Quantifying Changes in British Columbia Dungeness Crab (Cancer magister) Landings Using Intervention Analysis

Natural or man-induced interventions may cause temporary or permanent changes in the behaviour of biological systems. To determine the impact of future management decisions, it is first necessary to quantify the effects of past interventions on the system. Only then can rational decisions be made so as to ensure that the desired response is achieved. However, time series data are often autocorrelated and this precludes the use of standard statistical tests. The linear stochastic intervention model outlined in this paper takes this autocorrelation into account and provides a procedure for quantifying the impacts of particular interventions. This model is employed to determine if there has been a significant abrupt decrease in Dungeness crab (Cancer magister) landings in British Columbia.