<![CDATA[Abstract. Sulaksono N, Nayasilana IN, Hadiyan Y, Pradopo ST, Haryoso A, Wahyudi J, Adiningtyas DP, Putra RA. 2026. Identifying spatial interaction zones between humans and long-tailed macaques in the Mount Merbabu landscape, Indonesia. Biodiversitas 27 (4): d270414. https://doi.org/10.13057/biodiv/d270414. Human-long-tailed macaque (Macaca fascicularis) interactions are an increasing conservation and management concern in Indonesia, particularly within forest-agriculture-settlement mosaics surrounding protected areas. This study aimed to identify spatial interaction zones between humans and long-tailed macaques using habitat suitability modeling in the Mount Merbabu landscape. Habitat suitability was modeled using the Maximum Entropy (MaxEnt) algorithm based on occurrence records from direct field encounters and indirect evidence of macaque presence. Model complexity was optimized using the ENMeval framework based on the Akaike Information Criterion corrected for small sample sizes (AICc), cross-validated AUC, and omission rate. Eight environmental and anthropogenic predictors representing topography, vegetation, disturbance, and human proximity were included in the model. The final model showed stable predictive performance (cross-validated AUC: 0.7469±0.1023). Distance to burned areas emerged as the strongest predictor, followed by distance to settlements and land cover. Suitable habitat covered 4,131.71 ha (25.80%) of the interaction analysis area (16,010.88 ha) outside the national park and extended beyond forest interiors into surrounding agricultural and settlement landscapes. Moderate and high interaction zones (3,511.44 ha and 620.28 ha, respectively) formed localized clusters along forest-agriculture transition areas surrounding the national park. Sensitivity analysis showed that the spatial configuration of interaction zones remained consistent across alternative occurrence models. These mapped zones represent relative spatial interaction potential derived from macaque occurrence patterns and their overlap with human land use, rather than direct predictions of conflict incidence. The results demonstrate that MaxEnt-based spatial modeling provides a useful framework for identifying potential human-wildlife interaction zones and supporting spatially targeted mitigation in protected-area buffer landscapes.]]>
