Retrospective analysis of a classical biological control programme

1. Classical biological control has been a key technology in the management of invasive arthropod pests globally for over 120 years, yet rigorous quantitative evaluations of programme success or failure are rare. Here, I used life table and matrix model analyses, and life table response experiments to quantitatively assess a classical biological control programme for an invasive insect pest in the western United States. 2. Life tables and matrix models were developed for populations of Bemisia tabaci (sweetpotato whitefly) on cotton in Arizona before (1997-1999) and after (2001-2010) the permanent establishment of two exotic aphelinid parasitoids. Analyses tested multiple hypotheses relative to the expected outcome of a successful programme. 3. Marginal rates of parasitism, rates of irreplaceable mortality from parasitism, total generational mortality and finite population growth (lambda) were unchanged for B. tabaci populations before and after exotic parasitoid establishment. Prospective analyses showed that predation during the final nymphal stadium had the greatest influence on population growth rates regardless of parasitoid establishment. Retrospective LTREs showed that predation and unknown mortality contributed most to changes in lambda after parasitoid establishment. 4. Marginal parasitism acted weakly in a direct density dependence fashion after parasitoid establishment, and for all 14 years combined. However, this did not translate into an association between pest population density and marginal rates of parasitism for the 10-year period following establishment.