Agriculture expansion and intensification are one of the main drivers of biodiversity loss and ecological disturbances in natural habitats. Environmental DNA (eDNA) has emerged as a powerful, non-invasive, cost and time-effective tool for biodiversity assessment across a wide range of ecosystems, enabling the detection of both visible and cryptic taxa.

Here, we explore the use of airborne eDNA and surface-derived eDNA to assess how different agricultural practices influence biodiversity patterns and, consequently, ecological disturbance. Specifically, we collected airborne DNA using spider webs and surface eDNA from flower substrates to characterize local communities of metazoan.

Sampling was conducted across three regions in southern Kenya, near Lake Victoria and the Kakamega Forest. Natural habitat patches were selected as reference sites representing high expected biodiversity and were compared with agroecological and conventionally managed agricultural fields.

We evaluated the ability of eDNA-based approaches to detect disturbance gradients by assessing shifts in community composition across different levels of agricultural landscapes. Both sampling methods provided a well described picture of local biodiversity, with spiderwebs being the tool providing more diversity in terms of taxa.

Preliminary results suggest that overall species diversity is comparable between agroecological fields and natural patches, while conventionally managed fields show slightly reduced diversity. In particular, pollinator diversity appears higher in agroecological systems, supporting the hypothesis that intermediate levels of disturbance may promote more complex communities.