Estimating Critical Abundance Thresholds in Exploited Populations: A Simulation Approach Based on Species Resilience to Disturbance
Managers of exploited species too often assume that populations can be maintained at equilibrium abundances that will provide maximum yield. Most evidence to date suggests that populations seldom adhere to equilibria, but rather fluctuate stochastically between bounds. The last decade has revealed the consequences of not incorporating uncertainty around point estimates of equilibria, which has led to the decline of several fisheries. We used the sample importance re-sampling (SIR) algorithm to exhibit the uncertainties in point estimates generated by models for management of two Chinook salmon Oncorhynchus tshawytscha stocks and a bowhead whale Balaena mysticetus population. We then incorporated the cumulative uncertainties of each system into a simulation technique similar to population viability analyses (PVA) to provide decision support for establishing threshold abundances of each exploited population. The simulation presented was based upon the resilience (time to recover from perturbations to abundance) of each population, which was found to be relatively high for the Chinook stocks and low for bowhead whale. Various thresholds could be chosen depending on: (1) how much time should be allowed for the population to recover from a perturbation, (2) when should the stock be considered recovered (i.e., within 1%, 5%, 10%, and so on of what abundance would be had there been no perturbation), and (3) the maximum allowable risk that a threshold is too low. Reasonable thresholds for the Chinook stocks were 60% to 80% of abundances that provide maximum sustained yield (SMSY). Due to their low productivity, no clear threshold below the biomass point estimate was apparent for bowhead whale.
Rishi Sharma and Scott W. Raborn
Sharma, R., and S. Raborn. 2011. Estimating critical abundance thresholds in exploited populations: a simulation approach based on species resilience to disturbance. Computational Ecology and Software 1:189-207.
Dec 1st, 2011