Using pilot data to inform the design of a monitoring program for Bombus affinis

Effective species monitoring requires a sampling design that accounts for imperfect detection and spatial variation to support occupancy modeling across the species’ range. These models can guide and respond to the objectives of a monitoring program and recovery plan designed for an endangered species like the rusty patched bumble bee (Bombus affinis). We used data obtained in the field season of 2024 from May to October across 138 sites in Iowa, Illinois, Minnesota, and Wisconsin to evaluate survey effort, temporal patterns in detectability, and other key environmental covariates. Using these fields collected data, we conducted simulations to evaluate how alternative monitoring strategies influence estimates of local extinction, colonization, and occupancy within different sampling unit strata. We found that B. affinis was detected at 32% of surveyed sites, and occupancy modeling that accounted for imperfect detection estimated a true site occupancy of 47% (95% CI: 0.36–0.59). Detection probability varied with survey date, indicating strong seasonal effects. These findings showed the importance of survey timing and replication in estimating occupancy trends. Although temperature and humidity data were incomplete or inconsistent across sites, incorporating these covariates could be helpful in understanding when to detect the species. Our pilot analysis provides guidance for optimizing survey frequency, temporal survey, and environmental data collection, forming the foundation for a standardized, multi-state monitoring framework for B. affinis conservation and recovery plan.