Detecting range shifts from historical species occurrences: new perspectives on old data

Abstract

The difficulty of making valid comparisons between historical and contemporary data is an obstacle to documenting range change in relation to environmental modifications. Recent statistical advances use occupancy modeling to estimate simultaneously the probability of detection and the probability of occupancy, and enable unbiased comparisons between historical and modern data; however, they require repeated surveys at the same locations within a time period. We present two models for explicitly comparing occupancy between historical and modern eras, and discuss methods to measure range change. We suggest that keepers of historical data have crucial roles in curating and aiding accessibility to data, and we recommend that collectors of contemporary specimen data organize their sampling efforts to include repeated surveys to estimate detection probabilities.