Through the management of an estimated 8.8 million acres of land across the United States (U.S.), the U.S. Military is tasked with the surveillance of numerous species. Among arthropods, these include many threatened or endangered species of conservation concern, such as Taylor’s Checkerspot Butterfly, the Karner Blue Butterfly and the Rusty-Patched Bumble Bee, as well as invasive species, such as the Spotted Lanternfly. This management responsibility will only expand with continued species invasions and the near-term prospect of federal listing under the Endangered Species Act for candidates such as the Frosted Elfin, Regal Fritillary, Yellow-Banded Bumble Bee and American Bumble Bee. As a result, there is a vital need for advancements in cost-effective high-throughput biosurveillance methods to be used on military lands. Recent advances in metagenetics, including eDNA-based detection of floral-associated arthropods, represents a promising avenue for the development of robust and cost-effective means to meet this need. This project plans to optimize and demonstrate the capabilities of eDNA-based biosurveillance of floral-associated arthropods on Department of Defense lands. Further, the project team will evaluate relationships between sampling methods, sampling intensity, and detection probability for multiple endangered species in order to inform future surveys of U.S. Military installations.

Targeted next generation sequencing of metagenetic markers amplified from flower, pollen and leaf surface eDNA samples will be used to survey floral-associated arthropod communities. Metagenetic methods will be optimized for high-confidence detection of a broad diversity of threatened and endangered floral-associated arthropods prior to field deployment. Novel direct polymerase chain reaction (PCR) methods will be developed to circumvent the need for costly DNA extraction techniques that increase cross-contamination risks and decrease assay sensitivity through loss of trace eDNA across processing stages. Loss of sensitivity through off-target PCR amplification, an issue inherent to current eDNA methods, will be minimized using universal Arthropoda and Anthophila primer sets designed for selectivity against fungal and bacterial byproducts. Multiple genetic loci will be targeted to enable broad community-level characterization and greater detection confidence for species identifiable with multiple loci. A novel software for taxonomic classification of DNA sequences, based on generalized linear mixed modeling, will be further refined and used to obtain high accuracy estimates of probabilistic classification confidence during eDNA analysis. Lastly, metagenetic surveys will be implemented within a Bayesian multiscale occupancy modelling framework to ensure that surveys possess adequate statistical power for target species detection, thereby limiting the risk of false negative results.

Accurate and precise biosurveillance of threatened or endangered species requires the minimization of both false positive and false negative detection rates. To that end, it is important for methods development procedures to both maximize assay sensitivity and quantify statistical detection power across sampling methods and intensities. This project entails optimization of metagenetic eDNA analysis across nearly all implementation steps, from sample collection to bioinformatic analysis, and is integrated within an occupancy modelling framework to aid in the determination of the statistical power of eDNA surveillance. With the ultimate goal of producing an effective survey tool where sampling can be conducted quickly and by non-expert practitioners, this project exhibits great potential for increasing cost-efficiency, speed and scalability of eDNA metagenetic methods targeted toward threatened and endangered pollinator species.