Stage structure must inform recovery in fish populations: a case for lake sturgeon in the Namakan River, Ontario

Abstract

Lake sturgeon are a large bodied, slow maturing, intermittently reproducing acipenserid native to central North America. The species is a conservation concern range-wide because of population declines during the 19th and 20th centuries related to overharvest and habitat degradation. It is generally accepted that harvest limitations alone are not enough to recover the species, but a stock-recruitment model accounting for the effect of the species’ complex, partially migratory life history has been elusive so far. In this dissertation I explore the age distribution of a population of adult lake sturgeon in a Canadian Shield river system for evidence of a density-dependent population bottleneck in the population’s early life history by fitting an ARIMA autoregressive time series model. I find that there is an autoregressive relationship, whereby the presence of lake sturgeon from one to four years old is associated with lower recruitment of lake sturgeon into this juvenile stage. A slight improvement in recruitment is associated with five-year-old fish in the system. I infer that density negatively affects lake sturgeon in a short early-juvenile period. I describe characteristics of nursery areas that might partially account for this effect. I then adapt a Lefkovich Matrix stock-recruitment model with density dependence expressed as a Ricker-modeled carrying capacity on year classes 0 to 5. The model is run with stochastic larval survival and vital rates specific to five life history stages of the lake sturgeon. In the model, simulated changes to juvenile carrying capacity drove recoveries and declines in the overall population within 25 years.