Temporal and Spatial Genetic Consistency of Walleye Spawning Groups

Population genetic compositions of the three largest spawning groups of Lake Erie walleyes Sander vitreus (Maumee River, Sandusky River, and Van Buren Bay reefs) were tested for temporal and spatial consistency across 14 years using nine nuclear DNA microsatellite loci from 726 adult walleyes collected in 1995, 1998, 2003, 2007, and 2008. Previous genetic work focused on a one-time “snapshot”; an earlier study by our laboratory found genetic connectivity among the three spawning groups in 2003, whereas most other spawning runs across Lake Erie were genetically distinct. Present results show overall year-to-year genetic consistency of walleye spawning groups; no significant differences were found among collection dates within an annual run, between sexes, or among age-cohorts. Overall, walleyes spawning at the Van Buren Bay reefs were genetically divergent from those spawning in the Maumee and Sandusky rivers, reflecting geographic distance; the latter two groups were genetically closer, with slight differences that suggested more recent divergence, higher gene flow, or both. Individual year comparisons among the three sites showed some stochasticity, and the spawning groups appeared to be more similar in some years than in others. The Van Buren Bay spawning group in 1995 was the most divergent sample and had the greatest genetic self-assignment (100%); this may reflect some changes over time. Results demonstrate the importance of sampling over several years of spawning runs to understand overall patterns of walleye genetic stock structure, which show remarkable genetic consistency across an open-lake system. The walleye Sander vitreus is one of the most ecologically and economically important exploited fishes in the Laurentian Great Lakes, and Lake Erie supports the world’s largest freshwater fishery (Fielder 2002; Hubbs and Lagler 2004) that included an estimated 21.2 million walleyes in 2011 (WTG 2011). Maintaining and identifying healthy walleye genetic stocks have been identified as key to Lake Erie habitat restoration efforts (Ryan et al. 2003), with Hallerman et al. (2003) defining genetic stocks as, “population subunits that interbreed freely in given geographic locations, share a common gene pool, and differ significantly from other subunits.” Tagging studies have shown that although walleyes comingle during nonspawning months, they tend to return to the same spawning locations with an apparently high degree of philopatry (Ferguson and Derksen 1971; Wolfert and Van Meter 1978; Wang et al. 2007). Reproductive isolation *Corresponding author: [email protected] 1Present address: U.S. Fish and Wildlife Service, 4625 Morse Road, Suite 104, Columbus, Ohio 43230, USA. Received July 8, 2011; accepted November 14, 2011 among spawning groups can lead to genetically distinct fishery stocks, which adapt to local environments over time (Horrall 1981; Hauser and Carvalho 2008). Maintaining the genetic diversity of these spawning stocks may be important for long-term viability and adaptive potential of the walleye population across Lake Erie. However, since Lake Erie lacks geographic divisions among most walleye spawning sites there appears to be ample potential for this vagile species to interbreed, which would result in genetic homogeneity and little stock structure. These alternatives are tested here. Genetic studies of walleyes have shown both large-scale continental genetic divergence patterns that corresponded to their origins from glacial refugia and historic separations in major watersheds (Billington et al. 1992; Stepien and Faber 1998; Stepien et al. 2009), as well as discrete intralacustrine 660 D ow nl oa de d by [ C ar ol S te pi en ] at 1 7: 18 1 7 M ay 2 01 2 WALLEYE TEMPORAL GENETIC STRUCTURE 661 groups and fine-scale patterns among spawning groups from tributary and reef sites (Merker and Woodruff 1996; Stepien and Faber 1998; Strange and Stepien 2007; Stepien et al. 2009, 2010). Strange and Stepien (2007) discerned genetically divergent spawning groups of walleyes across Lake Erie, especially in eastern basin tributaries; they also identified greater genetic connectivity among some walleye spawning groups from 2003 along the southern lakeshore, extending from the western basin’s Maumee and Sandusky rivers to the eastern basin’s Van Buren Bay reefs. Since Strange and Stepien’s (2007) data were based on a single year, that genetic “snapshot” may or may not reflect a consistent pattern, which is tested here. We evaluated contemporary genetic population structure among walleye spawning groups to test for continuity of spatial patterns for 5 years of spawning run data spanning 14 years (1995, 1998, 2003, 2007, and 2008) within and among three of the largest Lake Erie spawning groups (Maumee River, Sandusky River, and Van Buren Bay reefs). Nine nuclear microsatellite (μsat) loci (the same ones used by Strange and Stepien 2007) were used to test the possible influence of spawning years, sampling dates, spawning locations, sex, and age-cohorts on genetic composition. The following null (or alternative) hypotheses were tested: (1) population genetic compositions of walleye spawning groups are (or are not) consistent from year to year; (2) genetic compositions are consistent (or differ) across the course of a given spawning run, i.e., from earlier versus later in the run; (3) males and females have the same (or different) genetic patterns; (4) genetic compositions among age-cohorts do not vary (or vary) within a spawning group; and (5) relationships among spawning groups at the three sites remain consistent (or differ) from year to year. Life History and Walleye Recruitment In early spring, walleyes migrate to their natal sites to spawn, aggregating in tributaries and shallow lake reefs (Ferguson and Derksen 1971; Wolfert and Van Meter 1978). Mark-andrecapture studies recovered most Lake Erie and Lake St. Clair walleyes near their original spawning sites during the subsequent spring spawning season or seasons (Wang et al. 2007). It is hypothesized that imprinting in walleyes occurs during the early life history, as walleyes have a highly developed olfactory system used to detect natal spawning sites and may be able to discern the pheromones of kin (see Horrall 1981; Gerlach et al. 2001). At night, the female walleye releases thousands of eggs that settle in crevices and are fertilized by one or a few males (Regier et al. 1969; Kerr et al. 1997). No parental care is provided (Zhao et al. 2009); after spawning, the adults move to adjacent bays and littoral areas, then travel offshore to summer feeding grounds (Kerr et al. 1997). Eggs hatch after about 3 weeks and where water temperature is warmer, development is more rapid and survival is better (Johnson 1961; Kerr et al. 1997; Roseman et al. 2005). The larvae primarily passively drift via currents and are dependent on tributary discharge rates for transport to nursery areas in shallow vegetated areas (Mion et al. 1998; Jones et al. 2003; Roseman et al. 2005). They school and feed on zooplankton in the nearshore nursery grounds, where food abundance and high water turbidity are believed to augment survival (Roseman et al. 2005). The young of the year (age 0) move into deeper water as juveniles in late summer and eat mayflies, amphipods, and fishes (Kerr et al. 1997). Males generally mature by age 2 and females by age 3 (Hatch et al. 1987). Walleye Spawning Habitats in Lake Erie Modern Lake Erie formed over the course of about 12,000 up to 3,000 years ago (ya) after the retreat of the Wisconsinan glaciers and has experienced several changes in lake levels (Holcombe et al. 2003). Lake Erie today contains three basins— western, central, and eastern—of which the western basin is the shallowest, youngest, and most eutrophic and consisted of a riverine system until about 3,600 ya (Holcombe et al. 2003; see Figure 1). Walleyes and other fishes are believed to have colonized western Lake Erie via the Wabash–Maumee River system, by migrating northward from the Mississippian glacial refugium (Bailey and Smith 1981; Strange and Stepien 2007). Today the western basin houses the largest walleye spawning runs in the Great Lakes (Mion et al. 1998; Einhouse and MacDougall 2010), including those located approximately 25 km upstream from the mouths of the Maumee and Sandusky rivers (Mion et al. 1998). The spawning runs begin shortly after the ice recedes and peak about the third week of April (Baker and Manz 1971). The Ballville Dam constructed in 1911 on the Sandusky River restricts upstream accessibility (Mion et al. 1998) and now is slated for removal (Roger Knight, Ohio Division of Wildlife [ODW], Sandusky, personal communication; Scudder Mackey, Habitat Solutions, Chicago, Illinois, personal communication), rendering these “before” genetic data an important baseline. The Maumee River has a mean discharge rate about four times greater than the Sandusky River. Both rivers have relatively low current velocities and gradients and empty into large bays 2–3 m in depth that serve as walleye nursery grounds (Mion et al. 1998). By comparison, the central basin (Hartman 1973) is deeper and contains limited spawning substrate (Nepszy et al. 1991); summer migrants from the western basin spawning groups support its fishery (Einhouse and MacDougall 2010). The eastern basin of Lake Erie, which formed about 12,000–10,000 ya (Bolsenga and Herdendorf 1993; Holcombe et al. 2003; Ryan et al. 2003; Figure 1), is the deepest and oldest basin and houses fewer walleyes (Einhouse and MacDougall 2010). Those fish spawn on a reef complex around Van Buren Bay and in several tributaries, including the Grand River, Cattaraugus Creek, and Smoke’s Creek (genetic stock structures of these walleyes were studied by Stepien et al. 2004, 2010; Strange and Stepien 2007). Spring warming of the eastern basin trails that of the western and central basins by several weeks, and its walleye spawning runs thus occur later (Hartman 1973). Both sexes D ow nl oa de d by [ C ar ol S te pi en ] at 1 7: 18 1 7 M ay 2 01 2