Evaluating approaches for integrating species distributions in spatial conservation planning

Abstract

Map-based decision support tools (DSTs) that use species distributions are an important means of identifying priority areas for conservation. The 2020 Wisconsin Waterfowl Habitat Conservation Strategy (WWHCS) uses a DST to identify priority ecological landscapes and watersheds to guide waterfowl habitat projects. The WWHCS DST relies on waterfowl habitat suitability layers derived through expert opinion in lieu of species distributions, a common approach in DSTs. Given the inherent subjectivity of expert opinion, data-driven species distributions such as those available from citizen science projects, could provide more accurate information and better identify areas important for waterfowl conservation. Here, we explore the application of relative abundance products available through the eBird Status and Trends project as an alternative to expert-derived habitat suitability layers in the WWHCS DST. Our objectives were to compare seasonal species distributions derived from habitat suitability models (expert-derived) and species distribution models (eBird-derived) and determine whether differences influenced DST prioritizations. Correlations between expert- and eBird-derived distributions were generally low to moderate for the breeding and fall layers (ρ: -0.03–0.76), and least for the spring (ρ: -0.49–0.72). There was minimal agreement among top-ranked ecological landscapes (40%) and watersheds (28%) between the expert- and eBird-derived versions of the DST. Overall, our results suggest the DST may benefit from incorporating data-driven species distributions. However, additional work validating eBird relative abundance against professional surveys (e.g., aerial counts) and empirical studies evaluating waterfowl habitat selection and vital rates are important future considerations for the WWHCS DST and waterfowl habitat conservation in Wisconsin.