Combining multi-state species distribution models, mortality estimates, and landscape connectivity to model potential species distribution for endangered species in human dominated landscapes

摘要: Abstract Species ranges are changing in response to human-related disturbances and often management conservation decisions must be based on incomplete information. In this context, species distribution models (SDMs) represent the most widely used tool, but they lack any reference demographic performance of population under study, spatial structure habitat patches, or connectivity at landscape level. Combining a multi-state SDM with pattern analysis mortality model, we developed spatially-explicit, integrated model assist inform planning for Apennine brown bear central Italy. We identified 15 critical areas, potentially hosting 76 adult female bears. Many these areas are, however, characterized by high levels mortality, making them attractive sink-like areas. Structural was higher northern part study area while only limited characterizes core area, where bears currently live. Our indicates that relict isolated is realistic goal, as estimated 192–270 could live across Apennines. modelling framework enhances biological realism traditional SDMs provides tool integrates suitability, risk (as component total performance) structural among patches scale. It particularly suited endangered living human-modified landscapes establishing spatially explicit goal would facilitate pro-active management.