Bias-Elimination in Fish Population Models with Stochastic Variation in Survival of the Young

Abstract

The addition of random variation in survival to a single prereproductive age class in a Leslie matrix population model can alter population growth in the modeled system. Methods are presented to characterize the stochastic variation in survival and to determine a correction factor that, when included in the model, will eliminate the propensity for the modeled population to systematically deviate from the initial conditions. This ability is important for applications of fish population models that seek to evaluate probability distributions of population trends that are associated with stochastic variations in survival, and to forecast the effects of changes in fishing mortality rates in fluctuating environments. Received October 2, 1981 Accepted May 10, 1984