Comparing sampling methods to monitor population abundance while accounting for imperfect detection: An application of N-mixture models on Orthoptera

Abstract

Human activities are profoundly altering natural ecosystems, impacting both the distribution and abundance of animal populations. Monitoring is the base for assessing the status and temporal changes in animal populations. Among the various considerations involved in establishing a monitoring program, the choice of sampling methods can significantly influence the project's outcomes. In this study, we compared two sampling methods to monitor Orthoptera communities: sweep netting and tube sampling (a modified version of box quadrats). Imperfect detection can strongly bias the outcome of statistical models. Therefore, to compare the two sampling methods, we performed N-mixture models to take into account detection probability and uncertainty in abundance estimates. Between June and September 2023, we sampled Orthoptera communities across 25 sites in northern Italy, including a broad variety of grasslands, which are ecosystems holding high ecological value in agricultural landscapes. We detected a total of 35 species, 33 through sweep netting and 30 through tube sampling. N-mixture models were performed on the 14 most abundant species detected with both sampling methods. While detection probability was similar between the two sampling methods, the precision of detection estimates was markedly higher when using sweep netting. Additionally, abundance estimates were generally higher and showed less uncertainty when using sweep netting instead of tube sampling. These results suggest that, in our study system, sweep netting seems to be the best method to monitor Orthoptera communities. Still, the limited area sampled with tube used here (2–4 m² per survey in each plot) possibly influenced the precision of abundance estimates, and larger sampled areas may provide more accurate measures. Our framework to compare sampling methods can be applied to a broad range of organisms. It can help in deciding which method is better to fulfill the aims of multi-season monitoring programs by conducting pilot surveys at a small set of sites.