Assigning Phenologically Asynchronous Moths to Source Populations Using Individual Genotypes

The spruce budworm (Choristoneura fumiferana; SBW) is a periodically outbreaking forest insect pest that affects the boreal forests of North America through extensive defoliation and tree mortality. Causes of widespread spatial synchrony of SBW outbreaks remain a key question in the ecology and management of this species. While the Moran effect (correlated favourable environmental conditions) and density-dependent dispersal (from epicentres of demographic explosions) have been proposed and supported as drivers of synchronised outbreaks, the relative contribution of long-distance dispersal is still poorly understood. In this study, we use a novel approach to distinguish resident from migrant moths and to assign migrants to likely source clusters with the goal of better characterising regional dispersal. First, we characterise the genetic diversity and structure of resident SBW larvae and three phenologically separated groups of moths over one flight season using Genotyping-by-Sequencing. Then, using a novel machine learning approach, we assign putative migrants to their likely source populations. We hypothesised that migrant moths and resident larvae would be genetically distinct and could be assigned to source populations. Our findings revealed complex patterns of moth dispersal and population differentiation within a single season, including two spatially overlapping genetic clusters. We observed subtle but significant genetic differences between resident larvae and migrant moths, supporting the hypothesis that long-distance dispersal contributes to outbreak dynamics and synchrony. These insights enhance our understanding of SBW population dynamics and suggest that effective management strategies, such as the Early Intervention Strategy (EIS), must account for the role of dispersal in mitigating the detrimental effects of major outbreaks.