Using empirical data analysis and expert opinion to identify farmland-associated bird species from their habitat associations

Agricultural expansion is a pervasive threat to biodiversity, and intensification of farming activities can reduce the diversity and abundance of farmland-associated species. Tracking these changes to monitor and manage the biodiversity impacts of farming requires identification of farmland-associated species - a well-established category in Europe and North America, but little-documented elsewhere. Here we develop an integrated approach using empirical data analysis and expert opinion to delineate farmland-associated bird species in Taiwan. We investigated the relationship between land use variables and the abundance of 129 bird species from the Taiwan Breeding Bird Survey. We also administered a questionnaire to 24 expert birdwatchers as an alternative method for estimating the habitat associations of the 129 bird species. The classifications of 104 species into habitat association classes using the two methods were well aligned, with 75 species (72.1%) classified consistently. Only two species (1.9%) were discordantly classified. We could not confidently assign 25 species to any category through empirical data analysis, but expert opinion allowed a qualitative identification of their habitat associations. The two different approaches enabled us to identify farmland-associated species in a consistent way, increasing confidence that where empirical data were insufficient, expert opinion might suffice. Identifying Taiwan's farmland-associated species using expert opinion, validated by the empirical analysis, paves the way for exploring the status of the group, how agricultural intensification affects it, and the effectiveness of conservation interventions in rapidly changing agricultural landscapes.