QUANTIFYING DISPERSAL OF SOUTHERN PINE BEETLES WITH MARK RECAPTURE EXPERIMENTS AND A DIFFUSION-MODEL

Pest management decisions should take into consideration quantitative information on dispersal of insect pests, but such information is ofte n lacking. The goal of this study was to measure intraforest dispersal in the southern pine beetle (SPB). We developed an analytical formula for interpreting data from mark-recapture studies of insect dispersal . The formula is obtained by postulating a simple model of diffusion w ith disappearance (e.g., as a result of death) for the spread of marke d insects from the release point. Parameters of the model (assumed to be invariant in space) are estimated by fitting the curve to the cumul ative number of recaptured insects as a function of the distance from release. We also derived an expression for the distribution of dispers al distances in terms of the fitted parameters. The proposed methodolo gy was applied to a mark-recapture study of SPB dispersal. Statistical analysis of recaptures-with-distance curves obtained in 11 replicate releases indicated that the proposed formula provided an accurate desc ription of the data. There were no systematic departures from the func tional relationship prescribed by the formula, and the model consisten tly outperformed another commonly used formula for fitting data on dis persal distances, the exponential curve. We explored the effect of spa tial heterogeneity in the host distribution on SPB movement by regress ing the deviation from the recapture rate predicted by the model in ea ch trap on the pine basal area around the trap. This correlation was s ignificantly greater than zero, indicating that beetles tended to aggr egate in localities where pines were dense. This result suggests that a diffusion model with spatially varying parameters may provide a more accurate description of the redistribution process in the SPB. Quanti tative results on SPB intraforest dispersal were summarized by calcula ting radii of circles enclosing a given proportion of SPB dispersal di stances. For example, we estimated that one half of released beetles d ispersed > 0.69 km. This result has important implications for evaluat ing the area-wide consequences of current or novel control tactics tha t rely on disrupting SPB movements, e.g., cut-and-leave and treatments with anti-congregation pheromone.