Seasonally Dynamic Diel Vertical Migrations of Mysis diluviana, Coregonine Fishes, and Siscowet Lake Trout in the Pelagia of Western Lake Superior

Diel vertical migrations are common among many aquatic species and are often associated with changing light levels. The underlying mechanisms are generally attributed to optimizing foraging efficiency or growth rates and avoiding predation risk (μ). The objectives of this study were to (1) assess seasonal and interannual changes in vertical migration patterns of three trophic levels in the Lake Superior pelagic food web and (2) examine the mechanisms underlying the observed variability by using models of foraging, growth, and μ. Our results suggest that the opossum shrimp Mysis diluviana, kiyi Coregonus kiyi, and siscowet lake trout Salvelinus namaycush migrate concurrently during each season, but spring migrations are less extensive than summer and fall migrations. In comparison with M. diluviana, kiyis, and siscowets, the migrations by ciscoes C. artedi were not as deep in the water column during the day, regardless of season. Foraging potential and μ probably drive the movement patterns of M. diluviana, while our modeling results indicate that movements by kiyis and ciscoes are related to foraging opportunity and growth potential and receive a lesser influence from μ. The siscowet is an abundant apex predator in the pelagia of Lake Superior and probably undertakes vertical migrations in the water column to optimize foraging efficiency and growth. The concurrent vertical movement patterns of most species are likely to facilitate nutrient transport in this exceedingly oligotrophic ecosystem, and they demonstrate strong linkages between predators and prey. Fishery management strategies should use an ecosystem approach and should consider how altering the densities of long-lived top predators produces cascading effects on the nutrient cycling and energy flow in lower trophic levels. An understanding of how and why diel vertical migration (DVM) patterns are dynamic among species, seasons, and years is important for understanding within-lake processes, such as *Corresponding author: [email protected] Received March 10, 2010; accepted April 21, 2011 Published online December 5, 2011 species interactions, energy transfer, and nutrient cycling (Iwasa 1982; Salonen et al. 1984). Diel vertical migrations occur in many aquatic ecosystems within most groups of organisms, 1504 D ow nl oa de d by [ U ni ve rs ity o f M in ne so ta L ib ra ri es , T w in C iti es ] at 0 8: 59 2 8 O ct ob er 2 01 4 DIEL VERTICAL MIGRATIONS OF THE LAKE SUPERIOR PELAGIC FOOD WEB 1505 including zooplankton (Worthington 1931; Gliwicz 1986), planktivores (Bohl 1980; Janssen and Brandt 1980), and piscivores (Weng and Block 2004; Hrabik et al. 2006). For most species, DVM patterns correspond to fluctuations in light levels (Appenzeller and Leggett 1995; Gal et al. 1999), while thermal gradients may influence the magnitude of many migrations (Gal et al. 2004; Boscarino et al. 2007). The mechanism underlying this migratory pattern is commonly attributed to optimization of foraging efficiency (Narver 1970; Levy 1990) or growth rates (McLaren 1963; Brett 1971; Wurtsbaugh and Neverman 1988) while avoiding predation risk (μ; Zaret and Suffern 1976; Lampert 1993). Consequently, seasonal and annual variation in the abiotic and biotic environments within lake ecosystems can cause differences in vertical migration patterns among species. These differences have a direct influence on predator–prey interactions, energy flow, and nutrient cycling rates, which can ultimately affect entire aquatic ecosystems (Iwasa 1982; Steinberg et al. 2002). In Lake Superior, the pelagic food web is composed of species that are known to perform—and to be well adapted for—DVMs. The offshore pelagic community is dominated by the opossum shrimp Mysis diluviana (formerly M. relicta), kiyi Coregonus kiyi, cisco Coregonus artedi, and siscowet lake trout Salvelinus namaycush (hereafter, “siscowet”); the shortjaw cisco Coregonus zenithicus, bloater Coregonus hoyi, and rainbow smelt Osmerus mordax are also present but at relatively low densities (Bronte et al. 2003; Yule et al. 2009). Vertical migrations by mysids are known to occur in many systems (Dakin and Latarche 1913; Beeton and Bowers 1982), including Lake Superior (Bowers 1988; Jensen et al. 2009). The DVM pattern exhibited by various coregonines has been documented in Lake Michigan (TeWinkel and Fleischer 1999) and Lake Superior (Hrabik et al. 2006; Yule et al. 2007). In addition, siscowets are known to migrate vertically in Lake Superior (Hrabik et al. 2006; Stockwell et al. 2010). Although vertical movements have been documented for M. diluviana, kiyis, ciscoes, and siscowets, most DVM research has focused on coregonines in Lake Superior (Hrabik et al. 2006; Jensen et al. 2006; Stockwell et al. 2010). Hrabik et al. (2006) provided the first empirical documentation that coregonine fishes in Lake Superior migrate vertically, although speciesspecific DVM patterns were not examined. Jensen et al. (2006) used foraging, growth, and predation models to explore the mechanisms driving movements of coregonines; those authors concluded that the movements were related to growth potential and μ, but the conclusions were based on limited field data and the study failed to incorporate a seasonal component. Stockwell et al. (2010) added detail by examining DVM patterns of kiyis and ciscoes separately; they found that kiyis have more extensive vertical movements than ciscoes. In addition, Stockwell et al. (2010) observed that spring migrations by kiyis were less pronounced than summer and fall migrations. It is clear that many of the important species in the Lake Superior pelagia migrate vertically, but the following remain unclear: (1) how migrations change seasonally for each dominant species, (2) how those movements are interrelated in a food web context, and (3) the mechanisms driving those changes for each species across seasons. Seasonal changes in temperature, light, prey availability, and predator distributions may alter DVM behavior for a particular species among seasons, thereby dictating how the trophic levels interact (Lampert 1993; Hays 2003). For example, M. diluviana migrate less extensively in spring than in other seasons and are found throughout the water column at night, probably because of isothermal conditions in the Great Lakes during that time period (Gal et al. 2004; Boscarino et al. 2009). For some fish species, such as juvenile kokanee Oncorhynchus nerka and walleye pollock Theragra chalcogramma, DVM patterns are influenced by seasonal changes in water temperature, light, and prey–predator densities and distributions (Hardiman et al. 2004; Adams et al. 2009). However, for the vendace Coregonus albula and Fontane cisco Coregonus fontanae, the DVM patterns do not change significantly by season (Mehner et al. 2007). Overall, seasonal changes in the physical and biological environment could modify DVM patterns for a particular species depending on the mechanisms driving those movements. Models of foraging, growth, and μ have been used to explore factors influencing movement patterns of many species (Wright and O’Brien 1984; Mason and Patrick 1993; Jensen et al. 2006). In general, foraging rate models predict where predators can maximize foraging opportunity based on prey densities, prey and predator swimming speeds, and predator reaction distances (Gerritsen and Strickler 1977). Growth rate models build upon the general foraging model framework by incorporating prey availability and water temperature to predict areas with the highest growth rate potential (GRP) for a given organism based on fundamental bioenergetic principles (Kitchell et al. 1977). Although foraging and growth models may provide explanations for the movements of top predators in aquatic ecosystems, they commonly fail to explain the movements of organisms at lower trophic levels, where μ may be a driving factor (Iwasa 1982; Werner et al. 1983). Therefore, it is important to consider models that incorporate foraging opportunity, GRP, and μ when trying to understand general movement patterns. The purpose of this study was to quantify characteristics of DVM patterns among seasons and years for three trophic levels in the Lake Superior food web. Our first objective was to empirically characterize diel movement patterns of the three trophic levels simultaneously by using hydroacoustics (for fish) and an optical plankton counter (OPC; for zooplankton); both methods enable the collection of data at a fine temporal resolution. Our second objective was to explore the mechanisms driving movement patterns by using a modeling approach, which was similar to that of Jensen et al. (2006) and used various abiotic (i.e., temperature and light) and biotic (i.e., predator–prey diets, distributions, and densities) inputs. An understanding of how and why organisms migrate in Lake Superior is important because it will help to elucidate nutrient and energy flow pathways through D ow nl oa de d by [ U ni ve rs ity o f M in ne so ta L ib ra ri es , T w in C iti es ] at 0 8: 59 2 8 O ct ob er 2 01 4 1506 AHRENSTORFF ET AL. the system and to clarify predator–prey linkages, thus aiding in fishery management decisions. Based on previous studies, we hypothesized that all four dominant species would migrate during all seasons and that less-extensive migrations would occur during spring, when the lake is isothermic and prey resources are dispersed throughout the water column (Gal et al. 2004; Stockwell et al. 2010). Alternatively, we hypothesized that ciscoes would exhibit less-consistent migration patterns among seasons because of their larger size and lowerμ. In addition, we hypothesized that foraging or bioenergetic efficiency and μ would drive the movement patterns of M. diluviana and kiyis, whereas μ would have a lesser influence on movements by ciscoes because of their decreased vulnerability to predation (Jensen et al. 2006; Stockwell et al. 2010). Lastly, we hypothesized that siscowets would migrate concurrently with kiyis during all seasons and would also feed upon deepwater sculpin Myoxocephalus thompsonii in bottom areas to satisfy their energetic requirements as the top predator in the pelagia of Lake Superior.