Bridging acoustic monitoring and field surveys to assess bird responses to solar farms

Solar energy infrastructure is expanding rapidly, yet its ecological effects on bird communities remain insufficiently documented. Robust assessment of these effects requires large-scale, long-term monitoring, often constrained by logistical and financial limitations. Passive acoustic monitoring (PAM) offers a scalable alternative, yet the reliability of acoustic indices as proxies for avian diversity in solar farm contexts remains uncertain. This study aims to (i) assess the ability of acoustic indices to reflect bird community attributes and (ii) evaluate the ecological effects of solar farms on these communities.As part of the ICARE project (2025–2029), 25 sites in western France were monitored, including 15 operational solar farms and 10 planned sites, each paired with a control site following a Before–After–Control–Impact (BACI) design. Bird communities were characterised using continuous acoustic recordings and standardised field surveys conducted between March and October 2025, with analyses focusing on daytime recordings during the breeding season (April–June 2025). Species richness and abundance were derived from field observations using the R package vegan, while acoustic indices (NDSI, ACI, ADI, entropy) were computed from recordings using soundecology. Relationships between acoustic and ecological metrics, as well as spatial patterns within sites, were examined using correlation tests, principal component analysis (PCA), clustering methods, and mixed-effects models.Preliminary results reveal significant associations between several acoustic indices and ecological metrics. NDSI was positively related to Shannon diversity, whereas entropy was negatively associated with both species richness and abundance, likely reflecting increased abiotic noise masking biotic signals. Clustering analyses showed no clear separation between built and unbuilt sites, suggesting that bird community patterns are jointly shaped by solar farm presence and the surrounding landscape context. Bird diversity and abundance were significantly lower in the centre of operational solar farms than at edges and paired control sites, indicating a spatial edge effect associated with infrastructure.Overall, acoustic indices provide only a partial representation of bird communities and should not be used as standalone biodiversity indicators. However, their integration with standardised field surveys can improve the reliability of biodiversity assessments. These preliminary results highlight the importance of combining complementary monitoring approaches and explicitly accounting for landscape context when evaluating the ecological impacts of solar energy projects.