Calibrating occupancy to density estimations to assess abundance and vulnerability of a threatened primate in Tanzania

The current decline of mammals worldwide makes quantitative population assessments crucial, especially for range-restricted and threatened species. However, robust abundance estimations are challenging for elusive or otherwise difficult to detect species. Alternative metrics requiring only presence/absence data, that is, occupancy, are possible but calibration with independent density estimates should be foreseen, although rarely performed. Here, we calibrated density estimates from acoustic surveys to occupancy estimates from camera-trapping detections to derive the abundance of the endangered Sanje mangabey (Cercocebus sanjei) across its entire range in the Udzungwa Mountains of Tanzania. We found marked occupancy-density relationships for the two forest blocks where this primate occurs and used them to derive spatially explicit density estimates. Occupancy increased in montane forest zones at mid-elevation but decreased slightly with proximity to forest borders. We predicted an average density (+/- SE) of 0.26 +/- 0.05 groups/km(2) in the national park and 0.24 +/- 0.06 in the nature reserve. Accordingly, and given the much larger area of the reserve, the average predicted individual abundance was 1555 +/- 325 and 2471 +/- 571 in the national park and nature reserve, respectively. We found higher density and abundance in the nature reserve compared with previous studies. Given the past disturbance and poorer protection in the nature reserve relative to the national park, our results instill optimism for the status of the species, although occupancy analysis highlighted the potential vulnerability of this primate to human disturbance. Our approach appears valuable for spatially explicit density estimations of elusive species, and provides robust assessments of vulnerability and identification of priority areas for conservation of threatened populations.