Protected areas are vital to conserve our planet’s biodiversity. However, to prioritise protected area placement, and assess the effectiveness of the conservation actions within them, we need robust biodiversity monitoring over large areas. Such monitoring is often difficult and expensive, particularly in the remote regions that contain our last remaining patches of wilderness, which are so important for many species and habitats. 

This study trialled the use of passive acoustic monitoring to collect data on multiple species groups over a vast area of Ukraine and Belarus, which is part of the Polesia region. This region holds some of Europe’s best preserved peatlands and natural floodplains. BTO researchers, in partnership with in-country experts, were able to collect recordings of bat, nocturnal bird, small mammal and invertebrate species at over 500 sites across Polesia. They then built cutting-edge machine-learning classifiers to help identify these species (available on the BTO Acoustic Pipeline) from the thousands of hours of recordings collected. With these data on species presence and acoustic activity (number of detections per day), predictive models were then developed to map the distribution and relative abundance of these taxa across the landscape. 

These models and maps confirmed the immense value of existing protected areas for conserving biodiversity, but also highlighted important gaps in the protected area network that should be prioritised for future protection. Despite the terrible political problems and ongoing war in the region, data from this study have already been used to help designate a new national park in Ukraine.

Aim

To combat the global biodiversity crisis, robust and scalable data are needed to target, monitor and evaluate conservation efforts, particularly in data-poor regions and for cryptic taxa. Passive acoustic monitoring (PAM) has the potential to provide solutions, but real-world examples are still rare. We demonstrate how PAM data can be used to rapidly and effectively map distributions of multiple taxa over large scales, in data-poor regions. We show how these data can be used to assess the importance of existing protected areas and prioritise future conservation efforts, including for rare and cryptic species often neglected in such assessments. 

Location 

Global with a case study from Polesia, Eastern Europe. 

Time Period 

Present. 

Major Taxa Studied 

Bats, Birds, Small mammals and Bush crickets. 

Methods 

Using machine-learning and manual verification, we identified bats, nocturnally active birds, small mammals and bush crickets from over 34,000 monitoring hours at 506 sites in the Polesia region of Belarus and Ukraine. Using multi-species generalised mixed models in a Bayesian framework, we then predicted occupancy and acoustic activity for these species and their associations with protected areas, over a 151,000 km2 project area. We identified areas of high conservation priority as measured by species richness and/or importance for globally or regionally threatened species. 

Main Conclusions 

Our approach provides a roadmap for collecting and processing large-scale, multi-taxa biodiversity data using passive acoustic monitoring. In our case study region, we show that although existing protected areas contain a relatively large proportion of high conservation priority areas, there are significant gaps in the protected area network. We also show low surrogacy of areas of high conservation priority between taxa at fine scales, but did at larger scales, showing the importance of multi-taxa monitoring to prioritise protected areas that conserve a wide variety of species.