Nonnative pest establishment: Spatial patterns and public detection

Abstract

Early detection of new pests can reduce their long-term impacts by enabling more rapid management response. Knowledge of pest establishment risk and background rates of detection (e.g., by the public) can help inform more cost-effective targeting of formal early detection survey programs. Here we quantify county-level locational attributes associated with pest establishment and detection by the public using data for 113 new pest incursions detected in the United States from 2010 through 2018. Aligning with expectations, we find a higher likelihood of new pest establishment in counties with higher human population numbers, nearer to ports (<250 km), and with amenable climate characteristics. Controlling for potential sample selection issues, we find that pests are less likely to be first detected by the public (e.g., homeowners, community members) versus by other sources (e.g., agency surveys, researchers, or agricultural operators) in counties with higher total crop sales values and lower human population number. The negative association between public detection and high agricultural values may reflect greater survey efforts by other sources (e.g., by agency surveillance programs, researchers, and agricultural operators) in high-value agricultural areas. The positive association between public detection and human population size may reflect larger numbers of public detectors (i.e., people) available to encounter the pests. Our models provide spatially explicit estimates of the likelihood of new pest establishment across U.S. counties and of the likelihood that an established pest would first be detected by the public. These estimates can serve as quantitative inputs to decision-support activities for new pest surveillance planning.