitoring activities is reserved for use by researchers and other professionals instead

of being shared with or used for the benefit of the affected communities.

All water quality monitoring programmes run by DWA and the WSAs at the moment do

not involve communities at any stage. DWA does not involve communities in its monitoring

of surface waters in the catchment. According to Mr. William Moasefoa, the former Assistant

Director (water quality) at the Polokwane Regional Office, communities do not even know

where the 31 monitoring points along the Luvuvhu River are located. They do not explain to

the affected communities the reasons for taking water samples. In fact according to him “the

communities think we are members of the Zion Church collecting water for their prayers”.

They do not give feedback to communities concerning the results of the monitoring done

along the river. The same situation prevails in the ground water quality monitoring

programme according to Mr. WH du Toit, the Information Manager in the geohydrology

section at the Polokwane Regional Office.

The monitoring of quality for domestic water supplies done by the DWA district office in

Vhembe district does not involve communities also. According to Mr. Nemaunzeni (Control

Eng. Technician) they do not disclose the designated points where they collect samples to the

communities because “people will be bothering us about why we concentrate on sampling

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 177

while the service itself is not consistent”. The results are not shared with communities but

compiled for the benefit of higher offices. The only time the results are used locally is when

there is a failure in the water quality system in which case the laboratory advises those in

charge of the water supply system to shut down the supply and supply the villages from

tankers. Even when the communities do not understand the explanation they are advised to be

patient while the system is being sorted out.

Communities confirmed in a survey that, they were not involved in water quality

monitoring and management activities in their villages. Although 91 – 98% (mean = 94; S =

2.4) of the groups interviewed said they knew that their water supplies were monitored/tested,

only 3 – 8% (mean = 5; S = 1.5) mentioned communities as participating in the process. The

percentage of the groups that said they got any feedback from the authorities on the results of

the water samples that are taken from their villages ranged from 3 – 9% (mean = 5; S = 2.2).

The communities were aware that there were water quality problems in their areas with 89 –

97% (mean = 93; S = 2.3) of the groups supporting this view. The percentage of the groups

that felt that the water quality problems in their areas needed to be monitored regularly ranged

from 92 – 99% (mean = 95; S = 2.1) and 91 – 98 (mean = 94; S = 2.2) said they are to be

involved in the monitoring. The percentage of the groups that said they were willing to

contribute financially towards monitoring of their water ranged from 87 – 97% (mean = 93; S

=3) and they were willing to pay amounts ranging from at least R10 – R150/month towards

this. More people were willing to pay R20/month, with the percentage ranging from 51 – 75%

(mean = 55; S =1.8). Communities held meetings to discuss development including water

quality issues as confirmed by 94 – 98% (mean = 96; S = 2.3) participating groups. The

percentage of the groups that said the water quality issues discussed at these meetings were

related to water quality monitoring and management ranged from 70 – 78% (mean = 74; S =

2.6). The communities acknowledged that they got external assistance whenever they had

problems with water quality, with 88 – 96% (mean = 92; S = 3.4) of the participating groups

confirming this. The percentage of the groups that said the external assistance came from

government/municipality system ranged from 62 – 72% (mean = 67; S =2.9) while those that

said the assistance came from NGOs ranged from 29 – 38% (mean = 33; S = 1.9).

In terms of the Health Belief Model, the communities in Luvuvhu Catchment appreciate

risks of using polluted water, appreciate the benefits of monitoring water quality and are

ready to take action to eliminate the risk. This could be the appropriate community where the

government could start experimenting with meaningful involvement and participation of

communities in development activities including management and protection of water

resources. The government can start experimenting with bottom – up approaches that can lead

to true empowerment of the communities in managing and protecting water resources in

Luvuvhu Catchment. This is expected to bring about meaningful participation of communities

in water resources management.

Luvuvhu Catchment falls under the Luvuvhu/Letaba WMA and the CMA for this WMA

is in the process of being formed. The process of setting up the water management institutions

as is underway. A proposal for its establishment was produced in early 2010 and it includes

structures that will in theory promote the participation by communities from grass root level

to WMA level (Figure 3). The CMA will take over most of the functions currently being

performed by the DWA Regional Office in Polokwne except for the regulatory ones.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

178

Figure 3. Proposed catchment management structures for Luvuvhu/Letaba Water Management Area.

The immediate problem with the proposed structure (Figure 3) is that, it does not connect

with local government administrative structures such as municipalities, wards, villages and

sub villages. This means that it cannot articulate the feelings and perceptions of the common

public but concentrate on those who are considered to be major water users such as farmers,

industry, etc. The second weakness lies on how the structures were established. DWA

contracted a private company to conduct “public participation” meetings and then establish

the structures. The contract was for six months with predetermined outputs. The schedules

were very tight and the private company had little time to deliver. Therefore the whole

objective on the part of the contracted company became meeting the targets instead of

allowing communities to understand the whole concept and then making suggestions on how

the whole issue should be conducted. At the end most of the meetings were poorly attended

and major decisions were taken by very few people on behalf of the whole community.

The next problem relates to sustainability. When the private company completed its work

and left the connection between the community and DWA was broken since there was no one

to follow up on issues that were raised in the meetings.

Figure 4. Local government structures in Luvuvhu Catchment.

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 179

The introduction of structures that will change fundamentally the way government

engages with communities would have required protracted, committed and sustainable

communication with communities to allow them time to adapt and adjust to the changes.

Instead of contracting out the task to private companies the government should have allowed

the decentralisation processes of water services to Vhembe District Municipality to be

completed and then build capacity within the proposed Institutional Social Development

(ISD) unit at that level.

The municipality would have the capacity to carry out the task since its structures are

already linked to the communities as illustrated in Figure4. Locally based NGOs could have

been involved in mobilizing the communities and establishing the structures since these

would remain for a long time in the community. They would offer the communities support

as they adapt and adjust to the changes.

4.4. Water Scarcity and Failure by Service Providers to Supply Adequate

Amounts of Water

Although Luvuvhu Catchment is viewed as the only catchment in South Africa that

potentially has surplus water, the water is not adequate to satisfy all users. The Thulamela

Local Municipality IDP (2008), showed there was a backlog of 93 121 households without

access to reticulated water supplies in the year 2008. When these households are finally

connected to the water supply system the surplus water is likely to be reduced. Luvuvhu

Catchment is already transferring water to Louis Tritchardt in Sand Catchment and some

areas in Letaba Catchment.

Studies carried out by students from the University of Venda in the catchment including

those by Masidiri (2008); Nemadodzi (2008); Malume (2010); Mvundlela, (2010) have

shown that at household level communities experience severe water shortages. The

communities only receive water for a limited number of days in a week and sometimes go on

for weeks without getting any water. Mvundlela (2010) determined that the per capita use of

water in Tshidembe Village which is part of the catchment was 17 l/c/d and was lower than

the figure given in the RDP document of 25 l/c/d. Although the three treatment plants

supplying water to the population within the catchment had large design treatment capacities,

their treatment efficiencies were relatively low and not adequate for the target population. For

example, the Mhinga treatment plant has a design capacity of 3 mega litres (ML) but its

treatment efficiency ranges from 30 to 90% (900 - 2700 m3) with a mean of 60% (Figure

5).Therefore, at its lowest it produces 900m3/day.

According to the DWAF Limpopo Water Services Regional Bulk Infrastructure Grant

document for the years 2008/2009 and after, the design domestic consumption levels for

different communities is as shown in Table 12. Using the guidelines for basic service (16

l/c/d, 25 l/c/d and 35 l/c/d), the minimum water demand for the villages under Mhinga

treatment plant would be 800, 1250 and 1750 m3/day respectively. At 30% efficiency, the

plant “produces” 900m3/d, barely enough to cover a supply required for survival (16 l/c/d) of

50000 people for a day (800 m3/d) especially considering that there are always unaccounted

for water losses along the distribution line. If the supply is operated at standard level (25

l/c/d) which is supported by the Reconstruction and Development Programme (RDP)

document the supply deficit becomes even bigger.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

180

Table 4. Design domestic consumption levels: DWAF (2008)

Service level

Scenario

Domestic consumption l/c/d

Basic Higher Urban

Survival 16 50 70

Standard (DWAF) 25 90 150

Higher 35 120 200

l/c/d = litres/capita/d.

Figure 5. Water treatment efficiency at the Mhinga treatment plant.

As the daily water demand exceeds the supply, the authorities resort to water rationing

tactics to even out the demand. Each sub village is supplied with water for less than 5 days in

a week. This forces communities to resort to the use of other sources including the

unprotected ones.

A community survey carried out in the area also confirmed the fact that not all people in

the area had access to a reticulated water supply all the time. The percentage of the groups

that confirmed that they sometimes used water from a combination of protected and

unprotected sources for drinking ranged from 20 – 29% (mean = 24; S = 2.6) while 67 – 75%

(mean = 68; S = 17) and 61 – 69% (mean = 64; S = 16) said that they used a combination of

protected and unprotected sources for bathing and cooking respectively.

Failure to access adequate water of sufficient quality increases community vulnerability

to water quality related problems. When people are not supplied with adequate amount of

water, they turn to other alternative sources of water which may cause diseases (Dungumaro,

2007). Each year about 4 million people die of diarrhea and more than 800 million people in

the world are malnourished due to insufficient water (Cosgrove and Rijsberman, 2000).

4.5. Exposure to Polluted Water

The second source of vulnerability faced by the communities in Luvuvhu Catchment

relates to the potential use and exposure to polluted water. Although literature on pollution

problems in the catchment is limited, review of available information relating to land use

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 181

patterns in the catchment showed a high potential for water pollution occurring in the

catchment. The upper part of the catchment is dominated by irrigated agriculture which uses

fertilizers and pesticides with high potential for polluting water. The middle part of the

catchment is dominated by human settlements. Most of the settlements are rural and use pit

latrines for their sanitation.

The high density of pit latrines poses a threat to the quality of groundwater in the

catchment. Contamination of groundwater with human waste can raise nitrate levels in water.

Figure 6 shows the results of groundwater quality monitoring from four points located in

various parts of the catchment over two years. The nitrate levels in some stations at some

points exceeded even the maximum allowable limit for nitrates given in SANS 241 of 20

mg/l. SANS 241 require that nitrate – nitrite concentration in water should remain below 10

mg/l under Class I limits and range between 10 and 20 mg/l under Class II limits provided the

consumers are not exposed for more than 7 years.

Nitrate levels from four stations along the Luvuvhu River monitored for a period of two

years also exceeded both the minimum and maximum limits in Thohoyandou and Waterfall

Sewage Outflow in some months (Figure 7).

Figure 6.4Nitrates levels in ground water from four monitoring points in Luvuvhu Catchment (2005 –

2007).

Figure 7.5Nitrate levels from four stations along Luvuvhu River (2005 – 2007).

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

182

In Thohoyandou, the nitrate level was always above the minimum acceptable limit. This

could be due to the influence of irrigated agriculture in the upper part of the catchment,

contribution from the sewage treatment facilities upstream of Thohoyandou or contamination

from groundwater. Groundwater can carry nitrogen (in the form of nitrate) into surface water

bodies through recharge and spring discharges (West, 2001).

Turbidity values in all the four stations analysed, were mostly found to be way above

even the maximum limits given in SANS 241 of 5 NTU (Figure 8). This could be attributed

to anthropogenic activities in the catchment such as agriculture and human settlements (urban

and rural) which have altered land cover and led to generation and introduction into water of

particulate matter that causes high turbidity in water. The middle part of the catchment has

two peri - urban centres of Thohoyandou and Malamulele and these have the potential of

causing pollution problems associated with raised turbidity values in water courses. Urban

settlements contribute towards water pollution through alteration of land cover, urban

drainage, phosphorous and sewage generation. The paved surfaces in urban areas tend to

increase the production of dissolved salts, suspended solids, and nutrients (Ahearn et al.,

2005). Sewage facilities can also discharge sewage that is not adequately treated and raise

turbidity levels in water.

Figure 8. Turbidity values from four stations along the Luvuvhu River (2005 -2007).

Figure 9. Ammonia (NH

) levels from four stations along Luvuvhu River (2005 – 2007).

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 183

Figure 10. Results of E. coli monitoring along Luvuvhu River (2005 – 2007).

Results of Ammonia monitoring from two stations along Luvuvhu River (Thohoyandou

and Vuwani Oxidation Ponds were way above the maximum limit for ammonia given in

SANS 241 (Figure 9). The recommended limits for ammonia in SANS 241 are 1 mg/l Class I

operational limits and 1 – 2 mg/l for Class II limits despite the duration of exposure.

Concentrations exceeding 10 mg/l are found in raw untreated sewage, since ammonia

concentrations tend to be elevated in waters where organic decomposition under anaerobic

conditions takes place. The sewage facilities at Vuwani could be malfunctioning and

releasing raw sewage into the river. The areas around Vuwani and rural areas upstream of

Thohoyandou have livestock which could also contribute to the high levels of Ammonia in

the water.

The discharge of raw sewage into the water could also account for the high numbers of E.

coli indicator bacteria recorded at the Thohoyandou monitoring station (Figure 10).

According to SANS 241, 95% of all water samples (count/100ml) analysed should yield no

count of E. coli for water to be declared safe.

Most of the samples from Thohoyandou monitoring point show viable counts of E. coli

up to over 350000/100ml. This shows heavy pollution with material containing faecal matter,

strengthening the suspicion that the sewage ponds upstream at Vuwani are discharging raw

sewage into the river. The water needs to be treated properly because there is a possibility of

pathogens passing through treatment and disinfection processes and posing a danger to public

health.

The COD values at all the four stations monitored along the Luvuvhu River mostly

exceeded by far the maximum limit given in SANS 241 (Figure 11). COD is a component of

Dissolved Organic Carbon (DOC).

According to SANS 241 DOC concentration in water should not exceed 10 mg/l for Class

I and range from 10 to 20 mg/l under Class II limits provided the consumers are not exposed

for more than 3 months. The results from the four stations indicate a possibility of serious

pollution from a rich source of organic carbon.

This further strengthens the possibility that sewage facilities along the river are

discharging raw sewage into the Luvuvhu River. The treatment system needs to be thorough

and reduce the COD to manageable levels. Excessive COD not only causes problems with

aesthetics in the treated water but could promote the formation of trihalomethanes (THMS)

which are carcinogenic substances.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

184

Figure 11. Results of COD monitoring along Luvuvhu River at four monitoring sites (2005 – 2007).

Review of literature and official documents revealed that there were other potential

sources of water pollution besides the ones discussed above. Roadsides and rooftops from the

two urban centres were implicated as sources of metals and salts which could pollute water

when it rains.

The lawns and other garden trimmings from the urban centres contribute nutrients such as

phosphorous into the water courses. The geology of the catchment has the potential to

contribute to pollution of water. Literature reviewed by Odiyo et al. (2009) showed that, the

geology of the Soutpanseberg Mountains contains minerals such as copper, iron, refractory

flint, salt, sillimate and coal. Since the Luvuvhu River starts from this mountain, it means

there is a potential for the minerals to pollute the water in the river. While Tshikondeni coal

mine may not be causing any pollution problems at the moment according to DWA (2004b),

literature has shown that mines can cause problems when they cease to operate (van Zyl,

2002).

Review of literature also revealed that the common practice by local communities to dig

and open up the ground to obtain materials such as quarry stones and pit sand can cause water

pollution. Oxidation of sulphide minerals, especially pyrite, during weathering can cause Acid

Mine Drainage (AMD) and release iron (Fe), sulphate (SO4), acid (H+) and trace elements,

such as nickel (Ni) and arsenic (As), into the aquatic environment (Pope et al., 2006b).

Masidiri (2008) found that water from boreholes in the Tshimbupfe area in the catcthment

had very high E. coli count and suggested that, this could have been caused by poor water

point hygiene. The boreholes were not fenced off from animals and they could have licked the

spouts and contaminated them. But critically, this shows that water at the point of collection

is vulnerable to contamination putting the health of the community at risk.

4.6. Weaknesses in the Contemporary Water Quality Monitoring and

Management Practices

The third source of vulnerability emanates from the weaknesses in the contemporary

water quality monitoring and management practices. While the quality of water from the

treatment plants was generally acceptable, treatment failures did occur during the rainy

season. For example, analysis of treatment efficiency data from the Mhinga plant between

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 185

January 2010 and February 2011 shows a failure in the removal of turbidity in treated water

(Figure 12). Turbidity in treated water in December 2010 reached 22 NTU and exceeded by

far the allowable limit of 5 NTU given in SANS 241.

Turbidity has significant implications on human health. Excessive turbidity or cloudiness

in drinking water is aesthetically unappealing and may also represent a health concern (Fox

and Tversky, 1995). It can provide food and shelter for pathogens and can promote their

regrowth in the distribution system, leading to waterborne disease outbreaks (Fox and

Tversky, 1995). Although turbidity is not a direct indicator of health risk, numerous studies

show a strong relationship between removal of turbidity and removal of protozoa. There is a

relationship between turbidity removal and pathogen removal. Low filtered water turbidity

can be correlated with low bacterial counts and low incidences of viral diseases (LeChevallier

et al., 1991). Positive correlations between removal (the difference between raw and plant

effluent water samples) of pathogens and turbidity have also been observed in several studies.

Data gathered by LeChevallier and Norton in 1992 from three drinking water treatment plants

using different watersheds indicated that for every log removal of turbidity, 0.89 log removal

was achieved for the parasites Cryptosporidium and Giardia.

The particles of turbidity can also provide “shelter” for microbes by reducing their

exposure to attack by disinfectants. Microbial attachment to particulate material or inert

substances in water systems has been documented by several investigators and has been

considered to aid in microbe survival (Marshall, 1976).

Even the communities expressed dissatisfaction with various issues surrounding the

quality of their reticulated water supplies. While the percentage of the groups that said the

taste of the water from taps was very good ranged from 51 – 59% (mean = 54; S = 2.4), 41 –

49% (mean = 45; S = 2.2) said the taste of the water was just good, meaning there was a

significant proportion of the communities that were not completely happy with the taste of the

water.

The same applied to the question of smell of the water where 43 – 49% (mean = 46; S =

1.8) of the groups were not entirely happy with it saying it was just good. The percentage of

the groups that said the water had floating particles soon after collection ranged from 72 –

79% (mean = 75; S = 2.2) and 68 – 78% (mean = 74; S = 2.4) said the containers where the

water was kept were stained brown after some time.

Figure 12. Turbidity removal efficiency at the Mhinga treatment plant.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

186

Although most of the results from a monitoring exercise for microbiological quality from

street taps in selected villages included in the study were satisfactory, there were occasional

failures for both Total coliform and E. coli counts in some of the villages.

According to SANS 241, 95% of all samples (count/100ml) during monitoring should

have no sign of E. coli. In some villages E. coli counts were recorded in more than 5% of the

samples.

Since the results from most of the villages were satisfactory, it would suggest that the

problem is not with the treatment plants but water could be getting contaminated at the point

of collection or along the distribution pipes. The street pipes are not fenced off and allowing

animals to lick the taps. This could compromise the hygiene of the tap and contaminate water.

This serves to emphasize the need for water quality to be monitored at the point of collection

to reduce the risk of communities getting sick from consuming water that has been

contaminated after treatment.

Statistics obtained from four clinics in the study area, suggest a correlation between the

rainy season and the increase in the incidences of diarrhoeal diseases (Figs. 13; 14; 15; 16).

Although there was no official threshold to declare an outbreak when diarrhoea cases

exceeded a certain value, the shapes of the graphs drawn from the statics obtained from the

clinics showed that there was an enormous rise in cases during the rainy season. From 2008 to

2010 the peaks of the graphs were always between September and January or between

January and April. This further confirms that the health of the public is under threat from

waterborne diseases.

It suggests that people are using water from unprotected sources such as streams and

pools during the rainy season when surface water is easily available. The statistics confirm

the findings of the community survey that said between 20 and 29% of the people used a

combination of protected and unprotected sources of water for drinking purposes sometimes.

The diarrhoea cases could not be attributed to food sources because according to the Health

Statistics Manual from the Department of Health, food poisoning can only be suspected when

four members of a family fall ill at the same time. The cases are recorded and reported

separately according to Ms Khangale, the Health Information Officer at the Vhembe district

Department of Health offices.

Figure 13. Health statistics from Mphambo Clinic (2008 -2010).

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 187

Figure 14. Health statistics from Mavambe Clinic (2008 – 2010).

Figure 15. Health statistics from Tshikonelo Clinic (2008 – 2010).

Figure 16. Health statistics from Xigalo Clinic (2008 – 2010).

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

188

4.7. Fragmented Structure of the Decentralised Health Services

The other source of vulnerability is related to the fragmented decentralised structure of

the health services in Vhembe district. The Environmental Health Services who are

responsible for investigating outbreaks of diseases including waterborne diseases have been

decentralised to Vhembe District and its local municipalities. The clinics which are the ones

that collect statistics of diseases have remained with the Department of Health. According to

Mr. Mukwevho a Senior Environmental Health Officer at Vhembe District Municipality,

there is a problem with the flow of statistics between the DoH and the municipality.

Sometimes the DoH sends the statistics to their Head Office and Environmental Health

Services can only access it through the mother Department of Local Government. This creates

serious time lapses between an outbreak occurring and the reaction time by Environmental

Health Services and increases community vulnerability to water quality problems.

CONCLUSION

The first objective which relates to providing proof that water quality problems exist in

Luvuvhu Catchment and are worth receiving attention from the authorities was achieved. The

potential water pollution sources in the catchment include the geology of the catchment,

irrigated agriculture in the upper part of the catchment, rural human settlements in the middle

part of the catchment, livestock and wildlife in the middle and lower parts of the catchment,

mining activities, motor vehicle emissions and peri – urban areas of Thohoyandou and

Malamulele with all the water pollution associated activities such as generation of both solid

and liquid waste. The results obtained from DWA stations along the Luvuvhu River

confirmed that water pollution was a threat to the health of the people who depended on the

Luvuvhu River system for water supply. Parameters such as turbidity, nitrates, COD and E.

coli exceeded by far the safety limits given in SANS 241. Results of ground water quality

monitoring also confirmed this position with nitrate levels exceeding the limits in SANS 241.

Results of a monitoring exercise conducted on selected street taps in the 15 study villages also

confirmed the vulnerability and risks faced by the communities when some of the taps failed

the E. coli tests. People resort to using unprotected surface water sources which become

abundant during the rainy season making them vulnerable to waterborne diseases such as

diarrhoea. The study proved that communities in the catchment are vulnerable to water

quality problems and that water quality problems in the catchment are worth receiving the

attention of authorities and researchers.

The second objective was to analyse the water quality monitoring frameworks in South

Africa and identify any gaps. Although South Africa has a well developed and extensive

policy, legal and technical framework relating to water quality monitoring and management,

Luvuvhu Catchment has not developed the institutional capacity to adequately monitor and

manage the quality of water resources in the catchment. The water quality monitoring

services are in transition with key functions and staff being moved from DWA and DoH to

Vhembe District Municipality as the Water Services Authority. This transition is currently

affecting the level and quality of services offered because Vhembe District Municipality has

not developed its structures to carry out this function. Although DWA still assists with

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 189

monitoring the quality of water for domestic purposes, they only monitor the quality of raw

water just before it enters the treatment plant and the quality of treated water at designated

points after it leaves the treatment plant. This is not adequate to guarantee the safety of the

consumers since water can get contaminated during distribution and needs to be ultimately

monitored at the point of use.

Environmental Health Officers who are supposed to monitor water quality at point of use

are currently in transition and are not able to carry out this function. Although DWA monitors

water in the Luvuvhu River and ground water in the catchment, the monitoring is mostly

irregular with many months going unmonitored in a year. The data generated from the

monitoring activities although readily available on internet, is not analysed and not easily

accessible to members of the public. The water quality programme in Luvuvhu Catchment

currently fails to fulfil the requirements of the national constitution that every citizen has a

right to a safe environment.

The partial decentralisation of health services from government to the municipalities is

increasing community vulnerability to water quality problems. The fragmented health

structures where the clinics are run by the Department of Health while the Environmental

Health Officers who are responsible for investigating incidences of disease outbreaks are in

the municipalities make it impossible for the system to react timeously when poor water

quality causes problems within communities.

The third objective was to evaluate indigenous knowledge, attitudes, practices and

perceptions and community participation in water quality monitoring and management. The

communities felt that water pollution was a problem in their areas and that water quality

needed to be monitored regularly. They had knowledge relating to various aspects of water

quality such as how pollution of water happens, how to identify water pollution and improve

water quality. Although they said they would recognise that water had been polluted through

its physical appearance, their perception of water safety based on different degrees of

turbidity did not match the expected scientific requirements. The communities accepted to use

water with turbidity values as high as 92 NTU and yet the recommended maximum limit of

turbidity in water for domestic use in SANS 241 is 5 NTU. These communities lack

knowledge on water safety and are exposed to the risk of contracting diseases. They need

education to raise their awareness to a level where they can make at least scientifically

acceptable judgements.

The study established that, there is no meaningful participation by communities in the

catchment in water quality monitoring and management despite the fact that there is

comprehensive policy, legal and technical framework that could promote community

participation. The institutional framework and the way water quality monitoring and

management is carried out at the moment do not promote participation by communities. A top

– down approach is employed in water quality monitoring and management with the

provision of the service being viewed as purely technical domain to be carried out by the

experts only.

The current practice where DWA engages private companies as consultants to engage

communities in development issues e.g. establishment of CMAs should be discouraged

because it is not sustainable and does not deliver the desired change at community level.

Private companies are profit oriented and do not allow for the time required for communities

to move through all the phases in the resistance to change continuum. Therefore as they pull

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

190

out the initiative collapses since communities will have nobody to continue helping them

building confidence in the initiative.

The current approach to water quality monitoring and management in the catchment does

not promote an integrated approach. The programme is implemented vertically by DWA

mostly and at times Department of Health and the municipalities are involved. Other critical

players like communities and government department which deal with land use related

functions such as Departments of Agriculture and Environment are not involved. This has

serious implication of sustainable use of water resources since it is the land uses that impact

on the quality of water.

The extension staff involved in water quality monitoring and management view

communities as passive recipients of the service who have nothing to contribute in terms of

ideas and materials. Although terms like stakeholders are used in the water sector in South

Africa to refer to communities, they are just words while the practical reality on the ground is

completely different.

Communication with communities is monologic (one sided) with “experts” depositing

their new and educated ideas into the “empty minds” of communities without expecting any

feedback from them. Although community participation has been adopted in policy and legal

documents and accepted as the official approach to development, the current implementation

system does not provide for dialogue where the communities can also make their own

contribution or seek clarification at least. For example, if there is water quality failure from

one of the treatment plants, the staff just responds by shutting down the supply and rectify the

situation without discussing with the communities.

The fifth objective was to assess the reasonableness and acceptability of the proposed

model to the communities. The communities currently feel left out in water quality

monitoring and management activities and they view the proposed model as a way to get

them involved. They view the model from a wider perspective within the democratic agenda

in the country.

The communities are even willing to contribute towards the financing of the proposed

community based water quality monitoring programme. The community leadership believe

that the proposed model will empower communities to participate in water resources

management at that level.

6. RECOMMENDED CONCEPTUAL MODEL FOR COMMUNITY

EMPOWERMENT IN WATER QUALITY MONITORING AND

MANAGEMENT

6.1. Conceptualised Participatory Community Based Water Quality

Monitoring and Management Model

In response to the weaknesses in the current water quality monitoring and management

model pointed out in chapter 4, an alternative conceptual community based water quality

monitoring and management model has been proposed and shown to be reasonable and

acceptable. It is anchored on three frameworks which allow it to achieve what the current

model has failed to achieve.

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 191

Figure 17. Community based water quality monitoring and management conceptual model.

The three frameworks are technical, community empowerment and communication

frameworks (Figure 17). The current model is anchored from the top by DWA and the

municipalities. Although the current policy and legal frameworks support community

participation, the institutional framework is not designed to do so. There is no opportunity in

the current model to empower communities through education and skills provision which are

cornerstones of the proposed model.

6.2. Technical Framework

The proposed main features of the technical framework include simple technologies and a

GIS spot map showing flashy spots which require more attention. For example, Figure 18

shows villages in the catchment with stand pipes that recorded some failures during

monitoring for bacteriological quality over a period of eight months. These villages would

therefore receive more attention in monitoring than the other villages that performed well.

Figure 18. GIS spot map showing villages that had stand pipes that failed microbiological tests during

monitoring.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

192

The framework proposes simple technologies to monitor water quality at community

level. The first line of water quality monitoring at community level has been proposed to rely

on human senses such as taste, smell, sight and touch. Hrudey et al. (1999) found evidence

that an aesthetic problem (an unpleasant odour, taste and colour) is usually translated into a

potential health risk. Gingras et al. (1999) showed that the taste of water and its source (lakes,

rivers, groundwater aquifers) influence perception on water quality. If the relationship

between these sensory parameters, their possible causes and the health implications as

outlined in various pieces of scientific literature can be formalised as shown in Table 5, then

communities can be trained to use their natural senses to monitor water quality.

Indigenous knowledge and practice can inform the technologies developed under this

framework. Koocheki (2007) found that indigenous communities have valuable knowledge

systems that can be incorporated into strategies for environmental management and that the

knowledge systems are an essential cultural and technological element of human societies.

The communities in Luvuvhu Catchment depend on physical appearance of water to decide

whether the water is polluted or not. This fact can be built upon to ensure that simple

technologies to measure turbidity levels in water are developed. The practice by the

communities that, whenever they suspect that, there is a problem with their water supply, they

report the problem to the municipality/government, can be built upon to form a community

based surveillance system. The national water quality monitoring and management can be

improved significantly if the communities acted as part of the “alert” system which will draw

the attention of the authorities to trouble spots in time to prevent any disaster.

The major challenge that African countries continue to face according to Kamara (2004)

is how to reconcile indigenous knowledge and modern science without substituting each

other, respecting the two sets of values, and building on their respective strengths. Therefore

in the case of Luvuvhu Catchment, this can be achieved by integrating the dependency of

communities on physical appearance of water to detect pollution into the national water

quality monitoring and management framework.

According to (Hens, 2006) indigenous knowledge is context specific and what works

successfully in one location for one community may not for another. It is unique to a

particular culture and acts as the basis for local decision making in agriculture, health, natural

resource management and other activities and it is embedded in community practices,

institutions, relationships and rituals. Therefore, there is need to develop technologies that

will help perpetuate the uniqueness of the knowledge and practices related to water quality

monitoring and management in Luvuvhu Catchment. This should motivate communities to

participate effectively in water quality monitoring and management.

The second line of monitoring is proposed to use simple equipment and technologies that

support the reasons for the sensory judgements. Instead of measuring particular water quality

constituents, the technologies measure the effects of those constituents in water and confirm

the sensory judgements. Measuring of physical and chemical water quality parameters such as

pH, turbidity and residual chlorine can explain such things as the taste, smell, physical

appearance and sense of touch of water, link it to possible causes and health implications. The

health implications of concentrations/levels of each parameter are given in tabular form under

the South African Water Quality Guidelines (Volume 1: Domestic Water Use) (DWAF,

2002). The technologies used to measure such parameters as turbidity, pH and

microbiological quality (H2S strip test) are relatively cheap and easy to use. Rural

communities can be trained to use them effectively with relative ease. For example the cost of

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 193

pH meter is estimated at R6 579 – 00 as shown in a quotation sent to the University of Venda

by Labotec (Pty) Ltd (Appendix C1).

Therefore if the kits are distributed to Malamulele, Tshikonelo, Xikundu and Mhinga as

the main villages in the study area and the sub villages under each village share a kit then for

pH only about R27 000 – 00 is needed for capital equipment. This is relatively cheap when

compared with treatment of people in hospitals when they fall sick from waterborne diseases

such as cholera, especially considering that sometimes they lose their lives. According to

DWAF (2005), there are various potential sources of funds that can be invested in community

based water quality monitoring and management as shown in Table 18. In addition to that, the

communities indicated their willingness to contribute towards the cost of monitoring and

managing water quality at their own level. This should add to the potential sources of funds

for the proposed program.

Experiences from elsewhere have shown that rural communities, if given necessary

training can handle and use simple technologies to monitor water quality. Community based

water quality monitoring has been experimented on in other parts of the world and the results

have been good and used to influence environmental policy.

Table 5. Summary of common water quality manifestations, possible causes and health

implications ( DWAF, 2005)

Manifestation Possible Causes Health Implications

Sour taste Low pH which could be due to pollution

from anthropogenic activities or addition

of excess chlorine during disinfection

pH below 6 will encourage corrosion of

cadmium which can be toxic to human beings

Bitter taste High pH which could be due to pollution

or addition of excess lime during

treatment of water

pH around 12 encourages corrosion of lead

which has adverse effects on human health

including growth retardation in infants

Soapy touch High pH

Salty taste Chlorides above 400 mg/l Excessive chlorides encourage corrosion of

metals which can be toxic to humans. Will

cause nausea between 2000m/l and 10000

mg/l and cause vomiting at 10000 mg/l and

above

Colour Oxidation of ferrous and manganous salts

by iron bacteria

Aluminium hydroxide floc due to

aluminium concentrations in excess of 0.1

– 0.2 mg/l remaining in water during the

water treatment process

Blooms of cyanobacteria and other algae

in reservoirs

Makes the water unsightly but also could be

due to other toxic chemicals which would be

injurious to public health

Some cyanobacterial products such as

cyanotoxins are also of direct health

significance.

Excessive

turbidity, or

cloudiness

Pollution from various sources including

sewage systems

Blooms of cyanobacteria and other algae

in reservoirs

Turbidity can provide food and shelter for

pathogens

Promotes regrowth of pathogens in the

distribution system, leading to waterborne

disease outbreaks.

Odours Presence of sulphur compounds leading to

the formation of hydrogen sulphide

Presence of decaying organic material

Odours could indicate pollution from

chemicals that could be harmful to human

health

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

194

In the 1990s a rural community in the Philipines worked side by side with researchers,

non-governmental and governmental workers over a five year period to develop science based

indicators for water quality monitoring relevant for developing environmental policy

(Deutsch et al., 1997). The results prompted the Lantapan Municipal Council to incorporate

community based water testing and some of the research findings and recommendations into

their Natural Resource Management Plan. In Umgeni, South Africa a GREEN (Global River

Environmental Education Network) water project involved school children and community

groups in monitoring water quality using simple scientifically valid methods (O’Donoghue et

al., 1993). The methods included the use of litmus paper, to test for pH and then taking action

to remedy the situation e.g. lobbying local administrative structures to build better toilets, step

up health education and rehabilitate wetlands.

The Department of Civil Engineering at the University of Cape Town in South Africa is

currently involved in collaboration with a number of international organisations in a project

called Aquatest Project seeking to develop a combination of low-cost water quality field kit

and a cellphone based data collection tool (Rivett, 2011). The aim of the project is to support

basic drinking water quality monitoring at supply level, as well as the collection of results in a

digital format. Although the implementation of the Aquatest project at Hantam Municipality

(Eastern Cape, South Africa) is in its early stages, it has already provided some valuable

lessons. The most important of these is that there is a clear need for systems that improve the

ability of staff in the field to conduct operational monitoring. However, in order to be

acceptable, systems should be well integrated with existing municipal processes and

structures. Therefore the proposed community based water quality monitoring model in

Luvuvhu Catchment provides an opportunity for staff to improve on their monitoring

coverage at community level since the communities will act as an “extension” of the staff and

provide preliminary surveillance services.

The results of the community based surveillance activities can then be plotted on a GIS

map as done in Figure 18 to help the under staffed and poorly resourced water quality

monitoring system in Vhembe District to focus its surveillance on the trouble spots. The

health and hygiene education can be directed and focussed onto those areas that need it more

than the others.

The community based water quality monitoring and management programme can serve

as part of an “alert system” described below;

− Alert Level I (no significant risk to health): Routine problems including minor

disruptions to the water system and single sample non-compliances (Internal WSA

response only).

− Alert Level II (potential minor risk to health): Minor emergencies, requiring

additional sampling, process optimisation and reporting/communication of the

problem (Internal WSA response only).

− Alert Level III (potential major risk to health): Major emergencies requiring

significant interventions to minimise public health risk (Engagement of a designated

Emergency Management Team).

The community based water quality monitoring and management programme can form

part of Alert Level I (no significant risk to health) which deals with routine problems

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 195

including minor disruptions to the water system and single sample non-compliances requiring

internal WSA response only. When monitoring programmes at community level raise an alert

signal, the WSA responds and investigates further. If there is a potential minor risk they

activate Alert Level ll which deals with minor emergencies requiring additional sampling,

process optimization and reporting/communication of the problem. This also requires internal

WSA response only. If the potential risk to health is a major one, they activate alert level lll

which deals with major emergencies requiring significant interventions to minimise public

health risk. This requires engagement of a designated Emergency Management Team.

6.3. Community Empowerment Framework

The community based water quality monitoring and management model can serve as a

model for empowering communities in development issues. The fact that the communities

can conduct preliminary water quality monitoring using their senses and confirm that with

scientific equipment, boosts their self esteem and gives them confidence when discussing

with the “experts”. When the results from the community based monitoring system are

interpreted against Tables 1, 2 and 5 the communities will not only be able to relate the results

to possible impacts on their health but will be motivated to take up remedial actions to

mitigate against the situation as envisaged in the Health Belief Model.

The possible remedial actions might include treating water at the point of use through

such methods as boiling, disinfection with chlorine products, allowing sedimentation to take

place before using the water or sand filtration. The most important issue is that this model

will be able to provide evidence at a local level that the health of a community is at risk and

as stated by Redding et al., (2000) individual perceptions of risk play a critical role in

influencing the probability of adopting protective behaviour to prevent illness. According to

Dogaru et al. (2009) perceptual ability heavily depends upon the amount of perceptual

practice and experience that the subject has already enjoyed, implying that perception is a

skill that can be improved tremendously through judicious practice and experience. Therefore

by continuously monitoring water quality at their level the communities will improve their

perceptions of risk related to use of water of poor quality and cue themselves to take action as

stated in the Health Belief Model.

The corrective action taken by the community can also include engaging the authorities

on the need to take action beyond the capacity of the community. For example if turbidity

levels in reticulated water are higher than expected, then it could mean that the filtration

process at the treatment plant could be defective. The monitoring results will empower

communities to take up the issue with those responsible for the treatment plant. As stated by

Mackintosh and Colvin (2003) the implementation of well considered, community accepted

drinking water quality management procedures can effectively change an unacceptable water

quality to one that satisfies drinking water specifications.

The community based water quality monitoring and management model can empower the

communities to deal with not only issues relating to water resources management but

environmental management issues in general in line with integrated water resources

management principles. As stated by Ferreyra et al. (2007) integrated water resources

management is one of the major bottom-up alternatives that emerged during the 1980s in

North America as part of the trend towards more holistic and participatory styles of

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

196

environmental governance. It aims to protect surface and groundwater resources by focusing

on the integrated and collaborative management of land and water resources at a watershed

scale.

For example if the pH of a water body is lower than expected, by referring to Tables 5

and 2 the communities can come up with possible causes and sources of pollution which they

can refer to the authorities for further investigation with confidence since they will be backed

by the monitoring results. An environmental management expert wanting to introduce a

programme on soil conservation to a community can start with a project monitoring total

suspended solids (TSS) and work with communities throughout the project cycle. This will

allow the communities to perceive and understand the extent of the problem in their area and

make it easy for the expert to mobilise the communities to take action.

While the use of simple technology empowers communities to monitor the quality of

their water resources, a supportive environment needs to be created for effective

empowerment to take place. An environment that allows the voice of the communities to be

heard needs to be created. Bottom – up development approaches need to be adopted when

dealing with issues relating to water quality monitoring and management. This will lead to

sustainable use of the limited water resources.

As stated by Haddad (2007), sustainability is the capacity to maintain services and

benefits both at the community and institutional levels without detrimental effects even after

external assistance has been phased out. Key aspects of sustainability include empowerment

of local people, self-reliance and social justice (Haddad, 2007). These reflect principles of

equity, accountability and transparency. One way to incorporate these principles into real-life

management is to move away from conventional forms of water governance, which have

usually been dominated by top-down approaches, and professional experts in the government

and private sector. The movement should be towards the bottom-up approach, which

combines the experience, knowledge and understanding of various local groups and people.

Given the right environment and orientation communities can participate effectively in

different aspects of water resources management including water quality monitoring and

management. The case study from the Philippines referred to in subsection 6.2., did not only

help communities to acquire new skills but empowered them to participate in the management

of the quality of their water resources. For example a community leader who wanted to tap

water from some mountain spring and convey the water to the community, requested the

services of the community water monitor to determine the bacterial level of the water prior to

making the final decision of installing pipes. The tests revealed that some of the springs had

unsafe levels of coliform bacteria and alternative sources were found, saving the government

funds and minimising the risk of waterborne diseases.

A study from northern Thailand with villages participating in the management of their

watersheds and other villages not participating showed that the villages participating had

higher income after completion of the project (Empandhu et. al., 1996). Although they had

higher opportunity costs due to the time spent for meetings and seminars, the villagers taking

care of their own resources benefited from improved water, soil and forest quality and could

make use of it through higher yields and thus through a higher income (Empandhu et al.,

1996).

While the above cases show that communities can participate effectively in development,

the Food and Agricultural Organization (FAO) has carried out controlled experiments to

prove that output from farming activities can be improved through community involvement

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 197

and participation in the planning of the project. The case studies outlined as case studies 1 and

2 below illustrate the potential for community participation and farmer training in improving

harvests.

Case study 1: Start – up of participatory community planning in Mexico

From: FAO. 1997. Communication for Rural Development in Mexico: In Good Times

and Bad. By Fraser, C. and Restrepo-Estrada Rome

What might be coined FAO’s first concerted venture into participatory planning by

intended beneficiaries of a project, occurred in Mexico under the PRODERITH (Programme

of Integrated Rural Development in the Tropical Wetlands Project), funded by the World

Bank and technically backstopped by the Development Support Communication Branch. The

first phase ran from 1978 to 1984 and was concerned with improving agricultural

development in Mexico’s wetlands that make up 23 percent of the country’s total land area.

Prior to PRODERITH, a large-scale integrated rural development project had been launched

in the wetlands which drained 83 000 hectares, built roads, new villages, schools and medical

centres, yet was never successful.

The peasants never identified with it nor did they use or maintain the infrastructure

properly. This was attributed to “a lack of effective mechanisms for the participation of the

beneficiaries”. The objectives of the new project, budgeted at US$149 million, were to

increase agricultural productivity in the tropical wetlands, improve the living standards of

peasant families, and conserve natural resources. People in the targeted communities were

involved in the planning process from the start. The mechanism to do this was imbedded in a

Central Rural Communication Unit created for the project. It worked principally with video

and support print materials to cover three types of communication needs: a) situation analysis

and participatory planning with peasants, b) education and training for peasants, and c)

information for project coordination and management.

Outreach field units were set up to work with communities. Video was used to record

local people’s attitudes and perceived needs and then played back to individual communities

as a basis for promoting internal dialogue about its past, present and future, and options for

improvement.

People began to articulate the realities of their situation, their priorities, and what they felt

capable of doing. This was followed up with a synthesis of collective perceptions and

elaboration of a “local development plan” for project concentration. During its

implementation, video was also extensively used for orientation and farmer training in a wide

range of agricultural and rural development topics.

At the end of its first phase in 1984 incomes of some 3 500 families in a 500 000 hectare

zone had increased by 50 percent over 1977 levels. And perhaps most significantly, it had put

in place a methodology for replication in a second phase involving 650 000 people in an area

covering 1.2 million hectares.

The World Bank considered PRODERITH to be among the most successful projects they

had supported up to that time, attributing much of its success to the participatory approach

adopted by the communication units in synthesizing community development priorities, with

follow-up skills training for farmers in its implementation. As for the bottom line, the

communication component for the first phase absorbed only 1.2 percent of the total costs,

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

198

while the internal rate of return, a measure of the economic success of the project, was 7.2

percent higher than originally foreseen.

Case study 2: Comparison of Inputs and Outputs of ten IPM versus ten Non-IPM Rice

Farmers in West Sumatra, Indonesia

From: FAO .1993. IPM Farmer Training: The Indonesian Case, Jogyakarta: FAO –

IPM Secretariat

A controlled study was conducted in West Sumatra, Indonesia, during the wet season of

1992-1993 (December to May). The study compared costs of rice farming inputs and outputs

among ten farmers who had participated in IPM farmer field schools during the previous wet

season with practices and outputs of l0 farmers who had never participated in FFSs. The two

groups of farmers were matched by location, farm size and land tenure. The only treatment

variable was the IPM-FFS training.

Observations and discussions with both sets of farmers were held on a weekly basis. IPM

training had stressed “Growing a Healthy Crop” (improved seed varieties, balanced

fertilization, proper plant spacing in straight rows), Conservation of Beneficial Insects” (low

pesticide use), and Weekly Field Observations to Determine Management Actions. The

foregoing training focus was determined to be the major difference between IPM and non-

IPM farmers. The comparative results on a number of key variables based on actual harvests

are tabulated below.

Overall, the IPM farmers achieved 21 percent more rice harvest yield on a per hectare

basis (6.9 tons versus 5.7 tons), for 97 percent of production costs, when compared to their

non-IPM farmer counterparts. The significantly lower “input” costs for IPM farmers were

largely attributed to minimal usage of commercial pesticides. Labour costs were also slightly

lower for IPM farmers, possibly because of better management actions.

The above two case studies prove that community involvement and participation can

improve results in a project and should be adopted in water quality monitoring and

management in Luvuvhu Catchment. The first step will be to make adjustments to the

national water quality monitoring information system to allow it to incorporate data generated

by the community based water quality monitoring system. This should not pose serious

challenges since the community based system will be using scientifically sound and accepted

field kits.

For example there will be no variation in the way experts measure turbidity and the way

the communities will do it. The only difference is that the new model will not only allow for

more water quality monitoring points to be included in the national data base but also for

“real – time” monitoring to take place at community level. Allowing the data from

community level to be incorporated into the national system, amounts to transmitting the

community’s voices throughout the system to the top. The authorities at different levels will

make decisions based on data that reflects the real conditions at community level.

The second action to encourage true or effective community participation will be for

authorities to finalise the establishment of the catchment management system in

Luvuvhu/Letaba WMA since the proposed structures are likely to enhance community

participation in water quality monitoring and management as shown in Figure 3. The structure

would need to be linked to local government structures at community level so that it can

articulate the opinions of the majority of the people. Links have been created between Figure

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 199

3 and Figure 4, resulting in Figure 19 so that information from the grass root level can flow

through to the national level.

Variable Average Budget for 10 IPM

Non-IPM Farmers (In Rupiahs)

Average Budget for 10 IPM

Trained Farmers

(In Rupiahs)

Pre-Harvest Labour/Ha 414 660 384 656

Harvest Labour/Ha 657 730 659 851

Total Inputs/Ha 163 268 139 819

Total Production Costs/Ha 820 998 799 670

Total Output in Kg/Ha 5 741 6 953

Figure 19. Proposed links between catchment management and local government structures.

Each sub village will operate as the smallest unit of water quality monitoring. A set of

field kits for monitoring, for example pH, turbidity, residual chlorine and microbiological

quality of water can be issued to a chief and all the sub villages under that chief can share the

same set. Each village will select at least two volunteers who will be trained by experts from

DWA/Vhembe District Municipality in the use of the kits. The volunteers will submit the

monitoring results to the staff at the nearest treatment plant for onward transmission to the

district level and incorporation into the national water quality information system. Rapport

between the volunteers and the staff in - charge of treatment plants should be encouraged so

that if the results of monitoring at community level are not satisfactory, the volunteers can

easily bring this to the attention of the staff at the treatments plants. The staff will either sort

out the problem if it is within their capacity or report to Vhembe District Municipality as the

WSA responsible for treating domestic water. This will be done as part of Alert Level

framework as discussed above.

The results from each of the sub villages should be discussed at the weekly meetings held

at the Tribal Office and chaired by the chief to allow the rest of the community to be aware of

the state of their water supply. The elected Ward Committee members from each village will

then take the results to the monthly Ward Committee meetings chaired by the Councillor.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

200

Decisions are then made in this committee. From this point two routes should be taken. First,

the Councillor as a member of the local municipality should take the results to municipal

meetings where the results and any suggestions from each ward should be discussed and

actions recommended. The results will then follow the normal local government reporting

structures up to national level. The second route is that, the councillor who should be a

member of the sub Catchment Forum should take the results for discussion at the sub

Catchment Forum. The chiefs should be part of the sub CF so that they can represent their

villages. The members from the sub CFs who attend the Catchment Forum present the results

from each sub catchment for discussion. The Catchment Forum is made of all stakeholders in

the water sector at catchment level. The results will then be taken up and discussed at WMA

level by the CMA and the Catchment Steering Committee before being passed onto the

national level. This should empower the communities to have “a voice” in how their supplies

are managed and deepen the democratic agenda within the catchment.

The third action will require investing time and resources to bring about a mindset change

at all levels. Currently the accepted norm of service delivery is that the government is

responsible for everything. The communities are passive recipients of services. The

approaches to development are top – down oriented and all the decisions are made from

above and transmitted down to the communities. Communities on the other hand believe that

their duty is to demand for services from above and wait for them to be provided. Therefore

for the proposed model to work there will be a need for change of mindset from all levels

starting from policy and decision makers to implementers and communities. Everyone has to

change and accept that communities are capable partners who can take responsibility for their

actions. Everyone involved has to be taken through the five phases of adapting and

overcoming resistance to change. All the people involved in the community based water

quality monitoring and management programme need to accept and believe in the SARAR

philosophy. Therefore, there will be need to ensure that mindset activities are carefully

planned and implemented at all levels.

Advocacy meetings and workshops should be conducted for policy makers (politicians,

directors and community leadership, etc.) while intensive training on development and

leadership theories should be carried out among implementers (managers and extension

workers). This process should ensure that, it assures all involved of the community’s capacity

and capability to implement and manage the proposed community based water quality

monitoring and management programme. All role players, especially extension workers need

to be assured that the proposed programme will not lead to any loses (jobs, allowances, status,

etc.) on their part to avoid generating resistance among them.

6.4. Communication Framework

For effective community participation in water quality monitoring and management to

take place, an appropriate communication framework has to be developed. The

communication framework should place the targeted communities at the centre of its three

key components as shown in Figure 20. The three fundamental components of the

communication framework are education and communication, research and extension.

The Health Belief Model dictates that before one attempts to change the behaviour of

individuals, one needs to understand how they perceive the situation under discussion. Hens

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 201

(2006) states that indigenous knowledge is context specific and what works successfully in

one location or for one community may not work for another. Koocheki, (2007) further states

that indigenous knowledge is unique to a particular culture and acts as the basis for local

decision making in agriculture, health, natural resource management and other activities and

it is embedded in community practices, institutions, relationships and rituals.

Figure 20. The Knowledge Triangle: Adopted from FAO (2003).

As already discussed under the empowerment framework, individual perceptions

influence the chances that an individual will take action against a perceived risk and that

perceptions can improve with experience. Therefore research will play a critical role in

shaping the whole water quality monitoring and management agenda at community level.

Besides the government agencies that are involved in water quality monitoring and

management, Luvuvhu Catchment has academic institutions such as the University of Venda,

Madzivhandila Agricultural College, Techniven etc. These can play an important role in

carrying out research which will focus on the education and communication component of the

framework. The trademark of this framework will be the dialogue between the communities

and the extension staff. Before the extension staff go out to give health and hygiene

education, the researchers should collect relevant data and analyse it and use it to come up

with the content of the health and hygiene education topic. In other words any education

programme will act as a “reply” to a community request for information.

While the academic institutions can play an important role of research, it remains critical

to realise that the extension workers themselves through their engagement with communities

collect a lot of data that can guide their education and mode of communication. The extension

workers should always use participatory tools to engage communities as these allow the

communities to discuss the issues amongst themselves and reach consensus before giving

feed back to the extension workers. People should always be divided into groups and given a

theme to work on and give feedback to the meeting.

The education and communication component is responsible for supplying information to

the communities. This involves simplifying all the data accumulated by researchers or

extension staff to a level that the communities where the data was collected can understand it

and use it. For example, the water quality monitoring results collected by DWA staff and the

academic institutions in the catchments can be classified and translated into colour codes as

shown in Table 9. These will allow even those who are not educated to understand the results

and the implication to public health.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

202

There are many ways of giving out information as outlined in section. Most of the

villages in Luvuvhu Catchment have electricity and it is possible to use electronic based

methods of communication such as videos which show villagers how other villagers

elsewhere have performed in similar projects or dissemination of information through the

internet. The researchers and extension workers can simplify water quality monitoring results

from each village and upload them onto internet and villagers through their representatives

can download the results from internet cafes and discuss them at their meetings. India has

used internet successfully in community education programmes (Bailur, 2007). Simplified

journalistic approaches can be utilized to disseminate information. For example weekly

values of turbidity, pH, and residual chlorine can be published on the wall at the tribal office.

Interpersonal communication methods such as extension workers explaining water quality

monitoring results at community meetings or the community members explaining results to

each other remain viable options.

Community radios have played a significant role in improving community perception of

different development scenarios. University of Venda has a radio station which is not a public

broadcaster and can play the role of a community radio in Luvuvhu Catchment. For example

researchers at the University can simplify results of water quality monitoring done in the

catchment over the years and disseminate these through Univen FM as a way of provoking

debate around water resources management in the catchment. Univen FM can also broadcast

regular updates on the quality of water in different villages.

Efficient management of existing limited water resources in terms of quality and quantity

requires real time flow of information both from managers to users and users to managers.

These processes can be made effective and efficient by using latest available open source

technology such as cell phones (Khan, 2010). Mobile phone networks extend over most of

South Africa and have an untapped potential for data capture and water quality management.

In 2008, South Africa had 45.68 million subscribers which came only second in Africa to that

of Nigeria with 45.89 million subscribers (Moorgas et al., 2010). Luvuvhu Catchment is

therefore in a position to speed up the dissemination of water quality monitoring by

employing simple technologies to monitor water quality at community level and then

uploading the results onto a cell phone for dissemination. One such technology is the

Electronic Mobile Water Application (eMWAP) an innovative concept harnessing the use of

appropriate field kits and mobile phones to support monitoring and management of water

quality in remote areas.

The eMWAP application works via the process of using simple field test kits, such as the

H2S strip test or World Water Monitoring Day Test Kits. The results of the test are then

immediately loaded onto a mobile phone which gives instant feedback to the user and to the

manager/authority, who can then ensure that problems are identified and resolved without

delay. eMWAP currently captures data for pH, turbidity, electrical conductivity, H2S strip test

and free chlorine residual (any determinants can be added on request).

eMWAP helps to:

• Capture water quality analysis related information in remote places by means of a

mobile phone

• Communities receive instant feedback on their water quality even though would not

have immediate access to a laboratory

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 203

• Understand what issues exist by receiving direct results

• Capture water quality results on a database without the need for internet access

• View data on a dedicated website that gives you the number of samples collected,

compliance and failures

• Inform your sampler, manager, authorities of any immediate water quality issues

• Know and understand the health status of rivers in your area

• Inform municipalities of water services related issues, e.g. burst water pipes

The extension component of the communication framework deals with issues of training,

imparting and transfer of skills. The communities will need to be trained and given skills to

monitor quality at their level. They need to be trained on how to use their natural senses to

monitor water quality and interpretation of these sensory judgements in a scientifically

acceptable manner. They need to be trained to use simple water quality monitoring field kits

and how to interpret the results. The ratio of EHOs to the people to be served is very low and

therefore the available manpower cannot offer quality extension service. Vhembe District

Municipality has also not appointed staff for water quality monitoring purposes.

To overcome manpower shortage, there is a need to select and train community

volunteers to give health and hygiene education and to monitor water quality at their villages.

In the Nzhelele area under Makhado Local Municipality, the Environmental Health staff has

created waste management fora at community level to mobilise communities to clean up the

area. These fora report any waste management problems to the staff based at Siloam Hospital.

The same concept can be employed in water quality monitoring activities in Luvuvhu

Catchment.

REFERENCES

Adams, W.M., Potkanski, T.P. and Sutton, J.E.G. (1994) Indigenous farmer managed

irrigation in Sonjo, Tanzania. GeogrlJ. 160: 17-32.

Agenbag, M., Gouws, M. (2004) Redirecting the role of environmental health in South

Africa. In: Paper presented at 8th World Congress on Environmental Health, Durban,

2004 Feb 23-27.

Agenbag, M. (2008) The Management and Control of Milk Hygiene in the Informal Sector by

Environmental Health Services in South Africa. Masters Thesis in Environmental Health

at the Central University of Technology, Free State, Bloemfontein, South Africa, 2008.

Unpublished.

Ahearn, D.S., Sheibley, R.S., Dahlgren, R.A., Anderson, M., Jonson, J. and Tate, K.W.,

(2005) Land use and land cover influence on water quality in the last free-flowing river

draining the western Sierra Nevada, California, Journal of Hydrology 313 (3): 10 – 13.

Anderson, B. A., Romani, J.H., Phillips, H.E., Wentzel, M., Tlabela, K., 2008. Exploring

Perceptions, Behaviors and Awareness: Water and Water Pollution in South Africa

www.hsrc.ac.za/Publication-Keyword-924. phtml.

Bailur S., (2007) The complexities of community participation in ICT for development

projects: the case of “our voices”. Proceedings of the 9th International Conference on

Social Implications of Computers in Developing Countries, São Paulo, Brazil, May 2007.

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

204

Barwell, L., (2006) Global Change and Ecosystems Research in South Africa: Opportunities

for Partnership with Europe2006 www.csir.co.za/nre/docs/ NREARR2009_FINAL.

Deutsch, M. W. (1997). Antibiotic use in necrotizing pancreatitis; 122:356-61. (AN :

97259146 ).

DWAF, (2002) Water Supply and Sanitation Policy: White Paper, Cape Town www.dwaf.

gov.za/documents.

Department of Health, (2005a) Diarrhoea and Typhoid outbreak in Delmas under control,

Sept 2005. www.hsr.iop.kcl.ac.uk/prism/tag/download/ DHFast.

Department of Health, (2005b) Protecting what’s underneath the tap www.

hsr.iop.kcl.ac.uk/prism/tag/download/DHFast-forwardingPCMH.pdf.

Dogaru, D., Zobrist J., Balteanu D., Popescu C., Sima M., Amini M. and Yang, H., (2009)

Community Perception of Water Quality in a Mining-Affected Area: A Case Study for

the Certej Catchment in the Apuseni Mountains in Romania. Environmental Management

43:1131–1145.

Dungumaro, W.E, and Madulu, N.F., (2002) Public Participation in Integrated Water

Resources Management: the Case of Tanzania: 3rd WaterNet/Warfsa Symposium 'Water

Demand Management for Sustainable Development', Dar es Salaam, 30-31 October 2002.

Dungumaro, W. E., 2007. Socioeconomic differentials and availability of domestic water in

South Africa: Physics and Chemistry of the Earth, Parts A/B/C, Issues 15 -18; 1141 –

1147.

DWAF, (2006). National Water Resource Policy (NWRP) Department of Water Affairs

Forestry: Pretoria www.dwaf.gov.za/documents.

DWAF, 2005. Luvuvhu/Letaba WMA: Internal Strategic Perspective: APPENDICES. Report

No. P WMA 02/000/00/0304. Department of Water Affairs and Forestry Directorate:

National Water Resource Planning (North).

Empandhu, D, Lakhaviwattanakul, T. and Kalyawongsa, 1996. A Case of Successful

Participatory Watershed Management in Protected Areas of Northern Thailand. In case

studies of people’s participation in Watershed Management in Asia. Part II : Sri Lanka,

Thailand Vietnam and Philipines, ed. Sharma, P.N Field Documents No. 5 Kathamandu,

Nepal.

FAO, (1993). IPM Farmer Training: The Indonesian Case, Jogyakarta: FAO – IPM

Secretariat. Published by Republic of Indonesia, IPM Secretariat, Jakarta; pp 512.

FAO, (1997). Communication for Rural Development in Mexico: In Good Times and Bad.

By Fraser, C. and Restrepo-Estrada Rome www.fao.org/ DOCREP/003/X8002E/

x8002e00.htm - 9k.

FAO, (2003) The state of world fisheries and aquaculture: Highlights of special FAO Studies

Editorial Group FAO Information Division www.fao.org/DOCREP/003/X8002E/

x8002e00.htm - 9k.

Ferreyra, C., de Loë, R. C. and Kreutzwise, R. D., (2007) Imagined communities, contested

watersheds: Challenges to integrated water resources management in agricultural areas:

Journal of Rural Studies; 304(24) 423 – 478.

Fox, C. R. and Tversky, A., (1995). Ambiguity Aversion and Comparative Ignorance. The

Quarterly Journal of Economics, Vol. 110, No.3. 585–603.

Friis-Hansen, R., (1999) The Socio-economic dynamics of farmers` management of local

plant genetic resources – A framework for analysis with examples from Tanzanian case

study. CRD Working Paper 99.3.

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 205

Genthe, B. and Steyn, M., (2006) Good Intersectoral Water Governance – A Southern African

Decision-Makers Guide: Chapter on Health and Water www.researchspace.csir.co.za/

dspace/bitstream/.../1/Genthe_2006_d.pdf.

Gingras AC., Gygi S. P., Raught B., Polakiewicz R. D., Abraham RT., Hoekstra MF.,

Aebersold R., Sonenberg N, (1999). Regulation of 4E-BP1 phosphorylation: a novel two-

step mechanism. Genes Dev. Jun 1;13(11): 1422-37.

GWP, (2000). Towards water security: A framework for action www.gwptoolbox.org/

index2.php?option=com...id...

Haddad, B ., (2007) The professional and intellectual challenges of sustainable water

management, in proceedings of the 3rd Dubrovnik conference on sustainable development

of energy. Water and Environment system. eds. Singapore: World scientific publication.

Hemson, D., and O'Donovan, M. (2005). Putting numbers to the scorecard: Presidential

targets and the state of delivery. In S. Buhlungu, J. Daniel, R. Southall and J. Lutchman

(Eds.), State of the nation: South Africa, 2005-2006 (pp. 11-45). Cape Town: HSRC

Press.

Hens, L., (2006) Indigenous Knowledge and Biodiversity Conservation and Management in

Ghana: Human Ecology Department, Vrije Universiteit Brussel, Laarbeeklaan 103B-

1090 Jette, Belgium: www.vub.ac.be/MEKO.

Hirji, R., Mackay, H., Maro, P., (2002) Defining and Mainstreaming Environmental

Sustainability in Water Resources Management in Southern Africa. SADC, IUCN,

SARDC, World Bank, Maseru, Harare, Washington, DC.

Hodgson, K. and Manus, L., (2006) A drinking water quality framework for South Africa)

ISSN 0378-4738 = Water SA Vol. 32 No. 5 (Special edn. WISA 2006) ISSN 1816-7950).

Kamara, J., (2004) Indigenous knowledge in natural disaster reduction in Africa, periodic

publication by UNEP/GRID-Arendal.

Kauzeni A. S. and Madulu N. S., (2000) Review of development programmes, Projects in

Serengeti, Consultant report to SIDA, Daressalam.

Koocheki, A.A., (2007) Indigenous knowledge in Agriculture with Particular Reference to

Saffron Production in Iran: I International Symposium on Saffron Biology and

Biotechnology May 20-26, 2005, Tehran (Iran).

LeChevallier, M.W., Norton, W.D. and Lee, R.G. (1991). “Giardia and Cryptosporidium in

Filtered Drinking Water Supplies.” Applied and Environmental Microbiology.

2617-2621.

McGhee, T.J. (1991). Water Supply and Sewerage, 6th Ed., McGraw-Hill, New York, 602.

Mackintosh, G. and Colvin C., (2003) Failure of rural schemes in South Africa to provide

potable water. Environmental Geology, CSIR Research Space > General science,

engineering and technology.

Malume, F. (2010) Evaluation of Water Demand in Makonde Village in Thulamela Local

Municipality of South Africa and the Implications on Water Resources Management,

Unpublished honours dissertation. School of Environmental Sciences University of

Venda.

Marshall, M. M., 1976. Waterborne protozoan pathogens. Clinical Microbiology Reviews Vol

10, No. 1, 67-85.

Masidiri, F., (2008) Evaluation of the Quality of Water Supplies and the Implications on

Domestic Uses in Malamulele area of Thulamela Local Municipality in South Africa,

Nova Science Publishers, Inc.

L. Nare and J. O. Odiyo

206

Unpublished honours dissertation, School of Environmental Sciences, University of

Venda.

Moorgas, S., Naidoo,V., Wensley, A., Mackintosh, G. and Charles, K. (2010) The use of

mobile based DWQ monitoring and management technology for remote areas. Feedback

on Kwa Zulu–Natal pilot initiative, 2nd DWQ Conference 7a4Paper www.ewisa.co.za/.../

DWQUALITY CONFERENCES.

Mvundlela, M.S., (2010) Evaluation of Actual Domestic Water Consumption in Tshidembe

Village under Thulamela Local Municipality in South Africa (Unpublished in honours

dissertation, School of Environmental Sciences, University of Venda.

Oberholster, P.J., Botha, A.M. and Cloete, E., (2007) Using a battery of bioassays, benthic

phytoplankton and the AUSRIVAS method to monitor long-term coal tar contaminated

sediment in the Cache la Poudre River. Colorado. Water Res. 39, 4913 -4924.

Nemadodzi, N., (2008) Evalution of Water Poverty in Mhinga Village in Thulamela Local

Municipality of South Africa, Unpublished honours dissertation, School of

Environmental Sciences, University of Venda.

Odiyo, J. O., Makungo R. and Muhlarhi, T. G. (2009) Investigating the impacts of

geochemistry and agricultural activities on groundwater quality in the Soutpansberg

fractured rock aquifers, Paper presented at the groundwater conference on pushing the

limits held at Somerset West, Cape Town, South Africa, 16 – 18 November 2009.

O’Donoghue., Habel., Normal., Michael., Maddox., Marion., (1993) Myth, ritual and the

sacred. Introducing the phenomena of religion, (Underdale: University of South

Australia, 1993).

Ong’or, D.O., (2005) Community Participation in Integrated Water Resource Management:

The Case of the Lake Victoria Basin: Department of Agriculture, Moi Institute of

Technology.

Pope J., Singh B. and Thomas D., (2006b) Mining related environmental database for West

Coast and Southland: Data structure and preliminary geochemical results, AUSIMM

2005 Annual Conference: Auckland, 7p.

Redding, C.A., Rossi, J. S., Rossi, S. R., Velicer, W. F. and Prochaska, J.O. (2000) Health

Behavior Models. The International Electronic Journal of Health Education, 3 (Special

Issue):180-193.

Tsiho S., (2007) Water Pollution in Southern Africa www.whois.ws/whois_ index/g/domain.

Moyo, N. and Mtetwa, S., (2002) Water Quality Management and Pollution Control. In Hirji,

J., Johnson, P., Maro, P. and Matiza Chiuta, T. (eds) 2002. Defining and Mainstreaming

Environmental Sustainability in Southern Africa. SADC, IUCN, SARDC, World Bank:

Maseru/ Harare/ Washington DC.

Silberbauer, M. J., Rossouw, J. N., Kamish, W., Chetty, K., Hadebe S. and Wildemans D.,

(2001) Building Capacity in Water Quality Modelling in South Africa: Report on a U. S.

- South Africa Binational Commission training visit by staff of Ninham Shand,

Department of Water Affairs and Forestry and Umgeni Water to the United States of

America, September 2001 Report number: N/0000/00/USA/0601.

Van Zyl, (1999) Impact on Diffuse Pollution? Proceeding of the international conference on

diffuse pollution in Australia centres. exeter.ac.uk/cws/ publications.

Van Zyl, B., (2002) Challenges for industry in reaching mine closure. Paper presented at the

WISA Mine Water Division, Mine Closure Conference, Randfontein, 23–24 October.

Nova Science Publishers, Inc.

Evaluation of Community Water Quality Monitoring ... 207

West, D., (2001) Nitrates in Ground Water: A Continuing Issue for Idaho Citizens: Ground

Water Quality Information Series No. 1: Idaho Department of Environmental Quality:

State Ground Water Program 1410 N: www.deq.idaho.gov/media/471641-nitrates_issue_

citizens.pdf.

Wigrup, I., (2005) The Role of Indigenous Knowledge in Forest Management : A Case Study

from Masol and Sook Division, West Pokot, Kenya: University essay from SLU/Dept. of

Silviculture ex-epsilon.slu.se:8080/.

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 8

THE FATE AND PERSISTENCE OF THE

ANTIMICROBIAL COMPOUND TRICLOSAN

AND ITS INFLUENCE ON WATE R QUALITY

Teresa Qiu1,3, Christopher P. Saint2,

∗

and Mary D. Barton1

1School of Pharmaceutical and Medical Sciences, University

of South Australia, Australia

2SA Water Centre for Water Management and Reuse, University

of South Australia, Australia

3Water Quality and Environment, South Australian Water Corporation, Australia

ABSTRACT

The Earth’s water resources are coming under increased stress from the combined

pressures of quality and quantity and wastewater has an increasingly important impact on

these issues. The quality of treated wastewater entering both the marine and freshwater

environments can greatly influence the health of natural ecosystems. In addition, a

shortage of freshwater supply in drought affected countries has led to an escalation in the

use of treated wastewater (“reclaimed water”) for the purpose of water supply

augmentation for agricultural, industrial and even potable uses. The fate and persistence

of chemical contaminants in wastewater is of considerable concern when considering the

potential environmental and health impacts of discharges, be they to the open

environment or captured for reuse. Of the many types of anthropogenic compounds found

in wastewater, pharmaceutical and personal care products (PPCPs) make up a significant

portion of encountered pollutants. Of these, triclosan is one of the prevalent compounds

known to be recalcitrant to removal or inactivation in wastewater treatment processes.

This chapter deals with the significance of triclosan in the wastewater cycle, its

persistence and fate, and examines the microbiological processes involved in the removal

of this compound. The significance of microbial resistance to triclosan is also considered

and how this relates to other types of microbial resistance to antibiotics, and we discuss

the potential role of wastewater as a source for the creation of new antibiotic resistant

strains of bacteria.

∗ Corresponding author.

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

210

1. INTRODUCTION

A recent study estimated that 80% of the world’s population is exposed to high levels of

threat to water security (Vörösmarty et al., 2010). This is compounded by the fact that many

of the world’s most densely populated areas are a significant distance from readily available

sources of freshwater.

Wastewater is not only a significant source of pollution in the environment but is now

increasingly being viewed as a resource with regard to providing an alternative water source

or “reclaimed water”. The greatest potential sources of industrial and domestic wastewater lie

in our cities, so in communities with well developed capture and treatment systems it is a

conveniently located source of water, primarily for irrigation and industrial purposes. Most

Western countries have well established sewer networks and advanced treatment plants that

can produce water of a good quality, however it is known that inorganic and some organic

contaminants can enter the environment post-treatment. This is an issue in terms of potential

environmental degradation but also increasingly of concern when one considers re-use of

treated wastewater with regard to its intended use, something often referred to as “fit for

purpose”. For example, reducing the amount of treated effluent released to sensitive marine

environments is of course a good thing but we also need to be mindful that we are not merely

reducing an environmental effect in one location and creating a new issue elsewhere through

inappropriate re-use.

Pharmaceutical and Personal Care Products (PPCPs) refers, in general, to any product

used by individuals for personal health or cosmetic reasons or used by agribusiness to

enhance growth or health of livestock. PPCPs comprise a diverse collection of thousands of

chemical substances, including prescription and over-the-counter therapeutic drugs,

veterinary drugs, fragrances, and cosmetics. The use of PPCPs is on the rise, for instance in

the USA between 1999 and 2009 the annual prescription rate for these compounds virtually

doubled from 2 billion to 3.9 billion (Tong et al., 2011). PPCPs largely enter the environment

through wastewater disposal whose original source is human or animal excretion or trade

waste from industry deposited directly into the environment or via a sewerage and treatment

network. PPCPs include the types of compounds referred to as endocrine disrupting

compounds (EDCs) and although effects on a variety of vertebrates have been demonstrated a

link is still to be made regarding the adverse human health effects of these compounds.

There is also concern about the fate and persistence of medical and veterinary antibiotics

in the environment. It is well documented that low level consistent exposure of

microorganisms to antibiotics can bring about increased levels of resistance. While the causes

of this effect are well documented in medical and veterinary applications, environmental

exposure (for example in the wastewater treatment process) and its significance with this

regard are largely unknown. Likewise the fate and persistence of antibiotics in the

environment and their influence on the overall burden of resistance traits and genes and their

dissemination is poorly understood.

A crucial factor in implementing the widespread use of reclaimed water is the economic

considerations when designing and operating treatment facilities to provide such water.

Whereas, in years to come, the value of any source of water may become so great as to

essentially neutralise the economic considerations; at present water reuse schemes are

generally expensive to construct, operate and maintain. The water industry and related health

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 211

authorities are conservative in their approach towards the level of treatment required prior to

use and rightly so – as the onus has to be on providing the science to demonstrate that reduced

levels of treatment do not result in increased levels of risk. Therefore, there is much interest in

researching the various treatment options that will provide a satisfactory level of removal of

contaminants such as PPCPs. There is interest in monitoring removal by existing wastewater

treatment processes and by alternative, and possibly more cost effective processes, such as

biofiltration.

In this chapter we consider the fate and persistence of a major PPCP, namely triclosan.

Triclosan has become widely used as an antiseptic and is of significant environmental

concern. We consider the history of its use and its mode of action and why there are concerns

regarding its continued use. For the reasons mentioned in this introduction we are also

interested in the presence of triclosan in wastewater, its removal by treatment processes and

persistence in the environment when these processes fail to remove it completely. With

regards to biodegradation we discuss what is known about the biochemical and genetic

processes that may be involved in triclosan dissimilation. Finally, we consider the

significance of triclosan resistance and biodegradation with regard to the role that wastewater

and other aquatic systems may play in the spread of antibiotic resistance.

2. WHAT IS TRICLOSAN AND WHY IS IT USED?

2.1. General Properties of Triclosan

Triclosan (TCS, C12H7Cl3O2, Irgasan) is a synthetic broad spectral antimicrobial

compound (General Chemistry Online, 2004) with a molecular weight of 289.5. Figure 1

below shows the molecular structure of this compound. It belongs to a group of 2-

hydroxydiphenyl ethers that is a class of compounds showing antimicrobial activity.

Triclosan was originally introduced to health care settings as a surgical scrub in 1972,

and was first included in toothpaste as an additive in Europe in 1985. Since then, triclosan has

been used as an active antimicrobial component in many PPCPs, which include detergents,

soaps, deodorants, cosmetics, lotions, toothpastes, mouthwashes and plastics for hospital and

household use. In recent years, there has been a remarkable increase in the popularity of the

addition of antimicrobial chemicals to consumer products, and triclosan is commonly used in

products ranging from chopping boards, pizza-cutters, mop handles and mattresses to bowling

ball finger inserts and baby toys.

Figure 1. Molecular structure of triclosan.

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

212

Triclosan exhibits a broad spectrum and clinically significant antimicrobial activity

(Regos et al., 1979, Vischer and Regos, 1974). It is acutely or chronically toxic to aquatic

organisms (Tatarazako et al., 2004), but has limited antiviral and antifungal properties

(Samsøe-Petersen et al., 2003).

The solubility of triclosan in water is poor under acidic and neutral conditions, and this

poor solubility significantly complicates the study of triclosan. However, triclosan is fat-

soluble, i.e. soluble in most organic solvents, and it is also water soluble in alkaline

conditions (pH>7). Triclosan is chemically stable, and has a decomposition temperature of

280 ˚C (Triclosan MSDS).

2.2. Antimicrobial Properties and Mechanisms of Action of Triclosan on

Bacteria

Triclosan is active against many, but not all, types of Gram-positive and Gram-negative

bacteria. It is bacteriostatic at low concentrations, but at higher concentration it is bactericidal.

Many studies have been done to evaluate the effectiveness of triclosan as a disinfectant for

dentifrices and hand-wash products as well as its use in clinical antibacterial settings (Barry et

al., 1984, Faoagali et al., 1999, Fine et al., 1998, Stephen et al., 1990, Walker et al., 1994,

Webster, 1992). Triclosan was found to be an efficient active component in toothpaste to

prevent the development of plaque on teeth and in hand washes and scrubs for general

hygiene (Aiello et al., 2004, McBain et al., 2003). One of the most significant findings about

the biocidal efficiency of triclosan has been its effectiveness in the control of outbreaks of

methicillin-resistant Staphylococcus aureus (MRSA) (Webster et al., 1994, Zafar et al.,

1995).

The efficiency of the antimicrobial activity of triclosan is dependent on several factors. A

bacterial population kinetic study (Gomez Escalada et al., 2005b) showed that generally

bacterial growth was delayed under triclosan exposure and the extent of delay corresponded

to the concentrations to which bacteria were exposed. Interestingly, instead of showing a

longer delay effect at higher concentrations, the study showed a longer delay at lower

concentrations. This is probably due to the low solubility of triclosan. It was observed in this

study that triclosan tends to crystallize from solution at high concentration (i.e. higher than

saturation concentration, triclosan solubility in water 0.01g/L at 20°C). Once the

crystallization process is initiated, the formed crystal will become a nucleus to promote

further crystallization even in an unsaturated solution. This will result in soluble triclosan in

the solution being present at a concentration lower than the saturated level. This explains the

reason for longer growth delay at low concentrations, where less crystal nuclei are formed

and more soluble triclosan is present in the solution. Only soluble triclosan is bio-available

and this accounts for the more efficient biocidal effect.

Bacterial growth phase is another key factor that influences the efficacy of triclosan

action. Bacteria in the log phase of growth are more susceptible to triclosan compared to

bacteria in stationary phase, and washed cells were found to be least affected (Tabak et al.,

2007).

Concentration (soluble component) is another very important factor in determining the

lethality of triclosan. At high concentrations, triclosan was found to be lethal regardless of the

growth phase of the bacterial population and this lethal activity could be effective within a

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 213

time period as short as 15 seconds (Gomez Escalada et al., 2005b). The effect of triclosan at

high concentration is positively related to contact time and triclosan concentration. However,

this effective triclosan concentration (represented by Minimal Inhibitory Concentrations)

varies among different isolates (Braoudaki and Hilton, 2004a, Gomez Escalada et al., 2005a,

McBain et al., 2004, Slayden et al., 2000, Suller and Russell, 2000).

These facts indicate that the mechanisms of triclosan bactericidal activity are complicated

and concentration related. Until a specific target was identified in 1998, triclosan had long

been thought to act as a non-specific biocide by attacking the bacterial cell envelope and

rendering it too porous to retain nutrients, which leads to the death of cells. The publication of

McMurry and co-worker’s study on the genetic target of triclosan (McMurry et al., 1998),

namely fabI, brought the study of mechanisms of triclosan resistance into a new era. This

study showed that triclosan acts on a specific target in the bacterial fatty acid biosynthetic

pathway, namely the NADH-dependent enoyl ACP (acyl carrier protein) reductase (FabI).

Subsequent studies investigated various areas of composition, kinetics and genetics of this

pathway (Heath et al., 2000, Heath et al., 1999, Kuo, 2003, Levy et al., 1999, McMurry et al.,

1999, Parikh et al., 2000, Slayden et al., 2000, Stewart et al., 1999).

Although the capacity of triclosan to target the fatty acid biosynthesis pathway has been

well established in recent years, triclosan biocidal action by targeting FabI in the bacterial

fatty acid synthesis pathway can only form part of the story of triclosan activity. This

mechanism applies only when the concentration of triclosan is not high enough to kill the

bacteria immediately, as the inhibition in bacterial fatty acid synthesis results in the inhibition

of bacterial proliferation. At this low triclosan concentration even when a complete inhibition

of fatty acid synthesis is achieved, triclosan will only inhibit the growth of organisms, but will

not kill them (Gomez Escalada et al., 2005a, Gomez Escalada et al., 2005b). Recent studies

have shown that the inhibitory factor is actually a ternary complex of triclosan, NAD+ and

InhA; therefore, triclosan can only act as a slow binding inhibitor of InhA due to the nature of

its requirement to form a complex with NAD+ (Kapoor et al., 2004, Kuo, 2003). However,

we can speculate that with the bacteriostatic effect, if lipid synthesis was prevented from

taking place over a long time, and no cell membranes could be renewed, the bacteria would

eventually die. Thus the bacteriostatic activity of triclosan at lower concentration can still be

used as a strategy in hygiene or bacterial growth control, yet it is not enough to achieve an

immediate lethal effect at this concentration range. Obviously there are some other more

efficient lethal actions of interest, which might be independent or related to fatty acid

biosynthesis, that remain to be elucidated.

3. WHY ARE THERE CONCERNS REGARDING

THE USE OF TRICLOSAN?

3.1. Use of Triclosan and Resulting Concerns

Triclosan is widely used as a constituent in PPCPs, and so is commonly discharged into

the water and wastewater environment via sewage and discharge of waste. Triclosan, at

various concentrations, has been detected worldwide in the water and wastewater

environment (Boyd et al., 2003, Jackson and Sutton, 2008, Kuster et al., 2008). As discussed

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

214

later in Section 4.1 triclosan concentrations vary from 0.02 µg/L to 0.21 µg/L in surface

waters, and much higher levels of 0.1-21.9 µg/L were detected in sewage plant influents

worldwide (Chau et al., 2008, Lishman et al., 2006, McAvoy et al., 2002, Miege et al., 2008,

Nagulapally et al., 2009, Okumura and Nishikawa, 1996). Although these concentrations may

not be significant enough to select triclosan resistance, the concentration of triclosan in the

digested sludge from wastewater treatment plants ranges from 0.5 to 15.6 μg/g (dry wt)

(McAvoy et al., 2002), which may be close to the MICs of some bacteria. In addition, a much

higher concentration of triclosan is expected to be present in the sewer immediately

downstream of households as well as major triclosan users, such as medical clinics. Concerns

have thus arisen in the water and wastewater management area that the biocidal property of

triclosan against many bacteria and the acute toxicity to algae and other aquatic organisms

will significantly influence the balance of the natural ecology in water environments. Another

key concern arising because of the presence of triclosan at sub-lethal levels is the potential for

selection of bacterial resistance to this bis-phenol (Thompson et al., 2005, Xia et al., 2005).

3.2. Bacterial Resistance to Triclosan

3.2.1. Triclosan Resistance and Adapted Resistance in Bacteria

Many bacteria, such as Neisseria subflava, Prevotellanigrescens, Porphyromonas

gingivalis and E. coli, Salmonella, Staphylococcus etc, are intrinsically susceptible to

triclosan at low concentrations (0.1–3.9 mg/L) (Copitch et al., 2010, Randall et al., 2004),

whereas Lactobacillus and Streptococcus mutans were less susceptible (MIC range 15.6–20.8

mg/L) (Randall et al., 2004). Most Pseudomonas strains are found to be resistant to triclosan

even at high concentrations (Chuanchuen et al., 2002, Ellison et al., 2007).

Acquired resistance to triclosan has also been widely observed across bacterial genera

(Ledder et al., 2006), and the trait varies in different species. A highly significant reduction

(400-fold) in triclosan susceptibility was achieved in a laboratory E. coli strain when exposed

to triclosan, whereas only minor (two-fold) decreases in triclosan susceptibility occurred for

Prevotella nigrescens and Streptococcus strains (Randall et al., 2004).

In a recent study, forty environmental and human isolates were tested for effects of

chronic triclosan exposure on antimicrobial susceptibility; however only 5 isolates gained

acquired resistance (Ledder et al., 2006).

The inconsistency seen in studies on acquired resistance in bacteria under prolonged

exposure of triclosan suggests that this adaption is not through a common pathway shared by

all bacteria, but it is confined to certain bacterial species that share similar specific pathways.

3.2.2. Mechanisms of Resistance to Triclosan

A recent study investigated the prevalence of decreased susceptibility to triclosan and a

panel of antibiotics in over 400 human and animal isolates of Salmonella enterica. Seventy

four percent of the isolates were shown to be susceptible to triclosan at a concentration of

0.12 mg/L. Of the 111 isolates showing growth on media containing 0.12 mg/L, only 16

strains showed consistent triclosan resistance, but mostly at a low level (MIC 0.25-0.5 mg/L).

Only 1 isolate had MIC of 4 mg/L. Of these strains showing reduced triclosan susceptibility,

56% were also multidrug-resistant, and expressed higher efflux activity compared to the

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 215

strains showing reduced susceptibility only to triclosan. No specific mutations within fabI

were detected for these multi-drug resistant (MDR) strains (Copitch et al., 2010). This

indicates that triclosan resistance maybe a result of generic resistance mechanisms that can be

promoted by a multidrug efflux system. However, this investigation was restricted to

Salmonella strains and mechanisms in other bacteria could be significantly different.

Pseudomonas is intrinsically resistant to triclosan, and the mechanisms involved in

triclosan resistance as well as cross resistance to other antibiotics have been studied

extensively. So far, the most commonly accepted mechanisms of reduced susceptibility to

triclosan and cross-resistance to antibiotics in Pseudomonas are mutation of fabI and over-

expression of a Resistance Nodulation Division (RND) efflux system with more emphasis on

the latter.

3.2.3. Resistance via Biodegradation?

Theoretically, if bacteria have the ability to decompose triclosan at a specific

concentration, then they will automatically be resistant to this compound at this concentration,

as for biodegradation to occur, bacteria have to physically interact with triclosan without

being killed. Resistance to antibiotics can occur in several ways: by mutations that render the

site of antibiotic action inaccessible, or alter the target site of action so the antimicrobial agent

cannot bind, or by the action of inactivation enzymes carried by the target bacterium.

Therefore, if one considers biodegradation a type of resistance this can only be proffered

via an enzymatic pathway. Depending on the pathways endowed the breakdown (inactivation)

may also provide a source of energy for the host microorganism thus imparting a dual benefit.

The advantage of such a mechanism over mere resistance is that it provides the

microorganism with a selective advantage allowing it to boost its population; this in turn

would actually result in greatly enhanced triclosan removal from the immediate environment.

However, it is worth noting that the intermediates formed during the biodegradation process

could also be toxic and so may demonstrate inhibition to the growth of the degraders

(Aranami and Readman, 2007, Canosa et al., 2005). In this situation, triclosan biodegradation

would actually function as an inhibitory mechanism for these bacteria.

4. TRICLOSAN IN THE ENVIRONMENT

4.1. Detection of Triclosan and its Effects on the Water and Wastewater

Environment

Presence of triclosan in water environments has been reported worldwide (Boyd et al.,

2003, Chau et al., 2008, Federle et al., 2002, Jackson and Sutton, 2008, Lishman et al., 2006,

McAvoy et al., 2002, Sabaliunas et al., 2003, Samsøe-Petersen et al., 2003, Singer et al.,

2002, Xia et al., 2005). Table 1 below summarizes the levels of triclosan reported recently in

water and wastewater environments.

It can be seen that triclosan is largely present in the wastewater environment and is

removed during wastewater treatment processes, which is indicated by the significant drop of

triclosan concentrations between the influent and effluent of sewage plants. Furthermore,

when discharged into a natural water body, the triclosan-containing effluent from sewage

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

216

plants is also diluted in the surface water by natural dilution and the self-purification capacity

of the natural water body. The resulting concentrations of triclosan in surface water after the

wastewater treatment process and the natural process thus fall far below inhibitory

concentration levels and thus will not have a lethal effect on environmental microorganisms

and other ecosystems.

However, the predicted no effect concentration (PNEC) of triclosan for algae is as low as

50 ng/L (Singer et al., 2002). Therefore, even after all the natural and artificial removal

processes, the triclosan levels listed in these studies are still significantly above this PNEC

level.

Therefore, the presence of triclosan in the natural water environment as well as in sewage

plants may potentially have an impact on the ecological environment (Tatarazako et al.,

2004). A recent study (Thompson et al., 2005) highlighted the necessity for further

investigations into the effect of residuals, at concentrations far below inhibitory

concentration, on bacterial populations and their role, if any, in the continued problem of

antibiotic resistance.

Concentrations of triclosan in local sewage collection systems leading to the sewage plant

could be much higher than those listed as influent concentrations of triclosan in Table 1 due

to the well documented adsorptive property of this substrate (Federle et al., 2002, Heidler and

Halden, 2007, Singer et al., 2002). The concentration of triclosan measured in sewage plant

influent may therefore be lower than that in the local sewer system, because of the adsorption

of triclosan onto the surface of organic material, such as polyvinyl chloride (PVC) pipes.

Therefore, the sub-lethal concentration of triclosan immediately or shortly after discharge

from homes could act as an inducer of the expression of bacterial multidrug resistance

determinants, such as multidrug efflux. The potential for triclosan to select for antibiotic

resistant bacteria has been addressed by scientists all over the world. Various studies have

reported an increase of triclosan resistance and a possible reduction in susceptibilities to

antibiotics after exposure to triclosan (Braoudaki and Hilton, 2004b, Chuanchuen et al., 2001,

Schweizer, 2001).

In the natural water environment, apart from the potential selection of antibiotic resistant

bacteria, the presence of triclosan has also raised various other concerns about the impact that

triclosan could have on the ecological environment. For example, triclosan is acutely or

chronically toxic to some aquatic organisms, as in the case of acute toxicity to algae at trace

levels, i.e. PNEC of 50 ng/L (Tatarazako et al., 2004).

Table 1. Occurrence of triclosan in the surface water and wastewater environment as

reported in the literature (Chau et al., 2008, Lishman et al., 2006, McAvoy et al., 2002,

Miege et al., 2008, Nagulapally et al., 2009, Okumura and Nishikawa, 1996)

Surface Water

(µg/L)

Sewage Plant

Influent (µg/L) Effluent (µg/L)

Europe 0.04-0.21 0.6-21.9 0.11-1.1

USA 0.021 3.8-16.6 0.2-2.7

Japan 0.05-0.15 NA NA

China 0.004-0.117 NA NA

Canada <4.01 <0.324

France 0.09-0.43

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 217

Triclosan is also found to have a high potential for bio-accumulation (Coogan et al.,

2007). In a Japanese study, a range of 0.89-2.5 µg/kg of triclosan was detected in fish tissues

(Okumura and Nishikawa, 1996), and a range of 0 to 2100 μg/kg in lipid was found in human

breast milk samples (Dayan, 2007). Despite the lack of knowledge on how this accumulation

of triclosan will affect human or environmental health, it is significant enough to cause an

alert on public health grounds.

Recently, the formation of dioxin has been reported in a natural water body as a result of

chlorination of triclosan in the presence of ultraviolet light (UV) (Aranami and Readman,

2007, Chung et al., 2007, Lores et al., 2005, Son et al., 2007, Yu et al., 2006). These much

more toxic triclosan derivatives will have greater impacts on the aquatic fauna.

The presence of triclosan in the wastewater environment may have similar impacts as that

in the natural water body as described above. In addition, it may also have significant impact

on the performance of the wastewater treatment plant.

Engineered wastewater treatment processes represent artificial enhancement of biological

processes that occur spontaneously in the natural water environment. It employs a biological

process as the principal treatment process, where the behaviour of the microbiological

community dictates the performance of the processes. In particular, for conventional

wastewater treatment processes, the nutrient removal efficiency relies largely on the activity

of microorganisms in the activated sludge bio-reactors. Therefore, a selective pressure from

the presence of the biocidal effect of triclosan will potentially provide an inhibitory impact on

the biological environment and affect the bacterial community profile as well as the viability

of the microorganisms. This will affect the efficiency of the treatment process, particularly

when triclosan enters the process as a matter of shock load, when a high level of triclosan

enters within a short period of time.

There are limited studies investigating the effect of triclosan on the performance of the

wastewater treatment process. A bench-top study showed that the load of triclosan at levels

expected in household and manufacturing wastewaters is unlikely to upset sewage treatment

processes (Federle et al., 2002).

However, it is noteworthy that in this study, the sewage treatment process tested was a

triclosan pre-acclimated system (dosed with additional triclosan), which was evaluated after

the adaptation to triclosan had already occurred. Therefore, the result does not reflect the real

situation.

A recent study (Heidler and Halden, 2007) revealed that a large proportion of particle-

active triclosan (composed of 80±22% on average of the total triclosan inflow) was bound to

particulates and eventually sequestered and accumulated in excess sludge in the wastewater

treatment process.

This study therefore raised concerns about the practice of applying large quantities of

triclosan-containing sludge to soils used for animal husbandry and crop production, and

consequent concerns over the ecological, environmental and health impacts. However, in this

study, no information on sludge digestion was given; therefore it is hard to determine if the

concern is valid. In another study (McAvoy et al., 2002) investigating triclosan in digested

sludge, it was revealed that triclosan was adequately removed from digested sludge prior to

disposal for land application.

To conclude, the presence of triclosan in the water and wastewater environment

represents a potential risk to public health and the natural environment. The efficient removal

through sewage treatment plants is thus critical to mitigate the risk.

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

218

4.2. Wastewater Treatment Process and Triclosan Removal Efficiency

A conventional wastewater treatment process train includes preliminary treatment,

primary treatment, secondary treatment and sludge treatment (Wiesmann et al., 2007). The

preliminary treatment normally includes screens and a silt settling basin, which removes large

objects and a proportion of the suspended solids. Most of the suspended solids are removed

through the primary sedimentation basin, where a proportion of organics that attached to the

suspended solids are also removed.

Secondary treatment is considered as the main process for removal of organics and other

nutrients. Conventional secondary treatment includes aeration and sedimentation. During the

start-up of a new system, the indigenous microorganisms in the source water are

concentrated; the microorganisms that are capable of utilizing organics and other nutrients in

the source water proliferate and outgrow other microorganisms. By working synergistically,

the dominant microorganisms gradually evolve the ability to utilize the organic compounds in

the source water as carbon sources and thus treat the water.

In the aeration basin, aerobic microorganisms utilize oxygen, organics and other nutrients

in the water to proliferate. As a result, biomass increases in amount and the water is treated.

Downstream to the aeration basin, biomass is settled in the secondary sedimentation basin. A

proportion of the settled biomass is removed from the water treatment system before further

treatment in the sludge treatment process prior to disposal, while another portion of the settled

sludge from the secondary sedimentation basin is recycled into the aeration basin as inoculum

to facilitate the biological process in the aeration basin. Excess sludge from the secondary

sedimentation tank is digested using aerobic and/or anaerobic processes to remove organics

and nutrients through nitrification.

Triclosan may be removed by wastewater treatment processes together with other organic

compounds in the water. Therefore its removal efficiency is related to the overall organics

removal efficiency of the treatment process. Table 2 lists a brief summary of reports on

triclosan removal from wastewater treatment plants in different geographic locations. It is

assumed that these wastewater treatment plants accepted influent streams composed of mostly

domestic wastewater and they all operated well. In all processes, Biological Oxygen Demand

(BOD) and Total Suspended Solids (TSS) removal were steady and efficient.

From the data listed it can be seen that a good removal rate (over 90%) of triclosan may

be achieved by most secondary biological treatment processes. Of the commonly used

wastewater treatment processes, conventional activated sludge treatment demonstrated

reliable performance for triclosan removal with an effluent concentration of around 1µg/L

even at a high triclosan load in the influent.

A high removal percentage of 95% can be achieved in most activated sludge processes.

This is probably due to the high Hydraulic Retention Time (HRT) and Sludge Retention Time

(SRT), commonly seen in most conventional activated sludge processes. In some activated

sludge processes designed for enhanced removal of phosphorus and nitrogen, the HRT can be

maintained up to 12-18 hours and the SRT up to 15-30 days. Trickling filters and biological

contactors normally have less HRT, and therefore are slightly less efficient in triclosan

removal. Other parameters of operational conditions like temperature at the Waste Water

Treatment Plant (WWTP) will also influence the efficiency of the treatment processes.

Nevertheless, efficient removal of triclosan from wastewater can be achieved in most

secondary treatment processes (as shown inTable 2).

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 219

Table 2. Triclosan Removal Efficiency from Sewage by Wastewater Treatment

Processes (Bester, 2003, McAvoy et al., 2002, Sabaliunas

et al., 2003)

Type of Works Percentage

removal

Influent conc.

(µg/L)

Effluent conc.

(µg/L) Source

Activated Sludge 95% 21.9 1.1 UK

Activated Sludge 83% 5.4 0.9 UK

Activated Sludge 98% 4.7 0.07 USA

Activated Sludge 96% 1.2 0.05 Germany

Activated Sludge 95% 5.2 0.2 USA

Activated Sludge 96% 10.7 0.41 USA

Trickling filter 58% 3.8 1.6 USA

Trickling filter 86% 16.6 2.1 USA

Trickling filter 83% 15.4 2.7 USA

Trickling filter 79% 2.8 0.6 UK

Trickling filter 95% 7.5 0.34 UK

Biological contactor 58-96% 0.15-1.1

However, we can also see from this table that downstream of a successful triclosan

removal treatment process, there are still hundreds or thousands of nano-grams of triclosan

per litre of effluent, and this concentration will still cast an influence on the natural water

environment as discussed previously. Therefore, investigations into triclosan removal

mechanisms in these wastewater treatment processes are necessary in order to develop

strategies for an enhanced triclosan removal during these processes or to identify some

advanced supplementary treatment process to ensure that the residual concentration of

triclosan in the wastewater environment has minimum adverse impact on the natural

environment.

4.3. Mechanisms Involved in Triclosan Removal from Activated Sludge

Wastewater Processes

4.3.1. Flow of Triclosan in Activated Sludge Wastewater Treatment Process

Triclosan enters wastewater systems in the form of soluble triclosan in the aquatic phase

and also particulate-active mode. A molecular level of adsorption and desorption process

occurs at all times, and the ratio of soluble and particulate triclosan remains kinetically

balanced according to the property of the carrying water and the performance of the treatment

during the whole process.

Triclosan in wastewater will undergo a biological treatment process as described

previously, where physical adsorption and biological degradation work synergistically for the

removal of triclosan. At the biological treatment process, primary wastewater is mixed with

activated sludge at the beginning of the process, where soluble triclosan is adsorbed on

activated sludge. This adsorption is expected to be effective due to the polarity of triclosan.

At the same time, the large percentage of triclosan in particle form will be mixed and trapped

into activated sludge. The active component of activated sludge is the consortium of

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

220

microorganisms that are able to utilize the contaminants in water as nutrients. Under optimal

conditions provided by the bioreactor, these microorganisms will utilize triclosan and other

organics to metabolize and proliferate. During this process, triclosan is biodegraded, and the

remaining triclosan will be separated and sediment to the bottom of the reactor and is

disposed as excessive sludge at the end of the water treatment process. The excessive sludge

will then undergo an aerobic and anaerobic digestion, where contaminants in the sludge are

further degraded prior to disposal.

4.3.2. Variable Contribution of Different Triclosan Removal Mechanisms

Despite the fact that the triclosan removal rate in most activated sludge treatment

processes remains consistently high (95-98%) (Bester, 2003, McAvoy et al., 2002, Sabaliunas

et al., 2003), the mechanisms of this removal have been subject to debate. In a recent mass

balance study on triclosan removal from conventional wastewater treatment processes

(Heidler and Halden, 2007), an overall of 50±19% of the triclosan entering the plant was

found to remain in the excessive sludge after the whole treatment process, and only 48±16%

of the triclosan intake was biodegraded or transformed through other mechanisms. However,

in a bench-top based study (Federle et al., 2002) on continuous activated-sludge systems,

biodegradation and removal of triclosan radio-labelled with C14 was monitored, and the

results showed that, over a series of triclosan concentrations tested, 94-99.3% of triclosan in

this system underwent primary biodegradation, and only 1.5-4.5% was adsorbed to the waste

solids. The conflict in this literature suggests that although the mechanisms involved in

triclosan removal are similar, the dominant mechanism may vary in different systems, and

they may to some extent influence the removal efficiency.

For an existing wastewater treatment plant, the efficiency of treatment depends on lots of

factors such as property of wastewater influent, property of activated sludge and operational

conditions including aeration rate, temperature etc. The adaptation of activated sludge to a

specific wastewater influent is very important for a successful treatment of this wastewater.

As described previously, hydraulic retention time (HRT) will substantially influence the

efficiency of secondary wastewater treatment.

4.3.3. Biodegradation as a Primary Mechanism to Be Promoted for Triclosan Removal

All studies on the efficiency of removal of triclosan from wastewater treatment systems

are based on the concentration of triclosan residual in the water phase of the effluent after

solids/liquid separation occurs in the secondary sedimentation process. Because of its

polarity, triclosan is readily absorbed into activated sludge, which guarantees a good removal

rate of triclosan (over 90%) (Bester, 2003, McAvoy et al., 2002, Sabaliunas et al., 2003) seen

in most conventional activated sludge processes.

However, better or complete removal of triclosan requires faster biodegradation rate and

longer HRT. Fast biodegradation also leads to a rapid and sustained decrease of triclosan

concentration in the excess sludge, where biodegradation also occurs. As discussed before, a

kinetic balance of triclosan exists between the water phase and the sludge phase, so more

triclosan from water will be absorbed into the sludge phase, if the degradation of triclosan in

the sludge phase takes place at a higher rate. This will consequently reduce the triclosan

concentration in the treated water.

This theory is consistent with published experimental results (Federle et al., 2002), which

described that adaptation of activated sludge to triclosan favoured its removal and it is related

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 221

to the overall treatment efficiency of the system. Based on the above theory, variable

efficiency in the removal of triclosan in treatment processes described in the previous section

could be largely attributed to the properties of activated sludge and variable HRTs in different

systems. These two factors determine the extent of biodegradation that occurs in the sludge.

The better acclimatised sludge has more degradation capacity, and decomposes triclosan more

efficiently, thus achieving a better removal in a system running at a specific HRT. For one

system with pre-established activated sludge composition, the longer the HRT, the more

biodegradation in sludge will occur, and a higher removal rate will be achieved.

This theory also explains the conflicting results of studies on the efficiency of triclosan

removal from excess sludge in the sludge treatment process. A recent mass balance

assessment of triclosan removal from the wastewater treatment process (Heidler and Halden,

2007) revealed that despite the good removal rate of triclosan from the liquid phase, a large

proportion of particle-active triclosan (composed of 80±22% on average of the total triclosan

inflow) was bound to particulates and eventually sequestered and accumulated in excess

sludge in the process. In contrast, in another study (McAvoy et al., 2002) investigating

triclosan in digested sludge, it was revealed that triclosan was adequately removed from

digested sludge particularly when using an aerobic digestion process. Removal of triclosan

from digested sludge under anaerobic conditions was much less efficient.

The above comparison indicates that an aerobic biodegradation of triclosan took place in

the McAvoy study, where a good triclosan removal was achieved in the sludge treatment;

whereas no or minimum biodegradation of triclosan occurred in the Heidler investigation and

triclosan accumulated in the excess sludge. It becomes obvious that biodegradation is a key

mechanism for assured triclosan removal efficiency in the wastewater and sludge treatment

process. Additionally, it is also a more environmentally friendly and sustainable mechanism

that should be promoted in the triclosan treatment process.

As described above, triclosan biodegradation occurs naturally as well as in wastewater

treatment plants. However, the performance varies between plants. Understanding of the

mechanisms involved in triclosan biodegradation will assist in providing information for

process optimization, improve triclosan removal efficiency, and consequently reduce the total

quantity of triclosan present in the environment.

5. THE BIOCHEMISTRY AND GENETICS OF TRICLOSAN

DEGRADATION–WHAT IS KNOWN?

5.1. Microorganisms Involved in Triclosan Biodegradation and Possible

Pathways

Transformation of triclosan in the natural water environment involves photolytic

degradation (Aranami and Readman, 2007), chlorination (Canosa et al., 2005), and

methylation (Coogan et al., 2007). Apart from these degradation pathways that occur

naturally in the water environment, a small scale laboratory based investigation was carried

out on sonochemical degradation of triclosan in various environmental samples including

seawater, urban runoff and influent domestic wastewater, as well as pure and saline water

(Sanchez-Prado et al., 2008). A complete removal of triclosan was achieved in 120 minutes of

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

222

reaction from most samples spiked with 5 µg/L of triclosan, except the influent domestic

wastewater for which only 60% of removal was achieved within 180 minutes. No chlorinated

and other toxic by-products were detected under the conditions studied (Sanchez-Prado et al.,

2008).

Recently, triclosan biodegradation in soil has been investigated under aerobic and

anaerobic conditions (Ying et al., 2007). In this study, a gradual decrease of triclosan

concentration in non-sterile soil was observed from 1.01mg/kg at the beginning of the

experiment to 0.08 mg/kg over 70 days of the investigation period. Interestingly, a

consistently higher dehydrogenase activity was also observed in the triclosan incorporated

soil compared to the trichlocarban incorporated soil, which showed much poorer removal of

trichlocarbon. Aerobic triclosan biodegradation was also discussed in a study on triclosan

removal during the wastewater treatment process (McAvoy et al., 2002). Much higher

triclosan contents in the anaerobically digested sludge than found in aerobic sludge indicated

a much more efficient triclosan biodegradation under aerobic conditions.

A wastewater consortium was found to be able to degrade triclosan (Hay et al., 2001).

Two bacterial strains of Pseudomonas putida and Alcaligenes xylosoxidans subsp.

denitrificans were isolated from soil and they were demonstrated to inactivate triclosan in

liquid and on solid substrates and to use triclosan as a sole carbon source (Meade et al., 2001).

Biotransformation of triclosan by fungi is also reported. The fungus Pycnoporus cinnabarinus

was found to methylate the hydroxyl group of triclosan during cultivation, and the fungus

Trametes versicolor metabolized triclosan and produced three metabolites, namely 2-O-

(2,4,4’-tichlorodiphenyl ether)-β-D-xylopyranoside, 2-O-(2,4,4’-trichlorodiphenyl ether)-β-

D-glucopyranoside, and 2,4-dichlorophenol (Hundt et al., 2000).

Reproduction permitted by JEM.

Figure 2. Degradability of chlorinated organic compounds by biological metabolism (Janssen et al.,

2005).

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 223

The degradability or recalcitrance of chlorinated organic compounds was reviewed by

Janssen (Janssen et al., 2005), where chlorinated compounds were classified by their

degradability as illustrated in Figure 2.

Based on these theories and studies, triclosan could be considered as a moderately

difficult compound that could be selectively degraded by bacteria after prolonged exposure.

Dehalogenation, irrespective of the mechanism (hydrolysis, oxidation or reduction) is an

important step in the biodegradation pathway of a halogenated compound. Often the loss of

halogen from a halogenated compound as a result of dehalogenase activity initiates a potential

catabolic pathway for this compound, although the toxicity of intermediates of this pathway

could also be key determinants in the maturity of this pathway, particularly for compounds

carrying multiple halogen groups (van Hylckama Vlieg, 2000). Chlorinated aliphatic

compounds can be utilized by both aerobic bacteria and strict anaerobic bacteria as growth

substrates (Altenschmidt and Fuchs, 1992, Biegert and Fuchs, 1995, Caldwell et al., 1999,

Leutwein and Heider, 1999, Li et al., 2010, Lovley and Lonergan, 1990, Pries et al., 1994).

Some compounds, such as chloroacetate and 2-chloropropionate can be directly utilized by

bacteria in pure cultures, while others, like di-chlorinated allyl alcohol and tri-chloroallyl

alcohol, were found to be involved in a co-metabolisation pathway (van der Waarde et al.,

1994). Chloroacetate was found to be only co-metabolized by one Pseudomonas strain, which

can only degrade this compound in the presence of another readily metabolisable carbon

source. By comparison, one proteobacteria from the r-subgroup was isolated and was

successfully adapted to live on trichloroacetate as the only carbon and energy source

(Laturnus et al., 2005, Yu and Welander, 1995). Interestingly, this bacterium cannot utilize

monochloroacetate or dichloroacetate as the sole carbon source (Yu and Welander, 1995).

Much less data is available on degradation of chlorinated compounds by anaerobic bacteria,

partly due to the difficulty in culturing and studying anaerobes. An inducible enzyme was

found to be responsible for transferring the methyl group of chloromethane onto

tetrahydrofolate and therefore provide methyl-tetrahydrofolate as a substrate in the acetate

synthesis pathway in a strict anaerobe (Meßmer et al., 1996). Some reductive dehalogenases

were isolated and identified in anaerobic bacteria, such as Dehalospirillum multivorans and

Desulfomonile tiedjei (Louie et al., 1997, Neumann et al., 1995).

In addition, it is well known that aromatic hydrocarbons and their substituted derivatives

must be modified into o-diphenols through different peripheral reactions before ring cleavage

can take place (Galli, 1996). Aromatic ring cleavage occurs through intradiol or extradiol ring

cleavage pathways and a few homologous catechol 2,3-dioxygenases in different bacterial

isolates have been cloned and characterised (Arensdorf and Focht, 1994, Bartilson and

Shingler, 1989, Di Gioia et al., 1998, Kitayama et al., 1996, Moon et al., 1995). Following

ring cleavage products are further metabolised via the TCA cycle.

5.2. Isolation and Growth of P. citronellolis F12 on Triclosan

We have performed enrichments from activated sewage sludge of triclosan degrading

bacteria. Five hundred mL of activated sludge was added to a 1 litre flask along with 1g of

triclosan (2g/L final concentration). The flask was incubated at 35oC with shaking at 160 rpm.

The disappearance of triclosan was followed using HPLC-UV and when levels reduced to

c.200mg/L the flask was augmented with an additional 1 g of triclosan.

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

224

Figure 3. Growth of P. citronellolis F12 in Triclosan minimal medium containing 300mg/L of triclosan.

This process was maintained for 3 months following which colony isolation was

performed on minimal agar containing 0.05% casamino acids and 1g/L triclosan. An isolate

(F12) was obtained from this enrichment that was identified by 16S rRNA gene sequencing as

Pseudomonas citronellolis and named P. citronellolis F12. Members of the genus

Pseudomonas are fast-growing nutritionally versatile bacteria that are capable of utilizing a

wide variety of carbon sources (Lloyd-Jones et al., 2005), and are capable of the degradation

of environmentally important organic compounds (Wackett, 2003).

Figure 3 shows the growth curve for strain F12 in triclosan minimal medium broth. An

increase in bacterial cell count and a corresponding decrease of triclosan concentration in the

culture medium was observed over a 29 day period as monitored by flow cytometric cell

counts. For the two control series, there was no bacterial growth evident in minimal medium

without triclosan and killed bacterial cells inoculated into triclosan minimal medium showed

a zero viable cell count and a constantly high concentration of triclosan in the culture

medium. This indicated that the decrease of triclosan in the culture was due to the growth of

isolate F12.

5.3. Tentative Identification of Biodegradation Intermediates

In this study we applied HPLC-UV, LC-MS and GC-MS to the identification and

quantification of triclosan in liquid medium and detection of putative intermediates

respectively. The putative triclosan biodegradation intermediates or by-products were

tentatively identified as 2’,4’-dichlorophenyl 4-chloro-6-oxo-2,4-hexadienoate, 4-methylene

but-2-en-4-olide (or proto-anemonin), 2,4-dichloro-phenol (Figure 4) and an unknown

compound having close mass spectrum to trichlorobenzene. The identified intermediates are

all possible breakdown products of triclosan. Ortho-cleavage of the hydroxyl substituted ring

of triclosan could yield 2’,4’-dichlorophenyl 4-chloro-6-oxo-2,4-hexadienoate, subsequent

hydrolysis 2,4-dicholorophenol and dehalogenation of the remaining 3-chloromuconic acid

would yield protoanemonin. However, at this stage this breakdown pathway is speculative

and needs to be supported by further high level analysis of intermediates, possibly by nuclear

magnetic resonance analysis.

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 225

Figure 4. Putative triclosan biodegradation intermediates or by-products. (a) 2’,4’-dichlorophenyl 4-

chloro-6-oxo-2,4-hexadienoate. (b) 4-methylene but-2-en-4-olide (or proto-anemonin). (c) 2,4-dichloro-

phenol.

5.4. Genes Involved in Triclosan Catabolism are Plasmid Encoded

It has been known for over 30 years now that genes involved in the catabolism of

xenobiotic compounds can be associated with extra-chromosomal genetic elements known as

plasmids (Downing and Broda, 1980). In fact, with regard to what is understood about the

degradation of chlorinated organics, genes for their dissimilation are almost always plasmid

encoded.

The interest in bacterial catabolic plasmids that are naturally present in soil or water

bodies has grown over this time, due to the increasing amounts of toxic halogenated organic

pollutants released to the environment in the last few decades (Bernhard et al., 2008, Feakin

et al., 1994, Fewson, 1988, Haro and de Lorenzo, 2001, Kobayashi et al., 2009, Okpokwasili

and Olisa, 1991, Sinton et al., 1986).

Plasmids, as well as other mobile genetic elements in environmental bacteria, have been

demonstrated to play an important role in the acquisition of new catabolic functions. Through

the evolution of different catabolic pathways such bacteria are important in the degradation of

a range of environmental pollutants (Nojiri et al., 2004).

We undertook an examination of P. citronellolis F12 to ascertain whether its ability to

degrade triclosan was directed by genes encoded on a catabolic plasmid. A plasmid was

isolated from F12 using standard DNA extraction procedures (Qiu, 2012) and subjected to

restriction enzyme analysis and gel electrophoresis, a plasmid of c.17kb in size was identified

(Figure 5).The plasmid in the wild type F12 strain was designated pF12.

Plasmid associated genes are often lost at relatively high frequencies by a process known

as plasmid curing. Essentially, this involves growing the host strain under non-selective

conditions that promote plasmid loss. The maintenance and replication of plasmid DNA in the

bacterial cell becomes a burden under non-selective conditions and cells that have

spontaneously lost the plasmid gain a competitive advantage and can become amplified in the

population as a whole due to their preferential growth rate. These “cured” cells can be

isolated and identified in the laboratory. Growth at elevated temperature is known to be

effective in removing plasmids from their hosts and we ascertained that strain F12 grew

slowly at 46oC but not at all at 48oC. The strain was inoculated into LB broth and subcultured

3 times allowing for growth through at least 30 generations. Screening of subsequent colonies

derived from nutrient agar revealed that approximately 10% of the colonies no longer grew on

triclosan.

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

226

Figure 5. Characterisation of pF12. From left to right: Lane 1 DNA molecular weight marker III (0.12-

21.2kbp) (λDNA•EcoRI and Hind III digested); Lane 2 Plasmid preparation from F12; Lane 3 Mock

digestion of pF12 (no enzyme but nuclear free water was added); Lane 4 EcoRI digestion of pF12; Lane

5 NotI digestion of pF12; Lane 6 EcoRI and NotI double digestion of pF12.

Screening of a selection of these colonies by DNA extraction and restriction enzyme

analysis revealed that all had lost the17Kb plasmid previously present in the F12 strain. One

of these isolates was confirmed by 16S rRNA gene sequencing as being otherwise identical to

the F12 strain. Subsequently, pF12 was completely DNA sequenced using standard

procedures (Qiu, 2012) and found to be 17.77Kb in size.

The DNA sequence derived for pF12 was compared to known sequences within the DNA

databases and a preliminary map outlining the functional regions of the plasmid constructed

(Figure 5). The catabolic region involved in triclosan degradation is situated in a region just

under 6 Kb in size. Analysis so far has revealed this region could encode several hypothetical

proteins with known homologies to enzymes involved in catabolic functions. However, it is

unlikely that this region could encode the entire suite of catabolic enzymes required for the

complete mineralisation of triclosan. This constitutes the first report of the isolation and

preliminary characterisation of a plasmid directly involved in triclosan dissimilation.

6. TRICLOSAN AND SELECTION OF ANTIMICROBIAL RESISTANT

BACTERIA

6.1. Evidence for Triclosan Induced Multiple Resistance

Apart from concerns about intrinsic and acquired resistance to triclosan in various

bacterial species, there is increasing concern that resistance to triclosan could select for

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 227

resistance to clinically important antimicrobials. In 1991 a study reported that methicillin-

resistant Staphylococcus aureus with low-level resistance to triclosan that had been isolated

from patients receiving daily triclosan baths were also resistant to mupirocin (Cookson et al.,

1991). Levy and his group (2002) investigated the link between mutations in efflux pump

related genes and significant increases in triclosan resistance in Esherichia coli. Efflux pumps

are associated with resistance to a diverse range of antimicrobials including fluoroquinolones,

tetracyclines and glycylcyclines, chloramphenicol and florfenicol, macrolides, lincosamides

and ketolides, β-lactams, aminoglycosides and oxazolinidones (Poole, 2005). However,

studies demonstrating a clinical link between triclosan resistance and resistance to other

antimicrobials are rare. In laboratory studies Chuanchuen et al (2001) found that

Pseudomonas aeruginosa strains resistant to triclosan because of the presence of specific

efflux pumps were also resistant to a number of antimicrobials. Similarly, resistance to

quinolones was found to be induced in Stenotrophomonas maltophilia by use of triclosan and

the up-regulation of efflux pumps (Hernández et al., 2011).

However, Cottell et al (2009) did not find any increased resistance to triclosan-tolerant

strains of E. coli, Staphylococcus aureus or Acinetobacter johnsonii. In fact they noted that

triclosan-tolerant strains of E. coli were more susceptible to gentamicin than wild-type strains.

Similarly, Suller and Russell (2000) could not replicate the work of Cookson et al. (1991) and

were unable to demonstrate any link between triclosan resistance and resistance to mupirocin

or other antimicrobials. Other workers have reported similar findings with S. aureus and

MRSA (Yazdankhah et al., 2006). It should be noted here that efflux pumps are not the only

mechanism of resistance to triclosan, with mutations in Fab1, an enzyme in the fatty acid

biosynthetic pathway, also important. In contrast to the S. aureus studies, there is a recent

report of sublethal exposure of Listeria monocytogenes to triclosan resulting in decreased

resistance to gentamicin and other aminoglycosides (Christensen et al., 2011). The detailed

molecular mechanisms remain to be elucidated.

However, of concern is a very recent report of an apparent link between triclosan

resistance and antibiotic resistance in a clinical setting. Strains of Pseudomonas aeruginosa

isolated from an epidemic in immunocompromised patients in an oncohaematology ward

were of identical genotypes and showed identical resistance patterns (resistant to

fluoroquinolones and aminoglycosides) as well as high level resistance to triclosan. The

source of infection was found to be a common soap dispenser which at the time of the

outbreak contained triclosan soap (D’Arezzo S et al., 2012). So while the evidence is still

scant it seems prudent to be mindful of the potential for selection of multiple antimicrobial

resistance in pathogens from exposure to triclosan.

6.2. Could Wastewater Play a Role in the Amplification of Antibiotic

Resistant Bacteria?

Antibiotic resistant bacteria and antibiotic residues have been found in wastewater

(Baquero et al., 2008). Many of the studies concentrated on faecal organisms and/or human

pathogens. Enterococci resistant to tetracyclines and fluoroquinolones and faecal coliforms

resistant to amoxicillin, tetracyclines and ciprofloxacin were found by Ferreira da Siva et al.,

(2006, 2007). Lefkowitz and Duran (2009) reported resistant strains of E. coli in the different

stages of wastewater treatment, with significant resistance detected to tetracyclines,

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

228

streptomycin and beta-lactams. They noted that the percentage of isolates with multiple

resistance increased through the treatment process. Faecal coliforms resistant to

fluoroquinolones and trimethoprim-sulphamethoxazole and enterococci (principally E.

faecalis) resistant to vancomycin were found by Nagulapally et al. (2009) and vancomycin

resistant enterococci have also been detected in sewage sludge (Sahlström et al., 2009).

Studies of hospital sewage treatment plants have yielded extended-spectrum β-lactamase

producing organisms (Prado et al., 2008, Uyaguari et al., 2011) and gentamicin resistant

strains of E. coli (Jakobsen et al., 2008).

Studies of environmental organisms in wastewater have also revealed microbial

resistance. An increase in the percentage of isolates resistant to clinically important

antibiotics such as amoxicillin/clavulanate, chloramphenicol and rifampicin (and multiple

resistance) through the steps in treatment has been reported in Acinetobacter spp. (Zhang et

al., 2009) and ticarcillin and quinolone resistance has been reported in Aeromonas spp.

(Figueira et al., 2011) and fluoroquinolone resistance in Pseudomonas spp. (Schwartz et al.,

2006). No studies have investigated transfer of antimicrobial resistance genes in wastewater.

However, a number of recent reports have studied the presence of plasmids carrying antibiotic

resistance genes, including class 1 and 2 integrons (Akiyama et al., 2010, Pellegrini et al.,

2011, Pignato et al., 2009).

If mobile resistance plasmids carrying resistance genes are present in the wastewater

microbial flora, then the presence of antimicrobial residues and/or triclosan (Saleh et al.,

2011, Schweizer, 2001) is likely to drive the transmission of these plasmids (and their

resistance genes) to other bacteria. This could contribute to the dissemination of antibiotic

resistant bacteria into the environment.

CONCLUSION

Microorganisms are highly adaptive and by a process of oxidation or reduction most

xenobiotics can be degraded, although rates are very much influenced by temperature and

generally proceed at a more rapid rate under oxygenic conditions. For more than 50 years

now there has been an interest amongst microbiologists and biochemists to charaterise the

microorganisms capable of degrading recalcitrant man-made compounds and elucidate the

biochemical pathways involved in their dissimilation.

Triclosan is one such mass produced compound that is ubiquitous in the developed world

and its degradation is of interest not merely because of its environmental significance but also

because of its implication in the promotion of antibiotice resistance amongst the microbial

community. In this review data has been summarized that deals with the concerns regarding

this compounds presence in the environment and the efficacy of engineered and natural

processes in dealing with its removal.

Microbial ecology is a growing field of interest due to our improved understanding of

microbial processes and the critical role these play in maintaining a natural equilibrium.

Something that microbiologists know little about is the influence that the presence of

xenobiotics can have on the microbial biodiversity of an ecosystem. Whilst microbial removal

of contaminants might be seen as a beneficial outcome, if the organisms that carry out these

conversions prosper to the detriment or demise of microorganisms that play an important part

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 229

in cycling processes then this could have potentially devastating effects on the equilibrium of

a particular ecosystem. This is of particular relevance when considering what role this may

play in further exacerbating the changes in natural processes being enacted through climate

change.

So whilst microbial removal of contaminants might ostensibly appear to be a good thing,

there could be unforeseen challenges created by unbalancing an established ecosystem.

Although we have not considered this in detail here, excitingly, the molecular tools are now

becoming available that will permit detailed study of natural microbial communities and

begin to answer some of the questions surrounding these issues.

ACKNOWLEDGMENTS

This work was funded by Water Quality Research Australia (formerly the CRC for Water

Quality and Treatment). The authors would like to acknowledge Dr. Alexandra Keegan (SA

Water), Dr. Michael Heuzenroeder (IMVS) and Dr Zuliang Chen (UniSA) for their kind

assistance and guidance on the laboratory work. We would also like to acknowledge the

scholarships awarded to TQ by the IPRS program and the CRC for Water Quality and

Treatment.

REFERENCES

Aiello, A. E., Marshall, B., Levy, S. B., Della-Latta, P. and Larson, E. 2004 'Relationship

between triclosan and susceptibilities of bacteria isolated from hands in the community',

Antimicrobial Agents Chemotherapy, vol 48, no 8, pp. 2973-2979.

Akiyama, T., Asfahl, K. L. and Savin, M. C. 2010 'Broad-host-range plasmids in treated

wastewater effluent and receiving streams', Journal of Environmental Quality, vol 39,

no 6, pp. 2211-2215.

Altenschmidt, U. and Fuchs, G. 1992 'Anaerobic toluene oxidation to benzyl alcohol and

benzaldehyde in a denitrifying Pseudomonas strain', Journal of Bacteriology, vol 174,

no 14, pp. 4860-4862.

Aranami, K. and Readman, J. W. 2007 'Photolytic degradation of triclosan in freshwater and

seawater', Chemosphere, vol 66, no 6, pp. 1052-1056.

Arensdorf, J. J. and Focht, D. D. 1994 'Formation of chlorocatechol meta cleavage products

by a pseudomonad during metabolism of monochlorobiphenyls', Applied and

Environmental Microbiology, vol 60, no 8, pp. 2884-2889.

Baquero, F., Martínez, J. L. and Cantón, R. 2008 'Antibiotics and antibiotic resistance in

water environments', Current Opinion in Biotechnology, vol 19, no 3, pp. 260-265.

Barry, M. A., Craven, D. E., Goularte, T. A. and Lichtenberg, D. A. 1984 'Serratia

marcescens contamination of antiseptic soap containing triclosan: implications for

nosocomial infection', Infection Control, vol 5, no 9, pp. 427-430.

Bartilson, M. and Shingler, V. 1989 'Nucleotide sequence and expression of the catechol

2,3-dioxygenase-encoding gene of phenol-catabolizing Pseudomonas CF600', Gene, vol

85, no 1, pp. 233-238.

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

230

Bernhard, M., Eubeler, J. P., Zok, S. and Knepper, T. P. 2008 'Aerobic biodegradation of

polyethylene glycols of different molecular weights in wastewater and seawater', Water

Research, vol 42, no 19, pp. 4791-4801.

Bester, K. 2003 'Triclosan in a sewage treatment process--balances and monitoring data',

Water Research, vol 37, no 16, pp. 3891-3896.

Biegert, T. and Fuchs, G. 1995 'Anaerobic oxidation of toluene (analogues) to benzoate

(analogues) by whole cells and by cell extracts of a denitrifying Thauera sp', Archives

of Microbiology, vol 163, no 6, pp. 407-417.

Boyd, G. R., Reemtsma, H., Grimm, D. A. and Mitra, S. 2003 'Pharmaceuticals and

personal care products (PPCPs) in surface and treated waters of Louisiana, USA and

Ontario, Canada', The Science of The Total Environment, vol 311, no 1-3, pp. 135-149.

Braoudaki, M. and Hilton, A. C. 2004a 'Adaptive resistance to biocides in Salmonella

enterica and Escherichia coli O157 and cross-resistance to antimicrobial agents',

Journal of Clinical Microbiology, vol 42, no 1, pp. 73-78.

Braoudaki, M. and Hilton, A. C. 2004b 'Low level of cross-resistance between triclosan and

antibiotics in Escherichia coli K-12 and E. coli O55 compared to E. coli O157', FEMS

Microbiology Letters, vol 235, no 2, pp. 305-309.

Caldwell, M. E., Tanner, R. S. and Suflita, J. M. 1999 'Microbial metabolism of benzene

and the oxidation of ferrous iron under anaerobic conditions: Implications for

bioremediation', Anaerobe, vol 5, no 6, pp. 595-603.

Canosa, P., Morales, S., Rodríguez, I., Rubí, E., Cela, R. and Gómez, M. 2005 'Aquatic

degradation of triclosan and formation of toxic chlorophenols in presence of low

concentrations of free chlorine', Analytical and Bioanalytical Chemistry, vol 383, no 7,

pp. 1119-1126.

Chau, W. C., Wu, J.-l. and Cai, Z. 2008 'Investigation of levels and fate of triclosan in

environmental waters from the analysis of gas chromatography coupled with ion trap

mass spectrometry', Chemosphere, vol 73, no 1, Supplement 1, pp. S13-17.

Christensen, E. G., Gram, L. and Kastbjerg, V. G. 2011 'Sublethal triclosan exposure

decreases susceptibility to gentamicin and other aminoglycosides in Listeria

monocytogenes, Antimicrobial Agents and Chemotherapy, vol 55, no 9, pp. 4064-

4071.

Chuanchuen, R., Beinlich, K., Hoang, T. T., Becher, A., Karkhoff-Schweizer, R. R. and

Schweizer, H. P. 2001 'Cross-resistance between triclosan and antibiotics in

Pseudomonas aeruginosa is mediated by multidrug efflux pumps: exposure of a

susceptible mutant strain to triclosan selects nfxB mutants overexpressing MexCD-

OprJ', Antimicrobial Agents and Chemotherapy, vol 45, no 2, pp. 428-432.

Chuanchuen, R., Narasaki, C. T. and Schweizer, H. P. 2002 'The MexJK efflux pump of

Pseudomonas aeruginosa requires OprM for antibiotic efflux but not for efflux of

Triclosan', Journal of Bacteriology., vol 184, no 18, pp. 5036-5044.

Chung, W. W. P., Rafqah, S., Voyard, G. and Sarakha, M. 2007 'Photochemical behaviour

of triclosan in aqueous solutions: kinetic and analytical studies', Journal of

Photochemistry and Photobiology A: Chemistry, vol 191, no 2-3, pp. 201-208.

Coogan, M. A., Edziyie, R. E., La Point, T. W. and Venables, B. J. 2007 'Algal

bioaccumulation of triclocarban, triclosan, and methyl-triclosan in a North Texas

wastewater treatment plant receiving stream', Chemosphere, vol 67, no 10, pp. 1911-

1918.

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 231

Cookson, B. D., Farrelly, H., Stapleton, P., Garvey, R. P. J. and Price, M. R. 1991

'Transferable resistance to triclosan in MRSA', The Lancet, vol 337, no 8756, pp. 1548-

1549.

Copitch, J. L., Whitehead, R. N. and Webber, M. A. 2010 'Prevalence of decreased

susceptibility to triclosan in Salmonella enterica isolates from animals and humans and

association with multiple drug resistance', International Journal of Antimicrobial

Agents, vol 36, no 3, pp. 247-251.

Cottell, A., Denyer, S. P., Hanlon, G. W., Ochs, D. and Maillard, J. Y. 2009 'Triclosan-

tolerant bacteria: changes in susceptibility to antibiotics', Journal of Hospital Infection,

vol 72, no 1, pp. 71-76.

D’Arezzo S, Lanini S, Puro V, Ippolito G and P, V. 2012 High-level tolerance to triclosan

may play a role in Pseudomonas aeruginosa antibiotic resistance in

immunocompromised hosts: evidence from an outbreak investigation. BMC Research

Notes.

Dayan, A. D. 2007 'Risk assessment of triclosan [Irgasan] in human breast milk', Food and

Chemical Toxicology, vol 45, no 1, pp. 125-129.

Di Gioia, D., Fava, F., Baldoni, F. and Marchetti, L. 1998 'Characterization of catechol- and

chlorocatechol-degrading activity in the ortho-chlorinated benzoic acid-degrading

Pseudomonas sp. CPE2 strain', Research in Microbiology, vol 149, no 5, pp. 339-348.

Downing, R. G. and Broda, P. 1980 'A cleavage map of the TOL plasmid of Pseudomonas

putida mt-2', Molecular and General Genetics, vol 177, no pp. 189-191.

Ellison, M. L., Roberts, A. L. and Champlin, F. R. 2007 'Susceptibility of compound 48/80-

sensitized Pseudomonas aeruginosa to the hydrophobic biocide triclosan', FEMS

Microbiology Letters, vol 269, no 2, pp. 295-300.

Faoagali, J. L., George, N., Fong, J., Davy, J. and Dowser, M. 1999 'Comparison of the

antibacterial efficacy of 4% chlorhexidine gluconate and 1% triclosan handwash

products in an acute clinical ward', American Journal of Infection Control, vol 27, no 4,

pp. 320-326.

Feakin, S. J., Blackburn, E. and Burns, R. G. 1994 'Biodegradation of s-triazine herbicides

at low concentrations in surface waters', Water Research, vol 28, no 11, pp. 2289-2296.

Federle, T. W., Kaiser, S. K. and Nuck, B. A. 2002 'Fate and effects of triclosan in activated

sludge', Environmental Toxicology and Chemistry, vol 21, no 7, pp. 1330-1337.

Ferreira Da Silva, M., Tiago, I., Verrissimo, A., Boaventura, R. A., Nunes, O. C. and

Manaia, C. M. 2006 'Antibiotic resistance of enterococci and related bacteria in an

urban wastewater treatment plant', FEMS Microbiology Ecology, vol 55, no 2, pp. 322-

329.

Ferreira Da Silva, M., Vaz-Moreira, I., Gonzalez-Pajuelo, M., Nunes, O. C. and Manaia, C.

M. 2007 'Antimicrobial resistance patterns in Enterobacteriaceae isolated from an

urban wastewater treatment plant', FEMS Microbiology Ecology, vol 60, no 1, pp. 166-

176.

Fewson, C. A. 1988 'Biodegradation of xenobiotic and other persistent compounds: the

causes of recalcitrance', Trends in Biotechnology, vol 6, no 7, pp. 148-153.

Figueira, V., Vaz-Moreira, I., Silva, M. and Manaia, C. M. 2011 'Diversity and antibiotic

resistance of Aeromonas spp. in drinking and waste water treatment plants', Water

Research, vol 45, no 17, pp. 5599-5611.

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

232

Fine, D. H., Furgang, D., Bonta, Y., DeVizio, W., Volpe, A. R., Reynolds, H., Zambon, J. J.

and Dunford, R. G. 1998 'Efficacy of a triclosan/NaF dentifrice in the control of plaque

and gingivitis and concurrent oral microflora monitoring', American Journal of

Dentistry, vol 11, no 6, pp. 259-270.

Galli, E. 1996 Alternative pathways for biodegradation of alkyl and alkenyl benzenes. In

Nakazawa, T., Furukawa, K., Haas, D. and Silver, S. (Eds.) Molecular biology of

Pseudomonads: Proceedings of the Fifth International Symposium on Pseudomonads:

Molecular Biology and Biotechnology, in Tsukuba, Japan, August 1995. Washington,

DC, ASM Press.

Gomez Escalada, M., Harwood, J. L., Maillard, J.-Y. and Ochs, D. 2005a 'Triclosan

inhibition of fatty acid synthesis and its effect on growth of Escherichia coli and

Pseudomonas aeruginosa', Journal of Antimicrobial Chemotherapy, vol 55, no 6, pp.

879-882.

Gomez Escalada, M., Russell, A. D., Maillard, J.-Y. and Ochs, D. 2005b 'Triclosan-bacteria

interactions: single or multiple target sites?' Letters in Applied Microbiology, vol 41, no

6, pp. 476-481.

Haro, M.-A. and de Lorenzo, V. 2001 'Metabolic engineering of bacteria for environmental

applications: construction of Pseudomonas strains for biodegradation of 2-

chlorotoluene', Journal of Biotechnology, vol 85, no 2, pp. 103-113.

Hay, A. G., Dees, P. M. and Sayler, G. S. 2001 'Growth of a bacterial consortium on

triclosan', FEMS Microbiology Ecology, vol 36, no 2-3, pp. 105-112.

Heath, R. J., Li, J., Roland, G. E. and Rock, C. O. 2000 'Inhibition of the Staphylococcus

aureus NADPH-dependent enoyl-acyl carrier protein reductase by triclosan and

hexachlorophene', Journal of Biological Chemistry, vol 275, no 7, pp. 4654-4659.

Heath, R. J., Rubin, J. R., Holland, D. R., Zhang, E., Snow, M. E. and Rock, C. O. 1999

'Mechanism of triclosan inhibition of bacterial fatty acid synthesis', Journal of

Biological Chemistry, vol 274, no 16, pp. 11110-11114.

Heidler, J. and Halden, R. U. 2007 'Mass balance assessment of triclosan removal during

conventional sewage treatment', Chemosphere, vol 66, no 2, pp. 362-369.

Hernández, A., Ruiz, F. M., Romero, A. and Martínez, J. L. 2011 'The binding of triclosan

to SmeT, the repressor of the multidrug efflux pump SmeDEF, induces antibiotic

resistance in Stenotrophomonas maltophilia', PLoS pathogens, vol 7, no 6, pp. 1-12,

e1002103.

Hundt, K., Martin, D., Hammer, E., Jonas, U., Kindermann, M. K. and Schauer, F. 2000

'Transformation of triclosan by Trametes versicolor and Pycnoporus cinnabarinus',

Applied and Environmental Microbiology, vol 66, no 9, pp. 4157-4160.

Jackson, J. and Sutton, R. 2008 'Sources of endocrine-disrupting chemicals in urban

wastewater, Oakland, CA', Science of The Total Environment, vol 405, no 1-3, pp. 153-

160.

Jakobsen, L., Sandvang, D., Hansen, L. H., Bagger-Skjøt, L., Westh, H., Jørgensen, C.,

Hansen, D. S., Pedersen, B. M., Monnet, D. L., Frimodt-Møller, N., Sørensen, S. J. and

Hammerum, A. M. 2008 'Characterisation, dissemination and persistence of gentamicin

resistant Escherichia coli from a Danish university hospital to the waste water

environment', Environment International, vol 34, no 1, pp. 108-115.

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 233

Janssen, D. B., Dinkla, I. J. T., Poelarends, G. J. and Terpstra, P. 2005 'Bacterial

degradation of xenobiotic compounds: evolution and distribution of novel enzyme

activities', Environmental Microbiology, vol 7, no 12, pp. 1868-1882.

Kapoor, M., Mukhi, P. L., Surolia, N., Suguna, K. and Surolia, A. 2004 'Kinetic and

structural analysis of the increased affinity of enoyl-ACP (acyl-carrier protein)

reductase for triclosan in the presence of NAD+', The Biochemical Journal, vol 381, no

Pt 3,pp. 725-733.

Kitayama, A., Achioku, T., Yanagawa, T., Kanou, K., Kikuchi, M., Ueda, H., Suzuki, E.,

Nishimura, H., Nagamune, T. and Kawakami, Y. 1996 'Cloning and characterization of

extradiol aromatic ring-cleavage dioxygenases of Pseudomonas aeruginosa JI104',

Journal of Fermentation and Bioengineering, vol 82, no 3, pp. 217-223.

Kobayashi, T., Murai, Y., Tatsumi, K. and Iimura, Y. 2009 'Biodegradation of polycyclic

aromatic hydrocarbons by Sphingomonas sp. enhanced by water-extractable organic

matter from manure compost', Science of The Total Environment, vol 407, no 22, pp.

5805-5810.

Kuo, M. R. 2003 'Targeting tuberculosis and malaria through inhibition of enoyl reductase',

The Journal of Biological Chemistry, vol 278, no 23, pp. 20851-20860.

Kuster, M., Liez de Alda, M. J., Hernando, M. D., Petrovic, M., Marti-Alonso, J. and

Barcel, D. 2008 'Analysis and occurrence of pharmaceuticals, estrogens, progestogens

and polar pesticides in sewage treatment plant effluents, river water and drinking water

in the Llobregat river basin (Barcelona, Spain)', Journal of Hydrology, vol 358, no 1-2,

pp.

112-123.

Laturnus, F., Fahimi, I., Gryndler, M., Hartmann, A., Heal, M., Matucha, M., Schöler, H. F.,

Schroll, R. and Svensson, T. 2005 'Natural formation and degradation of chloroacetic

acids and volatile organochlorines in forest soil. Challenges to understanding',

Environmental Science and Pollution Research, vol 12, no 4, pp. 233-244.

Ledder, R. G., Gilbert, P., Willis, C. and McBain, A. J. 2006 'Effects of chronic triclosan

exposure upon the antimicrobial susceptibility of 40 ex-situ environmental and human

isolates', Journal of Applied Microbiology, vol 100, no 5, pp. 1132-1140.

Lefkowitz, J. R. and Duran, M. 2009 'Changes in antibiotic resistance patterns of

Escherichia coli during domestic wastewater treatment', Water Environment Research,

vol 81, no 9, pp. 878-885.

Leutwein, C. and Heider, J. 1999 'Anaerobic toluene-catabolic pathway in denitrifying

Thauera aromatica : activation and β-oxidation of the first intermediate, (R)-(+)-

benzylsuccinate', Microbiology, vol 145, no 11, pp. 3265-3271.

Levy, C. W., Roujeinikova, A., Sedelnikova, S., Baker, P. J., Stuitje, A. R., Slabas, A. R.,

Rice, D. W. and Rafferty, J. B. 1999 'Molecular basis of triclosan activity', Nature, vol

398, no 6726, pp. 383-384.

Levy, S. B. 2002 'Active efflux, a common mechanism for biocide and antibiotic resistance'.

Symposium series (Society for Applied Microbiology), 2002; (31): 65S-71S

Li, C. H., Wong, Y. S. and Tam, N. F. Y. 2010 'Anaerobic biodegradation of polycyclic

aromatic hydrocarbons with amendment of iron(III) in mangrove sediment slurry',

Bioresource Technology, vol 101, no 21, pp. 8083-8092.

Lishman, L., Smyth, S. A., Sarafin, K., Kleywegt, S., Toito, J., Peart, T., Lee, B., Servos,

M., Beland, M. and Seto, P. 2006 'Occurrence and reductions of pharmaceuticals and

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

234

personal care products and estrogens by municipal wastewater treatment plants in

Ontario, Canada', Science of The Total Environment, vol 367, no 2-3, pp. 544-558.

Lloyd-Jones, G., Laurie, A. D. and Tizzard, A. C. 2005 'Quantification of the Pseudomonas

population in New Zealand soils by fluorogenic PCR assay and culturing techniques',

Journal of Microbiological Methods, vol 60, no 2, pp. 217-224.

Lores, M., Llompart, M., Sanchez-Prado, L., Garcia-Jares, C. and Cela, R. 2005

'Confirmation of the formation of dichlorodibenzo-p-dioxin in the photodegradation of

triclosan by photo-SPME', Analytical and Bioanalytical Chemistry, vol 381, no 6, pp.

1294-1298.

Louie, T. M., Ni, S., Xun, L. and Mohn, W. W. 1997 'Purification, characterization and

gene sequence analysis of a novel cytochrome c co-induced with reductive

dechlorination activity in Desulfomonile tiedjei DCB-1', Archives of Microbiology, vol

168, no 6, pp. 520-527.

Lovley, D. R. and Lonergan, D. J. 1990 'Anaerobic oxidation of toluene, phenol, and p-

cresol by the dissimilatory iron-reducing organism, GS-15', Applied and Environmental

Microbiology, vol 56, no 6, pp. 1858-1864.

McAvoy, D. C., Schatowitz, B., Jacob, M., Hauk, A. and Eckhoff, W. S. 2002

'Measurement of triclosan in wastewater treatment systems', Environmental Toxicology

and Chemistry, vol 21, no 7, pp. 1323-1329.

McBain, A. J., Bartolo, R. G., Catrenich, C. E., Charbonneau, D., Ledder, R. G. and Gilbert,

P. 2003 'Effects of triclosan-containing rinse on the dynamics and antimicrobial

susceptibility of in vitro plaque ecosystems', Antimicrobial Agents Chemotherapy, vol

47, no 11, pp. 3531-3538.

McBain, A. J., Ledder, R. G., Sreenivasan, P. and Gilbert, P. 2004 'Selection for high-level

resistance by chronic triclosan exposure is not universal', Journal of Antimicrobial

Chemotherapy, vol 53, no 5, pp. 772-777.

McMurry, L. M., McDermott, P. F. and Levy, S. B. 1999 'Genetic evidence that InhA of

Mycobacterium smegmatis is a target for triclosan', Antimicrobial Agents and

Chemotherapy, vol 43, no 3, pp. 711-713.

McMurry, L. M., Oethinger, M. and Levy, S. B. 1998 'Triclosan targets lipid synthesis',

Nature, vol 394, no 6693, pp. 531-532.

Meade, M. J., Waddell, R. L. and Callahan, T. M. 2001 'Soil bacteria Pseudomonas putida

and Alcaligenes xylosoxidans subsp. denitrificans inactivate triclosan in liquid and solid

substrates', FEMS Microbiology Letters, vol 204, no 1, pp. 45-48.

Meßmer, M., Reinhardt, S., Wohlfarth, G. and Diekert, G. 1996 'Studies on methyl chloride

dehalogenase and O-demethylase in cell extracts of the homoacetogen strain MC based

on a newly developed coupled enzyme assay', Archives of Microbiology, vol 165, no 1,

pp. 18-25.

Miege, C., Choubert, J. M., Ribeiro, L., Eusebe, M. and Coquery, M. 2008 'Removal

efficiency of pharmaceuticals and personal care products with varying wastewater

treatment processes and operating conditions - conception of a database and first

results', Water Science and Technology, vol 57, no 1, pp. 49-56.

Moon, J. H., Chang, H., Min, K. R. and Kim, Y. S. 1995 'Cloning and sequencing of the

catechol 2,3-dioxygenase gene of Alcaligenes sp. KF711', Biochemical and Biophysical

Research Communications, vol 208, no 3, pp. 943-949.

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 235

Nagulapally, S. R., Ahmad, A., Henry, A., Marchin, G. L., Zurek, L. and Bhandari, A. 2009

'Occurrence of ciprofloxacin-, trimethoprim-sulfamethoxazole-, and vancomycin-

resistant bacteria in a municipal wastewater treatment plant', Water Environment

Research, vol 81, no 1, pp. 82-90.

Neumann, A., Wohlfarth, G. and Diekert, G. 1995 'Properties of tetrachloroethene and

trichloroethene dehalogenase of Dehalospirillum multivorans', Archives of

Microbiology, vol 163, no 4, pp. 276-281.

Nojiri, H., Shintani, M. and Omori, T. 2004 'Divergence of mobile genetic elements

involved in the distribution of xenobiotic-catabolic capacity', Applied Microbiology and

Biotechnology, vol 64, no 2, pp. 154-174.

Okpokwasili, G. C. and Olisa, A. O. 1991 'River-water biodegradation of surfactants in

liquid detergents and shampoos', Water Research, vol 25, no 11, pp. 1425-1429.

Okumura, T. and Nishikawa, Y. 1996 'Gas chromatography--mass spectrometry

determination of triclosan in water, sediment and fish samples via methylation with

diazomethane', Analytica Chimica Acta, vol 325, no 3, pp. 175-184.

Parikh, S. L., Xiao, G. and Tonge, P. J. 2000 'Inhibition of InhA, the enoyl reductase from

Mycobacterium tuberculosis, by triclosan and isoniazid', Biochemistry, vol 39, no 26,

pp. 7645-7650.

Pellegrini, C., Celenza, G., Segatore, B., Bellio, P., Setacci, D., Amicosante, G. and Perilli,

M. 2011 'Occurrence of class 1 and 2 integrons in resistant enterobacteriaceae collected

from a urban wastewater treatment plant: First report from central Italy', Microbial

Drug Resistance, vol 17, no 2, pp. 229-234.

Pignato, S., Coniglio, M. A., Faro, G., Weill, F. X. and Giammanco, G. 2009 'Plasmid-

mediated multiple antibiotic resistance of Escherichia coli in crude and treated

wastewater used in agriculture', Journal of Water and Health, vol 7, no 2, pp. 251-258.

Poole, K. 2005 'Efflux-mediated antimicrobial resistance', Journal of Antimicrobial

Chemotherapy, vol 56, no 1, pp. 20-51.

Prado, T., Pereira, W. C., Silva, D. M., Seki, L. M., Carvalho, A. P. D. A. and Asensi, M. D.

2008 'Detection of extended-spectrum β-lactamase-producing Klebsiella pneumoniae in

effluents and sludge of a hospital sewage treatment plant', Letters in Applied

Microbiology, vol 46, no 1, pp. 136-141.

Pries, F., van der Ploeg, J. R., Dolfing, J. and Janssen, D. B. 1994 'Degradation of

halogenated aliphatic compounds: The role of adaptation', FEMS Microbiology

Reviews, vol 15, no 2-3, pp. 279-295.

Qiu, T. 2012 Biodegradation of triclosan as a representative of pharmaceuticals and

personal care products (PPCPs) in the wastewater environment, PhD thesis, Adelaide,

University of South Australia.

Randall, L. P., Cooles, S. W., Piddock, L. J. V. and Woodward, M. J. 2004 'Effect of

triclosan or a phenolic farm disinfectant on the selection of antibiotic-resistant

Salmonella enterica', Journal of Antimicrobial Chemotherapy, vol 54, no 3, pp. 621-

627.

Regos, J., Zak, O. and Solf, R. 1979 'Antimicrobial spectrum of triclosan, a broad-spectrum

antimicrobial agent for topical application. II. Comparison with some other

antimicrobial agents', Dermatologica, vol 158, no 1, pp. 72-79.

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

236

Sabaliunas, D., Webb, S. F., Hauk, A., Jacob, M. and Eckhoff, W. S. 2003 'Environmental

fate of triclosan in the River Aire Basin, UK', Water Research, vol 37, no 13, pp. 3145-

3154.

Sahlström, L., Rehbinder, V., Albihn, A., Aspan, A. and Bengtsson, B. 2009 'Vancomycin

resistant enterococci (VRE) in Swedish sewage sludge', Acta Veterinaria Scandinavica,

vol 51, no 24, pp. 1-9.

Saleh, S., Haddadin, R. N. S., Baillie, S. and Collier, P. J. 2011 'Triclosan - an update',

Letters in Applied Microbiology, vol 52, no 2, pp. 87-95.

Samsøe-Petersen, L., Winther-Nielsen, M. and Madsen, T. 2003 Chemicals: Fate and

Effects of Triclosan. Environmental Project No. 861 2003. Hørsholm, Danish

Environmental Potection Agency.

Sanchez-Prado, L., Barro, R., Garcia-Jares, C., Llompart, M., Lores, M., Petrakis, C.,

Kalogerakis, N., Mantzavinos, D. and Psillakis, E. 2008 'Sonochemical degradation of

triclosan in water and wastewater', Ultrasonics Sonochemistry, vol 15, no 5, pp. 689-

694.

Schwartz, T., Volkmann, H., Kirchen, S., Kohnen, W., Schön-Hölz, K., Jansen, B. and

Obst, U. 2006 'Real-time PCR detection of Pseudomonas aeruginosa in clinical and

municipal wastewater and genotyping of the ciprofloxacin-resistant isolates', FEMS

Microbiology Ecology, vol 57, no 1, pp. 158-167.

Schweizer, H. P. 2001 'Triclosan: a widely used biocide and its link to antibiotics', FEMS

Microbiology Letters, vol 202, no 1, pp. 1-7.

Singer, H., Muller, S., Tixier, C. and Pillonel, L. 2002 'Triclosan: occurrence and fate of a

widely used biocide in the aquatic environment: Field measurements in wastewater

treatment plants, surface waters, and lake sediments', Environmental Science and

Technology, vol 36, no 23, pp. 4998-5004.

Sinton, G. L., Fan, L. T., Erickson, L. E. and Lee, S. M. 1986 'Biodegradation of 2,4-D and

related xenobiotic compounds', Enzyme and Microbial Technology, vol 8, no 7, pp. 395-

403.

Slayden, R. A., Lee, R. E. and Barry, C. E. 2000 'Isoniazid affects multiple components of

the type II fatty acid synthase system of Mycobacterium tuberculosis', Molecular

Microbiology, vol 38, no 3, pp. 514-525.

Son, H. S., Choi, S. B., Zoh, K. D. and Khan, E. 2007 'Effects of ultraviolet intensity and

wavelength on the photolysis of triclosan', Water Science and Technology, vol 55, no 1-

2, pp. 209-216.

Stephen, K. W., Saxton, C. A., Jones, C. L., Ritchie, J. A. and Morrison, T. 1990 'Control of

gingivitis and calculus by a dentifrice containing a zinc salt and triclosan', Journal of

Periodontology, vol 61, no 11, pp. 674-679.

Stewart, M. J., Parikh, S., Xiao, G., Tonge, P. J. and Kisker, C. 1999 'Structural basis and

mechanism of enoyl reductase inhibition by triclosan', Journal of Molecular Biology,

vol 290, no 4, pp. 859-865.

Suller, M. T. and Russell, A. D. 2000 'Triclosan and antibiotic resistance in Staphylococcus

aureus', Journal of Antimicrobial Chemotherapy, vol 46, no 1, pp. 11-18.

Tabak, M., Scher, K., Hartog, E., Romling, U., Matthews, K. R., Chikindas, M. L. and

Yaron, S. 2007 'Effect of triclosan on Salmonella typhimurium at different growth

stages and in biofilms', FEMS Microbiology Letters, vol 267, no 2, pp. 200-206.

Nova Science Publishers, Inc.

The Fate and Persistence of the Antimicrobial Compound Triclosan … 237

Tatarazako, N., Ishibashi, H., Teshima, K., Kishi, K. and Arizono, K. 2004 'Effects of

triclosan on various aquatic organisms', Environmental Science, vol 11, no 2, pp. 133-

140.

Thompson, A., Griffin, P., Stuetz, R. and Cartmell, E. 2005 'The fate and removal of

triclosan during wastewater treatment', Water Environment Research, vol 77, no 1, pp.

63-67.

Tong, A. Y., Peake, B. and Braund, R. 2011 'Disposal practices for unused medications

around the world.' Environment International, vol 37, no 1, pp. 292-298.

Uyaguari, M. I., Fichot, E. B., Scott, G. I. and Norman, R. S. 2011 'Characterization and

quantitation of a novel β-lactamase gene found in a wastewater treatment facility and

the surrounding coastal ecosystem', Applied and Environmental Microbiology, vol 77,

no 23, pp. 8226-8233.

van der Waarde, J. J., Kok, R. and Janssen, D. B. 1994 'Cometabolic degradation of

chloroallyl alcohols in batch and continuous cultures', Applied Microbiology and

Biotechnology, vol 42, no 1, pp. 158-166.

van Hylckama Vlieg, J. E. T. 2000 'Detoxification of reactive intermediates during

microbial metabolism of halogenated compounds', Current opinion in microbiology, vol

3, no 3, pp. 257-262.

Vischer, W. A. and Regos, J. 1974 'Antimicrobial spectrum of triclosan, a broad spectrum

antimicrobial agent for topical application', Antibacterielles Wirkungsspektrum Von

Triclosan, Einem Breitband Antimicrobicum Zur Lokalen Anwendung, vol 226, no 3,

pp. 376-389.

Vörösmarty, C. J., McIntyre, P. B., Gessner, M. O., Dudgeon, D., Prusevich, A., Green, P.,

Glidden, S., Bunn, S. E., Sullivan, C. A., Reidy Liermann, C. and Davies, P. M. 2010

'Global threats to human water security and river biodiversity', Nature, vol 467, no

7315, pp. 555-556.

Wackett, L. P. 2003 'Pseudomonas putida - a versatile biocatalyst', National Biotechnology,

vol 21, no 2, pp. 136-138.

Walker, C., Borden, L. C., Zambon, J. J., Bonta, C. Y., DeVizio, W. and Volpe, A. R. 1994

'The effects of a 0.3% triclosan-containing dentifrice on the microbial composition of

supragingival plaque', Journal of Clinical Periodontology, vol 21, no 5, pp. 334-341.

Webster, J. 1992 'Handwashing in a neonatal intensive care nursery: product acceptability

and effectiveness of chlorhexidine gluconate 4% and triclosan 1%', Journal of Hospital

Infection, vol 21, no 2, pp. 137-141.

Webster, J., Faoagali, J. L. and Cartwright, D. 1994 'Elimination of methicillin-resistant

Staphylococcus aureus from a neonatal intensive care unit after hand washing with

triclosan', Journal of Paediatrics and Child Health, vol 30, no 1, pp. 59-64.

Wiesmann, U., Choi, I. S. and Dombrowski, E.-M. 2007 Fundamentals of Biological

Wastewater Treatment: Fundamentals, Microbiology, Industrial Process Integration,

Wiley-VCH Verlag GmbH.

Xia, K., Bhandari, A., Das, K. and Pillar, G. 2005 'Occurrence and fate of pharmaceuticals

and personal care products (PPCPs) in biosolids', Journal of Environmental Quality, vol

34, no 1, pp. 91-104.

Yazdankhah, S. P., Scheie, A. A., Høiby, E. A., Lunestad, B. T., Heir, E., Fotland, T.-Ø.,

Naterstad, K. and Kruse, H. 2006 'Triclosan and antimicrobial resistance in bacteria: an

overview', Microbial Drug Resistance, vol 12, no 2, pp. 83-90.

Nova Science Publishers, Inc.

Teresa Qiu, Christopher P. Saint and Mary D. Barton

238

Ying, G.-G., Yu, X.-Y. and Kookana, R. S. 2007 'Biological degradation of triclocarban and

triclosan in a soil under aerobic and anaerobic conditions and comparison with

environmental fate modelling', Environmental Pollution, vol 150, no 3, pp. 300-305.

Yu, J. C., Kwong, T. Y., Luo, Q. and Cai, Z. 2006 'Photocatalytic oxidation of triclosan',

Chemosphere, vol 65, no 3, pp. 390-399.

Yu, P. and Welander, T. 1995 'Growth of an aerobic bacterium with trichloroacetic acid as

the sole source of energy and carbon', Applied Microbiology and Biotechnology, vol 42,

no 5, pp. 769-774.

Zafar, A. B., Butler, R. C., Reese, D. J., Gaydos, L. A. and Mennonna, P. A. 1995 'Use of

0.3% triclosan (Bacti-Stat) to eradicate an outbreak of methicillin-resistant

Staphylococcus aureus in a neonatal nursery', American Journal of Infection Control,

vol 23, no 3, pp. 200-208.

Zhang, Y., Marrs, C. F., Simon, C. and Xi, C. 2009 'Wastewater treatment contributes to

selective increase of antibiotic resistance among Acinetobacter spp', Science of the

Total Environment, vol 407, no 12, pp. 3702-3706.

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 9

WATER QUALITY ASSESSMENT METHODS:

THE COMPARATIVE ANALYSIS

Tatyana I. Moiseenko1,

∗

, Alexandr G. Selukov2

and Dmitry N. Kyrov2

1Institute of Geochemistry and Analytical Chemistry, Russian

Academy of Sciences, Russia

2Tyumen State University, Russia

ABSTRACT

A comparative analysis of approaches and methods of biological assessment of water

quality is presented here. The method of ecotoxicological diagnostic of aquatic ecosystem

“health” and water quality evaluation based on the physiological state of fish is

substantiated as well in this chapter, based on our findings. Characteristics of the main

symptoms of diseases in fish inhabiting freshwater bodies of Russia and pathologic

disturbances in their organs and tissues, caused by water bodies' contamination with toxic

substances, are also presented. In this chapter, the method of ecotoxicological assessment

of water quality is shown to be both highly informative and easy-to-use in practical

monitoring. For example, the dose-effect dependencies and critical levels of water

pollution were determined for arctic lake Imandra on the basis of an ecotoxicological

approach.

1. INTRODUCTION

The problem of qualitative depletion of water resources caused by their contamination

has become especially acute for the last decades. The human factor affecting the formation of

the chemical composition of water is becoming as important as the natural geochemical and

biological processes. Transformation of catchment areas, transboundary ﬂows, discharge of

untreated industrial and domestic efﬂuents, as well as non-sewage efﬂuents lead to changes in

∗ Corresponding author.

Nova Science Publishers, Inc.

Tatyana I. Moiseenko, Alexandr G. Selukov and Dmitry N. Kyrov

240

the geochemical cycles of elements in the catchment area – water body system and the

occurrence of toxic substances in the aquatic environment, which entails water quality

deterioration.

The system of limitations, based on the concept of maximum permissible concentration

(MPC) of pollutants (or Guideline Concentration) in the water, cannot fully protect aquatic

ecosystems against degradation. Water quality criteria applied by water users differ

depending on the aim of water use: industrial use, drinking water supply, natural or

aquacultural fish reproduction.

From the viewpoint of the ecological paradigm, water is a vital resource for all live

organisms. At the same time, water is a habitat for aquatic organisms. In the process of their

vital activity, living water organisms, using water as both a resource and habitat, actively

affect its properties. The relative natural stability of properties of water and their seasonal

cycles in individual water bodies is supported due to the dynamic balance of physical,

chemical, and biological processes — both in the water body itself and within its catchment

area.

The habitat is characterized by certain conditions. According to E. Odum (1981), a

condition is an ambient environmental factor, changing in time and in space, to which an

organism responds depending on its strength. The conditions, necessary for living and

reproduction of organisms inhabiting various water bodies, can differ largely. For instance, if

typically northern species adapted to low-saline oligotrophic water are placed in

“background” southern lakes, ﬁlled with water typical of such lakes it becomes clear that such

water properties as low salinity and oligotrophic character will not be acceptable for them

(even if thermal conditions in the southern lakes could be made similar to those in the

northern ones); and vice versa. Specific ecosystems have formed in water bodies with unique

properties of water (e.g., geothermal or brackish). For such communities, the properties under

consideration are optimum. Therefore, while determining ecologically permissible human

impact on aquatic ecosystems, it is necessary to take into consideration both the regional

conditions of water formation and the sensibility of organisms and ecosystems as a whole to

pollution.

Disengaging ourselves from subjective requirements of individual water users to water

quality, we can say that the water quality definition from the viewpoint of the ecological

paradigm appears to be more universal: “Water quality is the totality of properties of water

formed in the process of chemical, physical, and biological processes, occurring both in the

water body itself and within its catchment area. Water quality in a water body can be

regarded as good where it meets the requirements of preservation of the most vulnerable

organism’s health of species adapted to the existence in the given aquatic environment”.

At present, the necessity of developing unbiased ecological criteria of determining

aquatic ecosystems state and assessment of water quality causes no doubt. Such criteria

should be substantiated based on the response of individual organisms, their populations, and

communities to the effect of pollutants. If properties of the water under consideration meet

the requirements of the life and reproduction of the most sensitive aquatic organisms, then

water quality (with the exception of certain cases) can be regarded as meeting the respective

requirements for human health preservation, too.

This work is aimed at comparing biological methods of water quality assessment,

substantiating the technique of ecotoxicological studies for water quality evaluation and

regulation of human-induced contamination of water bodies.

Nova Science Publishers, Inc.

Water Quality Assessment Methods 241

1.1. Methods of Biological Evaluation of Water Quality

Biological methods of water quality assessment are based on studying different impacts

on ecosystems and their structural elements (individual organisms, populations, and

communities). The main methods of an estimation of the quality of water is presented on the

block-diagram (Figure 1).

Biotesting methods understood as the evaluation of the potential hazard of pollutants (or

concrete efﬂuents or contaminated waters) entering the water body, based on ex-situ

experimental laboratory studies. This method allows us to determine the lethal and sublethal

concentrations of potential pollutants, industrial efﬂuents, or contaminated waters for living

organisms (test objects) under the laboratory conditions (Bioassey…, 1985; Canadian water

quality.., 1994; Methodological recommendation, 1998). This approach was used to evaluate

maximum permission concentration (MPCﬁsh ) for toxic substances in fish water body

(Bespamyatnov, Krotov, 1985; List of Fishery-related…, 1999). This method is based on

experimental determination of pollutant concentrations, causing most signiﬁcant and easy-to-

detect disturbances in aquatic organisms, such as morbidity, survivorship rate, physiological

or pathological disturbances. The “threshold” value, causing visible abnormalities in the most

sensitive group of organisms, is adopted as the MPCﬁsh for the given hazardous substance.

Organisms belonging to different systematic groups (bacteria, algae, invertebrates, ﬁsh) are

used as test objects (Bioassey…, 1985; Methodological recommendation…, 1998). The

greater part of toxicological studies is carried out at the level of individual organisms. These

experiments allow us to study the effect of most frequently occurring toxic substances on

hydrobionts belonging to different systematic groups.

The main advantage of biontesting is the possibility to quickly obtain information on the

toxicity of individual contaminants or industrial wastewater. However, it is not clear how

rightful it will be to extrapolate the experimental results, obtained under the laboratory

conditions, to natural objects.

Figure 1. Вlock-diagram of the methods of the water quality assessment.

Nova Science Publishers, Inc.

Tatyana I. Moiseenko, Alexandr G. Selukov and Dmitry N. Kyrov

242

The behavior of pollutants in natural water bodies and their toxicological properties can

largely differ from those demonstrated in aquaria; combined effects (both synergetic and

antagonistic) can manifest themselves. Under the laboratory conditions, it is difﬁcult to

determine the transfer of toxins along food chains and their cumulative effects. In addition,

individual organisms, used in the experiment, have very little in common with the respective

natural populations and communities. Therefore, the MPC system, based on biotesting studies

of individual elements and their compounds, does not present a scientiﬁc basis for the

ecological standardization of pollutants entering natural water bodies.

Bioindication is aimed at evaluating water quality in natural water bodies or in the zone

of contamination according to the state of indicator species or communities (in-situ). This

method is widely used in monitoring of aquatic environment. The hydrobiological monitoring

of freshwater ecosystems involves observations over the state of the major subsystems:

microﬂora, periphyton, phytoplankton, macrophytes, zooplankton, and zoobenthos (Manual

for hydrobiological…, 1992; Wang, Dixon, 1995; Environment quality.., 2001).

Detailed description of the techniques and advantages of using each group of organisms

in the bioindication is presented elsewhere (Manual for hydrobiological…, 1992). The expert

characteristic of the ecological state of a water body is based on the totality of all the features,

including structural ones (species composition, quantity, biological diversity, the ratio of

species of different ecological valency, and the characteristics of their saprobility) and

functional characteristics of aquatic communities (production and destruction parameters,

etc.).

Change in the structural and functional organization of communities is the most

signiﬁcant parameter describing variations in freshwater ecosystems under the impact of

human factors. However, quantitative methods of ecosystem state evaluation were not paid

due attention and did not become widely used in hydrobiological analyses. The indices and

parameters deduced basing on the account of the species composition of biocenosis are very

often subjective and depend on the biotope homogeneity and the season of the year. In

addition, populations of different species differ, as concerns the degree of their

polyfunctionality. The use of such indices becomes hampered where water bodies are

eutrophic and contaminated with organic substances: the number of some communities begins

to increase, while that of some others decreases. That is why, as a result, comparative

estimates, expressed as grades, rating points, marks, and indices are obtained. These estimates

occupy an intermediary position between the qualitative and quantitative parameters and

depend on the qualiﬁcation of experts.

Ecotoxicological diagnostic is aimed at obtaining an integrated assessment of water

quality, based on symptoms of disturbance in the ecosystem “health” (in situ). Such

diagnostic is based on theoretical principles of toxicology and ecology, synthesis of their

methods, allowing the determination of environmental effects and mechanisms of the impact

of hazardous substances on aquatic organisms. Aquatic ecosystems are stressed in all levels,

ranging from individual and up to the population and community levels. For ecosystem health

assessment the following four definitions have been used: i) cellular health, which describes

the structural integrity of cellular organels and the maintenance of biochemical processes; ii)

individual health, which presents structural and morphological health and functioning in

terms of physiology of the entire organism; iii) population health, which measures the

sustainability and maintenance of a population of a particular species; iv) community health,

which describes a group of organisms and the relationships between species in that group.

Nova Science Publishers, Inc.

Water Quality Assessment Methods 243

Each method has its limitations and advantages, and the type of method used defines how we

interpret the effect of a stressor on ecosystem health (Adams, Ryon, 1994; Cash, 1995; Arttril,

Depledge, 1997; Moiseenko et al., 2006, 2008; Yeom, Adams, 2007).

In general, indicators at the biochemical and physiological levels provide information on

the functional status of individual organisms, while intermediate-level responses, such as

histopathological condition, are indicative of the structural integrity of tissues and organs.

Symptoms of physiological changes and pathologic state of organisms, functional and

structural disturbances in the state of populations and communities in natural water bodies

reﬂect poor “health” of the ecosystem and unsatisfactory water quality. Community and

population level measurements integrate the responses to a variety of environmental

conditions, and therefore may be less reflective of toxic contaminant-induced stress in

comparison to the level of organisms (Yeom, Adams, 2007). Unlike traditional methods of

bioindication, ecotoxicological approach is aimed at revealing the dose-effect dependences

between water contamination parameters and disturbances in biological ecosystems, serving

as the basis for the determination of ecologically permissible concentrations of toxic

substances in the water.

1.2. Fish Health as Criteria of Water Quality

Many groups of organisms can be used as indicators of environmental and ecological

change. But numerous publications attest that fish (in situ) is a good indicator of

environmental change and ecosystem health, especially in case of toxic water pollution

(Whitfield, Elliott, 2002; Eliott et al., 2003; Moiseenko, 2009). Fish occupy the top level in

the trophic system of aquatic ecosystems. Pathological changes in fish organ enable us to

determine the toxicity of water and the potential danger of man-entering substances in water.

Fish, in comparison with invertebrate, are more sensitive to many toxicants and are the

convenient test-object for indication of ecosystem health. Physiological and biochemical

parameters of the state of ﬁsh in water bodies, allowing us to reveal both short-term and long-

term effects of the sublethal dose of pollutants, are also determined.

In the 1970s, methods of pathologic–physiologic research of ﬁsh were widely used in

connection with more and more frequent cases of large-scale ﬁsh poisoning caused by

natural-water contamination. Methods of clinical and postmortem examination of organisms

applied in veterinary and medicine were used to study ﬁsh organisms in order to assess the

consequences of water pollution with toxic substances.

Diagnostics of fish health requires system studies, which combine extensive data with

disease diagnoses. A diagnostic system for fish state, suitable for practical monitoring, is

presented elsewhere (Moiseenko et al., 2006; Moiseenko et al., 2008). The macro-level

examination of individuals involves diagnosing of diseases on the basis of a visual

examination of numerous organisms. Method of clinic and postmortem examination of

organisms is used for large-scale examination of ﬁsh, carried out not later than within one

hour after the ﬁsh has been caught. In the process of visual examination, special attention

should be paid to the following: the intensity of color (the state of pigment cells–

melanophores); the integrity of the ﬁn edge and somactids; the total amount of mucus on the

ﬁsh body; the state of squama, opercula, oral cavity, anus; the cases of hyperemia,

subcutaneous hemorrhages, sores, or hydremia of the body; deformation of skull and skeleton

Nova Science Publishers, Inc.

Tatyana I. Moiseenko, Alexandr G. Selukov and Dmitry N. Kyrov

244

bones; the state of eye crystalline lens and cornea. When the opercula are opened, gills are

examined, in particular, their color, the presence and the amount of mucus, the state of gills

petals (accretion, adhesion, dilatation, or thinning down). After the abdominal cavity is

dissected, the state of ﬁsh muscles is studied (color, consistence, hemorrhages, attachment to

bones), as well as the presence of exudate in the abdominal cavity, the amount of cavitary fat,

its color and density. The topographic location of viscera (liver, kidneys, gonads, spleen,

heart, stomach, intestines), their dimensions, color, density, edges, hemorrhages, zones of

necrosis, etc. are studied. Mucous membranes of dissected stomach and intestines are

examined, in addition to cerebrum, paying special attention to ﬁlling of vessels, their color

and density. At this level, a preliminary diagnosis is made based on clinical and postmortem

symptoms of intoxication.

The micro-level diagnostics include hematologic, histological, biochemical, instrumental,

and other methods. These are labor-intensive and cannot be widely used, however they are

essential to refine the diagnosis and estimate the consequences of pathologic changes in fish

organisms.

The degree of disturbance in individual organisms is very important to diagnose the

damage to fish organisms in the contaminated zone. For instance, up to 70% of individuals in

the zones of contamination may be in a state close to the lethal threshold. If the concentration

of toxic substances is not high, the percentage of affected individuals can be the same, but

disturbances in fish organisms can be insignificant and will not be life threatening. In order to

estimate the state of fish organism using the data of clinical and postmortem examination,

experts suggest using different rating systems. In the process of macro-diagnostics, three

stages of the disease can be singled out (0 denotes healthy individuals):

(1) Low disturbances, not threatening the life of the fish;

(2) Medium-level disturbances, causing a critical state of the organism;

(3) Distinct intoxication symptoms leading to inevitable death of the organism.

The overall index of morbidity in fish in the given zone of contamination can be

presented as:

Z = (N1 + 2N2 + 3N3) /Ntot.

Here N1, N2, and N3 are the number of fishes in the first, second, and third stages of the

disease, respectively; Ntot is the total number of the examined fishes, including healthy

individuals. Note that O ≤ Z ≤ 3. If all the fish in the given water body does not demonstrate

any intoxication symptoms, then Z = 0. The value of Z will increase with an increase in both

the number of sick fishes and the extent of their diseases.

The table 1 presents typical health pathologies detected in ﬁsh inhabiting different

polluted aquatic environment of lakes and rivers of Russia. These disturbances were

diagnosed using the method of clinic and postmortem examination.

When toxic substances are discharged from non-point sources, zones of contamination,

having different degree of hazard, are formed in the water bodies. Fish stock in the

contaminated zones is non-homogeneous, as concerns morbidity, which is related to the

concentration of pollutants in the zone under consideration and the duration of their effect, on

the one hand, and the tolerance of individual ﬁsh organisms, on the other hand.

Nova Science Publishers, Inc.

Water Quality Assessment Methods 245

Table 1. Pathology disturbance in organ and tissues f fish inhabiting in polluted aquatic

environment of lakes and rivers of Russian. This table is based on an analytical review

of published studies (Chinareva, 1988; Savaitova et al., 1995; Lukin, Sharova, 2004;

Moiseenko, Kudryavtseva, 2002; Moiseenko et al., 2006; Moiseenko et al, 2008,

Moiseenko 2009)

Organ and

tissues

Visual symptoms of

intoxication

Changes in the cell

structure and diagnosis

Fish species, source Disturbing factors

Eyes Opacity,

hemorrhages in the

cornea

Cataract or chemical scald Bream in the Volga and

Kama reservoirs;

whitefish in lakes in the

North Kola Peninsula;

loach in water bodies of

the Noril’sk

Pyasinskaya system

Contamination with a

set of toxic

substances;

wastewater of the

copper and nickel

metallurgical plant,

with a high

concentration of

metals

Tectorial

tissues

Change in the

natural color of the

body; specific tints

of skin (yellowish,

greenish); bristling

of squama

– Bream in the Volga and

Kama reservoirs;

whitefish in lakes in the

North Kola Peninsula

Contamination with a

set of toxic

substances; effluents

of ore-mining and

ore-concen-trating

plants

Heart Increased

dimensions of the

heart, flabby

structure of the

heart; symptoms of

obesity (adipose

cover)

Decomposition of fibers

and muscles of the heart,

dystrophy of muscular

fibers, lipoid dystrophy of

myocardium

Bream in the Volga and

Kama reservoirs;

whitefish in lakes in the

North Kola Peninsula;

whitefish in the Pechora

River; bream and

whitefish in the Ladoga

Lake

Contamination with a

set of toxic

substances; effluents

of mining; metals,

organic xenobiotic

compounds, oil

products; wastewaters

of a pulp-and-paper

plant

Gills Anemic ring around

the gill’s arc;

cyanotic tint of

gills, filled with

blood;

conglomerated

petals

Hyperplasia of epithelium

of petals and stamens;

exfoliation of epithelium;

overfilling and stagnation in

blood vessels; degeneration

of cartilage tissue

Bream in the Volga and

Kama reservoirs;

whitefish in lakes in the

North Kola Peninsula

and the Pechora River;

whitefish and bream in

the Ladoga Lake and

the Volkhov River

Contamination with a

set of toxic

substances; metals,

organic xenobiotic

compounds, oil

products; wastewaters

from a pulp-and-

paper plant

Skull bones

and spine

bones

Insufficient

ossification of

cranium; curvature

of the spine;

accretion of

vertebrae

Osteoporosis, scoliosis Whitefish in Imandra

Lake; whitefish in the

Neva and Volkhov

rivers in their lower

sections; loach in water

bodies of the Noril’sk

Pyasninskaya system

Effluents of the ore-

concen-trating plant,

with a high

concentration of Sr;

wastewaters of a

pulp-and-paper plant

Muscles Flabby tissues,

crumbled into

myocepts

Myopathy Whitefish in the

Imandra Lake; sturgeon

in the Caspian Sea

Contamination with a

set of toxic

substances: metals;

POP’s compounds,

oil.

Nova Science Publishers, Inc.

Tatyana I. Moiseenko, Alexandr G. Selukov and Dmitry N. Kyrov

246

Table 1. (Continued)

Organ and

tissues

Visual symptoms of

intoxication

Changes in the cell

structure and diagnosis

Fish species, source Disturbing factors

Gonads Constriction and

twisting of gonads;

insufficient

saturation of egg-

bearing plates with

fish eggs, till their

consistence

ecomes friable and

transparent;

hermaphroditism

Asynchronous development

of fish eggs; lipoid and

connective tissue

degeneration of ovaries and

testicles; resorption and

degeneration of oocytes;

lysis of their shells; laying

of oocytes without nuclei;

formation of cysts

Whitefish in the

Imandra Lake;

whitefish in the Pechora

River; bream and

whitefish in the Ladoga

Lake; sturgeon in the

Caspian Sea

The same

Liver Appearance of

bright spots or

change in the color

from brown

(normal) to light

yellow or sandy;

increase in the

dimension;

intumescence or

decrease in the

dimensions and

atrophy, change in

the configuration

Cirrhosis (zones of necrosis

inside the parenchyma,

excessive development of

the connective tissue,

decomposition of

hepatocyte nuclei,

excessive proliferation of

connective tissue,

hemorrhages); lipoid

degeneration (vacuolar

lipoid dystrophy of the

parenchyma; neoplasms

Whitefish in the

Imandra Lake;

whitefish in the Pechora

River; loach in water

bodies of the Noril’sk–

Pyasinskaya system;

bream and whitefish in

Ladoga and Karelian

lakes; bream in the

Volga River; sturgeon

in the Caspian Sea

The same

Kidneys Increase in the

dimension,

considerable

swelling of urine-

excreting canaliculi

in the distal section,

with calculi of the

diameter reaching 5

mm precipitated

inside. The kidneys

are of a bigger size

as compared to the

norm, have flabby

structure, the

uniform color

changes to visible

granulated structure

Nephrocalcinosis

(disturbances in the

structure of urine-excreting

canaliculi, hyperplasia or

disquamation of epithelium,

with extraneous occlusions

inside canaliculi);

nephropathy (zones of

necrosis); growth of

connective tissue inside the

parenchyma; neoplasms

Whitefish in Imandra

Lake; whitefish in the

Pechora River; loach in

water bodies of the

Noril’sk–Pya-sinskaya

system; Ladoga and

Karelian lakes;

sturgeon in the Caspian

Sea

The same

In case of acute intoxication, ﬁshes quickly die; if ﬁshes are exposed to a permanent

impact of toxic substances, physiological disturbances manifest themselves in degenerative

processes (sometimes, irreversible) occurring in ﬁsh organs and tissues. In the long run,

uninterrupted effect of toxic substances results in delayed death of ﬁshes.

Method of hematologic analysis is used to diagnose toxicoses in ﬁsh at early stages,

because the circulatory system of ﬁsh is a sensitive indicator of the impact of unfavorable

ambient conditions on organisms and responds to toxic agents by the appearance of

disintegrating cells and their pathologic forms (Ziteneva, 1989). Changes in blood parameters

Nova Science Publishers, Inc.

Water Quality Assessment Methods 247

manifest themselves earlier as compared to visual symptoms of pathologic diseases. In the

blood samples thus taken, hemoglobin concentration, erythrocyte sedimentation rate (ESR),

erythrocyte and leukocyte concentration. Blood smear examination allows the analysis of red

blood composition, differential blood count, and the detection of occurrence of pathologic

blood cells addition, other characteristics known in hematology can be used.

The exact interpretation of the results obtained needs taking into account the regularities

of the development of toxicosis in ﬁsh, in which four stages can be arbitrarily singled out

(Moiseenko, 1998):

i) mobilization,

ii) destabilization,

iii) degradation.

i) Investigation of the dynamics of hemathologic parameters (case study of whiteﬁsh

inhabiting contaminated northern water bodies) have shown that at the ﬁrst stage of toxicosis,

young and reserve cells ﬁnd their way to the blood channel, thus making the blood thicker.

The amplitude of changes in the hemoglobin concentration in the blood grows: whereas the

normal hemoglobin concentration in the ﬁsh blood varies from 80 to 130 g/l, certain

individuals can be found, whose hemoglobin concentration reaches 180 g/l. Polychromasia

and anacytosis can be diagnosed in ﬁsh, basing on blood smear analyses; abnormal cell

division appears; the concentration of leukocytes grows, and the leukocyte formula begins to

change, such changes manifesting themselves in an increase in the share of young

lymphocytes, neutrophiles, and monocytes. All this testiﬁes to the fact that, from the ﬁrst

stages of toxic substances impact, the system of hemogenesis in ﬁsh responds as a protective

function by means of mobilizing young and reserve sells.

iii) At the stage of destabilization, the process of blood cells destruction is partly

compensated for by reserve young cells ﬁnding their way in the blood channel, that is why

hemoglobin concentration in the blood can be close to the normal value, but numerous

pathologic and destroyed cell forms (schistocytes) begin to appear in the blood. Pycnosis of

cell nuclei begins to manifest itself; anacytosis, polychromasia, and some other blood diseases

become more pronounced. This stage can conventionally be adopted as a critical threshold of

irreversible changes. For instance, for whiteﬁsh inhabiting the northern Kola Peninsula water

bodies, this “threshold” state occurs, when hemoglobin concentration in the blood decreases

to the value less than 80 g/l, and destructive and pathologic cells appear. After that,

irreversible changes begin to occur in the system of hemogenesis.

iii) At the stage of degradation, when numerous cells are already destroyed, anemia,

accompanied by destructive changes in other physiological systems of the organism (kidneys,

liver, gills), begins to manifest itself. The above changes lead to the death of the organism.

The scheme of toxicosis development described above cannot always be observed in ﬁsh

inhabiting natural water bodies, because the adaptation threshold for different individuals and

their resistance to the effect of sublethal doses of pollutants can vary.

Regularities of intoxication process development allow the conclusion that the use of

such parameter as the average concentration of hemoglobin in the blood (without

investigation of blood smears) for a group of individuals may fail to reﬂect the initial stages

of intoxication. Variations in the feature and the number of ﬁshes with hemoglobin

concentration lower than the critical threshold are more informative.

Nova Science Publishers, Inc.

Tatyana I. Moiseenko, Alexandr G. Selukov and Dmitry N. Kyrov

248

Method of histological analysis is important for revealing disturbances in the

morphologic and functional structure of organs and tissues, degenerative processes, and

making the diagnoses of ﬁsh diseases according to the disturbances revealed. Pieces of organs

and tissues with visible changes are sampled and ﬁxed using traditional methods adopted in

histology. In order to establish the normal physiological state, it is necessary to take organ

and tissue samples from healthy individuals as well. The table 1 presents histopathological

changes in tissues and organs, typical for ﬁsh diseases. Destructive changes in the cell

structure of organs and tissues that develop most frequently in ﬁsh inhabiting natural water

bodies contaminated by toxicants are presented below (Fish pathology…, 1994; Chinareva,

1988; Lukin, Sharova, 2004; Savaitova et al., 1995; Moiseenko, Kudryavtseva, 2002;

Moiseenko et al., 2006; Moiseenko et al, 2008).

In liver, its cells change and its tissue becomes destroyed. Changes in liver histologic

structure testify to the development of cirrhosis or lipoid degeneration of liver. In kidneys,

pollutants cause hyperemia, dystrophic changes in the epithelium of ductules and capsules,

frequently complicated by necrobiosis. A similar histologic pattern of disease is diagnosed as

interstitial nephritis (ﬁbroelastosis). Considerable pathologic disturbances can be observed in

the structure of epithelium of gyrose kidney ductules: in some cases, epithelium can ﬂatten,

while in other cases it may increase even to become prismatic (the normal conﬁguration of

epithelium is cubic). In case of nephrocalcinosis, extraneous occlusions can be detected inside

urine-excreting canaliculi. For gills, swelling of respiratory epithelium, hyperemia of petals

and small petals, hemorrhages, desquamation, and necrosis of epithelium cells are typical. As

for the heart, its expansion and overﬁlling with blood is observed, in addition to small

hemorrhages in myocardium, dystrophic changes in muscular ﬁbers and loss of transverse

stripes, protein dystrophy and small-cell fatty degeneration of heart are also observed.

Pathologic changes in ﬁsh skeleton (osteoporosis or scoliosis) are most frequently related

to the disturbance in mineral substances metabolism, and insufﬁcient or low anabolism of

calcium. In muscular tissue, myopathy often manifests itself in delamination of tissue.

Disturbances in the reproduction system manifest themselves in lipoid and connective-

tissue degeneration of ovaries and testicles, resorption (observed in breams inhabiting the

Volga River), degeneration of oocytes, the formation of cysts, and the occurrence of

hermaphrodite individuals. Neoplasms, such as skin sarcoma in pike perch, as well as zones

of tumor tissues in liver and kidneys, were also revealed.

Biochemical methods are very important when it is relevant to study the mechanisms

responsible for the development of certain abnormalities in the organism. It is shown that,

under the conditions of low concentration of pollutants, ﬁshes inhabiting natural water bodies

demonstrate activation of physiological systems responsible for detoxiﬁcation. Intoxication is

characterized by an increased concentration of enzymes, dispersion of fatty acids, cholic

acids, triaglycerins (triacetins), lysosomatic ferments, acid phosphotasa forms, isoferments of

lactadehydrogenasa, free aminoacids as a “quick response” group (Bitton and Dutku, 1985

Sidorov, and Yurovitskii, 1991). The general regularity of metabolism re-organization in ﬁsh

under the impact of toxic substances corresponds to hemogenesis response. At the stage of

contact, the primary response of the endocrine system is characterized by a stimulation of

andrenoenergetic systems (leading to an increase in the concentration of catechilamines,

adrenalines, and noradrenalines) and hypothalamus-hypophysis-epinephros centers (entailing

an increase in the concentration of ATP and corticosterole), which testiﬁes to an increase in

the energy exchange rate as a response to toxic substance penetration in the organism.

Nova Science Publishers, Inc.

Water Quality Assessment Methods 249

Figure 2. Scheme. of fish response on toxic stress

The secondary response is related to metabolic adaptation, ensuring steady elevated

energy consumption for detoxiﬁcation: increased concentration of catechilamines (the supply

of energy to the organism by means of glycogenolysis) and cortisol (inhibiting protein

synthesis and energy supply to the organism by means of catabolism of glycogen, lipids and

proteins (Niimi, 1990; Sidorov, Yurovitsky, 1991; Werd, Komen, 1998; Nemova, Visotskaya

2004).

When the biochemical functions of detoxiﬁcation are exhausted (stage of destabilization),

these functions become de-regulated, which leads to the destruction and death of the

organism. A generalized mechanism of the development of toxicosis in ﬁsh is presented in

scheme (figure 2).

It is evident that methods of clinic and postmortem examination of ﬁsh inhabiting

contaminated water bodies are most convenient. They are readily-available and easy-to-apply

and can be widely used for express-diagnostic of the ecosystem’s “health,” because

disturbances in physiologic systems of ﬁsh testify to unsatisfactory water quality, and these

manifestations of poor water quality appear earlier than changes in the structural and

functional organization of the community.

Criteria of water quality assessment are substantiated based on the general regularity of

toxicosis development and parameters of ﬁsh stock heterogeneity in respect of morbidity.

Speciﬁc diseases (e.g., scoliosis or nephrocalcinosis) can give an idea of their etiology - water

contamination with heavy metals. While using hemathologic parameters, it is not the average

concentration of hemoglobin in ﬁsh blood, but the number of ﬁshes in the concrete habitat

with hemoglobin concentration lower than the critical threshold value, delineating the third

and the fourth stages of toxicosis (for whiteﬁsh, it is the number of ﬁshes whose blood

hemoglobin concentration is lower than 80%), in addition to the occurrence pathological

forms, that are the most informative criteria.

Nova Science Publishers, Inc.

Tatyana I. Moiseenko, Alexandr G. Selukov and Dmitry N. Kyrov

250

1.3. Dose-Effect Dependencies and Critical Levels of Water Pollution: Case

Study of Arctic Lake Imandra Served As Example

Among Arctic regions, the Russian Kola North (Murmansk region) is the most densely

populated and industrially developed. Lake Imandra being one of the most contaminated

lakes in Murmansk region served as an example. Development of copper-nickel industrial in

the catchment of Imandra began in the 1940s. Large amounts of pollutants entered the lakes

between 1940 and 1990. Transformation of economic conditions in Russia at the beginning of

1990s brought to a standstill of many industrial activities and, accordingly, slowed down

surface water pollution. Some revival of the economy in last decade is spurred by technology

modernization and more restrictions on pollution of the lake and the atmosphere. As a result,

pollution levels entering the lake have decreased over the past 20 years, but nowadays it still

remains high (Moiseenko et al., 2006).

The main pollutants were heavy metals (predominantly, nickel and copper), sulphates,

chlorides and nutrients. The catchment areas were also polluted by airborne contaminants,

metals and acid deposition. The highest contamination level of lakes in that region was

observed in 1970-1980’s when the copper and nickel production was the highest. The metal

content in lake Imandra reflects the historical dynamics of water contamination. The maximum

nickel contamination reached 290 μg/l (table 2).

Table 2. The metal concentration in the water of Lake Imandra in various years of

researches (numerator – average values, denominator – minimum and maximum

values)

Metal, μg/l Years of investigation

1981 1986 1991 1996 2003 2007

Ni 42

16-84

13-290

13-49

13-29

7-27

4-16

Cu 3

0-5

1-130

4-38

5-18

5.5

4-10

4.4

3-9.9

Cd n/d n/d n/d

0.27

0.12-0.60 <0.05 0.10

0.05-0.29

Pb n/d n/d n/d n/d <0.3 <0.3

<0.3-0.5

Table 3. The characteristic of whitefish diseases and metals accumulations in condition

of metal water pollution of Lake Imandra in various years of researches

Years of investigation 1981

1986

1991

1996

2003

2007

The main symptoms of fish diseases, % from number of the surveyed individuals

Nephrocalcitoses 52 47 45 14 - -

Fibroelastos 48 53 55 48 39 16

Lipoid degeneration of a liver and a cirrhosis 100 89 78 48 39 16

Anomalies of a structure gonads 34 27 8 - - 3

Number investigated fish n = 788 n=721 n=453 n=462 n=235 n=196

Nova Science Publishers, Inc.

Water Quality Assessment Methods 251

Unfortunately, there is no historical data on the content of cadmium and lead in water for

that time period. The reasonable assumption is that the content of all concomitant metals was

much higher.

Starting from 1980 there are studies aimed at the investigation of fish morbidity. As

bioindicator white fish was used (Coregonus lavaretus L.). Table 3 summarizes the data

about and fish morbidity in Imandra lakes and content of metals in fish in last 26 years.

The main symptoms of fish intoxication in the Kola lakes polluted by metals are as

follows: change of the integument colour (depigmentation), tousling of scales, oedema gills

and appearance of anemia rim, destructive changes of liver (increase of size, change colour

and friability) and kidneys (colour, granulation, thickening of renal and presence of nephritic

calculi), anomalies in gonad texture, etc. In the areas polluted by Kola copper-nickel smelters

the fish endemical pathology most frequently met is nephrocalcitosis and fibroelastosis of

kidney.

The kidney having nephrocalcitosis signs show clear changes in kidney

cytomorphological structure. There have been visible fibrous granulomas formations and

necrotic parts in the parenchyma. Connective tissue growing was found around the Boumen

capsules, blood vessels and excretory duct. Significant pathologies were observed in

epithelium texture of proximal tubule. There has also been found epithelial hyperplasia (up to

the prismatic epithelial cells, in norm - cubic); at other part - epithelial hypoplasia (up to

desquamation); inside it - inclusions of stones (Figure 3).

Figure 3. Liver histopathologies: 1 – drops of fat in the hepatocyte; 2 – lipoid degeneration of the liver;

3 – liver parenchyma hemorrhage; 4 – widening of inter-hepatocyte spaces.

Nova Science Publishers, Inc.

Tatyana I. Moiseenko, Alexandr G. Selukov and Dmitry N. Kyrov

252

Fish kidneys are capable of accumulating high levels of nickel. Copper being one of the

essential elements does not exhibit the same trend; moreover, the copper content in liver

during the period of highest contamination was low. The work of Moiseenko, Kudryаvtseva

(2002) reflects this phenomenon, which is explained by the degenerative alteration of the

functional tissue of this organ.

Hystological analysis of hepatic diseases shows lipoid degeneration of its cells, in some

places - signs of necrosis and degenerative damage caused by destruction of hepatocytes and

overgrow of connecting tissue (cirrhosis). Different abnormalities of liver (from initial to

debilitative) are also frequently detected in fish living in water polluted by metals (Figure 4).

Gill histopathology is presented on figure 5. In nature, aquatic organisms are exposed to a

combined dose of all contaminates.

It is important to find a general numeric parameter describing the total impact of

contaminates on biota. In case of Imandra lake the main pollution is heavy metals. An

integrated impact dose of metals is determined by the number of metals, their concentration

and toxic properties for each of them.

The values of Guideline Concentrations (GC) or Maximum Permeation Concentrations

(MPC) largely differ by country, in spite of the fact that experimental research techniques to

establish the MPCs are universal. For example, in Russia, the MPC values for Cu, V, Mn and

some other elements are unreasonably underestimated, whereas the MPCs for Cd, As, Pb, and

Al are overestimated.

Figure 4. Kidney histopatologies: 1 – focal renal parenchyma necrosis; 2 – conjunctive tissue

overgrowth; 3 – necrosis of glomerulus renalis vascular loops; 4 – renal channel necrosis.

Nova Science Publishers, Inc.

Water Quality Assessment Methods 253

Although accepted in Russia, as well as in other countries, water quality standards for

metals in water do not take into account the integrated impact dose. Using data about

toxicological properties of each metal based on GC, we can define the integrated impact dose

by summing the excess of real concentration for each of metals to their GC or known

threshold of impact as follows:

Itox = ∑(Ci /GCi-fish.).

Itox – is the integrated toxicity index, Ci – are concentrations registered in water; GCi-fish –

are GC for metals accepted in Russia for aquatic biota, which are more stringent than those

for drinking water. Water quality may be considered good if Itox-1 is no more than one (0 <Itox

≤1).

Despite our critique of the GC, since Imandra is in Russia, we used the following, legally

binding in Russia, concentrations, (µg/l): Cd = 5, Ni = 10, Cu = 1, Pb = 10, Zn = 10, As = 50;

Cd = 5 (List of Fishery-Related Standards…, 1999).

The rate of fish disease (measured in terms of all the three indicators considered above:

percent of fish with physiological deviances from norm, Z - the morbidity index of health,

and the percent of fish with anemia - decrease of hemoglobin in blood below 80 g/l) came out

as a very sensitive measure of toxic pollution (Figure 6).

Analysis of these dose-effect dependencies shows that when the integrated toxicity index

of water is more then 1 pathologies and dysfunctions are likely to appear in fish organisms.

Figure 5. Gill histopathology: 1 – gill epithelium hyperplasia; 2 – gill petal adhesion; 3 – gill petal

epithelium exfoliation; 4 – swelling of apical ends of gill petals.

Nova Science Publishers, Inc.

Tatyana I. Moiseenko, Alexandr G. Selukov and Dmitry N. Kyrov

254

Figure 6. Measuring ecosystem health: dependence between an integrated parameter of toxic loading

(Itox = ∑(Ci /GCi) and indicators of fish diseases (% of fish with deviances, % of fish with anemia

(decrease of haemoglobin in blood lower than critical 80% level).

Our results showed that for ecosystem health assessment the fish disease index is

sensitive, and also demonstrated that the standards for water quality assessment adopted in

Russia are inadequate for Arctic regions. The dose-effect dependencies (in case study of

arctic lake Imandra) clearly show that heavy metal pollution must be significantly decreased.

At the same time, for healthy ecosystem revitalization water pollution (integral parameter Itox-

fishl) must be decreased at least 5 times, first of all for nickel, which determines many water

properties.

CONCLUSION

The necessity of determining ecological standards for water quality requires an

improvement of the existing methods of its biological assessment and the development of

new ones, which would be effective and readily-available for large-scale application. At

present, two methods are used. The ﬁrst one is the bioassay technique, implying water quality

evaluation on the basis of the response of organisms to the impact of toxic substances during

the experiment (ex-situ). The second method is bioindication, which is based on studying the

state of the ecosystem’s structural elements (individu-als, populations, communities in situ).

Both these methods allow a qualitative assessment of water to be used for certain purposes.

At the same time, these methods do not allow the determination of critical water pollution and

standardization of toxic substances input in the water bodies.

At present, ecotoxicological approach to estimating the maximum permissible level of

water contamination, which allows a comprehensive characteristic of the ecosystem “health,”

gradually wins the recognition of researchers. Fish, regarded as the upper trophic level of the

aquatic ecosystem, is an indicator of toxic contamination of water bodies. Detection of

changes in ﬁsh physiologic parameters ensures forecasting the consequences of toxic

substances occurrence in the water for human beings as well. The state of ﬁsh can be

determined using all the available methods, such as clinic, postmortem, hemathologic, or

biochemical studies. As compared to parameters of structural and functional state of aquatic

Nova Science Publishers, Inc.

Water Quality Assessment Methods 255

communities, these methods allow us to characterize water quality in a shorter period and to

assess the effect of multicomponent long-term water contamination. Scientiﬁc substantiation

of criteria of estimating the biological effects of water pollution (the establishment of

ecological standards, the critical levels or hazard of water contamination) is the most

important element in the system of human impact management. Basing on dose–effect

dependences (between numerical indices of pathologic–physiologic states of ﬁsh and the

chemical parameters of water quality, in particular, the total concentration of toxic substances

in the water standardized to MPC), the critical water contamination can be determined.

In nature, aquatic organisms are exposed to a combined dose of all contaminates. It is

important to find a general numeric parameter describing the total impact of contaminates on

biota. In case of Imandra lake the main pollution is heavy metals. Parameters of fish

physiological conditions are even more directly related to water quality. During the period of

severe pollution there were mass fish diseases (nephrocalcinosis, lipid liver degeneration,

cirrhosis, anemia, scoliosis and others). Based on fish intoxication symptoms, we can

conclude that water quality is getting better in last period, but it is still far from recovery. In

case study of arctic lake Imandra used ecotoxicological approach clearly show that heavy

metal pollution must be significantly decreased.

ACKNOWLEDGMENTS

The work was supported by the Russian Foundation for Basic Research (Projects no 10-

05-00854) and grant of Russian Governments (№ 11G34.31.0036).

REFERENCES

Adams, S.M, Ryon, M.G.A. (1994). Comparison of health assessment approaches for

evaluating the effects of contaminant-related stress on fish populations. J. Aquat. Ecosyst.

Health 3, 15-25.

Attrill, M.J, Depledge, M.H. (1997). Community and population indicators of ecosystem

health: targeting links between levels of biological organization. Aquat. Toxicol. 38, 183 -

197.

Bespamyatnov GP, Krotov Yu.A. (1985). Maximum Permissible Concentrations of Chemical

Compounds in the Environment, Leningrad: Khimiya, 163 p. (in Russian).

Bioassay Methods for Aquatic Organisms. (1985). In: Standart Methods for the Examination

of Water and Wastewater, Washington, DC: Amer. Public Health Assoc., pp. 45–52.

Bitton, G. and Dutku, B.J. (1985). Introduction and Review of Microbial and Biochemical

Toxicity Screening Procedures. In: Standart Methods for the Examination of Water and

Wastewater, Washington, DC: Amer. Public Health Assoc., pp. 31–40.

Canadian Water Quality Guidelines. (1994). Published by Canadian Council of Ministry of

Environment.

Cash, K.J. (1995). Assessing and monitoring aquatic ecosystem health - approaches using

individual, population, and community/ecosystem measurements. In: Northern River

Basins Study Project, Report No 45,68 p.

Nova Science Publishers, Inc.

Tatyana I. Moiseenko, Alexandr G. Selukov and Dmitry N. Kyrov

256

Chinareva, I.D. (1988). Pathohistological changes in fish of the Ladoga basin. In: Pollution

Impact on the Ladoga Ecosystem. Leningrad: GosNIORKh, pp. 24–32 (in Russian).

Elliott M., Hemingway K.l., Krueger D., Thiel R., Hylland K., Arukwe A., Forlin l., Sayer M.

(2003). From the individual to the Population an Community Responses to Pollution. In:

Effects of Pollution on Fish (eds. Lawrence A.J., Hemingway) K.L. Blackwell Science

Ltd. pp. 221 -255.

Environmental Quality Objectives for Hazardous Substances in Aquatic Enviroment. (2001).

Berlin: UMWELTBUNDESAMT. 186p.

Fish Pathology, third edition. (1994). Ed./ Roberts, R.J. London.WB SAUNDERS.

Liney, K. E., Hagger, J. A., Tyler, Ch. R., Depledge, M. H., Galloway, T. S. and Jobling, S.

(2006). Health Effects in Fish of Long-Term Exposure to Effluents from Wastewater

Treatment Works. Environ. Health Perspect. 114, 81–89.

List of Fishery-Related Standards on Maximum Permissible Concentrations (MPC) and Safe

Reference Levels of Impact (SRLI) of Hazardous Substances for Water Bodies. (1999).

Used for Fishery and Aquatic Life). Moscow: VNIRO; 304 pp.

Lukin, A.A. and Sharova, Yu.N. (2004). Water Quality Estimation Based on Histological

Investigations of Fish: Case Study of Kenozero Lake. Water resour. 4, 481–489.

Manual on Hydrobiological Monitoring of Freshwater Ecosystems (Ed. Abakumov, V.A.)

(1992). St. Petersburg: Gidrometeoizdat, 308 p. (In Russian).

Methodological recommendations for establishing environmental and fishery standards for

pollutants in water for aquatic life and Fishery. (1998) (Ed. Filenko OF), Moscow:

VNIRO. (In Russian).

Moiseenko, T.I., and Kudryavtseva, L.P. (2002). Trace Metals Accumulation and Fish

Pathologies in Areas Affected by Mining and Metallurgical Enterprises, Environ. Poll,

114 (2), 285 - 297.

Moiseenko, T.I. (1998). Hematologic Characteristics of Fish in the Assessment of Their

Toxicoses, Russian Journal of Ichthyology, 2, 371–380.

Moiseenko, T.I., Voinov, A.A., Megorsky, V.V. (2006). Ecosystem and human health

assessment to define environmental management strategies: the case of long-term human

impacts on an Arctic lake. Sc. Tot. Environ. 369, 1-20.

Moiseenko T. I. Gashkina N., Sharova Yu., L.P. Kudryavtseva L. (2008). Ecotoxicological

assessment of water quality and ecosystem health: A case study of the Volga River,

Ecotoxicology and Environment Safety, 71, 837 -870.

Моисеенко T.I. (2009). Aquatic Ecotoxicology: Fundamental and Applied Aspects. M.:

Science. 367p. (in Russian).

Nemova, N.N. and Visotskaya, R.U. (2004). Biochemical indication of fish health М.: Nauka.

230 p (In Russian).

Niimi, A. J. (1990). Review of biochemical methods and other indicators to assess fish health in

aquatic ecosystems containing toxic chemicals. J. Great Lakes Res. 16, 529-541.

Odum, P. (1981). Fundamentals Ecology. EUGENE. Philadelphia-London-Toronto. 618 pp.

Savvaitova, K.A., Chebotarev, Yu.V., Pichugina, M.Yu., and Maksimov, S.V. (1995).

Anomalies in Fish Organisms as Indications to the Environmental Conditions. Russian

journal of Ichthyology, 2, 182–188.

Sidorov, V.S. and Yurovitskii, Yu.G. (1991). Perspectives of the Use of Biochemical

Methods for Recording Ecological Modulations. In: Ecological Modiﬁcations and

Nova Science Publishers, Inc.

Water Quality Assessment Methods 257

Environmental Standardization Criteria, Leningrad: Gidrometeoizdat, pp. 264–278. (In

Russian).

Werd J.H., Komen J. (1998). The effects of chronic stress on growth in fish: critical appraisal

// Comparative Biochemistry and Physiol. 120, 107-112.

Whitfield, AK, Elliott, M. (2002). Fish as indicator of environmental and ecological changes

within estuaries: a review of progress and suggestions for the future. J. Fish Biol. 61,

229-250.

Wong PTS, Dixon DG. (1995). Bioassessment of water quality. Environ. Toxicol. Water

Qual.10, 9–7.

Yeom D-H. and Adams S.M. (2007). Assessment effects of across level of biological

organization using an aquatic ecosystem health index. Ecotoxicol. Environ. Saf. 67,

286–295.

Zhiteneva, L.D., Poltavtseva, T.G., and Rudnetskaya, O.A. (1989). Atlas of Normal and

Pathologic Modiﬁed Blood Cells of Fish, Rostov-on-Don: Rostovskoe Kn. Izd. 110 p (In

Russian).

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 10

WATER QUALITY IMPACTS ON HUMAN POPULATION

HEALTH IN MINING-AND-METALLURGICAL

INDUSTRY REGIONS, RUSSIA

T. I. Moiseenko1,

∗

, N. A. Gashkina1, V. V. Megorskii2,

L. P. Kudryavtseva2, D. N. Kyrov3 and S. V. Sokolkova3

1Institute of Geochemistry and Analytical Chemistry,

Russian Academy of Sciences, Russia;

2Institute of Problems of Industrial Ecology of the North,

Kola Research Center, Russia

3Tyumen State University, Russia

ABSTRACT

The pollution of water sources and drinking water in some towns and settlements of

the Kola North by metals, wastewaters, and airborne pollution effluents of mining-and-

metallurgical industry is characterized in this chapter. Our statistical data on population

morbidity are provided, as well. Over the course of our studies, relationships were found

to exist between water quality indices and heavy metal accumulation in kidney and liver

of postmortem-examined patients, and the results of their histological, clinical, and

postmortem examination are given. The results of comprehensive studies are used to

assess the effect of drinking water pollution on the population health in the region.

1. INTRODUCTION

The Arctic region is a part of the Planet where the territory is covered by a very great

number of lakes and rivers. The high provision of the Arctic regions with water till recently

has not caused any trouble about the state of the latter. At the same time, intensive

development of the rich deposits of mineral recourses and trans-boundary pollutions lead to a

∗ Corresponding author.

Nova Science Publishers, Inc.

T. I. Moiseenko, N. A. Gashkina, V. V. Megorskii et al.

260

rapid disturbance in the fragile environmental equilibrium already in many urbanized and

industrial Arctic regions, which leads to qualitative depletion of the water resources.

The Kola peninsula of Russia is most densely populated and industrially developed

region of Arctic. The spectrum of anthropogenic impacts on the surface water is wide:

mining, metallurgy, refineries and chemical industries, nuclear power plants, etc. Industrial

development of copper- nickel, rich apatite-nephelinite and iron deposits in the Kola

Peninsula began in the 1930s. Large amounts of pollutants entered the lakes between 1940

and 1990. The main pollutants were heavy metals (predominantly nickel and copper),

sulphates, chlorides and nutrients. The catchment areas were also polluted by airborne

contaminants, metals and acid deposition. Since 1990, as a result of the economic crisis in

Russia, the anthropogenic pressure on the lakes has decreased.

For more than 70 years the lakes have been used as a source of the technical and drinking

water supply, for recreation, tourism and fishery. Paradoxically though it may seem, the water

used for drinking water supply to major towns is often being taken from surface water bodies,

receiving effluents from industrial plants or located in the zone of airborne pollution. The

recent recovery (2000 -2010) of the economy goes on simultaneously with technological

modernization and stricter controls of pollutant emissions into the lakes and the atmosphere.

The map of the region and location of basic industries in it is shown in figure 1.

The aquatic environments are final collectors of all kinds of pollution. Contents of

elements in water reflect airborne metal contamination and pathways of metals. Life in water

bodies, as opposed to terrestrial conditions, is characterized by stronger relations between

aquatic organisms and factors of the environment due to the high role of ecological

metabolism in water ecosystems and high mobility of polluting substances in water. Four

basic processes are distinguished that lead to high contents of metals in the surface water of

region: (i) in waste waters of metallurgic manufactures; (ii) distribution with smoke

emissions; (iii) acid leaching from surrounding rocks, especially from natural geochemical

formations; iv) transboundary pollution to Arctic. Ni, Cu, As, Cd, Pb, Hg, Co, Zn, Mn and

rare-earth metals will enter the environment as a result of industrial activities.

Many metals, derived from the rocks and enriched in technological treatment, will

become toxic when they enter the environment. It is known that a number of human diseases

are connected with increased metal concentrations. The surplus of trace elements in the

human organism results in specific diseases: Hg causes a neurological effect, Cd and Pb have

cancerogenic properties, Sr leads to pathologies of bone tissues, Cu - to anemia, etc.

(Handbook of Metals…, 1994; Handbook of Ecotoxicology, 2005; Spry, Wiener, 1992).

Difficulties in determination of dangerous metal levels for vitality are stipulated by the

following factors: (i) many elements (Cu, Zn, Co, Sr, Se, Ni etc.) are significant, i.e. inherent

to organisms and are present in organisms in microquantities; (ii) poisoning influence of

metals is formed both due to direct effects and ability to be accumulated in organisms,

causing remote consequences - mutagenic, embriotoxic, gonadotoxic, cancerogenic, etc.; (iii)

toxicological properties depend on metal speciation, combinations of elements (phenomena of

sinergetism and antagonism) and concomitant factors. The aquatic environments are final

collectors of all kinds of pollution. Contents of elements in water reflect airborne metal

contamination and pathways of metals. New geochemical provinces are shown to form under

current conditions (Moiseenko, Kudryatseva, 2002), and the environmental consequences of

this process require close attention, and especially—studying the effect of surface water

pollution on human health.

Nova Science Publishers, Inc.

Water Quality Impacts on Human Population Health … 261

Figure 1. Layout of (1) copper_nickel works, (2) mines and mining and concentration plants, (3)

population localities, and (4) drinking water intakes.

It is necessary to note, that hazardous substances enter the human organism with food,

water, and air; therefore, it is often difficult to establish correlation between drinking water

quality and population health. However, it should be taken into account that people in

transpolar regions largely eat products delivered from southern regions; therefore, the input of

pollutants with food made from local agricultural products has not significant effect on

population morbidity.This region can serve as a model region for understanding key

anthropogenically-inducted processes in lakes and its impacts for water quality and

population health.

The objectives of this study were as follows:

• the identification of pollution of surface water used for drinking water supply by toxic

heavy metals with six towns and settlements in the industrial Kola region of Arctic

used as an example;

• efficiency assessment of water purification of metals in the process of water

treatment;

• assessment of population morbidity in towns and settlements taking water for

drinking water supply from surface water bodies;

• assessment of the effect of drinking water pollution and heavy metals accumulation in

human organs on the population health, including the development of metal-induced

pathologies in humans.

Nova Science Publishers, Inc.

T. I. Moiseenko, N. A. Gashkina, V. V. Megorskii et al.

262

2. MATERIALS AND METHODS

Multidisciplinary studies were carried out in industrially developed towns (Monchegorsk,

Apatity, Polyarnye Zori, and Olenegorsk) and in more remote settlements (Alakurti and

Lovozero), which take water for drinking water supply from surface sources. The layout of

major industrial facilities and water intakes for drinking water supply is schematically shown

in Figure 1. Water for drinking water supply to Monchegorsk population is taken from

Monche Lake, which lies in the zone of aerotechnogenic pollution by emissions of

Severonikel industrial complex; the towns of Apatity and Polyarnye Zori consume water

originating from Lake Imandra.

The population morbidity in towns and settlements was assessed by using data from the

Committee of Public Health, Murmansk region; Murmansk Region Cancer Detection Centre;

and Moscow Research Oncological Institute.

Water intakes of Apatity and Polyarnye Zori are located in the zones of transit mixed

flow of wastewaters from various industrial facilities (copper-nickel smelter, mining-and-

processing of apatite and iron ores) and municipal wastes of Apatity and Monchegorsk towns.

The distances between the various discharge sites and the water intakes are 50–100 km for

Apatity and 100– 150 km for Polyarnye Zori. Water supply to Lovozero and Alakurti

settlements also relies on surface water bodies; however, they are far enough from industrial

centers (>200 km from the smoke emission plumes).

Drinking water samples for the analysis were taken from the source of water supply and

from networks for water supply to the population, allowing changes in water chemistry after

treatment and during its passage through pipes to be assessed.

Water samples were taken into Nalgen® polyethylene bottles, whose material has no

sorption capacity. The bottles were thoroughly cleaned in the laboratory and twice rinsed by

the lake’s water before sampling. Once taken, the samples were placed into dark containers

and cooled to about +4°C, at which temperature they were transported into the laboratory.

Water for trace element analysis was filtered in the field with the use of Milipore plant; both

filtered and unfiltered waters were acidified by nitric acid and in the prepared form send to

the laboratory for further analysis.

To study heavy metals accumulation in the tissues of residents of the examined populated

localities, samples were taken from the liver and kidneys of peoples who have lived in the

area for no less than 10 years and did not work immediately at plants with high health hazards

(110 postmortem-examined patients). In the case of chronic alcoholism or viral hepatitis, the

appropriate samples were rejected. The reference group or a “norm” of the trace element

composition of organs for the assessment of the extent of metal bioaccumulation was taken to

be appropriate tissues of fetuses (dead-born and immature births). It was assumed that the

transplacental barrier prevents metals from penetration into the organisms of developing

fetus. Simultaneously, postmortem samples were taken for histological examination, which

was further carried out by conventional methods. The consolidated conclusion regarding the

population morbidity was based on the results of analysis of clinical, histological, and

laboratory characteristics of the examined patients. The examined clinical features included

anamnesis data, the results of clinical examinations while alive and the pathologist’s report.

Determine of water chemistry are executed by uniform techniques according to the

recommendations (Standart method…, 1992) - Са2+, Mg2+, К+, Na+, Alk, SO42-, Сl-, color,

Nova Science Publishers, Inc.

Water Quality Impacts on Human Population Health … 263

NO3-, NH4, Ntot (ТN), РО4, Ptot (ТР), Si; pH – by a Metrohm® pH-meter; conductivity

(20oC) - by Metrohm®-conductivity; alkalinity - be using the Gran titration method, organic

matter content – by the Mn oxidation method. The applied analytical techniques and results of

the determination of the chemical composition of waters were verified using a common

system of standard solutions under permanent strict intralaboratory control.

Samples for determining trace element concentrations in biological sample were prepared

by wet decomposition in aquafortis with addition of hydrogen peroxide.

Concentration of elements (Sr, Al, Fe, Mn, Сr, Сu, Ni, Zn, Сd, Co, Рb, Аs) in water and

in sample the tissues of residents were determined twice: atomic absorption

spectrophotometry with graphite atomization Analyst-800 with Zeeman background corrector

and by plasma atomic emission spectroscopy «Plasma Quad 3» of firm «Fisons Instruments

Elemental Analisis» (Great Britain). Mercury concentration was determined on mercury

analyzer FIMS-100 (Perkin-Elmer).

3. RESULTS

3.1. Water Quality

Surface water of Kola ecoregion in reference condition was once characterized by very

clean water and oligotrophic conditions. Historical sampling indicated low concentrations of

suspended material (0.7 – 1.0 mg/l), microelements (<1 µg/l), and nutrients (e.g total

phosphorus < 2 µg/l). Biological uptake of any available nutrients was rapid, so phosphates

during the vegetation period were almost gone. Water transparency was about 8m

(Moiseenko, 1999).

Accordingly during these periods, the industries spawned the development of numerous

towns and cities to support their operations, including Monchegorsk, Olenegorsk, Apatity,

and Polyarnye Zori. Pollution from these areas, as well as from the industrial facilities, began

to affect the condition of the lake. Yet still, Lake Imandra was (and still is) used as a source of

drinking water for several cities, while serving as a sink of industrial and domestic

wastewater. The source of water supply Dumps of waste rock and processed ores, mine and

quarry water, wastewaters and dumped wastes of concentration plants are sources of

anthropogenic migration of some elements. Acid precipitation, which form under the effect of

atmospheric emissions of acid-forming agents by smelter and transboundary pollution,

accelerate the chemical leaching of elements and their migration.

3.2. Characteristic of Water Pollution Near Water Intakes and When

Supplied to the Population

Metal concentrations in water bodies involved in drinking water supply and in drinking

water supplied to the population of towns and settlements are given in Table 1.

Lake Monche lies in the zone of aerotechno-genic pollution from the source of smoke

emission of Severonikel mining-and-processing integrated works (<30 km). The lake’s water

is high in Ni, Cu, Cd, and other metals relative the background values.

Nova Science Publishers, Inc.

T. I. Moiseenko, N. A. Gashkina, V. V. Megorskii et al.

264

Table 1. Metal concentrations in natural water bodies at water intake sites for water

supply (top number) and in a distribution pipeline during the delivery to the population

(bottom number) in towns and settlements in Murmansk province

Water supply

source/populated

locality

Ni Cu Cd Pb Sr Cr Co ΣСi/MPCi

Lake Monche___

Monchegorsk

11.0

15.8

15.5

15.8

0.30

0.19

<0.5

0.6

0.2

0.4

0.5

0.345

0.233

Lake Imandra

Apatity

6.0

4.9

3.0

3.2

0.10

0.15

1.5

0.6

0.1

0.2

0.3

0.2

0.174

0.196

Lake Kuna____

Olenegorsk

1.4

0.9

2.3

1.6

0.10

0.09

<0.5

0.2

0.3

<0.2

0.113

0.103

River Virma____

Lovozero Settl.

0.5

0.8

0.7

1.4

0.10

0.13

<0.5

0.3

<0.2

0.116

0.144

River Tuntassaioki

Alakurti Settl.

0.5

0.4

0.2

0.6

0.11

0.14

<0.5

160

138

0.1

<0.2

0.136

0.161

The water intake of Olenegorsk Town is also situated within the propagation zone of

smoke emission, though at a larger distance (>50 km); accordingly, metal concentrations in

the lake’s water are lower. Lovozero and Alakurti settlements are supplied by water from

lakes where metal content of water is near the regional background level. The high

concentration of Sr in the water of Alakurti Settlment is accounted for by the geochemical

features of rocks in the area. Water to Apatity Town is supplied from Lake Imandra, which is

polluted by wastewater from Severonikel mining-and-processing integrated works. The

concentrations of metals in the water intake site in Lake Imandra is lower than in Lake Monche

because of the larger volume of the former lake, though they are far in excess of background

values. The concentrations of Ni and Cu at the wastewater discharge sites is 5–10 times

greater than those in the water intake zone (Moiseenko et al, 2006).

The water processed in the water treatment cycles is only slightly clearer than water in

natural water bodies. In population localities, Fe concentration in water is found to increase

during its passage through pipelines, because of Fe leaching from steel pipes. Higher Ni

concentrations were established for water supplied to the population in Monchegorsk Town.

Neither of metals was found to exceed the sanitary–hygienic standards for drinking water in

any population locality. However, it should be taken into account that water is polluted by a

complex of metals, which act in soft water against the background of very low Ca content, the

feature that enhances their penetrability and the adverse impact on the health of living

organisms, including humans.

3.3. Characteristics of Public Population Morbidity

According to statistical data, the most significant disease classes in the region are

diseases of blood circulatory system; neoplasms; diseases of respiratory organs, urogenital

system, and digestion organs, including hepatic cirrhosis. The highest morbidity rates are

Nova Science Publishers, Inc.

Water Quality Impacts on Human Population Health … 265

typical of the population of towns consuming water from lakes Imandra and Monche, where

the highest metal concentrations were recorded in drinking water (Figure 2).

According to data of the Committee of Public Health, the mortality rate in Monchegorsk

region increased by 14% over the past five years. The most unfavorable is the situation with

the growth of malignant neoplasms. Notwithstanding the relatively small percentage of

oncopathology among other diseases, its significance in the analysis of the population health is

high, since neoplasms have the highest death and invalidization rates as their consequences.

The identified classes of diseases, which are more inert to the effect of the etiologic factor,

can be largely dependent on the cumulative impact of chronic doses of heavy metals.

Clearly the effect of drinking water on humans is to one extent or another combined with

all environmental factors, as well as the social and economic conditions. Humans in the

Extreme North live under stress climate conditions.

Figure 2. Average statistical data on occurrence of major diseases in adult population in cities and

Murmansk region as a whole (a – disease of organs of digestion, b – diseases of genitourinary system, c

– urolithiasis, d – neoplasms). Occurrence measured as number of first diagnosed per 1000 population.

Nova Science Publishers, Inc.

T. I. Moiseenko, N. A. Gashkina, V. V. Megorskii et al.

266

It is possible that the complex of factors, including the effect of metals in low-

mineralization water, results in higher population morbidity in the towns under consideration

as compared with statistical data on Russia (Figure 2). The highest morbidity of population in

the towns consuming water from Imandra and Monche lakes suggests that human organisms

accumulate technogenically introduced metals, which can increase the population morbidity.

3.4. Accumulation of Heavy Metals and Accompanying Pathologies of

Organism Systems

More detailed studies in the system metals in water–their bioaccumulation–

histopathology–the diagnostics of pathologies with toxic etiology were carried out with the aim

to prove the reliability of the effect of drinking water pollution on the population health.

The concentrations of metals in the liver and kidneys of the examined patients are given in

Figure 3.

The highest indices of heavy metal accumulation in the liver were recorded in

Monchegorsk inhabitants, where the concentrations of many metals, especially, Ni, Cu, Cr,

Cd, and Pb, are 2–10 times larger than the norm. The highest concentration in the kidney

tissue were recorded for Cr and Cd – 10–50 times the norm.

Human organisms (liver and kidneys) mostly accumulate Cu, Cr, and Cd in Apatity

region; Cu, Cd, and Pb in Olenegorsk region; and Cu and Cd in Lovozero region and Alakurti

Settlement.

Despite the fact that Ni and Cd are the major drinking water pollutants in the region, Cd

features the largest ability to accumulate in human kidneys. It should be taken into account

that the Kola Region is subject to acid precipitation, which actively leach this element into

water. Lake Monche, where a municipal water intake is operated, lies in the zone subject to a

strong impact of acid precipitation. High concentrations of Cd were detected in people living

in Olenegorsk and Apatity.

By the complex accumulation of heavy metals in the liver and kidneys of inhabitants, the

towns can be ranked in the following order: Monchegorsk > Apatity > Olene-gorsk >

Lovozero > Alakurti. This series is in agreement with the order of decreasing HM

concentrations in the drinking water of the towns.

To assess the consequences of the chronic impact of subtoxic doses of metals and their

accumulation on human health, histologic specimens taken from extracted organs of

postmortem examined patients were analyzed for heavy metal content. Parallel to that,

clinical records and autopsy protocols were examined. The main objective was to identify the

latent forms (which were not revealed by life-time diagnostics) of diseases associated with

compromised liver and kidney function; it is therefore most likely that their etiology and

pathogenesis are associated with chronic metal intoxication.

Pathological processes in liver not diagnosed in the life-time were recorded in 24 cases out

of 110. The his-tological pattern of liver pathogenesis is most often represented by fatty

degeneration of hepatocytes, their sporadic necroses, autolytic decay in the periphery with the

formation of fat–protein detritus.

One case of hemosiderosis, five cases of toxic destruction, four different types of

dystrophy, and 10 cases of fatty degeneration were recorded. Most detected dystrophies are of

toxic etiology.

Nova Science Publishers, Inc.

Water Quality Impacts on Human Population Health … 267

Figure 3. Metal content in liver (1) and kidneys (2) of the examined patients in cities and settlements of

Murmans region.

The histologic picture of kidney pathogenesis is represented by nephrosclerosis at the

initial stage of its development, as well as by nephrosclerosis and amy-loidosis. All these

diseases are noninfectious and polyetiologic, which does not rule out their toxic patho-genesis

Nova Science Publishers, Inc.

T. I. Moiseenko, N. A. Gashkina, V. V. Megorskii et al.

268

under the effect of heavy metals. In addition to inflammatory changes in kidneys, specific

focal dystrophic and necrobiotic changes in blood vessels (capillaries, precapillaries, capillary

veins) were recorded. The morphological changes include small blood vessel diseases, an

increase in vascular permeability, resulting in tissue edema and serous or serosanguineous

inflammation in the tissues of the cortex and medulla of kidney. Moreover, 36 cases of

urolithiasis not revealed by life-time diagnostics were identified, accounting for 12% of the

specimens examined. These results are in agreement with statistical data on the incidence of

urolithiasis in the region. Latent cases of pathologies with toxic etiology can weaken the

organism and cause the development of accompanying diseases and an increase in population

death rates.

CONCLUSION

Hazardous substances enter the human organism with food, water, and air; therefore, it is

often difficult to establish correlation between drinking water quality and population health.

However, it should be taken into account that people in transpolar regions largely eat products

delivered from southern regions; therefore, the input of pollutants with food made from local

agricultural products has not significant effect on population morbidity.

Water bodies are the terminal accumulators of all types of pollution, including fallouts

from the polluted atmosphere. They determine, to one extent or another, the total dose of

pollutants entering the environment. In the northern region under study with widely

developed metallurgical and ore mining and processing industry, the leading environmental

factor adversely affecting the population health is the pollution of water bodies used as

sources for drinking water supply to the population. Pollutants can penetrate into the organism

with drinking water and accumulate in it, provoking diseases. According to medical–

demographic data, the Kola Region features higher incidence of urolithiasis and cholelithiasis

in people.

The toxic impact of chemical elements on human organism is governed by their chemical

nature, amount, and composition, as well as the individual features of the organism. The

threshold concentrations of individual elements vary depending on other elements present in

the environment and the organism. Surface continental waters in the extreme north regions

are generally low in Ca and mineralization. The toxicity of metals increases in low-

mineralization water, especially in ionic forms, which ensures their maximal penetrating

capacity and toxicity for living organisms.

The water processing system used at treatment plants is ineffective with respect to metals.

According to the authors' data, their concentrations in the water supplied to the population are

near and sometimes even higher than those in natural waters near the water intake. Earlier

detailed studies (Kudryavtseva, 1999) of drinking water and changes in metal concentrations

at different stages of water treatment showed that metal concentrations (in particular, Fe, Zn

and Mn) not only fail to drop, but even increase because of leaching from pipelines during

water delivery to the population (Figure 4).

Correlation analysis of population morbidity indices in towns and settlements with

averaged data on drinking water quality and the extent of metal accumulation confirm the

leading effect of drinking water pollution by metals on population health. Table 2 gives

Nova Science Publishers, Inc.

Water Quality Impacts on Human Population Health … 269

correlation coefficients between Ni, Co, and Cu accumulation in human liver and their

concentrations in drinking water.

Table 3 gives significant correlation coefficients between statistical data on human

morbidity and drinking water quality in the appropriate towns and settlements.

Figure 4. Dynamics of metal concentrations (MC) in drinking water in the process of treatment

(Kudryavtseva, 1999). (I) Water before treatment at the station, (II) water after treatment at the station,

(III) water from distribution hydrants, (IV) water from indoor water supply net-works.

Table 2. Correlation coefficients between metals accumulated in the liver and kidneys of

people and their concentrations in drinking water in different towns and settlements in

Murmansk region (here and in Table 3, significant correlation coefficients are shown in

bold type)

Organ Ni Cu Cd Pb Sr Cr Со

Liver 0.78 0.78 0.30 –0.45 –0.30 0.26 0.81

Kidney 0.87 –0.19 0.62 –0.29 –0.30 0.31 0.83

Table 3. Correlation coefficients between diseases and pathologies of systems and organs

in the population of Kola Peninsula towns and microelement concentrations in drinking

water

Organs Systems

kidney liver gastrointestinal hematopoietic cardiovascular

Ni 0.40

0.63 0.89 0.64 0.73

Cu 0.43

0.64 0.91 0.67 0.77

Co 0.93 0.75 0.85 0.90 0.91

Cr 0.60 0.53 0.51 0.76 0.70

Sr 0.56 0.18 -0.03 0.08 0.01

Cd 0.87 0.76 0.87 0.77 0.83

Pb 0.97 0.58 0.58 0.69 0.70

Zn 0.49 0.59

0.81 0.61 0.68

∑Ci/MPCi 0.76 0.77 0.95 0.79 0.86

Nova Science Publishers, Inc.

T. I. Moiseenko, N. A. Gashkina, V. V. Megorskii et al.

270

Interestingly, the incidence of kidney pathology has a high correlation coefficient with

the concentration in water of elements such as Cd, Co, and Pb, as well as with the index of

total exceedance of MPCdw. Similarly, the incidence of gastrointestinal, liver, and vascular

diseases correlates well with metal concentrations in drinking water.

Priority metal pollutants in surface continental waters of Kola region are Ni and Cu. Co is

an accompanying element of copper–nickel ores. Its excessive concentrations cause diseases

of heart, liver, and organs of vision and keratitis. Ni has a carcinogenic and gonadotoxic

effect (Sidorenko, Itskova, 1980). Epidemiologic examination of workers involved in Ni

production has shown that it can provoke cancerous diseases of nose, throat, and lungs.

Malignant tumors formed in warm-blood animals after Ni had been introduced in their

organisms. Higher incidence of nephrolithiasis and cholelithiasis was recorded in Kola

residents who use for drinking water taken from Ni-polluted water bodies. Cu, as well as Co is

not an acutely toxic element for humans.

When in very high concentration in water, Cu compounds can be toxic for warm-blood

animals and humans: liver diseases, anemia, and yellow sickness occur. MPCdw of Cu for

drinking water is 1 mg/l, while its concentration in drinking water in the Kola Region is two

orders of magnitude less. However, the correlations given above suggest that higher Cu

concentrations in drinking water have their effect on gastrointestinal diseases in people in

combination with other metals. Special attention should be paid to the high accumulation of

Cd in human kidneys, where its concentration exceeds those of other heavy metals.

This element is inessential and does not occur in the organism. The authors' studies have

shown its small amounts to occur in babies, which does not exclude its penetration through

the transplacental barrier. The body of data suggesting the high toxicity of Cd for humans has

increased in the recent years. Cd can cause pathological changes in organs and tissues; it affects

the progress of diseases, such as diabetes, hypertension, osteoporosis, leukemia, and the

development of neoplasms (Nishijo et al, 2000). Small doses of Cd result in its accumulation

in lungs, kidneys, and adrenal glands of humans, provoking kidney pathology (Nishijo, 1999).

The increase in cardiovascular diseases, including hypertension, in the group of people, is

associated with kidney pathology provoked by Cd. Measurements of Cd concentration in

urine covering a group of 22000 people in USA showed about 2.3% of them to have Cd

concentration in the organism exceeding the admissible level (Satarug et al., 2003).

Thus, there is a good reason to suppose that the inadequate quality of drinking water in

industrial regions of Extreme North provokes diseases in the population, a decrease in the

immunity, and perhaps, higher death rate.

Six towns and settlements in the Kola North are used to study the effect of drinking water

pollution on population health. Relationships are established in the system metal content of

drinking water → metal accumulation in human liver and kidneys → metal-induced

pathologies in organisms.

It was established that waters in the region are polluted by heavy metals because of the

ore mining and smelting works functioning here, the major polluting element there being Ni.

The higher concentrations of Cd (against the background) in waters of the region are

associated with this element leaching into surface water sources. The existing water

processing system fails to remove technogenic elements from the withdrawn water. Water in

Extreme North region has low mineralization and a high capacity for leaching elements from

pipelines (the presence of free carbon dioxide in the water), thus facilitating the high

migration activity of metals.

Nova Science Publishers, Inc.

Water Quality Impacts on Human Population Health … 271

The fact of accumulation of metals in human liver and kidneys has been established. The

high level of metal concentrations in the liver and kidneys of people (relative to a reference

group) is typical of residents of Monchegorsk Town, whose water intake is located in the

propagation zone of smoke emissions from copper–nickel smelting works. Notwithstanding

the fact that Ni concentrations are the highest in drinking water, the most abundant element in

human kidney tissue is Cd (50 times the norm). This element is highly toxic and provokes the

development of some pathologies in the organism. The highest morbidity rate in the

population (neoplasms, urolithiasis, cholelithiasis, glouronephritis) was shown to be

characteristic of Monchegorsk, where drinking water are higher in Ni, Cd, Pb, and other

elements. The population morbidity rate in other towns decreases with decreasing

concentrations of elements in drinking water.

Though the concentrations of toxic metals in water supplied to the population does not

exceed the standards accepted in Russia for drinking water, their long action and

accumulation in the systems of human organisms can cause pathologies and increase the

population morbidity.

The most significant classes of diseases in that region are blood circulatory system

diseases, neoplasms, and diseases of respiratory organs, urogenital. Analysis of statistical data

on drinking water quality, population morbidity rate, metal accumulation in kidneys and liver

in postmortem examined patient, as well as on their pathogenesis gives good grounds to

suppose that a direct relationship exists between the increase in the morbidity rate and

drinking water quality indices, and the etiology of the diseases depends on a prolonged

impact of metals on human organisms. These conclusions suggest the need to carry out more

thorough studies with the aim to correct drinking water quality standards with allowance

made for the specific regional conditions and factors of environmental pollution and to

develop more elaborate water treatment systems to reduce the risk of diseases in the

population.

ACKNOWLEDGMENTS

The work was supported by the Russian Foundation for Basic Research (Projects no 10-

05-00854), grant of Russian Governments (№ 11G34.31.0036) and Program 21 of Presidium

RAS “Basic Sciences to the Medicine”.

REFERENCES

Handbook of Ecotoxicology. (2005). / eds. Hoffman D.J., Rattner B.A., Burton G.A, Cairnce

J.Jr. N.Y.: Lewis Publishers, 501-556.

Handbook of Metals in Clinical and Analytical Chemistry. (1994). / eds. Seile, H.G.., Sigel,

A.., and Sigel, N.Y.: Dekker. 471 p.

Kudryavtseva, L.P. (1999). Assessment of Drinking Water Quality in the City of Apatity,

Water Resour. (Engl. Transl.), 26(6), 659-665.

Moiseenko, T.I., (1999). A Fate of Metals in Arctic Surface Waters. Method for Defining

Critical Levels. The Science of the Total Environ. 236, 19-39.

Nova Science Publishers, Inc.

T. I. Moiseenko, N. A. Gashkina, V. V. Megorskii et al.

272

Moiseenko, T.I., and Kudryavtseva, L.P. (2002). Trace Metals Accumulation and Fish

Pathologies in Areas Affected by Mining and Metallurgical Enterprises, Environ. Poll,

114 (2), 285-297.

Moiseenko, T.I., Kudryavtseva, L.P., and Gashkina, N.A. (2006). Trace Elements in Surface

Continental Waters: Technophilia, Bioaccumulation, and Ecotoxicology. Moscow:

Nauka, 261 p. (in Russian).

Nishijo M., Nakagawa H., Morikaw M. et al. (1999). Relationship between Urinary Cadmium

and Mortality among Inhabitants Living in a Cadmium Polluted Area in Japan, Toxicol.

Lett., 108. 321-327.

Nishijo, M., Hakagawa, H., and Kid, T. (2000). Environmental Cadmium Exposure and

Hypertension and Cardiovascular Risk, in Metals in Biology and Medicine, Paris: JA

John Libbey Eurotext, 6, 365-367.

Satarug, S., Baker, J.R., Urbenjapol, S., et al. (2003). A Global Perspective on Cadmium

Pollution and Toxicity in Non-Occupationally Exposed Population, Toxicol. Lett., 137,

65-83.

Sidorenko, G.I. and Itskova, A.I. (1980). Nickel: Hygienic Aspects of Environmental

Protection, Moscow: Meditsina. 176 p. (in Russian).

Spry D.J., Wiener J.G. (1991). Metal Bioavailability and toxicity to fish in low-alkalinity

lakes: a critical review. Environ. Pollution, 71, 243-304.

Standard Methods for the Examination of Water and Wastewater, (1992). American Public

Health Association, Washington. (D.C.), 1195 р.

Nova Science Publishers, Inc.

In: Water Quality ISBN: 978-1-62417-111-6

Chapter 11

CHITOSAN BIOPOLYMER FOR WATE R QUALITY

IMPROVEMENT: APPLICATION AND MECHANISMS

Xinchao Wei

∗

and F. Andrew Wolfe

Dept. of Engineering, Science and Mathematics, The State

University of New York Institute of Technology, Utica, NY

ABSTRACT

Chitosan is a cationic biopolymer derived from chitin, the second most abundant

natural fiber (next to cellulose), which is found in the shells of shrimp and crab. Chitosan

has drawn great attention as an effective biosorbent for various dissolved contaminants

mainly due to its high density amino groups (-NH2) and hydroxyl groups (-OH). Chitosan

has the highest adsorption capacity among the biopolymers. Chitosan can be a low-cost

alternative to granular activated carbon (GAC) and has been used to remove heavy metals

(Cu, Cd, Hg, Pb, Cr, As, and Se etc.) and anionic contaminants (phosphate, nitrate,

fluoride, perchlorate). This chapter highlights the recent research and development in the

application of chitosan for water quality improvement. It summarizes the advances and

results in removal of various contaminants using chitosan as an adsorbent. The

mechanisms involved in the adsorption of cationic and anionic species by chitosan are

presented and discussed.

INTRODUCTION

To ensure safe drinking water, governmental agencies undertake important actions to

assess and manage risks posed by various water quality contaminants, including man-made

and naturally occurring chemicals in surface water, groundwater, and wastewater. Water and

wastewater treatment plays a critical role in protecting public health and safeguarding the

environment. To meet the water quality standards and accelerate the advancement of

sustainable drinking water protection, innovative treatment technologies or materials are

∗ E-mail address: weix@sunyit.edu.

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

274

needed for public water systems to remove or mitigate groups of contaminants or contaminant

precursors from the water source. There are a wide range of technologies available to remove

contaminants from water, including physical, chemical, and biological processes. The specific

type of the treatment process or process combination to be employed depends upon the nature

of the contaminants, the scale of the plants, and the treatment costs. In general, the adsorption

process with granular activated carbon (GAC) is considered the most versatile method, which

can remove organics/inorganics, metals/nonmetals, and cations/anions. Moreover, the GAC

adsorption process can be conveniently operated [1-4].

However, the wide use of GAC in water treatment systems is often restricted due to its

relatively high cost [5] and the large carbon footprint during the production and regeneration

of GAC [6]. Therefore, it is desired to develop effective and economical adsorption processes

employing low-cost adsorbents. There are many nonconventional materials which have been

tested to adsorb various contaminants with some success. In particular, some abundant and

environmentally friendly biosorbents have demonstrated great promise in water and

wastewater treatment. Chitosan is a cationic biopolymer derived from chitin, the second most

abundant natural fiber (next to cellulose), which is found in the shells of shrimp and crab.

Chitosan has drawn great attention as an effective biosorbent for various dissolved

contaminants mainly due to its high density amino groups (-NH2) and hydroxyl groups (-OH)

[5, 7-10]. Chitosan has the highest adsorption capacity among the biopolymers [7, 11].

Chitosan can be a low-cost alternative to GAC and has been used to remove dissolved organic

contaminants and dissolved inorganics, such as metals and anions. Considering the wide

spectrum of contaminants which can be present in the water supply, there is a great advantage

to using chitosan as an adsorbent water and wastewater treatment.

CHITOSAN AND ITS PROPERTIES

Chitosan is typically derived from chitin, which is the second most abundant natural

polymer, next to cellulose, with a wide range of sources [5, 12]. However, chitosan is

primarily manufactured from crustaceans (crab and shrimp), because of the easy availability

of a large amount of the crustacean exoskeletons as a by-product of food processing [13, 14].

Chitosan is a partially deacetylated polymer of acetylglucosamine and its structure is shown

in Figure 1. Chitosan is a poly(aminosaccharide) consisting mainly of poly(1 → 4)-2 amino-

2-deoxy-D-glucose units [8]. The mole fraction of deacetylated units (glucosamine), defined

as the degree of deacetylation, is usually 70–90%. The molecular weight of commercial

chitosan varies significantly within the range 10,000–1,000,000 Da, depending on the

processing conditions.

While chitin is insoluble in most solvents, chitosan is soluble in water at low pHs (<6).

The water solubility of chitosan in acidic solutions is primarily due to the presence of amino

groups with a pKa value of ~6.3. At low pHs, the amino radicals undergo protonation and

become a water soluble polymer with positive charges. At high pHs (>6), the amino radicals

deprotonate and the cationic polymer becomes insoluble and subsequently precipitates as

solids. Chitosan can be solubilized in inorganic acids such as hydrochloric acid and nitric

acid. However, chitosan is insoluble in sulfuric and phosphoric acids [15]. Chitosan can also

be solubilized in organic acids such as, acetic, formic, lactic, and oxalic acids.

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 275

Figure 2. Chemical structure of chitin (a) and chitosan (b).

The solubility change of chitosan might affect its stability as an adsorbent in treating

water with low pHs. On the other hand, the solubility change with pH provides a convenient

way to prepare small chitosan beads for adsorption by spraying solubilized chitosan into an

alkaline solution media [16-18] or to coat other adsorbents with a layer of chitosan polymer

[11, 19, 20].

The adsorptive properties of chitosan depend on its deacetylation degree, crystallinity,

and molecular weight [10]. The free electron doublet of nitrogen on amino groups are the

active sites for adsorption of contaminants by chitosan and the degree of deacetylation

dictates the fraction of free amino groups available [21, 22]. The residual crystallinity in

chitosan after derivation from chitin influences the accessibility of contaminants to the

sorption sites and low crystallinity is preferred for adsorption.

One unique property of chitosan is that it can be readily modified physically or

chemically to improve its mechanical or chemical properties. With respect to the use of

chitosan as an adsorbent for contaminant removal, its disadvantage is due to its relatively

lower specific surface area (2-30 m2/g), as opposed to that of most commercial GAC based

adsorbents (800-1,500 m2/g)[23]. For example, chitosan flakes and powders with low surface

area and no porosity are typically not effective in adsorbing contaminants from water.

However, the adsorptive property can be improved when chitosan is prepared in such forms

as beads, fibers (solid or hollow), sponges, membranes, and nanoparticles [10, 23]. Chitosan

gel beads with controlled drying processes have enhanced volumetric adsorption capacity and

improved kinetic performance for metal removal [24]. The adsorption selectivity can be

improved by template formation or imprinting methods [25, 26].

Chemical modification is done to enhance the surface properties of chitosan as an

adsorbent, while maintaining its fundamental skeleton. For example, chemical modification

can prevent chitosan from dissolution into solutions, which is particularly important in

removing metals from acidic waters. Chemical modification can also lead to increased

adsorption capacity or enhanced selectivity for some specific contaminants. Chemical

adsorption is typically achieved by grafting new functional groups in the chitosan backbone

[10, 23]. The cross-linking agents may include bi-functional group (e.g. glutaraldehyde),

mono-functional group (e.g. epichlorhydrin), tri-polyphosphate groups, carboxylic groups,

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

276

amine groups and sulfonic groups [23]. The choice of cross-linking agents is determined by

the specific use and application in dealing with water with different characteristics.

REMOVAL OF HEAVY METALS

Heavy metals are the most ubiquitous and persistent contaminants in the aquatic

environment. They are toxic to aquatic plants and animals and they pose a threat to human

health either through direct consumption or bioaccumulation through the food-chain pyramid.

Heavy metals occur naturally in surface water and groundwater, and increasing amounts of

metals are released into the environment by anthropogenic sources such as the mining or

chemical industry. Many processes can be used to remove heavy metals from water including

precipitation, cationic exchange, reverse osmosis, and adsorption [27]. Chitosan is one of the

most extensively studied adsorbents for the removal of heavy metals from water and

wastewater.

Chitosan actually has the highest adsorption capacity for metals among the biopolymers

[11], mainly due to its high density amino and hydroxyl functional groups with high

adsorptive potential for a host of aquatic contaminants [5, 12]. The high adsorption capacity

of chitosan for metals can be attributed to (1) the high hydrophilicity due to a large number of

hydroxyl groups of glucose units, (2) the presence of a large number of functional groups

(acetamido, primary amino and/or hydroxyl groups) (3) the high chemical reactivity of these

groups and (4) the flexible structure of the polymer chain [12, 28]. Two mechanisms have

been clearly established for the interpretation of metal adsorption on chitosan materials, i.e.

chelation of metal cations to amine groups and ion exchange/electrostatic attraction [7].

Copper Removal

Copper (Cu) as one of heavy metals in the environment has been of great concern

because of their increased discharge, toxic nature and other adverse effects on receiving

waters [29]. Excessive Cu intake by humans can result in Cu accumulation in the livers and

Cu is toxic to aquatic organisms even at very small concentrations in natural waters [30] The

potential sources of copper pollution include metal cleaning and plating baths, pulp, paper

and paper board mills, wood pulp production, and the fertilizer industry [29].

There are many studies on Cu removal using various chitosan products with different

adsorption performances. Popuri et al. [31] coated polyvinyl chloride (PVC) beads with a

layer of chitosan to develop a biosorbent and the chitosan modified beads were examined in

batch and column experiments to remove Cu from solutions. It was found that the maximum

monolayer adsorption capacity of the biosorbent was 87.9 mg/g for Cu(II). Kyzas et al. [32]

prepared chitosan sorbents by cross-linking and grafting with amido or carboxyl groups. The

calculated maximum sorption capacity of the carboxyl-grafted chitosan was 318 mg/g at pH

6. The desorption experiments demonstrated that the sorbent can be regenerated without

significant loss in sorption capacity. Kannamba et al. [33] studied the removal of Cu(II) using

chitosan which were first cross-linked with epichlorohydrin and then chemically modified as

xanthate chitosan. The Cu(II) adsorption by chitosan was pH and temperature dependent The

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 277

maximum adsorption capacity of modified chitosan was 43.47 mg/g at pH 5.0 and 50 °C. The

desorption tests of loaded chitosan by sulfuric acid, hydrochloric acid and EDTA indicated

the adsorbent can be used for many adsorption and desorption cycles. The adsorption

mechanism involved the complexation of Cu(II) ions with the amine and thiol functional

groups in the modified chitosan. Futalan et al. [34] immobilized chitosan onto the surfaces of

bentonite and used the fixed-bed column to study the adsorption of Cu(II) from the aqueous

solutions. Chitosan immobilized bentonite was effective in the removal of Cu(II) in the

packed bed systems. The breakthrough curves and adsorption capacity was strongly

dependent upon the bed height, flow rate and influent concentrations. At the optimum

conditions, the adsorption capacity value is 14.92 mg/g with breakthrough time of 24 hr and

exhaustion time of 35 h [34]. The fixed bed results could provide essential information in the

practical design of permeable reactive barriers for the removal of heavy metals from

groundwater. Lopes [35] chemically modified chitosan through reaction with ethylenediamine

and diethylenetriamine. They found the effectiveness of such surface modification depended

on a prior reaction to incorporate cyanuric chloride as an intermediate biomaterial. In general,

chitosan derivatives were more effective than the original chitosan in interacting with the

copper cation, demonstrating that the chemically modified chitosans could be successfully

employed for copper removal from wastewater or industrial effluents [35]. In order to

increase the Cu sorption capcity of raw chitosan beads, Gandhi et al. [36] chemically

modified chitosan into protonated chitosan beads (PCB), carboxylated chitosan beads (CCB)

and grafted chitosan beads (GCB). The sorption capacities for PCB, CCB, and GCB were 52,

86 and 126 mg/g, respectively, as opposed to that of raw chitosan beads (40 mg/g).

Thermodynamic studies revealed that copper sorption onto chitosan bead products is

spontaneous and endothermic.

The mechanism of copper sorption onto the modified chitosan beads is governed by

adsorption, ion-exchange and chelation [36]. Chen and Wang [37] prepared magnetic

chitosan nanoparticles by chemical co-precipitation of Fe2+ and Fe3+ ions by NaOH in the

presence of chitosan, followed by hydrothermal treatment. The nanoparticles were super-

paramagnetic and the size was in the range of 8–40 nm. The Langmuir maximum sorption

capacity for Cu(II) was calculated to be 35.5 mg/g and more than 90% Cu(II) ions could be

desorbed from magnetic chitosan nanoparticles using 0.02–0.1 M EDTA. FTIR analysis

suggested that the removal mechanism of Cu(II) by chitosan involved –NH2 and –OH groups

[36]. Al-Karawi et al. [37] investigated the graft copolymerization of acrylamide onto

chitosan using potassium persulfate as initiator, and the grafted chitosan was used to remove

Cu(II). It was found that the Cu(II) adsorption equilibrium from Cu(II) followed the

Langmuir isotherm.

Cadmium Removal

Cadmium (Cd) is a heavy metal widely used in many industries including metallurgy,

surface treatment, dye synthesis and battery production. Due to its toxicity, Cd in aqueous

effluents has been a great concern for many decades. Cd in industrial wastewater can usually

be removed through chemical precipitation, leading to the production of highly toxic sludges,

which must be further treated before environmentally safe disposal. Other treatment options

include ion exchange, adsorption, electrodeposition and membrane systems [38].

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

278

Erosa et al. [38] studied the kinetic and equilibrium of Cd adsorption by chitosan and

found that chitosan was an effective adsorbent in removing Cd through chelation mechanisms

involving amine groups of chitosan. The adsorption capacity exceeded 150 mg/g at optimum

pH 7 and both intraparticle diffusion and external diffusion contributed to the kinetic control

of Cd(II) adsorption by chitosan. Hasan et al. [39] prepared chitosan-coated perlite beads via

the phase inversion of a liquid slurry of chitosan dissolved in oxalic acid and perlite to an

alkaline bath. The perlite beads revealed a porous chitosan structure. The Cd(II) adsorption on

chitosan was pH-dependent, and the maximum adsorption capacity of the chitosan-coated

perlite beads was determined to be 178.6 mg/g of bead at the pH 6.0 (the capacity was 558

mg/g of chitosan) [39]. Desorption was achieved by dilute HCl or EDTA solution. A novel

chitosan adsorbent was obtained via chemical modification by introducing a xanthate group

onto the backbone of chitosan [40]. The chemically modified chitosan showed superior

adsorption capacity (357.14 mg/g at the optimum pH 8) to the plain chitosan (85.46 mg/g).

The regeneration experiments demonstrated the chemically modified chitosan could be used

in Cd laden wastewater. The same adsorbent was also tested in fixed bed columns for Cd

removal from electroplating wastewater [41]. The adsorption column treated 367 bed volumes

of electroplating wastewater using the adsorbent, reducing the concentrations of Cd(II) from

10 to 0.1 mg/L [41]. Filho et al. [42] chemically modified chitosan with ethylene sulfide,

under solvent free conditions, to obtain an adsorbent with a high content of thiol groups. The

modified chitosan showed a good adsorption for Cd(II) with a capacity of 1.94 mmol/g.

Copello et al. [43] prepared a layer-by-layer silicate–chitosan composite biosorbent. The

films were evaluated on their stability regarding polymer leakage and their capability in the

removal of Cd(II) from an aqueous solution. For Cd(II) adsorption, the greatest adsorption

was obtained at pH 7. This non-covalent immobilization method allowed chitosan surface

retention and did not affect its adsorption properties [43]. Chitosan was also used to modify

the surfaces of zeolite [44], which functions as a porous support for chitosan. The modified

zeolite was used to remove Cd(II) from a micro-polluted water source. It was demonstrated

Cd (II) was more effectively removed by the modified zeolite powder than by natural zeolite

powder [44].

Mercury Removal

Mercury (Hg) is one of the most toxic heavy metals which persists in the environment

and creates long term contamination problems in the water, land and air [12]. The primary

anthropogenic sources of Hg pollution are urban runoff, mining discharge, coal combustion,

and industrial discharge. Once released into the environment, Hg typically undergoes

complex physical, chemical and biological transformation. The Hg toxicity depends strongly

on its redox states [45], while the most toxic form of Hg in the aquatic environment is the

high reactive Hg(II)[12]. Although there are a variety of technologies which can be used to

remove Hg(II) from water and wastewater, including precipitation, ion exchange, coagulation,

membrane process and adsorption, adsorption with different adsorbents seems to be the most

versatile and widely used technology for Hg removal.

Extensive research has been performed using chitosan as an adsorbent to remove Hg(II)

from the aqueous phase, and the Hg removal performance depends on the chitosan source, the

degree of deacetylation, and the experimental conditions such as pH, particle size, and ionic

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 279

strength, etc [12]. Benavente [46] studied the adsorption kinetic and equilibrium using

chitosan derived from shrimp shells and found that chitosan had the largest adsorption

capacity for Hg(II) compared with those for Cu, Zn and As. The Hg adsorption on chitosan

followed the pseudo-second-order equation and Langmuir isotherm model with a maximum

adsorption capacity of 106.4 mg/g at pH 4. NaCl solutions of 1M were effective to desorb Hg

for chitosan adsorption regeneration. Shafaei et al [47] evaluated Hg(II) removal using

chitosan with >85% deacetylation of three different sizes: 0.177mm, 0.5mm, and 1.19mm. In

general, higher pHs and smaller particle sizes favored Hg adsorption on chitosan. The

chitosan of 0.177mm showed the Langmuir maximum capacity of 1,127.1 mg/g (or 5.62

mmol/g) at pH 6, which was significantly higher than those reported by other researchers

[47]. Gamage and Shahidi [48] prepared three types of chitosan with different degrees of

deacetylation from fresh crab processing discards: 91.3%, 89.3%, and 86.4%. The chitosan

products were evaluated for their adsorption performance for several divalent metals at

different pHs in batch and column tests. It was found pH 7 served best for Hg adsorption and

higher degree of deacetylation resulted in better Hg(II) adsorption. As opposed to other

metals ( Mn(II), Co(II), Cd(II), Pb(II), Cu(II) and Zn(II)), chitosan demonstrated the best

adsorption potential for Hg(II). Chitosan (molecular weight of 54,000 g/mol) of a low degree

of deacetylation (20%) was prepared from red lobster shells by Polimeros et al [49] and used

for Hg(II) removal. The chitosan product was in the form of flakes with an average length

from 0.35 mm to 0.45 mm. The Langmuir maximum adsorption capacity of the chitosan for

Hg(II)was 1.8 mmol/g and 2.3 mmol/g ( or ~360 mg/g and ~460 mg/g), respectively. FTIR

analyses revealed both amine (-NH2) and –OH groups were involved in the retention of

Hg(II) on chitosan, due to complex formation. Lopes et al. [50] used the chitosan of 93%

deacetylation to prepare membrane pieces and investigated the Hg(II) adsorption on the

chitosan membrane. The Hg(II) adsorption kinetic did not follow the traditional Lagergren

model. The control step of Hg(II) adsorption by chitosan seemed to be the migration of Hg

anions into the pores of the chitosan membrane and the interaction of the ions with the

available adsorptive sites on the interior of the membrane. Peniche-Covas et al. [51] studied

the removal of Hg(II) from aqueous solutions using chitosan derived from lobster shells. They

found the intraparticle diffusion was the rate-limiting step in the adsorption of Hg(II) on

chitosan. The adsorption followed the Langmuir isotherm with a maximum capacity of 429.3

mg/g. The column experiments demonstrated that chitosan could be used to remove Hg(II)

from solutions in the absence of high levels of chlorides. In order to increase the adsorption

capacity of chitosan for Hg(II), Jeon and Holl [52] evaluated several chemical modifications

and prepared chitosan beads cross-linked with glutaraldehyde. They found the aminated

chitosan beads prepared via chemical reaction with ethylenediamine had the best adsorption

capacity of 2.3 mmol/g (~460 mg/g) at pH 7, which was attributed to the increased number of

amine groups. The chitosan beads showed the characteristic of competitive sorption between

Hg (II) and hydrogen ions [52].

Lead Removal

Lead (Pb) is one of the ubiquitous toxic metals which can be released into the aquatic

environment from natural and anthropogenic sources. The effect of Pb on neurobehavioral

development and brain cell function has been long recognized [53] and the bioaccumulation

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

280

of Pb in bones, muscles, kidney and brain can interfere with the normal physical activities of

humans. Due to its high toxicity, the U.S. Environmental Protection Agencies has set the

maximum contaminant level (MCL) of Pb ions in drinking water as 0.015 mg/L. Different

technologies has been developed for Pb removal from water and wastewater including

chemical precipitation, ion exchange, and adsorption.

Ng et al. [53] evaluated the removal of Pb(II) from aqueous solutions using chitosan of

three size ranges: 250–355 µm, 355–500 µm and 500–710µm. Chitosan with the smallest

sizes showed the best adsorption of Pb(II), and at pH 4.5, the Langmuir maximum adsorption

capacity was 0.558 mmol/g (~116 mg/g). Rangel-Martdez et al. [54] prepared chitosan from

shrimp heads by homogenous deacetylation and evaluated its effectiveness in removing

Pb(II). At pH 4, the chitosan adsorption for Pb(II) was better than that for Cd(II) and Cu(II).

Asandei et al [55] studied the adsortpon of Pb(II) using chitosan of 20.8 % deacetylation at

different conditions. The optimum pH for Pb(II) adsorption was found to be pH 6 with a

Langmuir maximum capacity of 47.39 mg/g. To address the high cost of chitosan beads, Wan

et al. [11] studied the immobilization of chitosan on a low-cost material (sand). The chitosan

coated sand demonstrated high efficiency in removing Pb(II) and higher pH favored Pb(II)

adsorption. The maximum adsorption capacity for Pb(II) was 13.32 mg/g, and the binding

strength between metal ions and sorbents showed high stability under neutral pHs. The

chitosan coated sand could potentially be used to build inexpensive large-scale filters as a

permeable reactive barrier for Pb(II) removal in contaminated groundwater. In order to

enhance adsorption performance of chitosan for Pb(II), chemical modification of cross-linked

chitosan through the introduction of xanthate was performed to remove Pb(II) in acidic

solutions [56]. Optimum adsorption was observed at pH 4 to 5 with a maximum capacity of

322 mg/g at pH4, which is higher than those from other reports. The enhanced adsorption was

attributed to the involvement of xanthate groups beside amine groups in the adsorption

process. The developed chitosan was also successfully applied to remove Pb(II) from actual

battery wastewater [56]. Yan and Bai [57] examined the adsorption behavior of chitosan

hydrogel beads for Pb(II) and the effect of the presence of humic acid (HA). Lead adsorption

on chitosan hydrogel beads was found to be strongly pH-dependent in the pH range of 5-7.5.

The amine group in chitosan was attributed to the adsorption of Pb(II) in water. Previously

adsorbed HA on chitosan beads led to enhanced Pb(II) adsorption while previously adsorbed

Pb(II) resulted in decreased HA adsorption. The formation of chitosan-HA-Pb structure had a

more favorable structure than that of chitosan-Pb-HA [57]. In a mixed Pb(II)-HA solutions,

the adsorption of chitosan for both species was significantly reduced, indicating the

interference of HA for Pb(II) removal. Jin and Bai [58] also investigated the mechanisms of

lead adsorption on chitosan/PVA (poly(vinyl alchohol)) beads in aqueous solutions. Lead

adsorption on chitosan/PVA beads was strongly pH-dependent with a maximum uptake

capacity at pH ~4. The chitosan/PVA beads exhibited positive ζ potentials at pH <6.3 and

negative ζ potentials at pH > 6.3. The adsorption mechanisms involved complexation, ion

exchange, and electrostatic interactions. Futalan et al. [34] studied the comparative and

competitive adsorption of Pb(II) with Cu(II) and Ni(II) using chitosan immobilized on

bentonite (CHB). It was observed that the adsorption capacity is highest for Pb(II) over Cu(II)

and Ni(II) for single and binary systems. The competition between metal ions for the binding

sites on CHB was strong as indicated from the low adsorption isotherm constants in binary

systems. To enhance the adsorption of Pb(II) on chitosan granules, Li and Bai [59]

successfully grafted poly(acrylic acid) (PAAc) on cross-linked chitosan granules (CTS)

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 281

through a simple two-step reaction in a solution. The PAAc modified CTS demonstrated

significantly greater adsorption capacity for Pb(II) (294.12 mg/g) and shorter equilibrium

time (5 hr), in contrast to CTS (95.15 mg/g and 8 hr). The enhanced performance of Pb(II)

adsorption of CTS-PAAc was attributed to the carboxyle groups grafted on CTS. In addition,

the adsorbed Pb(II) could be effectively desorbed and the CTS-PAAc could be regenerated

and reused without any loss of adsorption capacity [59].

Chromium Removal

Chromium (Cr) can be present in the aqueous environment in its trivalent (Cr(III)) and

hexavalent (VI) states. While Cr(III) is an essential nutrient, Cr(VI) is highly toxic and

considered carcinogenic. Unlike the Cr(III) cation, Cr(VI) typically exists in water as

chromate anion (CrO42-) or dichromate anion (Cr2O72-). Cr is a very important metal with a

very broad application such as metal finishing, hide tanning, and the manufacturing of alloys,

pigments and dyes. Typical Cr treatment processes include reduction, precipitation, ion

exchange, reverse osmosis and adsorption with activated carbon.

Rojas et al. [60] evaluated the Cr(III) and Cr(VI) removal from aqueous solutions using

chitosan powder (21-850 µm) and cross-linked chitosan flakes (850-2000 µm) with a

molecular weight of 125,000 Da and 87% deacetylation. Cross-linked chitosan was efficient

in removing Cr(VI) ( adsorption capacity of 215 mg/g) and less efficient in eliminating Cr(III)

(adsorption capacity 6 mg/g) due to the protonated active sites interacting mainly with anions

at low pHs. de Castro Dantas et al. [61] examined the Cr(III) removal from aqueous solutions

by chitosan impregnated with a microeulsion. A significant increase in Cr(III) adsorption

capacity was achieved compared to the untreated chitosan. Dynamic column experiments

demonstrated that Cr(III) adsorption is pH-dependent and the amount of Cr(III) sorbed

increased with increasing initial concentrations. The presence of Cu(II) interfered with Cr(III)

adsorption. Schmuhl et al. [62] studied the Cr(VI) adsorption by chitosan and chitosan cross-

linked by epichlorohydrin. The adsorption behavior of Cr(VI) was described by the Langmuir

isotherm. The maximum capacity was 78 mg/g for non-cross-linked chitosan and 50 mg/g for

cross-linked chitosan, indicating cross-linking did not improve Cr(VI) adsorption. When

chitosan was cross-linked, some of the adsorptive sites were used for cross-linking [62].

Baroni et al. [63] tested the Cr(VI) removal using chitosan and chitosan cross-linked by

glutaraldehyde and epichlorohydrin. They found both types of cross-linked chitosan showed

significantly higher adsorption capacity for Cr(VI) than the plain chitosan, contradictory with

the findings of Schmuhl et al. [62]. The maximum Cr(VI) adsorbed was about 1,400 mg/g for

epichlorohydrin cross-linked chitosan at pH 6.0, which is the highest number reported in the

literature for Cr(VI). The desorption tests with NaCl also showed chitosan cross-linked by

epichlorohydrin was more effective for Cr(VI) removal at pH 2 than pH 6. Spinelli et al. [64]

chemically modified the surface of chitosan with glycidyl trimethyl ammonium chloride in

order to produce a chitosan quaternary ammonium salt that works as a strongly basic

exchanger. In order to study the adsorption of Cr(VI) ions in the form of an oxyanion, they

cross-linked the chitosan salt with glutaraldehyde making it insoluble in aqueous solution.

The maximum adsorption capacity was 68.3 mg/g (1.31 mmol/g) at pH 4.5 and the sorbed

Cr(VI) can be desorbed by 1 M NaCl/NaOH solution [64]. Similarly, a novel chitosan

adsorbent was developed by chemically modifying chitosan through grafting glutaraldehyde,

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

282

xanthante group onto the backbone of chitosan [65]. Both modified chitosan beads and flakes

showed improved adsorption for Cr(VI) with maximum capacities of 625 mg/g and 256 mg/g,

respectively. The desorption test indicated that the modified chitosan products can be reused

after at least 10 cycles without any significant change in adsorption capacity. Boddu et al.

[66] developed a new composite chitosan biosorbent by coating chitosan onto ceramic

alumina and studied the Cr(VI) adsorption with the adsorbent in both batch isothermal and

continuous column adsorption experiments. The effect of competing anions such as sulfate

and chloride was also investigated as well as that of pH. The ultimate capacity obtained from

the Langmuir model is 153.85 mg/g, indicating great adsorption capacity for Cr(VI). Hasan et

al. [67] prepared chitosan coated perlite beads by drop-wise addition of chitosan solution to

an alkaline bath. The beads containing 23% chitosan showed a good affinity for Cr(VI) with

an adsorption capacity of 104 mg/g based on the beads (452 mg/g based on chitosan itself),

which was higher than that of the chitosan only adsorbents (11.3-78 mg/g).

Arsenic Removal

Arsenic (As) is a highly toxic and carcinogenic metalloid which can be released into the

aquatic environment naturally or anthropogenically. The World Health Organization (WHO)

recommends the maximum As level of 10 µg/L in drinking water. Long-term consumption of

arsenic contaminated water can lead to severe and permanent impairment of human health. In

the aquatic environment, the most common and mobile As species are As(III) and As(V),

existing as arsenite (As(III) or AsO33-) and arsenate (As(V), or AsO43-) anions. Due to the

stringent As regulations, there has been extensive research on arsenic removal technologies

demonstrating varying levels of success. As removal can be achieved using precipitation,

coagulation, oxidation, ion exchange and adsorption with iron-based adsorbents. In general,

As(V) is relatively easy to remove from the aqueous phase while As(III) is more troublesome

and requires oxidation as a pretreatment to achieve the satisfactory results.

Boddu et al. [68] prepared a composite chitosan biosorbent (CCB) by coating chitosan on

ceramic alumina via a dip-coating process and evaluated the adsorption capacity of CCB for

As(III) and As(V) under equilibrium and dynamic experimental conditions. CCB contained

21.1 wt % of chitosan with a specific surface area of 125.24 m2/g. The adsorbent showed an

impressive adsorption capacity for As(III) and As(V) at pH 4 (56.50 and 96.46 mg/g,

respectively, based on Langmuir isotherm data).The better adsorption for As(V) was

explained by the speciation of arsenic at pH 4. The adsorbent was also suitable for column

application with breakthrough bed volumes of 40 and 120 for As(III) and As(V), respectively,

and the column bed could be regenerated using 0.1M NaOH solutions. Gupta et al. [69]

prepared iron coated chitosan flakes (ICF) and iron doped chitosan beads (ICB) and used

them as adsorbents to remove As(III) and As(V) from actual groundwater in batch and

column experiments. The monolayer adsorption capacities of ICF were 22.47 mg/g for As(V)

and 16.15 mg/g for As(III), which were considerably higher than those obtained for ICB (2.24

mg/g for As(V) and 2.32 mg/g for As(III)). The competing anions such as sulfate, phosphate

and silicate at the levels typical of groundwater did not cause serious interference in the

adsorption behavior of As(III) and As(V). The column regeneration studies were carried out

for two sorption–desorption cycles for both As(III) and As(V) using ICF and ICB as sorbents.

The adsorbents were also successful in treating the actual arsenic contaminated groundwater

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 283

below <10 μg/L (total As) and the adsorbent could be regenerated using 0.1M NaOH

solution. Gerente et al [70] investigated the sorption As(V) onto chitosan. They proposed the

fixation of As(V) onto chitosan included the following steps: (i) the increase of amine

protonation and zeta potential at the surface of the chitosan by a decrease of the pH, (ii) the

fixation by electrostatic attraction between a positive surface charge located on the amine

function and the anionic form H2AsO4−, (iii) thermodynamically spontaneous and exothermic

biosorption reaction. dos Santos [71] evaluated the effectiveness of chitosan-Fe-crosslinked

complex (Ch-Fe) in removing As(III). The maximum adsorption capacity of Ch-Fe was 13.4

mg/g at the optimum pH 9.0 and adsorption kinetics followed the pseudo-first-order kinetic

equation. Recently, Miller et al. [72] investigated As(III) and As(V) removal using TiO2-

impregnated chitosan beads (TICB) to maximize the capacity and kinetics of arsenic

adsorption. It was found that the adsorption capacity of TICB was influenced by pH and TiO2

loading within the bead and As removal was enhanced with exposure to UV light. The

effective pH for As(V) removal was pH < 7.25 while it was pH < 9.2 for As(III). The

reduction in bead size and exposure to UV light improved the adsorption rate. The common

background groundwater ions, except phosphate, did significantly interfere As(V) adsorption

onto TICB. Gupta et al. [73] prepared chitosan iron nanoparticles (CIN) by reducing Fe(III)

with NaBH4 in the presence of chitosan as a stabilizer and evaluated the As adsorption on

CIN. Remarkably high adsorption capacities of CIN was observed in a wide pH range for

both As(III) and As(V), i.e. 94 mg/L and 119 mg/L, respectively. The interfering anions

commonly found in groundwater such as sulfate, phosphate and silicate only marginally

influenced the adsorption of both As species onto CIN. The CIN adsorbent was successfully

regenerated for five cycles with 0.1M NaOH solution without a significant sacrifice in

adsorption capacity, indicating CIN could be a potential candidate for arsenic filtering units

for groundwater treatment or remediation. Gang et al. [74] developed an iron-impregnated

chitosan granular adsorbent and evaluated its ability to remove As(III) from water through

batch and column studies. The impregnation of iron into chitosan significantly increased the

As(III) adsorption capacity. The maximum adsorption capacity increased from 1.97 to 6.48

mg/g at pH = 8 as the initial concentration of As(III) increased from 0.3 to 1 mg/L.

Selenium Removal

Selenium (Se) is an emerging contaminant for many regions worldwide. It is an essential

micronutrient for humans and animals, but considered toxic when ingested in amounts higher

than those needed for optimum nutrition. Se regulations vary from country to country. For

drinking water, most countries adopt the 10 µg-Se/L limit of World Health Organization

(WHO) guideline. In the US, 50 μg-Se/L is still the current Environmental Protection Agency

(EPA) limit for both the maximum contaminant level (MCL) and the MCL goal (MCLG),

although a new limit of 5 μg-Se/L is being proposed. For surface water, the US Clean Water

Act (CWA) lists Se as a priority toxic pollutant and adopts freshwater acute and chronic

criteria of 20 µg-Se/L and 5 µg-Se/L, respectively. Se levels exceeding the freshwater criteria

can pose a serious risk to aquatic life and humans due to the likelihood of bioaccumulation

through the food chain. In addition to its natural occurrence in the environment, Se can be

released as a result of anthropogenic processes such as the mining of minerals, combustion of

coal, metal smelting, oil refining and utilization, and agricultural irrigation In aquatic

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

284

environments, Se can be present in four different oxidation states viz. selenide (Se2-),

elemental selenium (Se0), selenite (SeO32-), and selenate (SeO42-), of which selenite or Se(IV)

(SeO32-) and selenate or Se (VI) (SeO42-) are more soluble and mobile. Technologies in

removing Se from water and wastewater include: 1) conventional water treatment practices

such as lime neutralization, softening and ferric coagulation , 2) ion exchange and membrane

processes such as reverse osmosis, nanofiltration, and emulsion liquid membranes, 3) Se

reduction, 4) biological processes, and 5) adsorption processes using such adsorbents as

alumina, activated carbon, manganese nodule leached residues sulphuric acid-treated peanut

shell and various iron oxides/hydroxides [75].

Bleiman and Mishael [76] prepared chitosan-montmorillonite composite to remove

Se(VI) from water. Compared with Al-oxide and Fe-oxide, the chitosan composite exhibited

higher capacity for Se(VI) (18.4 mg/g). Furthermore, Se(VI) adsorption by the composite was

not pH dependent while its adsorption by the Fe- or Al- oxides decreased at high pHs. The

experiments with actual well water indicated that the chitosan composite was able to reduce

Se level to below 10 µg/L and the adsorption of Se was relatively selective in the presence of

13 mg/L sulfate. Yang et al. [77] developed chitosan-coated quartz sand and examined its

performance in removing Se(IV) from groundwater. The effects of pH, adsorption time,

temperature, and initial Se concentration were studied. It was found, at a dose of 40 g/L, the

chitosan coated sand reduced the Se(IV) to 8.66 µg/L and monolayer adsorption is the

primary mechanism for Se(IV) adsorption. Sabarudin et al. [78] synthesized a chitosan resin

functionalized with 3,4-diamino benzoic acid (CCTS-DBA resin) using a cross-linked

chitosan (CCTS) as base material. It was found that Se(VI) was strongly adsorbed at pH 2 and

pH 3 as an oxyanion of SeO42−, while selenium(IV) as HSeO3− was adsorbed on the resin at

pH 3. The adsorption capacities were 64 and 88 mg/g for Se(IV)and Se(VI), respectively. The

common anions such as chloride, sulfate, phosphate and nitrate did not affect the Se

adsorption when their concentrations were less than 20 ppm.

REMOVAL OF ANIONIC CONTAMINANTS

Nitrate Removal

Nitrate (NO3-) is one of the most common contaminants in surface water and

groundwater. Excessive amount of nitrate can cause eutrophication problems in surface

waters and outbreaks of infectious disease. Nitrate concentration higher than the regulation

allowed level (10 mg/L, U.S. EPA) in drinking water can lead to potential human health

issues especially for babies and children. Nitrate removal methods include biological

denitrification, chemical reduction, reverse osmosis, electrodialysis, and ion exchanges [79].

Chatterjee and Woo [79] prepared chitosan hydrobeads of 2.5 mm by a drop-wise

addition of chitosan solution to an alkaline coagulating mixture and examined the nitrate

adsorption on the chitosan hydrobeads. The adsorption process was pH and temperature

dependent. Lower pHs favored nitrate adsorption because a pH decrease in solution resulted

in more protons available for the chitosan amine group for a more positive charge at the

chitosan surface. The enhanced adsorption of nitrate was due to electrostatic interactions

between negatively charged nitrate groups and positively charge amine groups. When the pH

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 285

was above 6.4, the physical forces were attributed to the appreciable amount of nitrate

adsorbed by chitosan hydrobeads. The maximum adsorption capacity was 92.1 mg/g at 30ºC.

The kinetic results corresponded well with the pseudo-second-order rate equation and intra-

particle diffusion played an important part at the initial stage of the adsorption process. The

loaded nitrate was desorbed by increasing pH to the alkaline range and 87% desorption was

achieved at pH 12. To improve the nitrate adsorption capacity, chitosan beads were modified

by cross-linking with epichlorohydrin (ECH) and surface conditioning with sodium bisulfate

[80]. The maximum adsorption capacity was found at a cross-linking ratio of 0.4 and

conditioning concentration of 0.1mM NaHSO4. The maximum adsorption capacity was

104.0mg/g for the conditioned cross-linked chitosan beads at pH 5, as opposed to 90.7mg/g

for normal chitosan beads. The high adsorption capacity values for all adsorption systems in

acidic solutions (pH 3–5) were attributed to the strong electrostatic interactions between its

adsorption sites and the nitrate. Arora et al. [81] modified the zeolite surfaces with a layer of

chitosan to prepare chitosan coated zeolite (Ch-Z). The analyses found that chitosan did not

provide a full coating of the zeolite particles. Ch-Z exhibited a comparable capacity to other

weak anion exchangers with a nitrate ion exchange capacity 0.74 mmol/g (~46 mg/g).

Athough the nitrate exchange capacity was reasonable, Ch-Z was shown to be more selective

towards sulfate and chloride than nitrate. Jaafari et al. [82] prepared chitosan gel beads and

cross-linked with glutaraldehyde and evaluated nitrate adsorption by the chitosan beads. It

was found protonated cross-linked chitosan was able to remove nitrate from model water and

contaminated surface and groundwater to meet the drinking water standard. The reactive

process involved the total volume of the chitosan beads, not only the surfaces. Fluoride was

the main competing anions while chloride and sulfate did not interfere with nitrate adsorption.

Phosphate Removal

Excess phosphate (PO43-) is believed to be the cause of eutrophication problems in

surface waters and can also lead to other water quality problems. The sources of phosphate

pollution include domestic source, agricultural runoff, urban runoff, industrial effluent,

municipal wastewater discharge and mining drainage etc. Typical phosphate removal

technologies include biological process, chemical precipitation, and adsorption.

Dai et al [83] examined the feasibility of phosphate removal by using the spent chitosan

beads after copper adsorption. It was found that the spent chitosan beads loaded with copper

were stable and suitable for phosphate removal in a wide pH range and the maximum

adsorption capacity (28.86 mg-P/g) was achieved at pH ~5. Among the species of phosphates,

dihydrogen phosphate was the preferred species for Cu(II)-loaded chitosan beads to adsorb.

The effect of competing anions on adsorption capacity indicated that chloride and sulfate

interfered with phosphate adsorption on the chitosan beads. The adsorption equilibrium and

kinetic study indicated that the adsorption behavior was mainly chemical monolayer

adsorption for phosphate facile to bind with Cu(II). Fierro et al. [84] investigated the use of

chitosan immobilized algae for phosphate removal from water. Dark green microalgal

colonies were observed in the outer and inner portions of chitosan beads. The chitosan

immobilized algae showed a very high efficiency of phosphate removal(>94%) at an initial

phosphate concentration of 6 mg-P/L, which was better than plain chitosan bead (60%

removal)[84].

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

286

Chung et al. [85] evaluated the feasibility of using chitosan of different molecular

weights to simultaneously remove various pollutants from the discharge of an eel culture

pond. Experimental results indicated chitosan of a low molecular weight excelled at removing

phosphate as well as ammonium from the wastewater. The best performance of chitosan was

99.1% for phosphate removal. The best removal of chitosan for turbidity, suspended solids,

BOD, COD, NH3 and bacteria was 87.7%, 62.6%, 52.3%, 62.8%, 91.8%, and 99.998%,

respectively.

These results indicated that it was feasible to use chitosan to treat the aquaculture

effluents to minimize deterioration of receiving water quality. Shimizu et al. [86] studied the

removal of phosphate with a chemically modified chitosan/metal-ion complex. It was found

that Cu(II) and Fe(III) based chitosan complexes had the highest adsorption ability toward

phosphate. Fagundes et al. [87] investigated the adsorption of phosphate on iron(III)-cross-

linked chitosan in an experiment of solid-phase extraction. A batch study showed that

phosphate adsorption reached equilibrium and pH 7.0 is the optimal pH for phosphate

adsorption.

The maximum adsorption capacity determined by Langmuir equation was 131 mg/g. For

the column study, an increase in the flow-rate resulted in reduced breakthrough volume (180

and 230 mL) and breakthrough sorption capacity (18 and 23 mg/g). Zheng and Wang [88]

prepared a chitosan-g-poly(acrylic acid)/vermiculite composite by aqueous dispersion

polymerization using chitosan as the stabilizer, acrylic acid as the monomer and vermiculite

as the inorganic additive. The composite was then cross-linked with common divalent or

trivalent cations to obtain adsorbents with a higher affinity for phosphate ions. The results

demonstrated that the trivalent ion cross-linked hybrid exhibited a phosphate potential for the

removal.

The adsorption of phosphate ions onto the developed adsorbent was pH-dependent and a

lower pH led to a higher adsorption capacity. The maximum adsorption capacity was 22.64

mg/g, comparable with those reported for other adsorbents. Desorption studies indicated that

the adsorbent was relatively difficult to regenerate for reuse.

Fluoride Removal

While fluoride (F-) is an essential element for dental health, fluoride concentration above

1.5 mg/L in drinking water can be detrimental to human health, leading to dental or skeletal

fluorosis [23]. Consequently, the World Health Organization (WHO) has set a desirable and

permissible fluoride limit range of 0.5 – 1.0 mg/L for drinking water [23]. The main fluoride

source for drinking water is naturally occurring fluoride minerals in geological formations.

Anthropogenic release of fluoride into the environment is from various engineering processes

including semiconductor manufacturing, coal power generation, electroplating, rubber and

fertilizer production, etc. [23, 89]. Fluoride can be removed from water by various methods

such as ion exchange, coagulation, membrane process, and adsorption with various

adsorbents.

Miretzky and Cirelli [23] have provided an excellent review on the use of chitosan for

fluoride removal. Various chitosan and chitosan derivatives has been used for drinking water

treatment and they report maximum adsorption capacity ranges from 2.22 mg/g to 44 mg/L

[23]. The high variability in the performance of chitosan is mainly due to the disuniformity in

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 287

chitosan property, chitosan processing, and modification (physical and chemical). This review

is focused on the recent research and development in fluoride removal using chitosan since

2010.

Thakre et al. [90] synthesized lanthanum incorporated chitosan beads (LCB) using a

precipitation method under optimized conditions and examined their adsorption performance

in fluoride removal from drinking water. It was found that parameters for the synthesis of

LCB such as complexation time, precipitation time, ammonia strength and lanthanum loading

had a significant effect on fluoride removal. LCB effectively reduced the fluoride

concentration below the level of 1.5 mg/L. The fluoride adsorption capacity of LCB was 4.7

mg/g which was much greater than that of the commercially used activated alumina (1.7

mg/g).

LCB also possessed other advantages such as relatively fast kinetics, high chemical and

mechanical stability, high resistance to attrition, negligible lanthanum release, and suitability

for column applications. The same group of researchers also prepared lanthanum incorporated

chitosan flakes by lanthanum impregnation to enhance the fluoride removal capacity of

chitosan [91].

It was observed that the synthesis parameters have significant influence on development

of LCF and in turn on fluoride removal capacity. The LCF prepared under the optimal

condition showed a maximum adsorption capacity of 1.27 mg/g, which was lower than LCB.

SEM of LCF showed the presence of spherical particles spread over the chitosan matrix [91].

Viswanathan and Meenakshi [92] prepared a hydrotalcite/ chitosan composite using a co-

precipitation method to enhance fluoride removal. The composite showed an adsorption

capacity of 1,255 mg/kg, which was better than plain hydrotalcite (1,030 mg/kg) and plain

chitosan (52 mg/kg). Field trial studies indicated the hydrotalcite/chitosan composite was an

effective defluoridation agent. The same researchers prepared an alumina/chitosan composite

by incorporating alumina particles in the chitosan polymeric matrix [93]. The composite

displayed a maximum adsorption capacity of 3,809 mg /kg than the alumina (1566 mg/kg)

and chitosan (52 mg F−/kg). The fluoride removal by the chitosan composite was mainly

governed by electrostatic adsorption/ complexation mechanism.

Jagtap et al. [94] used chitosan as a template to prepared mesoporous alumina (MA450)

with improved properties fluoride removal from water. MA450 showed highly porous

structure of amorphous alumina with some partially converted chitosan residue. It was

observed that MA450 was effective over a wide range of pH (3-9) and showed a maximum

adsorption capacity 8.264 mg/g at an initial fluoride concentration of 5 mg/L, much better

than the conventional alumina. The order of anions interfering fluoride adsorption was

observed as HCO3− > SO4− > NO3− > Cl. MA450 also demonstrated significantly high

fluoride removal in field water. Vijaya et al. [95] developed a novel biosorbent, chitosan

coated calcium alginate (CCCA), by coating chitosan onto an anionic biopolymer calcium

alginate for the fluoride removal from aqueous solutions under batch equilibrium and column

flow experimental conditions. The adsorption process was optimized through the study of the

effects of pH, contact time, concentration of fluoride, and biosorbent dosage. The maximum

monolayer adsorption of fluoride on plain calcium alginate and CCCA were found to be 29.3

and 42.0 mg/g, indicating the enhancement of chitosan for fluoride adsorption. The

breakthrough curves were obtained from column flow tests and the experimental results

demonstrated that chitosan coated calcium alginate beads could be used for the defluoridation

of drinking water.

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

288

Perchlorate Removal

Perchlorate (ClO4-) is an emerging contaminant that has been detected in soil, surface

water and groundwater. As a toxic species, perchlorate can inhibit iodine uptake by the

thyroid gland and disturb normal metabolism, leading to physical or mental retardation or

other diseases such as neurological damage or anemia [96, 97]. Perchlorate is both a naturally

occurring and man-made contaminant and the anthropogenic source is the main concern.

Perchlorate has been used in products such as explosives, rocket fuels, fireworks, air bags,

bleaches, and fertilizers. Perchlorate, once released into the environment, can persist for

several decades due to its high solubility, non-reactivity, and poor adsorption to soil matrix

[96, 98]. Currently, U.S. EPA does not have regulation limit for perchlorate. However, it is

planning to develop national limits due to its toxicity. The technologies for perchlorate

removal include anionic exchange, biological reduction, chemical reduction, membrane

filtration, and adsorption with activated carbon.

Xie et al. [96] examined the perchlorate removal from aqueous solutions using chitosan

cross-linked with glutaradehyde in batch and column tests. The maximum monolayer

adsorption capacity was 45.5 mg/g and the optimum pH was determined to be pH 4. Column

adsorption indicated that the proper contact time was 8.1 minutes, indicating a rapid

adsorption. The effluent perchlorate concentration was kept below 24.5 µg/L for up to 95 bed

volumes with the influent perchlorate concentration of 10 mg/L. However, the presence of

competing anions, sulfate in particular, negatively influenced the perchlorate adsorption. The

adsorbent could well be regenerated with NaOH solution at pH 12 and reused for 15 cycles.

Electrostatic attraction as well as physical forces was believed to be the driving force for

perchlorate adsorption on cross-linked chitosan.

REMOVAL OF OTHER CONTAMINANTS

Ammonium Removal

Ammonium (NH4+) is commonly present in various wastewaters and it can be

transformed into nitrite and nitrate in the aquatic environment. Therefore, ammonium, as part

of the nitrogen nutrient along with phosphorus, is an important contaminant responsible for

eutrophication problems in surface waters. At high concentrations, ammonium itself is toxic

for many aquatic plants, fishes, and animals. The main source of ammonium pollution is from

wastewater effluents and agricultural and urban runoffs. Ammonium can be removed from

water by biological nitrification/denitrification, supercritical water oxidation, ion exchange

and adsorption.

Zheng and Wang [99] evaluated the removal of ammonium ions from aqueous solution

using a hydrogel composite chitosan grafted poly (acrylic acid)/ rectorite prepared from in-

situ copolymerization. The ammonium adsorption equilibrium can be reached within 3–5

min, indicating rapid ammonium adsorption kinetics. The hydrogel composite had a higher

adsorption capacity for ammonium (from 62 mg-N/g to 109 mg-N/g) in a wide pH levels

ranged from 4.0 to 9.0. No significant changes in the adsorption capacity were found over the

temperature range studied. Multivalent cations coexisting with ammonium in the solution had

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 289

some negative effects on the ammonium adsorption capacity. The regeneration condition for

the composite adsorbent was mild and the regenerated adsorbent was suitable for reuse in

ammonium removal. The electrostatic attraction between –COO− and NH4+ was believed to

be the main adsorption mechanism. The incorporation of inorganic clay particles improved

the hydrogel strength and enhanced the thermal stability. The same research group [100] also

developed a composite adsorbent with three-dimensional cross-linked polymeric networks

based on chitosan and attapulgite via in situ copolymerization in aqueous solution. The

efficacy of the composite adsorbent was examined for removing ammonium from synthetic

wastewater using batch adsorption experiments. At natural pH, the composite adsorbent had

an adsorption capacity of 1.0 mg-N/g, far higher than the other adsorbents (such as clay and

powdered activated carbon). Desorption of ammonium was achieved using 0.1 M NaOH

within 10 min. The results demonstrated the as-prepared composite was a fast-responsive and

high-capacity adsorbent for ammonium removal [100]. Chitosan was also used as the

backbone to prepare halloysite hydrogel composite adsorbents with an adsorption capacity of

32.87 mg-N/g [101].

Humic Acid Removal

Humic acid (HA) is a subclass of humic substances commonly present in surface waters.

It is macromolecular material possessing both hydrophobic and hydrophilic groups as well as

other functional groups such as carboxyl, phenolic, carbonyl and hydroxyl groups [102].

Humic acid in the aquatic environment predominantly carries negative charges due to the

existence of carboxylic and phenolic groups [103]. The presence of humic acid in surface

water has been of great concern in the water supply community, mainly due to the

consumption of disinfectant and the generation of disinfection byproducts. Therefore, it is of

practical significance to minimize the humic acid level in drinking water.

Maghsoodloo [104] investigated the equilibrium and kinetic adsorption of HA onto

chitosan treated granular activated carbon (GAC). The adsorption performance was compared

with that of GAC. The HA adsorption onto chitosan treated GAC followed the Langmuir

isotherm while Freundlich isotherm was better fitted for HA adsorption onto plain GAC,

indicating the chitosan coating changed the predominant adsorption mechanisms. Monolayer

HA capacities onto plain GAC and chitosan modified GAC were 55.8 mg/g and 71.4 mg/g,

respectively. Film diffusion and intraparticle diffusion were simultaneously participating in

the HA adsorption onto chitosan coated GAC. Ngah et al. [105] evaluated the HA adsorption

onto chitosan-H2SO4 beads in a batch study.

Based on the Langmuir isotherm model, the maximum adsorption capacities attained by

chitosan-H2SO4 beads were from 342 mg/g to 377 mg/g with low temperature favoring HA

adsorption. The same research group also examined HA adsorption onto chitosan beads cross-

linked by epichlorohydrin [106]. The optimum HA adsorption on cross-linked chitosan beads

was obtained at pH 6.0. The maximum adsorption capacity determined from the Langmuir

model was 44.84 mg/g. Zhang and Bai [107] coated the surfaces of polyethyleneterephthalate

(PET) granules with a layer of chitosan through a dip and phase inversion process and used

them as an adsorbent for HA removal. The uniform coverage of chitosan onto PET granules

with numerous open pores on the surface was confirmed by scanning electron microscopic

(SEM) images. The maximum adsorption capacity was 0.407 mg per gram of chitosan coated

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

290

PET granules. Adsorption of humic acid onto chitosan-coated granules was pH dependent and

significant amounts of humic acid could be adsorbed under acidic and neutral pH conditions.

Chitosan-coated granules were found to have positive zeta potentials at pH < 6.6, mainly due

to the protonation of the amino groups in chitosan. The adsorption process involved

protonation of the amino groups in chitosan followed by attachment of humic acid onto the

protonated amino sites on the surface. Under low-pH conditions, the adsorption process is

transport-controlled, but under high-pH conditions, both transport and attachment [107].

CONCLUSION

Improving water quality to protect human health and the environment is a challenging

task in view of ever increasing contaminants (in type and quantity) from industrial,

agricultural, and domestic sources. Chitosan, the second most abundant natural polymer, has

been demonstrated to be an effective biosorbent for many contaminants including metals,

anions, and other contaminants. The versatile adsorption properties of chitosan are mainly due

to its high content of amine and hydroxyl functional groups exhibiting high affinity to various

water contaminants.

Although chitosan has a very low specific surface area (2-30 m2/g) compared to

commercial activated carbon (800-1,500 m2/g) [12], the outstanding adsorption capacity

makes chitosan an attractive, low cost alternative to activated carbon because of its unique

attributes such as physicochemical characteristics, high reactivity, excellent chelation ability,

and high selectivity towards contaminants. Chitosan can be easily made into different shapes

to satisfy different applications such as beads, flakes, microspheres, films, membranes,

nanoparticles, and magnetic particles. Chitosan can also be modified by chemical or physical

processes to improve the mechanical and chemical properties for specific applications.

However, the disadvantages of chitosan as an adsorbent are its high solubility in acidic water

and the lack of selectivity for some contaminants as indicated by the interferences of

competing cations or anions.

REFERENCES

[1] R.C. Bansal, M. Goyal, Activated Carbon Adsorption, CRC press, Bota Raton, FL

(2005).

[2] EPA, Technologies for Upgrading or Designing New Drinking Water Treatment

Facilities, EPA/625/4-89/023, EPA Office of Drinking Water, Cincinnati, OH (1990).

[3] EPA, Small Community Water and Wastewater Treatment, EPA/625/R-92/010, EPA

Office of Research and Development, Washington, DC (1992).

[4] EPA, Drinking Water Treatment for Small Communities, EPA/640/K-94/003, EPA

Office of Research and Development, Washington, DC (1994).

[5] Bhatnagar and M.Sillanpää, Adv. Colloid Interf. Sci. 152, 26 (2009).

[6] P. Bayer, E. Heuer, U. Karl and M. Finkel, Water Res. 39, 1719 (2005).

[7] G. Crini, and P.-M. Badot, Prog. Polym. Sci. 33, 399 (2008).

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 291

[8] C. Gerente, V.K.C. Lee, P.Le. Cloirec and G. McKay, Cri. Rev. Environ. Sci. Technol.

37, 41(2007).

[9] A.J. Varma, S.V. Deshpande and J.F. Kennedy, Carbohydr. Polym. 55, 77 (2004).

[10] E. Guibal, Sep. Purif. Technol. 38, 43 (2004).

[11] M.-W. Wan, C.-C. Kan, B.D. Rogel and M.L.P. Dalida, Carbohyd. Polym. 80, 891

(2010).

[12] P. Miretzky and A.F. Cirelli, J. Hazard. Mater. 167, 10 (2009).

[13] E. Guibal, M.Van. Vooren, B.A. Dempsey and J. Roussy, Sep. Sci. Technol. 41, 2487

(2006).

[14] M.N.V.R. Kumar, React. Funct. Polym. 46, 1 (2000).

[15] C.K.S. Pillai, W. Paul and C.P. Sharma, Prog. Polym. Sci. 34, 641 (2009).

[16] S. Chatterjee and S.H. Woo, J. Hazard. Mater. 164, 1012 (2009).

[17] X.-F. Sun, S.-G. Wang, X.-W. Liu, W.-X. Gong, N. Bao and Y. Ma, Colloids. Surf. A

Physicochem. Eng. Asp. 324, 28 (2008).

[18] W.L. Yan and R. Bai, Water Res. 39, 688 (2005).

[19] M. Arora, N.K. Eddy, K.A. Mumford, Y. Baba, J.M. Perera and G.W. Stevens, Cold

Reg. Sci. Technol., 62, 92 (2010).

[20] V.M. Boddu, K. Abburi, J.L. Talbott, E.D. Smith and R. Haasch, Water Res. 42, 633

(2008).

[21] M.S.D. Erosa, T.I.S. Medina, R.N. Mendoza, M.A. Rodriguez, and E. Guibal,

Hydrometallurgy 61, 157 (2001).

[22] B. Benguella and H. Benaissa, Colloids Surf. A: Physicochem. Eng. Aspects 201, 142

(2002).

[23] P. Miretzky and A.F. Cirelli, J. Fluo. Chem. 132, 231 (2011).

[24] M.A. Ruiz, A.M. Sastre and E. Guibal, Sep. Sci. Technol. 37, 2143 (2002).

[25] T.W. Tan, X.J. He and W.X. Du, J. Chem. Tech. Biotechnol. 76, 191 (2001).

[26] Y. Baba, K. Masaaki and Y. Kawano, React. Funct. Polym. 36, 167(1998).

[27] W.W. Eckenfelder, Industrial Water Pollution Control, 3rd ed., McGraw-Hill, New

York, NY (2000).

[28] G. Crini, Prog. Polym. Sci. 30, 38 (2005).

[29] W.S. Wan Ngah, C.S. Endud and R. Mayanar, React. Funct. Polym. 50, 181 (2002).

[30] V.K Gupta. Ind. Eng. Chem. Res. 37, 192 (1998).

[31] S.R. Popuri, Y. Vijaya, V.M. Boddu and K. Abburi, Bioresour. Technol. 100, 194

(2009).

[32] G.Z. Kyzas, M. Kostoglou and N.K. Lazaridis, Chem. Eng. J. 152, 440 (2009).

[33] B. Kannamba, K. Laxma Reddy and B.V. AppaRao, J. Hazard. Mater. 175, 939 (2010)

[34] C.M. Futalan, C.-C. Kan, M.L. Dalida, C. Pascua, M.W. Wan, Carbohydr. Polym. 83,

697 (2011).

[35] E.C.N. Lopes, K.S. Sousa and C. Airoldi, Thermochim. Acta 483, 21 (2009).

[36] Y. Chen and J. Wang, Chem. Eng. J. 168, 286 (2011).

[37] J.M. Al-Karawi, Z.H.J. Al-Qaisi, H.I. Abdullah, A.M. A. Al-Mokaram and D.T.A. Al-

Heetimi, Carbohydr. Polym. 83, 495 (2011).

[38] M.S.D. Erosa, T.I.S. Medina, R.N. Mendoza, M.A. Rodriguez and E. Guibal,

Hydrometallurgy 61, 157 (2001).

[39] S. Hasan, A. Krishnaiah, T.K. Ghosh and D.S. Viswanath, Ind. Eng. Chem. Res., 45,

5066 (2006).

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

292

[40] N. Sankararamakrishnan, A.K. Sharma and R. Sanghi, J. Hazard. Mater. 148, 353

(2007).

[41] N. Sankararamakrishnan, P. Kumar and V.S. Chauhan, Sep. Purif. Technol. 63, 213

(2008).

[42] E.C.D. S. Filho, P.D.R. Monteiro, K. S. Sousa and C. Airoldi, J. Therm. Anal. Calorim.

106, 369 (2011).

[43] G.J. Copello, F. Varela, R.M. Vivot and L.E. Diaz, Bioresour. Technol. 99, 6538

(2008).

[44] M. Li, X. Zhu, D. Lin, W. Chen and G. Ren, Environ. Eng. Sci. 28, 735 (2011).

[45] T.W. Clarkson, Environ. Health Perspect. 100, 31, (1992).

[46] M. Benavente, Adsorption of metallic ions onto chitosan: equilibrium and kinetic

studies, Licentiate Thesis, Royal Institute of Technology, Department of Chemical

Engineering and Technology, Stockholm, Sweden (2008).

[47] Shafaei, F.Z. Ashtiani and T. Kaghazchi, Chem. Eng. J. 133, 311 (2007).

[48] Gamage and F. Shahidi, Food Chem. 104, 989 (2007).

[49] E. Taboada, G. Cabrera and G. Cardenas, J. Chile Chem. Soc. 48, 7 (2003).

[50] E. Lopes, F. dos Anjos, E. Vieira and A. Cestari, J. Colloid Interf. Sci. 263, 542 (2003).

[51] Peniche-Covas, L.W. Alvarez and W. Arguelles-Monal, J. Appl. Polym. Sci. 46, 1147

(1992).

[52] Jeon and W.H.H. Holl, Water Res. 37, 4770 (2003).

[53] J.C.Y. Ng, W.H. Cheung and G. McKay, Chemosphere 52, 1021 (2003).

[54] J.R. Rangel-Mendez, R. Monroy-Zepeda, E. Leyva-Ramos, P.E. Diaz-Flores and K.

Shirai, J. Hazard. Mater. 162, 503 (2009).

[55] Asandei, L. Bulgariu and E. Bobu, Cellulose Chem. Technol. 43, 211 (2009).

[56] Chauhan and N. Sankararamakrishnan, Bioresour. Technol. 99, 9021 (2008).

[57] W.L. Yan and R. Bai, Water Res. 39, 688 (2005).

[58] L. Jin and R. Bai, Langmuir 18, 9765 (2002).

[59] N. Li and R. Bai, Ind. Eng. Chem. Res. 45, 7897 (2006).

[60] Rojas, J. Silva, J.A. Flores, A. Rodriguez, M. Ly and H. Maldonado, Sep. Purif.

Technol. 44, 31 (2005).

[61] T. N. de Castro Dantas, A. A. Dantas Neto, M. C. P. de A. Moura, E. L. Barros Neto,

and E. de Paiva Telemaco, Langmuir 17, 4256 (2001).

[62] P. Baroni, R.S. Vieira, E. Meneghetti, M.G.C. da Silva and M.M. Beppu, J. Hazard.

Mater. 152, 1155 (2008).

[63] R. Schmuhl, H.M. Krieg and K. Keizer, Water SA 27, 1 (2001).

[64] V. A. Spinelli, M.C.M. Laranjeira and V.T. Fávere, React. Funct. Polym. 61, 347

(2004).

[65] N.Sankararamakrishnan, A. Dixit, L. Iyengar and R. Sanghi, Bioresour. Technol. 97,

2377 (2006).

[66] V.M. Boddu, K. Abburi, J.L. Talbott and E.D. Smith, Environ. Sci. Technol. 37, 4449

(2003).

[67] S. Hasan, A. Krishnaiah, T.K. Ghosh, D.S. Viswanath, V.M. Boddu and E.D. Smith,

Sep. Sci. Technol. 38, 3775 (2003).

[68] V.M. Boddu, K. Abburi, J.L. Talbott, E.D. Smith and R. Haasch, Water Res. 42, 633

(2008).

[69] Gupta, V.S. Chauhan and N. Sankararamakrishnan, Water Res 43, 3862 (2009).

Nova Science Publishers, Inc.

Chitosan Biopolymer for Water Quality Improvement 293

[70] Gérente, Y. Andrès, G. McKay and P. Le Cloirec, Chem. Eng. J. 158, 593 (2010).

[71] H.H. dos Santos, C.A. Demarchi, C.A. Rodrigues, J.M. Greneche, N. Nedelko and A.

Slawska-Waniewska, Chemosphere 82, 278 (2011).

[72] S.M. Miller, M.L. Spaulding and J.B. Zimmerman, Water Res. 45, 5754 (2011).

[73] Gupta, M. Yunus and N. Sankararamakrishnan, Chemosphere 86, 150 (2011)

[74] D.D. Gang, B. Deng and L.S. Lin, J. Hazard. Mater. 182, 156 (2011).

[75] X. Wei, S. Bhojappa, L.-S. Lin and R.C. Viadero, Environ. Eng. Sci. (2011) (in press).

[76] N. Bleiman and Y.G. Mishael, J. Hazard. Mater. 183, 590 (2010).

[77] W. Yang, H. Chi, B. Sun, H. Zhao and Z. Wei, J. Shenyang Jianzhu Univ. 26, 744

(2010).

[78] Sabarudin, K. Oshita, M. Oshima and S. Motomizu, Anal. Chim. Acta 542, 207 (2005).

[79] S. Chatterjee and S.H. Woo, J. Hazard. Mater. 164, 1012 (2009).

[80] S. Chatterjee, D.S. Lee, M.W. Lee , S.H. Woo, J. Hazard. Mater. 166, 508 (2009).

[81] M. Arora, N.K. Eddy, K.A. Mumford, Y. Baba, J.M. Perera and G.W. Stevens, Cold

Reg. Sci. Technol. 62, 92 (2010).

[82] K. Jaafari, S. Elmaleh, J. Coma and K. Benkhouja, Water SA 27, 9 (2004).

[83] J. Dai, H. Yang, H. Yan, Y. Shangguan, Q. Zheng and R. Cheng, Chem. Eng. J. 166,

970 (2011).

[84] S. Fierro, M. del, Pilar Sánchez-Saavedra and C. Copalcúa, Bioresour. Technol. 99,

1274 (2008).

[85] Y.-C. Chung, Y.-H. Li and C.C. Chen, J. Environ. Sci. Health A 40, 1775 (2005).

[86] Y. Shimizu, S. Nakamura, Y. Saito and T. Nakamura, J. Appl. Polym. Sci. 107, 1578

(20080.

[87] T. Fagundes, E.L. Bernardi and C.A. Rodrigues, J. Liq. Chromatogr. Related Technol.

24, 1189 (2006).

[88] Y. Zheng and A. Wang, Adsorpt. Sci. Technol. 28, 89 (2010).

[89] Lee, C. Chen, S.T. Yang, W.S. Ahn, Micropor. Mesopor. Mater. 127, 152 (2010).

[90] D. Thakre, S. Jagtap, A. Bansiwal, N. Labhsetwar and S. Rayalu, J. Fluo. Chem. 131,

373 (2010).

[91] S. Jagtap, M.K.N. Yenkie, S. Das and S. Rayalu, Desalination 273, 267 (2011).

[92] N. Viswanathan and S. Meenakshi, Appl. Clay Sci. 48, 607 (2010).

[93] N. Viswanathan and S. Meenakshi, J. Hazard. Mater. 178, 226 (2010).

[94] S. Jagtap, M.K.N Yenkie, N. Labhsetwar and S. Rayalu, Microporous Mesoporous

Mater. 142, 454 (2010).

[95] Y. Vijaya, S.R. Popuri, G.S. Reddy and A. Krishnaiah, Desalin. Water Treat. 25, 159

(2011).

[96] Y. Xie, S. Li, F. Wang and G. Liu, Chem. Eng. J. 156, 56 (2010).

[97] F.X. Li, L. Squartsoff and S.H. Lamm, J. Occup. Environ. Med. 43, 630 (2001)

[98] B.E. Logan, Environ. Sci. Technol. 35, 482A (2001).

[99] Y. Zheng and A. Wang, J. Hazard. Mater. 171, 671 (2009).

[100] Y. Zheng and A. Wang, Chem. Eng. J. 171, 1201 (2011).

[101] Y. Zheng and A. Wang, Ind. Eng. Chem. Res. 49, 6034 (2010).

[102] K. Ghosh and M. Schnitzer, Soil Sci. 129, 266 (1980).

[103] P.K. Cornel, R.S. Summers and P.V. Roberts, J. Colloid Interface Sci. 110, 149 (1986).

[104] S. Maghsoodloo, B. Noroozi, A.K. Haghi and G.A. Sorial, J. Hazard. Mater. 191, 380

(2011).

Nova Science Publishers, Inc.

Xinchao Wei and F. Andrew Wolfe

294

[105] W.S.W. Ngah, S. Fatinathan and N.A. Yosop, Desalination 272, 293 (2011).

[106] W.S.W. Ngah, M.A.K.M Hanafiah and S.S. Yong, Colloids Surf. B 65, 18 (2008).

[107] X. Zhang and R. Bai, J. Colloid Interface Sci. 264, 30 (2003).

Nova Science Publishers, Inc.

INDEX

Abraham, 205

abstraction, 112

access, 128, 162, 163, 164, 168, 170, 173, 179, 180,

188, 202, 203

accessibility, 275

accountability, 196

accounting, 17, 268

acid, 59, 102, 113, 118, 135, 144, 162, 184, 213,

224, 227, 231, 232, 236, 238, 248, 250, 260, 262,

263, 266, 274, 277, 278, 280, 284, 286, 289, 290

acidic, 74, 75, 131, 141, 212, 274, 275, 280, 285, 290

acrylic acid, 280, 286, 288

activated carbon, 51, 52, 273, 274, 281, 284, 288,

289, 290

active site, 275, 281

AD, 128

adaptation, 217, 220, 235, 247, 249

adhesion, 244, 253

adipose, 245

adrenal gland, 270

adrenal glands, 270

adsorption, 63, 64, 67, 68, 69, 70, 71, 72, 88, 89, 91,

92, 93, 96, 110, 111, 115, 121, 131, 216, 219,

273, 274, 275, 276, 277, 278, 279, 280, 281, 282,

284, 285, 286, 287, 288, 289, 290

Adsorption, 290

advancement, 273

adverse effects, 82, 193, 276

advocacy, 164

aerobic bacteria, 223

aesthetic(s), 49, 50, 174, 175, 183, 192

Africa, 108, 161, 162, 163, 172, 188, 194, 202, 203,

205, 206

agar, 224, 225

age, 157

agencies, 201, 273

aggregation, 4, 5

agriculture, 22, 56, 108, 110, 123, 132, 143, 165,

181, 182, 188, 192, 201, 235

alcoholism, 262

alcohols, 237

algae, 52, 54, 55, 59, 193, 214, 216, 241, 285

algorithm, 47, 125

aliphatic compounds, 223, 235

alkalinity, 48, 54, 55, 102, 117, 263, 272

aluminium, 193

amine, 276, 277, 278, 279, 280, 283, 284, 290

amine group, 276, 278, 279, 280, 284

amino, 142, 273, 274, 275, 276, 290

amino acid, 142

amino groups, 273, 274, 275, 290

aminoglycosides, 227, 230

ammonia, 23, 26, 53, 65, 67, 68, 83, 142, 175, 183,

287

ammonium, 88, 175, 281, 286, 288

amplitude, 1, 3, 4, 5, 6, 13, 14, 15, 16, 17, 18, 20, 21,

24, 25, 26, 27, 28, 247

anabolism, 248

anaerobe, 223

anaerobic bacteria, 223

anaerobic digestion, 220

anemia, 247, 251, 253, 254, 255, 260, 270, 288

animal husbandry, 217

antagonism, 260

antibiotic, 209, 211, 215, 216, 227, 228, 229, 230,

231, 232, 233, 235, 236, 238

antibiotic resistance, 211, 216, 227, 228, 229, 231,

232, 233, 235, 236, 238

anus, 243

aquaculture, 120, 204, 286

aquaria, 242

aquatic life, 256, 283

aquatic systems, 211

aqueous solutions, 230, 277, 279, 280, 281, 287, 288

aquifers, 100, 102, 107, 108, 114, 143, 155, 192, 206

Nova Science Publishers, Inc.

Index

296

arithmetic, 3, 25, 140

aromatic hydrocarbons, 223

arsenic, 107, 184, 282

Asia, 204

assessment, 1, 2, 3, 5, 6, 13, 22, 31, 32, 102, 108,

120, 127, 130, 157, 221, 231, 232, 239, 240, 241,

242, 249, 254, 255, 256, 261, 262

assessment tools, 1, 2, 5

atmosphere, 154, 250, 260, 268

ATP, 248

atrophy, 246

attachment, 185, 244, 290

attitudes, 164, 165, 167, 168, 169, 189, 197

audit, 173

authority(s), 164, 173, 174, 177, 180, 188, 192, 195,

196, 198, 202, 203, 211

automate, 11

autopsy, 266

awareness, 163, 189

bacteria, 68, 83, 183, 193, 196, 209, 212, 213, 214,

215, 216, 223, 224, 225, 227, 228, 229, 231, 232,

234, 235, 237, 241, 286

bacterial cells, 224

bacterial strains, 222

bacteriostatic, 212, 213

bacterium, 215, 223, 238

barium, 54

barriers, 277

base, 94, 95, 118, 164, 198, 284

baths, 227, 276

Belgium, 129, 205

beneficiaries, 197

benefits, 55, 57, 173, 177, 196

benzene, 230

bias, 7, 8, 12, 14, 16, 17, 19, 20, 27, 28, 32, 137

bicarbonate, 117, 121, 122, 151, 152

bioaccumulation, 230, 262, 266, 276, 279, 283

bioassay, 254

biochemical processes, 65, 242

biodegradation, 211, 215, 220, 221, 222, 223, 224,

225, 230, 232, 233, 235

biodiversity, 228, 237

biological processes, 64, 217, 239, 240, 274, 284

biomass, 66, 67, 68, 218

biopolymer(s), 273, 274, 276, 287

bioremediation, 230

biosynthesis, 213

biotic, 17

births, 262

bis-phenol, 214

blood, 245, 246, 247, 248, 249, 251, 253, 254, 264,

268, 270, 271

blood smear, 247

blood vessels, 245, 251, 268

bone(s), 244, 245, 260, 280

boreholes, 113, 135, 184

bottom-up, 195, 196

bounds, 39, 48

brain, 279

Brazil, 203

breakdown, 215, 224

breast milk, 217, 231

by-products, 222, 224, 225

Ca2+, 100, 102, 103, 104, 106, 109, 110, 117, 119,

120, 121, 135, 142, 150

cadmium, 193, 251

calcium, 115, 117, 121, 122, 123, 135, 151, 152,

248, 287

calcium, 115, 152

calculus, 236

calibration, 22, 88, 89, 90, 91, 94, 95, 113

cancer, 32, 57, 142

capillary, 268

carbon, 51, 54, 65, 66, 68, 69, 183, 218, 222, 223,

224, 238, 270, 274, 290

carbon dioxide, 65, 270

carboxyl, 276, 289

carboxylic groups, 275

cardiovascular disease(s), 270

cartilage, 245

case study(s), 1, 5, 21, 34, 107, 130, 196, 197, 198,

204, 247, 254, 255, 256

Caspian Sea, 245, 246

catabolism, 225, 249

catchments, 162, 171, 201

category a, 22, 25, 28

cation, 100, 116, 117, 118, 121, 123, 135, 142, 150,

277, 281

cell division, 247

cell membranes, 213

cell phones, 202

cellulose, 135, 273, 274

Central Europe, 157

ceramic, 282

cerebrovascular disease, 58

cerebrum, 244

challenges, 56, 198, 205, 229

chemical characteristics, 114, 128

chemical industry, 276

chemical properties, 275, 290

Nova Science Publishers, Inc.

Index 297

chemical reactivity, 276

children, 194, 284

Chile, 292

China, 128, 216

chitin, 273, 274, 275

chitosan, vi, ix, 273, 274, 275, 276, 277, 278, 279,

280, 281, 282, 284, 285, 286, 287, 288, 289, 290,

292

chloride, 284

chlorination, 217, 221

chlorine, 54, 173, 174, 192, 193, 195, 199, 202, 230

chlorophyll, 23, 26, 66, 69, 83, 84, 88, 89, 90, 91, 94,

95, 96

cholelithiasis, 268, 270, 271

cholera, 163, 172, 193

cholic acid, 248

chopping, 211

chromatography, 135, 230, 235

circulation, 77, 154

cirrhosis, 248, 250, 252, 255, 264

cities, 210, 263, 265, 267

citizens, 164, 170, 174

City, 97, 100, 108, 271

classes, 110, 123, 264, 265, 271

classification, 29, 102, 121, 123, 131, 142, 148, 150,

151, 158, 176

clay minerals, 117, 123

Clean Water Act (CWA), 283

cleaning, 276

cleavage, 223, 224, 229, 231, 233

climate(s), 32, 56, 101, 110, 111, 121, 162, 229, 265

climate change, 162, 229

clinical examination, 262

closure, 75, 206

cluster analysis, 111, 123, 125, 126

clustering, 4, 125

clusters, 110

CO2, 117

coal, 184, 206, 278, 283, 286

coal tar, 206

cocoa, 134

collaboration, 194

color, 243, 245, 246, 262

combined effect, 139, 242

combustion, 278, 283

commercial, 132, 165, 166, 198, 274, 275, 290

communication, 162, 179, 191, 194, 195, 197, 200,

201, 202, 203

Community, 162

comparative analysis, ix, 239

competition, 280

competitive advantage, 225

complexity, 19, 47

compliance, 1, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 16, 17,

20, 21, 23, 24, 25, 26, 27, 28, 203

composition, 100, 106, 110, 114, 118, 129, 153, 155,

159, 213, 221, 237, 239, 242, 247, 262, 263, 268

compost, 233

compounds, 174, 193, 209, 210, 211, 223, 225, 228,

231, 233, 236, 237, 242, 245, 270

computation, 4, 19, 20, 73, 115, 119

computer, 123, 170

conception, 234

conceptual model, 191

conditioning, 285

conductivity, 135, 138, 142, 151, 263

conference, 205, 206

configuration, 246

conflict, 220

Congress, 97, 108, 203

connective tissue, 246

consensus, 201

conservation, 79, 110, 196

conserving, 164

constituents, 63, 64, 65, 68, 69, 76, 77, 79, 87, 89,

90, 95, 96, 99, 119, 150, 192

Constitution, 164, 174

construction, 232

consumers, 49, 50, 52, 171, 181, 183, 189

consumption, 139, 163, 170, 172, 174, 179, 180,

276, 282, 289

contact time, 213, 287, 288

containers, 185, 262

contaminant, 108, 243, 255, 274, 275, 280, 283, 288

contaminated water, 49, 142, 241, 249, 282

contamination, 49, 127, 133, 139, 140, 144, 145,

147, 162, 182, 184, 229, 239, 240, 242, 243, 244,

249, 250, 252, 254, 260, 278

Continental, 272

continuous function, 14, 16

contour, 114

cooking, 168, 174, 180

coordination, 197

copolymerization, 277, 288

copper, 184, 245, 250, 251, 252, 260, 261, 262, 270,

271, 276, 277, 285

cornea, 244, 245

correlation(s), 50, 115, 125, 131, 148, 154, 185, 186,

261, 268, 269, 270

correlation coefficient, 154, 269, 270

corrosion, 145, 193

cortex, 268

cortisol, 249

cosmetic(s), 210, 211

Nova Science Publishers, Inc.

Index

298

cost, 31, 32, 33, 34, 39, 40, 41, 42, 43, 44, 45, 46, 49,

50, 51, 52, 54, 55, 56, 57, 192, 193, 194, 211,

273, 274, 280, 290

cost curve, 51

Council of Ministers, 3, 29

covering, 134, 153, 168, 197, 270

cranium, 245

critical state, 244

critical value, 146

CRM, 137

crop(s), 121, 123, 134, 150, 151, 217

crop production, 217

crust, 150

crystalline, 135, 159, 244

crystallinity, 275

crystallization, 212

cultivation, 134, 222

cultural beliefs, 163

culture, 192, 201, 224, 286

culture medium, 224

cumulative distribution function, 16

customers, 51, 57

cyanotic, 245

cycles, 5, 14, 95, 240, 264, 277, 282, 288

cycling, 229

cytochrome, 234

danger, 165, 175, 183, 243

data analysis, 32, 107

data collection, 194

data set, 123, 125, 166, 173

database, 203, 206, 234

DDT, 59

deacetylation, 274, 275, 278, 280, 281

death rate, 268, 270

deaths, 163

decay, 81, 142, 143, 266

decentralisation, 179, 189

decision makers, 3, 5, 6, 96, 132, 200

decomposition, 65, 68, 69, 73, 76, 102, 183, 212,

246, 263

decomposition temperature, 212

deficit, 179

deformation, 243

degradation, 107, 219, 220, 221, 223, 224, 225, 226,

228, 229, 230, 233, 236, 237, 238, 240, 247

degradation process, 107

Dehalogenation, 223

Delta, 64, 82, 97

democracy, 161

demographic data, 268

demography, 110

dendrogram, 125

denitrification, 67, 68, 69, 284, 288

denitrifying, 229, 230, 233

Department of Agriculture, 206

Departments of Agriculture, 190

deposition, 151, 250, 260

deposits, 259, 260

depression, 111

deprivation, 57

depth, 66, 67, 74, 76, 77, 80, 81, 82, 83, 84, 89, 94,

104, 107, 113, 115

derivatives, 217, 223, 277, 286

desorption, 63, 64, 67, 68, 69, 70, 71, 72, 88, 89, 91,

92, 93, 96, 219, 276, 281, 282, 285

destruction, 242, 247, 249, 252, 266

detection, 58, 137, 224, 236, 247

detergents, 211, 235

developing countries, 164

deviation, 44, 102, 114, 136, 137, 141, 145, 155

diabetes, 270

diarrhea, 180

diffusion, 69, 76, 79, 278, 279, 285, 289

diffusivity, 74

digestion, 217, 221, 226, 264, 265

dimensionality, 102

dioxin, 217, 234

directors, 200

disaster, 192

discharges, 134, 182, 209

diseases, 163, 167, 172, 173, 180, 186, 188, 189,

193, 196, 239, 243, 244, 247, 248, 249, 250, 252,

260, 264, 265, 266, 267, 268, 269, 270, 271, 288

diseconomies of scale, 49

disinfection, 51, 120, 183, 193, 195, 289

dispersion, 10, 118, 121, 248, 286

dissatisfaction, 185

dissolved oxygen, 23, 29, 54, 63, 64, 65, 73, 74, 96

distilled water, 135

distribution, 7, 10, 14, 15, 16, 17, 19, 28, 32, 33, 34,

36, 37, 44, 45, 47, 56, 77, 79, 101, 115, 116, 117,

147, 171, 179, 185, 186, 189, 193, 233, 235, 260,

264, 269

District of Columbia, 158

diversity, 242

DNA, 225, 226

DOC, 183

DOI, 58, 157

dominance, 104, 127

dosage, 287

drainage, 101, 110, 121, 127, 162, 182, 285

drawing, 118

Drinking water, 57, 262

Nova Science Publishers, Inc.

Index 299

drought, 6, 110, 209

drug resistance, 231

drugs, 210

drying, 275

Dublin Convention, 164

dyes, 281

E.coli, 51

ecology, 55, 214, 228, 242

economic crisis, 260

economic development, 162

economic growth, 109

economies of scale, 49

ecosystem, 3, 65, 87, 228, 229, 237, 239, 242, 243,

249, 254, 255, 256, 257

Ecosystem Health Monitoring Program (EHMP), 3

ecotoxicological, 239, 240, 243, 254, 255

edema, 268

education, 172, 173, 174, 189, 191, 194, 197, 200,

201, 202, 203

effluent(s), 104, 143, 144, 147, 162, 185, 210, 215,

218, 219, 220, 229, 233, 235, 245, 259, 260, 277,

285, 286, 288

egg, 246

elaboration, 197

electrical conductivity, 135, 151, 202

electricity, 202

electrodeposition, 277

electron, 275, 289

electrophoresis, 225

electroplating, 278, 286

emergency, 49

emission, 262, 263, 264

emitters, 61

employment, 110

empowerment, 161, 162, 164, 177, 191, 196, 201

encoding, 229

endocrine, 210, 232, 248

endocrine system, 248

endocrine-disrupting chemicals, 232

endothermic, 277

energy, 69, 110, 137, 205, 215, 223, 238, 248, 249

energy consumption, 249

energy supply, 249

engineering, 205, 232, 286

environmental change, 243

environmental conditions, 56, 243

environmental degradation, 210

environmental effects, 242

environmental factors, 265

environmental issues, 163

environmental management, 165, 192, 195, 196, 256

environmental policy, 193, 194

environmental protection, 164

Environmental Protection Agency, 30, 64, 96, 283

environmental quality, 29, 163

enzyme(s), 215, 226, 227, 233, 234, 248

EPA, 51, 58, 283, 284, 288, 290

epidemic, 175, 227

epithelial cells, 251

epithelium, 245, 246, 248, 251, 253

equilibrium, 70, 71, 72, 73, 88, 92, 228, 260, 277,

278, 279, 281, 282, 285, 286, 287, 288, 289, 292

equipment, 175, 192, 193, 195

equity, 196

erythrocyte sedimentation rate, 247

ESR, 247

ethers, 211

ethylene, 278

etiology, 249, 266, 268, 271

Europe, 211, 216

evaporation, 105, 132, 135, 154, 155

evidence, 32, 58, 107, 132, 144, 155, 192, 195, 227,

231, 234

evolution, 100, 107, 108, 128, 225, 233

exchange rate, 248

excretion, 210

exercise, 168, 186, 188

exertion, 65

experimental condition, 137, 278, 282, 287

exploitation, 110, 128, 132

explosives, 288

exposure, 180, 183, 185, 210, 212, 214, 216, 223,

227, 230, 233, 234, 283

extraction, 112, 225, 226, 286

extracts, 230, 234

exudate, 244

facies, 99, 107, 127

factor analysis, 4, 106

families, 197

farm land, 94

farm size, 198

farmers, 178, 197, 198, 204

farmland, 83

fat, 212, 244, 251, 266

fatty acids, 248

fauna, 217

feelings, 167, 178

fertilization, 198

fertilizers, 142, 144, 162, 181, 288

fetus, 262

Nova Science Publishers, Inc.

Index

300

fiber(s), 245, 273, 274, 275

films, 278, 290

filters, 135, 218, 280

filtration, 195, 288

financial, 2

financial budgetary constraints, 2

finite element method, 78

fish, 17, 83, 94, 134, 217, 235, 239, 240, 241, 243,

244, 245, 246, 249, 250, 251, 252, 253, 254, 255,

256, 257, 272

fish diseases, 250, 254, 255

fisheries, 204

fixation, 283

flame, 135

floods, 17

flora, 228

flow field, 75, 87, 90

fluid, 76

fluorine, 118

fluoroquinolones, 227

food, 172, 185, 186, 193, 242, 261, 268, 274, 276,

283

food chain, 242, 283

food poisoning, 186

force, 76, 77, 78, 87, 288

forecasting, 254

formation, 117, 150, 158, 183, 193, 217, 230, 233,

234, 239, 240, 246, 248, 266, 275, 279, 280

formula, 4, 42, 69, 71, 74, 77, 80, 95, 121, 139, 168,

247

France, 216

freshwater, 3, 5, 13, 100, 110, 209, 210, 229, 239,

242, 283

FRP, 15, 18, 23, 26

FTIR, 277, 279

funds, 193

fungi, 222

fungus, 222

Gaussian random variables, 58

gel, 225, 275, 285

genes, 210, 225, 227, 228

genetics, 213

genotyping, 236

genus, 224

Geochemistry, 108, 157

geology, 55, 135, 145, 184, 188

germanium, 136

Germany, 136, 219

gill, 245, 253

gingivitis, 232, 236

GIS, 108, 191, 194

glomerulus, 252

glucose, 274, 276

glycogen, 249

gonads, 244, 246, 250

governance, 164, 196

government funds, 196

grades, 28, 242

granules, 280, 289

granulomas, 251

graph, 125, 154, 155

graphite, 263

grass, 177, 199

grassroots, 163

grazing, 67

Great Britain, 128, 263

Greece, 108

Groundwater salinization, 108

Groundwater, 99, 107

grouping, 55, 56, 147

growth, 63, 64, 65, 66, 67, 68, 69, 83, 89, 110, 132,

162, 163, 193, 210, 212, 213, 214, 215, 223, 224,

225, 232, 236, 246, 257, 265

growth rate, 66, 89, 225

guidance, 229

guidelines, 2, 3, 6, 10, 11, 13, 14, 15, 18, 19, 28, 29,

30, 140, 145, 157, 179

Guinea, 134

habitat, 240, 249

halogen, 223

hardness, 54, 55, 121

hazardous substances, 242, 261

hazards, 32, 262

health care, 50, 211

health care costs, 50

health education, 194

health effects, 120, 175

health risks, 58

health services, 188, 189

health status, 203

health/environmental guidelines, 2

heavy metals, 54, 131, 139, 140, 144, 249, 250, 252,

255, 260, 261, 262, 265, 266, 268, 270, 273, 276,

277, 278

height, 277

hematology, 247

hemoglobin, 247, 249, 253

hemorrhage, 251

hepatitis, 262

hepatocytes, 252, 266

Nova Science Publishers, Inc.

Index 301

hermaphrodite, 248

heterogeneity, 249

high density polyethylene, 135

histological examination, 262

histology, 248

historical data, 14, 17, 19, 20, 31, 48, 251

history, 46, 163, 211

HM, 266

Holocene, 129

homes, 216

homogeneity, 242

host, 215, 225, 229, 276

housing, 134, 157

HPLC-UV, 223, 224

human, 108, 139, 163, 170, 175, 181, 182, 185, 188,

192, 193, 210, 214, 217, 227, 231, 233, 237, 239,

240, 242, 254, 256, 260, 261, 266, 268, 269, 270,

271, 276, 282, 284, 286, 290

human health, 175, 185, 193, 210, 240, 256, 260,

266, 276, 282, 284, 286, 290

human organisms, 266, 271

humidity, 134, 153

hybrid, 286

hydrogen, 102, 154, 193, 263, 279

hydrogen peroxide, 263

hydrolysis, 68, 223, 224

hydrophilicity, 276

hydroxide, 193

hydroxyl, 222, 224, 273, 274, 276, 289, 290

hydroxyl groups, 273, 274, 276, 289

hygiene, 111, 172, 173, 174, 184, 186, 194, 201,

203, 212, 213

hyperemia, 243, 248

hyperplasia, 246, 251, 253

hypertension, 270

hypoplasia, 251

hypothalamus, 248

ICS, 102, 135

ideal, 46, 138, 141, 168

identification, 107, 128, 224, 261

images, 289

immobilization, 278, 280

immunity, 270

immunocompromised, 227, 231

impact assessment, 29

impregnation, 283, 287

imprinting, 275

in transition, 172, 188, 189

in vitro, 234

inappropriate sampling frequency, 2

incidence, 163, 268, 270

income, 49, 50, 196

independence, 28, 38

indexing, 4

India, 107, 108, 128, 129, 130, 158, 164, 202

indigenous knowledge, 162, 163, 164, 165, 167, 168,

173, 189, 192, 201

individual perception, 195, 201

individuals, 168, 200, 210, 243, 244, 247, 248, 250

Indonesia, 198, 204

inducer, 216

inducible enzyme, 223

industry(s), 22, 110, 162, 178, 206, 210, 259, 260,

263, 268, 276, 277

infants, 193

infection, 227, 229

inflammation, 268

infrastructure, 51, 197

ingestion, 58

inhibition, 213, 215, 232, 233, 236

inhibitor, 213

inoculum, 218

institutions, 177, 192, 201

integration, 155

integrity, 17, 242, 243

integument, 251

intensive care unit, 237

interaction process, 96

interface, 64, 65, 73, 74, 76, 89, 96, 100, 104, 107,

173

interference, 280, 282

internal rate of return, 198

interstitial nephritis, 248

intoxication, 244, 245, 246, 247, 251, 255, 266

invertebrates, 241

investment, 3

iodine, 288

ion-exchange, 277

ions, 99, 100, 102, 109, 115, 116, 117, 123, 125,

132, 144, 175, 277, 279, 280, 281, 283, 286, 288,

292

Iran, 128, 129, 205

iron, 184, 193, 230, 233, 234, 260, 262, 282, 284,

286

irradiation, 136, 137

irrigation, 109, 110, 111, 113, 115, 120, 121, 122,

123, 124, 127, 129, 130, 132, 133, 145, 150, 151,

152, 155, 156, 158, 162, 203, 210, 283

irrigation, vii, 110, 113, 123, 150, 152, 155

Islam, 157

isolation, 224, 226

isoniazid, 235

isotherms, 72

Nova Science Publishers, Inc.

Index

302

isotope, 108, 131, 135, 136, 137, 155, 156, 159

issues, 1, 2, 3, 50, 110, 164, 165, 176, 177, 178, 185,

189, 195, 196, 201, 203, 209, 229, 284

Italy, 107, 235

Japan, v, ix, 99, 100, 101, 103, 106, 108, 165, 216,

232, 272

Jordan, 127

K+, 100, 102, 103, 104, 106, 109, 110, 119, 120, 122,

135, 142

Kenya, 207

kidney(s), 244, 246, 247, 248, 251, 252, 259, 262,

266, 267, 268, 269, 270, 271, 280

kill, 213

kinetic studies, 292

kinetics, 64, 65, 68, 70, 213, 283, 287, 288

Kola Peninsula, 245, 247, 260, 269

Korea, 107

laboratory studies, 227, 241

Lactobacillus, 214

lakes, 63, 64, 78, 83, 95, 96, 192, 240, 244, 245, 246,

250, 251, 259, 260, 261, 264, 265, 266, 272

land tenure, 198

landscapes, 165

languages, 163

lanthanum, 287

lateral motion, 101

laws, 65

LC-MS, 224

leaching, 144, 260, 263, 264, 268, 270

lead, 49, 155, 175, 177, 193, 196, 200, 239, 247,

251, 259, 260, 275, 280, 282, 284, 285

leadership, 167, 174, 190, 200

leakage, 278

legislation, 174

lens, 244

leukemia, 270

leukocytes, 247

lifetime, 176

light, 18, 63, 64, 66, 69, 83, 84, 85, 87, 88, 89, 95,

96, 217, 246, 283

limestone, 113, 117

Limpopo, 163, 165, 171, 172, 179

linear function, 4, 9, 12, 17, 19

lipids, 249

lipoid, 245, 246, 248, 251, 252

liquid phase, 221

Listeria monocytogenes, 227, 230

liver, 244, 247, 248, 250, 251, 252, 255, 259, 262,

266, 267, 269, 270, 271

liver disease, 270

livestock, 56, 134, 183, 188, 210

lobbying, 194

local community, 164

local government, 178, 198, 199, 200

logging, 150

Louisiana, 230

Luo, 238

Luvuvhu Catchment, vi, ix, 161, 192

lying, 41

lymphocytes, 247

lysis, 246

Mackintosh, 165, 195, 205, 206

magnesium, 58, 113, 115, 117, 122, 123, 135, 151,

152

magnesium, 115, 151

magnetic particles, 290

magnitude, 6, 13, 167, 270

major decisions, 178

majority, 109, 127, 153, 198

malaria, 172, 233

man, 228, 243, 273, 288

manganese, 23, 175, 284

manpower, 171, 203

manufacturing, 217, 281, 286

manure, 233

marine environment, 210

Maryland, 29

mass, 57, 71, 76, 78, 79, 102, 136, 220, 221, 224,

228, 230, 235, 255

mass spectrometry, 102, 136, 230, 235

materials, 68, 106, 132, 136, 142, 144, 173, 184,

190, 197, 273, 274, 276

matrix, 30, 44, 45, 100, 125, 148, 149, 287, 288

matter, 52, 157, 174, 175, 182, 183, 217

maximum sorption, 276, 277

measurement(s), 2, 3, 4, 16, 32, 33, 38, 39, 40, 42,

46, 47, 54, 56, 63, 69, 70, 72, 73, 74, 77, 80, 83,

84, 87, 89, 95, 102, 121, 135, 236, 243, 255

media, 207, 214, 275

median, 7, 14, 15, 19, 114

medical, 197, 210, 214, 268

medicine, 243

Mediterranean, 108, 110, 111

Nova Science Publishers, Inc.

Index 303

medulla, 268

membranes, 244, 275, 284, 290

mental retardation, 288

Mercury, 59, 263, 278

Metabolic, 232

metabolism, 222, 229, 230, 237, 248, 260, 288

metabolites, 222

metabolized, 222, 223

metal ion(s), 175, 280

metallurgy, 260, 277

metals, 54, 60, 109, 113, 135, 136, 141, 148, 149,

175, 184, 193, 245, 250, 251, 252, 253, 259, 260,

261, 262, 263, 264, 266, 268, 269, 270, 271, 274,

275, 276, 279, 290

meter, 84, 104, 107, 135, 193, 263

methodology, 1, 6, 12, 17, 24, 56, 119, 197

methyl group, 223

methylation, 221, 235

Mexico, 108, 197, 204

Mg2+, 100, 102, 103, 104, 106, 109, 110, 119, 120,

121, 135, 142, 150, 262

microbial communities, 229

microbial community, 228

microorganism(s), 210, 215, 216, 217, 218, 220, 228

Microsoft, 137

microspheres, 290

migration, 100, 263, 270, 279

mineral water, 128

mineralization, 67, 68, 110, 125, 266, 268, 270

Miocene, 111

Missing values, 19

mixing, 75, 81, 100, 105, 106, 107, 114, 155

mobile phone, 202

modelling, 17, 19, 20, 31, 32, 34, 46, 47, 50, 56, 238

models, 10, 33, 34, 35, 38, 46, 47, 48, 56, 57, 58, 64,

70, 73, 75, 92, 93

modern science, 192

modernization, 250, 260

modifications, 21, 279

moisture, 154, 155

mole, 274

molecular structure, 211

molecular weight, 211, 226, 230, 274, 275, 279, 281,

286

momentum, 76

monitoring design, 1, 2, 5, 20, 26

monolayer, 276, 282, 284, 285, 287, 288

Moon, 223, 234

morbidity, 241, 244, 249, 251, 253, 259, 261, 262,

264, 266, 268, 269, 271

morphology, 83

mortality, 89, 265

mortality rate, 89, 265

Moscow, 256, 262, 272

MR, 129

mucus, 243

multivariate analysis, 123

muscles, 244, 245, 280

muscular tissue, 248

mutant, 230

mutation(s), 215, 227

myocardium, 245, 248

myopathy, 248

Na+, 100, 102, 103, 104, 106, 109, 110, 119, 120,

121, 122, 123, 135, 142, 150, 262

NaCl, 279, 281

NAD, 213, 233

NADH, 213

nanoparticles, 275, 277, 283, 290

natural disaster, 205

natural resource management, 192, 201

natural resources, 2, 164, 165, 197

nausea, 193

necrosis, 244, 246, 248, 252

negative effects, 176, 289

Nepal, 204

nephrocalcinosis, 248, 249, 255

nephrolithiasis, 270

nephropathy, 246

Netherlands, 96

neutral, 115, 117, 131, 141, 144, 212, 280, 290

New South Wales, 31

New Zealand, 57, 129, 234

NGOs, 177, 179

NH2, 273, 274, 277, 279

nickel, 184, 245, 250, 251, 252, 254, 260, 261, 262,

270, 271

Nigeria, 129, 130, 202

nitrates, 181, 188, 207

nitrification, 67, 68, 142, 218, 288

nitrite, 57, 142, 181, 288

nitrogen, 23, 26, 53, 54, 64, 65, 66, 67, 68, 88, 89,

90, 91, 95, 96, 143, 182, 218, 275, 288

nitrogen fixation, 143

nitrogen gas, 67, 69

nodes, 80, 83

normal distribution, 37

North Africa, 109, 110

North America, 195

nuclear magnetic resonance, 224

nuclei, 212, 246, 247

nucleus, 212

Nova Science Publishers, Inc.

Index

304

nutrients, 29, 63, 64, 65, 66, 69, 70, 71, 72, 73, 74,

76, 83, 88, 89, 92, 93, 94, 95, 96, 162, 182, 184,

213, 217, 218, 220, 225, 250, 260, 263, 281, 288

nutrient concentrations, 70, 83, 93

nutrient enrichment, 162

nutrition, 283

obesity, 245

oceans, 155

oedema, 251

officials, 167, 168

OH, 277, 279, 290

oil, 245, 283

oligotrophic conditions, 263

operating costs, 33, 39, 41

operations, 263

opportunities, 57, 176

opportunity costs, 196

optimization, 195, 221

oral cavity, 243

ores, 262, 263, 270

organ, 243, 245, 248, 252

organic compounds, 65, 218, 222, 223, 224

organic matter, 102, 115, 143, 233, 263

organic solvents, 212

organism, 234, 240, 242, 244, 247, 248, 249, 260,

261, 268, 270, 271

organs, 239, 243, 246, 248, 261, 262, 264, 265, 266,

269, 270, 271

ossification, 245

osteoporosis, 248, 270

ovaries, 246, 248

overtime, 50

oxidation, 68, 118, 144, 223, 228, 229, 230, 233,

234, 238, 263, 282, 284, 288

oxygen, 26, 65, 67, 68, 69, 73, 76, 136, 154, 159,

218

pain, 101

pancreatitis, 204

parasites, 175, 185

parenchyma, 246, 251, 252

participants, 174

partition, 92, 93

pathogenesis, 266, 267, 271

pathogens, 51, 54, 183, 185, 193, 205, 227, 232

pathologist, 262

pathology, 248, 251, 270

pathways, 214, 215, 221, 223, 225, 228, 232, 260

PCA, 110, 111, 115, 125

PCR, 234, 236

penetrability, 264

percentile, 3, 13, 15, 16, 18, 19

perchlorate, 58, 273, 288

permeability, 121, 123, 131, 133, 150, 152, 268

permission, 241

permit, 229

Perth, 98

pesticide, 51, 198

PET, 289

pharmaceutical(s), 209, 233, 234, 235, 237

phase inversion, 278, 289

phenol, 224, 225, 229, 234

Philadelphia, 97, 127, 256

Philippines, 196

phosphate(s), 65, 68, 72, 73, 88, 92, 93, 115, 263,

273, 282, 283, 284, 285, 286

phosphorous, 182, 184

phosphorus, 23, 26, 64, 65, 66, 68, 72, 74, 75, 83, 88,

89, 90, 91, 92, 95, 96, 97, 218, 263, 288

phosphorylation, 205

photodegradation, 234

photolysis, 236

photosynthesis, 65, 68, 69

physical properties, 121

physicochemical characteristics, 290

physico-chemical parameters, 102, 113, 119, 141

Physiological, 243

physiology, 242

phytoplankton, 63, 64, 65, 66, 67, 68, 69, 83, 84, 89,

96, 206, 242

pipeline, 264

Piper, 118, 129

plants, 68, 121, 127, 142, 161, 166, 167, 168, 169,

171, 174, 179, 184, 185, 186, 190, 199, 210, 214,

215, 216, 217, 218, 221, 228, 231, 234, 236, 245,

260, 261, 262, 263, 268, 274, 276, 288

plaque, 212, 232, 234, 237

plasmid, 225, 226, 231

plasmid DNA, 225

plastics, 211

playing, 155

Pliocene, 100

PM, 84

polar, 233

polarity, 219, 220

policy, 164, 167, 170, 174, 188, 189, 190, 191, 200

policy makers, 200

pollutants, 63, 82, 143, 144, 161, 209, 225, 240, 241,

242, 243, 244, 247, 248, 250, 256, 260, 261, 266,

268, 270, 286

Nova Science Publishers, Inc.

Index 305

Pollution, 238, 272

polycyclic aromatic hydrocarbon, 233

polyethyleneterephthalate, 289

polymer, 274, 275, 276, 278, 290

polymer chain, 276

polymerization, 286

polythene, 113

polyvinyl chloride, 216, 276

ponds, 183

pools, 67, 186

population, 110, 134, 157, 162, 163, 168, 172, 179,

210, 212, 215, 225, 234, 242, 243, 255, 259, 261,

262, 263, 264, 265, 266, 268, 269, 270, 271

population growth, 162

porosity, 74, 275

Portugal, 123, 128

potassium, 113, 115, 117, 127, 277

potassium persulfate, 277

potassium, 115

power generation, 286

power plants, 260

precipitation, 99, 101, 105, 118, 132, 143, 152, 154,

263, 266, 276, 277, 278, 280, 281, 282, 285, 287

preparation, 226

preservation, 165, 240

principal component analysis, 4, 125, 127

principles, 164, 195, 196, 242

private sector, 196

probability, 1, 5, 14, 15, 25, 33, 34, 35, 36, 37, 38,

39, 40, 49, 195

probability density function, 34

probability distribution, 34

production costs, 198

professionals, 173, 176

profit, 189

project, 17, 82, 130, 164, 194, 196, 197, 198

proliferation, 213, 246

propagation, 264, 271

proportionality, 138

protection, 2, 29, 150, 164, 177, 273

protein synthesis, 249

proteins, 226, 249

protons, 284

Pseudomonas aeruginosa, 227, 230, 231, 232, 233,

236

public health, 2, 6, 28, 57, 174, 183, 193, 194, 195,

201, 217, 273

pulp, 245, 276

pumps, 145, 227, 230

pure water, 66, 69

purification, 216

purity, 136

PVA, 280

PVC, 216, 276

pyrite, 184

quality control, 173

quality of service, 188

quality standards, 31

quantification, 224

quantitative research, 170

quartz, 284

quaternary ammonium, 281

Queensland, 1, 3, 6, 7, 22, 29, 30

questionnaire, 167

radiation, 84

radicals, 274

radio, 202, 220

rainfall, 17, 22, 56, 111, 115, 132, 134, 153, 154,

155, 156, 163

Random sampling, 169

rating scale, 119, 140

RDP, 179

reaction rate, 70

reaction time, 188

reactions, 118, 150, 223

reactivity, 288, 290

real time, 202

reality, 190

recognition, 254

recommendations, 26, 194, 256, 262

recovery, 132, 255, 260

recreation, 22, 260

reflectivity, 69

regenerate, 286

regeneration, 274, 278, 279, 282, 289

regression, 14, 17, 18, 27, 29, 30, 37, 47, 86, 154

regression line, 86, 154

regression model, 37, 47

regrowth, 185, 193

regulations, 2, 282, 283

rehabilitation, 145

relevance, 170, 229

reliability, 266

religion, 206

remedial actions, 195

remediation, 283

replication, 197, 225

repressor, 232

reproduction, 88, 89, 240, 248

Nova Science Publishers, Inc.

Index

306

requirements, 189, 240

researchers, 4, 21, 72, 78, 164, 176, 188, 194, 201,

202, 254, 279, 287

residuals, 38, 42, 216

residues, 227, 228, 284

resistance, 189, 200, 209, 210, 211, 213, 214, 215,

216, 226, 227, 228, 230, 231, 234, 235, 237, 247,

287

resolution, 78, 80, 136

resources, 110, 111, 132, 162, 163, 164, 170, 177,

195, 196, 200, 204

respiration, 67, 68, 89

response, 10, 47, 50, 52, 70, 94, 168, 190, 194, 195,

240, 248, 249, 254

restriction enzyme, 225, 226

restrictions, 250

retardation, 193

reverse osmosis, 276, 281, 284

risk(s), 1, 5, 6, 31, 32, 33, 34, 37, 40, 43, 44, 46, 48,

49, 50, 53, 54, 55, 56, 57, 58, 142, 161, 166, 167,

174, 175, 177, 184, 185, 186, 188, 189, 192, 194,

195, 196, 201, 211, 217, 271, 273, 283

risk assessment, 31, 32, 57

risk management, 32, 46

risk profile, 32

river basins, 132

river flows, 147

RNA, 58

Romania, 204

root, 91, 121, 177, 199

routes, 200

Royal Society, 157

rubber, 286

rules, 47

runoff, 22, 63, 83, 94, 162, 163, 221, 278, 285

rural areas, 183

rural development, 197

rural population, 173

Russia, vi, ix, 239, 244, 250, 252, 253, 254, 259,

260, 266, 271

safety, 2, 58, 167, 168, 174, 188, 189

saline water, 100, 129, 221

salinity, 108, 110, 113, 115, 121, 127, 162, 240

salinity levels, 115

Salmonella, 214, 230, 231, 235, 236

salt tolerance, 121, 127

salts, 110, 115, 116, 162, 182, 184, 193

saltwater, 100, 108

sample design, 18

SANS, 163, 174, 181, 182, 183, 185, 186, 188, 189

saturation, 9, 66, 89, 212, 246

scaling, 13, 15, 19

scarcity, 162

scatter, 105

scatter plot, 105

school, 194, 197, 198

science, 163, 165, 168, 194, 205, 211

scoliosis, 245, 248, 249, 255

scope, 3, 4, 55

scripts, 32

seasonal changes, 16, 19, 20

seasonal effects, 2

seasonality, 5, 14, 17, 27

security, 205, 210, 237

sediment(s), 52, 54, 63, 64, 65, 66, 67, 68, 69, 70,

71, 72, 73, 74, 76, 77, 79, 82, 83, 84, 86, 87, 88,

89, 91, 92, 94, 95, 96, 100, 101, 206, 220, 233,

235, 236

sedimentation, 195, 218, 220

seed, 41, 198

selectivity, 275, 290

selenium, 284

self esteem, 195

semiconductor, 286

seminars, 196

senses, 192, 195, 203

sensitivity, 94, 95

sensors, 113

sequencing, 224, 226, 234

services, 162, 163, 164, 171, 172, 179, 188, 194,

196, 200, 203

settlements, 165, 181, 182, 188, 259, 261, 262, 263,

264, 267, 268, 269, 270

sewage, 104, 143, 162, 169, 182, 183, 193, 213, 215,

216, 217, 223, 228, 230, 232, 233, 235, 236, 239

shape, 15, 19, 36, 37, 57, 58

shear, 77, 78

shelter, 185, 193

shock, 217

shortage, 203, 209

showing, 7, 103, 104, 105, 106, 115, 122, 125, 132,

155, 166, 191, 211, 212, 214

shrimp, 273, 274, 279, 280

signs, 251, 252

silica, 65, 99, 104

simulation, 70, 80, 87, 88, 89, 94, 107

simulations, 64, 95

Singapore, 205

SiO2, 103, 106

SIP, 79

skeleton, 243, 248, 275

skills training, 197

skin, 245, 248

Nova Science Publishers, Inc.

Index 307

sludge, 214, 217, 218, 219, 220, 221, 222, 223, 228,

231, 235, 236

smoothing, 17

smoothness, 11, 12, 18

SO42-, 100, 102, 103, 104, 106, 109, 110, 119, 120,

135, 142, 143, 262

social justice, 196

society, 56, 161

sodium, 110, 115, 116, 118, 121, 122, 131, 133, 150,

151, 152, 285

Sodium adsorption ratio SAR, 121

software, 17, 102, 114, 137

solubility, 175, 212, 274, 275, 288, 290

solution, 64, 70, 71, 79, 81, 82, 113, 136, 212, 275,

278, 281, 283, 284, 288

solvents, 274

sorption, 262, 275, 276, 277, 279, 282, 286

South Africa, vi, ix, 107, 161, 162, 163, 164, 165,

167, 169, 170, 172, 173, 174, 175, 179, 188, 190,

192, 194, 202, 203, 204, 205, 206

South-East Queensland (SEQ), 1, 3

Spain, 233

speciation, 260, 282

species, 23, 110, 175, 214, 226, 240, 242, 245, 246,

273, 280, 282, 283, 285, 288

specific surface, 275, 282, 290

specifications, 165, 195

spectra analysis, 137

spectrophotometry, 263

spectroscopy, 263

spine, 245

spleen, 244

Sri Lanka, 204

SS, 64, 69, 72, 84, 87, 88, 89, 90, 92, 93, 94, 96

St. Petersburg, 256

stability, 240, 275, 278, 280, 287

staffing, 172

stakeholders, 164, 167, 168, 173, 190, 200

stamens, 245

standard deviation, 34, 36, 37, 38, 44, 102, 137, 170,

173

standard error, 38, 45

standardization, 4, 242, 254

state(s), 2, 65, 68, 77, 95, 106, 163, 199, 201, 204,

205, 239, 240, 242, 243, 244, 247, 248, 254, 255,

259, 278, 281, 284

statistics, 19, 114, 186, 187, 188

steel, 264

sterile, 222

stimulation, 248

stock, 244, 249

stomach, 142, 244

storage, 7, 38

stress, 75, 76, 77, 78, 87, 209, 243, 249, 255, 257,

265

stretching, 80

strong interaction, 63

strontium, 107

structure, 20, 33, 40, 42, 75, 100, 110, 114, 121, 150,

171, 172, 178, 188, 198, 206, 211, 245, 246, 248,

250, 251, 274, 275, 276, 278, 280, 287

subscribers, 202

substrate(s), 216, 222, 223, 234

sulfate, 117, 118, 282, 284, 285, 288

sulfuric acid, 277

sulphates, 250, 260

sulphur, 193

Sun, 291, 293

supplier, 31

supply chain, 53, 55

support staff, 171

surface area, 275

surface modification, 277

surface properties, 275

surface treatment, 277

surfactants, 235

surplus, 162, 179, 260

surveillance, 39, 56, 192, 194

survival, 179, 185

susceptibility, 214, 215, 230, 231, 233, 234

sustainability, 2, 109, 164, 178, 196, 242

sustainable development, 204, 205

Sweden, 292

swelling, 246, 248, 253

Switzerland, 130

symptoms, 239, 242, 244, 245, 246, 247, 250, 251,

255

syndrome, 142

synthesis, 197, 213, 223, 232, 234, 242, 277, 287

tactics, 180

Tanzania, 203, 204

target, 5, 33, 35, 36, 37, 167, 179, 213, 215, 232, 234

target population, 167, 179

techniques, 4, 17, 109, 111, 123, 128, 129, 234, 242,

252, 262

technology(s), 162, 174, 191, 192, 193, 196, 202,

205, 206, 250, 273, 278, 280, 282, 285, 288

teeth, 212

temperature, 46, 54, 56, 66, 67, 68, 73, 74, 89, 96,

101, 102, 109, 113, 115, 134, 135, 218, 220, 225,

228, 262, 276, 284, 288, 289

temporal variation, 2, 9, 12, 19, 89, 104

tensions, 110

Nova Science Publishers, Inc.

Index

308

territory, 259

testing, 49, 50, 194

tetracyclines, 227

texture, 251

Thailand, 196, 204

thermal stability, 289

thinning, 244

Third World, 165

threats, 173, 237

three-dimensional model, 64, 75

thyroid, 288

thyroid gland, 288

time periods, 47

time series, 82

tissue, 245, 246, 248, 251, 252, 266, 268, 271

tissue degeneration, 248

Togo, 135

toluene, 229, 230, 233, 234

top-down, 196

total costs, 45, 197

tourism, 260

toxic contamination, 254

toxic effect, 175

toxic metals, 162, 271, 279

toxic substances, 239, 240, 241, 243, 244, 245, 246,

247, 248, 254

toxicity, 175, 214, 216, 223, 241, 243, 253, 268, 270,

272, 277, 278, 280, 288

toxicology, 242

toys, 211

trace elements, 102, 113, 118, 157, 184, 260

trade, 210

training, 75, 193, 197, 198, 200, 203, 206

traits, 210

transformation(s), 5, 15, 17, 21, 25, 76, 77, 79, 102,

278

transmission, 175, 199, 228

transparency, 196, 263

transport, 63, 69, 76, 79, 96, 156, 290

transportation, 132

trial, 287

Triclosan in wastewater, 219

Triclosan, 213, 237

trimmings, 184

Tsonga, 165

tuberculosis, 233, 235, 236

tumor(s), 248, 270

turbulence, 75, 77

turbulent flows, 75

turnover, 132

typhoid, 163

U.S. Army Corps of Engineers, 96

U.S. Department of Agriculture, 158

U.S. Geological Survey, 83

UK, 219, 236

uniform, 28, 64, 78, 80, 143, 246, 262, 289

unit cost, 49, 50

United, 82, 83, 128, 206

United States, 82, 83, 128, 206

urban, 29, 51, 162, 165, 182, 184, 188, 221, 231,

232, 235, 278, 285, 288

urban areas, 182, 188

urbanization, 132

urine, 246, 248, 270

USA, 96, 97, 102, 206, 210, 216, 219, 230, 270

USDA, 158

USGS, 97

USSL diagram, 110

UV, 217, 283

UV light, 283

vacuum, 135

validation, 22, 80, 90, 91, 94, 95

vancomycin, 228, 235

variables, 2, 3, 5, 6, 11, 18, 27, 32, 33, 38, 46, 47, 53,

65, 68, 83, 95, 102, 109, 125, 170, 198

variations, 115, 118, 242

varieties, 198

vascular diseases, 270

vegetables, 134

vegetation, 111, 263

velocity, 47, 67, 74, 75, 76, 77, 78, 79, 80, 81, 87, 89

vertebrae, 245

vertebrates, 210

vessels, 244

videos, 202

Vietnam, 204

viral diseases, 185

viruses, 175

viscera, 244

viscosity, 74, 75, 76, 77, 78, 79, 80, 87

vision, 270

volatility, 33, 36

vomiting, 193

vulnerability, 161, 166, 167, 180, 184, 188, 189

war, 178

Nova Science Publishers, Inc.

Index 309

Washington, 128, 158, 205, 206, 232, 255, 272, 290

waste, 132, 147, 181, 188, 203, 210, 213, 220, 231,

232, 260, 263

waste disposal, 132

waste management, 203

waste water, 231, 232, 260

wastewater, 128, 209, 210, 211, 213, 215, 216, 217,

218, 219, 220, 221, 222, 227, 228, 229, 230, 231,

232, 233, 234, 235, 236, 237, 241, 245, 263, 264,

273, 274, 276, 277, 278, 280, 284, 285, 286, 288,

289

water chemistry, 100, 262

water ecosystems, 22, 260

water purification, 261

water quality index (WQI), 3, 111

water quality standards, 133, 253, 271, 273

water resources, 3, 110, 132, 157, 162, 163, 164,

165, 176, 177, 188, 190, 195, 196, 202, 204, 209,

239, 260

water shortages, 179

water supplies, 49, 165, 167, 176, 177, 179, 185

watershed, 83, 110, 196

weakness, 8, 178

wealth, 162, 173

welfare, 49

well-being, 164

wells, 101, 104, 109, 111, 113, 115, 116, 117, 125,

126, 135, 144

Western Australia, 97, 98

Western countries, 210

wetlands, 129, 194, 197

White Paper, 204

WHO, 102, 106, 108, 114, 115, 116, 117, 118, 119,

127, 139, 142, 143, 145, 146, 157, 162, 164

wild type, 225

wildlife, 56, 188

wind speed, 78

withdrawal, 100, 112

wood, 276

wood pulp production, 276

workers, 194, 200, 201, 202, 227, 270

workload, 171

World Bank, 197, 205, 206

World Health Organization (WHO), 102, 114, 130,

157, 172, 282, 283, 286

worldwide, 213, 215, 283

worry, 145

worst case scenarios, 16, 28

worst-case scenario (WCS), 3

yield, 100, 123, 150, 151, 183, 198, 224

zinc, 236

zooplankton, 67, 242

Zulu, 206

Nova Science Publishers, Inc.

Ecosystem health report cards: An overview of frameworks and analytical methodologies

Article

Full-text available

Nov 2019
ECOL INDIC

Ecosystem health report cards have become increasingly more important tools for communicating environmental state and assessing progress towards management goals. We provide an overview of the major analytical methods underpinning the translation of observed data into robust health indices. In particular, we outline the process of indicator selection, illustrate a variety of index metrics and describe index aggregation with consideration for weighting and the propagation of uncertainty.

Predicting Water Quality Parameters in Lake Pontchartrain using Machine Learning: A comparison on K-Nearest Neighbors, Decision Trees, and Neural Networks to Predict Water Quality

Conference Paper

May 2021

THE MANAGEMENT AND CONTROL OF MILK HYGIENE IN THE INFORMAL SECTOR BY ENVIRONMENTAL HEALTH SERVICES IN SOUTH AFRICA

Thesis

Full-text available

Jul 2007

Michael H.A. Agenbag

Local government (LG) is under increasing pressure from the milk industry and consumers regarding their ability and willingness to carry out their mandate with regard to the quality control of milk, especially in the informal sector. The government and the milk industry currently have programmes underway to stimulate economic activities in the informal sector, targeting emerging cattle farmers for the production of milk as part of government’s Accelerated Shared Growth Initiative of South Africa (ASGISA). These initiatives further increase the number of informal milk producers and distributors, which holds a further challenge to regulatory authorities. At the same time, the quality of milk from the informal milk-producing sector poses a serious public health concern. Most of the milk produced and sold by the informal sector is raw (unpasteurised), which does not meet the minimum statutory requirements, and the milking practices applied by the informal sector also do not comply with best practice compliance standards. Local authorities (LAs) are statutorily responsible for registering milking parlours and controlling milk hygiene quality from production stage to purchase stage in order to ensure safe and wholesome dairy products to the consumer. Therefore, LG should play an increasingly important role in ensuring that safe and wholesome milk is produced and distributed to the consumers. All metropolitan municipalities (metros) and district municipalities (DMs) should be authorised by the Ministry of Health to enforce the Foodstuffs, Cosmetics and Disinfectants Act, 1972 (Act 54 of 1972) through their authorised officials – mainly environmental health practitioners (EHPs). Secondly, LG should have specific programmes, systems and resources to register, monitor, evaluate and control milk production and distribution outlets for continued compliance. The main aim of this dissertation is to determine the legal compliance of LG in controlling food hygiene in general, and the approach of municipal health services (MHS) to monitoring and controlling milk hygiene at LG level. A further aim is to determine specific the availability of resources and systems to sustain their activities in this regard. This study was conducted amongst all participating metros and DMs in South Africa, targeting specifically the municipal health service managers. In the study the legal compliance and authorisation status of metros and DMs by the Ministry of Health and their respective EHPs was determined. The estimated number of informal milk producers in each metro and DM area was determined, as was MHS’ awareness of such. The availability of certain resources and the approach of MHS towards milk hygiene quality control in general were established in order to determine the MHS’ capacity to properly monitor and control milk hygiene in the informal sector. By September 2006 the majority (69.6%) of DMs and one metro had not yet been authorised by the Ministry of Health to enforce Act 54 of 1972. Accordingly, most of the EHPs had not been authorised by their statutorily mandated metros and DMs as required by that particular Act. It was noted that a few municipalities had authorised their EHPs, though they themselves had not yet been authorised by the Ministry of Health. Old disestablished municipalities, which were not supposed to be authorised after July 2004, were nonetheless still being authorised. For an LG to allow the sale of raw milk in its area of jurisdiction, application should be made to the Ministry of Health to be listed in Annexure C of Regulation R1555 of 21 November 1997, and proof should be given of its ability to exercise sufficient control over the selling of raw milk. However, according to the actual listing of relevant authorised LAs in the government notices, only the West Coast District Municipality is listed in Annexure C, allowing the sale of raw milk in its area, as statutorily required, together with local municipalities (LMs) and disestablished municipalities that are still listed, yet should not be. There are two tools that should assist metros and DMs, as well as the Ministry of Health, to determine the relevant municipality’s capacity to deliver MHS (including food control, of which milk hygiene control forms an integral part). The first tool is the approved report of a Section 78 (S.78) assessment, which was done in accordance with Sections 76, 77 and 78 of the Municipal Systems Act, 2000 (Act 32 of 2000). The legislation makes it compulsory for metros and DMs to conduct such an assessment to determine the authority’s current and future ability to render MHS and also to identify shortcomings. At the time of the survey (January 2006) only 25% (n=7) of the respondents indicated that their municipality had completed an S.78 assessment. The second tool is to ascertain that a project for milk hygiene control in the informal milk-producing sector is part of the municipality’s Integrated Development Plan (IDP) and subsequently part of the council’s budget. Unfortunately a specific question in this regard was not asked, but the Karoo DM indicated that milk hygiene monitoring and control was part of their district municipality’s IDP. Although just over half (55.3%) of the respondents were aware of informal milk-producing sources in their respective areas of jurisdiction, only 20% were making an effort to control them. A total of 68.1% (n=32) of the respondents stated that resources were not sufficient for the effective monitoring and control of milk hygiene, while a corresponding number of respondents (n=15 [48.4%]) stated that funds and the number of EHPs were regarded as their key reasons, and 35.5% (n=11) were of the opinion that a lack of basic equipment was contributing to insufficient resources. More than half (57.4% [n=27]) of the respondents were of the opinion that MHS were not applying effective monitoring and control of milk hygiene from the production stage to the consumer. In summary, the reasons involved a lack of systems, lack of fixed programmes, lack of a standardised approach or system to capture visits to premises and sampling results, and lack of a database in terms of milking parlours and distributors. When all the inputs from the respondents with regard to their reasons for the MHS not having proper control over milk hygiene are analysed and grouped in appropriate categories, 96.8% (n=30) of the reasons are management-related issues. Although food quality control was high on the agenda of the MHS’ daily activities, 63.6% (n=28) of respondents indicated that they were taking milk samples on an ad hoc basis, whereas 22.5% (n=9) disclosed that they were conducting planned premises evaluations, and 78.8% (n=26) of respondents stated that they were carrying out their health and hygiene education on an ad hoc basis. Only 16.3% (n=7) of the respondents indicated that they were integrating their inspections and sampling. The results therefore suggest that there is no audit- and risk-based approach to evaluating the premises. This means that most of the respondents were not planning their work in advance, resulting in superficial and inefficient MHS delivery. Various authors remind us that there is little value in this kind of monitoring and control activities at food premises in order to determine the safety of foodstuffs, and the approach should rather be outcomes driven. In conclusion, it is evident that MHS do not properly manage and control milk hygiene in the informal sector due to a lack of management capacity, as well as a lack of resources, standardised programmes, systems and so forth to optimally use the available resources in order for MHS interventions to serve their purpose and to contribute towards the building of consumer trust. There is thus a need for guidance and assistance from relevant role-players such as the National Directorates of Food Control and Environmental Health, the Department of Provincial and Local Government (DPLG), the South African Local Government Association (SALGA), the South African Institute of Environmental Health (SAIEH), tertiary institutions, the milk industry and other interested parties, to assist metros and DMs in the development of the abovementioned LG and MHS capacity. The Ministry of Health should ensure that all metros and DMs are authorised as legally required. Municipal health service managers should ensure that milk hygiene monitoring and control, especially of the informal sector, is included in their councils’ IDPs and subsequent linked programmes to ensure the availability of the necessary resources required to properly monitor and control the informal milk-producing sector.

Water Pollution: Causes Effects and Control

Book

Aug 1997

Pradip Kumar Goel

Water Pollution: Causes, Effects And Control Is A Book Providing Comprehensive Information On The Fundamentals And Latest Developments In The Field Of Water Pollution.The Book Is Divided Into 28 Chapters Covering Almost All The Aspect Of Water Pollution Including Water Resources And General Properties Of Water; History Of Water Pollution And Legislation; Origin, Sources And Effects Of Pollutants; Bioaccumulation And Biomagnification; Toxicity Testing And Interaction Of Toxicities In Combination; Water Quality Standards; Biomonitoring Of Water Pollution; Bacteriological Examination And Purification Of Drinking Water; Monitoring And Control Of Pollution In Lakes, Rivers, Estuaries And Coastal Waters; Physical And Biological Structure Of Aquatic Systems; And Structure, Properties And Uses Of Water.Some Important Topics Like Eutrophication, Organic Pollution, Oil Pollution And Thermal Pollution Have Been Discussed In Detail. The Water Pollution Caused By Pesticides, Heavy Metals, Radio Nuclides And Toxic Organics And Inorganic Along With The Water Quality Problems Associated With Water-Borne Pathogens And Nuisance Algae Have Also Been Dealt With Extensively.The Book Covers In Detail The Flow Measurement And Characterization Of Waste Waters In Industries, And Control Of Water Pollution By Employing Various Techniques For Treatment Of Biological And Nonbiological Wastes. The Considerations For Recycling And Utilization Of Waste Waters Have Also Found A Place In The Book. Special Topic Has Also Been Given On Water Pollution Scenario And Water Related Policies And Programmes In India.The Book Shall Be Of Immediate Interest To The Students Of Environmental Science, Life Science And Social Sciences Both At Undergraduate And Postgraduate Levels. People From A Wide Variety Of Other Disciplines Like Civil, Chemical And Environmental Engineering; Pollution Control Authorities; Industries; And Practicing Engineers, Consultants And Researchers Will Also Find The Book Of Great Interest.

Waterborne protozoan pathogens.

Article

Jan 1997

Protozoan parasites were the most frequently identified etiologic agents in waterborne disease outbreak from 1991 to 1994. The waterborne parasites Giardia lamblia, Naegleria fowleri, Acanthamoeba spp., Entamoeba histolytica, Cryptosporidium parvum, Cyclospora cayetanesis, Isospora belli, and the microsporidia are reviewed. For each parasite, the review includes history, life cycle, incidence, symptoms, and therapy. Clinical detection methods are compared, and emerging technologies are discussed. Information on the association of these parasites with waterborne outbreaks is reviewed. Current information on protozoan parasites identified as etiological agents in waterborne outbreaks is discussed. Water industry issues related to recent disease outbreaks are examined in the context of water quality testing regulations for G. lamblia and those proposed for C. parvum. The review identifies the limitations of the American Society of Testing and Materials water-testing method for these parasites. An overview of federal regulations affecting the water industry and laboratories that test for water quality is also provided. The article highlights the importance of the clinical laboratory as a frontline defense for the detection of infectious organisms. The review points to the need for clinical laboratories, physicians, and public health personnel to cooperatively plan and assess the challenge of meeting this potential public health threat.

Introduction and Review of Microbial and Biochemical Toxicity Screening Procedures

Chapter

Aug 2019

Cometabolic degradation of chloroallyl alcohols in batch and continuous cultures

Article

Nov 1994

The biodegradation of chloroallyl alcohols by pure and mixed bacterial cultures was investigated. Only 2-chloroallyl alcohol and cis- and trans-3-chloroallyl alcohol served as growth substrate for pure cultures. The other chloroallyl alcohols could be cometabolically degraded during growth on 2-chloroallyl alcohol. Cometabolic degradation of trichloroallyl alcohol, which was the most recalcitrant congener, by a Pseudomonas strain isolated on 2-chloroallyl alcohol resulted in 60% dechlorination. Efficient degradation of a mixture of chloroallyl alcohols in continuous culture could only be achieved in the presence of a satellite population. The mixed culture degraded 99% of the total chloroallyl alcohols added with 71% chloride release. The culture contained strains with a new catabolic potential. The results indicate the importance of mixed cultures and genetic adaptation for efficient chloroallyl alcohol removal.

Anaerobic oxidation of toluene (analogues) to benzoate (analogues) by whole cells and by cell extracts of a denitrifying Thauera sp.

Article

Jun 1995

Toluene and related aromatic compounds can be mineralized to CO2 under anoxic conditions. Oxidation requires new dehydrogenase-type enzymes and water as oxygen source, as opposed to the aerobic enzymatic attack by oxygenases, which depends on molecular oxygen. We studied the anaerobic process in the denitrifying bacterium Thauera sp. strain K172. Toluene and a number of its fluoro-, chloro- and methyl-analogues were transformed to benzoate and the respective analogues by whole cells and by cell extracts. The transformation of xylene isomers to methylbenzoate isomers suggests that xylene degradation is similarly initiated by oxidation of one of the methyl groups. Toluene oxidation was strongly, but reversibly inhibited by benzyl alcohol. The in vitro oxidation of the methyl group was coupled to the reduction of nitrate, required glycerol for activity, and was inhibited by oxygen. Cells also contained benzyl alcohol dehydrogenase (NAD+), benzaldehyde dehydrogenase (NADP+), benzoate-CoA ligase (AMP-forming), and benzoyl-CoA reductase (dearomatizing). The toluene-oxidizing activity was induced when cells were grown anaerobically with toluene and also with benzyl alcohol or benzaldehyde, suggesting that benzyl alcohol or benzaldehyde acts as inducer. The other enzymes were similarly active in cells grown with toluene, benzyl alcohol, benzaldehyde, or benzoate. This is the first in vitro study of anaerobic oxidation of an aromatic hydrocarbon and of the whole-cell regulation of the toluene-oxidizing enzyme.

Geochemical classification of mine drainages and natural drainages in mineralized areas

Article

Jan 1992

Environmental Cadmium Exposure, Hypertension and Cardiovascular Risk

Article

Feb 1996

The socio-economic dynamics of farmers' management of local plant genetic resources: A framework for analysis with examples from a Tanzanian case study

Article

Jan 1999

Esbern Friis Hansen

This paper discusses the debate around farmers' management of local plant genetic resources. It seek to develop a theoretical framework for analysing farmers management of plant genetic resources using examples from fieldwork carried out in 1995-1997 among farmers in Tanzania with a focus on the 1994/95 growing season. The study is part of a research programme involving several institutions and coordinated by CDR, titled Sustainable Agriculture in Semi-Arid Tanzania (SASA). Two very different systems of plant genetic resources management have developed in most developing countries, including Tanzania. One is a formal science-based plant genetic resource management system consisting of state institutions for plant conservation, breeding and seed multiplication and distribution. The other is an informal plant genetic resource management system based on farmers' local knowledge. Neither of the two plant genetic resource management systems are adequate for ensuring sustainable plant genetic resource use by small-scale farmers. The research suggests that the following three elements must be included when studying farmers management of local plant genetic resources: (i) an understanding of the concept of local knowledge vis a vis scientific knowledge, (ii) a conceptualisation of farmers institutions for managing plant genetic resources, and (iii) an analysis of the social, gender and cultural contexts of local plant genetic resource management. Farmer's local experimentation and collection of knowledge are not based on scientific theoretical models, but emerge from strong interests in knowledge with practical local application and which provides options for improving peoples' livelihood. Local technical knowledge, skills and capacities to manage plant genetic resources are embedded in the cultural, ecological and sociological context of the community. In the case of plant genetic resource management, such local knowledge evolves around the crop cycle and the three main groups of management practices are (i) seed selection, (ii) treatment and storage of seed and (iii) seed multiplication and exchange. The strength of local seed selection is its ability to adapt varieties to the specific cultural, economic and social requirements of local farming systems. Seed is often, but not always, stored separate from the harvested crop and a range of local seed treatment and storage techniques is used by farmers to ensure viable seed for future planting seasons. While most households are self-sufficient with seed of frequently used varieties, new or less common varieties are frequently exchanged within or between communities. Community seed exchange also serves as a back-up source of seed in years of deficit harvest or following droughts or other emergencies. Local institutions for managing plant genetic resources only seldom take the shape of organisations and commonly consist of shared norms and practices within a community or ethnic group of people. While local institutions for natural resource management are often associated with conflict over access and use, plant genetic resources are renewable and therefore are seldom the cause of conflict over access at the local level. Plant genetic resource management is largely the product of social processes and institutions in the community level. Limitations to local plant genetic resource management include, (i) constraints in terms of knowledge about genetics (lack of support from scientists), (ii) limited access to globally or nationally available plant genetic resources, and (iii) declining capacity of local resource management institutions. There are potentials for synergy if the weak links between the two systems of plant genetic resource management were enhanced to develop a participatory partnership, which takes advantage of the comparative strengths of each of the two systems.

Assessment of drinking water quality in the city of apatity

Article

Jan 1999

L.P. Kudryavtseva

An industrially developed center in the Kola North is used as an example to consider the problem of drinking water quality. A low content of dissolved salts along with the presence of free carbon dioxide is shown to make natural water corrosive with respect to metals. It is found that the mobility and toxic properties of anthropogenic solutes increase in waters with low salt content, and the quality of such water deteriorates.

Water quality: Indicators, human impact and environmental health

Abstract

Recommended publications

The European union’s 2010 target: Putting rare species in focus

Indicators and assessment methods for measuring the ecological integrity of semi-aquatic terrestrial...

Methods to Assess the Effects of Chemicals on Ecosystems.

The Economic Value of the Greater Montreal Blue Network (Quebec, Canada): A Contingent Choice Study...