The dataset of Liquidambar orientalis for species distribution models

The primary objective of this study was to predict the existing geographic range of Liquidambar orientalis, commonly known as the oriental sweetgum. To gain insights into the potential effects of climate change on the oriental sweetgum, the study employed species distribution models to project the model to future periods. Considering two Shared Socioeconomic Pathways (SSP1-2.6 and SSP5-8.5), the ensemble modeling approach utilized the biomod2 package in the R programming language to analyze the alterations in the spatial distribution of the species in forthcoming periods (namely, for the years 2035s, 2055s, and 2070s). 

Methods
1. Occurrence data

81 occurrence data were obtained from two reputable sources: the Global Biodiversity Information Facility (GBIF 2023, www.gbif.org) and the European Forest Genetic Resources Program (EUFORGEN 2023). In the dataset obtained from the Global Biodiversity Information Facility (GBIF), erroneous and redundant records were removed.

2. Environmental data

The dataset used in this study consisted of nineteen bioclimatic variables (BIO1 to BIO19) obtained from the CHELSA version 2.1 (https://chelsa-climate.org/). These variables as .tiff format represented climatic and environmental factors and were downloaded at a spatial resolution of 30-arc seconds. The dataset covered four temporal ranges: 1981-2010, 2011-2040, 2041-2070, and 2071-2100. The bioclimatic variable values of the grid cell were obtained using QGIS 3.18.2. The data utilized in this study were obtained from two Global Circulation Models (GCMs), namely the Max Planck Institute Earth System Model (MPI-ESM1-2-HR) and the Meteorological Research Institute Earth System Model Version 2.0 (MRI-ESM2.0). 

Variance Inflation Factors (VIF) were computed for climatic variables to avoid the issue of multi-collinearity. The usdm package in R was utilized for this purpose. 

3. Species distribution modeling 

The species distribution models, including the ensemble, were built using the R programming language (https://www.r-project.org/) and the biomod2 package version 4.2-4 (https://cran.r-project.org/web/packages/biomod2/biomod2.pdf). The following techniques were employed: Generalized Linear Model (GLM), Random Forest (RF), Generalized Boosted Model (GBM), Generalized Additive Model (GAM), Maximum Entropy (Maxent), and the ensemble model.