Predicting the Potential Distribution of Camellia osmantha in Response to Climate Change Using the MaxEnt Model

Camellia osmantha is an economically valuable woody oil plant with increasing cultivation in southern China, yet its potential distribution under climate change remains poorly understood. Here, we used an optimized MaxEnt model with 27 occurrence points and 11 environmental variables (retained from 27 candidate predictors via correlation filtering and MaxEnt-based contribution and permutation importance) to predict its potential distribution in Guangxi under current and future climate scenarios (2050s, 2090s; SSP126, SSP370, SSP585). The model demonstrated acceptable predictive performance (training AUC = 0.7784, test AUC = 0.6935) with the test AUC falling within the acceptable/fair range (0.6–0.7) according to standard evaluation criteria. The dominant drivers were available water capacity (24.3%), altitude (24.1%), and temperature annual range (21.9%), with suitability declining when AWC exceeded 3.0, altitude exceeded 543 m, or bio7 fell below 23.06 °C. Current suitable habitats cover 18.53 × 104 km2 (89.7% of Guangxi), with high suitability areas concentrated in the northeast and southwest. Under future scenarios, total suitable area remained stable (−1.2% to +3.2%), while high suitability areas expanded under SSP370/SSP585 (+39.0%/+41.1%) and contracted under SSP126 (−30.3%) by 2090s; centroid shifts were limited (2.95–14.93 km), with core habitats remaining in central Guangxi. These findings provide a quantitative basis for conservation and sustainable cultivation strategies for this species under climate change, and highlight the importance of soil water availability alongside topographic and climatic factors in shaping its distribution.

Перевод пока недоступен