<![CDATA[The Javan banteng (Bos javanicus) persists on Java mainly in a small number of protected-area strongholds, making robust climatic niche characterization important for conservation planning and for evaluating potential management or restoration targets. Here, we modeled banteng climatic suitability in southwestern Java using a MaxEnt (maxnet) framework calibrated with bioclimatic predictors from CHELSA and benchmark occurrence records from extant populations in Ujung Kulon National Park (UKNP) and Alas Purwo National Park (APNP). To contextualize transferability to non-occupied protected habitat, we also projected suitability to Cagar Alam Pananjung Pangandaran (CAPP) and quantified environmental novelty using the Multivariate Environmental Similarity Surface (MESS). Univariate comparisons indicated that all eight bioclimatic variables differed significantly among UKNP, APNP, and CAPP, supporting strong site-level climatic differentiation. Tuned MaxEnt models showed good discrimination under cross-validation, and projections revealed pronounced contrasts among sites: mean suitability was low in UKNP (≈0.13), high in APNP (≈0.80), and intermediate in CAPP (≈0.48). MESS values indicated that UKNP and APNP projections largely remained within the training climatic envelope, whereas CAPP exhibited localized environmental novelty (negative MESS), implying higher extrapolation risk and greater uncertainty in inference. Overall, our results suggest that APNP currently aligns most closely with the modeled climatic niche, while CAPP may contain partially suitable conditions but requires cautious interpretation and additional ecological validation (e.g., habitat structure, disturbance, and prey–human interactions) before being considered in conservation decision-making.]]>
Copyrights © 2026
