Evaluating Methods to Describe Dietary Patterns of Lake Michigan Salmonids

Documenting trophic relationships in aquatic ecosystems can facilitate understanding of not only system processes, but also the potential responses of food webs to stressors. Often, trophic studies assume consistent behavior and trophic roles among individuals in a population, but intraspecific diet variation, such as individual specialization, can play a critical role in food web complexity and can promote ecosystem resilience. In Lake Michigan, the introduction of invasive species (e.g., zebra mussel, Dreissena polymorpha; quagga mussel, Dreissena bugensis; round goby, Neogobius melanostomus) and reduced nutrient loading has resulted in changes in nutrient dynamics, system productivity, and community composition over the past two decades. As a result, abundances of many forage fish have declined, including alewife (Alosa pseudoharengus) which have historically supported the five dominant salmonid species of Lake Michigan (brown trout, Salmo trutta; Chinook salmon, Oncorhynchus tshawytscha; Coho salmon, Oncorhynchus kisutch; lake trout, Salvelinus namaycush; rainbow trout, Oncorhynchus mykiss). With these ecosystem changes, there is uncertainty as to the extent of how different species of salmonids will transition to alternative prey items (e.g., round goby). Common methods for examining diet patterns and trophic linkages include stomach content analysis, stable isotope ratios (e.g., δ13C and δ15N), and fatty acid composition, but these methods vary in temporal resolution and have differential biases. Furthermore, elucidating agreement of these trophic indicators and whether or not agreement is consistent across species can improve their use in future food web studies. The first research chapter of this thesis investigated the diet complexity of Lake Michigan salmonids by evaluating stomach content composition, diversity, and potential specialized consumption of different alewife lengths. Stomach contents revealed that Chinook salmon almost exclusively consumed alewife and had a lower diet diversity compared to the other four species, which consumed round goby (brown trout and lake trout), aquatic invertebrates (Coho salmon), and terrestrial invertebrates (rainbow trout) in addition to alewife. Although there were clear spatio-temporal and size-related feeding patterns for each species, much of the variation in diet composition and diet diversity was present at the individual-level. Additionally, salmonid species appeared to consume the entire size range of alewife that were available to them and individually specialized on alewife lengths. Due to their reliance on alewife, it is likely that Chinook salmon may be more negatively impacted than other salmonid species if alewife abundance continue to decline in Lake Michigan. The second research chapter assessed the agreement of multiple trophic indicators. Although we found agreement among trophic indicators across the five salmonid species using linear and logistic models, particularly between stomach contents, δ13C, and fatty acid 16:1n-7, there was significant variation in relationships across species, potentially due to variation among salmonids in specific prey items consumed (e.g., alewife and round goby) and species-specific regulation of fatty acids. Additionally, δ15N estimated from stomach contents using linear mixing models were typically greater relative to observed δ15N, which may suggest small alewife were underrepresented in stomachs of 2016 angler-caught salmonids. Lastly, stomach contents underestimated benthic resource use by rainbow trout, which may be related to biases associated with fish collection methods and stomach content analysis. Overall, the results of trophic indicator comparisons indicate that caution should be taken when generalizing trophic relationships across species and to consider biases associated with trophic indicators, especially when relying on a single diet metric.