An empirical and expert‐knowledge hybrid approach to implement farmland habitat assessment for birds.

Bird species survival and breeding depend on necessary abiotic and biotic components of farmland habitat. A hybrid model based on expert knowledge using Analytic Network Process (ANP) integrated with empirical data is a capable approach for conservation scientists to determine the habitat suitability for bird species. Our study assessed the suitable distribution in the ranges of farmland habitats for bird richness throughout Taiwan, focusing on the regularly occurring birds on farmlands. We established a framework for analyzing potential habitat indicators developed by expert interviews. Then, ANP was employed to compute the weight‐based priorities. By combining habitat indicators, we assessed and presented the spatial distribution of habitat suitability. The results show that southern and eastern regions of Taiwan with high habitat suitability scores cover 60,576.51 ha (27.98%) and 40,531.42 ha (30.09%), respectively. The model has an overall accuracy of 70.40%, showing a capable and reliable approach to assessing habitat suitability for bird richness on farmland. Our study highlights a valuable tool for the policy‐makers through exploring suitable farmland habitat spots, where further conservative practices and prudent biodiversity management should be applied. [ABSTRACT FROM AUTHOR]