Abstract
Caterpillars Count! is a citizen science project that allows participants to collect data on the seasonal timing, or phenology, of foliage arthropods that are important food resources for forest birds. This project has the potential to address questions about the impacts of climate change on birds over biogeographic scales. Here, we provide a description of the project’s two survey protocols, evaluate the impact of survey methodology on results, compare findings made by citizen scientist participants versus trained scientists, and identify the minimum levels of sampling frequency and intensity needed to accurately capture phenological dynamics. We find that beat sheet surveys and visual surveys yield similar relative and absolute density estimates of different arthropod groups, with beat sheet surveys recording a higher frequency of beetles and visual surveys recording a higher frequency of flies. Citizen scientists generated density estimates within 6% of estimates obtained by trained scientists regardless of survey method. However, patterns of phenology were more consistent between citizen scientists and trained scientists when using beat sheet surveys than visual surveys. By subsampling our survey data, we found that conducting 30 foliage surveys on a weekly basis led to 95% of peak caterpillar date estimates to fall within one week of the “true” peak. We demonstrate the utility of Caterpillars Count! for generating a valuable dataset for ecological research, and call for future studies to evaluate how training and resource materials impact data quality and participant learning gains.