A structural and functional assessment of the microbial foodweb of Lakes Superior, Huron, Erie and Ontario was undertaken during late summer (August) between 2001 and 2004. One lake was sampled in each year. Our analysis included microscopic enumerations of bacteria, autotrophic picoplankton, phytoplankton, heterotrophic nanoflagellates and ciliates coupled with radioisotope tracer measurements of primary productivity (14C) and bacterial growth (3H). Phytoplankton biomass was highest on average in Lake Erie (≈1.9 g m−3) and lowest in Lake Ontario (≈0.2 g m−3), whereas microbial loop biomass was highest in Lake Ontario (≈2.4 g m−3) and lowest in Lake Huron (≈0.1 g m−3). The organic carbon pool was found to be predominantly autotrophic (>80%) in Lakes Superior (≈280 mg C m−3), Huron (≈195 mg C m−3) and Erie (≈660 mg C m−3) and smaller picoplankton had the highest carbon turnover rates (≈0.4–1.5 d−1). However, in Lake Ontario (≈335 mg C m−3) the carbon pool was about 75% heterotrophic and larger net plankton had the highest carbon turnover rates (≈6.8 d−1). Despite differences in trophic state, the microbial foodwebs of Lakes Superior, Huron and Erie appeared to function in a similar and efficient manner. In contrast, the microbial foodweb of Lake Ontario appeared to be unhealthy with autochthonous production being sequestered by heterotrophic nanoflagellates. More detailed work is needed to understand the foodweb linkages both within the Great Lakes and among them.
The North American Great Lakes are one of the world's largest repositories of freshwater. Covering approximately 250,000 km2, these lakes–Superior, Michigan, Huron, Erie and Ontario–contain roughly 20% of the global supply of surface freshwater and their watersheds are home to 37 million people (Environment Canada and United States Environmental Protection Agency, 1995). The vast majority of the population resides in the southern portion of the watershed which includes Lakes Erie and Ontario as well as the southern reaches of Lakes Huron and Michigan. Not coincidentally, these areas of the Great Lakes have been the most heavily impacted by multiple anthropogenic stressors including eutrophication and phosphorus abatement, contaminants, the expansion of invasive species and climate change (Munawar et al., 2005). While all of these stressors have had immediate impacts on the lowest trophic levels, their long term impacts have been shown to reverberate through the entire foodweb (Mills et al., 2003; Fahnenstiel et al., 2010). In contrast, the northern portion of the watershed which includes Lake Superior and most of Lakes Huron and Michigan is sparsely populated and the effects of anthropogenic stressors are limited to nearshore zones (Grigorovich et al., 2003; Niemi et al., 2011), but climate change in the upper lakes may be a very significant stressor (Austin and Colman, 2007). The question remains as to how these stressors affect autochthonous carbon dynamics in the Great Lakes.
The transfer of autochthonous energy from lower to higher trophic levels is the most fundamental component of any ecosystem. The role of the microbial loop in facilitating the transfer of organic carbon as well as recycling essential nutrients is an increasingly important topic in aquatic research (Pomeroy, 1974; Sherr et al., 1986, 1988; Munawar et al., 1999). Moreover, the short generation times of micro-organisms makes them useful bioindicators for detecting ecosystem change (Munawar et al., 1987, 1999). Microbial foodweb studies have highlighted some of the knowledge gaps regarding the underlying biogeochemical processes in aquatic ecosystems. For example, Bidanda et al. (2001) found that 90% of planktonic respiration in oligotrophic lakes is heterotrophic, primarily bacterial, in contrast to the more traditional view that planktonic respiration in oligotrophic lakes is net autotrophic and that eutrophic lakes are net heterotrophic (del Giorgio and Peters, 1994; Prairie et al., 2002). Several studies have also highlighted the relative importance of heterotrophic organisms to community respiration (Dodds and Cole, 2007) and in regulating the transfer of authochthonous carbon through the foodweb (Tadonléké et al., 2004; Brett et al., 2009; Munawar et al., 2010). Understanding the health of aquatic ecosystems is increasingly about understanding the dynamics of the lowest trophic levels.
There is no coherent picture emerging as to how the observed multiple anthropogenic stressors will affect the health of the Great Lakes, but there is ample evidence of stress at the lowest trophic levels. Rapid oligotrophication of the deeper offshore waters of Lakes Michigan, Huron and Ontario has been reported by many researchers and attributed to cumulative impacts of phosphorus abatement and dreissenid mussel grazing (Mills et al., 2003; Dove, 2009; Evans et al., 2011; Barbiero et al., 2012) suggesting a net loss of autochthonous production. However, other research has suggested that autochthonous carbon production is stable over longer periods of time and carbon may be sequestered by smaller micro-organisms rather than passed on to higher trophic levels (Fitzpatrick et al., 2007; Munawar et al., 2010). Evidence for climate change is also becoming apparent with several studies reporting increased water temperatures throughout the Great Lakes basin (McCormick and Fahnenstiel, 1999; Jones et al., 2006; Austin and Colman, 2007). Because phytoplankton are extremely sensitive to changes in temperature, changes in the thermal regime would also affect the production of autochthonous carbon. Since autochthonous carbon is the energy which sustains foodwebs, it is critical to understand how this energy flows between trophic levels in order to assess ecosystem health.
The first biological surveys of the lower trophic levels of the Great Lakes were organized by the Fisheries Research Board of Canada (later Fisheries and Oceans Canada) in the 1970s and emphasized phytoplankton and zooplankton dynamics in relation to nutrient concentrations and loading (Vollenweider et al., 1974; Watson and Carpenter, 1974; Munawar and Munawar, 1982). In the late 1980s, initial surveys of the microbial loop (bacteria, picoplankton, heterotrophic nanoflagellates, ciliates) of the Great Lakes were conducted (Munawar and Weisse, 1989) followed by more extensive work in Lakes Erie and Ontario in the 1990s (Munawar et al., 1999; Munawar and Munawar, 2001). These first microbial foodweb studies reported that Lakes Superior and Huron had lower levels of bacteria, APP and HNF compared to Lakes Michigan, Erie and Ontario. However, comparative studies of the Great Lakes microbial foodweb have been relatively sparse. Fahnenstiel et al. (1998) surveyed all of the Great Lakes during the spring and summer of 1995 and found broad similarities in the foodweb structure of all of the lakes except for Lake Erie which was attributed to higher phosphorus concentrations. The logistic and financial constraints inherent in sampling the Great Lakes mean that such comparisons are extremely valuable.
Our lab has previously developed a model for ecosystem health assessments which addresses the concepts of “net autotrophy” and “net heterotrophy” from structural measurements of bacteria, autotrophic picoplankton, phytoplankton and ciliates combined with functional measurements of size fractionated primary productivity and bacterial productivity (Munawar et al., 2009, 2010). This approach is quite different from traditional community respiration studies and builds on the contributions of various autotrophic and heterotrophic organisms to the organic carbon pool. The purpose of our model is to map out the underlying structure of the foodweb in order to make assumptions about the direction of energy flow. From this model, we consider a healthy ecosystem as one where the organic carbon pool is predominantly autotrophic coupled with high picoplankton carbon turnover rates. The current article will use this model to make a comparative study of the microbial-planktonic foodweb of Lakes Superior, Huron, Erie and Ontario conducted during the summers of 2001–2004 in order to make a preliminary yet robust assessment of the health of the Great Lakes.
Materials and Methods
The physical characteristics of each of the Great Lakes including volume, area depth and retention time are described in Table 1. Synoptic surveys were conducted during the summer by Fisheries & Oceans Canada and its research vessel CCGS Limnos over 4 consecutive years. Lake Superior was sampled during 2001, Lake Huron during 2002, Lake Ontario during 2003 and Lake Erie during 2004. Lake Michigan was not included in these surveys as no portion of it falls under Canadian jurisdiction. Three sites were sampled in each lake (Figure 1) to assess microbial loop (bacteria, autotrophic picoplankton, heterotrophic nanoflagellates and ciliates), phytoplankton, size fractionated primary productivity and bacterial growth. With the sole exception of size fractionated primary productivity at Lake Huron station 48, all parameters were sampled at all sites. Water was collected with an integrated sampler (Schroeder, 1969) to a maximum depth of 20 m or 1 m above the knee of the thermocline when stratified. Maximum sampling depths were 20 m in Lake Superior, 7–13 m in Lake Huron, 8.5–17 m in Lake Erie and 9–10 m in Lake Ontario. Samples were typically collected during daylight hours. Phytoplankton, microbial loop and ciliate samples were preserved immediately for later analysis and size fractionated primary productivity and bacterial growth experiments were conducted immediately upon collection. During each research cruise, many physical and chemical parameters were measured by the vessel's technical staff following standard methods (NLET, 1997). Of these, temperature, total phosphorus and chlorophyll a were obtained from Environment Canada's STAR database for use in this article.
Microbial loop samples, including bacteria, autotrophic picoplankton and heterotrophic nanoflagellates, were fixed with 1.6% formaldehyde and enumerated using DAPI staining (Porter and Feig, 1980) under epi-fluorescence microscopy (Munawar and Weisse, 1989). Organic carbon was estimated as 200 fg C cell−1 for APP, 10 fg C cell−1 for Bacteria and 14 pg C cell−1 for HNF (Sprules et al., 1999). Ciliate samples were preserved with Lugol's iodine (Lynn and Munawar, 1999) and enumerated following the Quantitative Protargol Staining technique (Montagnes and Lynn, 1993). Ciliate organic carbon was calculated as 11% of the fresh weight biomass (Turley et al., 1986).
Phytoplankton samples were fixed immediately with Lugol's iodine. Identification and enumeration followed the Utermöhl (1958) inverted microscope technique and the counting procedures of Lund et al. (1958). Organic carbon was estimated for the various groups of phytoplankton using equations from the literature, namely: Diatomeae (Strathmann, 1967), Chlorophyta, Chrysophyceae, Cryptophyceae and Dinophyceae (Verity et al., 1992), and Cyanophyta (Lee and Furman, 1987). Size structure was determined as Equivalent Spherical Diameter (ESD) as per the procedures given in Munawar and Munawar (1996, 2000).
Size fractionated primary productivity was estimated for three size categories of phytoplankton (<2 μm, 2–20 μm and >20 μm) by the 14Carbon technique as per the standard protocol of Munawar and Munawar (1996). Whole water samples were spiked with Na14CO3, incubated for 4 h at surface temperature and exposed to a constant light level of 240 μE s−1 m−2. Because light and temperature levels are constant in these experiments, the results should be interpreted as potential rather than actual since there is likely to be more variation in situ. After incubation, size classes were determined by filtration of the sample through polycarbonate filters, all filters were rinsed with hydrochloric acid (0.5N) in order to remove excess 14Carbon. Radioactivity was determined by liquid scintillation counting.
Bacterial growth rates were estimated by 3H-Leucine incorporation into bacterial proteins following the protocol of Jørgensen (1992). Radioactivity was determined by liquid scintillation counting. Detailed procedures are available in Heath and Munawar (2004).
Daily carbon turnover rates were estimated as the ratio of productivity to carbon (also known as P/B ratios) for bacteria and each respective size class of phytoplankton including picoplankton (<2 μm), nanoplankton (2–20 μm) and net plankton (>20 μm). Bacteria were assumed to be active for 24 h while phytoplankton were assumed to be active for 10 h.
Throughout this article, data are generally reported as the arithmetic mean and the standard error (i.e. mean ± 1 S.E.). The structure and function of the microbial foodweb was compared using a series of ANOVA post hoc Tukey tests (P < 0.05).
Results and Discussion
The terms “net autotrophic” and “net heterotrophic” have been used to describe ecosystems based on the ratio of gross photosynthesis to gross respiration in traditional community respiration studies. The problem with such descriptors is that they largely describe the physiological processes of autotrophic organisms and don't necessarily account for the contributions of heterotrophs (Dodds and Cole, 2007; Munawar et al., 2010). In order to provide a more comprehensive assessment of the autotrophs and heterotrophs at the base of the foodweb, we undertook microscopic analyses of bacteria, autotrophic picoplankton, phytoplankton, heterotrophic nanoflagellates and ciliates combined with radioisotope tracer measurements of primary productivity and bacterial growth. The results are summarized in Table 2. The Great Lakes vary considerably in terms of their respective trophic status, carbon budgets and potential energy transfer. We use this model to compare and contrast the microbial – planktonic foodweb of the Great Lakes in order to make preliminary comparisons of the health of the lakes.
Lake Superior is a large, deep mixing ultra-oligotrophic ecosystem and the largest lake in the Great Lakes system. During August 2001, surface temperatures ranged from 17–19°C, total phosphorus concentrations averaged 3.4 μg l−1 and mean chlorophyll a concentrations were 1.3 μg l−1. As shown in Table 2, microbial loop biomass was 360.5 mg m−3 dominated by heterotrophic nanoflagellates (HNF). Likewise, phytoplankton biomass was roughly 880 mg m−3 with an average ESD of 14 μm and contained a mixture of Diatomeae, Chrysophyceae and Dinophyceae. Primary productivity was approximately 0.8 mg C m−3 h−1 with nanoplankton being the most productive size fraction and the bacterial growth rate was 0.01 mg C m−3 h−1.
The organic carbon pool based on microscopic assessment was overwhelmingly autotrophic with phytoplankton supplying almost 90% of the organic carbon; phytoflagellates, notably species of Chrysophyceae, Chlorophyta and Cryptophyceae, accounted for more than half of the organic carbon (Figure 2a). Carbon turnover rates for picoplankton (0.37 ± 0.002 d−1) were 14 times greater than that of bacteria (0.026 ± 0.003 d−1), shown in Figure 3, suggesting that APP are a major vector for energy transfer. However bacterial rates were not significantly different from that of net plankton (0.023 ± 0.002) or nanoplankton (0.029 ± 0.004). The data and analyses given here is a small but representative subset of the more detailed study found in Munawar et al. (2009). Our results support the hypothesis that the organic carbon pool of Lake Superior is primarily autochthonous (Cotner et al., 2004; Urban et al., 2005). Our results also show that Lake Superior is a healthy, pristine ecosystem with an efficiently operating foodweb as evidenced by the largely autotrophic organic carbon pool and the high picoplankton turnover rates.
Lake Huron is also a large deep mixing ultra-oligotrophic lake as well as the second largest in Great Lakes. Our study area was limited to the main body of the Lake Huron only. Surface temperatures were 20–22°C during August of 2002, total phosphorus concentrations averaged 3.7 μg l−1 and mean chlorophyll a was 1.0 μg l−1. Microbial loop biomass was 135.2 mg m3 with bacteria contributing more than half of the biomass. Phytoplankton biomass (≈650 mg m−3) included a mixture of Chrysophyceae, Diatomeae and Dinophyceae and ESD was 15.1 μm. Primary productivity was 0.8 mg C m−3 h−1, dominated by nanoplankton and bacterial growth was 0.04 mg C m−3 h−1 (Table 2).
The organic carbon pool was over 90% autotrophic with Chrysophyceaean flagellates and Chlorophyta together accounting for more than half of the organic carbon (Figure 2b). Carbon turnover rates (Figure 3) for picoplankton (≈0.67 ± 0.15 d−1) were 5 times greater than that of bacteria (0.13 ± 0.07 d−1). Similar to our findings in Lake Superior, bacterial carbon turnover rates were not significantly different from net plankton turnover rates (0.07 ± 0.02) but were significantly higher than nanoplankton rates (0.02 ± 0.006). Our findings suggest that the carbon pool of Lake Huron also is autochthonous and energy transfer is largely autotrophic with the base of the foodweb resembling that of Lake Superior (Barbiero et al., 2012).
While these findings might lead us to conclude that the foodweb is healthy, there is recent evidence of foodweb disruption having occurred in Lake Huron. A dramatic change in the structure of the zooplankton community including crashes in populations of Daphnia mendotae, Bosmina longirostris and cyclopod copepods was observed in the 2003–2006 period (Barbiero et al., 2009). Both grazing by Bythotrephes (Bunnell et al., 2011) and continuing oligotrophication (Barbiero et al., 2012) have been offered as hypotheses. Unfortunately, our study predates the crash in zooplankton, and comprehensive microbial foodweb studies were not undertaken in the intervening years so it is impossible to assess whether or not changes in the structure and function of the organic carbon pool affected the zooplankton community.
Lake Erie is considerably smaller than Lakes Superior and Huron, and is the shallowest of the Great Lakes. Lake Erie itself has three limnologically distinct basins with the west basin being meso-eutrophic, the central mesotrophic and east mesotrophic (Munawar et al., 2008). Our study included 1 station in each of the 3 basins in order to capture the range of conditions. Surface temperatures varied from 20–24°C during August 2004, total phosphorus averaged 9.0 μg l−1 and mean chlorophyll a was 1.6 μg l−1. Microbial loop biomass was approximately 645 mg m−3 with HNF and Bacteria each contributing about a third of the total. Mean phytoplankton biomass was about 1950 mg m−3 and contained a mixture of Diatomeae, Cryptophyceae and Chrysophyceae with the average cell size (ESD) being 12.5 μm. Primary productivity in Lake Erie averaged 3.7 mg C m3 h−1, again dominated by nanoplankton and bacterial growth was 0.4 mg C m3 h−1 (Table 2).
We found the organic carbon pool of Lake Erie to be 90% autotrophic, with phytoflagellates (Cryptophyceae, Chrysophyceae and Dinophyceae) contributing 60% of the planktonic carbon (Figure 2c). Autotrophic Picoplankton had the highest carbon turnover rates (1.5 ± 0.9 d−1) which were triple that of bacteria (0.4 ± 0.16 d−1), however bacterial carbon turnover was significantly higher than both net plankton (0.14 ± 0.05) and nanoplankton (0.03 ± 0.006) (Figure 3). The comparatively high rates of bacterial carbon turnover are likely attributable to the lake's elevated trophic state and the role of bacteria in assimilating allochthonous matter (Pace, 1993). Previous studies of Lake Erie have shown that bacteria are an important food resource for both Daphnia and Heterotrophic Nanoflagellates (Hwang and Heath, 1997, 1999) and that allochthonous carbon may be an important source of energy supporting the lake's valuable commercial fisheries (Fitzpatrick et al., 2008). However, autochthonous carbon is generally regarded as being more important for sustaining zooplankton populations (Brett et al., 2009) and by extension, a planktivore based fishery (Munawar et al., 2010). While the base of the Lake Erie foodweb appears to be net autotrophic, heterotrophic bacteria appear to have an important role in energy transfer to higher trophic levels.
Lake Ontario is the smallest of the Great Lakes by area, although it is larger than Lake Erie by volume due to its depth. We have previously reported on the structure and function of the microbial-planktonic foodweb of Lake Ontario (Munawar et al., 2010) and the results and conclusions presented here are based on a smaller subset of that study. During August 2003, surface temperatures were fairly consistent at 23–24°C, Total Phosphorus concentrations were on average, 9.0 μg l−1 and chlorophyll a was 1.9 μg l−1. Microbial loop biomass was highest in Lake Ontario at almost 2400 mg m−3 and approximately 90% of that biomass was HNF. Phytoplankton biomass, at 218 mg m−3 was the lowest observed in this study. The phytoplankton community contained a mixture of Cyanophyta, Chlorophyta and Diatomeae and the ESD was 7.2 μm. Primary productivity averaged 2.4 mg C m−3 h−1 dominated by nanoplankton and bacterial growth was 0.03 mg C m−3 h−1 (Table 2).
Unlike Lakes Superior, Huron and Erie, the planktonic carbon pool of Lake Ontario was 70% heterotrophic and overwhelmingly dominated by heterotrophic nanoflagellates (Figure 2d). High abundances of heterotrophic nanoflagellates, tentatively identified as “colourless chrysomonads,” were reported in the first planktonic surveys of Lake Ontario during 1970 (Munawar and Nauwerck, 1971) and in later surveys during the 1980s (Pick and Caron, 1987). With respect to carbon turnover rates (Figure 3), net plankton (6.8 ± 3.9 d−1) and picoplankton (2.1 ± 1.6 d−1) had the highest rates and were not significantly different from each other, but were at least 10 times greater than nanoplankton (0.3 ± 0.12 d−1) and more than 50 times greater than bacteria (0.035 ± 0.026 d−1). The importance of net plankton in energy transfer and the dominance of heterotrophs observed in Lake Ontario are unique in the Great Lakes and somewhat unexpected. We would have predicted a greater role for bacteria in transferring organic carbon to higher trophic levels in what appears to be a net heterotrophic system (Heath et al., 2003). Nonetheless, our findings support independent observations that Lake Ontario is net heterotrophic during the late summer (Bocaniov and Smith, 2009), although there is evidence for net autotrophy at other times of the year (ibid). The high proportion of heterotrophic nanoflagellates observed in our study suggests at first glance that the organic carbon pool may be allochthonous however such high proportions of allochthonous matter have only been observed in the nearshore waters of Lake Ontario which are subject to riverine inputs (Hiriart-Baer et al., 2008). Moreover, the physical transport of allochthonous matter from near shore to offshore in such a large lake, especially under thermal bar conditions, seems unlikely.
In our original paper, we hypothesized that HNF were sequestering autochthonous carbon and inhibiting the flow of energy from lower to higher trophic levels (Munawar et al., 2010). Briefly, we argued from observations of Lake Ontario during 2003 that: prey fishes such as Alewife were considerably stressed as evidenced by a relatively low Catch Per Unit Effort (O’Gorman et al., 2008) and malnourishment (Schlectriem et al., 2008); crustacean zooplankton biomass (particularly Daphnia) was at or near historic lows (Holeck et al., 2008), and phytoplankton biomass was also at historic lows (our data) which indicated that autochthonous energy was not being passed along a traditional grazing food chain. Moreover, the high biomass of HNF coupled with high carbon turnover rates for picoplankton (compared to bacteria), provided evidence that HNF were grazing on the smaller phytoplankton. We did not find evidence that the energy bound in HNF was being passed on to higher trophic levels. However we also acknowledged multiple uncertainties and called for further research into the dynamics of HNF.
It is important to define what constitutes a healthy lake according to our structural and functional assessment of the microbial foodweb. From our methodology, we see two key indicators that are fundamental for healthy large lake ecosystems. The first is a predominantly autotrophic carbon pool. The second is high carbon turnover rates for picoplankton. With respect to the first indicator, a high proportion of autotrophs suggest that the foodweb is sustained on autochthonous carbon and the lake is self sufficient in the sense that it is able to generate its own energy. In regards to the second indicator, high picoplankton turnover rates imply optimal growth conditions for these tiny organisms, including low phosphorus and high light penetration, in addition to significant grazing pressure. Of course, if picoplankton is being grazed by other microbes, we would expect the carbon pool to contain a higher proportion of heterotrophic carbon. From the base of the foodweb, we consider a largely autotrophic organic carbon pool coupled with a high potential for energy transfer through picoplankton to be the hallmarks of a healthy ecosystem.
The offshore waters of Lake Superior and Lake Huron appear to be healthy, pristine ecosystems (Munawar et al., 2009; Munawar and Munawar, 2001) characterized by a relatively low microbial-planktonic biomass dominated by autotrophic organisms with energy transfer mainly through autotrophic picoplankton followed by heterotrophic bacteria. That being said, it must be noted that both ecosystems are increasingly stressed with Lake Superior showing signs of rapid warming (Austin and Colman, 2007) and nitrification (Sterner et al., 2007; McDonald et al., 2010) and Lake Huron showing signs of foodweb disruption (Barbiero et al., 2009; Bunnell et al., 2011). Lake Erie, on the other hand, is considerably more stressed than the upper lakes and has been affected by a long list of anthropogenic stressors including eutrophication, phosphorus abatement and exotic species, all of which have impacted the lower trophic levels. Nonetheless, we observe the same basic pattern in Lake Erie that we do in the upper lakes, specifically an organic carbon pool dominated by autotrophs and energy transfer dominated by picoplankton. The major difference is the size of the organic carbon pool – ≈660 mg C m−3 in Lake Erie, compared to ≈280 mg C m−3 in Lake Superior and ≈195 mg C m−3 in Lake Huron (Table 2) – which is attributable to the mesotrophic – eutrophic conditions that still persist in Lake Erie.
Lake Ontario is a complete anomaly. Like the other Great Lakes, it continues to be affected by multiple anthropogenic stressors, however the amount of autotrophic carbon (≈ 80 mg C m−3) was the lowest observed in the Great Lakes (Table 2). The organic carbon pool was dominated by heterotrophic nanoflagellates (>60%), and net plankton was the main vector of energy transfer. We have offered our own hypothesis as to why we consider this lake to be unhealthy (i.e. HNF sequestering organic carbon; see Munawar et al., 2010) but we cannot explain why this lake is so different. At the top of the aquatic foodweb, Lake Ontario has also been observed to have lower than expected fish production and sustainable yields compared to the other Great Lakes (Leach et al., 1987; Leach, 2003). The independent findings of Leach et al. (1987) support our contention that significant amounts of autochthonous carbon are not being passed on to higher trophic levels and underscores the need for further research on trophic transfer efficiencies.
The Great Lakes vary considerably in terms of their physical, chemical and biological characteristics, yet they are connected and share the same source water. In this article we show how trophic state is reflected in the size of the organic carbon pool and highlight the similarities and differences with respect to the structure and function of the microbial – planktonic foodweb. Despite differences in the magnitude of the organic carbon pool, the microbial – planktonic foodweb in oligotrophic Lakes Superior and Huron and mesoeutrophic – eutrophic Lake Erie appears to be similar in both structure and function. The organic carbon pool in these three lakes is primarily autotrophic and picoplankton appears to be the main vector for energy transfer as evidenced by high carbon turnover rates (P/Bs). The foodweb of Lake Ontario, on the other hand, was primarily heterotrophic and energy transfer was primarily through net plankton. More work needs to be done to elucidate the linkages between the microbial – planktonic organisms at the base of the foodweb and the fisheries they support. This would include a better understanding of spatial and temporal variations in the flux of organic carbon and the relative importance of autochthonous and allochthonous carbon sources.
We thank the crew of the CCGS Limnos and Environment Canada's Technical Operations for sample collection and vessel support. H. Niblock and J. Lorimer assisted with the experimental work and data analysis. Constructive reviews were provided by Drs. M. Legner (University of Toronto), R. T. Heath (Kent State University), J. H. Leach (Ontario Ministry of Natural Resources) and E. L. Mills (Cornell University).