BookPDF Available

SPEAR: sustainable options for people catchment and aquatic resources

Authors:

Abstract and Figures

This book describes the approach and main result obtained in the Sustainable options for People, Catchment and Aquatic Resources (SPEAR) Project, together with complementary case studies focusing on realted work in China and in the United States. Each chapter in this book is designed to be readable by itself and contains enough information for the reader to understand both the methodologies applied and the key outcomes. Wherever possible, those outcomes have been developed into products, with the objective of leveraging their usability as a legacy of SPEAR
Content may be subject to copyright.
J.G. Ferreira, H.C. Andersson, R.A. Corner, X. Desmit, Q. Fang, E.D. de Goede, S.B. Groom,
H. Gu, B.G. Gustafsson, A.J.S. Hawkins, R. Hutson, H. Jiao, D. Lan, J. Lencart-Silva, R. Li,
X. Liu, Q. Luo, J.K. Musango, A.M. Nobre, J.P. Nunes, P.L. Pascoe, J.G.C. Smits,
A. Stigebrandt, T.C. Telfer, M.P. de Wit, X. Yan, X.L. Zhang, Z. Zhang, M.Y. Zhu, C.B. Zhu
S.B. Bricker, Y. Xiao, S. Xu, C.E. Nauen, M. Scalet.
IMAR – Institute of Marine Research
http://www.imar.pt
Sustainable Options for People,
Catchment and Aquatic Resources
The SPEAR Project,
an International Collaboration
on Integrated Coastal Zone Management
3
SPEAR TEAM
J.G. Ferreira
J. Lencart-Silva
A.M. Nobre
J.P. Nunes
Y. Xiao
IMAR - Institute of Marine Research,
Centre for Ocean and Environment, New University of Lisbon, Portugal
A.J.S. Hawkins
S.B. Groom
R. Hutson
P.L. Pascoe
Plymouth Marine Laboratory, United Kingdom
T.C. Telfer
R.A. Corner
Institute of Aquaculture, University of Stirling, United Kingdom
A. Stigebrandt
H.C. Andersson
B.G. Gustafsson
University of Gothenburg, Sweden
J.G.C. Smits
X. Desmit
E.D. de Goede
Deltares, Netherlands
M.Y.Zhu
R. Li
X. Liu
X.L. Zhang
Z. Zhang
First Institute of Oceanography, State Oceanic Administration, China
X. Yan
H. Jiao
Q. Luo
S. Xu
Ningbo University, China
D. Lan
Q. Fang
H. Gu
Third Institute of Oceanography, State Oceanic Administration, China
J.K. Musango CSIR, South Africa
M.P. de Wit De Wit Sustainable Options (Pty) Ltd, South Africa
C.B. Zhu South China Sea Fisheries Research Institute, CAFS, China
S.B. Bricker Center for Coastal Monitoring and Assessment, NOAA, USA
C.E. Nauen European Commission, Brussels, Belgium
M. Scalet European Commission, Brussels, Belgium
4
FOREWORD
This book describes the approach and main results obtained in the Sustainable options for
PEople, catchment and Aquatic Resources (SPEAR) Project, together with complementary
case studies focusing on related work in China and in the United States. Each chapter in this
book is designed to be readable by itself, and contains enough information for the reader
to understand both the methodologies applied and the key outcomes. Wherever possible,
those outcomes have been developed into products, with the objective of leveraging their
usability as a legacy of SPEAR.
From a technical standpoint, these products currently represent the state-of-the-art in coast-
al management, featuring web-based models, hybrid ecological-economic approaches, and
management tools to be used at a variety of scales. Technological developments will mean
that the tools themselves will evolve fairly rapidly, but the underlying scientic paradigms are
expected to change more slowly.
Models do not by themselves lead to robust management, without a complementary invest-
ment in appropriate environmental data. Management-oriented work in Integrated Coastal
Zone Management (ICZM) often results in multidisciplinary actions, rather than an interdis-
ciplinary approach. As a consequence, there is often a lack of integration which is limiting to
coastal management. Three aspects of this merit further analysis:
(i) The lack of effective interaction between natural and social scienc-
es, the acknowledgement of the limitations and errors of each, and
the recognition that ICZM can only be appropriately addressed by a
well integrated approach;
(ii) The understanding that environmental baselines are shifting, in
some cases rather rapidly, and that the record of that shift is often at
best anecdotal;
(iii) The realisation that tools such as those developed in SPEAR are of
maximum utility when all social agents, such as environmental and
fisheries agencies, farm stakeholders, non-governmental agencies
and other parties are actively involved.
This book does not aim to provide an exhaustive account of all the research executed in
SPEAR, and the reader is directed to the ofcial project website, available in English at http://
www.biaoqiang.org/ and in Chinese at http://www.spear.cn/
A digital copy of this book is available on the site, together with links to databases, models
and all other resources made available by this research.
6
7
8
9
10
11
oyster
cage area
coastal features
coastal pond
12
1600
1400
1200
1000
800
600
400
200
0
0 200 400 600 800 1000 1200 1400 1600
R2 = 0.9233
R2 = 0.9515
Modelled mean growth (g)
Observed mean growth (g)
13
14
Shellfish filtration
Phytoplankton removal
2608 Kg C y-1
Detritus removal
972 Kg C y-1
N removal (kg y-1)
Algae -406
Detritus -151
Excretion 24
Faeces 129
Mass balance -404
Shellfish farming: 77.3 k y-1
Nutrient treatment: 36.7 k y-1
Total income: 114.0 ky-1
Density: 20 oysters m-2
Cultivation period: 365 days
20% mortality
3.3 kg N y-1 PEQ
ASSETS INCOME PARAMETERS
Chl a
O2
Score
Population equivalents
122 PEQ y-1
EUTROPHICATION CONTROL
15
16
17
Aquaculture zones
Fishcages
Kelp
Bivalves
Scenario 2
Scenario 3
Scenario 1
18
19
EXECUTIVE SUMMARY
In 2004, the European Union nanced a research project entitled Sustainable options for
PEople, catchment and Aquatic Resources (SPEAR). This project was framed in the INterna-
tional COoperation for DEVelopment (INCO-DEV) programme, with its focus on mutually
benecial and equitable partnership in research between the Community and its Member
States on the one hand and INCO target countries on the other.
The general objective of SPEAR was to develop and test an integrated framework for man-
agement of the coastal zone, using two test cases where communities depend primarily
upon marine resources.
Two contrasting coastal systems in China were used as study areas. Sanggou Bay is in a rural
area in the North, and Huangdun Bay is in an industrialized area south of Shanghai, subject
to substantial human pressure at both local and regional levels. The common denominator
for both is that aquatic resources, i.e. cultivated species of seaweeds, shellsh and nsh, are
of paramount importance for community income and livelihood, both locally and regionally.
This book describes the approach and main results obtained, together with complementary
case studies focusing on related work in China and in the United States.
The book is divided into seven chapters, followed by a Conclusions section. Each of the
chapters is designed to be readable by itself, allowing different publics to nd their sections
of interest without needing to refer extensively to other material. A brief description of each
chapter is given below.
20
Rationale for the SPEAR Project
The motivation and objectives for this work are explained here, together with an overview of
the way in which this project ts into the wider context of coastal zone management.
The key objectives of SPEAR are outlined below, focusing on integration of disciplines and
tools as a primary element, and specically associating ecology and economics as the two
major components of the framework.
Later chapters illustrate how the various parts were brought together, and provide specic
examples of applicability.
Main objectives of SPEAR
Develop an integrated framework that simulates the dynamics of the
coastal zone accounting for basin effects (exchanges of water, sediment
and nutrients), ecological structure and human activities
Test this framework using research models, which assimilate dispersed
local and regional data, and develop screening models which integrate key
processes and interactions
Examine ways of internalizing environmental costs and recommend re-
sponse options such as optimisation of species composition and distribu-
tions, thereby restoring ecological sustainability
Evaluate the full economic costs and benefits of alternative management
strategies, and societal consequences
Provide managers with quantitative descriptors of environmental health,
including simple screening models, as practical diagnostic tools
There is a chapter dedicated to an overview of the various tools, which will guide a reader
who is fundamentally interested in knowing what scientic and technical instruments are
available to support coastal management.
21
Tools
This chapter provides a description of data, remote sensing and modelling tools, cross-
cutting the different scientic disciplines.
oyster
cage area
coastal features
coastal pond
FIGURE 1. Huangdun Bay: (Left) Landsat band 5, near IR scene showing coastal features and sh cages.
(Right) Diagram of main aquaculture provided by Chinese partners.
Figure 1 shows an example of these tools, illustrating how satellite data were combined with
ground truthing data and local knowledge to generate nal aquaculture maps.
Systems
This chapter provides the grounding for the system-scale modelling work. Models can only
ever be as good as the data that drive them, so a robust data programme is a key element
for success.
China Shandong
province Sanggou Bay ZheJiang
province Huangdun Bay
(Ningbo City)
Population
(million) 1 300 92 0.15 47 6
Urban per
capita dispos-
able income
(USD)
1 290 860 2000
(Weihai)
560 (Jinan)
460 (Yantai)
1 260 3 300
Primary sec-
tor share of
economy (%)
15% 11% n/a 7% 7%
Fish production
(tons) 47 061 064 7 062 244 188 227 4 935 288 903 301
22
Total sheries
value (RMB
millions)
332 341 37 600 5 897 13 859 n/a
Related indus-
try value (RMB
millions)
126 186 40 208 9 212 3 083 n/a
Related servic-
es value (RMB
millions)
119 357 22 505 500 292 n/a
Marine farming
value (RMB
millions)
73 375 16 797 1 364 n/a n/a
Total sheries
jobs 7 007 564 n/a n/a n/a n/a
Fish farming
jobs 4 324 174 n/a n/a n/a n/a
Marine farming
jobs n/a n/a 11 100 n/a n/a
FIGURE 2. Selected socio-economic indicators for the regions (compiled from FAO 2004, China Data Centre,
National Bureau of Statistics of China).
Data collation was a key element of the work, data collection was only carried out as required,
with a view to optimising costs and leveraging previous work. A horizontal approach to this
part of the work meant that both the catchment and waterbody were considered, and that
there was a close integration of the various scientic disciplines involved.
Figure 2 shows an example of the data collected for the two bays. Both Shandong and Zheji-
ang provinces are less dependent on the primary sector than China in general, but Shandong
is more dependent on primary production than Zhejiang. The total value of all sheries pro-
duction and the value of marine farming is also markedly higher in Shandong where 45% of
the total sheries value is in marine farming.
In China as a whole 4.2 million people are employed in the sh farming (inland and marine)
business. This means that almost 19 direct sh farming jobs are created per RMB 1 million of
value in sh farming. Comparative statistics for the provinces were not available, but for ma-
rine farming only in Sanggou Bay this gure is around 9 direct jobs per RMB 1 million in total
marine farming value.
23
Aquaculture
This chapter provides a full description of aquaculture activity within the two bays, highlight-
ing the areas involved, key features of the culture practice, and socio-economic aspects. In
addition, the models used to describe growth and waste production for seaweeds, shellsh
(both bivalves and shrimp), and nsh are described here.
1600
1400
1200
1000
800
600
400
200
0
0 200 400 600 800 1000 1200 1400 1600
R2 = 0.9233
R2 = 0.9515
Modelled mean growth (g)
Observed mean growth (g)
FIGURE 3. Modelled growth (TGC model) against empirical data for Japanese seabass (black) and yellow
croaker (blue) grown in Huangdun Bay.
As an example, Figure 3 shows a comparison of modelled and observed sh growth of spe-
cies in Huangdun Bay, simulated using a Thermal Growth Coefcient model.
24
Ecosystem models
The system-scale data were combined with aquaculture data to provide the information re-
quired to run ecosystem-scale models.
SPEAR system-scale models
Fine-scale models, simulating the three-dimensional water circulation in
both bays
Broader-scale water quality models, simulating key features of water and
sediment properties
Coarse-scale ecological models, which represent the systems using a few
dozen boxes, but contain all the necessary elements of the ecological and
human components, and simulate multi-year periods
Economic models, coupled to the ecological models, in order to explicitly ac-
count for the interactions between the human system and the ecosystem
The integration of the various models was carried out both online (with models running
together and interacting with each other) and using ofine coupling, with results from one
model being used to drive another model. The essence of this was to capture the scales at
which important phenomena occur, since it is clearly impossible to use the same time and
space scales to simulate the detailed water circulation over a tidal cycle and the decadal
production of oysters.
Box 1 Box 2 Box 3 Box 4 Box 5 Box 6 Box 7 Box 10 To tal
Oyster (model) 9817 8948 1534 6912 8810 ---36021
Oyster (data) 9764 8976 1500 6860 7220 ---34320
Razor (model) 808 568 471 52 115 46 - - 2058
Razor (data) 812 588 415 57 95 30 - - 1997
Manila clam (model) 134 102 91 26 26 739 6431
Manila clam (data) 116 84 108 26 27 837 5410
Muddy clam (model) 355 316 174 -35 23 - - 903
Muddy clam (data) 394 286 199 -26 14 - - 920
FIGURE 4. EcoWin2000 shellsh harvest results and comparison with data for Xiangshan Gang (ton yr-1).
25
A synthesis illustrating nal outputs from these models, which are of clear management
interest, is shown in Figure 4 for Xiangshan Gang. The data compare the results from the
EcoWin2000 ecological model to reported production for four shellsh species in Xiangshan
Gang.
In order to produce such results, a diverse set of models needs to be combined, including hydro-
dynamic models such as Delft3D, shellsh individual growth models such as ShellSIM, models
of sh production and waste such as MOM, and economic models such as MARKET.
Screening models
SPEAR also focused on the development and implementation of screening models, which
have a totally different objective and user-base than the research models referred above.
A screening model is a tool that may be useful for a sh farmer, farm manager or coastal
manager. Typically these models are easy to use, run in minutes, and support decisions
such as the assessment of the impact of a particular planning option or classication of the
ecological status of an aquatic system. Two examples of application of this type of model are
provided below.
Particulate waste models
At the local scale, screening models may be used to look at aquaculture yields, local impacts
of sh farming, and water quality. A good example is a particulate waste distribution model
developed for sh culture in Huangdun Bay (Figure 5) using GIS, which provides a footprint
of organic enrichment beneath sh farms.
FIGURE 5. Screening model for carbon input to sediments from sh culture.
26
FARM
The Farm Aquaculture Resource Management (FARM) modelling framework applies a com-
bination of physical and biogeochemical models, bivalve growth models and screening mod-
els for determining shellsh production and for eutrophication assessment. Requirements
for input data have been reduced to a minimum, since the model is aimed at the shellsh
farming community and local managers. Model inputs may be grouped into data on: (i) farm
layout, dimensions, species composition and stocking densities; (ii) suspended food enter-
ing the farm; and (iii) environmental parameters. The FARM model is publicly available at
http://www.farmscale.org/
Shellfish filtration
Phytoplankton removal
2608 Kg C y-1
Detritus removal
972 Kg C y-1
N removal (kg y-1)
Algae -406
Detritus -151
Excretion 24
Faeces 129
Mass balance -404
Shellfish farming: 77.3 k y-1
Nutrient treatment: 36.7 k y-1
Total income: 114.0 ky-1
Density: 20 oysters m-2
Cultivation period: 365 days
20% mortality
3.3 kg N y-1 PEQ
ASSETS INCOME PARAMETERS
Chl a
O2
Score
Population equivalents
122 PEQ y-1
EUTROPHICATION CONTROL
FIGURE 6. Application of FARM to calculate a mass balance for oyster culture in Sanggou Bay
The application of FARM to an oyster farm in Sanggou Bay is shown in Figure 6, and il-
lustrates the substitution value of shellsh with respect to land-based control of nutrient
emissions. In a scenario of integrated catchment management of discharges of nitrogen and
phosphorus, such as already occurs in parts of the U.S. and in Scandinavia, shellsh farmers
in Southeast Asia may in future be able to sell nutrient credits to their land-based counter-
parts in much the same way as carbon credits are traded today.
27
Management and case studies
The tools developed within SPEAR are designed to support decision-makers in success-
fully implementing integrated castal zone management (ICZM), harmonising various (often
competing) uses into a framework for sustainable development.
The project team worked closely with stakeholders to develop a set of example scenarios
(Figure 7) which could be examined using the SPEAR models.
System Scenario description Too ls
Huangdun Bay Assess impact of change to sh cage numbers and sizes GIS, EcoWin2000
Assess impact of nutrient discharge reduction from waste water
treatment plants SWAT, Delft3D,
EcoWin2000
Combination of the two scenarios above As above
Sanggou Bay Reduce culture densities for shellsh alone by 50% (achieved by
increasing distance between longlines and/or droppers, to assess
consequences for total production value
GIS, EcoWin2000
Alter species composition: currently there are 450 Mu1 of sh
cages, 50,000 Mu of Laminaria, 40,000 Mu of shellsh, proposed
change to a 70:20:10 (kelp:lter:nsh)
GIS, EcoWin2000
Replace oyster culture (1500 Mu) with abalone culture (1000 Mu)
and sh cages (400 Mu) MOM, FARM
FIGURE 7. Development scenarios for Huangdun Bay and Sanggou Bay
The tools that are appropriate for each type of scenario differ, highlighting the importance of
multiple models, tailored to appropriate issues and scales. A key nding from this integrated
project has been that the combination of models running at widely varying time and space
scales is at the core of a successful analysis.
Results from two scenarios are provided here to illustrate the potential for this kind of ap-
plication.
Huangdun Bay – Changes to fish cages and nutrient reduction
The stakeholder community in Huangdun Bay identied this as a major management ques-
tion. It is thought that the sh farms have a substantial impact on the water quality in Huang-
dun Bay, due to excessive organic loading. To further improve water quality there are also
plans for sewage water treatment plants.
1 The Mu is the Chinese unit of area. In aquaculture, the Culture Mu is used for licensing, and although
nominally rated as 1/15 of one hectare, its size is variable according to the productivity of the system, i.e. a less
productive system has a larger Culture Mu. Typical values range from 1000-5000 m2.
28
CoBEx-Eco has been used to estimate changes in water quality for the three different sce-
narios (Figure 8):
Reduction of sh farms (proposed by stakeholders to be about 40% of the total production)
Reduction in sewage discharge
The two reduction scenarios combined
The scenario simulations indicate that there are only minor changes in net primary produc-
tion for all three cases, signifying that reduction in loads will have little effect on the water
quality. The reason for this is that inorganic nutrient concentrations remain high during the
whole year and therefore nutrient availability does not limit production. Thus, reducing nu-
trient concentration slightly will not cause any signicant change in primary production. To
further check the sensitivity of this statement, a simulation with all land based loads reduced
to zero (i.e. no loads from land, sewage or pond cultures) was carried out, which resulted
in a 10% reduction of primary production. Although phosphate concentrations are reduced
by 50-75%, there was still no nutrient limitation. These results seem reasonable as long as it
can be veried that inorganic nutrient concentrations are indeed high during the productive
season. For the standard case this appears to be reasonable, but for such large perturbations
as in the no load case, further validation of the model is probably necessary to exclude the
possibility of nutrient limitation.
Bay Fish cage
reduction Increased sewage
treatment Combination of sh cage reduc-
tion and sewage treatment No loads from land,
shrimp ponds or sewage
Xize 0.2 1.0 1.2 6.2
Wushe 0.3 1.4 1.8 9.2
Tie 0.3 1.5 1.8 11.6
Huangdun 0.2 1.3 1.5 10.4
FIGURE 8. Reduction in net primary production (in %) relative to the standard case for the three scenarios
and the no land load case.
Locally, sh farms may have an impact on the sediment quality beneath the farms due to
accumulation of particulate matter. Estimates from the MOM model indicate that 60-80%
of the outputs from a farm originate from uneaten food due to overfeeding of the sh. There-
fore, local improvements can probably be attained by improving feeding routines rather than
by reducing the size of a farm or the number of farms. Measures in this direction will of
course also increase prots as feed can be a substantial part of the total production cost.
29
Sanggou Bay – Changes to culture combinations
A farm was selected as a demonstration site, located in Box 4 of Sanggou Bay (Figure 9),
where Pacic oyster (Crassostrea gigas) raft culture, Japanese Flounder (Paralicthys olivaceus)
and Puffer sh (Fugu rubripes) cage culture coexist.
Box 4 is a polyculture area as a whole, but may be divided into three types of farms: shell-
sh monoculture farms (located in the northern part of Box 4), shellsh farms separated by
navigation channels (located in the middle of Box 4), and shellsh – sh cage IMTA farms
(located in the southeastern part of Box 4). The three types of farms are adapted into three
setup options, and have been simulated in the FARM screening model.
Aquaculture zones
Fishcages
Kelp
Bivalves
Scenario 2
Scenario 3
Scenario 1
FIGURE 9. Box layout of Sanggou Bay and the location of FARM simulation area
Figure 10 shows the three different options considered. The rst setup considers oysters in
all sections, at a density of 20 animals m-2, the second considers only oysters in the two end
sections, and the third adds sh cages to the middle section of the farm.
NºDescription TPP
(ton TFW) APP Prot
(K)Nitrogen
removal (kg y-1)PEQ
(y-1)Total
income1
(K y-1)
1 Oysters in sections 1, 2
and 3, 15.5 875.4 404 122 114
2 Oysters in sections 1 and 3,
section 2 empty 10.4 850.7 270 82 76.5
3 Oysters in sections 1 and 3,
section 2 sh cages only 18.1 14 89.1 350 106 122.2
FIGURE 10. Setup options and results of Sanggou Bay FARM scenario
1 Includes substitution costs of nutrient removal on land, e.g. by reducing application in agriculture. Does not
include the additional revenue from sh cages in scenario 3.
30
The addition of sh cages in the middle section of the farm (Setup 3) provides an additional
source of revenue, and the additional input of particulate organic matter to the downstream
part of the farm substantially increases oyster production, which in total exceeds that in the
uniform distribution used in Setup 1, and has the added advantage of reducing the local
organic deposition effects of sh aquaculture. Overall, Setup 3 provides both the highest
prot from production activities and the highest potential income when considering also
the environmental costs of nutrient treatment, or (alternatively) the resale value of nitrogen
credits as a catchment management option.
Case studies
Over the life-cycle of SPEAR, a number of complementary activities were carried out, known
collectively as the SPEAR Leverage Programme. One example of these was the development
of a project for Trophic Assessment in Chinese Coastal Waters, or TAICHI, which aims to
execute a full assessment of eutrophication in Chinese coastal waters. A case study of the
application of the ASSETS assessment method to Jiaozhou Bay is presented as an example,
together with a case study from the United States which brings together eutrophication
assessment and sheries, in an integrated approach to coastal management from outside
China.
Conclusions
Integrated assessment of the different components of coastal systems, contemplating land-
based drivers and pressures, uses such as aquaculture and sheries and impacts such as
eutrophication is a necessary pre-requisite to successful coastal management.
Work carried out in SPEAR, which is described in detail in this book, represents one ap-
proach to address this requirement. The outputs from multi-year models are not only useful
in themselves, as highlighted previously, but serve to drive farm-scale models and other
screening models of various types, which are of interest to both the farmer and regulator.
The possibility of operating coarser scale models, such as the EcoWin2000 implementations
described in this work, allows users to deal with manageable amounts of data and accept-
able run-times. This trade-off between mutiple-year simulation and spatial complexity, whilst
preserving acceptable levels of accuracy, is essential in building a bridge with microeconomic
models, which require simulations at the decadal scale.
Future developments of simulation approaches must include the linkage of both the natural
and social sciences, if possible with explicit feedbacks. This will allow changes in pricing
linked to production, supply and demand, to be reected in the attractiveness of commer-
cial cultivation, and provide indicators on employment and other aspects of social welfare.
Additionally, by factoring in the non-use value of ecosystems, with respect e.g. to the valua-
tion of biodiversity, a more complete mass balance of the effective gains to society may be
31
computed. An holistic assessment of aquaculture on the basis of people, planet and prot,
as has been applied elsewhere should become central to studies of sustainable coastal zone
management, particularly in areas such as Southeast Asia where such activities are highly
developed. This concept, sometimes termed the triple bottom line, is a goal that is at pres-
ent challenged by the application of fragmented approaches. The work we have described
in the framework of SPEAR allows managers to examine the consequences of development
for biodiversity, conservation and habitat protection, water quality and yield, including prot
maximisation through the use of marginal analysis.
The integration of basin-scale models such as SWAT which allow for the effects of changes
in land use agricultural practice to be explicitly simulated in this framework, provides a link
to the drivers and pressures of nutrient loading to the coastal zone. The explicit connec-
tion with economic models, including incorporation of dynamic feedbacks, is also an area
where exciting developments are expected in the near future. The challenge of bringing the
various components of the People-Planet-Prot equation together as a holistic indicator of
sustainable carrying capacity in coastal areas appears both achievable and appropriate for
integrated coastal management.
32
33
RATIONALE FOR THE SPEAR PROJECT
Summary
In 2004, the European Union nanced a research project entitled Sustainable options for
PEople, catchment and Aquatic Resources (SPEAR). This project was framed in the INterna-
tional COoperation for DEVelopment (INCO-DEV) programme, with its focus on mutually
benecial and equitable partnership in research between the Community and its Member
States on the one hand and INCO target countries on the other.
The general objective of SPEAR was to develop and test an integrated framework for manage-
ment of the coastal zone, using test cases where communities depend primarily upon ma-
rine resources. Two contrasting coastal systems in China were used as study areas. Sanggou
Bay is in a rural area in the North, and Huangdun Bay is in an industrialized area South of
Shanghai, subject to substantial human pressure at both local and regional levels. In both
systems, cultivated species of seaweeds, shellsh and nsh are of paramount importance
for community income and livelihood, both locally and regionally.
34
This chapter describes the rationale for the work and the main objectives of SPEAR:
Develop an integrated framework that simulates the dynamics of the coastal zone ac-
counting for basin effects (exchanges of water, sediments and nutrients), ecological
structure and human activities;
Test this framework using research models, which assimilate dispersed local and regional
data, and develop screening models which integrate key processes and interactions;
Examine ways of internalizing environmental costs and recommend response options
such as optimisation of species composition and distributions, thereby restoring ecologi-
cal sustainability;
Evaluate the full economic costs and benets of alternative management strategies, and
societal consequences;
Provide managers with quantitative descriptors of environmental health, including simple
screening models, as practical diagnostic tools.
35
Problem definition
In 2004, the European Union nanced a research project entitled Sustainable options for
PEople, catchment and Aquatic Resources (SPEAR). This project was framed in the INterna-
tional COoperation for DEVelopment (INCO-DEV) programme, with its focus on mutually
benecial and equitable partnership in research between the Community and its Member
States on the one hand and INCO target countries on the other.
The general objective of SPEAR was to develop and test a structurally integrated conceptual
framework for interpretation of coastal zone structure and dynamics, within areas where
communities depend primarily upon marine resources.
Main objectives of SPEAR
Develop an integrated framework that simulates the dynamics of the
coastal zone accounting for basin effects (exchanges of water, sediments
and nutrients), ecological structure and human activities
Test this framework using detailed research models, which assimilate
dispersed local and regional data, and develop screening models which
integrate key processes and interactions
Examine ways of internalizing environmental costs and recommend re-
sponse options such as optimisation of species composition and distribu-
tions, thereby restoring ecological sustainability
Evaluate the full economic costs and benefits of alternative management
strategies, and societal consequences
Provide managers with quantitative descriptors of environmental health,
including simple screening models, as practical diagnostic tools, innova-
tively combining local and regional datasets
Two contrasting coastal systems in China were used as study areas. Sanggou Bay is in a rural
area in the North, and Huangdun Bay is in an industrialized area south of Shanghai, subject to
substantial human pressure at both local and regional levels. The common denominator for
both is that aquatic resources, i.e. cultivated species of seaweeds, shellsh and nsh, are of
paramount importance for community income and livelihood, both locally and regionally.
36
Impact of science on
sustainability in the coastal
zone
The Johannesburg Plan of Implementation
(JPoI) illustrates that ten years on from the
Rio Earth Summit, the motto of the 2002
World Summit on Sustainable Development
was no longer ‘environment and develop-
ment’, even less ‘environment or develop-
ment’, even though that contraposition still
lingers in many mindsets. It uses the em-
blematic title of ‘sustainable development’
to underscore the need for a durable balance
between social, environmental and econom-
ic dimensions of development in a mode
that is inspired by the Asian thinking that
we borrow the Earth from our children. The
habitual language of ‘conserving’ or ‘protect-
ing’ marine resources was replaced by time-
bound restoration of degraded marine eco-
systems, to the extent still possible, by 2015.
In addition to the usual technical measures,
the prescription for action includes the es-
tablishment of networks of marine protected
areas by 2012 as a pathway towards achiev-
ing the objective. Since then the Millennium
Ecosystem Assessment (2005) has been in-
strumental in producing the most compre-
hensive compilation and interpretation of
the state of the world’s ecosystems, includ-
ing coastal zones.
37
Reconciling multiple demands on coastal zones
The time-bound objectives of the JPoI require that each country and region examines how
these can be articulated in practice, resulting in mechanisms to take effective action for their
achievement. Between 1996 and 1999, 36 demonstration projects and six thematic studies
explored integrated approaches to coastal zone management to counter increasing coastal
degradation.
The European Marine Directive in the context of ICZM in China
The European Marine Directive is a component of the European Sustainable Development
Strategy and the 6th Environment Action Programme (2002-2012) and intended to be the
environmental pillar of a European Maritime Policy. A Green Paper for developing a mari-
time policy was under public consultation through much of 2006 and 2007 and will lead to
more follow-up work, including research touching on the environment, food production,
greater energy efciency in transport and capacity building and climate change mitigation
and adaptation as cross-cutting issues. Key ecosystem functions and recreational value of
beaches and other parts of the coastal zone are already supported by some legislation. In
Europe, among others, a modest network of Natura 2000 nature protection areas was rst
established under the 1992 Habitats Directive. Yet the European Environment Agency re-
port on coastal zones published in 2006 raises the spectre of continuous degradation of
Europe’s coasts, threatening living standards. It illustrates how sums of individual decisions
predominantly focused on economic gain not only compromise environmental and social
well-being, but backre on economic opportunities in the medium- to long-term and can
lead to structural decline.
Although coastal regions in China are more prosperous than most inland regions, the in-
crease in environmental problems is being felt strongly there as well. Developing suitably
located and sequenced mariculture is one response to excess nutrients from terrestrial
sources, which has turned a potential liability into an asset. However, heavy metals, organo-
chlorine compounds and other pollutants reaching coastal zones together with nutrients
from insufciently treated point sources may pose risks for human consumers.
Human health protection aspects have been pursued through different routes for some time
in the EU, e.g. through establishing safety standards of bathing waters and sandy beaches,
which have recently been signicantly overhauled. Even more far-reaching at international
level has been to turn the traditional system of end-of-pipe safety standards for food into
process standards through the rules on sanitary and phytosanitary standards and measures,
now underlying international trade of food, including seafood.
38
There is also an on-going debate about seek-
ing more integration of environmental stan-
dards into the WTO disciplines currently
under negotiation in the Doha Round, par-
ticularly in relation to subsidies. This would
be a major step forward to overcome the
sectoral divisions which have hampered
more integrated approaches to ensure sys-
tem sustainability.
The OECD published a report in 2006
analysing China’s environmental perfor-
mance from the perspective of sustainable
development, because many of these envi-
ronmental issues have strong international
dimensions. It highlights the serious pollu-
tion problems of inland waters and coastal
zones, which pose health risks and start
having negative economic effects. Some 51
recommendations aim at improving perfor-
mance and recognising that economic tools
are starting to show some results. Addition-
ally, protected areas at different administra-
tive levels have increased signicantly in the
last twenty years, but marine and coastal ar-
eas are not sufciently represented and sub-
ject to excessive pressure.
During the 10th Five-Year Plan period the Chi-
nese government allocated $90.51 billion to
environmental expenditures, a 107% increase
over the previous planning period. The cur-
rent 10th Five-Year Plan period (2006-2010)
improves on this with an expected expendi-
ture of $174.05 billion, but will require signi-
cant improvement of enforcement down to
local levels to make the much needed decou-
pling of economic growth from environmen-
tal degradation more pervasive.
Environmental levies are also being explored
in China and some EU member states to
bring down CO2 discharge and improve other
air and water quality parameters. However,
they have not yet been proven to be very ef-
fective given the overall balance, and more
rapid progress would be desirable.
39
Trends in aquaculture development
Shellsh and nsh aquaculture has grown
rapidly over the last two decades and has
been largely responsible for the worldwide
increase in global aquatic production. Dur-
ing the rst half of the 1990’s, the annual
contribution of cultivated shellsh and n-
sh to total marine production increased
linearly from 12% to 19%, and presently
supplies 25% of sh consumption. A large
proportion of this growth is attributed to
Far Eastern countries, and particularly to
China, which now accounts for 89% of glob-
al aquaculture tonnage. Annual production
in Asia is in excess of 37 million tons, over
30 million of which are cultivated in China.
There has been a steadily increasing effort over the last decades to understand coastal zone
structure and dynamics, and the highly complex relationships between the state of coastal
zone ecosystems and the various pressures exerted upon them. Ecosystem effects such as
altered nutrient ratios, anoxic or hypoxic episodes, increased occurrences of nuisance and/
or toxic algal blooms, modied primary production patterns or abnormal mortalities of sh
and shellsh are all examples of undesirable changes in state, which are largely identied
with human pressure, and which have signicant economic costs.
Adapted from Bricker et al, 2007
Pressure State Response
worsening
improving
Pressure-state relationships within coastal systems are reasonably well understood at local
scales, such as those resulting from organic enrichment and deoxygenation of sediment
underlying sh cages, or discrete blooms of opportunistic algae linked to coastal sewage dis-
Concerns regarding ecological balance
and sustainability in areas used for the
culture of marine resources have gained
visibility in three ways:
From the shery perspective by the
occurrence of decreased yields, in-
cluding increased disease and mor-
tality rates
Changes in localised chemical and
biological structure and function
From a more holistic ecosystem per-
spective of eutrophication and toxic
algal effects
40
charges. However, at a system scale, our grasp of pressure-state interactions is less robust,
requiring integration in both space and time.
On a system scale, the various components which need to be included in an integrated
framework are also well researched. There is a body of literature on watershed uses and
pressure on coastal systems, describing the alterations in discharge of water, sediments
and dissolved nutrients. Models have been developed which incorporate watershed geomor-
phology, land cover and use, and hydrology.
The processes and parameters that determine the growth and survival of key coastal ecosys-
tem components such as shellsh and nsh are well described. Physical processes which
may be responsible for aggravating or mitigating the impacts of pressures on state, such as
residual current patterns, boundary exchanges and water residence time, vertical stratica-
tion, and sediment dynamics have been described and modelled in many coastal systems.
The socio-economic impacts of the changes in habitat of these various species are also
known.
Aquaculture operations nd it increasingly difcult to maximise their prots, unless, at least
for the short term, larger volumes are being demanded, which appears to be the general
case. However in some cases demand for certain species is on the decline due to changes in
taste and perceived quality of aquaculture, as for sea bass in some Chinese systems and for
salmon in Europe. Many of the environmental impacts of aquaculture operations are recipro-
cal, meaning that more production would lead to more impacts on the habitat, and thus to
more pressure on the ability to produce. The timing of these impacts determines when the
capacity of the habitat to provide the necessary services for maintaining ecosystem integrity
will be exceeded, thus compromising economically sustainable aquaculture.
Despite the accumulated knowledge on the various building blocks of a holistic framework
for integrated interpretation of coastal zone structure and dynamics, there are areas which
need to be addressed, particularly at the interfaces between ecological compartments and
scientic disciplines. Three broad topics are identied below, which have been specically
addressed by SPEAR, in order to progress beyond the current state-of-the-art.
41
Internal feedbacks: Study of internal feedbacks, e.g. multispecies inter-
actions, and how these can significantly affect the relationship between
pressure and state. The aim is to help optimise relative densities, distribu-
tions and species composition of cultured algae, shellfish and finfish, with
respect both to waste removal and harvest value.
Integrated models: Development of an integrated natural sciences-social
sciences approach which cross-cuts scaling issues and is capable of aggre-
gation, in order to bring the different (mostly known) parts together on a
multi-year scale.
Management tools: (a) A holistic approach where quantifiable environ-
mental health and socio-economic descriptors are used as management
metrics; (b) A screening model approach used for selecting key para-
meters, including derived parameters calculated using research models,
for system scale decision making; and (c) A combination of these into
practical tools for management.
European Commission, 2006a. Towards a future
maritime policy for the Union. A European vision for
the oceans and seas. Luxembourg, Ofce for Ofcial
Publications of the European Communities: 56 p.
European Environment Agency, 2006. The changing
faces of Europe’s coasts. Copenhagen, European
Environment Agency, Environmental Issue Report:
107 p.
European Environment Agency, 2001. Late les-
sons from early warnings 1896-2000. Copenhagen,
European Environment Agency Report, 6/2006: 210 p.
Humborg, C., Ittekkot, V., Cociasu, A. and Bodungen,
B. V., 1997. Effect of Danube River dam on Black Sea
biogeochemistry and ecosystem structure. Nature,
386: 385-388.
McIsaac, G.F., David, M.B., Gertner, G.Z. and
Goolsby, D.A., 2001. Eutrophication. Nitrate ux in
the Mississippi River. Nature, 414: 166-167.
Millennium Ecosystem Assessment, 2005. http://
www.millenniumassessment.org/en/index.aspx
Milliman, 1997. Blessed dams or damned dams?
Nature, 368: 325-326.
Nauen, C.E. (ed.) in collaboration with Bogliotti,
C., Fenzl, N., Francis, J., Kakule, J., Kastrissianakis,
K., Michael, L., Reeve, N., Reyntjens, D., Shiva, V.,
Spangenberg, J.H., 2005. Increasing impact of the
EU’s international S&T cooperation for the transi-
tion towards sustainable development. Luxembourg,
Ofce for Ofcial Publications of the European
Communities: 26 p.
OECD, 2006. Environmental performance review of
China. Conclusions and Recommendations. OECD
Working Party on Environmental Performance,
Beijing, 8-9 November 2006. http://www.oecd.org/
dataoecd/58/23/37657409.pdf
KEY REFERENCES
42
43
TOOLS
Summary
This chapter describes the various tools used during SPEAR for data storage, visualisation
and analysis, modelling, outreach and the economic concepts utilised in the project. In situ
data obtained prior to or during the project have been stored in the BarcaWin2000 database
to enable search and analysis. Satellite data were obtained from archives at higher resolution
for specic studies such as catchment modelling or mapping of aquaculture sites. Lower
resolution (~1-km) data were received from agencies such as the NASA, ESA and NOAA
agencies and processed in near-real time for continuous synoptic scale monitoring through-
out the project. In situ and satellite data were inserted into a Geographical Information
System (GIS) to enable geospatial data visualisation and analyses. Internet access to data
was provided through the main project web site and visualisation of satellite data through a
dedicated satellite image web site. Visualisation of in situ and satellite data was also investi-
gated using web-based systems such as the proprietary, commonly used GoogleEarth, and
more specialised web-based GIS systems such as Adobe SVG viewer. A variety of numerical
models enabled investigation of: nutrient inputs from agricultural and urban sources into
Sanggou and Huangdun bays (SWAT); the interaction of shellsh with ecosystem processes
(ShellSIM); the physical movement of water in the bay and its impact upon the biogeochem-
istry (Delt3D and Delft3D-WAQ models); aquatic ecology and management (EcoWin2000;
FARM); and eutrophic assessment/trophic status (ASSETS).
44
Introduction
This chapter describes the tools used to store and visualise data obtained prior to or during
SPEAR and the tools used to provide some inputs to the models. It also contains details of
outreach through web-sites and web-GIS.
Databases
A widely used relational database software (BarcaWin2000) was employed for water quality
data assimilation and analysis (Figure 11). The software provides:
FIGURE 11. Snapshot of the BarcaWin2000 database and description of main features
(http:// www.barcawin.com)
Organisation of information in a state-of-the-art relational model
Security for ve levels of user access
Easy import of data from formatted MS-Excel spreadsheets
Robust data entry validation
Data query outputs to MS-Excel
Open architecture and easy export to Oracle, SQL server, etc
Both historical and data collected in the context of the SPEAR were assimilated for the study
sites.
45
Remote Sensing
Landcover and aquaculture
classification and mapping
The SPEAR project used remote sensing
techniques to map the location of relevant
land cover features in the catchment area
surrounding the coastal regions under study,
as well as the relevant aquaculture struc-
tures in place in the systems. Images from
the Landsat satellite were used as a basis for
mapping. Relevant features were located in
the map using a supervised classication
method; the spectral signature patterns of
the satellite image were compared with those
characteristic for different types of landcover
and aquaculture, taken during a ground
truthing survey.
Near-real time data processing
Throughout the SPEAR project, near-real
time satellite data covering the Yellow Sea to
the east of China were provided from a num-
ber of satellites. The AVHRR and MODIS sen-
sors provide sea-surface temperature while
MODIS and MERIS provide observation of
ocean colour that can be related to phyto-
plankton chlorophyll or suspended particu-
lates. Data were obtained from various agen-
cies including NOAA for AVHRR and NASA
for MODIS by specic subscriptions estab-
lished for SPEAR; MERIS data were obtained
from the global ESA rolling archive. Data
products were usually available 7 hours (me-
dian estimate) after reception on-board the
spacecraft. Examples are shown in Figure
12: the Sea Surface Temperature (SST) im-
age (Figure 12a) shows warmer water to the
South, cooler to the North with mesoscale
variability. After processing, images were
placed on the web site (see below). Dur-
ing the 3.25 years of SPEAR approximately
5000 AVHRR scenes, 2300 MERIS and 2300
MODIS scenes were processed (represent-
ing about 6TBytes of raw data) providing
comprehensive day to day observation of
SST and chlorophyll a of the Chinese coast.
However, the coast is often covered with
cloud and haze restricting observation of
the sea-surface so composite images can be
more useful since they combine all the clear
components over a 7 day period.
46
FIGURE 12. Image composite for 3-9 September 2007 a) AVHRR SST; b) MODIS chlorophyll a and c) false-colour.
The chlorophyll a image (Figure 12b) shows apparently higher concentrations along the coast
but Figure 12c the false colour composite shows that the chlorophyll a estimates are probably
erroneous and affected by the very high suspended particulate levels along the coast.
FIGURE 13. Example web site screenshot of sea-surface temperature for late June 2007.
47
Time series
Time series of sea-surface temperature and ocean colour were available during SPEAR and
were used to investigate temporal changes offshore of the two bays. SST from the AVHRR
instrument has been available since 1981 providing a 27 year time series whereas ocean
colour from SeaWiFS, MODIS and MERIS together provide a ten year series (1997 – date).
Figure 14 shows monthly SST offshore of Sanggou Bay from 1985 to 2006 extracted from the
NASA Pathnder 9-km dataset. The annual cycle can be seen with varying maximum and
minimum annual temperatures. Removing the long-term monthly mean produces residual
(anomaly) SST and these show a general warming between 1985 and the late 1990’s with a
decline since then.
Temperature at 37.1° N 122.7° E
Average SST in 3x3 box
Temperature, C
SST anomaly, K
4
3
2
1
0
-1
-2
-3
-4
1985 1990 1995 2000 20051985 1990 1995 2000 2005
30
25
20
15
10
5
0
FIGURE 14. time plots of SST and SST anomaly West of Sanggou Bay
48
Geographical Information Systems
Arc/Info
A commercial GIS software package was used to store and analyse the spatial data acquired
in the project (Figure 15).
FIGURE 15. Example of GIS project in ArcGIS.
The use of spatial data is very valuable for integrated modelling projects, for:
Denition of spatial domains, such as model boxes.
Spatial visualisation of water quality data, by loading the GIS project with databases such
as BarcaWin2000.
Data extraction using common GIS functions (reclassication of grid cells, geostatistical
analysis, map algebra), for:
Area and volume calculations
Assessment of benthic biodiversity
Distribution of aquaculture features
General spatial data visualisation
49
Web–based visualisations
Google Earth
GoogleEarth is a visualisation tool for geospatial data that has signicant worldwide usage
and easily accessed. It is proprietary technology as opposed to open source. As a demonstra-
tion example SPEAR data were incorporated as layers in GoogleEarth (see Figure 16). Figure
16a shows a high resolution (250m) MODIS radiance image showing Sanggou Bay. Figure
16b shows an oblique view of the land cover classication draped over GoogleEarth with
exaggerated topography. Figure 16c and Figure 16d present two approaches for integrating
and visualising point data in GoogleEarth: the rst (Figure 16c) shows sampling locations
used in SPEAR for surface samples of ammonium and the capability developed at PML to
link points to a simple database to extract the values pertinent to the point. Figure 16d shows
chlorophyll a concentration measured in situ at the same locations as Figure 16 through two
methods, height and colour of a column viewed obliquely. A time plot was also produced
that showed how the chlorophyll a measured through the year changed at each location.
These gures, although only for investigation purposes, demonstrated the value of providing
project data in geospatial form for wider dissemination and outreach.
FIGURE 16. Example screenshots from GoogleEarth.
50
Web catalogue of SPEAR spatial data
The SPEAR spatial data, which range from simple point location of the sampling stations to
land use classications of the watershed were consolidated in a web map catalogue available
at http://www.biaoqiang.org/gis (Figure 17).
FIGURE 17. Snapshot of the SPEAR web-GIS.
51
Project Web site
Main project websites
A number of web-based models and tools have been developed by the SPEAR consortium
and are described below.
Site Description
Ofcial website:
http://www.biaoqiang.org/ in
English
http://www.spear.cn/ in Chinese
Includes information about the SPEAR project, study sites, the consor-
tium, a document retrieval zone and useful links.
There is a public website, aimed for a wide audience and a restricted
website aimed for SPEAR partners
FARM screening model:
http://www.farmscale.org/ The FARM screening model is an example of how carrying capacity of a
shellsh farm can be modelled with respect to production and environ-
mental impact
MOM model:
http://ancylus.net/ MOM screening model estimates the holding capacity of a sh farm
together with the outputs of particulate and dissolved nutrients from
production.
SPEAR remote sensing web site:
http://www.npm.ac.uk/rsg/proj-
ects/mceis/ys
Web site that includes near-real time ocean colour and sea-surface tem-
perature data and an archive of data from the start of SPEAR.
SPEAR GIS data catalogue:
http://www.biaoqiang.org/gis/ Web GIS browser that includes SPEAR spatial data
52
Models
S WAT
The Soil and Water Assessment Tool (SWAT) catchment model was used to simulate nu-
trient inputs from agricultural and urban sources into Sanggou and Huangdun bays. The
model simulates processes such as vegetation growth (taking into account agricultural and
grazing activities), river ow, soil erosion and nutrient transport from elds and wastewater
discharge points into the bays. The physical equations which form the backbone of SWAT
allow its application to investigate scenarios of climate, land use and agricultural manage-
ment changes in order to predict consequences for water discharge, nutrient and sediment
loadings to aquatic systems.
ShellSIM
To model the complex feedbacks, whereby mussels and oysters interact with ecosystem
processes, experimental measurements of physiological responses were undertaken in each
species over conditions that spanned full normal ranges of food availability and composition
in each bay.
53
Mathematical equations were then derived that dene functional inter-relationships between
the component processes of growth, integrating those interrelations within a dynamic model
structure (ShellSIM) developed to simulate time-varying rates of individual feeding, metabo-
lism and growth in these and other species.
Delft3D-FLOW
The Delft3D-FLOW hydrodynamic model was used to simulate the tidal, wind and ocean
currents in the study areas. This ne-grid model provides a detailed description of the cir-
culation, and is coupled with other models to provide an appropriate description of mass
transport for detailed water quality models such as Delft3D-ECO and for broader-scale mod-
els such as EcoWin2000.
Delft3D-WAQ/ECO
The Delft3D-ECO model has been adopted for detailed simulation of water and sediment
quality as well as algae growth and species composition. The model links dynamic ow elds
simulated with Delft3D-FLOW to water quality processes and the algal primary production
optimizing sub-model BLOOM. An important feature of Delft3D-ECO is that it explicitly
computes sediment and pore water quality in bottom sediment, which allows for taking into
account sediment-water interaction optimally. Water and sediment quality processes con-
cern organic matter, nutrients, dissolved oxygen, suspended sediment, salinity and several
other inorganic substances. The model also contains sub-models for microphytobenthos
and grazers like shellsh. For the present study shellsh have been imposed as forcing func-
tions for shellsh biomass.
CoBEx-ECO
CoBEx-ECO was used to simulate concentrations and uxes of water, salinity, nutrients, car-
bon and phytoplankton and bivalve shellsh in Huangdun Bay. In addition, the model cal-
culates the impact of sh farms on the water shed. The marine system was simulated using
four coupled basins which are horizontally homogenous but vertically resolved in density
layers. The model uses well-founded empirical formulations for the physical, biological and
chemical processes and is suitable to assess the impact on water quality by loads from land,
sh farms and exchange with adjacent seas.
EcoWin2000
EcoWin2000 is an ecological model for aquatic systems, developed using an object-oriented
approach. It resolves hydrodynamics, biogeochemistry and can incorporate population dy-
namics for target species. The various components consist of a series of self-contained ob-
jects, rather than multiple sub-models.
54
The EcoWin2000 model consists of two basic parts: a shell module and “ecological” objects.
The shell is responsible for communication with the various objects, for interfacing with the
user, supplying model outputs and general maintenance tasks.
Objects have “attributes” (variables) and “methods” (functions) – see Figure 18.
Each object groups together related state variables, and may at any time, be extended to
contain a new state variable without affecting the code of any other part of EcoWin2000.
Object Sample attributes Typical active methods Typical passive methods
Transport Salt Advection-diffusion -
Dissolved
substances Forms of DIN, PO43-, SiO2,
D.O. Nitrication, formation of
particulates Mineralisation of detritus,
exudation
Phytoplankton Phytoplankton, toxic algae Production, respiration,
senescence, exudation,
production of toxins
Grazing by zooplankton,
sh, benthic lter-feeders
Phytobenthos Microalgae, macroalgae, salt
marsh ora Production, respiration,
senescence Grazing by zooplankton,
sh, harvesting of seaweeds
Zooplankton Zooplankton, copepods Eat, grow reproduce,
excrete, natural mortality,
swim, settle (for benthic
larvae)
Predation by other objects
and within the object
Zoobenthos Filter-feeders, deposit-
feeders Filter, grow, reproduce,
excrete, natural mortality,
swim, settle (for benthic
larvae)
Fisheries, predation by sev-
eral other objects
Nekton Fish, large - invertebrates
(e.g. Sepia) Hunt (including select),
grow, reproduce, excrete,
natural mortality, swim,
migrate
Fisheries, hunting by birds
Man Various socio-economic
attributes Seed and harvest shellsh -
FIGURE 18. Attributes and methods (active and passive) for some objects of EcoWin2000 modelling platform.
Similarly, the methods which control interactions among state variables within objects may
be easily changed, due to inheritance (which is a property of object-oriented programming
languages).
55
FIGURE 19. Screenshot of the EcoWin2000 model, as applied to a SPEAR bay.
EcoWin2000 uses a range of equations depending on the application requirements, and
may be used as a research model to examine nutrient loading and aquaculture development
scenarios. It has been extensively tested, and is a potentially useful tool for supporting an
ecosystem approach to sustainable aquaculture development.
In the SPEAR project, the EcoWin2000 modelling platform was used to implement an eco-
logical model for each bay to estimate aquatic production and simulate relevant manage-
ment scenarios. The main features modelled for these systems were the hydrodynamics,
suspended matter transport, nitrogen cycle, phytoplankton and detrital dynamics, shellsh
growth and human interaction.
FARM
The Farm Aquaculture Resource Management (FARM) model is a web-based tool for as-
sessment of coastal and offshore shellsh aquaculture at the farm-scale. directed both at
the farmer and the regulator, and has three main uses: (i) prospective analyses of culture
location and species selection; (ii) ecological and economic optimisation of culture practice,
such as timing and sizes for seeding and harvesting, densities and spatial distributions (iii)
environmental assessment of farm-related eutrophication effects (including mitigation).
The modelling framework applies a combination of physical and biogeochemical models,
bivalve growth models and screening models for determining shellsh production and for
eutrophication assessment. Shellsh species combinations (i.e. polyculture) may also
be modelled.
56
ASSETS
The Assessment of Estuarine Trophic Status (ASSETS) evaluates inuencing factors, overall
eutrophic condition and future outlook, and combines them into a single overall rating called
ASSETS. Each of the component ratings is determined using a matrix approach.
Inuencing factors (IF) is a combination of a system’s natural susceptibility (i.e. ushing
and dilution characteristics) and the nutrient load to the system. Loads are estimated as
the ratio of land (i.e. human-related) and ocean based inputs.
Overall eutrophic condition (OEC) is a combined assessment of ve symptoms based on
occurrence, spatial coverage and frequency of problem occurrences. The rating is deter-
mined from a combination of the average scores for chlorophyll and macroalgae, primary
symptoms indicating the start of eutrophication, and the worst score of the three more
serious secondary symptoms (dissolved oxygen, submerged aquatic vegetation, and nui-
sance/toxic algal blooms).
Future outlook (FO) predicts what future eutrophic conditions will likely be by combining
susceptibility and expected changes in nutrient loads to determine whether conditions
will worsen, improve, or remain the same.
The ASSETS synthesis combines the IF, OEC and FO ratings into a single score falling into
one of ve categories that are colour coded following international convention: “High”,
Good”, “Moderate”, “Poor”, or “Bad”.
Borja, A., Bricker, S.B., Dauer, D.M., Demetriades,
N.T., Ferreira, J.G., Forbes, A.T., Hutchings, P., Jia,
X., Kenchington, R., Marques, J.C., Zhu, C.B., 2008.
Overview of integrative tools and methods in as-
sessing ecological integrity in estuarine and coastal
systems worldwide. Mar. Pol. Bull., In Press.
Bricker, S.B., Ferreira, J.G., Simas, T., 2003. An
Integrated Methodology for Assessment of Estuarine
Trophic Status. Ecological Modelling 169 (1): 39-60.
Ferreira, J.G., 1995. EcoWin - An object-oriented
ecological model for aquatic ecosystems. Ecol.
Modelling, 79, 21-34.
Franco, A.R., Ferreira, J.G., Nobre, A.M., 2006.
Development of a growth model for penaeid shrimp.
Aquaculture, 259, 268-277.
Neitsch SL, Arnold JG, Kiniry JR, Williams JR, Kiniry
KW, 2002. Soil and Water Assessment Tool theoreti-
cal documentation. TWRI report TR-191, Texas Water
Resources Institute, College Station.
Nobre, A.M., Ferreira, J.G., Newton, A., Simas, T.,
Icely, J.D., Neves, R., 2005. Management of coastal
eutrophication: Integration of eld data, ecosystem-
scale simulations and screening models. Journal of
Marine Systems, 56 (3/4), 375-390.
Shutler, J.D., Smyth, T.J., Land, P.E., and Groom,
S.B., 2005. A near real-time automatic MODIS data
processing system, International Journal of Remote
Sensing, 25 (5): 1049-1055.
KEY REFERENCES
57
58
SYSTEMS
Summary
The SPEAR project uses two contrasting coastal study sites; Sanggou Bay in a northern rural
area and Huangdun Bay in a heavily industrialized area with substantial human pressure on
both local and regional levels. Huangdun Bay is located in Zhejiang Province, with the largely
industrialised centre of Ningbo City, while Sanggou Bay is in Shandong Province, within the
jurisdiction of the smaller city Rongcheng. Weihai is the closest larger city with a population
of 2.5 million. Both provinces are known for their valuable marine resources. The culture
production is 263 500 and 50 000 ton y-1 in Sanggou Bay and Huangdun Bay respectively.
In China as a whole 4.2 million people are employed in the sh farming (inland and marine)
business. This means that almost 19 direct sh farming jobs are created per RMB 1 million
of value in sh farming. Comparative statistics for the provinces were not available, but for
marine farming only in Sanggou Bay this gure is around 9 direct jobs per RMB 1 million in
total marine farming value.
Sanggou Bay is in temperate climate zone and rainfall and runoff are concentrated in the
summer months. The Huangdun Bay area has a humid subtropical climate with constant
rainfall throughout the year and a typhoon season in summer. The main sources of nutri-
ent loads into Sanggou Bay are cropland fertilisation (65%) and urban wastewater (35%).
The Huangdun Bay catchment is mostly forested (65%), with a signicant presence of rice
croplands (20%).
59
Huangdun Bay is a semi-enclosed estuary with a tidally dominated coastal water exchange
with the East China Sea, through a connecting channel, the Xiangshan Gang. The water col-
umn is almost vertically homogenous due to the tidal mixing but run off inuence is seen
in the longitudinal salinity gradient, where the salinity at the head of the bay is lower than at
the mouth. Sanggou Bay is semi-circular bay with an open boundary to the sea. The water
exchange is also here chiey forced by the tides, and the bay is well-mixed, both horizon-
tally and vertically. The residence time for Huangdun is in the order of 1-4 months and for
Sanggou Bay of 5-20 days. To asses the state of the marine systems in Huangdun Bay, nine
stations were designed for water quality investigation. Water samples were taken at the sur-
face and bottom of the water column. Sediment samples were taken both in the basin and in
the intertidal areas. The survey was conducted during May, 2005 to April, 2006.
60
Introduction
The SPEAR project uses two contrasting coastal study sites. Sanggou Bay is an open bay with
rapid water exchange in a rural area in the north, whereas Huangdun Bay, south of Shanghai,
is located in a heavily industrialized area with substantial human pressure on both local and
regional levels. It also has a more limited exchange with coastal waters. Both bays are exten-
sively used for marine culture of nsh, shellsh and algae (Figure 20).
Lat.: 37° N
Long.: 122° E
Lat.: 30° N
Long.: 122° E
Sanggou Bay
Huangdun Bay
FIGURE 20. Location of the coastal study sites.
Data sets and surveys
Marine system
One of the priorities in the SPEAR project was the collection of existing water quality re-
cords in both study sites. In Sanggou Bay there was already a large number of water quality
sampling campaigns (Figure 21), coupled with sediment quality, intertidal and oyster culture
sampling campaigns. These data were compiled from different data sources and introduced
into a common project database. The historical dataset for Sanggou Bay was sufcient to
study the characteristics of the system, and therefore no additional sampling campaigns
were needed.
61
Kilometers
0 1 2 4 6 8 10
1983-1984
1983-1984, 1989-1990
1993-1994
1999-2000
2003-2004
Sampling campaign
FIGURE 21. Map of sampling stations in Sanggou Bay.
In contrast with the previous study site, data
for Huangdun Bay prior to the project start
were scarce. During SPEAR, data were col-
lected to investigate water and sediment
quality in the system (Figure 23).
Nine stations were dened for water quality
investigation (Stations 1-9; see Figure 22).
Surface and bottom samples were collected
in all stations except station 3, which also
had a mid-water sample. In total 19 water
samples were collected monthly. The survey
lasted from May, 2005 to April, 2006.
A total of 13 parameters for seawater were
analysed: water depth, water temperature,
salinity, transparency, pH, suspended par-
ticulate matter (total suspended particulate
matter, organic matter, inorganic matter),
ammonium, nitrate, nitrite, soluble reactive
phosphate, soluble silicate, phytoplankton di-
versity and abundance, primary production.
62
A3
A2
A1 B1
B2
9
8
7
3
2
6
1
4
5
Xiangshan Bay
29° 30' 17''
29° 26' 40''
121° 29' 39'' 121° 33' 56'' 121° 38' 12''E
N
FIGURE 22. Map of sampling stations in Huangdun Bay and Xiangshan Gang.
Nine stations were also dened for a sediment quality investigation, which used the same
locations as the water quality investigation. A total of 11 parameters for sediment were analy-
sed: grain size, water fraction, organic nitrogen, inorganic nitrogen, total nitrogen, inorganic
phosphorus, organic phosphorus, total phosphorus, total organic carbon, diatom diversity
and abundance.
Two sections, A and B, were designed for investigation of the intertidal zone (Figure 22). Sur-
face sediment samples (top 10 cm) were collected in the high, medium and low tide zones.
Additionally, water samples were collected in the low tide zone. A total of 12 parameters were
analysed: grain size, water fraction, organic and inorganic nitrogen and phosphorus, total
nitrogen and phosphorus, total organic carbon, primary productivity, chlorophyll a, diatom
diversity and abundance.
63
FIGURE 23. Water quality measurements in Huangdun Bay.
For both bays data were stored in the BarcaWin2000 database (Figure 24).
Database table Sanggou Bay Huangdun Bay
Historical data Historical data SPEAR data
Stations 86 11 20
Parameters 137 33 487
Campaigns 95 16 24
Samples 23 621 207 225
Results 55 384 2 024 5 594
FIGURE 24. Properties of the water quality databases for Sanggou and Huangdun bays.
Catchment survey
The SPEAR project also investigated the distribution of land cover in the watershed around
the Sanggou and Huangdun bays. Land cover affects rates of runoff and the concentrations
of nutrients and pollutants in the water as it drains into a bay. Initial estimates were obtained
from remote sensing using low spatial resolution classications, for example, from global
MODIS data, but these did not provide the level of detail necessary for this study so higher
resolution satellite data was required.
There are various high resolution satellite sensors such as SPOT and Landsat. The very high
resolution, less than 10m, data are generally very expensive. For this reason Landsat ETM+
data were chosen as they provided 30m ground resolution with visible and infra-red bands.
It was also possible to test the method using free historical scenes for the study sites from
NASA’s Global Land Cover Facility (GLCF).
64
The classication was tested on a pre-project image (from 2000) and applied to scenes pur-
chased covering the project study period with good image quality, low cloud cover and full
coverage for the study region including the watershed around the bays. Scenes on 18 May
2005 for Sanggou Bay and 28 June 2005 for Huangdun Bay were chosen. The images were
purchased from the USGS website (http://glovis.usgs.gov/) and the gap-ll product was cho-
sen. Fortunately, the gap lled regions are fairly minimal around the areas of interest.
Land classication was generated using the supervised classication tool in ENVI. This pack-
age can import the geo-referenced Landsat scenes and display the various bands. The rst
draft classication was implemented by visually dening training regions for regions such as
water, cloud, urban areas and forest as these are fairly easy to identify.
To improve the accuracy of the classication it was also necessary to get as much ground
truth data as possible. This was provided by Chinese partners from who went into the eld
with GPS units to make observations. Additionally, it was necessary to remove hill-shading
effects from the classication. The sun-scene-satellite geometry at the time of image cap-
ture gives rise to shading effects in hilly regions. This effect was countered by generating a
vegetation index, which takes account of solar shading and providing this as an input to the
classication.
0 1 2 4 6 8 10
Kilometers
Aquaculture
Urban
Cropland
Orchard
Forest
Water
Wetland
Bare wetland
Wetland
Burnt land
Forest
Cloud
Urban
Cropland
Low tide water
Water
Rice
Shrimp pond
0 2 4 8 12 16 20
Kilometers
FIGURE 25. Final land classications cropped down for SWAT for Huangdun Bay (left) and Sanggou Bay
(right)
65
FIGURE 26. Part of the NDVI scene generated for Huangdun Bay to improve the land classication.
This highlights the forest/crop regions, whereas urban and bare regions are darker.
The rst classication had some noticeable problems especially with the over-representation
of rice cropland. Additionally, some of the classes seemed to be following the shape of hill-
shading features, especially around Huangdun Bay. To counter the rst problem extra ground
truth was requested from a few regions, especially where forest/crop lands were confused.
For the hill-shading a vegetation index image was created for each bay and added to the clas-
sication process. These updates made a large improvement to the land classication.
In Huangdun bay, this classication was complemented with a river water quality sampling
campaign, running from July 2005 to June 2006. Water quality was sampled in the rivers
Dajia, Fuxi and Yangong; monthly samples were taken of 12 biogeochemical parameters in
two points at each river, including temperature, salinity, pH, dissolved oxygen, ow rate,
chlorophyll a, total nitrogen, NH3, NO3, total phosphorus, PO4 and silicate.
66
Catchment description
The catchment areas draining into Sanggou and Xiangshan bays present contrasting charac-
teristics in terms of climate, land cover and nutrient export to the coastal systems.
200
180
160
140
120
100
80
60
40
20
0
30
25
20
15
10
5
0
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Temperature (°C)
Rainfall and Runoff (mm)
Rainfall
Runoff
Average daily temperature
200
180
160
140
120
100
80
60
40
20
0
30
25
20
15
10
5
0
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Temperature (°C)
Rainfall and Runoff (mm)
Rainfall
Runoff
Average daily temperature
FIGURE 27. 30-years climate and runoff normals for the catchment areas of Sanggou (left) and Xiangshan
bays (right).
Sanggou has a temperate climate with warm summers (Figure 27); rainfall and runoff are
concentrated in the summer months, leading to peak river ows in this period. Xiangshan
has a humid subtropical climate (Figure 27), with constant rainfall throughout the year and a
typhoon season in summer. This leads to a constant runoff into the coastal system through-
out the year, although the occurrence of typhoons can lead to signicant peaks e.g. ty-
phoon Matsa in August 2005 led to 800 mm rainfall and an estimated 600 mm runoff in
four days, which also led to a peak in suspended matter and nutrient inputs.
The Sanggou bay catchment is mostly covered by wheat and maize croplands (68%; Fig-
ure 28), resulting in a signicant specic nutrient export from diffuse sources (Figure 29).
This is added to the wastewater generated by the main urban center – Rongcheng city, with
200 000 inhabitants contributing to the hydrological network – of which 2/3 is discharged
in wetlands for natural treatment, and the remaining is untreated. Overall, the main sources
of nutrient loads into Sanggou bay are cropland fertilisation (65%) and urban wastewater
(35%). Peak discharges are in summer, with 50% of nitrogen and 75% of phosphorus reach-
ing the system in dissolved inorganic forms.
67
68%
12%
2%
11%
7% Cropland
Urban
Forest
Orchard
Wetland
Urban
Forest
Wetland
Burnt land
Cropland
Rice
8%
61%
1%
4%
6%
20%
FIGURE 28. Landcover in the catchment areas of Xiangshan (left) and Sanggou bays (right).
The Xiangshan bay catchment is mostly forested (61%; Figure 28), with a signicant pres-
ence of rice croplands (20%); the subset draining into Huangdun bay has similar distribu-
tions. This still results in a high amount of agricultural pollution, due both to the high runoff
and the nutrient exports from forest litter decomposition (c. 30% of total diffuse pollution
exports; Figure 29). The wastewater generated by the urban population – 420 000 inhabit-
ants, mostly living in and around Ninghai city – is discharged mostly into Huangdun bay; al-
most 2/3 is treated, while the remainder is expected to be delivered to wastewater treatment
plants already under construction.
Overall, the main sources of nutrients into Xiangshan Gang are urban wastewater (55%),
cropland fertilisation (25%) and forest residue (15%). The picture is slightly different in
Huangdun bay, with a greater relative importance of cropland fertilisation (30%) and forest
residue (20%) at the expense of urban wastewater (40%). The discharges occur mostly dur-
ing spring, summer and autumn, with c. 50% of the nitrogen and 60% of the phosphorus
reaching the system in dissolved inorganic forms.
68
Parameter Sanggou Bay Xiangshan Bay Huangdun Bay
Rainfall 735 mm.y-1 1310 mm.y-1 1310 mm.y-1
Runoff 384 mm.y-1 705 mm.y-1 705 mm.y-1
2 x 108 m3.y-1 1 x 109 m3.y-1 5 x 108 m3.y-1
Nitrogen loads Diffuse 9.0 kg.ha-1.y-1 20.0 kg.ha-1.y-1 20.0 kg.ha-1.y-1
520 ton.y-1 2800 ton.y-1 1240 ton.y-1
Point-source 430 ton.y-1 650 ton.y-1 650 ton.y-1
Tot al 950 ton.y-1 3450 ton.y-1 1890 ton.y-1
Phosphorus
loads Diffuse 2.5 kg.ha-1.y-1 3.5 kg.ha-1.y-1 3.5 kg.ha-1.y-1
145 ton.y-1 480 ton.y-1 200 ton.y-1
Point-source 90 ton.y-1 135 ton.y-1 135 ton.y-1
Tot al 235 ton.y-1 615 ton.y-1 340 ton.y-1
FIGURE 29. Rainfall, Runoff and nutrient loads in the coastal systems.
Marine system
The Xiangshan Gang, and its inner part Huangdun Bay, is a semi-enclosed estuary with
coastal exchange with the East China Sea (Figure 30). The exchange is dominated by tidal
ows. The tides are of a semi-diurnal character and the amplitude of the tidal wave is ampli-
ed from mouth to head in the bay due to the decreasing cross-sectional area. The water
column is almost vertically homogenous due to the tidal mixing; the little freshwater input to
the area is not enough to inuence the vertical stratication. Inuence of river run off is, how-
ever, seen in the longitudinal salinity gradient, where the head of the bay displays the lowest
salinity. Both temperature and salinity displays an annual cycle with peak values during the
summer months. Although the average condition of the bay is well-mixed, there are periods
where the water column is more stratied which enhance the biological production. The
nitrate to phosphate (N/P) ratio is about 30 throughout the year, with a mean phosphorus
concentration of about 2.3 mol.L-1. Primary production is therefore limited by phosphorus.
69
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
-22
-24
-26
-28
-30
30
29.8
29.6
29.4
121.6 121.8 122 122.2
FIGURE 30. Bathymetric map of Xiangshan Gang
and its inner part Huangdun Bay.
Sanggou Bay is a semi-circular bay with an
open boundary to the Yellow Sea (Figure
31). There also, the water exchange is chiey
forced by the tides. The bay is well-mixed,
both horizontally and vertically. The tem-
perature varies from a few degrees in winter
to about 25ºC in summer. The annual varia-
tion in salinity is on the order of one psu.
The nutrient data in Sanggou Bay indicates
a long-term increase. Concentrations display
a seasonal cycle with minima during sum-
mer. The N/P ratio has increased more than
tenfold from 1984, with values at about 5, to
1994 where values of 20-40 have been re-
corded.
FIGURE 31. Map of Sanggou Bay. The height and
depth contours are 1 m apart. The intertidal area of
the bay is marked by light blue colouring.
70
Some properties of the SPEAR systems are summarized in Figure 32.
Sanggou Bay Xiangshang Gang Huangdun Bay
Volume 9 x 105 m32.7 x 109 m31.6 x 107 m3
Area 130 km2511 km26.2 km2
Max Depth 15 m 30 m 13 m
Mean Depth 8 m 9 m 3 m
Catchment area 590 km21430 km2625 km2
Temperature 0-25°C 8-30°C 8-30°C
Salinity 30.0-32.6 psu 24-29 psu 24-26 psu
Mean Salinity 31.83 psu 26.26 psu 25.44
Freshwater input 10 m3.s-1 32 m3.s-1 15 m3.s-1
Tidal RangeNeap/spring 0.6 m/1.36 m 1.49 m/3.85 m -
Residence Time 5-21 days 1-4 months -
FIGURE 32. Summary of SPEAR system properties.
Data from the existing databases for both bays, from the SPEAR survey in Huangdun Bay and
from modelled loads from land, waste water, sh farms and shrimp and crab ponds were used
to assess the integrated net effects of the biogeochemical processes in the systems. This can
be a useful way to gain information on the export, import, sinks and sources of biogeochemi-
cally conservative and nonconservative substances in the systems. The resulting nutrient bud-
gets for Sanggou Bay are shown in Figure 33 and for Xiangshang Bay in Figure 34.
236
284
Boundary
PO4-P=0.63
249PO4-P=0.54
35
620 338
kelpshellfishfinfish
996
sink/source
309
949
822
Boundary
tot-N=4.04
1685tot-N=8.10
863
4185 2282
kelpshellfishfinfish
5904
sink/source
-477
Sanggou Bay total phosphorus balance (ton y-1) Sanggou Bay total nitrogen balance (ton y-1)
FIGURE 33.Nutrient budgets for Sanggou Bay.
71
The nutrient loads from the sh farms are
quite signicant. The same order of magni-
tude is bound to the seaweed and lter feed-
er production. Measurements of DIN dur-
ing the 80’s and 90’s indicate that primary
production in the area was nutrient limited
and that therefore the bivalve shellsh pro-
duction was nutrient limited. Between 2003
and 2004 the nsh production increased
to 15 000 ton y-1, which increased the nutri-
ent loads to the area to about 3720 ton N
y-1. At the same time the oyster production
increased to about 60 000 ton . y-1. This pro-
duction would need about 1620 ton N y-1. The
combined information indicates that shellsh
production before 2003 was nutrient limited.
However, recent data show declining growth
of oysters and increasing nutrient concentra-
tions which implies a shortage of food due to
grazing control of primary production.
The tidal ows in Xiangshan Gang are large
and the properties of the inowing water
thereby have a large input on the system.
There are few data available from the adja-
cent sea which introduces uncertainties to
the budgets. It is also plausible that there are
some unaccounted loads to the system.
72
648
570
Boundary
tot-P~1.09
603
497
Tie
Tot-P=2.40
shellfish
44-24 78
95
Huangdun
Tot-P=2.55
78-100 106
10
198-218
4535
3347
Wushe
Tot-P=2.17
-804--824 1187
64
243-263
2776
1833
Xize
Tot-P=1.62
439-524 943
175
370-455
194-216
sink/source
Total Phosphorus balance (ton . y-1)
System: -136--282
8596
6892
Boundary
tot-N~46
7536
6013
Tie
Tot-N=70.49
shellfish
-1043--1163 1704
568
Huangdun
Tot-N=70.55
-390--505 1523
63
1109-1229
5484
5381
Wushe
Tot-N=58.12
4147-4267 103
385
1408-1528
44624
34937
Xize
Tot-N=57.67
-7945--8459 9687
1050
2175-2689
1091-1206
sink/source
Total Nitrogen balance (ton/year)
System: -5111--5980
FIGURE 34. Nutrient budget for Xiangshan Gang.