Project Context and Opportunity

A primary focus of landscape restoration is facilitating the recovery of biodiversity and ecological functioning, however, monitoring changes in biodiversity over extended time periods, geographical areas and in the context of other restoration projects is challenging. Low-cost methods, deliverable across different scales, are therefore necessary to properly assess the impacts of restoration projects on biodiversity.

Acoustic surveys are becoming increasingly popular for monitoring biodiversity due to the low-cost of deployment, lack of observer bias and the wide range of taxa they can survey. However, processing acoustic data manually by looking through spectrograms is prohibitively time-consuming and needs expert knowledge to identify species, which may be difficult to standarise and reproduce through the typical duration of landscape restoration monitoring.

Project Aim

The project will enhance the ability of acoustic monitoring to measure the effects of restoration on biodiversity, with broader implications for biodiversity monitoring in Europe. We will build on an existing platform to enable automated acoustic identification of bats, focal bird species and species of bush-crickets and small mammals across Europe, processing data from acoustic surveys across ELP projects to help assess the effects of restoration on these taxa.

Project impact

• Develop classifiers to allow automated species identification of bats from acoustic survey data throughout Europe. We will extend our existing machine learning classifiers (Random-forest and convolutional neural networks), that cover the UK, Polesia and parts of Central Europe, to identify bats across Europe.
• Build new classifiers for owls, other bird species, and taxa.
• Incorporate the classifiers detailed above into an existing platform (Acoustic Pipeline: bto.org/pipeline) to broaden its taxonomic and geographical coverage to meet the needs of the ELP projects. This will allow researchers and citizen scientists to upload recordings from acoustic surveys to a cloud-based processing environment for automated identification.

Support acoustic sampling of restoration projects of ELP partners and analyse findings in a project-based and ELP-wide context.

Project Partners