Natural Mortality in Relation to Age and Fishing Mortality on Walleye in Escanaba Lake, Wisconsin, During 1956–2009

To evaluate the dynamics of fishing and natural mortality, we used a catch-age cohort analysis model of walleye population dynamics in Escanaba Lake, Wisconsin, during 1956–2009.  The catch-age model was based on angling catches at age from a compulsory creel census, abundance at age from mark-recapture studies, and recruitment at age-0 from mark-recapture studies.  Harvest of age-3-and-older (age-3+) walleye averaged 12.5 fish/ha and varied 55% among years during 1956–2002, prior to imposition of a 71-cm (28-in) minimum-length limit that eliminated harvest.  Population density of age-3+ walleye averaged 34.8 fish/ha and varied 51% among years during 1956–2009.  Annual exploitation on age-3+ walleye averaged 34% during 1956–2002 and varied 35% among years during 1956–2002.  Variation in population density explained only 0.5% of the variation in annual exploitation during 1956–2002.  Natural mortality was best described by a model with decreasing age-specific vulnerability and a negative relationship with angling exploitation.  Relative vulnerability to natural mortality declined gradually with age from 100% at age-0 to 83% at age-3, 50% at age-9.4 and 24% at age-17.  Relative vulnerability to fishing mortality increased sharply from <1% at age 0 to 98% at age 3, peaked at 100% at age 4, and declined gradually to 66% at age 17.  Fully-vulnerable natural mortality (M) declined from 0.70 to 0.10 (mean = 0.41) as annual angling exploitation (u) on age-3+ walleye increased from 0.0 to 0.65 (mean = 0.30).  Our findings suggest that natural mortality declines with age and is inversely related to fishing mortality, so varies among ages and years.  We conclude that stock-assessment models for walleye fisheries should evaluate the effects of year-specific and age-specific natural mortality on model predictions.