<![CDATA[Abstract. Aisyahtunnisa AR, Rabbani AH, Zain AN, Saputra ADP, Indrawan M, Sugiyarto, Dianti, Setyawan AD. 2026. GIS-based biophysical screening for agroforestry in the volcanic highland landscape of Selo, Central Java, Indonesia. Biodiversitas 27 (8): d270819. https://doi.org/10.13057/biodiv/d270819. Agroforestry is widely promoted for production and soil conservation in intensively cultivated highlands, yet a suitability map is only as reliable as the data and assumptions used to build it. We examined the volcanic highland landscape of Selo Sub-district, Central Java, Indonesia, using a screening-level GIS model rather than a parcel-scale prescription. Five biophysical criteria were combined by weighted linear combination: a soil-fertility proxy based on soil organic carbon and cation-exchange capacity (40%), slope (30%), rainfall (10%), Normalized Difference Vegetation Index (NDVI; 10%), and elevation (10%). The soil surface came from inverse-distance interpolation of 10 published observations; elevation and slope were derived from DEMNAS, rainfall from mean annual CHIRPS precipitation for 2015-2024, and NDVI from a June 2024 Sentinel-2 Level-2A scene. Of the 6,080.75-ha analytical footprint, 3,077.15 ha (50.60%) fell in S3, 2,426.67 ha (39.91%) in S2, 395.90 ha (6.51%) in N, and 181.03 ha (2.98%) in S1. The S1-S2 area totaled 2,607.70 ha (42.88%), concentrated mainly in central and eastern Selo. Rainfall became spatially non-discriminating after reclassification. The mapped classes matched 42 of 48 field reference labels (87.5%; Wilson 95% CI: 75.3-94.1%; kappa = 0.833), but these observations were not an independent probability-based validation of agroforestry performance. The map is best used to identify broad zones for further assessment. Decisions at farm or parcel level still require legal, hazard, land-use, tenure, socioeconomic, and system-specific information.]]>
Copyrights © 2026
