SPATIAL VARIATION IN LAKE WHITEFISH RECRUITMENT IN LAKE MICHIGAN: THE POTENTIAL ROLES OF ZOOPLANKTON PREY AND THE RELATIVE FITNESS OF AGE-0 FISH

Abstract

Since the early 2000s, Lake Whitefish Coregonus clupeaformis stocks in many regions of Lake Michigan have demonstrated declines in abundance that likely reflect prolonged declines in recruitment. Conversely, recruitment is still occurring in southern Green Bay, albeit at levels lower than peaks observed in the mid-2010s. Reasons for contrasting trends in Lake Whitefish recruitment are unclear, yet understanding factors contributing to sustained recruitment in southern Green Bay and broad-scale spatial variation in recruitment trends throughout Lake Michigan is critical to implementing appropriate management actions and ensuring the sustainability of all stocks. At a broad scale, spatial patterns in Lake Whitefish recruitment appear to correspond with broad spatial trends in system productivity and zooplankton availability, forming the basis of a conceptual model where more eutrophic conditions in areas with sustained Lake Whitefish recruitment (i.e., southern Green Bay) could result in higher rates of secondary production (i.e., higher zooplankton prey densities) and age-0 Lake Whitefish with higher measures of relative fitness (i.e., weight-at-length and energy density) that are better-equipped to grow, survive, and recruit. In contrast, more oligotrophic conditions in areas that have shown signs of declining recruitment (i.e., northern Green Bay and much of the main basin of Lake Michigan proper) could result in lower rates of secondary production and age-0 Lake Whitefish that are unable to obtain adequate energetic stores needed for growth and survival. However, this conceptual model is largely untested. Therefore, the objectives of my study were to determine if primary productivity, zooplankton density, diets, and relative fitness (i.e., weight-at-length and energy density) of larval and post-larval Lake Whitefish vary among regions of Lake Michigan showing contrasting trends in Lake Whitefish recruitment. In 2021 and 2022, larval Lake Whitefish were collected in shallow embayments of southern Green Bay (SGB), northern Green Bay (NGB) and lakeside of the Door Peninsula (LDP) using hand-towed ichthyoplankton nets in March-May. Post-larval Lake Whitefish were collected from larval sampling regions as well as northern Lake Michigan (NLM), northeastern Lake Michigan (NELM), and southeastern Lake Michigan (SELM) using a beach seine during June-July. I collected water and zooplankton samples concurrently during each larval or post-larval sampling event and used generalized linear models (GLM) and t-tests to determine if productivity (total phosphorous; TP) and zooplankton metrics varied among regions and between years for each sampling period. Analysis of similarities (ANOSIM) and a similarity percentages (SIMPER) analysis were used to assess spatiotemporal variation in diet assemblages of larval and post-larval Lake Whitefish. Dummy variable regression was used to determine if loge weight- loge length relationships for larval and post-larval Lake Whitefish differed among sampling regions. Energy density was measured using bomb calorimetry. Dummy variable regression was used to determine if loge transformed energy density-weight relationships for post-larval Lake Whitefish differed among regions and between years. Larval Lake Whitefish energy density did not co-vary with fish weight and was compared among regions and between years for each sampling period using a GLM and t-tests, respectively. Mean loge TP concentration during the larval Lake Whitefish sampling period was higher in SGB than LDP in 2021 but did not vary significantly among sampling regions in 2022. During the post-larval sampling period, loge TP concentration in SGB was higher than in NELM during 2021 but did not differ from other sampling regions during 2022. Total and taxon-specific zooplankton densities did not significantly differ among sampling regions during larval Lake Whitefish sampling periods in 2021 or 2022 but density of total zooplankton in SGB was 47-164% higher than NGB and 302-493% higher than LDP depending on year. In 2021, loge calanoid copepod density in SGB was lower than NELM and in 2022, loge cyclopoid and Bosmina spp. densities were higher in SGB than LDP and NELM. Few between-year differences were noted in total or taxon-specific zooplankton density. Diet composition of larval Lake Whitefish did not vary among regions or between years; cyclopoid and harpacticoid copepods accounted for > 80% of diets by number, regardless of region or year. Despite similarities in zooplankton densities and diet composition among sampling regions, the percent of larvae with empty stomachs was lowest in SGB during 2021 (SGB = 15%; NGB = 27%; LDP = 21%) and 2022 (SGB = 11%; NGB = 26%; LDP = 34%), potentially reflecting spatial differences in prey availability or feeding efficiency. Diet assemblages of post-larval Lake Whitefish differed significantly among most combinations of regions although the degree of separation among regions varied (R-value range = 0.138-0.759). Diets within Green Bay were comprised primarily of Bosmina spp., while chironomids were a major component of diets elsewhere. Only a single instance of an empty stomach (% empty < 1%) was observed in post-larval diets regardless of region or year. Spatial patterns in productivity and zooplankton availability during the larval and post-larval stage did not manifest in the same spatial patterns in relative fitness. Although loge weight- loge length relationships of larval and post-larval Lake Whitefish varied among regions, slope and back-transformed weight of larvae and post-larvae in SGB were often similar to or less than the same metrics in regions experiencing recruitment declines. During both years, slopes of weight-length relationships for larval whitefish were steeper in SGB and NGB than in LDP but did not differ between regions within Green Bay. Similarly, energy density did not vary among regions during the larval period and slope and back-transformed energy density of post-larval fish in SGB were often similar to or less than other regions. Results from this study suggest that while measures of primary and secondary productivity are often greater in SGB than other regions, these differences did not manifest in greater relative fitness of larval and post-larval Lake Whitefish. Potentially, contrasting trends in recruitment originate from spatial variation in feeding efficiency during the larval stage, which may allow for increased survival of larvae. During both years of this study, ratios of zooplankton:fish in the environment and the percentage of non-empty stomachs in larvae were higher in SGB when compared to other regions. Future research to describe spatial differences in feeding efficiency and survival of larval Lake Whitefish between SGB and other regions may be warranted as greater survival during early life stages may influence year-class strength and spatial differences in survival may lead to the observed spatial patterns in Lake Whitefish recruitment. Despite the perceived importance of factors operating during early life stages for influencing recruitment of Lake Whitefish, many aspects of whitefish early life history are still poorly understood. For example, relatively little life history information exists for the period between when post-larvae can be captured using seines and when recruitment to recreational and commercial fisheries occurs at ages ≥ 5. Anecdotally, my study collected large numbers of larval and post-larval Lake Whitefish in all study regions, suggesting that hatching and survival to age-0 occurs in many areas of Lake Michigan, even those where recruitment declines are apparent at later life stages. This observation may indicate that a recruitment bottleneck occurs at life stages not considered in my study. Examination of relationships between subsequent life stages may help identify the timing of a recruitment bottleneck and subsequently provide a directive for the focus of future research.