Designing multi‐scale hierarchical monitoring frameworks for wildlife to support management: a sage‐grouse case study

Abstract Population monitoring is integral to the conservation and management of wildlife; yet, analyses of population demographic data rarely consider processes occurring across spatial scales, potentially limiting the effectiveness of adaptive management. Therefore, we developed a method to identify hierarchical levels of organization (i.e., populations) to define multiple spatial scales, specifically intended to help guide appropriate conservation and management actions. This approach can support mobile species with high site fidelity where surveys occur on birthing/breeding grounds or migratory stopovers. Our approach used a graph‐based clustering algorithm (Spatial K'luster Analysis by Tree Edge Removal) that explicitly included habitat selection information at multiple scales and further refined with constraint‐based rules. We applied these concepts to greater sage‐grouse leks (breeding grounds), a species of conservation concern, in two different ecological contexts (Nevada and Wyoming, USA ). The constraint‐based rules accounted for inter‐lek movement distances based on literature and field studies in Nevada from 2012 to 2016, included methods to support a spatially balanced monitoring design, and identified barriers to movements among leks based on resistance surfaces. We evaluated the performance of our hierarchical clusters in Nevada using independent data from radio‐marked sage‐grouse, and we found the finest‐scaled cluster level captured ~90% of sage‐grouse movements and mid‐level scales captured ~97 – 99% of movements. We expected comparable performance for Wyoming, where we lacked radio‐marked sage‐grouse for an evaluation, because genetic studies estimate similar dispersal distances to our ~15 km inter‐lek movement distance in Nevada. For sage‐grouse and other mobile species with high site fidelity, our approach to defining these frameworks could prove valuable for conservation and management applications, such as improving estimation of scale‐dependent population trends and guiding the prescription of management actions at spatial scales that align with identified threats. Specific to sage‐grouse, our analysis sets the stage for designing a monitoring framework that relies on comparison of short‐ and long‐term population trends across our defined spatial scales and identifies and disentangles factors driving local (e.g., habitat quality) and regional (e.g., climate) population changes, thereby supporting scale‐dependent management and research needs for adaptive management practices.