نبذة مختصرة : International audience ; 1. Modern farming practices threaten wildlife in different ways, and failure to identify the complexity of multiple threats acting in synergy may result in ineffective management. To protect ground-nesting birds in farmland, monitoring and mitigating impacts of mechanical harvesting is crucial. 2. Here, we use 6 years of data from a nationwide volunteer-based monitoring scheme of the Montagu's harrier, a ground-nesting raptor, in French farmlands. We assess the effectiveness of alternative nest protection measures and map their potential benefit to the species. 3. We show that unprotected nests in cultivated land are strongly negatively affected by harvesting and thus require active management. Further, we show that protection from harvesting alone (e.g. by leaving a small unharvested buffer around the nest) is impaired by post-harvest predation at nests that become highly conspicuous after harvest. Measures that simultaneously protect from harvesting and predation (by adding a fence around the nest) significantly enhance nest productivity. 4. The map of expected gain from nest protection in relation to available volunteers' workforce pinpoints large areas of high expected gain from nest protection that are not matched by equally high workforce availability. This mismatch suggests that the impact of nest protection can be further improved by increasing volunteer efforts in key areas where they are low relative to the expected gain they could have. 5. Synthesis and applications. This study shows that synergistic interplay of multiple factors (e.g. mechanical harvesting and predation) may completely undermine the success of wellintentioned conservation efforts. However, identifying areas where the greatest expected gains can be achieved relative to effort expended can minimize the risk of wasted volunteer actions. Overall, this study underscores the importance of citizen science for collecting large-scale data useful for producing science and ultimately informs large-scale evidence-based ...
No Comments.