Stoichiometric controls of mercury dilution by growth
Roxanne Karimi*†, Celia Y. Chen*, Paul C. Pickhardt‡§, Nicholas S. Fisher‡, and Carol L. Folt*
*Department of Biological Sciences, Dartmouth College, Hanover, NH 03755; and‡Marine Sciences Research Center, Stony Brook University,
Stony Brook, NY 11794
Edited by G. David Tilman, University of Minnesota, St. Paul, MN, and approved March 18, 2007 (received for review December 19, 2006)
Rapid growth could significantly reduce methylmercury (MeHg)
concentrations in aquatic organisms by causing a greater than
proportional gain in biomass relative to MeHg (somatic growth
dilution). We hypothesized that rapid growth from the consump-
tion of high-quality algae, defined by algal nutrient stoichiometry,
reduces MeHg concentrations in zooplankton, a major source of
MeHg for lake fish. Using a MeHg radiotracer, we measured
changes in MeHg concentrations, growth and ingestion rates in
juvenile Daphnia pulex fed either high (C:P ? 139) or low-quality
(C:P ? 1317) algae (Ankistrodesmus falcatus) for 5 d. We estimated
Daphnia steady-state MeHg concentrations, using a biokinetic
model parameterized with experimental rates. Daphnia MeHg
assimilation efficiencies (?95%) and release rates (0.04 d?1) were
unaffected by algal nutrient quality. However, Daphnia growth
rate was 3.5 times greater when fed high-quality algae, resulting
in pronounced somatic growth dilution. Steady-state MeHg con-
centrations in Daphnia that consumed high-quality algae were
one-third those of Daphnia that consumed low-quality algae due
to higher growth and slightly lower ingestion rates. Our findings
show that rapid growth from high-quality food consumption can
in freshwater food webs.
contaminants ? food quality ? heavy metals ? nutrient stoichiometry ?
standing the key factors driving MeHg accumulation in fish has
become a global priority (1). Rapid somatic growth rates are
hypothesized to reduce mass-specific MeHg concentration (bur-
den) in fish and other aquatic organisms (2, 3) by the process of
somatic growth dilution (SGD). SGD occurs when rapid growth
results in a disproportionate increase in the net rate of biomass
gain relative to MeHg gain. The relative quality of food con-
sumed can strongly influence growth rates in aquatic organisms.
Moreover, food quality for aquatic consumers varies widely
across lakes and seasons (4), thus potentially contributing to the
large variation in Hg concentrations observed in lake fish in situ.
MeHg accumulation has not been well examined.
At present, evidence for somatic growth dilution of MeHg,
whether due food quality or other factors, is sparse and some-
what contradictory. Thus far, our understanding of SGD has
somatic growth rates and Hg concentrations in fish. In most of
these studies, the many possible factors driving SGD (e.g.,
temperature, food availability, food quality, activity level, and
stress) are not controlled and are often confounded. Negative
correlations between somatic growth rate and concentrations of
total Hg (5–8) and other contaminants, such as Pb (9), have been
found for fish. However, other studies have found no (10), or
positive correlations between fish growth and Hg concentrations
(11). This discrepancy may be because the effects of growth on
Hg accumulation strongly depend on the particular mechanism
driving growth, which is often difficult to determine in situ. For
example, in the studies by Stafford and Haines (10) and Dutton
net biomass and net Hg gain through increased food-borne Hg
ethylmercury (MeHg) poses a serious human and wildlife
health risk primarily through fish consumption, so under-
ingestion, effectively maintaining, or even increasing mass-
specific Hg concentrations. Thus, a complete mass-balance
approach that accounts for all inputs, assimilation and outputs
of both Hg and total biomass is necessary to identify the
conditions under which SGD can occur.
SGD is likely to occur when growth increases with relatively
little or no change in Hg gain. A disproportionate increase in
biomass gain relative to Hg gain can occur when activity or
respiration rates are relatively low (12, 13) or when food quality
is high. Organisms consuming high-quality food gain more
biomass per unit food consumed, hence per unit Hg consumed,
than from low-quality food. Thus, our general hypothesis is that,
all else being equal, consumption of high-quality food can cause
greater SGD of Hg compared with the consumption of low-
Somatic growth dilution of Hg in zooplankton due to food
quality may explain variation in fish Hg concentrations across
lakes that differ in nutrient availability. Recent studies have
found a positive correlation between the relative availability of
N and P and fish mercury concentrations from hundreds of lakes
(14, 15). High availability of P relative to N (i.e., low N:P) results
in a low algal N:P ratio (indicative of high algal quality as food),
causing rapid growth in zooplankton, particularly Daphnia (4,
16). A large body of work has found that Daphnia consistently
grow more efficiently on phytoplankton with high P content
relative to N or C (16–18) because of their high P demand for
protein synthesis (19). Hence, SGD may cause zooplankton in
lakes with low N:P algae to have lower MeHg concentrations.
Fish consuming these zooplankton, in turn, should accrue less
MeHg, thereby propagating the effects of SGD through the food
web. In this way, differences in growth rates of zooplankton due
considerable differences in mercury concentrations in fish from
different lakes even if mercury concentrations in the water do
This study mechanisticly demonstrates somatic growth dilu-
tion of mercury in consumers. We experimentally test the
general hypothesis that phytoplankton nutrient stoichiometry
(C:P) affects MeHg accumulation in juvenile Daphnia because
of SGD. We fed juvenile Daphnia either high-quality (HiQ, low
falcatus) radiolabeled with Me203Hg. We nondestructively mea-
sured Daphnia MeHg assimilation and MeHg release, or efflux
rates, as well as growth and ingestion rates over the juvenile
growth period and parameterized a highly predictive biokinetic
model (20) with these rates to estimate steady-state MeHg
concentrations in Daphnia. We predicted that steady-state
performed research; P.C.P. contributed new reagents/analytic tools; R.K. analyzed data;
and R.K. and C.L.F. wrote the paper.
The authors declare no conflict of interest.
Abbreviations: HiQ, high-quality; LoQ, low-quality; MeHg, methylmercury; SGD, somatic
†To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
§Present address: Department of Biology, Lakeland College, Sheboygan, WI 53082.
This article is a PNAS Direct Submission.
© 2007 by The National Academy of Sciences of the USA
May 1, 2007 ?
vol. 104 ?
no. 18 ?
MeHg concentrations would be significantly lower in Daphnia
fed high-quality, low C:P phytoplankton than Daphnia fed
low-quality, high C:P phytoplankton because of somatic growth
Phytoplankton Treatment Conditions. We established A. falcatus
treatments with atomic ratios of 139:1 C:P, 15:1 N:P (HiQ), and
1317:1 C:P, 92:1 N:P (LoQ) (‘‘experimental conditions’’ in Table
1). Me203Hg uptake by A. falcatus cells was similar in both the
HiQ and LoQ treatments (Fig. 1). After harvesting radioactively
media, algal Me203Hg concentration based on cell dry weight was
higher in HiQ cells (114 ng mg?1) than in LoQ cells (62 ng mg?1)
(‘‘experimental conditions’’ in Table 1), because LoQ cell den-
cell density (2.75 104cells ml?1). These differences in algal
Me203Hg concentrations have no effect on Daphnia Me203Hg
assimilation and efflux estimates, which are normalized to initial
Daphnia Growth and Me203Hg Dynamics. Daphnia specific growth
diet (t10? 2.23; P ? 0.0001) (‘‘Daphnia responses’’ in Table 1).
As a result, the biomass of individuals fed HiQ algae at day 5 was
?3 times higher than those fed LoQ algae. Overall, Me203Hg
depuration (percent Me203Hg remaining) over 5 d was similar
between Daphnia fed high- and low-quality algae (Fig. 2A).
There was no significant difference in Daphnia Me203Hg assim-
ilation efficiency (?95%), efflux rate (?0.041 d?1), or biological
half-life (?17 d) between HiQ and LoQ treatments (‘‘Daphnia
responses’’ in Table 1). However, at the end of the 5-d depura-
tion period, Daphnia mass-specific Me203Hg concentration (ng
Hg mg?1dry weight) was twice as reduced when consuming HiQ
algae (16% remaining) than when consuming LoQ algae (34%
remaining) (t10 ? 2.23; P ? 0.017) (Fig. 2B and ‘‘Daphnia
responses’’ in Table 1).
Daphnia Ingestion Rates. Overall, Daphnia specific ingestion rates
were higher on LoQ algae than on HiQ algae (F1,15? 24.45; P ?
0.0002) (Fig. 3 and ‘‘Daphnia responses’’ in Table 1). However,
differences in ingestion between treatments were smaller than
repeated measures analysis also revealed significant differences
Table 1. Experimental treatment conditions and Daphnia physiological responses
(n ? 3)
(n ? 3)
pg d.w. cell?1
(n ? 4)
(n ? 15)
ng Hg mg?1*
(n ? 3)
MeHg AE, %
[n ? 6 (HiQ),
[n ? 6 (HiQ),
[n ? 6 (HiQ),
[n ? 36 (HiQ),
(n ? 6)
at day 5,
139 ? 14
15.25 ? 0.56
0.029 ? 0.004
150.29 ? 38.43
113.54 ? 0.56
96.02 ? 11.15
0.0415 ? 6 ? 10?5
16.70 ? 0.03
1.296 ? 0.110
0.254 ? 0.011
1,317 ? 174
91.84 ? 2.73
0.044 ? 0.006
156.50 ? 24.56
62.48 ? 0.54
94.75 ? 12.83
0.0413 ? 8 ? 10?5
16.77 ? 0.03
1.694 ? 0.108
0.069 ? 0.004
Values are means ? SE.
*Algae MeHg concentrations were based on ng Hg mg?1dry weight after resuspension.
based on ng MeHg associated with cells per liter of culture. Open circles
represent LoQ cells; filled circles represent HiQ cells. The dashed line indicates
when cells were harvested at 67.5 h.
A. falcatus Me203Hg uptake over 5 d. Values are means ? SE (bars)
www.pnas.org?cgi?doi?10.1073?pnas.0611261104Karimi et al.
in specific ingestion rate over time (F1.18, 17.72 ? 28.25; P ?
0.0001). There was no significant time-by-treatment interaction.
Steady-State MeHg Concentrations Based on Experimental Condi-
tions. The biokinetic model indicated that differences in inges-
tion and growth rate result in a steady-state MeHg concentration
(MeHgss) that is 3.5 times higher in Daphnia feeding on LoQ
algae than in Daphnia consuming HiQ algae (Fig. 4). Due to the
reduced growth rate alone, MeHgssin Daphnia fed LoQ algae is
2.5 times higher than MeHgssin Daphnia fed HiQ algae (Fig. 4).
Similarly, because of increased ingestion rate alone, MeHgssin
Daphnia fed LoQ algae is 1.3 times higher than the MeHg
concentration in Daphnia fed HiQ algae (Fig. 4). Thus, growth
rate differences between HiQ and LoQ algae consumption are
predicted to have a larger effect on MeHgssthan ingestion rate
differences at these nutrient levels.
This study shows that the consumption of high-quality food
reduces MeHg accumulation in consumers. Specifically, we
showed that consumption of high-quality phytoplankton, de-
fined by nutrient stoichiometry, reduces MeHg accumulation in
Daphnia. This result has several implications. First, even though
Daphnia have little regulatory control over the tendency of
MeHg to persist in somatic tissue, consumption of high-quality,
low C:P phytoplankton can considerably reduce Daphnia MeHg
concentrations by somatic growth dilution. Second, reduced
consumption rates, which can accompany feeding on high-
quality, low C:P algae, may further reduce MeHg accumulation.
Together, the effects of SGD and reduced consumption of
high-quality food are likely to propagate through the food web,
with widespread observations of lower fish Hg concentrations in
relatively productive aquatic systems that have relatively high P
availability (14, 15).
Our primary result is that somatic growth driven by food
quality can strongly influence MeHg accumulation. In our study,
Daphnia MeHg assimilation and efflux rates were unaffected by
food quality. However, Daphnia grew faster on high-quality, low
C:P algae (‘‘Daphnia responses’’ in Table 1) resulting in a
significant reduction in the concentration of MeHg in their
tissues. Daphnia MeHgeffluxrates(?0.041d?1)wereconsistent
with those found in other studies (21, 22) and were lower than
and inorganic Hg (24). Low MeHg efflux rates result in the
buildup of high levels of MeHg in somatic tissue. Thus, SGD may
pulse (%). (A) Total MeHg (ng Hg) remaining in individual Daphnia. (B) MeHg
concentration (ng Hg per mg of dry weight) remaining in Daphnia. Filled
Values are means ? SE (bars).
Depuration of the Me203Hg pulse over 5 d, normalized to T ? 0 after
(white) A. falcatus cells. Values are means ? SE (bars).
experimentally measured rates (growth, ingestion, growth and ingestion)
from the consumption of HiQ (shaded) and LoQ (white) algae. Error bars
represent the range of values (minimum, mean, maximum) estimated based
on the observed confidence limits of growth and ingestion rates.
Daphnia steady-state MeHg concentration based on response to
Karimi et al.
May 1, 2007 ?
vol. 104 ?
no. 18 ?
be a particularly important process by which organisms abate the
accumulation of MeHg and other biologically persistent sub-
stances such as other metals (25) and chlorinated hydrocarbon
Unlike MeHg assimilation and retention, daphnid growth
varies greatly with algal nutrient stoichiometry. Moreover, the
effect of algal stoichiometry on growth is greater than the effect
of the quantity of algae consumed (Fig. 3), as shown in other
studies (27–29). This is largely because Daphnia consuming
high-nutrient-quality algal food (low C:P) incorporate a greater
fraction of ingested C into tissue (thus increasing net biomass
gain) than do Daphnia consuming low nutrient quality algae
(high C:P) (30). This can occur through differences in C assim-
ilation through the gut (29) or C respiration (31). Either way, the
consumption of high nutrient quality algae can increase Daphnia
growth without a concomitant increase in the quantity of MeHg
ingested. The trend of faster growth from high-quality, low C:P
algae has been found for numerous daphnid and algal species
(16, 17, 32). Therefore, stoichiometric controls of SGD may
occur in many other Daphnia species that commonly dominate
aquatic food webs, making SGD an important, yet generally
overlooked factor capable of explaining much variation in zoo-
plankton mercury concentrations across lakes.
Reduced consumption of high-quality algae can act in concert
with rapid growth to further dilute Daphnia MeHg concentra-
tions. Because of lower ingestion rates, Daphnia consuming
high-quality algae may reduce their intake of food-borne MeHg.
Together, lower ingestion and increased growth from a high-
quality diet reduced steady-state Hg concentration to a value
one-third that of Daphnia fed the LoQ diet. Similarly, when
compensatory feeding of low-quality food occurs, higher inges-
tion and slower growth would both increase MeHg accumula-
tion. Compensatory feeding has been found in other studies of
Daphnia (33, 34) although it is not universal. For example, some
rates (18, 31) or lower ingestion of low-quality food (27, 35).
Nevertheless, the effect of ingestion rate on Hg accumulation
was minimal compared with the growth dilution effect in our
study. This suggests that even a reversal of ingestion rates, i.e.,
higher consumption of high-quality food, would only slightly
reduce the net dilution of Hg concentration. Moreover, com-
pensatory feeding has been found in a variety of organisms
(36–38) and its potential to enhance the accumulation of MeHg
and other contaminants merits further study.
In summary, our results clearly demonstrate that a low algal
C:P ratio substantially reduces the trophic transfer of MeHg to
Daphnia through somatic growth dilution and, to a lesser extent,
reduced consumption rates. A reduction in MeHg in Daphnia is
most likely to reduce fish MeHg concentrations in lakes where
Daphnia are a primary food source for fish. SGD of mercury in
Daphnia has the potential to reduce mercury accumulation in
food webs of relatively productive lakes with relatively high P
availability. Additionally, when high nutrient availability stimu-
lates rapid population growth of algae (39) and zooplankton
(28), the processes of algal bloom dilution (40–42) and zoo-
plankton density dilution (43), respectively, may further reduce
zooplankton MeHg concentrations. Consistent with our findings
for MeHg, studies of other biologically persistent contaminants
have found negative relationships between total P and fish
concentrations for PCBs (44, 45) and other chlorinated hydro-
carbons (46). Thus, low C:P stoichiometry may cause a combi-
nation of multiple dilution processes that influences the accu-
mulation of MeHg and other persistent contaminants in
freshwater organisms. Other, system-specific metrics of food
quality or nutrient limitation may influence the accumulation of
MeHg and other biologically persistent contaminants in any
organism through SGD. Hence, exploring the prevalence of
SGD under a range of field conditions could help predict
conditions leading to high or low contaminant concentrations in
organisms across a broad spectrum of ecosystems.
Algae and Zooplankton Culturing. Cultures of the chlorophyte, A.
falcatus var. acicularis (UTEX clone 101; University of Texas,
Austin, TX) were maintained under two different nutrient
conditions, a high nutrient quality treatment (HiQ, 15:1 atomic
N:P) and a low nutrient quality treatment (LoQ, 110:1 atomic
N:P). These nutrient levels are known to have contrasting effects
on Daphnia growth and reproduction (27, 28) and are well within
the range typically found in lakes throughout the northeastern
United States (14). Algae were cultured according to treatment
by using a modified Woods Hole MBA medium without buffer
(47) enriched with a vitamin mixture (48) [see supporting
information (SI) Table 2 for media composition]. All culture
flasks, tubing, filters, and other culturing supplies were auto-
claved. Culturing media were filtered through a sterilized
0.22-?m filter into five replicate flasks for each treatment. A
sterile inoculum of A. falcatus was added to each flask. Cultures
were continually aerated through a 0.22 ?m membrane filter and
incubated under continuous light at 20°C. After ?8 d, algal
cultures were in log-phase growth, at which time cells were
harvested for experimental feeding. For each treatment, har-
vested cells were filtered, rinsed, and resuspended in fresh media
and stored in a common flask. This algae storage medium was
the same as the Woods Hole medium without EDTA, vitamins,
or trace elements. At this time, algal cells were sampled for
nutrient concentrations, cell density, and cell volume.
Before the experiments, cultures of a clonal isolate of a
Daphnia pulex/Daphnia pulicaria hybrid (log52 clone; Indiana
University, Bloomington, IN) had been maintained in modified
Daphnia COMBO media (49) (see SI Table 3 for media com-
position) and fed on HiQ A. falcatus for 65 generations. Same-
age neonates (?24 h old) were isolated from maintenance
cultures one generation before experiments. These ‘‘brood fe-
males’’ provided neonates for experiments. For both the radio-
label feeding-depuration and ingestion rate experiments, ?24-
h-old neonates from the third brood clutch were isolated and
alternately assigned to either the HiQ or LoQ treatment. Ex-
perimental animals were kept in fresh COMBO media without
P (P-free media) to ensure that algae were the only source of
or LoQ algae according to treatment for 24 h before the
experiments to allow individuals to acclimate to their food.
Algae Me203Hg Radiolabeling. To track Daphnia assimilation and
depuration of mercury, we used an organic, methylated form of
the ?-emitting radioisotope203Hg. We chose to examine the
trophic transfer of methylmercury (CH3203Hg?, or MeHg),
because this particular form of mercury is known to biomagnify
through food webs (50), and therefore has a greater potential for
toxicity through food consumption than inorganic mercury.
MeHg was synthesized from203Hg according to methods de-
scribed (ref. 51 and references therein). The specific activity of
the resulting Me203Hg was 127 kBq ?g?1.
In labeling A. falcatus cells with Me203Hg, our goal was to
minimize differences in MeHg uptake between HiQ and LoQ
algae to isolate the effects of algal nutrient stoichiometry on
Daphnia SGD of Me203Hg. We did not test for the effects of
nutrient stoichiometry on A. falcatus Me203Hg uptake. For each
treatment, A. falcatus cells were added to six replicate flasks of
250 ml of HiQ or LoQ algae storage media to a density of 0.1 mg
of dry weight liter?1. Three control flasks per treatment con-
tained only media to control for the adsorption of the Me203Hg
label to the sampling filters. Each replicate flask received
Me203Hg to give an aqueous concentration of 0.58 nM (?115
ng/liter) at the initial time point. Whereas this concentration is
www.pnas.org?cgi?doi?10.1073?pnas.0611261104 Karimi et al.
higher than those typical of unpolluted lakes (1–2 ng liter?1, 50),
it allowed us to monitor the Me203Hg over a number of days and
is not known to cause toxic effects over short-term exposure (52,
53). To minimize differences in algal Me203Hg uptake and cell
concentrations between treatments, cell growth was minimized
cultures were incubated at 17°C for 5 d.
A. falcatus Me203Hg uptake was monitored at multiple time
points over 5 d. At each time point, radioactivity associated with
the algal cells was assessed by filtering 10-ml aliquots from each
flask onto 1-?m polycarbonate membranes, following the
method of Fisher et al. (54). After 67.5 h, ?72% of the label had
been taken up by HiQ and LoQ algae. To expose the Daphnia
to radiolabeled algae without aqueous exposure to Me203Hg, the
labeled algal cells were separated from their radioactive water by
filtering 20 ml of labeled A. falcatus cell suspension from each
flask onto polycarbonate membranes and resuspending them
into a common flask with fresh, unlabeled algal storage media
for each treatment. Samples of resuspended algae were analyzed
for radioactivity (see Radioassays), and cell density was deter-
mined by using a hemocytometer.
Daphnia MeHg Exposure and Bioenergetics. Radiolabeled algae of
different nutrient qualities (HiQ or LoQ) were pulse-fed to
Daphnia after which Daphnia were fed unlabeled HiQ or LoQ
algae for 5 d during the juvenile growth period. The depuration
of the label was followed in live Daphnia over the 5 d to quantify
Me203Hg assimilation efficiency and efflux rates. Daphnia of the
same size and age (48-h-old) were added to each of six replicate
borosilicate containers per treatment with 17 individuals per
container. This density of individuals was sufficient for radio-
activity measurements in the Daphnia while remaining below
crowding conditions (55). Each replicate contained 100 ml of
P-free Daphnia culturing media. Additional 48-h-old Daphnia
were measured for initial dry weights to quantify growth rate.
Daphnia were allowed to clear their guts in the absence of food
for 2 h. Then, 2.2 ? 106labeled HiQ or LoQ cells, corresponding
to 3.3 ? 107?m3, were added to each replicate. To minimize
cross-contamination between treatments, animals from the LoQ
replicates were removed from their feeding chambers before
HiQ animals. As a result, Daphnia in the LoQ treatment fed on
radiolabeled algae for a shorter time (35–78 min) than those in
the HiQ treatment (86–128 min). Both exposure times were
comparable with the gut passage time of these organisms (56) to
minimize recycling of the radiolabel.
After radioactive feeding, Daphnia were rinsed twice in fresh
P-free media, and five individuals from each replicate were
analyzed for initial radioactivity. To monitor depuration of the
label, all individuals were placed into new replicate containers
with fresh P-free media and fed nonradioactive HiQ or LoQ
algae at a daily ration of 1.6 ? 106?m3per Daphnia for 5 d.
Radioactivity in Daphnia was measured nondestructively at
multiple time points, and media was renewed every 24 h. At each
time point, five Daphnia from each replicate were placed into
counting vials, measured for radioactivity and returned to their
experimental containers. At the final time point, Daphnia were
measured for dry weights to calculate somatic growth rates.
In a parallel experiment, we monitored Daphnia ingestion
rates of HiQ or LoQ algae. Daphnia were fed nonradioactive
algae according to the same design used for the Me203Hg
depuration period. In addition, we monitored changes in algal
density in three control containers with no Daphnia for each
treatment. Daphnia were transferred daily to new containers
with fresh P-free media and algae according to treatment. Algal
cell densities were measured in each container before Daphnia
were added and 24 h later, after individuals were transferred to
a new container. Thus, ingestion rate was measured every 24 h
for 5 d.
Radioassays. Radioactivity of Me203Hg in all samples was deter-
mined by using an LKB Amersham Pharmacia Wallac (Gaith-
ersburg, MD) 1282 Compugamma with a NaI(T1) well detector.
Gamma-emissions were assayed at 279 keV and counting times
were 10 min, yielding typical propagated counting errors of
?5%. All counts were corrected for decay and background
radioactivity, using appropriate standards and blanks.
Calculations and Statistical Analyses. The assimilation efficiency of
MeHg (AE, the proportion of ingested Me203Hg assimilated into
tissue) was calculated as the y-intercept of the regression be-
tween the natural log of the percent Me203Hg retained in
Daphnia and time for the slowly exchanging pool during the 5-d
of assimilated Me203Hg) was calculated as the slope of the
regression (57). The biological half-life (tb1?2) of Me203Hg was
made for each replicate and averaged for each treatment.
Differences in resuspended algae Me203Hg concentrations,
Daphnia Me203Hg AE, Ke, tb1?2, and percent Me203Hg retained at
Daphnia specific growth rate was calculated as (Ln(final
weight) ? Ln(initial weight) time?1) for each replicate sepa-
rately, and compared between treatments by ANOVA. Signifi-
cant differences in Daphnia ingestion rates between treatments
and over time were tested with multivariate ANOVA-repeated
measures. To meet the assumptions of the multivariate
ANOVA-repeated measures approach, a significant lack of
sphericity in the variance-covariance matrix indicated by a ?2test
was accounted for by reporting the Geisser–Greenhouse F test
correction value (58). All statistical tests were conducted by
using JMP 5.01.
Modeling Steady-State MeHg Concentrations in Daphnia. We calcu-
lated steady-state MeHg concentrations in Daphnia (MeHgss, ng
g?1dry weight), using a biokinetic model fit with experimentally
measured rates given by the equation
MeHgss?AE ? SIR ? Cf
(57, 59) where AE ? the assimilation efficiency of MeHg (%),
SIR is the specific ingestion rate (mg mg?1d?1), Cfis the MeHg
concentration in the algal food (ng g?1), Keis the efflux loss rate
mg?1d?1). MeHg accumulation from the aqueous phase is
assumed to be negligible (60). Site-specific predictions of steady-
state concentrations of numerous metals in diverse aquatic
animals, using this model and lab-derived kinetic rates have
closely matched independent field measurements for a variety of
organisms and ecosystems (20), including crustacean zooplank-
ton (61). This match suggests that we can account for the major
factors governing metal concentrations in aquatic animals and
that the kinetic parameters quantified in lab experiments are
applicable to natural waters.
To model the effects of HiQ and LoQ algal nutrient quality on
Daphnia MeHg concentrations, we compared the response of
Daphnia MeHgsswith the observed variation in Daphnia inges-
tion and growth rates from HiQ and LoQ treatments. We used
the grand mean of AE and Kefor these analyses, because these
values were similar between treatments. For Cf, we used the
average MeHg concentration of phytoplankton typical of un-
polluted freshwater lakes (Cf? 34 ng g?1dry weight) (ref. 50)
to apply model predictions to natural systems. To compare the
magnitude of response in MeHgsswith differences in growth and
Karimi et al.
May 1, 2007 ?
vol. 104 ?
no. 18 ?
mean, upper, and lower confidence limits of specific ingestion Download full-text
rate and growth (averaged over the 5 d) for each treatment.
We thank J. Shaw, S. Glaholt, B. Mayes, L. Keyes, S. Baines and S. Palma
for lab assistance. We also gratefully acknowledge K. Cottingham, M.
Ayres, S. Kilham, R. Sterner, and an anonymous reviewer for helpful
comments. This research was supported by National Institutes of Health
Grant P42 ESO7373–7 (to C.L.F. and C.Y.C.), the National Institute of
Environmental Health Sciences, National Science Foundation Grant
CHE-0221934, and CALFED 03WRAG0038 (to N.S.F.).
1. United Nations Environment Programme (2002) Global Mercury Assessment
(Inter-Organization Programme for the Sound Management of Chemicals,
2. Stow CA, Carpenter SR (1994) Environ Sci Technol 28:1543–1549.
3. Sunda WG, Huntsman SA (1998) Sci Total Environ 219:165–181.
4. Brett MT, Muller-Navarra DC, Park SK (2000) Limnol Oceanogr 45:1564–
5. Schindler DW, Kidd KA, Muir DCG, Lockhart WL (1995) Sci Total Environ
6. Doyon JF, Schetagne R, Verdon R (1998) Biogeochemistry 40:203–216.
7. Stafford CP, Hansen B, Stanford JA (2004) Transactions of the American
Fisheries Society 133:349–357.
8. Simoneau M, Lucotte M, Garceau S, Laliberte D (2005) Environ Res 98:73–82.
9. Hogan LS, Marshall E, Folt CL, Stein RA (2007) J Great Lakes Res 33:46–61.
10. Stafford CP, Haines TA (2001) Environ Toxicol Chem 20:2099–2101.
11. Dutton MD (1997) in Biology (University of Waterloo, Waterloo, ON, Can-
ada), pp 77.
12. Essington TE, Houser JN (2003) Transactions of the American Fisheries Society
13. Trudel M, Rasmussen JB (2006) Can J Fish Aquat Sci 63:1890–1902.
14. Stemberger RS, Miller EK (1998) Environ Monit Assess 51:29–51.
15. Chen CY, Stemberger RS, Kamman NC, Mayes BM, Folt CL (2005) Ecotoxi-
16. Sterner RW, Hessen DO (1994) Annu Rev Ecol Syst 25:1–29.
17. Urabe J, Clasen J, Sterner RW (1997) Limnol Oceanogr 42:1436–1443.
18. Hessen DO, Faerovig PJ, Andersen T (2002) Ecology 83:1886–1898.
19. Elser JJ, Sterner RW, Gorokhova E, Fagan WF, Markow TA, Cotner JB,
Harrison JF, Hobbie SE, Odell GM, Weider LJ (2000) Ecol Lett 3:540–550.
20. Luoma SN, Rainbow PS (2005) Environ Sci Technol 39:1921–1931.
21. Tsui MTK, Wang WX (2004) Environ Toxicol Chem 23:1504–1511.
22. Tsui MTK, Wang WX (2004) Aquat Toxicol 70:245–256.
23. Yu RQ, Wang WX (2002) Limnol Oceanogr 47:495–504.
24. Tsui MTK, Wang WX (2004) Environ Sci Technol 38:808–816.
25. Karimi RK, Folt CL (2006) Ecol Lett 9:1273–1283.
26. Jorgenson JL (2001) Environ Health Perspect 109:113–139.
27. Sterner RW, Hagemeier DD, Smith WL (1993a) Limnol Oceanogr 38:857–871.
28. Kilham SS, Kreeger DA, Goulden CE, Lynn SG (1997) Freshw Biol 38:639–
29. DeMott WR, Gulati RD, Siewertsen K (1998) Limnol Oceanogr 43:1147–1161.
30. Hessen DO, Agren GI, Anderson TR, Elser JJ, De Ruiter PC (2004) Ecology
31. Darchambeau F, Faerovig PJ, Hessen DO (2003) Oecologia 136:336–346.
32. DeMott WR, Pape BJ (2005) Oecologia 142:20–27.
33. Plath K, Boersma M (2001) Ecology 82:1260–1269.
34. Darchambeau F, Thys I (2005) J Plankton Res 27:227–236.
35. Sterner RW, Smith RF (1993b) Bull Marine Sci 53:228–239.
36. Sibley RM (1981) in Physiological Ecology: an Evolutionary Approach to
Resource Use, eds Townsend CR, Callow P (Blackwell Scientific, Oxford), pp
37. Chen CY, Folt CL (1993) J Plankton Res 15:1247–1261.
38. Yearsley J, Tolkamp BJ, Illius AW (2001) P Nutr Soc 60:145–156.
39. Sterner RW (1993c) Ecology 74:2351–2360.
40. Pickhardt PC, Folt CL, Chen CY, Klaue B, Blum JD (2002) Proc Natl Acad Sci
41. Folt CL, Chen CY, Pickhardt PC (2002) in Biomarkers of Environmentally
Associated Disease: Technologies, Concepts, and Perspectives, eds Wilson SH,
Suk WA (CRC/Lewis, Boca Raton, FL), pp. 287–304.
42. Pickhardt PC, Folt CL, Chen CY, Klaue B, Blum JD (2005) Sci Total Environ
43. Chen CY, Folt CL (2005) Environ Sci Technol 39:115–121.
44. Larsson P (1992) Environ Sci Technol 26:346–352.
45. Berglund O, Larsson P, Ewald G, Okla L (2001) Ecology 82:1078–1088.
46. Taylor WD, Carey JH, Lean DRS, McQueen DJ (1991) Can J Fish Aquat Sci
47. Nichols HW (1973) in Handbook of Phycological Methods, ed Stein JR
(Cambridge Univ Press, Cambridge, England), pp 7–24.
48. Goulden CE, Comotto RM, Hendrickson, Jr, JA, Hornig LL, Johnson KL
(1982) in Aquatic Toxicology and Hazard Assessment: Fifth Conference, Amer-
ican Society for Testing and Materials Special Technical Publication 766, ed
Pearson JG, Foster RB, Bishop WE (American Society for Testing and
Materials, Philadelphia, PA), pp 139–160.
49. Kilham SS, Kreeger DA, Lynn SG, Goulden CE, Herrera L (1998) Hydrobio-
50. Watras CJ, Back RC, Halvorsen S, Hudson RJM, Morrison KA, Wente SP
(1998) Sci Total Environ 219:183–208.
51. Pickhardt PC, Stepanova M, Fisher NS (2006) Environ Toxicol Chem 25:2132–
52. Tsui MTK, Wang WX (2005) Environ Toxicol Chem 24:2927–2933.
53. Tsui MTK, Wang WX (2006) Environ Sci Technol 40:4025–4030.
54. Fisher NS, Bjerregaard P, Fowler SW (1983) Limnol Oceanogr 28:432–447.
55. Burns CW (1995) Oecologia 101:234–244.
56. Peters RH (1984) in A Manual on Methods for the Assessment of Secondary
Productivity in Fresh Waters, eds Downing JA, Rigler RH (Blackwell Scientific,
Oxford), pp 336–412.
57. Wang WX, Fisher NS (1999) Environ Toxicol Chem 18:2034–2045.
58. Geisser S, Greenhouse SW (1958) Ann Math Stat 29:885–891.
59. Reinfelder JR, Fisher NS, Luoma SN, Nichols JW, Wang WX (1998) Sci Total
60. Hall BD, Bodaly RA, Fudge RJP, Rudd JWM, Rosenberg DM (1997) Water
Air Soil Pollut 100:13–24.
61. Fisher NS, Stupakoff I, Sanudo-Wilhelmy S, Wang WX, Teyssie JL, Fowler
SW, Crusius J (2000) Mar Ecol Prog Ser 194:211–218.
www.pnas.org?cgi?doi?10.1073?pnas.0611261104Karimi et al.