Unintended introductions of invasive species in both aquatic and terrestrial systems have increased in recent decades, resulting in growing economic burdens from product losses and management costs. Early eradication, or preemptive detection to impede establishment, would be a viable solution if not for the difficulty of detecting species when they are rare. New molecular approaches in aquatic systems have overcome this by utilizing environmental DNA (eDNA), biological material that individuals deposit as they move through their environment. Optimized eDNA approaches do not require capture of target species, thus allowing detection of early infestations when encounters through conventional detection methods are unlikely. To date, eDNA-based approaches have not been implemented in terrestrial systems, nor have they been compared with conventional surveillance techniques. Here, we adapt aquatic eDNA techniques for use in terrestrial systems, by developing an eDNA-based surveillance protocol for a globally invasive insect pest, the brown marmorated stink bug (BMSB – Halyomorpha halys). We test our protocol by directly comparing eDNA detection with conventional trapping methods at two agricultural sites, one in a known high abundance location (New Jersey) and one near the distribution’s northern edge (New Hampshire), and evaluating differences in sensitivity and detection effectiveness.
We collected 16 filtered samples from four peach trees in the New Jersey site over two visits, and 58 filtered samples from eight different crop plots in the New Hampshire site over two visits (four days per visit). Conventional traps (blacklight and pheromone) were set out and collected daily during each period. We found eDNA evidence of BMSB via a positive result in a high-resolution real-time PCR on all samples from New Jersey, and in numerous crop plots across all eight days in New Hampshire. Conversely, in New Hampshire we found only a single detection (one nymph) with pheromone traps on the last sampling day, and no detections on any days with the blacklight trap. Our work demonstrates that eDNA-based surveillance is effective in terrestrial systems, and is much more sensitive than conventional methods in detecting target species at low abundance. We conclude eDNA-based surveillance provides a high-resolution framework that can enable precise management actions when target pest populations are still very small, and thus more easily eradicated or controlled with minimal use of pesticides. While this strategy was implemented in agriculture, we foresee its application in a number of different settings with myriad target species of interest.