James H. Selgeby

Population Recovery and Natural Recruitment of Lake Trout at Gull Island Shoal, Lake Superior, 1964–1992

We documented an increase in the abundance of wild lake trout (Salvelinus namaycush) at Gull Island Shoal in western Lake Superior and examined the relationship between parental-stock size and recruitment of age-0 fish in 1964–1992. Abundance of adult wild female lake trout and densities of age-0 fish both increased during the 28-year period. A significant positive, linear relationship (P = 0.0002) was found between the abundance of wild females on the spawning reef in the fall and density of age-0 lake trout on adjacent nursery grounds in August and September of the following year. The abundance of hatchery-origin females did not explain significant amounts (P = 0.107) of variation in recruitment. We concluded that most recruitment in 1965–1992 was the result of natural reproduction of wild females. After 28 years of recovery the Gull Island Shoal lake trout population appears to have additional capacity to increase because the stock-recruitment relationship is still linear. Therefore, restoration periods on the order of 30 years may be needed for other lake trout populations in the Great Lakes. We recommend that the refuge established to protect this population be maintained to allow further study of the relationship between parental stock and recruitment, and to provide a major source of recruitment to the lake trout population in the surrounding waters.

Hatching, Dispersal, and Bathymetric Distribution of Age-0 Wild Lake Trout at the Gull Island Shoal Complex, Lake Superior

We studied age-0 lake trout (Salvelinus namaycush) associated with spawning and nursery areas of the Gull Island Shoal complex in western Lake Superior. Post-emergent age-0 lake trout were captured on rocky spawning substrate with a 3-m beam trawl and at the nursery area with a bottom trawl from June to September 1990 and June to August 1991. Catch data suggested that age-0 lake trout move distances of 7–11 km to the nursery area over a 3-month period. Water currents, measured at Gull Island Shoal, may be a part of the transport mechanism. Examination of daily-growth increments on the sagittae and back-calculation from the date of capture revealed that most fish hatched between 6 June and 19 July in 1990 and between 30 April and 30 May in 1991. The duration of the hatch was 100 days in 1990 and 120 days in 1991, and the estimated incubation period is about 7 months for lake trout eggs at this site. Similar hatch-date distributions of age-0 captured on different sampling dates suggested that natural mortality was low.

Lake Trout Restoration in the Great Lakes: Stock-Size Criteria for Natural Reproduction

Abstract We examined the question of whether the lake trout restoration program in the Great Lakes has developed brood stocks of adequate size to sustain natural reproduction. Stock size criteria were developed from areas of the Great Lakes where natural reproduction has been successful (defined as detection of age-1 or older recruits by assessment fishing). We contrasted them with stocks in areas with no natural reproduction. Based on the relative abundance of spawners measured in the fall and the presence or absence of natural reproduction in 24 areas of the Great Lakes, we found three distinct sets of lake trout populations. In seven areas of successful natural reproduction, the catch-per-unit-effort (CPE) of spawners ranged from 17 to 135 fish/305 m of gillnet. Stock sizes in these areas were used as a gauge against which stocks in other areas were contrasted. We conclude that stock densities of 17–135 fish/305 m of gill net are adequate for natural reproduction, provided that all other requirements are met. No natural reproduction has been detected in seven other areas, where CPEs of spawners ranged from only 3 to 5 fish/305 m. We conclude that spawning stocks of only 3–5 fish/305 m of net are inadequate to develop measurable natural reproduction. Natural reproduction has also not been detected in ten areas where CPEs of spawners ranged from 43 to 195 fish/305 m of net. We conclude that spawning stocks in these ten areas were adequate to sustain natural reproduction, but that some factor other than parental stock size prevented recruitment of wild lake trout.

Reestablishing a Spawning Population of Lake Trout in Lake Superior with Fertilized Eggs in Artificial Turf Incubators

Abstract Fertilized eggs from lake trout Salvelinus namaycush were placed in artificial turf incubators and deployed on Devils Island Shoal, Lake Superior, in an attempt to reestablish a spawning population on this once important spawning area. Efficacy was measured by the changes in catch rates, age composition, and origin of adult lake trout returning to the shoal in the fall in subsequent years. The abundance of lake trout spawners without fin clips, which implies that these fish hatched in the lake, increased throughout the sampling period, whereas the abundance of hatchery-reared fish (indicated by one or more fin clips) stocked for restoration purposes remained low. Year-class-specific stock–recruitment analysis suggested that the recruitment of unclipped spawners was related to the number of eggs planted in previous years rather than to spawning by the few adult lake trout visiting the reef. Increases in adult fish at Devils Island Shoal were independent of trends at adjacent sites, where unclipped...

Gill-Net Saturation by Lake Trout in Michigan Waters of Lake Superior

Abstract We conducted experimental fishing for lake trout Salvelinus namaycush in Michigan waters of Lake Superior to determine the importance of soak time on catch per effort (CPE) in numbers per kilometer of standard gill net. We modeled CPE as a nonlinear function of the number of nights between setting and lifting (soak time), in which the nets fill at a certain rate toward some maximum after which the nets cannot hold more fish. We found that lake trout CPE increased with soak time at a rate that varied with lake trout density toward a saturation level that was independent of lake trout density. The CPE values of nets soaked 2–5 nights divided by the CPE of nets soaked 1 night were significantly lower than would be expected had CPE increased as a linear function of the number of nights soaked. We derived a means for correcting gill-net CPE values for differing soak times to a common base of 1 night soaked. We concluded that it is inappropriate to assume lake trout catches in gill nets will increase i...

Declining survival of lake trout stocked during 1963-1986 in U.S. waters of Lake Superior

Abstract The average catch per effort (CPE) values for the 1963–1982 year-classes of stocked lake trout Salvelinus namaycush caught at age 7 in gill nets and for the 1976–1986 year-classes caught at ages 2–4 in trawls declined significantly in U.S. waters of Lake Superior. The declines in CPE were not explained by reduced stocking, but rather by significant declines in survival indices of the year-classes of stocked lake trout. Increases in mortality occurred in year-classes before the fish reached ages 2–4, before they were recruited into the sport and commercial fisheries, and before they reached sizes vulnerable to sea lamprey predation. We conclude that declining abundance of stocked lake trout resulted from increased mortality, which may have been caused by competition, predation, or by a combination of these and other factors. Restoration of lake trout in Lake Superior may now depend on prudent management of naturally reproducing stocks rather than on stocking of hatchery-reared fish.

Predation by Ruffe (Gymnocephalus cernuus) on Fish Eggs in Lake Superior

Ruffe (Gymnocephalus cernuus) were introduced to North America in the Duluth, Minnesota—Superior, Wisconsin harbor, which is the westernmost point on the Laurentian Great Lakes. The species proliferated in the harbor and became the subject of research which has gradually revealed certain characteristics of the biology and population growth of the ruffe. In this study ruffe in Southwestern Lake Superior were found to have eaten benthic organisms and eggs of lake herring (Coregonus artedii). Overwinter predation by ruffe on eggs of lake herring and of other fall spawning Great Lakes fishes might pose a substantial new source of overwinter mortality.

Density-Independent Survival of Wild Lake Trout in the Apostle Islands Area of Lake Superior

The lake trout (Salvelinus namaycush) stock at Gull Island Shoal in western Lake Superior was one of only a few stocks of lean lake trout in the Great Lakes that survived overfishing and predation by the sea lamprey (Petromyzon marinus). Since the mid 1960s, the abundance of wild recruits measured at age 0 and the number of age-7 to -11 wild fish recruited to the fishable stock have increased. We used the Varley-Gradwell method to test for density-dependent survival between these life stages. Survival from age-0 to ages 7–11 was not affected by increasing density, which suggests that further increases in recruitment and stock size are still possible. We suggest that testing for the existence of density-dependent survival can be used to indicate when lake trout populations are rehabilitated.

A Bioenergetics Modeling Evaluation of Top-down Control of Ruffe in the St. Louis River, Western Lake Superior

Ruffe (Gymnocephalus cernuus), were accidentally introduced into the St. Louis River estuary, western Lake Superior, in the mid 1980s and it was feared that they might affect native fish through predation on eggs and competition for forage and habitat. In an effort to control the abundance of ruffe and limit dispersal, a top-down control strategy using predators was implemented in 1989. We used bioenergetics modeling to examine the efficacy of top-down control in the St. Louis Riverfront 1991 to 1994. Five predators––northern pike (Esox lucius), walleye (Stizostedion vitreum vitreum), small-mouth bass (Micropterus dolomieui), brown bullhead (Ictalurus nebulosus), and yellow perch (Perca flavescens)––were modeled to determine their consumption of ruffe and four other native prey species––spottail shiner (Notropis hudsonius), emerald shiner (Notropis atherinoides), yellow perch (Perca flavescens), and black crappie (Pomoxis nigromaculatus). Although predators ate as much as 47% of the ruffe biomass in 1 year, they were not able to halt the increase in ruffe abundance. The St. Louis River is an open system that allows predators to move freely out of the system, and the biomass of managed predators did not increase. A selectivity index showed all five predators selected the native prey and avoided ruffe. The St. Louis River has several predator and prey species creating many complex predator-prey interactions; and top-down control of ruffe by the predators examined in this study did not occur.

Winter Diet of Lake Herring (Coregonus artedi) in Western Lake Superior

Lake herring (Coregonus artedi) and zooplankton samples were simultaneously collected through the ice in the Apostle Islands region of western Lake Superior to provide information on the winter feeding ecology of lake herring. Zooplankton constituted the entire diet of the 38 lake herring collected for this study. We found no evidence of piscivory, although it has been reported by anglers. Diet selectivities were calculated using a Wilcoxon signed-ranks test and showed a preference of lake herring for larger zooplankton, especially Diaptomus sicilis, whereas the smaller copepod, Cyclops bicuspidatus thomasi, and immature copepod stages were selected against. These data document that overwintering copepods are food for a broad size range of lake herring in winter.