Cooperative Agreement for affiliated Partner with the Great Rivers Cooperative Ecosystem Studies Unit

The U.S. Geological Survey (USGS) invites proposals from Cooperative Ecosystem Studies Unit (CESU) Great Rivers network partners for novel research focusing on the development and application of Bayesian statistical methods. This funding opportunity seeks to advance methodologies for analyzing passive bat acoustic monitoring data, specifically for estimating species occurrence and relative abundance at un-surveyed locations and assessing their associations with meaningful environmental predictors. Proposed methods should consider accounting for spatial correlation and integrating of multiple observer information to address the misclassification process inherent in acoustic data.
Passive acoustic monitoring represents a cost-effective and technologically advanced approach for collecting extensive biological observations. The primary scientific challenge lies in translating these massive datasets into dynamic and interactive maps of species occurrence and relative abundance. Enhancing the ability to generate these maps directly supports land management and state agency conservation and restoration efforts, aligning with the Department of the Interior's (DOI) strategic goals to restore lands and waters, protect species, and provide reliable scientific information through collaborative partnerships (SO 3356: Shooting, and Wildlife Conservation Opportunities and Coordination with States, Tribes, and Territories (September 15, 2017)). These efforts will directly support EO 14303 "restoring gold standard science" because fundamental to science is using statistical models that account for all sources of uncertainty when drawing conclusions from data.
This research will contribute significantly to monitoring the distributions and activity rates of the 47 bat species across North America. Given the estimated $3.7 to $53 billion/year in savings bats provide as agricultural pest eaters [Boyles et al. 2011], accurate monitoring is crucial. While current count detection models developed by the USGS account for species classification errors and estimate species occurrence and relative activity [Wright et al., 2020], they currently do not explicitly model spatial dependence among observations. Incorporating spatial processes in these models is known to improve predictive accuracy for species distributions [Wright et al., 2021] and will enable a more complete understanding of underlying ecological patterns such as animal movement and dispersal [Levin, 1992].
It is expected that the developed methods will lead to more robust and accurate predictions of bat species distributions, directly informing critical conservation and resource management decisions.