Investigating the correlation between environmental variables and species distribution should be performed using data acquired from appropriate spatial scales to meet adaptive management requirements in a changing environment. This research aimed to model the influence of climate change on the spatial distribution of Brant’s oak (Quercus brantii Lindl.) via presence data acquired from local (Ilam province, western Iran), regional (Zagros ecoregion), and global (whole distribution extent of Brant’s oak) extents. To project the potential habitat of Brant's oak, general circulation models (CCSM4, HADGEM2-ES, BCC-CSM1–1 and GISS-E2-R) under the 2.6 and 8.5 representative concentration pathways (RCP) for 2050 and 2070 were used. To model the distribution of Brant’s oak, artificial neural network (ANN), random forest (RF), generalized linear model (GLM), and maximum entropy (MaxEnt) were compared. To validate the models, random-holdback cross-validation, whereby 80% of the data was randomly selected to calibrate the model and the remaining 20% was used to validate the models, was carried out. The results revealed that enhancing the modeling extent increased the accuracy of the model; hence, a model trained using the global dataset performed better than local and regional datasets. In all three geographical extents, RF and MaxEnt had the best performance in modeling the spatial distribution range of Brant’s oak. The main predictors of Brant’s oak distribution were different in local, regional, and global models. The mean temperature of driest quarter (bio9), at the local extent; precipitation of wettest month (bio13), at the regional extent; and temperature annual range (bio7), at the global extent were the most important climatic variables. The findings also indicated that the potential habitat of Brant’s oak will decline in the future under climate change scenarios (i.e., RCP 2.6 and RCP 8.5) and across all three geographical extents compared to the current habitat. Using the findings of this study, it is possible to identify the suitable habitats of Brant's oak forests with more certainty and take measures to manage and protect them.
Read full abstract