Comparing pest survey methods through detection probability: seasonal and abundance-dependent detection of spotted lanternfly (Lycorma delicatula, Hemiptera: Fulgoridae)

BACKGROUND: Invasive species management demands accurate information on the distribution of targeted pests, but surveys have imperfect detection. Understanding the factors influencing detection probability can guide improved management planning. For the invasive spotted lanternfly (Lycorma delicatula), several sampling methods have been used. We evaluated three widely applied methods: 5-minute timed visual search, BugBarrier sticky band traps, and modified circle traps. These methods were used to quantify abundance in 5 x 20 m plots spanning the core to margins of five outlying satellite populations in Pennsylvania U.S.A. Methods were assessed every two weeks over the growing season for two years. RESULTS: Using a hierarchical Bayesian model, we found seasonal variation in all methods' detection probabilities, with timed visual search increasing in sensitivity over the course of the season and both tree-based traps showing a U-shaped pattern, with lowest detection probability around midseason. For timed visual search in plots used by ten L. delicatula individuals, the estimated probability of detecting at least one individual ranged from 10% to 53%. For tree-based traps, detection was higher on larger trees and on Ailanthus altissima, a preferred hostplant. For traps installed on A. altissima with diameter at breast height of 20 cm in a plot with the same L. delicatula density (N=10), estimated mean detection probability ranged from 33% to 99% for BugBarrier and from 28% to 84% for circle traps. CONCLUSION: Informed with estimated detection probabilities, managers can compare alternative survey methods and design strategic management plans that properly incorporate realistic survey sensitivity.