Recent implementation of size-based regulations in recreational fisheries for walleye Sander vitreus have led to more released walleyes and presumably to more losses of released fish. We conducted this study to estimate hooking mortality in Mille Lacs, Minnesota, and to determine factors that influence the survival of released walleyes. Volunteers and Minnesota Department of Natural Resources employees sampled walleyes with common angling methods in 2003 and 2004 on Mille Lacs (n ¼ 1,246). Simple hooking mortality rates ranged from 0% (95% confidence interval ¼ 0-1.8%; n ¼ 204) in May, when lake water temperatures were less than 208C, to 12.2% (9.2-15.9%; n ¼ 392) in the July-August period, when lake water temperatures were at least 208C. We used logistic regression within generalized linear or additive models to determine influential variables. Hooking mortality was most associated with water temperature, bleeding, fish length, hook location, and fish floating upon release. Mortality increased as the water warmed above 188C and was higher for fish that bled at temperatures less than 248C but similar for both bleeding and nonbleeding fishes at temperatures of 248C or more. Fish hooked in the throat or stomach died at higher rates than fish hooked in the jaw, inner mouth, or gills and those that were externally foul-hooked, especially when they were smaller. Although fish of medium length (300-600 mm) were more likely to be deep hooked, they died less frequently than walleyes of other lengths. Cutting the line did not significantly improve survival in deeply hooked fish. Mortality was similar between live bait jigs and live bait regular hooks. Most observed hooking mortality was caused by damage to major internal organs. Hooking mortality is minimized when anglers fish in cool water, use active fishing methods, and catch medium-length walleyes.

We estimated prey consumption by burbot (Lota lota) based on diet, mortality, growth, maturity, thermal history, population density and a bioenergetics model derived for a similar, cold-water gadoid, the Atlantic cod (Gadus morhua). In Green Bay, Lake Michigan, burbot >400 mm fed primarily on fish; smaller burbot probably fed mostly on invertebrates and sculpins (Cottus sp.). Our calculations indicate that burbot of age ≥1 consumed 16 kg/ha of prey (12.2 kg/ha of fish) in 1988 in the Wisconsin waters of Green Bay including 3.3, 2.1, 1.9, 1.2, and 0.8 kg/ha of rainbow smelt (Osmerus mordax), sculpins, alewife (Alosa pseudoharengus), yellow perch (Perca flavescens), and bloater (Coregnus hoyi), respectively. On an areal basis, piscivory by burbot in Green Bay was higher than the reported lake-wide average for consumption by all salmonids in Lake Michigan. Burbot consumed about 25% of the lake-wide salmonid consumption of alewife per unit area and close to the estimated combined commercial and sport harvest of yellow perch in the Bay the same year (271 vs. 325 tons). Thus, burbot should be included when considering the balance between predatory demand and forage fish production in Green Bay and probably also in other areas of Lake Michigan.

We found no evidence that deflating the expanded swim bladder of burbot Lota lota affected survival. In a mark-recapture study of the species in Green Bay, Lake Michigan, many burbot caught from depths greater than 10 m floated at the surface when tagged and released because of expanded swim bladders. We deflated the swim bladders by forcing a hollow needle through the body wall to puncture the expanded organ, and we conducted two experiments to assess the effect of our procedures on survival. In one experiment 103 burbot were tagged and deflated and then held in nets on the bottom; 99% were alive after 1-8 d (mean, 4 d). In another experiment 316 burbot were tagged, deflated, and released at the surface, and 337 control burbot were tagged, not deflated, but forced to the bottom and released in a weighted net with an open bottom. Most of these fish that were recaptured were at large more than 5 months, and returns did not differ between the two groups; 6.3% of the deflated fish were recaptured compared to 5.9% of the controls. Recaptures of tagged and deflated burbot that were not part of these experiments indicated that swim bladder healing began within the first week of release and was complete in 8 weeks.