Improving the understanding of bat population trends in Great Britain

Understanding population trends is important in achieving global targets to restore and conserve biodiversity. Bats are considered indicators of biodiversity health, yet population trends for many species are poorly understood. In this thesis, I first reviewed evidence for socio-ecological drivers of European bat population trends (Chapter 2) and identified significant evidence gaps for many proposed drivers, such as climate change and agricultural practices. Addressing these gaps required a robust understanding of bat population trends spatially and temporally. I investigated impacts of spatial biases of bat populations trends (Chapter 3), using data from the long-term Britain-wide National Bat Monitoring Programme’s (NBMP) mobile passive acoustic survey – the Field Survey. I applied a Bayesian hierarchical modelling method, integrated nested Laplace approximation (INLA), to NBMP Field Survey data on four bat species in Great Britain. Whilst trends were broadly robust to biases, the small differences to the original trends could propagate over time. Disaggregating trends to national-levels highlighted clear differences compared to the Britain-wide trends. I then applied INLA under a spatially-explicit framework to assess the impact of climate, land-cover change, and anthropogenic pressures on the four bat species (Chapter 4). Responses varied between species, with associations between climate, and increasing woodland and water, and agricultural land cover. However, predictive mapping highlighted climate was likely most important in Ella Browning – Improving the understanding of bat population trends (doctoral thesis) 4 driving current abundance distributions. The passive acoustic sensors used by the NBMP Field Survey limits the number of species monitored. I therefore reviewed opportunities and challenges in the field of passive acoustic monitoring (PAM) for ecological research (Chapter 5). I concluded data analysis tools currently limit inference from large-scale PAM, yet PAM has huge potential for identifying human impacts on ecosystems. Finally, in Chapter 6 I discussed future perspectives for combining contemporary PAM technology with statistical methods, such as INLA, to ensure bat population trends are robust to biases, drivers of population trends identified, and conservation strategies are effective