Characterizing eastern spruce budworm’s large-scale dispersal events through flight behavior and stable isotope analyses

Abstract

Eastern spruce budworm moth (Choristoneura fumiferana (Clem.)) mass outbreaks have widespread economic and ecological consequences. A key explanation for the large-scale spread and synchronization of these outbreaks is the long-distance dispersal (up to 450 km) of moths from hotspots (high-density populations) to lower-density areas. These events have proved difficult to monitor because dispersal flights occur only a few times a year, have no consistent routes, and commonly used tracking methods (e.g., population genetics, mark-recapture, radio telemetry) are inadequate for this system. Confirming immigration and distinguishing between local and immigrant individuals are crucial steps in identifying the physical and ecological drivers of moth dispersal. Here, we test whether isotopes of hydrogen (i.e., delta notation: δ2H) and strontium (i.e., strontium isotope ratios: 87Sr/86Sr), known to independently vary in space in a predictable manner, can be used to show that an immigration event occurred and to distinguish between local and immigrant adult spruce budworm moths. We used an automated pheromone trap system to collect individuals at six different sites in eastern Canada within and outside the current outbreak area of budworm moths. We first use moth flight behavior and time of capture, currently the best available tool, to determine putative local vs. immigrant status, and then evaluate whether individual 87Sr/86Sr and δ2H differ between putative classes. At two sites, we detect immigrant individuals that differ significantly from putative locals and thus confirm immigration has occurred. Saliently, sites where putative locals were sampled before the occurrence of potential immigration events (~10 days) showed the strongest differences between immigrant individuals’ and local 87Sr/86Sr and δ2H values. Sites where the collection of putative locals was close in time (hours) or following an immigration event (days) had a less-clear distinction between putative immigrants and locals, and showed signs of mixing between these two groups. We speculate that recent immigration could have led to the misclassification of immigrants as putative locals. 87Sr/86Sr and δ2H data generally support the adequacy of current approaches using capture-time to detect immigration events, and provide enhanced resolution to distinguish between local and immigrant individuals and to confirm an immigration event. We discuss the broader implication of adding isotopes to the toolkit to monitor spruce budworm dispersal and suggest next steps in implementing these tools.