Tracing the effects of diets in freshwater fish using fatty acid profiles

The aim of the research included in this dissertation is to contribute to our knowledge of how fatty acids can be used as diet indicators in freshwater systems. More precisely this work was an effort to explore a tool that can be used to describe what salmonids have consumed during the foraging period of the year and relate that to the materials allocated to embryos. As stomach contents of salmonids are often empty due to stress of capture and since any items that are present offer data only on the most recent meal, other methods of gleaning diet information are gaining popularity. Biochemical tracers, such as fatty acids, accumulate in consumers in patterns that reflect diet compositions, thus providing a view of what has been consumed and assimilated into tissues over many months. Through the use of controlled feeding experiments, this dissertation explores the efficacy of using fatty acid profiles to both describe and quantify the composition of diets of salmonid species. Much of the work conducted exploring fatty acids in reference to a diet have used diets formulated from dry ingredients whereas the experiments described within this dissertation use natural prey to create diets. As such, data described herein offers a unique look at how fatty acids in a piscivore reflect that of their potential prey items in the wild. In Chapter 1, I illustrate how fatty acid profiles of prey are reflected in lake trout (Salvelinus namaycush) after 14 weeks of feeding. Juvenile lake trout were fed one of three prey species and individuals were sampled at 4, 8 and 14 weeks. Through comparison of consumer fatty acid proportions to those of their diet, it was determined that each diet is assimilated into a consumer in slightly different patterns. Such a finding is important as it suggests that one cannot simply compare fatty acids in a predator to those of prey without accounting for diet-specific patterns of assimilation. In Chapter 2, data from the feeding experiment described in Chapter 1 plus two others were used to explore a model that can predict the composition of prey items that have been consumed. Quantitative fatty acid signature analysis (QFASA) involves mathematically combining fatty acid profiles of prey items to yield a profile as similar to a consumer’s as possible. During this modeling exercise, I found that the most accurate data would be gained if data for each predator-prey relationship was modelled through controlled feeding experiments. This exercise was conducted with three different consumer species to ensure our results were not species specific. As a follow-up to Chapter 2, prey fish collected from two different systems were used to test how variability within a prey species affects the ability to estimate a consumer’s diet composition in Chapter 3. Data from this experiment illustrates that trophic indicators established in one freshwater system are applicable to other freshwater food webs. Data also indicated that variability among the species at a trophic level decreases as the trophic level increases. This would suggest that at higher trophic levels it might not be possible to differentiate between similar prey species consumed. As previous experiments herein used simplistic diets composed of a single species at a time, Chapter 4 explores the efficacy of fatty acid mixing models when diets consisted of combinations of prey items. Specifically, a feeding experiment was conducted that included diets of one of three invertebrate species along with even mixtures of each pair of invertebrates and a mixture of all three. Using lake trout fed the mixed diets as pseudo-wild or individuals with unknown diets, we compared their fatty acid profiles to those of the individuals fed monotypic diets. Data from this experiment indicated that diets of mixed compositions may predictably alter fatty acid profiles of consumers but quantitative models may not be as accurate as previously assumed. As a final study, in Chapter 5, adult brown trout (Salmo trutta) were used to investigate how diets of nonnative fish species from the Great Lakes may affect the fatty acids allocated to embryos. Data described within this chapter indicate that the fatty acid profiles of salmonid eggs are drastically altered by the composition of the diet. Such a result indicates that maternal diets may act as a parental care mechanism, and that genetic strains of salmonids that disfavor or are behaviorally unlikely to consume certain items may help alleviate the demand on stocking programs