Temporal variation in the niche partitioning of Lake Michigan salmonines as it relates to alewife abundance and size structure

Abstract

Stable isotope analyses offer a useful means for quantifying ecological niche dimensions, though few studies have examined isotopic response of an ecological community with respect to resource gradients such as fluctuations in prey availability. Stable carbon and nitrogen isotopes were measured for Lake Michigan salmonines and their prey collected from 2014 to 2016. Bayesian ellipse and mixing model analyses were used to quantify isotopic niche characteristics and diets, respectively, among species and years. During the three-year study period, abundance and size structure of preferred alewife prey changed substantially and offered an opportunity to explore predator isotopic niche response and diet shifts along a prey resource gradient. Results suggested increased reliance on alewives, especially small alewives, over the study period and were consistent with greater availability of this prey. However, differential use of alewife size classes and alternative prey sources by salmonine predators was apparent, which suggested possible resource partitioning. Characterization of ecological niche overlap using stable isotopes likely requires consideration of shared resource availability as well as specific prey and habitat preferences.

Bidirectional energy subsidies for fish in river mouths of a large lake as revealed by stable isotopes and fatty acids

“…Mixing model including δ 13 C and δ 15 N : Resource subsidies for fish across habitats were estimated using Bayesian mixing models in MixSIAR package (version 3.1.12; Stock et al 2018) in R version 4.0.3 (R Core Team 2020). MixSIAR models were modified from the standard rjags package to operate using the runjags package in R which facilitated parallel processing (i.e., four cores) and extension of Markov Chain Monte Carlo (MCMC) chain length (Turschak et al 2022). Models for each fish species collected in each habitat were structured with one random factor—“Habitat” (e.g., “Lake” or “River”).…”

“…Models were run with four parallel MCMC chains of length 50,000, burn‐in length 20,000, and chains were thinned by 10, with uninformed priors. Model convergence was evaluated using trace plots, and Gelman–Rubin Diagnostic not exceeding 1.05 for more than 5% of parameter estimates as described in Turschak et al (2022). Model end‐members from nearshore Lake Michigan and river mouth habitats were weighted by assumed equal mean contribution of benthic and pelagic prey (50 : 50).…”

Updated decision analysis to inform multi-species salmonine management in Lake Michigan

“…Although this approach is reasonable, it has not been fully vetted and lacks measures of uncertainty. In addition, the Lake Michigan food web has undergone important changes (Kornis et al 2012;Madenjian et al 2016) since the 2013 decision analysis was implemented, the most notable of which is the expansion of invasive round goby (Neogobius melanostomus) throughout the lake, which has become an important prey source for species like lake trout (Happel et al 2020;Leonhardt et al 2020;Kornis et al 2020;Turschak et al 2022). In addition, there have been changes to the contribution of other species to total stocking effort in Lake Michigan, as well as changes in the movement patterns of Chinook salmon between Lakes Michigan and Huron (Clark et al 2017;Kornis et al 2017).…”

Seasonality can affect ecological interactions between fishes of different thermal guilds

“…We lack data about the seasonal ecology of prey in our study lakes and in general little is known about the seasonal ecology of many prey fish species. Further study of prey fish abundance, habitat, and diet and how these factors change through time would improve our understanding of top predator seasonal niches (Turschak et al, 2022).…”