Standard deviation: Standardized bat monitoring techniques work better in some ecosystems

Standardized monitoring strategies are often used to study spatial and temporal ecological patterns and trends. Such approaches are applied for many study taxa, including bats (Mammalia, Chiroptera). However, local characteristics of individual field sites, including species assemblages, terrain, climatic factors, and presence or lack of landscape features, may affect the efficacy of these standardized surveys. In this paper, we completed mist-netting surveys for bats in two widely separated field sites, Calakmul Biosphere Reserve (CBR), a Mexican lowland tropical forest, and Krka National Park (KNP), a Mediterranean dry scrub forest in Croatia. Standardized surveys were conducted along predefined transects for six hours. We also completed targeted surveys in KNP that focused on the key bat activity period (the first two to three hours after sunset), with nets being deployed at sites of known or assumed value to bats (independent of predefined transects). We analyzed how survey success differed in standardized surveys between CBR and KNP and between standardized and targeted surveys in KNP. Survey success was measured through three parameters: capture rate = the number of individual bats captured per net hour, inventory rate = the number of unique bat species recorded per net hour, and inventory efficacy = the percentage of known species assemblage recorded per net hour across all surveys. Results for all three parameters indicate that standardized surveys in CBR were vastly more effective than those in KNP (e.g., mist-netting in CBR detected 69.8% of the species assemblage, compared to just 8.3% in KNP), and it was only by employing targeted mist-netting in KNP that meaningful capture rates could be achieved. This study contributes further evidence to discussions around how and when standardized survey methods should be employed, and the alternative approaches that can be taken in ecosystems where generally effective methods underperform.