Landscape context mediates the physiological stress response of birds to farmland diversification

Farmland diversification practices are increasingly adopted to help reverse biodiversity declines in agroecosystems. However, evidence for the effectiveness of this approach often comes from documenting the species attracted to particular farming systems or landscapes, rather than their underlying physiological states that ultimately determine population growth or decline over the longer term. Across 38 organic, mixed‐produce farms spanning the US west coast, we quantified three physiological biomarkers that are widely used to capture variation in short‐ and long‐term stress responses for nine bird species with diverse life‐history traits. While controlling for other potentially confounding variables, we used multilevel models to examine the association between bird physiological conditions, landscape context and local farm management practices, including the integration of livestock, and cropland composition and configuration. Birds generally had lower stress responses on more‐locally diverse farms and in landscapes with higher amounts of semi‐natural cover. However, interactions between farm diversity and landscape context suggested birds were less stressed on more diverse farms in simpler landscapes, but more stressed and in poorer condition on more diverse farms embedded within complex landscapes. We found no differences in stress responses among birds in relation to their degree of human association (synanthropy), which suggests generality in our findings. Synthesis and applications. Birds were often less stressed on more diverse farms. However, these patterns were mediated by landscape context: birds were less stressed on more diverse farms in simplified landscapes, but more stressed on more diverse farms within complex landscapes. This means land managers might not see uniform effects of local diversification schemes across different landscape contexts. Our results underscore the need to consider responses related to physiological health status, in addition to species richness and abundance, to determine the effectiveness of farm management for biodiversity conservation.