The Use of Birds as Biodiversity Indicators of Climatic Change: Downscaling European Indicators to Regional and National Trends

Abstract

This thesis examines the impacts of recent climatic change on bird species across Europe and develops robust indicators of ecosystem change at national and supra-national levels. The performance of three methods of species distribution modelling of European bird populations in relation to climatic change over the last 60 years is evaluated; the models being generalized additive models (GAM), climate response surface models (CRS) and maximum entropy models (MaxEnt). European breeding bird distributions were simulated using models that were parameterised using: (1) distribution data from the Climatic Atlas of European Breeding Birds and (2) augmenting these European data with distribution data from Turkey, Cyprus and North Africa. Including data from a wider geographical area improved the fit of SDMs; this was especially marked for some species with relatively poor fits based on the Europe-only dataset. Of the three SDM models tested CRS best simulated current species range data (mean AUC=0.982), closely followed by GAM (mean AUC=0.950), with MaxEnt performing worst (mean AUC=0.741).

The most robust of the modelling techniques (CRS and GAM) were used to produce climate suitability trends (CST) for European breeding birds, using population data provided by the Pan European Common Bird Monitoring Scheme (PECBMS). Population trends in European breeding birds significantly correlated with SDM retrodicted trends in 5 of 11 countries considered. Biological variables were assessed to examine their impacts on recent abundance trends; of these, habitat preference was the only biological variable found to impact upon the relationship between CST and the PECBMS trends. We generated indicators of the impact of climatic change by contrasting species trends of those projected to increase versus those projected to decline from the SDMs based on regional and national level data. Indices were also produced for individual species based on their observed and simulated trends among countries. Monitoring duration had a substantial effect on the strength of the indicator; therefore, the continuous updating of monitoring schemes is vital to ensure the accuracy of such indicators. Downscaling the continental indicator produced informative and reliable indicators that can inform policy decisions at a national level, helping to preserve biodiversity.