Biodiversity and ecosystem services have become major sociological, ecological and economic issues worldwide. However, understanding the complex relationships between landscape structure, biodiversity, management, ecological functions, and ecosystem services remains a challenging research question. An important example of such complexity is the small, fragmented woodlands of agricultural landscapes, which might be influenced by both local factors (e.g. management), and the surrounding landscape, including other wooded habitats and the farmland matrix. In this work, we investigated how several aspects of agricultural landscape, woodland ecosystem and ecosystem services interact, using partial least square ‐ path modelling (PLS‐PM). 28 woodland patches were sampled for biodiversity, management intensity, and supporting, regulating and producing ecosystem services. The 1km² landscape surrounding the sampled woodland patches were described as well, using 3 landscape representations: the usual anthropological landscape (land cover/ land use), the bio‐physical continuous landscape (based on remote sensing data), and the ecological landscape (as perceived by bird communities). All three representations accounted for the farmland mosaic heterogeneity. PLS‐PM was constructed as a set of linear relationships between woodland ecosystem components: biodiversity, management, ecological functions, and ecosystem services. Each conceptual block is represented by a latent variable (the concept) linked to manifest variables (the observed measures). PLS‐PM enable to account for correlations and crossed relationships between the latent variables and to assess the direct and indirect effects of a latent variable on another. We compared the results obtained from the different landscape representations. Preliminary results showed that relevant and significant PLS‐PM models can be constructed and that significant relationship exist between biodiversity, ecosystem functions and ecosystem services.