Matters Arising: Response to Crossley et al. 2020: Meta-analysis of Insect Temporal Trends Must Account for the Complex Sampling Histories Inherent to Many Long-term Monitoring Efforts

Abstract

In an article recently published in Nature Ecology & Evolution (Crossley et al. 2020 “No net insect abundance and diversity declines across US Long Term Ecological Research sites”), sampling effort within Long-Term Ecological Research (LTER) datasets was assumed to be consistent across years. Given the complex history of many long-term datasets at LTER sites, this assumption does not often hold and we believe this assumption led to errors in Crossley et al.’s analysis. Here we first use the Konza Prairie grasshopper dataset as an example of how changes in sampling locations and effort can cause errors when data are assumed to be collected with invariant sampling. Second, we describe similar and additional errors in data use from 7 of the 13 LTER sites included in Crossley et al. (2020)’s analysis.