Abstract: Long-term experiments are important for ecological monitoring, but come with unique challenges. Due to the length of time they span, a project can see changes of personnel or funding, and even changes to sampling protocols due to availability of materials. Yet, some measurements are not robust to even slight changes, for instance insect monitoring is notoriously sensitive to changes in protocol. At the Kellogg Biological Station Long-Term Ecological Research Site (KBS LTER), one of many NSF-funded long-term research stations, an over 30-year old project recently experienced a supply-chain-induced change in insect sampling methodology. Since 1989, lady beetles have been sampled weekly over the growing season in the Main Cropping System Experiment using yellow sticky cards. In 2021 the supplier discontinued the traps. While a similar trap was identified to replace the former standard, there were noticeable differences in trap hue and glue composition. In 2021 we conducted a field study which paired the legacy traps with the new model to examine if this change would impact the observed biodiversity patterns at the site, while otherwise maintaining identical protocol. The study was conducted with standard protocol in 2020 using entirely old traps and in 2022 using entirely new traps. We completed whole community and individual taxa analyzes to examine within-year differences in performance between the trap types, adding additional taxa identified in the lab to increase comparisons between the two traps. Identical analyses to assess typical between-year variation in patterns for each trap type separately were conducted. We found no difference in the overall insect community metrics between the two trap types in 2021, nor did we find differences in the insect community on the same sticky card types between years, once variation by plant treatment was accounted for. However, some statistical differences between the trap types were observed when data was examined by taxon, particularly for rare species. However, the magnitude of difference was less than the difference observed between sampling years for that same taxa using the same trap design. Thus we attribute these apparent differences in trap performance for certain taxa to chance: with low sample sizes, rare events are often ‘significant’ even when population-level effects are minimal. However, if sustained changes in the response of rare taxa are observed, future scientists may use our observations to specifically quantify and correct for these discrepancies related to the protocol change.
