Abstract

Conservation initiatives strive for reliable and cost-effective species monitoring.1,2,3 However, resource constraints mean management decisions are overly reliant on data derived from single methodologies, resulting in taxonomic or geographic biases.4 We introduce a data integration framework to optimize species monitoring in terms of spatial representation, the reliability of biodiversity metrics, and the cost of implementation, focusing on tigers and their principal prey (sambar deer and wild pigs). We combined information from unstructured ranger patrols, systematic sign transects, and camera traps in Sumatra's largest remaining tropical forest and used integrated community occupancy models to analyze this multifaceted dataset in a unified way. Data integration improved the precision of species occupancy estimates by 14%-42%, enhanced the accuracy of species inferences, expanded the spatial scope of inference to the landscape level, and cut operational costs up to 51-fold. Our framework demonstrates the underappreciated value of integrating unstructured observations with monitoring data derived from traditional wildlife surveys.