Limnol. Oceanogr., 31(l), 1986, 79-88
0 1986, by the American
Society of Limnology
and Oceanography,
Inc. .
Sources of carbon and sulfur nutrition for consumers in
three meromictic lakes of New -York State’
Brian Fry2
Department of Biology, Jordan Hall 142, Indiana University, Bloomington 47405
Abstract
The trophic importance of bacterioplankton as a source of carbon and sulfur nutrition for
consumers in meromictic lakes was tested using stable carbon (613C)and sulfur (c?~~S)
isotopic
measurements. Studies in three lakes near Syracuse, New York, showed that most consumers
ultimately derive their C and S nutrition from a mixture of terrestrial detritus, phytoplankton, and
littoral vegetation, rather than from bacterioplankton. Food webs in these meromictic lakes are
thus similar to those in other lakes that lack dense populations of bacterioplankton.
annual primary production
occurs at the
plate, a food web based largely on bacterioplankton has been postulated (Culver and
Brunskill 1969).
Measurements of natural abundances of
stable isotopes can be used to test such hypotheses about food web structure. Stable
isotopic studies show which primary producers at the base of food webs are the most
important general sources of nutrition for
consumers (Fry and Sherr 1984). Although
direct trophic connectedness between species
is not revealed by 613C and 634S measurements, general pathways of carbon, sulfur,
and energy transfer from sources of primary
production to consumers of all trophic levels
become evident. Consumers are expected to
have isotopic compositions similar to those
of bacterioplankton
if bacterioplankton
is
the major source of C and S nutrition in
meromictic lakes. Slight deviations from an
exact correspondence between primary producer and consumer isotopic compositions
are expected for C, but not for S, since isotopic compositions increase about 1% per
trophic level for C, but not for S (Fry and
Sherr 1984).
Physical measurements were made in
FGL to show the persistence of meromictic
conditions in the 1983-1984 study period.
Bacterioplankton
and various aquatic and
terrestrial plants were collected to ascertain
the isotopic compositions of sources of organic C and S available to consumers. Because of the large samples required for 634S
analyses (> 100 mg dry wt per sample), fish
consumers in particular were collected to
test for food chains based on bacterioplankton. Since fish are top carnivores, their tis-
Meromictic lakes can support dense populations of anaerobic, photosynthetic sulfur
bacteria. Permanently
stagnant bottom
waters of the monimolimnion
are often rich
in nitrogen, phosphorus, and sulfide necessary for bacterial growth; a further growth
requirement, light, dictates that photosynthetic bacteria be positioned at the top of
the monimolimnion.
Fayetteville Green Lake (FGL), 13 km
east of Syracuse, New York, in Green Lakes
State Park, is a classic meromictic lake (Eggleton 1956; Brunskill and Ludlam 1969).
Small, deep, and protected from winds that
might cause mixing and overturn, the lake
also has a high salt concentration that enhances stratification. Anoxic bottom waters
contain high concentrations of N and P nutrients, sulfate (N 15 mM), and sulfide (O1.2 mM) (Deevey et al. 1963; Turano and
Rand 1967; Torgersen et al. 198 1). A bacterial plate is found at the top of the monimolimnion
at 17-20 m from spring to fall;
divers’ observations indicate that the upper
boundary of the plate is well defined, with
an abrupt transition from clear epilimnetic
water to dark, bacteria-rich,
monimolimnetic water occurring in a vertical distance
of < 10 cm (Frey 1967; T. Field pers.
comm.). Dense populations of zooplankton
are associated with the upper surface of the
bacterial plate (Harman 1967; Culver and
Brunskill 1969) and because most of the
l This research was supported by NASA grant NGR
15-003-l 18 to J. M. Hayes and NSF grant PCM 7%
10747 to H. Gest.
’ Present address: Ecosystems Center, MBL, Woods
Hole, Mass. 02543.
79
80
Fry
sues can be regarded as time-averaged indicators of the sources of organic matter
utilized in lake food webs. Invertebrates were
also analyzed to test the most-expected
(zooplankton)
and least-expected (littoral
crayfish, snails) food web dependence on
bacterioplankton.
A few samples were also
collected from two other nearby meromictic
lakes: Round Lake (RL), about 110 m upstream of FGL, and Green Lake (GL), about
13 km southwest of the other two in Clark’s
Reservation State Park (Effler et al. 198 1).
I thank J. Favinger and W. Ruf for assistance in the field; D. Baas identified zooplankton samples. I also thank T. Maxian,
W. Murray, and the staffs of Green Lakes
State Park, and the New York State Office
of Parks, Recreation and Historic Preservation for their cooperation in the sampling
program. B. Culligan provided information
about trout stocking and also samples of
hatchery foods. J. Hasset and M. Melsor
provided facilities and assistance for continuous flow centrifugation
of bacterial
samples; E. Ripley and J. M. Hayes provided mass spectrometer facilities for isotopic
determinations.
Methods
Temperature, light, and oxygen profiles
were taken at noon on clear, calm days by
lowering probes from a rowboat anchored
in the center of FGL. Temperature was
measured to -tO.O5”C with a Whitney Underwater Instruments
thermistor
(model
TC-5A). Light intensities were measured to
a limit of 0.1 lux with a Photomatic underwater photometer and oxygen was measured with a Yellow Springs oxygen meter
to LO.1 ppm. Water casts were made with
a precision of about 1 m with a 37-liter
Plexiglas sampler. Silver nitrate was added
immediately to water samples upon retrieval from the lake; Ag,S was recovered by
filtration and sulfate in the filtrate precipitated with BaCl,.
Bacteria were harvested by continuous
flow centrifugation of water collected at 18
m. Organic sulfur of the bacteria was analyzed after So had been repeatedly extracted
by heating with CN- to form soluble SCN(Steinmetz and Fischer 198 1); extracted cells
were pelletized by centrifugation
and then
thoroughly
washed with distilled water.
Bromine water was used to oxidize supernatant SCN- to S042- which was then precipitated with barium for later isotopic determinations.
Fish were caught from shore with hook
and line in August and October 1983; crayfish and plants were collected by hand along
shorelines. Muscle tissue was dissected from
animals for isotopic analyses. Zooplankton
was collected in 15-30-min horizontal tows
with a weighted 70-pm-mesh net and held
alive for 12 h to allow gut clearance; after
this period, elemental sulfur that could have
been associated with bacteria in zooplankton guts was not detectable (cyanide test:
Steinmetz and Fischer 198 1). Zooplankton
were then separated by size with a 250+mmesh net. Plants and animals were dried,
ground to a fine powder, washed twice with
0.1 M LiCl to remove inorganic sulfate, and
dried again. The washing procedure involved shaking the sample for 30 min in 20
volumes of 0.1 M LiCl, centrifuging, discarding the supernatant, then repeating.
Washed, dried samples were combusted under 30 atm of 0, in a Parr bomb, and BaCl,
was added to the washings of the bomb interior to precipitate BaSO,.
For isotopic determinations, BaSO, samples were decomposed to SO2 in quartz tubes
at > 1,600”C (Fry et al. 1982). Ag,S samples
were combusted to SO, in quartz tubes with
V20, as an oxidant. For carbon, 0.5-2.0mg samples were combusted to CO2 in a
Carlo-Erba elemental analyzer. CO, and SO,
were purified with cryogenic and vacuum
techniques, and the gases analyzed with isotope ratio mass spectrometers. Results are
given in 6 notation where
613C = [(‘3R,ample/~3Rstd)- l] x 1,000
and
634s = [(34Rsamp,c/34Rstd)
- I] x 1,000;
= 13c:
12C, 34R= 34S: 32Sand values are
reported relative to PDB (carbon) and CDT
(sulfur) standards. Precision of individual
measurements is about -L0.2?4&~for carbon
and + 0.3%0 for sulfur.
13R
Results and discussion
Meromixis and bacterioplankton -The
meromictic
conditions
and the bacterial
plate reported earlier in FGL persisted at
81
Food webs in meromictic lakes
LIGHT
INTENSITY (%I
1.0
45#
e
1 a . ..I.
IO
100
p45
1 . . . . I . . . I.
I . ..I
7
9
II
13 15 I7
I9 21
I . . . ..I
0
2
4
.I
6
02 8
1,.
8
IO
1
I2
. . .
23 25
0
C
PPm
Fig. 1. Light and temperature profiles, 2 1 August 1983, Fayetteville Green Lake.
the time of this study (August 1983-June
1984). For example, the strong vertical
stratification accompanying meromixis was
still present in August 1983 (Fig. 1). Temperatures were 7”-7.5”C in the monimolimnion below 20 m; below 17 m, oxygen was
absent and sulfide present (data not shown).
A subsurface temperature maximum was
present at 15-16 m; Jelacic (197 1) called
this deep temperature maximum a “nose”
and found that it was generally located l-2
m above the bacterial plate. In August 1983
the light-absorbing bacterial plate was present in the region just below the temperature
nose (Fig. 1).
Mild centrifugation (1,000 x g, 5 min) of
water from 18 m yielded two highly colored
fractions, a bright purple pellet and a dark
chocolate brown supernatant fluid. The pellet contained several morphological
types
(N = 5) of bacteria, prominent among which
were purple photosynthetic bacteria (Chromatiaceae); these cells sediment easily due
to the high density of their internal So globules. Triiper (in Field 1974) identified the
following
purple sulfur bacteria from FGL:
Chromatium vinosum, Chromatium sp.,
Thiocystis violacea, Thiocystis gelatinosa,
Lamprocystis roseopersicina, and Thiocapsa roseopersicina. The brown supernatant
fraction contained only two dominant morphological types of bacteria: rods (2.5 x 1
,um) and spheres (l-pm diam). The latter
may be the green sulfur bacterium, Chlorobium phaeobacteroides, which is actually
brown and has been isolated from FGL
(Triiper and Genovese 1968).
Water taken from 18 m appeared visibly
pink while water taken from 19 m was
brownish;
a vertical habitat segregation
seems to exist, with purple Chromatiaceae
occurring above brown Chlorobiaceae (green
sulfur bacteria). This vertical stratification
of Chromatiaceae above Chlorobiaceae is
common in meromictic lakes (Caldwell and
Tiedje 1975).
Overall, temperature, oxygen, light, microscopic, and sulfide and sulfate isotopic
data (below) all indicate that meromictic
conditions reported in earlier studies of FGL
82
Fry
Table 1. 634Sof sulfate and sulfide in Fayetteville
Green Lake.
Depth
(m)
0
5
10
15
18
20
25
30
35
40
45
50
* Separate
PS CDT
SO,=
22.9
23.2
23.1
23.3
24.4
24.8, 25.9*
25.9
25.4
25.8
27.8
28.4
27.8
s2-
-32.5
-31.8
-31.7
-31.6
-30.4
-28.6
-28.2
-27.1
A
56.9
56.6, 57.7
57.6
57.0
56.2
56.4
56.6
54.9
R = 56.5kO.8
water casts.
persisted at the time of my study. One change
is that the upper boundary of the bacterial
plate (Fig. 1) had advanced to 16- 18 m from
the 18-20-m level reported in the late 1960s
(Culver and Brunskill 1969).
634S-Both sulfate and sulfide are important sources of sulfur for primary producers
in meromictic lakes. Sulfides are typically
depleted in 34Srelative to sulfate because of
large isotopic fractionations accompanying
bacterial sulfate reduction (Chambers and
Trudinger 1979). The isotopic difference between sulfate and sulfide in FGL was large
and approximately
constant with depth
(mean difference = 56.5o/oo:Table l), in excellent agreement with a 56Y& average difference found by Deevey et al. (1963). My
closer-interval sampling showed three well
mixed layers in FGL, with the sulfate isotopic values essentially constant within these
zones: the epilimnion
(O-15 m, 634S + 23o/oo);the upper monimolimnion
(20-3 5
m 634S - + 25.7o/oo); and the lower monimblimnion (40-50 m, 634S- +28Y&). These
sulfate isotopic measurements confirm a
two-tier monimolimnetic
structure proposed by Jelacic (197 1) on the basis of temperature measurements.
Primary producers in meromictic lakes
have access to sulfate only or to sulfate plus
sulfide. Macroalgae of the epilimnion have
access only to sulfate. Macroalgae in FGL
and RL had + 18.3 to + 19.4 634S values,
only slightly depleted in 34S relative to surface sulfate (634S = +22.9 and +22.8Ym in
FGL and RL). Similar small 34S depletions
relative to environmental
sulfate have been
observed in marine and freshwater macroalgae (Kaplan et al. 1963; Mekhtiyeva and
Pankina 1968). These small 34S depletions
arise during the general process of assimilatory sulfate reduction and are characteristic of unicellular algae as well as macroalgae (Kaplan and Rittenburg 1964). Because
of this similarity in 634S values of macroalgae and microalgae, I would expect phytoplankton in FGL and RL to average about
+ 18 to + 2Oa/oo,as do macroalgae (Table 2).
Because of the large samples needed and
possible contamination
from particulate
terrestrial debris I did not collect phytoplankton for direct confirmation of this expectation.
Bacterioplankton at 17-20 m in FGL have
two possible sulfur sources, sulfate and sulfide (634S = +24.4 and - 32.57~: Table 1).
Photosynthetic
bacteria oxidize sulfide to
elemental sulfur during carbon fixation; elemental sulfur extracted by cyanolysis from
both the Chromatiaceae and Chlorobiaceae
fractions had 634S values of -27.0
to
- 30.5%. These values are slightly enriched
in 34Srelative to - 32. Sa/oo
sulfide, but clearly
related to sulfide rather than to the +24.4
sulfate. However, organic sulfur from the
same bacteria was much less closely related
to sulfide. Organic sulfur averaged about
- 10%0 (Table 2), intermediate between the
sulfide and sulfate values. In the absence of
any isotopic fractionation during sulfide uptake, the intermediate
- 1Oo/oovalue indicates that about 60% of bacterial organic
sulfur is derived from sulfide and the remainder from sulfate.
Rooted plants in the littoral zone may
also incorporate sulfides. In marine plants,
34Sdepleted sulfur can enter the roots from
the sediments (Carlson and Forrest 1982;
Fry et al. 1982). Sulfides were readily detectable in the littoral zone of FGL and RL
by odor and the black color of subsurface
sediments. Incorporation of sulfides by roots
could account for the relatively low (vs.
+22.9o/oo surface sulfate) 634Svalues of -3.9
to + 8.77~ found for an unidentified
submerged rooted plant and the emergent rooted plant Typha ZatzjXa (Table 2). A last
sulfur source important for vegetation is
Food webs in’ meromictic lakes
83
Table 2. W3Cand 634Svalues for organic carbon and organic sulfur in samples collected from Fayetteville
Green Lake and Round Lake.
Bacterioplankton, 18 m
Purple fraction (Chromatiaceae)
Brown fraction (Chlorobiacesre)
Unseparated suspension
Plants
Ulothrix*
Chara, near inlet
Chara, near outlet
Cladophora
Ruppia-like (rooted submerged plant,
unidentified)
Typha Iatifolia
Corn
Surface particulate org C
Tree leaves
Zooplankton
16 m, >250 pm
16 m, >70 pm, ~250 pm
0-5m,>70pm
16 m, >70 pm, ~250 pm
16 m, ~250 pm
Fish
Ambloplites rupestris
Fundulus sp.
Lepomis gibbosus
Lepomis gibbosuP
Micropterus dolomieui
Micropterus dolomieuP
Notemogonius crysoleucas
Salmo gairdneri (std length, mm)
1 (686)
2
3
4
5
6
(220)
(264)
(235)
(230)
(254)
i!i (260)
Crayfish, Orconectes obscurus
Crayfish, 0. obscurus**
Gastropods, Goniobasis livescens, flesh
Frog, Rana pipiens
Amphipods from littoral algae
Acidified sediment from 52 m
-39.3(1983); -41.1(1984)
-30.9(1983); -32.6(1984)
-34.8(1983)
- 11.8(1983); - 10.6(1984)
-9.4(1983); -9.7(1984)
-
-25.0
-16.4
- 22.0
-29.2
+ 19.4
+18.3
-
-15.3
-29.4
-10.1
-30.2
-29.0f
+8.7; +7.1t
-3.9
+3.4
I+2.8+0.9 (N = 6)
1.2 (N = 6)
-39.3Ik 1.2(2)
-37.0f 1.3(2)
-35.6* 1.1(5)
-
+7.1(1983)$
+ ;.6(1984)§
+ 12.1(1984)§
-26.4t- 1.6(8)
-27.3(1)11
-26.2+0.2(3)
-26.7+ 1.5(6)
-26.8* 1.4(8)
-26.4(l)
-26.1(l)
+6.0( 1)
+ lO.Of 1.9(6)
+9.2-t-0.8(6)
+10.6(l)
+5.9(l)
- 19.6 (liver)
- 18.8 (muscle)
- 17.7 (skin)
-23.4
-26.5
-26.6
- 30.28
-33.2#
-33.7
-36.7
-27.2&0.2(2)
-26.3( 1)
-29.4(l)
-24.2(l)
-29.0(1)11
-28.4
+ 10.6
+10.1
+8.5
+6.9
+7.8
+6.8
+7.8
+4.4
+4.7
+3.6
+7.4+0.2(2)
+4.2( 1)
+12.3(l)
-20.1
* Sample from Round Lake; other samples are from Fayetteville
Green Lake.
t Two separate samples.
$ ~-75% Daphnia sp.
6 295% Diaptomus
sp.
I( Composite of four or more individuals.
If Gut contents measured -27.6 and consisted of insects.
# Corn in gut (6W = - 10.1; 6% = +3.4).
** From small creek connecting
Round Lake and Fayetteville
Green Lake.
+7.7f 1.9(7)
+W1)ll
84
.
Fry
Table 3. 613Cand 634Svalues of samples collected in Green Lake at Clark’s Reservation.
6’“C
Surface S042Plants
Leaf detritus
Surface particulate org C
Fish
Lepomis gibbosus
PUB
cm+7.5
PS
-24.8
-26.9
-27.7+ 1.1(3)
+2.6+0.3(5)
-27.0& 1.2(6)
+ 1.9+ 1.8(5)
Crayfish
Orconectes obscurus
sulfate deposited from the atmosphere. Tree
leaves in the FGL watershed averaged
+2.8% (Table 2), a value similar to +2.9
to + 3.7%0 average values for soils, tree
leaves, and lacustrine sulfate in the Adirondack Mountains of northeastern New York
state (unpubl. data), and seems representative of atmospheric sulfate compositions.
The relatively
34S-enriched
values of
- +23%~ found in RL and FGL surface sulfates arise from dissolution of ancient evaporites (Deevey et al. 1963) and not from
rainfall.
A variety of consumers was collected in
FGL and RL, but none had the - loo/o0634S
value to be expected if bacteria were essentially the sole important food resource in
the lake. Mixed Daphnia and Diaptomous
samples collected at 15-l 6 m in horizontal
tows above the upper surface of the bacterial
plate had values of +7.1 to 13.6% (Table
2), much closer to the approximately + 2Oa/oo
value expected of a pure algal diet than to
the - 1OVmvalue of organic sulfur in bacterioplankton.
These findings indicate that
food resources other than bacterioplankton
are important for the mixed zooplankton
samples, although individual
zooplankton
species may specialize as bacteriovores. Fish
and other invertebrate consumers all had
634Svalues between + 3.6 and + 12%~in FGL
and RL (Table 2), values that again did not
value of organic
closely approach the - 1O%CJ
sulfur in bacterioplankton.
In the third meromictic lake, GL, consumers had 634S values between 0 and +4%0, close to both terrestrial leaves (+2.8%, Table 2) and lake
sulfate (+ 7.5o/oo,Table 3); no evidence for
food web dependence on bacterioplankton,
expected to be strongly depleted in 34S relative to surface sulfate, was found.
PC-The
two bacterioplankton
fractions from 18 m in FGL had distinctly different carbon isotopic values. Chromatiaceae had values of -39.3 to -41.1%0,
whereas Chlorobiaceae were enriched in ’ 3C
by about 8.5%0 and had values of -30.9 to
- 32.6%0. This carbon isotopic difference has
been observed between cultures of Chromatium and Chlorobium (Sirevag et al.
1977) and is related to differences in carbon
fixation pathways. Other plants in RL and
FGL had 613C values of - 10 to -3OV& (Table 2) and were thus substantially enriched
in l 3C relative to the - 40%0 Chromatiaceae.
The wide - 16 to - 29?& range of algal 613C
values found in FGL (Table 2) is similar to
that found in other freshwater lakes (Oana
and Deevey 1960; Deevey and Stuiver 1964;
LaZerte and Szalados 1982); further investigations are needed to elucidate the reasons
for this broad isotopic variation among algae in a single lake. Terrestrial C3 and C,
plants were also analyzed in this study: C4
corn used as bait by fishermen had a value
of - 1Oo/oowhile tree leaves and Typha had
values of - 27 to - 30%0 typical of C3 plants
(Smith and Epstein 1971).
In spite of the carbon isotopic range of
- 10 to -4OVm among available food resources, P3C values of most consumers were
rather narrowly clustered in the -25 to
- 3Oymrange in all three lakes (Tables 2 and
3). More negative values of about -40 to
- 35%0 would be expected if the food web
in FGL were based solely on -40%0 bacterioplankton and if up to five trophic levels
were present (assuming a 1%1 increase in
Food webs in meromictic lakes
85
Fig. 2. Isotopic food web diagram, Fayetteville Green Lake. Values for most consumers (0) fall within two
mixing envelopes constructed from three kinds of foods (terrestrial leaves, phytoplankton, and macrophytes)
indicating that these are the important sources of carbon and sulfur for consumers in the lake; bacterioplankton,
G and P, appear unimportant since no consumers have negative 634Svalues. Symbols: O-individual
fish or
crayfish; S-gastropods; C-corn taken from trout gut; T-terrestrial leaves from trees in the watershed; PPsurface phytoplankton, 613Cestimated from POC and 634Sassumed +20 to +22.9%, 0 to -3o/oorelative to
+22.9% surface sulfate; P-purple fraction of bacterioplankton (Chromatiaceae); G-brown fraction of bacterioplankton (Chlorobiaceae); Z-zooplankton; M-macrophyte. Vertical crosses give SD about mean for both
tree leaves and phytoplankton.
P3C per trophic level: Fry and Sherr 1984).
The observed values of -25 to - 30%0 of
most consumers are thus not consistent with
a food web based on bacterioplankton,
but
are consistent with a food web based on a
mixture of algae and terrestrial detritus that
have 613Cvalues in the -25 to - 307~ range
(Tables 2 and 3). Zooplankton were exceptional in that their 613C values reached
- 39.3Ym (Table 2) and approached the values of Chromatiaceae, about -4Oo/oo, at the
top of the bacterial plate. The question of
zooplankton nutrition is considered further
below.
I also analyzed several trout (Salmo
gairdneri, stocked in FGL from a hatchery.
Feed rations used in the hatchery measured
- 19.7 to -20.2Ym P3C and +8.4 to +9.2Ym
634S.One large breeder trout collected in the
lake had stable isotopic compositions similar to the hatchery rations (Table 2: trout
No. 1); such trout are stocked as mature
adults and may not feed substantially after
being stocked, accounting for a continuing
isotopic similarity to hatchery foods. Several smaller trout (Table 2: individuals
28) had been stocked as juveniles more than
6 months before the collection in October
(B. Culligan pers. comm.). Since 613C values
of these trout changed as much as - 16.7o/oo
from the -20%0 hatchery foods (Table 2),
feeding and rapid tissue turnover were evidently occurring.
Combined 634Sand V3C food web analysis-The
613C and 634S data can be combined to show which food resources are im-
86
Fry
portant in food webs of meromictic lakes.
Isotopic values of most consumers from
FGL and RL fall within two mixing envelopes (Fig. 2) that are constructed from three
sources: phytoplankton,
submerged macrophytes, and terrestrial leaves. Isotopic
values of consumers most closely resemble
values for terrestrial materials, with small
admixtures of macrophyte and phytoplankton C and S. Since no consumers had negative 634S values, bacterioplankton
appear
unimportant as food resources.
Two comparisons substantiate this conclusion. First, littoral gastropods and crayfish can be expected to have much less access to a pelagic food web based on
bacterioplankton
than do fish consumers,
yet gastropods, crayfish, and fish showed the
same range of g13C and 634S values (Table
2). This isotopic similarity between littoral
invertebrates and fish indicates a minor importance of bacterioplankton in the food web
leading to fish. Second, the -23 to - 3 lo/o0
carbon isotopic values of most consumers
in the three meromictic lakes are essentially
identical to - 24 to - 3 1% isotopic values
of bivalves from other nonmeromictic
New
England lakes (Oana and Deevey 1960;
Keith et al. 1964). A significant lowering of
613C values toward the - 40% bacterioplankton value was thus not observed in the
meromictic lakes, again indicating little food
web importance for these bacteria. In summary, both P3C and 634S measurements
showed that most animals collected in this
study ultimately depend on terrestrial detritus, phytoplankton,
and submerged macrophytes for their carbon and sulfur.
Carbon isotopic values for zooplankton the animals most likely to consume bacterioplankton
directly- matched the - 4Oo/oo
value of bacterioplankton
(Table 2). This
result seems in contradiction to the organic
sulfur results that do not show a close correspondence between isotopic values of
zooplankton (7- 14o/oo)and bacterioplankton
(- 10%). Two hypotheses can be advanced
to explain this discrepancy between carbon
and sulfur results. First, zooplankton may
depend on bacterioplankton
for carbon, but
not sulfur. This seems unlikely since most
sulfur in bacteria is in the form of protein
(Cuhel et al. 1982), and assimilation of pro-
tein sulfur would normally entail assimilation of protein carbon as well. Second, an
unsampled, subsurface phytoplankton
population may exist within the lake that has a
-40% carbon isotopic composition that is
identical to that of bacterioplankton,
but the
usual +20Y~ 634Scomposition that is characteristic of algae in FGL. Zooplankton
feeding on a 2 : 1 mixture of such phytoplankton and bacterioplankton
would have
carbon and sulfur isotopic values near those
observed. This hypothesis is speculative but
may apply for the following reasons.
Zooplankton
in nonmeromictic
lakes
often have 613C values between -32 and
-47o/oo (Oana and Deevey 1960; Rau 1978,
1980), largely because algae have similar
very negative isotopic compositions. Such
13C-depleted values arise especially in stratified lakes when carbon dioxide respired by
benthic organisms is added in quantity to
bottom waters. This lowers the isotopic
composition of dissolved inorganic carbon
(DIC) from an air-equibrated value of - 0%
toward the average value of benthic COz,
- 30% (Oana and Deevey 1960). An overall
lowering of isotopic values in the food chain
DIC + algae + zooplankton occurs under
these conditions (Rau 1978). Low values of
DIC 613C occur in FGL (to - 2 lo/oo:Deevey
et al. 1963; Takahashi et al. 1968; Torgersen
et al. 198 l), indicating significant inputs of
benthic respired carbon. Algal populations
with -40%0 P3C values would be most likely to occur near the upper surface of the
bacterial plate where isotopic compositions
of epilimnetic DIC are lowest, - 11 to - 1~?J&o
(Deevey et al. 1963; Torgersen et al. 198 1).
Field observations
support zooplankton
feeding in this region: Culver and Brunskill
(1969) observed that Daphnia collected near
the bacterial plate had diatoms as well as
reddish sulfur bacteria in their guts.
In summary, analyses of organic sulfur
show little evidence for a food web in Fayetteville Green Lake of the type bacterioplankton ---) zooplankton
-+ intermediate
consumers -+ large fish. The carbon isotopic
analyses show that although zooplankton
may be feeding on bacterioplankton,
common fish consumers do not derive their tissue carbon from a food web based on bacterioplankton
and zooplankton.
These
Food webs in merorqictic lakes
results were unexpected since >80% of the
annual primary productivity
in FGL is due
to bacterioplankton,
zooplankton are concentrated at the upper surface of the bacterial plate, and several studies show that
zooplankton can feed on and efficiently assimilate bacteria (Harman 1967; Culver and
Brunskill 1969; Sorokin 1969; Gophen et
al. 1974; Matsuyama and Shirouzu 1978).
It is possible that more extensive analyses
of other fish species and individual
zooplankton species would reveal some animals that rely heavily on bacterioplankton
for their nutrition.
However, most fish
species common in FGL were analyzed in
this study (B. Culligan pers. comm.).
top carnivores such as smallmouth
bass
(Micropterus dolomieui) should be good indicators of which primary producers are
generally important as sources of carbon and
sulfur nutrition
for consumers in meromictic lakes. Analyses of such fish did not
support the concept of a bacterioplanktonbased food web in FGL; rather, food webs
of the three meromictic lakes are based more
strongly on algae, macrophytes, and terrestrial detritus than on bacterioplankton,
and
so closely resemble food webs in nonmeromictic lakes. The large standing stocks of
bacterioplankton
and zooplankton present
at the top of the monimolimnion
and base
of the epilimnion in FGL may exist as relatively isolated units. Low oxygen and the
presence of sulfide in this region may deter
further predation, allowing development of
the observed high planktonic densities.
microorganisms: Considerations for the application of sulfate incorporation into protein as a measurement of natural population protein synthesis.
Appl. Environ. Microbial. 43: 160-168.
CULVER, D. A., AND G. J. BRUNSKILL. 1969. Fayetteville Green Lake, New York. 5. Studies of primary production and zooplankton in a meromictic
marl lake. Limnol. Oceanogr. 14: 862-873.
DEEVEY, E. S., N. NAKAI, AND M. STUIVER. 1963.
Fractionation of sulfur and carbon isotopes in a
meromictic lake. Science 139: 407-408.
-,
AND M. STUIVER. 1964. Distribution of natural isotopes of carbon in Linsley Pond and other
New England lakes. Limnol. Oceanogr. 9: l-l 1.
EFFLER, S. W., D. A. WILCOX, AND S. D. FIELD. 198 1.
Meromixis and stability at Green Lake, Jamesville, NY, Sept. 1977-Nov. 1978. J. Freshwater
Ecol. 1: 129-139.
EGGLETON, F. E. 1956. Limnology of a meromictic,
interglacial, plunge-basin lake. Trans. Am. Microsc. Sot. 75: 334-378.
FIELD, T. C. 1974. Vertical migration of Diaptomus
sicilis Forbes in Fayetteville Green Lake. Ph.D.
thesis, SUNY-Syracuse. 102 p.
FREY, D. G. 1967. Biological characteristics of meromictic lakes, p. 63-95. In D. F. Jackson ted.],
Some aspects of meromixis. Dep. Civ. Eng., Syracuse Univ.
FRY, B., R. S. SCALAN, J. K. WINTERS, AND P. L. PARKER. 1982. Sulphur uptake by saltgrasses, mangroves and seagrasses in anaerobic sediments.
Geochim. Cosmochim. Acta 46: 112 l-l 124.
-,
AND E. SHERR. 1984. 613Cmeasurements as
indicators of carbon flow in marine and freshwater
ecosystems. Contrib. Mar. Sci. 27: 15-47.
GOPHEN, M., B. Z. CAVARI, AND T. BERMAN. 1974.
Zooplankton feeding on differentially labelled algae and bacteria. Nature 247: 393-394.
HARMAN, W. N. 1967. A late winter survey of macroscopic invertebrates in Green Lake, Fayetteville,
New York, p. 188-214. In D. F. Jackson [ed.],
Some aspects of meromixis. Dep. Civ. Eng., Syracuse Univ.
JELACIC, A. J. 197 1. Physical limnology of Green and
Round Lakes, Fayetteville, New York. Ph.D. thesis, Univ. Rochester. 295 p.
KAPLAN, I. R., K. 0. EMERY, AND S. C. RITTENBERG.
1963. The distribution and isotopic abundance
of sulphur in recent marine sediments off southern
California. Geochim. Cosmochim. Acta 27: 297331.
AND S. C. RITTENBERG. 1964. Microbiological fractionation of sulphur isotopes. J. Gen. Microbiol. 34: 195-2 12.
KEITH, M. L., G. M. ANDERSON, AND R. EICHLER. 1964.
Carbon and oxygen isotopic composition of mollusk shells from marine and freshwater environments. Geochim. Cosmochim. Acta 28: 17571786.
LAZERTE, B. D., AND J. E. SZALADOS. 1982. Stable
carbon isotope ratios of submerged freshwater
macrophytes. Limnol. Oceanogr. 27: 4 13-4 18.
MATSUYAMA, M., AND E. SIIIROUZU. 1978. Importance of photosynthetic sulfur bacteria, Chroma-
Also,
References
BRUNSKILL, G. J., AND S. D. LIJDLAM.
1969. Fayetteville Green Lake, New York. 1. Physical and
chemical limnology. Limnol. Oceanogr. 14: 8 17829.
CALDWELL, D. E., AND J. M TIEDJE. 1975. The structure of anaerobic bacterial communities in the hypolimnia of several Michigan lakes. Can. J. Microbiol. 21: 377-385.
CARLSON, P, R., JR., AND J. FORREST. 1982. Uptake
of dissolved sulfide by Spartina alterniflora: Evidence from natural sulfur isotope abundance ratios. Science 216: 633-635.
CHAMBERS, L. A., AND P. A. TRUDINGER. 1979. Microbiological fractionation of stable sulfur isotopes: A review and critique. Geomicrobiol. J. 1:
249-293.
CUHEL, R. L., C. D. TAYLOR, AND H. W. JANNASCH.
1982. Assimilatory sulfur metabolism in marine
87
88
Fry
tium sp. as an organic matter producer in Lake
Kaiike. Jap. J. Limnol. 39: 103-l 11.
MEKHTIYEVA, V. L., AND R. G. PANIUNA. 1968. Isotopic composition of sulfur in aquatic plants and
dissolved sulfates. Geochem. Int. 5: 624-627.
OANA, S., AND E. S. DEEVEY. 1960. Carbon 13 in lake
waters, and its possible bearing on paleolimnology. Am. J. Sci. 258: 253-272.
RAU, G. H. 1978. Carbon- 13 depletion in a subalpine
lake: Carbon flow implications. Science 201: 901902.
-.
1980. Carbon-13/carbon- 12 variation in
subalpine lake aquatic insects: Food source implications. Can. J. Fish. Aquat. Sci. 37: 742-746.
SIREVAG, R., B. B. BUCHANAN, J. A. BERRY, AND J. H.
TROUGHTON. 1977. Mechanisms of CO2 fixation
in bacterial photosynthesis studied by carbon isotope fractionation technique. Arch. Microbial. 112:
35-38.
SMITH, B. N., AND S. EPSTEIN. 197 1. Two categories
of 13C/12Cratios for higher plants. Plant Physiol.
47:380-384.
SOROKIN, Y. I. 1969. On the trophic role of chemosynthesis and bacterial biosynthesis in water bodies, p. 189-205. Zn C. R. Goldman [ed.], Primary
productivity in aquatic environments. Univ. Calif.
STEINMETZ, M. A., AND U. FISCHER. 1981. Cytochromes ofthe green sulfur bacterium Chlorobium
limicolu. Arch. Microbial. 130: 3 l-37.
TAKAHASHI, T., W. BROECKER, Y. H. LI, AND D. THURBER. 1968. Chemical and isotopic balances for a
meromictic lake. Limnol. Oceanogr. 13: 272-292.
TOKGERSEN, T., D. E. HAMMOND, W. B. CLARKE, AND
T. H. PENG. 198 1. Fayetteville, Green Lake, New
York: 3H-3He water mass ages and secondary
chemical structure. Limnol. Oceanogr. 26: 1IO122.
TR~PER, H. G., AND S. GENOVESE. 1968. Characterization of photosynthetic sulfur bacteria causing
red water in Lake Faro (Messina, Sicily). Limnol.
Oceanogr. 13: 225-232.
TURANO, V. S., AND M. C. RAND. 1967. Some chemical observations on Fayetteville Green Lake, New
York, p. 151-187. Zn D. F. Jackson [ed.], Some
aspects of meromixis. Dep. Civ. Eng., Syracuse
Univ.
Submitted: 21 January I985
Accepted: 1 August 198.5