Estimating mortality of Lake Sturgeon in the Lake Winnebago System using traditional age-based approaches and capture-recapture models
College of Natural Resources, University of Wisconsin-Stevens Point
MetadataShow full item record
The lake sturgeon Acipenser fulvescens population in the Lake Winnebago System (LWS) supports a culturally and economically important spear fishery. Harvest from the spear fishery is closely monitored and managed by the Wisconsin Department of Natural Resources (WDNR) through use of a safe harvest cap system with a 5% annual exploitation limit reference point. Harvest caps for all males and adult females are based on mark-recapture estimates of population abundance that are adjusted for natural mortality. The current instantaneous natural mortality rate (M = 0.055) used to determine harvest caps was estimated using a statistical catch at age (SCAA) model formulated using corrected ages estimated from fin rays. Correction is required because fin ray age estimates are not accurate for older fish. The WDNR marks lake sturgeon captured in annual assessments with passive integrated transponders (PITs). Recapture of fish with PITs during spawning sampling and spearing harvest provides information that can be used to estimate apparent survival and mortality rates using capture-recapture models as an alternative to traditional age-based approaches. Additionally, the WDNR has implanted acoustic transmitters into fish that were detected on receivers located throughout the LWS and its tributaries. The primary objectives of my research were to determine if: 1) total and natural mortality rates are similar among estimation methods that rely on corrected fin ray ages or capture-recapture methods and 2) potential differences in mortality rate estimates would affect safe harvest caps for the spear fishery. Three separate catch curve methods, along with a suite of eleven indirect natural mortality estimators, were used with each age correction method to estimate total mortality for fish handled from 2009 to 2019. Cormack-Jolly-Seber models were created using the encounter histories of fish tagged with PITs from 1999-2020 and for fish with acoustic transmitters detected from 2004 to 2019. Mortality estimates from these models were used to simulate harvest caps from 2009 to 2020 for males and females. Estimates of total and natural mortality rates varied among catch curve and capture recapture models (male: Z = 0.100 – 0.207, M = 0.054 - 0.226; female: Z = 0.075 - 0.201, M = 0.050 – 0.191). Cormack-Jolly-Seber models, when built separately for both PITs and acoustic telemetry, converged on similar estimates of mortality for each sex and appear to be the most effective way to estimate mortality for lake sturgeon in the LWS. A simulation of harvest caps over an eleven-year period using estimates of M obtained from a suite of age-based and capture-recapture models showed that female harvest could exceed an exploitation rate of 5% in some scenarios. Use of capture-recapture methods for estimating mortality appear to provide the best available approach for describing population dynamics and demographics for the LWS lake sturgeon population.
Natural Resource Management