Managing the farmscape for connectivity increases conservation value for tropical bird species with different forest-dependencies

Land clearing for agricultural use is a primary driver of biodiversity loss and fragmentation of natural ecosystems. Restoring natural habitat connectivity by retaining quality habitats and increasing on-farm tree cover contributes to species' mobility and persistence in agricultural landscapes. Nonetheless, remarkably few studies have quantified the impacts of on-farm practices for species' mobility measured as functional connectivity within the context of farm and broader spatial levels of landscape organization. We tested how adding and removing trees in different configurations on a farm comprised of coffee plantations and cattle pastures can help evaluate species’ mobility at the farmscape level (an area comprising the farm plus a 1.5 km buffer area). We coupled bird capture data and scenario modeling to assess species mobility of five neotropical bird species with distinct life history characteristics representing a gradient of forest dependency. We used seven years of mist-netting data to estimate species habitat affinity and to predict species mobility using the Circuitscape model across a 4371 ha farmscape in Costa Rica. Circuitscape allowed us to estimate changes in movement probability and relative changes in resistance to movement that species experience during dispersal (measured as resistance distance and passage area through which species can move) under four farmscape management scenarios. The four land-use scenarios included: (a) the 2011 farmscape land-use composition and configuration, b) converting all existing live fences to post-and-wire fence lines in the farm c) converting simplified coffee agroforests to multistrata coffee agroforests in the farm, and d) placing multistrata live fences around the perimeter of every parcel and roads on the farm. Model results suggest that existing multistrata live fences maintain the sporadic movement of all five species irrespective of forest dependence. Likewise, adding multistrata live fences around individual fields presents a more efficient strategy for increasing species mobility than multistrata coffee agroforestry systems in the assessed farmscape, by doubling the passage areas available to all species, although it created labyrinths with “dead-ends” for two species. While retaining large habitat patches remains important for conservation, managing on-farm connectivity complements these efforts by increasing movement probability and reducing dispersal resistance for forest-dependent bird species