Bioaccumulation of persistent organic pollutants
from floodplain lake sediments:
linking models to measurements
Prof. dr. A.A. Koelmans
Waterkwaliteitsbeheer, Wageningen Universiteit
Dr. J.J.G. Zwolsman
Adviseur waterkwaliteit, KIWA Water Research, Nieuwegein (voorheen RIZA).
Prof. dr. J. de Boer
Prof. dr. H.J.P. Eijsackers
Prof. dr. ir. D. van de Meent
Prof. dr. ir. W.H. Rulkens
Dit onderzoek is uitgevoerd binnen de onderzoekschool SENSE.
Hoogleraar bij de leerstoelgroep Aquatische Ecologie en
Vrije Universiteit Amsterdam
Radboud Universiteit Nijmegen
Bioaccumulation of persistent organic pollutants
from floodplain lake sediments:
linking models to measurements
Caroline T.A. Moermond
ter verkrijging van de graad van doctor
op gezag van de rector magnificus
van Wageningen Universiteit,
Prof.dr. M.J. Kropff,
in het openbaar te verdedigen
op maandag 10 september 2007
des namiddags om half twee in de Aula.
CIP-data Koninklijke Bibliotheek, Den Haag
Bioaccumulation of persistent organic pollutants from floodplain lake sediments:
linking models to measurements / Moermond, C.T.A.
Thesis Wageningen University, Wageningen, The Netherlands – with references –
with summary in Dutch.
The main research questions of this research were (1) what is the extent and nature of
bioavailability of sediment-bound polychlorobiphenyls (PCBs) and polyaromatic
hydrocarbons (PAHs) and (2) what are the effects of lake ecosystem structure on fate
and bioaccumulation of PCBs and PAHs. Fast-desorbing fractions in the sediment of
floodplain lakes were estimated by the 6-h Tenax-extractable fractions with a
correction factor. These fractions varied between 1 and 40% and did not show a
clear trend with log KOW. This means that contaminants in these sediments were
available, but to a smaller extent than total concentrations would suggest. The 6-h
Tenax extractable concentration often correlated better with bioaccumulation than
the total extractable concentration in sediment. Despite the reduced availability,
benthivorous fish and invertebrates in floodplain lakes still rapidly accumulated
substantial amounts of PCBs. PAHs were accumulated relatively less because PAHs
were relatively less available than PCBs due to their stronger sorption to
carbonaceous materials, also referred to as soot or black carbon. For fish, metabolic
transformation caused even lower PAH concentrations. Contaminants that have been
present in the sediment for longer periods of time (years to decades), were less
available for Tenax extractions as well as for uptake by biota in different parts of the
food web than contaminants that were recently added. Thus, aging may translate
directly into reduced uptake at higher trophic levels. Nutrient additions in enclosures
with benthivorous fish had a positive effect on PCB accumulation by these fish.
Measured bioaccumulated concentrations of PCBs and PAHs in invertebrates in flood
plain lakes were not influenced greatly by seasonal effects or ecological structure.
Although effects were statistically significant, their magnitude in terms of
accumulation factors was small, which may have been caused by the similar
sediment composition and bioavailability of contaminants in our systems. Differences
between compounds were much larger than differences due to ecosystem structure,
seasons, or species composition. As for total masses of PCBs and PAHs in certain
compartments however, lake ecosystem structure appeared to have a large influence
on the biomass of biota and therefore also on the mass distribution of PCBs and PAHs
in biotic compartments. Thus, changes in ecosystem structure strongly influenced
PCB and PAH dynamics, although concentrations within the biotic compartments
were not significantly influenced by biotic biomass. As for bioaccumulation
modelling, when aquatic exposure concentrations were quantified accounting for
sorption to carbonaceous materials, model results improved substantially. Including
metabolic transformation and sediment uptake in the model accounted for a further
improvement of the model fit. Implications are discussed for food chain
bioaccumulation modeling, bioavailability assessment, sediment policy making and
floodplain lake management.
Table of Contents
Chapter 2. Uptake of sediment-bound bioavailable
polychlorobiphenyls by benthivorous carp (Cyprinus
Chapter 3. Black carbon and ecological factors affect in situ BSAFs for
hydrophobic organic compounds in flood plain lakes.
Chapter 4. Aging limits bioavailability and bioaccumulation of PCBs
and PAHs in model ecosystems with different ecological
Chapter 5. Impact of macrophytes and fish on partitioning of sediment-
originating PCBs and PAHs in freshwater model ecosystems.
Chapter 6. Modeling decreased food chain accumulation of PAHs due
to strong sorption to carbonaceous materials and metabolic
Addendum: Steady state modeling of PCB accumulation in
aquatic food chains.
Chapter 7. Summarizing discussion
Background, research questions and thesis outline
Contaminated sediments are a major problem in many countries, and may limit
development of ecosystem health . At present, to determine if sediment meets
quality criteria, total concentrations of sediment-bound contaminants are analyzed.
However, particular fractions of sediment-bound toxicants seem to be unavailable to
organisms, irrespective of the uptake route [2-5]. Consequently, bioavailable
concentrations in sediment may be much lower in the field than what is estimated
using total sediment concentrations and laboratory-derived partitioning coefficients
[2-5]. To decide if sediment needs to be remediated, actual risks of these sediments
need to be determined. To assess this risk, it could therefore be better to use
bioavailable rather than total concentrations of contaminants. How to measure or
predict bioavailability of organic compounds and to assess which factors influence
bioavailability, has been the focus of research projects world wide (e.g., [2,4-12]).
It is not only necessary to develop practical concepts to accurately determine
bioavailability. There is also a great need for knowledge on the effects of food web
structure on bioavailability and bioaccumulation, and models that can describe these
processes. Food web structure partly depends on the amount of nutrients present in
the system (trophic state). Interactions between trophic state or food web structure
and contaminant fate and effects (Figure 1.1) have recently been the focus of several
research projects in the USA [13,14], Canada [15,16], Sweden , and the
Netherlands [18-21]. The Dutch NWO-sponsored Stimulation Programme for
System-oriented Ecotoxicological Research, which includes the research described in
this thesis, also addressed these issues.
Figure 1.1. Interactions between nutrients and contaminants. After Koelmans et al., 1999 .
These interactions may play a large role in floodplain lakes in the Rhine catchment
area in the Netherlands, which are historically polluted with trace metals and organic
chemicals such as PAHs, PCBs, and mineral oil , together with nutrients. Many of
these pollutants cannot be degraded by microorganisms and thus are very persistent
in the environment. Although river water quality has improved the last decades ,
floodplain lake sediment remains polluted with these persistent contaminants, and
may have changed from sink to source. The extent and nature of bioavailability of
sediment-bound persistent organic pollutants such as polycyclic aromatic
hydrocarbons (PAHs) and polychlorobiphenyls (PCBs) in floodplain lakes still needs
to be identified.
Sediments in floodplain lakes do not only contain contaminants, but are also rich in
nutrients [23-26]. Many floodplain lakes can be categorized as mesotrophic or
eutrophic, and are dominated by algal blooms, although shifts to a clear,
macrophyte-dominated structure may occur [27,28]. Of course, when compared with
other lakes, dynamics in floodplain lakes are influenced very much by inundations,
which can occur several times a year, usually in winter and spring. Insight in the way
food web structure affects fate and bioaccumulation of PCBs and PAHs in these
shallow floodplain lakes is lacking.
Because of their unique combination of varying ecological conditions and pollution
characteristics, floodplain lakes are a suitable area to study the interactions between
contaminants and nutrient cycles on the ecosystem level, to assess the effect of these
interactions on bioavailability of sediment-bound contaminants and to apply existing
models to analyze these interactions.
The objectives of this research were to assess the interactions between nutrients or
ecological structure and contaminant cycles and effects in floodplain lakes, and to
model contaminant flows in these systems. The main research questions were:
What is the extent and nature of bioavailability of sediment-bound PCBs and
- What are the effects of lake ecosystem structure on fate and bioaccumulation of
PCBs and PAHs?
Sub goals described in this thesis are:
- To measure and model uptake of PCBs and PAHs by benthivorous fish in field
enclosures treated with different nutrient regimes.
- To determine bioavailable fractions of PCBs and PAHs in sediment and
evaluate the 6h Tenax-extraction method.
- To estimate fate and availability of PCBs and PAHs in flood plain lakes with
different ecological structures in different seasons and in model ecosystems
with different ecological structures.
- To determine the effect of aging on uptake of PCBs and PAHs by biota.
- To develop and improve models that describe bioaccumulation of organic
In Chapter 2, an experiment is described in which the effect of nutrient additions on
bioaccumulation of PCBs by benthivorous carp (Cyprinus carpio) in enclosures in a
floodplain lake is assessed. In this chapter, also the 6h Tenax-extraction method is
evaluated. Accumulated amounts of PCBs were compared to these 6h Tenax-
extractable amounts, which were used as a measure for bioavailability. The effect of
nutrient additions on bioaccumulation was also assessed. Finally, an uptake model
was applied, including uptake of PCBs through sediment ingestion.
Chapter 3 describes bioavailability and bioaccumulation of PCBs and PAHs by
benthic invertebrates in three ecologically different floodplain lakes. A monitoring
study that lasted from September 2000 until March 2002 was described. A model
was developed to describe BSAFs for invertebrates, including a term to account for a
lesser bioavailability due to very strong sorption to carboneceous materials such as
black carbon, coal and kerogens, and a term to account for uptake of the
bioavailable fraction of the contaminants through ingested sediment.
In Chapter 4, the effect of aging (and the subsequent reduction of bioavailability) and
ecosystem structure on bioaccumulation in biota are described. To assess the effect of
ecosystem structure on fate and bioaccumulation of PCBs and PAHs, polluted
floodplain lake sediment was brought into indoor model ecosystems, in which four
different ecological structures were created. To determine the effect of aging on
bioavailability and bioaccumulation in these systems, the sediment was spiked with
two PCBs and a deuterated PAH, which subsequently could be compared with their
‘native’ counterparts. The 6h Tenax-extraction method was evaluated with respect to
its capability to capture PCB and PAH fractions that are available to various biotic
In Chapter 5, fate of all analyzed PCBs and PAHs in these model ecosystems is
described. An extensive statistical analysis was performed on lipid-normalized
concentrations and total mass distribution of the contaminants in biota, suspended
solids and sediments. Also, the effect of ecosystem structure on species composition
and the amount of PCBs and PAHs in the biological compartments was analyzed
In Chapter 6, model results from earlier chapters are combined with existing models
to obtain a bioaccumulation model for PAHs and PCBs in all biological
compartments of the model ecosystems described in chapters 4 and 5. This new
model included a term for strong sorption to hard carbonaceous materials and was
optimized for metabolic transformation rates for PAHs. Finally, a sensitivity analysis
was performed to assess model sensitivity of the newly introduced parameters
compared to the most sensitive parameters in the original model.
Finally, in Chapter 7 this thesis ends with some concluding remarks and the main
implications of the results.
Sorption and bioavailability of HOCs to field sediments
Figure 1.2 shows a schematic overview of the partitioning of a hydrophobic organic
compound between sediment and water, and water and biota.
Sorption of HOCs (hydrophobic organic compounds) to sediments has been
described by several authors (e.g., [3,9,11,29-31], starting with Karickhoff et al. .
Karickhoff, 1979  described the partitioning of a compound between sediment
and water using a partitioning coefficient Kp:
with Csed the concentration in sediment and Cw the freely dissolved concentration in
the water phase. When it is taken into account that organic contaminants are mainly
present in the organic matter fraction of sediment (fom, or if given on an organic
carbon basis: fOC) and concentrations in the mineral fraction (fmin) can be neglected,
concentrations are usually given on an organic carbon-basis (Coc). Partitioning of the
compound between water and sediment is then described using an organic-carbon-
to-water partitioning coefficient KOC :
In general, KOC correlates with the octanol-water partitioning coefficient (KOW) of a
compound (an overview of different correlations is given in Ten Hulscher, 2005 .
Figure 1.2. Equilibrium partitioning of an organic compound between sediment and/or
suspended solids, water, and biota. Csed = concentration in sediment/suspended solids; Cmin =
concentration in mineral fraction; COC = concentration in organic carbon fraction; Cbc =
concentration in black carbon fraction; Clip = concentration in lipid fraction of biota; Cwater =
freely dissolved concentration in water; CDOC = concentration in dissolved organic carbon
(DOC); Klip = lipid/water partitioning coefficient; KOC = organic carbon/water partitioning
coefficient; KF,bc and nF,bc= Freundlich coefficient and exponent for sorption to the black carbon
fraction; KDOC = DOC/water partitioning coefficient; g = food chain magnifier for digestive
uptake of sediment.
Sediment/suspended solids Sediment/suspended solids