LAUSR

Abstract Wildlife managers need reliable information on species distributions (i.e. patterns of occurrence and abundance) to make effective decisions. Historically, managers have relied on harvest records (collected at broad spatial extents but coarse resolution) to monitor wildlife populations. However, emerging citizen-science datastreams can potentially supplement harvest-based monitoring by providing fine-resolution data that permit identification of species-environment relationships needed to predict occurrence and abundance. We combined harvest records and citizen-science camera-trap data in integrated species distribution models (iSDMs) to estimate species-environment relationships and distribution patterns of six wildlife species in Wisconsin, USA. We expected that iSDMs would more precisely estimate species-environment relationships and predict spatial abundance patterns intermediate between camera- and harvest-only SDMs. We also conducted simulations to explore the consequences of incomplete knowledge of harvest effort for estimates of abundance and species-environment relationships. Integrated models produced more precise species-environment relationships than camera-only models in 53% of the relationships we tested; all harvest-only models failed to converge. Moreover, integrated and camera-only models showed low agreement (mean: 19.67%) in identifying abundance “hotspots” but considerably higher agreement (mean: 45.17%) in identifying abundance “cold spots”. Our simulations showed that abundance patterns estimated by iSDMs may suffer from imprecision if harvest effort is poorly measured. We recommend that harvest records be collected at finer spatial resolutions and be paired with in-depth effort reporting. Our work demonstrates the potential for integrating an existing datastream (harvest records) with an emerging one (citizen-science camera-trap monitoring) for modeling species distributions and providing support for wildlife management decisions.