Quantifying long-distance dispersal of an outbreaking insect species using trap capture data and phenology

Abstract

Long-distance dispersal plays an important role in insect population dynamics and is a key factor in insect outbreaks. However, accurately measuring the dispersal distance of insect species can be challenging. We present a simple and reliable method to quantify dispersal capacity using insect phenology and time-series of field-capture data. This open-source method is developed in R and tested using data collected on the eastern spruce budworm (Choristoneura fumiferana). Empirical data were collected in the summer of 2014 in Québec, in forest areas defoliated during the ongoing outbreak. Sampling dates were informed by BioSIM phenological models and government of Canada weather data where adult moths were sampled before, during, and after expected flying periods. A sensitivity analysis showed that the uncertainty of these models due to weather station interpolation did not affect general trends. Individuals sampled outside of their expected phenological window were assumed to be migrants. Pairwise Euclidean distances were measured between the sampling locations of these migrants and all potential areas of origin as predicted by BioSIM at the time of their sampling. The shortest distances were then selected to generate dispersal kernels. According to the dispersal kernels derived using this method, most flying individuals captured outside of their phenological window travelled up to 450 km, matching previous work on eastern spruce budworm dispersal. Here, we were able to demonstrate that with known phenology of a species and relatively simple to collect data (presence of flight capable individuals), it is possible to characterize insect dispersal using a relatively simple algorithm and spatial model. This method can be easily applied to other outbreaking species and used to inform management decisions in a variety of systems where outbreaks represent a risk to resources of interest.