Finding the needle in the haystack: iterative sampling and modeling for rare taxa

Much like finding a needle in a haystack, the effort required to detect a rare and endangered species increases inversely with limited taxa distribution. The infrequency of detections combined with limited fiscal resources often leaves scientists with knowledge gaps about the ecological niche and habitat requirements necessary for conserving rare species. The Arsapnia arapahoe snowfly ( A. arapahoe ), was thought to be a rare and cryptic aquatic invertebrate for which only 13 individuals from two locations were known to exist in Colorado. In response to potential listing by the US Fish and Wildlife Service as a threatened species, we sought to implement an improved sampling protocol and tested an iterative predictive modeling approach. Species distribution models successively employed annual presence data collected from 2015 to 2017 and detections improved. Although now understood to be a hybrid taxa, the model predicted the locations of seven additional localities while concurrently narrowing the search area and expanding the known geographic range of A. arapahoe . Given our results, we recommend an iterative species distribution modeling and sampling strategy to refine search areas and improve detection rates for rare and endangered species.