A state-of-the-art review on birds as indicators of biodiversity: Advances, challenges, and future directions

The current loss of biodiversity has been broadly acknowledged as the main cause of ecosystem change. To halt this trend, several international agreements have been made, and various biodiversity metrics have been developed to evaluate whether the targets of these agreements are being met. The process of developing good indicators is not trivial. Indicators should be able to synthesize and communicate our current knowledge, but they also need to meet both scientific and practical criteria. Since it would not be practical to monitor all species, indicators are mainly built on the monitoring of some well-known taxa, such as birds. Here we systematically review the wide spectrum of bird biodiversity indicators (hereafter indicators) available to: i) evaluate recent methodological advances; ii) identify current knowledge gaps jeopardizing indicator interpretation and use in guiding decision-making; and iii) examine challenges in their applicability across different spatial and temporal contexts. We pay particular attention to indicator characteristics such as site and species selection, spatial, seasonal and habitat coverage, and statistical issues in developing indicators and tools to tackle them, to provide specific recommendations for the future construction of indicators. Several methodological advances have recently been made to enhance the process of indicator development, including multiple ways to select sites and species to increase their robustness. However, we found that there are strong spatial, seasonal and habitat biases among the selected indicators. Most of them are from Europe, using mainly census data from the breeding season and typically covering farmland and forest habitats. The major challenges that we detected in their applicability were related to the modelling of the statistical uncertainty associated to the indicator. We recommend the use of quantitative methods in site and species selection procedures whenever possible. Current indicators should be expanded to areas outside Europe and to less studied habitats and should not neglect monitoring work outside the breeding season. Time-series analyses studying temporal trends and using multi-species data should in general account for temporal autocorrelation as well as for phylogenetic correlation. Multi-species hierarchical models are a good alternative for analysing and constructing indicators, but they need to include annual random effects allowing for unexplained annual variation in the average status of the community, i.e. the indicator target. Despite methodological and context-specific differences in the indicators reviewed, most of them seem to highlight the urgent need of devising strategic climate and conservation policies to improve the status and trends of biodiversity.