Full text of DOI:10.1002/etc.5620211020

Environmental Toxicology and Chemistry, Vol. 21, No. 10, pp. 2165–2171, 2002
2002 SETAC
Printed in the USA
730-7268/02 $9.00 .00
᭧
0
ϩ
RESPONSES OF BENTHIC INVERTEBRATES TO COMBINED TOXICANT AND FOOD
INPUT IN FLOODPLAIN LAKE SEDIMENTS
E
LSKE M. DE HAAS,*† BAS REUVERS,† CAROLINE T.A. MOERMOND,‡ ALBERT A. KOELMANS,‡ and
M
ICHIEL H.S. KRAAK†
†
Department of Aquatic Ecology and Ecotoxicology, Institute of Biodiversity and Ecosystem Dynamics, Faculty of Science,
University of Amsterdam, Kruislaan 320, 1098 SM, Amsterdam, The Netherlands
‡
Aquatic Ecology and Water Quality Management Group, Department of Environmental Sciences, Wageningen Agricultural University,
P.O. Box 8080, 6700 DD, Wageningen, The Netherlands
(
Received 26 November 2001; Accepted 20 March 2002)
Abstract—Benthic communities in floodplain lake ecosystems are often exposed to varying levels of both food and toxicants.
Inhibition through toxicants of sensitive species and stimulation through increased amounts of food of opportunistic species have
been observed in separate studies. The aim of this study was therefore to assess the responses of benthic invertebrates to combined
food and contamination input. Hence, seven floodplain lakes located along the River Waal, The Netherlands, with different levels
of food (being either phytoplankton or macrophyte dominated) and toxicants were selected. The responses of the sensitive mayfly
Ephoron virgo and the opportunistic midge Chironomus riparius to these sediments were assessed in 10-d growth bioassays with
both species and a 28-d emergence experiment with C. riparius. A decrease in both survival and growth of E. virgo was observed
with increasing contaminant levels. In contrast, C. riparius responded to the food quantity and quality in the sediments in spite of
the toxicants present. Therefore, we conclude that the midge C. riparius is not a suitable test organism for the assessment of
sediment toxicity. Alternatively, it proved to be an appropriate test organism to determine the nutritional value of sediments. The
mayfly E. virgo turned out to be a much more appropriate test organism for sediment toxicity bioassays because it responds to the
toxicant levels in the sediments rather than to the nutritional value. Our results demonstrate that the trophic state of an ecosystem
(
macrophyte or plankton dominated) influences the ecological risk of toxicants to benthic invertebrates in a species-specific way.
It is concluded that not the toxicant load but the combination of food and contaminants determines the persistence of benthic
invertebrates and therewith the benthic invertebrate composition in complexly polluted ecosystems.
Keywords—Contamination
Eutrophication
Combined effects
Chironomus riparius
Ephoron virgo
INTRODUCTION
the sediment, feeding on fine particulate organic matter. In
later stages they burrow U-shaped tubes in the sediment. They
filter food, such as detritus and algae, from the water by gen-
erating wavelike movements with their feathered tracheal gills
The trophic state of an ecosystem influences the amount of
food available for the benthic detritivorous community. Si-
multaneously, the trophic state also influences the sedimen-
tation of abiotic and biotic particles and therewith the depo-
sition of xenobiotic compounds due to sorption of organic
toxicants [1] and metals [2]. Analyzing the consequences of
varying food and toxicant availability to the benthos therefore
requires an inventory of the positive and negative terms in the
regulation of benthic processes. Inhibition through toxicants
of sensitive species [3] and stimulation through increased
amounts of food of opportunistic species [4,5] have been ob-
served in separate studies. These opportunistic species, how-
ever, are not necessarily more tolerant to toxicants but may
exploit at a higher rate the higher food levels [4,5]. The aim
of this study is therefore to disentangle the mechanisms that
determine the response of benthic invertebrates to combined
toxicant and food input.
[
9]. Chironomid larvae, such as C. riparius, are widely used
test organisms in acute and chronic (sediment) toxicity tests
10–13]. The larvae of C. riparius are opportunistic tube-
[
dwelling deposit feeders, which prefer eutrophic and organic
enriched waters [14]. They feed mainly on detritus and organic
matter present in the sediment.
During the past few decades, water quality of the large
western European rivers has improved [2], but the floodplain
lake sediments of these rivers still contain high concentrations
of xenobiotic compounds [15]. Therefore, the floodplain lake
sediments may now act not only as a sink but also as a source
of a wide range of chemical substances, such as nutrients,
metals, polychlorinated biphenyls (PCBs), and polycyclic ar-
omatic hydrocarbons (PAHs), that were deposited in high con-
centrations in the 1960s and 1970s [16]. The benthic com-
munities of these floodplain lakes are thus exposed to a diffuse
flux of sediment-bound toxicants, nutrients, and organic matter.
This makes these lakes an ideal area to study the combined
effects of contaminants and available food. Furthermore, flood-
plain lakes tend to be dominated by either phytoplankton or
macrophytes, which affects the supply of food to the sediment.
The objective of this study was to compare the responses
of the mayfly E. virgo and the midge C. riparius to floodplain
lake sediments with varying food and contamination levels.
Because of the opposite responses of sensitive and oppor-
tunistic species, a representative of both groups has been se-
lected: the sensitive mayfly Ephoron virgo and the opportu-
nistic midge Chironomus riparius. The mayfly E. virgo, a
benthic species present in the River Waal, proved to be a
sensitive test organism in acute and chronic toxicity tests [6–
8
]. The first-instar nymphs of E. virgo live freely on and in
*
To whom correspondence may be addressed
edehaas@science.uva.nl).
(
2
165

2
166
Environ. Toxicol. Chem. 21, 2002
E.M. de Haas et al.
a core sampler. The upper 5 cm of the core sample were isolated
and frozen at
Ϫ20ЊC within 6 h after sampling in either glass
jars for organic contaminant analyses or polyethylene bottles
for metal analyses. Sediments for bioassays and food quantity
and quality analyses were collected in December 2000. About
5
L of sediment were collected using an Ekman–Birdge grab,
which was adjusted to sample the upper 5 cm of the sediment.
The sediments were transported to the laboratory, where large
debris was picked out by hand. Next the sediment was ho-
mogenized and stored at
Ϫ20ЊC in 500-ml polyethylene bottles
within 6 h after sampling in order to eliminate autochthonous
organisms. Twenty to 25 L of lake water were collected in
jerry cans. The water was filtered twice over a 30-
␮
m filter
in order to remove the zooplankton and stored at 4
Њ
C in the
dark under constant aeration.
Sediment analyses
Metals (Cd, Cu, and Zn),
cene, fluoranthene, pyrene, benzo[
⌺
13PAHs (phenantrene, anthra-
]anthracene, chrysene,
]fluoranthene, benzo[ ]fluoranthene,
]pyrene, benzo[ghi]perylene, dibenzo[a,h]anthracene,
15PCBs (PCB18, PCB20, PCB28,
a
benzo[
benzo[
e
a
]pyrene, benzo[
b
k
Fig. 1. Sampling locations in the River Waal floodplain areas (The
Netherlands).
indeno[123]pyrene),
⌺
PCB31, PCB44, PCB52, PCB101, PCB105, PCB118,
PCB138, PCB149, PCB153, PCB170, PCB180, PCB194), and
total P were measured by Koelmans and Moermond [15].
The OM content was measured as loss-on-ignition by com-
To this purpose, whole-sediment bioassays with both species
were conducted, being an appropriate tool to assess the effects
of contaminated sediments on benthic invertebrates [11–
1
3,17–20]. In such bioassays, adverse effects on the test or-
bustion of dried sediment samples (60
50 C for 6h [26] in triplicate.
Chlorophyll and phaeophytin were measured according
ЊC until constant wt) at
ganisms are usually attributed to the presence of contaminants
in the sediments. However, if effects are observed in whole-
sediment bioassays, this will be the joint effect of all sediment
characteristics (e.g., contaminants, physical–chemical param-
eters, and food quantity and quality parameters). Food avail-
ability is a major factor in the development of benthic inver-
tebrates [21,22]. Thus, if food is deficient, the test organisms
may suffer from starvation, which may lead to a false inter-
pretation of sediment toxicity. On the other hand, feeding of
the test organisms during exposure to sediments may mask the
toxic effects, also resulting in a false interpretation [18,23–
5
Њ
a
to Lorenzen [27] in triplicate using freeze-dried sediment sam-
ples. The acetone solution was centrifuged in closed test tubes
to avoid optical disturbance by suspended sediment. Chloro-
phyll
sediments chl
Lipids were extracted from 1 g of wet sediment with 6 ml
methanol containing 2.5% H SO for 90 min at 80 C. The 400
l hexane and 1 ml 0.9% NaCl were added to the samples
and placed for 1 min on a shaker and centrifuged at 12,000
rpm for 1 min. The 200 l of the supernatant was transferred
to a 200- l vial. Fatty acid methyl esters (FAMEs) were mea-
sured using a gas chromatographer (GC) 8000 top (CE In-
struments, Milan, Italy) by injecting a 2- l aliquot in a polar
0-m DB WAX column (0.25-mm i.d.; 0.5- m film).
The labile fraction of the sediment organic matter was mea-
a
and phaeophytin contents were summed because in
a
is already partly degraded into phaeophytin.
Њ
2
4
␮
2
5]. We therefore exposed both test species to sediments in
␮
the presence and absence of additional food in order to de-
termine the effects of sediment-bound chemicals and exclude
mortality and reduction in growth due to food deficiency.
For this study, seven floodplain lakes located along the
River Waal, a branch from the River Rhine, with different
levels of contamination and trophic state were selected [15].
The species-specific preferences for these sediments were as-
sessed using 10-d whole-sediment bioassays with both species
and a 28-d emergence experiment with C. riparius. Survival,
growth, and emergence were related to contaminant levels
␮
␮
3
␮
sured by using microbial mineralization as a value for CO2
production according to Vos et al. [22]. A bacterial inoculum
was prepared from the surface layer of a sediment containing
a bacterial mat. Bacteria were detached by ultrasonification,
and large particles were removed by centrifuging for 5 min at
(
metals, PAHs, PCBs) and food quantity and quality param-
5
00 rpm. Four milliliters of homogenized wet sediment (du-
eters (organic matter content [OM], chl a, labile fraction of
plicate) were suspended in 11 ml of a 55-mM phosphate buffer
in a 60-ml serum bottle, and 1 ml of the inoculum was added.
After 30 min of aeration, pH was measured, and the bottle
was capped airtight. The bottles were placed on a rotary shaker
the OM [CO production], fatty acids [Fas], and polyunsatu-
2
rated fatty acids [PUFAs]). This enabled us to specify
eco(toxico)logical profiles of these benthic invertebrates and
to gain insight in key factors structuring benthic macrofauna
communities in the field.
at 20
ЊC in the dark, and after 1 h the CO in the headspace
2
was measured using a Carlo Erba
௡
GC (Milan, Italy). After
2
4 and 48 h of incubation, the CO concentration and the pH
2
MATERIALS AND METHODS
were again measured.
Sample collection, storage, and treatment
Bioassays
From seven floodplain lakes located along the River Waal,
The Netherlands, sediment and water were collected (Fig. 1).
Sediments for chemical analyses were collected in September
All experiments were conducted at 20
light ( 10 m/s) and a 16:7-h light:dark regime with 30 min
of twilight before and after each light period. During the ex-
Ϯ 1ЊC, moderate
ϳ
␮
2
000. Sediments were sampled at three sites in each lake using

Combined effects of food and toxicants
Environ. Toxicol. Chem. 21, 2002
2167
periments, the test systems were not aerated. Temperature,
dissolved oxygen, and pH were monitored every 5 d. At the
beginning and the end of each experiment, samples were taken
Trouvit
௡
(Trouw, France) and Tetraphyll (Tetrawerke, Melle,
௡
Germany) (20:1, wt/wt) suspension (250 mg Trouvit–Terta-
phyll in 100 ml of Elendt-M7 medium) daily. After 10 d, larvae
from five test beakers were collected from the sediment by
ϩ
Ϫ
for determination of NH4 and NO2 concentrations in the
water with Quantofix (D u¨ ren, Germany) test kits for am-
௡
sieving the sediment over a 250-␮m sieve. Surviving larvae
monium and nitrate/nitrite. During the experiments, water
quality conditions were satisfactory in all systems.
were counted, and body length was measured. Growth was
calculated by subtracting the average initial length from the
individual final length.
E. virgo
The first 10 d, the emergence experiment was carried out
exactly as the growth test. At day 10, a net trap was placed
on the test beakers, and emerged adults were removed, count-
ed, and sexed daily. The control experiments were fed 1 ml
of a Trouvit–Tetraphyll (20:1) suspension (250 mg Trouvit–
Tertaphyll in 100 ml of Elendt-M7 medium) daily. After 28
d, the experiment was terminated, and surviving larvae and
pupae were removed from the sediment by sieving over a 250-
Survival and growth of the mayfly E. virgo on floodplain
lake sediments were determined in a 10-d experiment. First-
instar nymphs (
eggs, kept in artificial diapause at 4
Ͻ
48 h) were obtained from field-collected
C in our laboratory. Seven
Њ
days prior to the start of the experiments, several glass slides
containing E. virgo eggs were placed in petri dishes containing
Elendt-M7 medium [28] and transferred to 20
terminate the artificial diapause [29]. Sediments were thawed
at 4 C 4 d before the start of the experiment.
ЊC in order to
␮
m sieve, counted, and measured for body length. Growth of
Њ
nonemerged larvae was calculated by subtracting the average
initial length from the individual final length.
Both experiments were repeated with addition of food to
all treatments, according to the feeding regime in the control.
One day before the start of the experiment, three replicate
glass jars (150 ml) with 25 ml wet homogenized sediment and
1
00 ml filtered site water were prepared and aerated overnight.
Control treatments contained 25 ml quartz sand (Sibelco
M32, Antwerp, Belgium), with a 100- to 400- m grain size
௡
Data analyses
␮
and 100 ml Elendt-M7 medium. Twenty nymphs were ran-
domly transferred into each test vessel using a blunt Pasteur
pipette (WU, Mainz, Germany). In addition, the body length
One-way analysis of variance (ANOVA) tests followed by
Scheff e´ ’s post hoc test were conducted to test for significant
differences between treatments and controls for each endpoint
of 20 larvae was measured.
(
survival, growth, or emergence). If data were not homogenous
or normally distributed, a Kruskall–Wallis was performed, fol-
lowing a Student’s test when permitted. Significant differ-
ences between the absence and presence of food were analyzed
per sediment with a Student’s test. Differences were consid-
The control treatment was fed 1.0.107
␮
m of small diatoms
3
2
(
Navicula atomus:Nitzscia perminuta, 1:1) per cm and two
t
drops of an Urtica suspension (0.75 g Urtica in 25 ml Elendt-
M7 medium) at day 0. At days 3, 5, and 7, the controls were
t
7
3
2
fed 0.5.10
␮m of diatoms per cm and one drop of Urtica
ered significant between the test categories at the 0.05 prob-
ability level. Emergence is given as the time (days) at which
7
suspension. At day 7, the controls were additionally fed 3.0.10
3
␮
m
Chlamydomonas monoica per ml.
5
0% of the adults had emerged (EmT50). This EmT50 was
At the end of the test, nymphs were collected from the
calculated according to a logistic response model adopted from
Haanstra et al. [31].
Correlations between survival, growth, time of emergence,
and sediment characteristics were determined with Pearson
correlation, using the average survival, growth, and time of
sediment by using a floating technique according to Boivin et
al. [30]. Surviving nymphs were counted and body length was
measured with an automatic image analyzer using the com-
puter program Research Assistant 3 (RVC, Hilversum, The
Netherlands). Growth was calculated by subtracting the av-
erage initial length from the individual final length.
5
0% emergence and the mean of the repeated analyses for the
different sediment parameters. All statistical analyses were
The experiment was repeated with addition of food to all
treatments, according to the control-feeding regime.
conducted using the computer program SPPS௡ 9.0 for Win-
dows (SPSS, Chicago, IL, USA).
C. riparius
RESULTS
Two bioassays were performed: Larval survival and growth
were determined in a 10-d experiment, and larval development
and adult emergence were determined in a 28-d emergence
Sediment characteristics
The sediment characteristics of the seven floodplain lake
sediments are listed in Table 1, in which the sediments are
ranked from relatively clean to contaminated. Concentrations
test. Both tests were started with first-instar larvae (Ͻ24 h),
obtained from a culture maintained in our laboratory. Three
days prior to the test, three newly deposited egg ropes were
removed from the culture and transferred into a petri dish with
of metals,
⌺PAHs, and ⌺PCBs were lowest in DeO2. Metal
concentrations were highest in DeO3a, DeO3b, and DeO4, and
the sum of PAHs and PCBs were highest in, respectively, DeO4
Elendt-M7 medium and placed at 20ЊC. Sediments were
thawed at 4 C 4 d before the start of the experiment.
Њ
(10.52 mg/kg dry wt) and DeO3b (133.22 g/kg dry wt). The
␮
For the 10-d bioassay, five replicate 400-ml glass beakers
with 75 ml homogenized wet sediment and 300 ml filtered site
water were prepared 1 d before the start of the experiment and
aerated overnight. Control experiments contained 75 ml quartz
concentrations of Cd, Zn, PAHs, and ⌺PCBs all correlated
⌺
positively with each other, and Cu was positively correlated
with Cd, Zn, and PCBs. Hence, a clear gradient in contaminant
concentrations was observed, allowing a ranking of the sed-
iments from relatively clean (DeO2) to heavily contaminated
(DeO4).
Sediment organic matter content was lowest in DeO2
(2.7%) and Gen1 (3.9%), while the organic matter content in
the other sediments ranged from 8.3 to 11.0%. The lowest total
sand (Sibelco M32) with a 100- to 400-␮m grain size and 300
ml Elendt-M7 medium. Five larvae were carefully transferred
into each test vessel with a blunt Pasteur pipette. Twenty larvae
were additionally measured for body length using an automatic
image analyzer. The control experiments were fed 1 ml of a

2
168
Environ. Toxicol. Chem. 21, 2002
E.M. de Haas et al.
Table 1. Sediment characteristics: Cd, Cu, Zn, sum of polycyclic aromatic hydrocarbons (
acids ( FAs) and sum of polyunsaturated fatty acids ( PUFAs) in mg/kg dry weight, sum of polychlorinated biphenyls (
weight, organic matter content (OM) and CO production (mmolCO d/kg dry wt). Sediments are ranked from relatively clean to contaminated
⌺
PAHs), total P (P), chlorophyll
a
(chl
a
), sum of fatty
⌺
⌺
⌺
PCBs) in
␮
g/kg dry
2
2
DeO2
Gen1
0.41
Gen3
0.90
Och2
1.25
DeO3a
DeO3b
DeO4
1.61
Cda
0.17
12
1.41
72
2.19
82
a
Cu
Zn
⌺
⌺
28
45
52
57
a
42
194
386
199
322
507
544
PAHsa
0.55
4.37
2.7
0.94
7.17
3.9
1.57
20.19
9.7
2.97
33.41
11.0
5.76
74.34
8.3
5.87
133.22
9.0
1,703
43.33
1.76
0.69
20.23
A
10.52
109.35
9.5
PCBsa
OM
P
Chl
a
434
1,226
1,137
1,312
1,458
30.07
1.23
0.37
19.45
A
1,198
a
19.31
0.99
0.44
9.46
16.44
0.89
0.22
9.10
A
76.50
2.81
1.12
74.05
3.65
1.34
30.75
M
55.40
3.08
1.23
24.33
M
⌺
FAs
⌺
PUFAs
CO2
Vegetation
23.51
b
c
A
M
a
Koelmans and Moermond [15].
b
c
A
ϭ
phytoplankton-dominated lake.
macrophyte-dominated lake.
M
ϭ
P concentration was also measured in DeO2 (434 mg/kg dry
wt) and varied from 1,137 to 1,703 mg/kg dry weight in the
other sediments. Macrophytes were present in the lakes DeO4,
nymphs, except for growth in DeO2, which was significantly
higher in the presence of food ( 0.01). Again, in the pres-
ence of food, both survival and growth were negatively cor-
related with Cu, Zn, PAHs, and PCBs. No correlations were
found with any of the food quality parameters.
p
Ͻ
Gen3, and Och2. Chlorophyll
PUFA content, and the labile fraction of the organic matter
CO production) in these macrophyte-dominated lakes were
a concentrations, FA content,
(
2
1
0-d survival and growth of C. riparius
higher than in phytoplankton-dominated lakes, with the lowest
concentration of chl (16.44 mg/kg dry wt) in Gen1 and the
highest (76.50 mg/kg dry wt) in Gen3. The labile fraction of
the organic matter ranged from 9.10 in Gen1 to 30.75 mol
a
Average control survival was 99%, and average growth of
control larvae was 7.8 mm. Survival of the midge C. riparius
was above 90% in the three macrophyte-dominated lakes (Fig.
␮
CO /d/kg dry weight in Och2. The lowest FA and PUFA con-
2). In contrast, significant lower survival (p Ͻ 0.05) compared
2
tents were found in Gen 1 (0.87 and 0.22 mg/kg dry wt), and
the highest concentrations were found in Och2 (3.59 and 1.34
mg/kg dry wt).
to the control was observed in three of the four phytoplankton-
dominated lakes (Gen1, DeO3a, and DeO3b). In all sediments,
growth was significantly lower than in the control (Fig. 2),
except for Gen3, the least polluted macrophyte-dominated
lake. Larval growth in the sediments positively correlated with
The CO production was positively correlated to OM, chl
2
a
content,
chl positively correlated with OM,
rophyte abundance. Organic matter content was positively cor-
related to FAs and PUFAs. Macrophyte abundance was
positively correlated to chl a, CO2 production, FA, and
PUFAs. Therefore, the sediments can be grouped in either
⌺
FAs,
⌺
PUFAs, and macrophyte abundance. The
a
⌺
FAs, PUFAs, and mac-
⌺
chl a,
⌺PUFAs, CO production, and the presence of macro-
2
phytes.
⌺
⌺
⌺
⌺
sediments with a low food quantity and quality (phytoplank-
ton-dominated lakes) or sediments with a high food quantity
and quality (macrophyte-dominated lakes).
No correlations were found between contaminant levels and
food quality parameters, expressing the contaminant levels
based on OM did not change these results. However, the rel-
atively clean sediment (DeO2) has the lowest food quantity
and quality and the most heavily polluted sediment (DeO4)
the highest food quantity and quality.
Survival and growth of E. virgo
Average control survival was 86.3%, and average growth
of control nymphs was 220 ␮m. With increasing contamination
levels, a decrease in both survival and growth of the E. virgo
nymphs was observed (Fig. 2). In the most polluted sediments
(
DeO3a, DeO3b, and DeO4), significant (
survival and growth was observed compared to the control.
Growth in DeO2 sediment was also significantly lower (
.01) than control growth. Survival was negatively correlated
p Ͻ 0.01) lower
Fig. 2. Survival (%) and growth (%) of Ephoron virgo and Chiron-
omus riparius larvae after 10 d on sediments in the absence and
presence of additional food, given as percentages of the corresponding
controls. White bars represent sediments without food, black bars
p
Ͻ
0
represent sediments with food. Error bars
ϭ standard deviation; * ϭ
with Cu, Zn, PAHs, and PCBs (Table 2), and growth was
negatively correlated with PAHs and PCBs.
significant difference from control ( 0.05); # indicate significant
p
Ͻ
difference between the absence and presence of additional food (
0.05).
p Ͻ
Additional food did not alter survival or growth of E. virgo

Combined effects of food and toxicants
Environ. Toxicol. Chem. 21, 2002
2169
Table 2. Correlations between growth and survival of Ephoron virgo and Chironomus riparius after 10 d and growth and EmT50 of C. riparius
2
after 28 d with sediment characteristics. Pearson coefficient
r
and significant level.
FAs sum of fatty acids,
macrophyte abundance; ϩ ϭ positive correlation with 0.05; ϩϩ ϭ positive correlation with
correlation with 0.05; ϪϪ ϭ negative correlation with 0.01
⌺
PAHs
ϭ
sum of polycyclic aromatic hydrocarbons,
sum of polyunsaturated fatty acids, CO2
0.01; Ϫ ϭ negative
⌺PCBs
ϭ
sum of polychlorinated biphenyls, chl
a
ϭ
chlorophyll a,
⌺
ϭ
⌺
PUFAs
ϭ
ϭ
Ϫ
CO production, M
ϭ
p
Ͻ
p Ͻ
2
p
Ͻ
p Ͻ
Parameter
Cd
Cu
Zn
⌺
PAHs
⌺
PCBs
Chl
a
⌺
FAs
⌺
PUFAs
CO2
M
E. virgo
Survival
Survival
Ϫ
ϩ
food
food
food
food
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
ϪϪ
ϪϪ
ϪϪ
ϪϪ
ϪϪ
ϪϪ
Ϫ
Growth
Growth
Ϫ
ϩ
Ϫ
Ϫ
Ϫ
ϪϪ
C. riparius
Growth
Growth
Ϫ
ϩ
food
food
ϩ
ϪϪ
ϩ
Ϫ
ϩ
ϪϪ
ϩ
ϪϪ
Ϫ
2
8-d growth
food
Ϫ
ϩ
ϪϪ
ϩ
Ϫ
ϩ
Ϫ
EmT50
Ϫ
food
Ϫ
ϪϪ
Addition of food diminished the mortality in Gen1, DeO3a,
and DeO3b. As a result, significant lower survival in com-
parison to the control was found only for DeO3b and, unex-
pectedly, Och2. In the presence of food, increased growth was
observed in all sediments except for Gen3. Growth negatively
tively correlated with growth after 10 d on sediments without
additional food and with chl a, FAs, and PUFAs.
Addition of food resulted in faster development in all sed-
iments; time of 50% emergence ranged from 17.5 d in DeO2
to 21.5 d in Och2, and all midges emerged within 28 d. The
⌺
⌺
correlated with chl a,
⌺
FAs,
⌺
PUFAs, CO production, and
EmT50 value in DeO2 was significantly lower (p Ͻ 0.05) than
2
the abundance of macrophytes. No correlations were found
with any of the contaminant concentrations.
in the control, while the EmT50 values in all other sediments
were comparable to the control EmT50. No correlations were
found between the EmT50 in the presence of additional food
and any of the sediment characteristics.
2
8-d development of C. riparius
Survival exceeded 80% in all control treatments. Average
EmT50 in the control was 20.3 d.
DISCUSSION
In the experiments without additional food, adult emer-
gence was observed only in the sediments from macrophyte
dominated lakes (Fig. 3). In the phytoplankton-dominated lake
sediments, little (4 and 8% in DeO2 and DeO3a) to no emer-
gence occurred, and therefore no reliable EmT50-values for
these lakes could be calculated. In the macrophyte-dominated
The present study demonstrated that the sensitive mayfly
E. virgo and the opportunistic midge C. riparius responded
completely opposite to the varying food and toxicant levels.
Ephoron virgo was far more sensitive to the toxicants present
in the sediments than C. riparius. An increase in contaminant
loading in the sediments resulted in a decrease of both survival
and growth of E. virgo. Addition of food did not enhance
survival and growth of E. virgo in any of the sediments, except
for growth in DeO2 sediment. These results indicate that re-
duction in survival and growth observed in the sediments with-
out additional food is caused by the toxicants present in the
sediment rather than to food deficiency. The reduction in
growth on DeO2 sediment in the absence of additional food,
however, can be explained by deficiency of food due to the
low OM of this sediment since growth of the nymphs in the
presence of additional food strongly increased. As reported in
several studies, mayflies are sensitive to many classes of con-
taminants [3,17,32,33]. In addition, the species used in the
present study, E. virgo, is very sensitive to several contami-
nants in the laboratory [6–8]. Our results clearly demonstrated
that it is also a sensitive test organism in sediment toxicity
bioassays. This sensitive response of E. virgo to sediment-
bound toxicants can partly explain their slow recolonization
in the River Rhine [34] since the onset of water quality im-
provement.
lakes, EmT50 values were significantly higher (
control EmT50. Emergence was positively correlated with chl
a, FAs, PUFAs, CO production, and macrophyte abun-
p Ͻ 0.05) than
⌺
⌺
2
dance. Growth of nonemerged larvae (not shown) was posi-
In contrast to E. virgo, C. riparius responded mainly to
the food quantity and quality in the sediments rather than to
the toxicants present. A reduction in both survival and growth
was observed in phytoplankton-dominated lake sediments in
the absence of additional food. In the presence of additional
food, however, survival and growth was acceptable in almost
all sediments. Thus, it can be assumed that the reduced survival
Fig. 3. Emergence of adult Chinomus riparius after 28 d of exposure
to sediments in absence (
pressed as EmT50. Error bars
EmT50 could not be calculated because of low or zero emergence.
▫
) and presence (
●) of additional food ex-
ϭ
95% confidence limits; no emergence
ϭ

2
170
Environ. Toxicol. Chem. 21, 2002
E.M. de Haas et al.
and growth was caused not by toxicants but by food deficiency.
However, not only food quantity but also food quality may
play an important role in the development of C. riparius as
indicated by the positive correlation between larval growth in
toxicants present. On the other hand, E. virgo performs best
on uncontaminated sediments containing relatively low
amounts of food, where the performance of C. riparius would
be low because of starvation. Thus, the trophic state of an
ecosystem influences the ecological risk of toxicants to benthic
invertebrates in a species-specific way. It is concluded that not
the toxicant load but the combination of food and contaminants
determines the persistence of benthic invertebrates and there-
with the benthic community composition in complexly pol-
luted ecosystems.
the absence of additional food and chl a,
⌺PUFAs, and CO2
production. The composition and nutritional value of the or-
ganic matter can indeed influence the growth of chironomids
[
35,36], and lower survival, growth, and reproduction of chi-
ronomids in nutritionally poor substrates has been observed
in several studies [13,24,25].
In the presence of additional food, growth of larvae in
phytoplankton-dominated lake sediments was higher than
growth of larvae in macrophyte-dominated lake sediments.
Growth of chironomids has been shown to correlate well with
the availability of algae or detritus from algal origin in their
food [37], indicating that the principal food source of C. ri-
parius larvae is detritus derived from algal and plant origin
Acknowledgement—We thank B. Verbree (Free University, The Neth-
erlands) for her help with the PUFA analyses and J.M. Westerveld
(University of Amsterdam, The Netherlands), who supported the CO
2
analysis. The investigations were subsidized by The Netherlands Or-
ganization for Applied Scientific Research (SSEO 014.23.013).
REFERENCES
[
22]. Hence, in the sediments where food was abundant, the
1
2
. Koelmans AA, Van der Heijde A, Knijff LM, Aalderink RH. 2001.
Integrated modeling of eutrophication and organic contaminant
fate and effects in aquatic ecosystems: A review. Water Res 35:
larvae could have fed selectively on the organic matter present
in the sediment instead of the highly nutritive food added to
the test system [38].
3
517–3536.
. Admiraal W, Van der Velde G, Smit H, Cazemier WG. 1993. The
rivers Rhine and Meuse in The Netherlands: Present state and
signs of ecological recovery. Hydrobiologia 265:97–128.
Food deficiency of C. riparius larvae was even more pro-
found in the 28-d emergence test in the absence of additional
food, where emergence was observed only in macrophyte-
dominated lake sediments, whereas in the presence of addi-
tional food, no differences were found between phytoplankton-
and macrophyte-dominated lake sediments. In an experiment
with natural sediments with different food quantities, Ristola
et al. [13] found no emergence of C. riparius without addition
of food, and only larvae in the sediment with the highest
nutritional value survived. With increasing food levels added,
a decrease in time of emergence was observed, and no dif-
ferences were observed between sediments with different
quantities of natural food. Thus, food limitation or poor food
quality in natural sediments can reduce survival, growth, and
development of C. riparius, which can be counteracted by
adding an additional food source [18,19,22,39].
3. Sibley PK, Legler J, Dixon DG, Barton DR. 1997. Environmental
health assessment of the benthic habitat adjacent to a pulp mill
discharge. I. Acute and chronic toxicity of sediments to benthic
macroinvertebrates. Arch Environ Contam Toxicol 32:274–284.
4
. Dub e´ MG, Culp JM. 1996. Growth responses of periphyton and
chironomids exposed to biologically treated bleached-kraft pulp
mill effluent. Environ Toxicol Chem 15:2019–2027.
. Stuijfzand SC, Helms M, Kraak MHS, Admiraal W. 2000. Inter-
acting effects of toxicants and organic matter on the midge Chi-
ronomus riparius in polluted river water. Ecotoxicol Environ Saf
5
4
6:351–356.
6. Van der Geest HG, Greve GD, Kroon A, Kraak MHS, Admiraal
W. 2000. Sensitivity of characteristic riverine insects, the cad-
disfly Cyrnus trimaculatus and the mayfly Ephoron virgo, to
copper and diazinon. Environ Pollut 109:177–182.
7
. Van der Geest HG, Greve GD, Boivin ME, Kraak MHS, Van
Gestel CAM. 2000. Mixture toxicity of copper and diazinon to
larvae of the mayfly Ephoron virgo judging additivity at different
effect levels. Environ Toxicol Chem 19:2900–2905.
. Van der Geest HG, De Haas EM, Boivin ME, Admiraal W. 2001.
Effects of zinc on larvae of the mayfly Ephoron virgo. Report.
University of Amsterdam, Amsterdam, The Netherlands.
When food was not the limiting factor, survival, growth,
and development of C. riparius were satisfactory in spite of
the toxicants present. However, C. riparius is not necessarily
very tolerant to toxicants [6,40]. Apparently, in the present set
of sediments, the advantages of increased organic enrichment
prevail against the potential adverse effects of the toxicants
8
9. Kureck A, Fontes RJ. 1996. The life cycle and emergence of
Ephoron virgo, a large potamal mayfly that has returned to the
River Rhine. Arch Hydrobiol Suppl 113:319–323.
[
4,5]. Therefore, the amount and quality of available food,
1
0. Harkey GA, Landrum PF, Klaine SJ. 1994. Comparison of whole-
sediment, elutriate and pore-water exposures for use in assessing
sediment-associated organic contaminants in bioassays. Environ
Toxicol Chem 13:1315–1329.
either natural or added to the test system, may affect the re-
sponse of the test organism to sediment-bound toxicants or
might even neutralize their effects [10]. This will cause un-
1
1. Kemble NE, Brumbaugh WG, Brunson EL, Dwyer FJ, Ingersoll
CG, Monda DP, Woodward DF. 1994. Toxicity of metal-contam-
inated sediments from the upper Clark Fork River, Montana, to
aquatic invertebrates and fish in laboratory exposures. Environ
Toxicol Chem 13:1985–1997.
desired side effects in sediment bioassays using C. riparius
.
Addition of food in midge bioassays is needed to exclude
reduction of survival, growth, and reproduction due to food
deficiency but may lead to masking of toxic effects. Therefore,
we conclude that the midge C. riparius is not a suitable test
organism for the assessment of sediment toxicity. Alterna-
tively, it proved to be an appropriate test organism to determine
the nutritional value of sediments. The mayfly E. virgo turned
out to be a much more appropriate test organism for sediment
toxicity bioassays because it responds to the toxicant levels
in the sediments rather than to the nutritional value.
1
1
2. Ristola T, Pellinen J, Leppanen M, Kukkonen J. 1996. Charac-
terization of Lake Ladoga sediments. 1. Toxicity to Chironomus
riparius and Daphnia magna. Chemosphere 32:1165–1178.
3. Ristola T, Pellinen J, Ruokolainen M, Kostamo A, Kukkonen
JVK. 1999. Effect of sediment type, feeding level, and larval
density on growth and development of a midge (Chironomus
riparius). Environ Toxicol Chem 18:756–764.
1
1
4. Armitage PD, Cranston PS, Pinder LCV. 1995. The Chironomi-
dae: Biology and Ecology of Non-Biting Midges. Chapman &
Hall, London, UK.
CONCLUSIONS
5. Koelmans AA, Moermond CTA. 2000. Stofstroom-analyse. Zwa-
re metalen, PCBs en PAKs in uiterwaardplassen. T Vastlegging
nulsituatie. Report. Wageningen Agricultural University, Wag-
eningen, The Netherlands.
0
The present study demonstrated that C. riparius performs
best on macrophyte-dominated lake sediments because it ex-
ploits the higher food levels at higher rates in spite of the
16. Beurskens JEM, Mol GAJ, Barreveld HL, Vanmunster B, Winkels

Combined effects of food and toxicants
Environ. Toxicol. Chem. 21, 2002
2171
HJ. 1993. Geochronology of priority pollutants in a sedimentation
area of the Rhine River. Environ Toxicol Chem 12:1549–1566.
Sediment-water chironomid toxicity test using spiked sediment.
OECD Guideline 218. Paris, France.
29. Greve GD, Van der Geest HG, Stuijfzand SC, Kureck A, Kraak
MHS. 1999. Development and validation of an ecotoxicity test
1
1
7. Giesy JP, Rosiu CJ, Graney RL, Henry MG. 1990. Benthic in-
vertebrate bioassays with toxic sediment and pore water. Environ
Toxicol Chem 9:233–248.
8. Ankley GT, Benoit DA, Balogh JC, Reynoldson TB, Day KE,
Hoke RA. 1994. Evaluation of potential confounding factors in
sediment toxicity test with three freshwater benthic invertebrates.
Environ Toxicol Chem 13:627–635.
9. Liber K, Call DJ, Dawson TD, Whiteman FW, Dillon TM. 1996.
Effects of Chironomus tentans larval growth retardation on adult
emergence and oviposition success: Implications for interpreting
freshwater sediment bioassays. Hydrobiologia 323:155–167.
0. Huuskonen SE, Ristola TE, Tuvikene A, Hahn ME, Kukkonen
JVK, Lindstr o¨ m-Sepp a¨ P. 1998. Comparison of two bioassays, a
fish liver cell line (PLHC-1) and a midge (Chironomus riparius),
in monitoring freshwater sediments. Aquat Toxicol 44:47–67.
1. Suedel BC, Rodgers JH. 1994. Responses of Hyalella azteca and
Chironomus tentans to particle-size distribution and organic mat-
ter content of formulated and natural freshwater sediments. En-
viron Toxicol Chem 13:1639–1648.
using field collected eggs of the riverine mayfly Ephoron virgo
.
Proceedings, Experimental and Applied Entomology, NEV, Am-
sterdam, The Netherlands 10:105–110.
3
0. Boivin ME, De Haas EM, Van der Geest HG, Admiraal W. 2001.
A: Short-term effects of zinc on larvae of the riverine mayfly
Ephoron virgo; B: Development of a long-term toxicity test using
larvae of the riverine mayfly Ephoron virgo. Report. University
of Amsterdam, Amsterdam, The Netherlands.
1. Haanstra L, Doelman P, Oude Voshaar JH. 1985. The use of
sigmoidal dose response curves in soil ecotoxicological research.
Plant Soil 84:293–297.
32. Diamond JM, Winchester EL, Mackler DG, Gruber D. 1992. Use
of the mayfly Stenonema modestum (Heptageniidae) in subacute
toxicity assessments. Environ Toxicol Chem 11:415–422.
33. Lowell RB, Culp JM, Wrona FJ. 1995. Stimulation of increased
short-term growth and development of mayflies by pulp-mill ef-
fluent. Environ Toxicol Chem 14:1529–1541.
4. Bij de Vaate A, Klink A, Oosterbroek F. 1992. The mayfly, Epho-
ron virgo (Olivier), back in the Dutch parts of the Rivers Rhine
and Meuse. Hydrobiol Bull 25:237–240.
5. Johnson RK, Bostr o¨ m B, Van den Bund W. 1989. Interactions
between Chironomus plumosus (L.) and the microbial community
in surficial sediments of a shallow, eutrophic lake. Limnol Ocean-
ogr 34:992–1003.
6. Vos JH, Ooijevaar MAG, Postma JF, Admiraal W. 2000. Inter-
actions between food availability and food quality during growth
of early instar chironomid larvae. J North Am Benthol Soc 19:
158–168.
1
2
2
3
3
2
2
2. Vos JH. 2001. Feeding of detritivores in freshwater sediments.
PhD thesis. University of Amsterdam, Amsterdam, The Nether-
lands.
3. Day KE, Kirby RS, Reynoldson TB. 1994. Sexual dimorphism
in Chironomus riparius (Meigen)—Impact on interpretation of
growth in whole-sediment toxicity tests. Environ Toxicol Chem
3
3
1
3:35–39.
2
2
2
2
2
4. Harkey GA, Landrum PF, Klaine SJ. 1994. Preliminary studies
on the effect of feeding during whole sediment bioassays using
Chironomus riparius larvae. Chemosphere 28:597–606.
5. Lacey R, Watzin MC, McIntosh AW. 1999. Sediment organic
matter content as a confounding factor in toxicity tests with Chi-
37. Kajak Z, Warda J. 1968. Feeding of benthic nonpredatory Chi-
ronomidae in lakes. Ann Zool Fenn 5:57–64.
3
8. Pinder LCV. 1995. The habitats of chironomid larvae. In Armitage
P, Cranston PS, Pinder LCV, eds, The Chironomidae: The Biology
and Ecology of Non-Biting Midges, 1st ed. Chapman & Hall,
London, UK, pp 107–135.
9. Bridges TS, Farrar JD, Duke BM. 1997. The influence of food
ration on sediment toxicity in Neanthes arenaceodentata (An-
nelida: Polychaeta). Environ Toxicol Chem 16:1659–1665.
0. Phipps GL, Mattson VR, Ankley GT. 1995. Relative sensitivity
of 3 freshwater benthic macroinvertebrates to 10 contaminants.
Arch Environ Contam Toxicol 28:281–286.
ronomus tentans. Environ Toxicol Chem 18:231–236.
6. Luczak C, Janquin MA, Kupka A. 1997. Simple standard pro-
cedure for the routine determination of organic matter in marine
sediment. Hydrobiologia 345:87–94.
7. Lorenzen CJ. 1967. Determination of chlorophyll and pheo-pig-
ments: Spectrophotometric equations. Limnol Oceanogr 12:343–
3
4
3
46.
8. Organization for Economic Cooperation and Development. 2001.