Declines in the abundance and condition of Great Lakes Alewives have been reported periodically during the last two decades, and the reasons for these declines remain unclear. To better understand how food web changes may influence Alewife growth and Wisconsin growth model predictions, we fed Alewives isocaloric diets high in omega-6 fatty acids (corn oil) or high in omega-3 fatty acids (fish oil). Alewives were fed the experimental diets at either 1% (“low ration”) or 3% (“high ration”) of their wet body weight per day. After six weeks, Alewives maintained on the high ration diets were significantly larger than those fed the low ration diets. Moreover, Alewives given the high ration fish oil diet were significantly larger than those maintained on the high ration corn oil diet after six weeks of growth. Body lipid, energy density and total body energy of Alewives on the high ration diets were significantly higher than those fed the low ration diets, and total body energy was significantly higher in Alewives given the high ration fish oil diet compared to those on the high ration corn oil diet. The current Wisconsin bioenergetics model underestimated growth and overestimated food consumption by Alewives in our study. Alewife thiaminase activity was similar among treatment groups. Overall, our results suggest that future food web changes in Lake Ontario, particularly if they involve decreases in the abundance of lipid rich prey items such as Mysis, may reduce Alewife growth rates and total body energy due to reductions in the availability of dietary omega-3 fatty acids.
The amount and quality of lipids in the diet has a particularly strong influence on growth and condition of fishes. Lipids are stored by fishes when food is plentiful and are mobilized when food is scarce (Halver and Hardy, 2002). Stored lipids provide energy for normal growth and metabolism, for production of gametes, and for surviving extended periods of food deprivation (Henderson and Tocher, 1987). Moreover, the specific types of lipids and fatty acids available to fishes play a key role in fish condition, growth and health. Highly unsaturated fatty acids (HUFA) such as eircosapentaenoic acid and docosahexaenoic acid: EPA (C20:5n-3) and DHA (C22:6n-3) are essential fatty acids for most fishes, meaning that they cannot be synthesized endogenously and must instead be obtained through diet (Bell et al., 1986). Food sources available to fishes vary in their fatty acid composition, and marine food webs typically contain much higher levels of omega-3 HUFA than freshwater food webs, where a lack of essential fatty acids can present significant challenges for fishes (Sargent et al., 2002). HUFA are known to be important in maintaining cell membrane function, improving cold tolerance, mitigating stress responses and contributing to healthy cardiovascular and immune system function in fish and higher vertebrates (Snyder and Hennessey, 2003; Tocher, 2003). Given the vital role of these compounds, it is not surprising that dietary deficiencies of omega-3 HUFA in fishes can result in decreased growth and survival, increased incidences of disease, and a variety of other metabolic disorders (Sargent et al., 2002; Tocher, 2003).
Alewives (Alosa pseudoharengus) are presently an important component of the Lake Ontario and Lake Michigan food webs. They are the primary prey of salmon and trout, and support a valuable recreational fishery (Madenjian et al., 2005). Alewives also have significant impacts on zooplankton communities through predation, on native fishes through competition for food and via predation on larvae, and on reproductive success of piscivores due to the detrimental effects of the thiaminase that they contain (Madenjian et al., 2002; Honeyfield et al., 2005). Declines in the abundance and condition of Lake Ontario (Murry et al., 2010) and Lake Michigan Alewives (Madenjian et al., 2003, 2005) have been reported periodically during the last two decades, and the reasons for these declines remain unclear. Given their broad ecological impacts, understanding key factors that influence Alewife growth remains a priority in Great Lakes fisheries management.
The diet of Alewives has been altered by continuing changes in the Lake Ontario food web. Feeding habits of Alewives shifted from a diet consisting largely of native copepods and cladocerans containing 40–70% lipid (Goulden et al., 1998) to one containing significant numbers of Bythotrephes containing lower lipid concentrations (10–19%; Bilkovic and Lehman, 1997) when this non-native cladoceran invaded Lake Ontario in the mid-1980's (Mills et al., 1992). Although the occurrence of Bythotrephes in Lake Ontario is now sporadic, another related non-native predatory cladoceran, Cercopagis pengoi, is well-established and is an important component of Alewife diets, especially during summer months (Storch et al., 2007; Walsh et al., 2008; Stewart et al., 2010a). The invasion of Lake Ontario by dreissenid mussels in the 1990's altered nearshore nutrient and food web dynamics (Hecky et al., 2004), and the effects of dreissenids coupled with predation from Bythotrephes and Cercopagis may be responsible for a significant decline in nearshore zooplankton abundance in the lake (Stewart et al., 2009). This reduced availability of food in nearshore areas may not adequately support Alewife growth, and Alewives may now need to rely even more heavily on Mysis in offshore areas to obtain sufficient energy and essential nutrients (Stewart et al., 2009) especially with the continuing disappearance of the amphipod Diporeia (Nalepa et al., 2006). Mysis and Diporeia contain high concentrations of lipids (average 25%–40%; Cavaletto and Gardner, 1998) and essential omega-3 fatty acids. With the loss of Diporeia, Mysis has become an increasingly important component of the diet of Alewives over the last two decades in Lake Ontario (Walsh et al., 2008; Stewart et al., 2009). Although Mysis are currently abundant in Lake Ontario, any future declines could limit the ability of Alewives to obtain sufficient energy and essential fatty acids through their diet. Over the long term, reduction in the availability of Mysis could result in reduced abundance of Alewives, or instead, Alewife abundance might remain high but growth rates and condition levels could be greatly reduced.
Bioenergetics models simulate growth or estimate food consumption of fishes, and these models have many applications in contemporary fisheries management. The Wisconsin bioenergetics model for Alewives (Stewart and Binkowski, 1986) has been used extensively in Great Lakes fisheries management for a variety of purposes, including optimization of salmonine stocking rates, modeling transfer of contaminants, and studying the impacts of Alewives on their zooplankton prey. Despite the usefulness of bioenergetics models, a recent review indicates that many bioenergetics models are inaccurate when predicting fish growth, and the authors conclude that more studies on the effects of low food rations and prey type are needed to improve the predictions of current models (Bajer et al., 2004). Hence, a better understanding of Alewife nutrition, coupled with knowledge of Alewife diets in the Great Lakes, should improve the accuracy of current growth models for this species.
In summary, continued invasions by non-native species and ongoing alterations of food webs in Lake Ontario and throughout the Great Lakes point toward dramatic changes in sources and availability of essential nutrients such as fatty acids. Herein we report our findings on a laboratory study examining the effect of dietary omega-3 and omega-6 fatty acids on Alewife growth. The results will improve our ability to understand the impacts of continued food web alterations on their growth, which is especially important as long as Alewives are an important prey fish of top predators in the Lake Ontario food web and throughout the Great Lakes basin.
Materials and methods
We obtained adult (age 1–2 years) Alewives from Waneta Lake, an inland freshwater lake in central New York, USA. Two hundred eighty Alewife were lightly sedated with tricaine methanesulfonate (MS-222) and individual weights recorded. Fish were distributed among eight 760 litre circular tanks at densities of 35 fish tank−1, and the temperature of the water was maintained at 16.0 ± 0.5°C with a 12L:12D photoperiod. We formulated two isocaloric, isonitrogenous diets with either fish oil or corn oil as the added source of lipids. One diet had marine fish oil as the primary source of fatty acids; this diet was rich in omega-3 fatty acids and had elevated levels of highly unsaturated fatty acids (HUFA) including EPA (C20:5n-3) and DHA (C22:6n-3). The second diet contained corn oil as the primary source of fatty acids; this diet was rich in omega-6 fatty acids and provided high amounts of linoleic acid (C18:2n-6) and low amounts of EPA and DHA (Table 1). In terms of natural prey of Alewives, the fish oil diet would resemble Mysis or Diporeia (high in HUFA), while the corn oil diet would be more similar to cladocerans or other prey that typically have low HUFA levels. Each diet was provided to the fish at either a low or a high amount (1% or 3% of wet body weight per day, respectively) assigned to replicate tanks. Feeding rates were adjusted after three weeks based on the number and weight of fish in each tank. Midway through the experiment (week 3) and at the conclusion of the study (week 6), all 35 fish from each tank were sedated with MS-222 (50 mg l−1) and individually weighed. No mortality occurred during the experiment (either as a result of sedation or from other causes), and fish exhibited normal feeding behavior the day after receiving anesthesia. At six weeks, fish were euthanized with an overdose of MS-222 (275 mg l−1) and individual fish were weighed, wrapped in plastic wrap followed by aluminum foil, and frozen at −20 C for later analysis.
Feed and whole body lipids were extracted for fatty acid analysis following standard procedures (Folch et al., 1957) with chloroform/methanol (2:1 v/v) and 0.01% BHT as an antioxidant. Fatty acid methyl esters (FAMES) were prepared using mild alkaline methanolysis and were identified via gas chromatography following the methods of Wei (2007). We determined energy content for six composite samples (two fish per sample) from each diet/ration treatment. Each composite sample was dried to a constant weight at 60 C, ground with a mortar and pestle, and energy content of two 1 g subsamples was determined with a Parr Digital Bomb Calorimeter (Model 1356). We expressed energy content in two ways: as energy density (J g−1 wet weight) and as total body energy (energy density × wet weight). Thiaminase activity was measured using the 4-nitrophenol (4NTP) assay of Hanes et al. (2007) as modified by Honeyfield et al. (2010). Whole Alewives collected at the end of the study (6-weeks) were ground frozen along with dry ice to produce powdered homogenate to be used in the thiaminase analysis. Alewife thiaminase activity is reported as the mean of three fish per tank.
We used the Wisconsin fish bioenergetics modeling software version 3.0 (Hanson et al., 1997) to predict weight of Alewives after 3 weeks and after 6 weeks of growth; inputs for the model included water temperature, starting weight, energy density of the feeds, and food consumption. The model used to predict total food consumption was based on holding temperature, starting weight, and ending weight after each 3-week interval of growth. For the simulations, only data from the low ration treatments was used, since the Alewives at this ration level routinely consumed all food provided. Data from the high ration groups were not included in the modeling runs since uneaten food sometimes remained in the holding tanks and hence direct consumption could not always be accurately estimated.
We used ANOVAs with post-hoc Tukey tests to compare mean body weight among treatments midway through the experiment (week 3); ANOVAs with Tukey tests were also used at the conclusion of the experiment (week 6) to compare final body weight,% body lipid, energy density, and total body energy. Individual fatty acids and fatty acid indices in dietary and body lipids were compared using t-tests. Data were examined for normality and homogeneity of variances and were transformed as necessary, and significance was assumed at the p < 0.05 level. Mean values are presented as mean ± SE.
Fatty acid composition of whole body lipids from Alewives fed the fish oil diet and the corn oil diet reflected dietary composition. With respect to fatty acid indices, the fish oil diet had elevated levels of omega-3 fatty acids, saturated fatty acids (SAFA), and omega-3 highly unsaturated fatty acids (HUFA), and lower levels of omega-6 fatty acids and polyunsaturated fatty acids (PUFA) compared to the corn oil diet. These same trends were found in whole body lipids from the fish themselves, with Alewives fed the fish oil diet showing significantly higher levels of omega-3 fatty acids, SAFA, and n3 HUFA, and significantly lower levels of omega-6 fatty acids and PUFA, compared to Alewives maintained on the corn oil diet (Table 1). With respect to individual fatty acids, the fish oil diet contained high levels of a number of fatty acids including C16:0, EPA (C20:5n3), and DHA (C22:6n3) compared to the corn oil diet, and these fatty acids were significantly higher in Alewives maintained on the fish oil diet compared to those given the corn oil diet (Table 1).
At the initiation of the study, mean body weight did not differ significantly among treatment groups and ranged from 11.42 – 11.52 g (Figure 1). After 3 weeks of feeding, fish weight was significantly lower in the low ration (1%) treatments compared to the high ration (3%) treatments, but there were no differences between diet treatments (fish oil and corn oil) at either ration level (Figure 1). After 6 weeks of growth, the differences in weight between the low ration and high ration treatments were more pronounced; moreover, for the high ration treatments, body weight of the Alewives maintained on fish oil were significantly higher than those of the Alewives given the corn oil diet (Figure 1). Mean Alewife thiaminase activity was 17.55 ± 1.93 μmol g−1 min−1 and did not differ by oil type or ration level.
At the conclusion of the study, lipid content (% body lipid) was significantly higher in the high ration treatments (12.6 ± 1.03%) compared to the low ration treatments (7.1 ± 0.64%), but there were no significant differences between Alewives given the corn oil and fish oil diets. Energy density (J g−1) was also significantly higher in the high ration treatments (8759 ± 206 J g−1) compared to the low ration treatments (6794 ± 216 J g−1), with no significant differences between corn oil and fish oil treatments. Total body energy was significantly higher in the high ration treatments (162,004 ± 2845 J) compared to the low ration treatments (93,725 ± 2983 J). In addition, for the high ration treatments, total body energy was significantly higher in Alewives given the fish oil diet (167,190 ± 3041 J) compared to Alewives maintained on the corn oil diet (156,818 ± 2649 J).
The Wisconsin fish bioenergetics model underestimated the increase in weight after 3 weeks by 9.2% for Alewives on the corn oil diet and 10.0% for Alewives on the fish oil diet; after 6 weeks, model predictions underestimated growth by 14.7% for the corn oil diet and 17.4% for the fish oil diet. The model also overestimated the consumption required to produce the changes in weight we observed over the 6 weeks of the study. A typical low-ration Alewife maintained on the fish oil diet was given 4.95 g of feed over the course of the study, while the model predicted that 6.34 g would have been consumed; the model therefore overestimated consumption by 28.1%. Similarly, a typical low-ration Alewife maintained on the corn oil diet was given 4.97 g of feed while the model predicted consumption of 6.15 g, overestimating actual consumption by 23.7%.
Growth of fishes is strongly influenced by the quantity of food they consume, i.e. their daily caloric intake (Halver and Hardy, 2002); hence the large difference in growth we observed between the low and high ration groups is to be expected. However, despite consuming an equal number of calories, Alewives in the high ration groups achieved significantly larger sizes when fed the fish oil diet compared to the corn oil diet. The difference in weight between these groups at the end of the study (mean body weight of 19.2 g in the fish oil group compared to 17.8 g in the corn oil group; Figure 1) represents a difference of approximately 8%. Projected over a normal growing season of 3–5 months for Alewives in the Great Lakes region, this difference in growth could result in a 20% difference in body size due simply to the effects of nutritional quality, not quantity. The fish oil diet contained high levels of essential fatty acids, especially highly unsaturated omega-3 fatty acids such as EPA (C20:5n3) and DHA (C22:6n3); in contrast, the corn oil diet contained low levels of these fatty acids, and much higher levels of omega-6 fatty acids, especially lenoleic acid (C18:2n6). Increased dietary levels of essential omega-3 fatty acids such as EPA and DHA have been found to increase growth in a number of fish species (Sargent et al., 2002), and these fatty acids have other positive health effects in fishes including mitigating stress responses and contributing to healthy cardiovascular and immune system function (Arts and Kohler, 2009). Although the mechanism by which EPA and DHA enhance growth in fishes is still uncertain, these fatty acids may reduce resting metabolic rates and allow more energy to be used for somatic growth (McKenzie, 2001).
Under natural conditions, energy density of Alewives can vary greatly. For Alewives in Lake Ontario, Rand et al. (1994) reported energy densities ranging from approximately 4,200–7,000 J g−1 over a two-year period. In our study, energy densities ranged from 6558–8810 J g−1 after 6 weeks of growth in the laboratory. These values suggest that even in our low ration treatments, the Alewives were relatively well-fed, and this is consistent with the fact that they were able to achieve positive growth over the course of the study. When energy content is expressed as total body energy (energy density x body weight), Alewives consuming the fish oil diet at the high ration level had energy values that were approximately 7% higher than Alewives consuming the corn oil diet at the same ration level. Thus, reduced growth rates and lower total body energy of Alewives can clearly be brought about by a change in the nutritional quality of the food they consume. Under such conditions, piscivores feeding on Alewives would have to increase foraging efficiency and rates of capture of individual Alewives to avoid a reduction in caloric intake and possible limitations on their own growth rates. Chinook salmon may have employed this strategy in response to declines in average size and energy density of Alewives in Lake Ontario in the 1980's, when it is estimated that Chinook salmon increased their average daily consumption of Alewives from 1.5 fish d−1 to 3.7 fish d−1 (Rand et al., 1994).
Using information provided in Rand et al. (1994) and Stewart et al. (2010b), we can estimate that Lake Ontario Alewives of similar size to those used in our study typically have short-term specific growth rates in the range of 0.30–0.40%. The increases in body size that we observed over the 6-week duration of our laboratory study correspond to specific growth rates in the range of 0.40–0.50% for the low ration treatments and 1.0–1.25% for the high ration treatments. This comparison of our laboratory growth rates to those estimated from field data suggest that Lake Ontario Alewives have a much higher growth potential than is typically observed in wild fish, and significantly higher growth rates may be observed in the future if daily consumption rates (on a per-fish basis) increased either through greater availability of food or decreases in Alewife population size.
When applied to data from our study, the current Wisconsin Alewife bioenergetics model underestimated growth by 9.2–17.4% and overestimated consumption by 23.7–28.1%. These results may help to explain previous modeling results suggesting that predator demand in Lake Ontario exceeded Alewife production (Jones et al., 1993), although assumptions regarding Chinook salmon consumption, temperature preferences, and degree of natural reproduction may also have played a role in biasing these estimates (Murry et al., 2010). Overall, the results of our study clearly demonstrate the importance of nutritional quality on growth of Alewives, and it is likely that incorporating both nutritional quality and caloric intake in Alewife bioenergetics models would improve growth estimates. For example, nutritional parameters could be added to existing Alewife bioenergetics models to reflect higher growth rates in Alewives consuming prey rich in essential fatty acids. Determining specific values for such nutritional parameters would require more detailed study of the fatty acid composition of Alewife prey, and how specific fatty acids influence Alewife growth at various ration levels.
Thiaminase activity was not affected by diet in this study, which compared two sources of lipid fed at a low or high feeding rate. As previously observed, the thiaminase activity of Alewife held in captivity (Lepak et al., 2008) is about two-fold higher than the thiaminase activity in wild-caught Alewife (Fitzsimons et al., 2005; Tillitt et al., 2005). The reason for elevated thiaminase in captive-held Alewife is not clear.
Disruption of the lower food web in Lake Ontario due to invasion by non-native invertebrates has not yet resulted in sustained negative effects on growth and condition of Alewives (O’Gorman et al., 2008). However, Mysis has been available and has become an increasingly important component of Alewife diets in Lake Ontario since the 1990's, providing an important source of dietary lipids and essential fatty acids. In contrast, in Lake Michigan in the mid-1990s and in Lake Huron several years later, the virtual disappearance of Diporeia (another rich source of dietary lipids and essential fatty acids) resulted in dramatic declines in abundance and condition of Alewives (Madenjian et al., 2003; Nalepa et al., 2006; Pothoven and Madenjian, 2008). The results of our study strongly suggest that a diet rich in essential fatty acids, in particular omega-3 highly unsaturated fatty acids such as EPA and DHA, will contribute to high growth rates in Alewives. Hence, although Mysis are currently abundant in Lake Ontario, any future declines in this important food item could limit the ability of Alewives to obtain sufficient energy and essential fatty acids from their diet, perhaps leading to declines in abundance, growth, and condition similar to those that have already occurred in Lakes Michigan and Huron. As the food web in Lake Ontario continues to change, it is clear that detailed knowledge of the nutritional status of both Alewife and their prey will be vital to understanding the long-term impact of these changes.
This research was supported by a grant from the SUNY Research Foundation to RJS. We thank W. Schregel (SUNY College at Buffalo) and W. Ridge (USGS) for their assistance with laboratory work.