TheRoles ofFoodandW aterinthe
Bioaccum ulationofO rganochlorine
Com pounds inH ighM ountainLake
J O R D I C A T A L A N * A N D
Center for Advanced Studies of Blanes (CEAB-CSIC), Acce Âs
Cala St. Francesc, 14, 17300-Blanes, Catalonia, Spain
M A R CV E N T U R A
I N G R I DV I V E S A N DJ O A N O . G R I M A L T
Institute of Chemical and Environmental Research
(ICER-CSIC), Jordi Girona, 18,
08034-Barcelona, Catalonia, Spain
An integrated study encompassing the distribution of
organochlorine compounds (OC) in water, food web
(chironomids, terrestrial insects, cladocerans, mollusks,
and cyanobacteria), and fish (brown trout) froma high
in these compartments have been determined to assess
by analysis of fish stomach content and food web stable
isotopes (δ13C and δ15N). OCs withoctanol-water partition
tions infoodthanexpectedfromtheoretical octanol-water
does not reach equilibriumwithin the life span of the
food web organisms (ca. 1 year). On the other hand, the
in the case of the largest compound analyzed (seven
chlorine substituents, PCB # 180). OC exchange at fish gill
and gut has been evaluated using a fugacity model
based on the water, food, and fish concentrations. All
compounds exhibited a net gill loss and a net gut uptake.
for the compounds in apparent steady state gave values
of days to a few weeks for HCHs, 1 year for HCB and 4,4′-
DDE, and 2-3 years for 4,4′-DDT and PCB# 28 and
PCB# 52. Residence times longer than one decade were
found for the more chlorinated PCB.
atlocationsclosertotheemission sites.However, persistent
organic pollutants including some organochlorine com-
dilution effect. Natural distillation and condensation pro-
cesses concurrent with atmospheric transport lead to their
(1-4) or high elevations (5). OCs are mobilized in areas of
warm temperatures (ca. mean annual temperature > 5 °C
(6)). The more volatile compounds, such as hexachloroben-
zene (HCB), hexachlorocyclohexanes (HCH), and low chlo-
in cold areas located beyond 60° N, with mean annual air
temperatures below -5 °C (6). In contrast, the less volatile
vapor pressure <10-2.5Pa) and DDTs, are also selectively
trappedin mountain coldareas(5),which donotreach such
low temperatures as the Arctic zone.
Mountain lakes are relatively small in size and very
oligotrophic (7), food isscarce, food websareshort, and fish
show an opportunistic behavior related to the seasonal
web pathways to fish in these environments is scarce (8),
In thepresent paperwereport an assessment ofthefood
pathways to brown trout in a mountain lake using stable
isotopes, diet evaluation, and OC content of the food web.
Concentrations of OC in food and fish are also compared to
thetheoretical values expected from their bioconcentration
from water levels. Finally, a fugacity model based on the
measured OC concentrations in water, food, and fish has
been used toevaluatetherolesofthegill and gutexchanges.
of OC bioaccumulation in fish from high mountain lakes.
M aterials andM ethods
Study Site. Lake Redon (42°38′N, 0°46′E) is situated at 2240
m above sea level in Central Pyrenees (Catalonia, Spain). It
has a surface area of 24 ha, a maximum depth of 73 m, a
mean water residence time of about 4 yr, and is usually ice-
covered from late December to June (9). The lake is
oligotrophic because most of its small watershed (155 ha) is
bare rock, and the rest are alpine meadows with scarcely
developed soils. The productivity patterns and seasonal
changes in the water column are typical for high mountain
lakes (7). The lake contains a population of brown trout
(Salmo trutta) (10), from which specimens up to 15 years
have been collected (11). OC inputs are only related to
atmospheric deposition (12). The composition of OC in the
waters (13, 14), sediment, and fish (5) of this lake has been
described in previous studies.
Sample Collection and Handling. Fish were collected
with a series of eight individual bottom gillnets of different
mesh sizes(10-46mm) designedtogivethebesttheoretical
catch of brown trout over a range of 10-45 cm. The nets
were set perpendicular to the shore at various depths and
exposed in the lake for 120 min just at sunrise and sunset.
site. Muscle fillets and stomach contents were wrapped in
precleaned aluminum foil and kept frozen (-20 °C) until
analysis. Brown trout analyzed for OCs (n ) 10) averaged
(mean (SD) 265 ( 59 mm in length, 204 ( 118 g in weight,
0.99 ( 0.09 in condition factor, and 7 ( 6 years in age.
A survey of the main food chain lake components was
carriedoutin parallel duringthesamedaysoffish sampling.
species, and plankton nets for zooplankton. Samples were
kept cold during transport and were later identified and
separated in the lab into distinct classes for stable-isotope
and OC analysis.
in the field and kept cold until arrival to the lab where they
content was determined mostly up to genus or family level,
and the relative percentage was estimated on volume basis
*Corresponding author phone: 34 972 33 61 01; fax: 34 972 33
78 06; e-mail: firstname.lastname@example.org.
Environ. Sci. Technol. 2004, 38, 4269-4275
10.1021/es040035p CCC: $27.50
Published on Web 07/15/2004
2004 American Chemical SocietyVOL. 38, NO. 16, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY94269
for each fish stomach. The degree of stomach fullness was
categorized between 0 (empty) and 5 (full).
Dry Weight and Lipid Content. The percentage of water
content in muscle (74.2 ( 1.8%, n ) 8) was estimated by
The brown trout lipid content in muscle was determined
from C and N elemental analysis assuming that the main
body constituentswerelipids, proteins, ash, and chitin (15).
Lipid percentage was estimated from elemental C content
after subtraction of protein, ash, and chitin weight. Protein
content was calculated from elemental N composition after
multiplying by 6.25. Literature values were used for ash (16,
17) and chitin content (18). Since cyanobacteria use carbo-
hydrates as energy reserve, the lipid content for Nostoc was
taken from the literature (19).
Stable Isotope Analysis. Stable isotope ratios were
analyzed using a Delta C Finnigan MAT mass spectrometer
coupled onlinewith a Carlo Erba CHNS elemental analyzer,
via a Finnigan conflo 2 interface. Atmospheric nitrogen and
Peedee Belemnite carbonate were used as reference. Re-
producibility was better than 0.1½ and 0.3½ for δ13C and
Organochlorine Compounds Analysis. Fish muscle tis-
sues(5g wet weight, about 0.75g dry weight) wereextracted
and analyzed for OCs as described elsewhere (20). OCs in
invertebrates and Nostoc were determined by grouping
individuals in common samples until enough material was
accumulated for proper quantification. Combination was
carried out at the species level when enough material was
available. Organisms that were too small or scarce in the
were composite based on common feeding habits, way of
living, and exposure to fish predation (Table 1). Samples
were analyzed following a slightly modified method as
previously described (21). The isolated OC fractions of fish
and food web were analyzed in a gas chromatograph
equipped with an electron capture detection (GC-ECD;
Hewlett-Packard 5890 Series II) and a 50 m × 0.25 mm i.d.
DB-5 capillary column coated with 5% phenyl 95% meth-
ylpolysiloxane (film thickness 0.25 µm; J&W Scientific,
Folsom, CA). The injector operated in splitless mode, and
the oven temperature program started at 90 °C (held for 1
min), to 120°C at 10°C/min, and then to 310°C at 4°C/min
(holding time 15 min). Injector and detector temperatures
were 270 °C and 310 °C, respectively. Stringent precautions
were observed for maintenance of the injector under clean
from linearity and increase the limits of detection and
quantification. Helium and nitrogen were used as carrier
(0.33mL/min) andmakeup(60mL/min) gases,respectively.
Some samples were examined by negative ion chemical
(GC-MS-NICI) for structural identification. These analyses
were performed in an Agilent Technologies 6890A gas
chromatograph equipped with a nonpolar fused silica
capillary column HP5-MS (30 m × 0.25 mm i.d. × 0.25 µm
film thickness) coated with 5% phenyl 95% methylpolysi-
loxane and coupled to a MS detector 5973N. Ion source and
Ammonia was chosen as ionization gas (1.6 Torr). Helium
analyzed for every set of six samples. The recovery of the
of all studied compounds were performed by injection of
external standards at different concentrations. Relative
responses to tetrachloronaphthalene and octachloronaph-
thalene were used in order to correct for instrumental
TABLE1. FishFood-WebCom ponents Analyzedfor O rganochlorine Com poundsa
Salmo trutta (brown trout)
pelagic and littoral
several species but all
included because found in
the fish stomachs
aThe degree of taxonomomic resolution was conditioned by the amount of available material for analyses.
42709ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 16, 2004
of the surrogate standards.
Brown Trout Diet. Sampling was performed in June and
November 2000 to cover two extremes of the fish feeding
food components (e.g. chironomids) decreases with water
temperature as autumn advances. Daily energy (food)
requirements also decrease in parallel. These 2 months are
thereforerepresentativecasesofhigh and low food demand
andavailability,respectively.In June,when larvaeofaquatic
insects were abundant, the average value of the stomach
fullness index was 4.4, and no fish with empty stomach was
found. In November, during lake overturn and low water
temperature, the average value of the index was 1.9, and 9%
of the fish had empty stomach. The food items found in the
stomachs and their relative contribution were also distinct
in the two periods (Table 2). In June, chironomids, either
larvae or pupae, were by large the most frequent and
abundant. Other organisms, such as terrestrial insects, the
but their contribution to the food volume was low. In some
stomachs chironomid pupae were nearly the sole content,
probably because this transient and passive stage facilitates
the capture by trout. In November, the more frequent and
abundant food components were cladocerans, the pelagic
Daphnia, and the littoral Eurycercus. However, they were
less dominant that chironomids in June. Other items were
also relevant either in abundance or frequency (e.g. chi-
ronomids, Radix, Pisidium, Sialis). During this period, the
number of unidentifiable items increased, although their
food items such as colonies of the cyanobacterium Nostoc
were also found.
These two snapshots of the trout diet at contrasting
periods of the year suggest that these fish mainly feed on
particularly small cladocerans such as Daphnia, when they
are scarce. The large variety of other food items is comple-
and nitrogen stable isotope ratios between the distinct food
web components provide information on trophic relation-
ships. Commonly used trophic fractionation values are 1½
for δ13C and 3.4½ for δ15N (22, 23), which are similar to
mean values found in recent studies of the variation in the
trophicfractionation(0.05(0.63½ δ13C,3.49(0.23½ δ15N)
(24). Due to its lower trophic fractionation, carbon is
considered to indicate primary energy sources (e.g. benthic
vs pelagic photosynthesis), and nitrogen is used for the
discrimination among trophic levels.
The isotopic signatures of the organisms and primary
carbon sources involved in thefood web pathways to fish in
LakeRedon areshown in Figure1.Themain primarycarbon
sourcessnamely littoral (epilithon and Nostoc), pelagic
(seston), and organic detritus (sediment)sshowed signifi-
cantly distinct isotopic signatures. The δ13C depletion was
larger in seston and deep sediment than in littoral algae,
which may reflect the predominant occurrence of primary
production in the hypolimnion, below the seasonal ther-
(9).Growingtemperaturein thehypolimnion issignificantly
lowerthan in thelittoral (ca.5-10°C),andavailableCO2has
a larger contribution from within lake respiration (25). On
theotherhand, benthic algaetend tobeenriched in13C, due
to a boundary layer effect, involving a limitation of CO2
diffusion to the cells that favors the use of bicarbonate as
carbon substrate (26).
Fractionation during nitrogen assimilation by algae (phy-
toplankton and phytobenthos) can be -4 to -5½ (27). In
our data, pooled epilithon (mainly diatoms) and Nostoc
agreed with thesevaluesassumingnitrogen sourcescloseto
δ15N as expected from its predominant atmospheric
origin. However, seston was slightly richer in the heavy
and allocthonous matter in the former. Since Daphnia, a
a δ15N similar to other herbivores feeding on littoral algae,
a common δ15N baseline for the herbivore food web around
-4½ was assumed.
TABLE2. BrownTrout Diet inLake Redonduring TwoDistinct
Periods of the Ice-Free Season(June andN ovem ber)a
aCommonness inthedietis indicatedby thefrequency thatacertain
item was found in the stomachs examined (n )21and 22, respectively,
for high and low food periods). Contribution to the diet is indicated by
the percentage of food volume, obtained by weighting percentages of
food volume in individual stomachs to the degree of stomach fullness.
Annualaveragecontributiontodietwas calculatedassuming 4months
of high requirements, 4 months of low, and 4 months of very low
consumption (stomach fullness <20%) during the ice covered period
with a diet similar to the low food requirement period of the ice-free
season. Other reasonable assumptions (i.e., different monthly periods
for each type of consumption) do not significantly change the average
FIGURE1. Isotopic signatureofthemainfood-webcomponents in
Sample description in Table 1.
VOL. 38, NO. 16, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 94271
As expected, brown trout appear to be the unique top
predator (Figure 1). However, the average food chain from
an enrichment factorof3.5½ δ15N pertrophic level change,
the average number of energy transfer steps from primary
producers to trout is only of 2.2. This is not surprising for a
high mountain lake because food availability is scarce due
to the oligotrophy of the system, the low inputs from the
watershed, and the small lake size (28).
δ15N gradient indicates a high degree of omnivory. The
differences between successive organisms in any trophic
show higher δ15N values than larvae, although in both cases
the species measured were herbivorous. The isotope dis-
crimination could be due to metamorphosis from larvae to
pupae. The new form is rebuilt from old tissues, and the
transformation may cause an enrichment in15N in a similar
way as it occurs in starving animals (29) because there is no
diet, particularly thebivalvePisidium. Theδ15N signatureof
some supposed predators (Arcynopterix, Platambus) do not
agree with an exclusive diet of macroinvertebrates, being
littoral, and sediment carbon sources at higher food web
The isotopic signatures are consistent with the stomach
content observations indicating that trout mainly predate
on the herbivore level, constituted by chironomids and
cladocerans, with some contribution from other inverte-
Table2, and themeasured isotopic signaturesofthedistinct
food items (δ13Ci, δ15Ni) the expected isotopic signature for
brown trout was calculated as follows:
The resulting isotopic values are -22.7 ( 1.8½ δ13C and 3.4
δ15N, which are quite similar to the direct fish
determinations,-22.6(1.5½ δ13C and3.5(2½ δ15N.Thus,
the above assumptions on average diet composition are
feasible and can be considered to provide a good estimate
for the annual average composition.
differences in OC concentrations were observed among the
distinct food-web components reflecting in part their large
heterogeneity in lipid content (Table 3). However, when
TABLE3. O rganochlorine Concentrations (ng g-1Dry Weight) inthe M ost Significant O rganism s Involvedinthe BrownTrout Diet
r-HCH γ-HCH HCB4,4′-DDE 4,4′-DDT PCB-28 PCB-52 PCB-101 PCB-118 PCB-153 PCB-138 PCB-180
ng g-1dry weight
Salmo trutta (muscle)
FIGURE 2. Concentrations of the organochlorine compounds infishandfood-webstandardizedby lipidcontent. Horizontal bars indicate
the expected values according to the concentrations in water (n ) 8) and Kow (31) (values are indicated for high (maximum) and low
(minimum) food periods, and mean diet is calculated as described in Table 2).
δ13Ctrout) Σpiδ13Ci+ 0.05 and δ15Ntrout) Σpiδ15Ni+ 3.5
42729ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 16, 2004
normalized to lipid concentration, brown trout showed the
in the trout diet, Daphnia showed lower OC concentration
valuesthan average, perhapsbecauseofitsshorterlifespan.
Combination of the concentrations of the individual organ-
isms to estimate OC content in food shows that the
the low feeding period (Figure 2). This difference was
to the diet. Mean OC intake was therefore calculated by
weighting the lipid normalized concentrations in each food
item according to the respective mean contributions to diet
Comparison ofthemean OC pooled food concentrations
and Kow(30) (Table 4) provides an estimate of the deviation
of OC in food web content (normalized to lipids) from
thermodynamic equilibrium. A significant number of OCs
show concentrations that are close to those expected at
equilibrium, namely HCHs, HCB, 4,4′-DDT and PCB con-
geners#28and #52(Figure2).4,4′-DDE exhibit valuesmuch
higherthan expectedatequilibrium (Figure2),but4,4′-DDT
of 4,4′-DDT into 4,4′-DDE within the organisms (31). In this
respect, 4,4′-DDT is always significantly higher than 4,4′-
DDE in water (Table 4).
The more chlorinated PCB congeners also deviate from
equilibrium since, according to waterconcentrations, lower
values than expected are found. The deviation increases
larger at higher Kow(Figure 3a). Equilibrium is therefore not
reached for OCs with log(Kow) values above ca. 6, indicating
that fortheseOC it could not bereached within thelifespan
(<1 yr) of the organisms more relevant in the trout diet.
In contrast to food, fish show OC concentrations signifi-
cantly higher than those expected at equilibrium which
of food intake. The ratio between the fish and food con-
3.8 for PCB #28 to 16.5 for PCB #138. The ratios between OC
in fish and food are also related to Kow(Figure 3b), since the
higher is Kowthe lower is its loss through the gills. PCB #180
is an exception to this trend (Figure 3b) since it exhibits
Theanomaly could reflect thelowermembranepermeation
of large molecules (32). In consequence, trout uptake
efficiency for PCB#180 seems to be significantly lower than
for the other congeners.
The Roles of Food and Water in OC Bioaccumulation.
OC concentration in fish results from the balance between
exchanges at the gills during fish respiration, uptake from
diet, elimination by fecal egestion, and metabolism and
dilution by fish growth. Among thesemetabolic elimination
The relative significance of the other flux components
can be evaluated using a fugacity approach (34) in combina-
tion with a brown trout food intake model (35) and the
measured concentrations in water, food, and fish. For this
purpose, a ªstandard fishº weighting 204 g submitted to the
seasonality of feeding over year periods divided in three
ªweather conditionsº each involving 4months with average
water temperatures of 8, 4, and 2 °C in an environment of
ca. 9 mg O2L-1has been taken as lake representative (9). All
water flows with oxygen and OCs were transferred inside
and outside by diffusion (30). Conductivities (Dw) for gill
uptake and loss were equally considered to be dependent
(Fg) was determined by the difference between water (fW)
and fish (fF) fugacities.
The exchanges in the gastrointestinal tract are more
complex. Between food uptake and egestion a fraction of
matter is removed and there is, in addition, lipid digestion.
Therefore, the intestine flux (Fi) required the separate
TABLE4. Concentrations of O rganochlorine Com pounds
DissolvedinWater of Lake Redona
aData from refs 13 and 14 and unpublished.bAverage values of
data collected in J uly 96 at 1, 5, and 60 m depth; J une 97 at 1 m, 5, 25,
and 59 m depth (n ) 4); and November 00 at 1 m depth (n ) 8).
FIGURE 3. (A) Departure fromoctanol-water equilibriumof the
organochlorine compounds in the fish food. (B) Biomagnification
ratioof the organochlorine compounds betweenfishandaverage
Fg ) Dw(fW- fF) (2)
VOL. 38, NO. 16, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 94273
consideration of uptake (DA) and loss (DE) conductivities
where fAis food fugacity.
rate (GA) and gut absorption efficiency (EA) (20, 33, 36). DE
was taken proportional to GA(1-?), where? was thefraction
of ingested diet absorbed by the organism. Development of
the two equations according to the definitions in Table 5
provided the following expressions for the two fluxes:
Apart from water and food concentrations, the relative
flux differences result from the values of the Kow and EA
coefficients, the latter being particularly relevant for dif-
ferentiating the behavior of PCB#180. Gw and GAdetermine
the absolute flux values. The two rates depend on the fish
daily energy requirements (37), which under optimal condi-
tions result from the body weight and temperature in a
nonlinear way. GA was estimated using Elliot's model for
mentioned above. GW, in addition to oxygen consumption
as determined by the energy requirement, was made de-
pendent on the oxygen in the water and uptake efficiency
(36) (Table 5).
The resulting calculations show that a number of OCs
were close to a steady state, namely HCB, 4,4′-DDE, PCB
#28, and PCB #52 (Figure 4). These are the compounds that,
in turn,arealsoin equilibrium between waterandfood.This
parallelism gives ground to the assumptions for the calcula-
Since HCHs and 4,4′-DDT have also log(Kow) < 6, steady
state for these OC should be expected, but this is not the
case. In 4,4′-DDT gut uptake appears to be higher than gill
the fish. In HCHs, gill loss is much higher than gut uptake,
correspond to the one presently experienced by fish. This
can certainly be the case as the calculations are based on
long time water concentrations averages, and fish renewal
time of the more volatile compounds is a matter of days.
An average residence time in fish can be calculated for
the compounds in apparent steady state. For HCHs it is in
it is about 1 year, and for 4,4′-DDT, PCB #28, and #52 it is
about 2 or 3 years. For the rest of the compounds steady
state is not achieved, but the present turnover indicates
characteristic times around a decade for PCB #101 and two
or three decades for PCBs #110 to #153. In the case of PCB
#180, a fish could hardly achieve a steady state at present
exposures unless it lived for centuries.
This is a contribution of the Limnology Group (UB-CSIC).
Financial support from the EU projects EUROLIMPACS
(GOCE-CT-2003-505540) and EMERGE (EVK1-CT-1999-
00032) and ACA-CIRIT (Generalitat de Catalunya) is ac-
for a Ph.D. fellowship and acknowledges Universitat Au-
to Ánoma de Barcelona. Guillem Carrera is thanked for the
analysis of OC in water. N. Garcõ Â a-Reyero (IBMB, CSIC), J.C.
Massabuau (University of Bordeaux 1), and R. Lackner
(University ofInnsbruck) arethanked fortheirvaluablehelp
TABLE5. Definitionof Sym bols andSum m ary of the Param eters U sedinthe O C FishFlux Calculations
fW, fA, fF
Pa L pg-1
water, food and fish fugacities, fi) CiZi-1
fugacity capacity, Zi) LiKowH-1
octanol-water partition coefficient
Henry's law constant
conductivity at gills, Dw ) Gw Zw
gill ventilation rate, Gw ) K keoO2w-1Eox
fish daily intake requirement, K) f(T,W) (33)
energy to oxygen consumption coeff, 0.047
water oxygen concentration
efficiency of oxygen uptake, 0.45
gut uptake conductivity, DA) EAGAZA
food consumption rate, GA) K kef
energy to food volume consumption coeff, 10-6
gut uptake efficiency, 0.75 (except for
PCB #180, 0.45)
gut loss conductivity, DE) GA(1-?) ZF
fraction of ingested diet absorbed by the fish, 0.8
Fi ) DAfA- DEfF
Fg ) Gw(Cw - CF/(LFKow))(4)
Fi ) GA(CAEA- CF(1-?))(5)
FIGURE 4. Comparison of the calculated net gill loss and net gut
uptake of organochlorine compounds in a fish of 204 g in Lake
Redon according to the measured concentrations in water, food,
42749ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 16, 2004
(1) Simonich, S. L.; Hites, R. A. Science 1995, 269, 1851-1854.
(2) Wania, F.; Mackay, D. Ambio 1993, 22, 10-18.
D. B.; Rosenberg, B. Nature 1998, 395, 585-588.
(4) Fellin, P., Barrie, L., Doughtery, D.; Toom, D.; Muir, D.; Grift,
N.; Lockhart L.; Billeck, B. Environ. Toxicol. Chem. 1994, 15,
(5) Grimalt,J.O.;Fernandez,P.;Berdie Â,L.;Vilanova,R.M.;Catalan,
J.; Psenner, R.; Hofer, R.; Appleby, P. G.; Rosseland, B. O.; Lien,
L.; Massabuau, J. C.; Battarbee, R. W. Environ. Sci. Technol.
2001, 35, 2690-2697.
(7) Catalan, J.; Ventura, M.; Brancelj, A.; Granados, I.; Thies, H.;
Nickus, U.; Korhola, A.; Lotter, A. F.; Barbieri, A.; Stuchlõ Â k, E.;
Lien, L.;Bitus Ïõ Â k, P.;Buchaca, T.;Camarero, L.;Goudsmit, G.H.;
Kopa Âc Ïek, J.; Lemcke, G.; Livingstone, D. M.; Mu Èller, B.; Rautio,
M.; S ˇis Ïko, M.; Sorvari, S.; S ˇporka, F.; Strunecky Â, O.; Toro, M. J.
Paleolimnol. 2000, 28, 25-46.
(8) Knapp, R. A., Matthews, K. R., Sarnelle, O. Ecol. Monogr. 2001,
(9) Ventura, M.; Camarero, L.; Buchaca, T.; Bartumeus, F.; Living-
stone, M. D.; Catalan, J. J. Limnol. 2000, 59, 97-108.
(10) Encina, L., Rodrõ Â guez-Ruiz, A. Hydrobiologia 2003, 493, 35-45.
(11) Vives, I.; Grimalt, J. O.; Ventura, M.; Catalan, J.; Rosseland, B.
Environ. Pollut. 2003, in press.
(12) Carrera,G.;Ferna Ândez,P.;Grimalt,J.O.;Ventura,M.;Camarero,
Ll.; Catala Ân, J.; Nickus, U.; Thies, H.; Psenner, R. Environ. Sci.
Technol. 2002, 36, 2581-2588.
(13) Vilanova,R.;Ferna Ândez,P.;Martõ Â nez,C.;Grimalt,J.O.J.Environ
Qual. 2001, 30, 1286-1295.
(14) Vilanova, R. M.; Fernandez, P.; Grimalt, J. O. Sci. Total Environ.
2001, 279, 51-62.
(15) Sterner, R. W.; Elser, J. J. Ecological Stoichiometry. TheBiology
Press: Princeton, 2002; p 439.
(16) Salonen, K.; Sarvala, J. Verh. Internat. Verein. Limnol. 1978, 20,
(17) Stelzer, R. S.; Lamberti, G. A. Ecology 2002, 83, 1039-1051.
(18) Cauchie, H.-M. Hydrobiologia 2002, 470, 63-96.
(19) Ahlgren, G.; Gustafsson, I.-B.; Boberg, M. J. Phycol. 1992, 28,
(20) Berdie Â, L.; Grimalt, J. J. Chromatogr. A 1998, 823, 373-380.
(21) Vives, I.; Grimalt, J. J. Chromatogr. B 2002, 768, 247-254.
(22) Minagawa, M.; Wada, E. Geochim. Cosmochim. Acta 1984, 50,
(23) Michener, R. H.; Schell, D. M. In Stableisotopesin ecology and
environmental science; Lajtha, K., Michener, R., Eds.; Blackwell
Scientific: 1994; pp 138-157.
(24) Vander Zanden, M. J.; Rasmussen, J. B. Limnol. Oceangr. 2001,
(25) Rau, G. H. Can. J. Fish. Aquat. Sci. 2001, 37, 742-746.
(26) Hecky, R. E.; Hesslein, R. H. Can. J. Fish. Aquat. Sci. 1995, 52,
production. In Organic Geochemistry; Engel, M. H., Macko, S.
A., Eds.; Plenum: 1993; pp 73-100.
(28) Post, D. M. Pace, M. L.; Hairston, N. G., Jr. Nature 2000, 405,
(29) Hobson, K. A.; Alisauskas, R. T.; Clark, R. G. Condor 1993, 95,
(30) Clark, K.E.;Gobas, F.A.P.C.;Mackay, D.Environ.Sci.Technol.
1990, 24, 1203-1213.
(31) Lund, B. Organohalogen Compd. 1998, 37, 341-344.
(32) Mackay, D. Environ. Sci. Technol. 1982, 16, 274-278.
(33) Campens, J.; Mackay, D. Environ. Sci. Technol. 1997, 31, 557-
(34) Mackay, D.; Fraser, A. Environ. Pollut. 2000, 110, 375-391.
(35) Elliott, J. M.; Hurley, M. A. Freshwater Biol. 1998, 39, 339-349.
G. D. Environ. Sci. Technol. 1997, 31, 3267-3273.
(37) Elliott, J. M. Quantitative ecology and the brown trout; Oxford
University Press: Oxford, U.K., 1994; p 286.
Received for review March 8, 2004. Revised manuscript re-
ceived June 4, 2004. Accepted June 4, 2004.
VOL. 38, NO. 16, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 94275