Prediction of the potentially suitable areas of Leonurus japonicus with the optimized MaxEnt model
Data files
Sep 26, 2023 version files 48.16 KB
Abstract
Leonurus japonicus Houtt. is a traditional Chinese medicinal plant with high medicinal and edible value. Wild L. japonicus resources have been reduced dramatically in recent years. This study predicted the response of distribution range of L. japonicus to climate change in China, which provided the scientific basis for the conservation and utilization. In this study, 489 occurrence points of L. japonicus were selected based on GIS technology and spThin package. The default parameters of the Maxent model were adjusted by using ENMeva1 package of the R environment, and the optimized Maxent model was used to analyze the distribution of L. japonicus. When the feature combination in the model parameters is hing and the regularization multiplier is 1.5, the Maxent model has a higher degree of optimization. With the AUC of 0.830 our model showed a good predictive performance The results showed that L. japonicus was widely distributed in the current period. The maximum temperature of the warmest month, the minimum temperature of the coldest month, the precipitation of the wettest month, the precipitation of the driest month and altitude were the main environmental factors affecting the distribution of L. japonicus. Under the three climate change scenarios, the suitable distribution area of L. japonicus will range-shift to high latitudes, indicating that the distribution of L. japonicus has a strong response to climate change. The regional change rate is the lowest under the SSP126-2090s scenario and the highest under the SSP585-2090s scenario.
README
Prediction of the potentially suitable areas of Leonurus japonicus habitability zones with maxent
occurrence points：By sorting out the information of Leonurus japonicus specimens recorded in the Chinese Digital Herbarium (CVH, http://www.cvh.ac.cn/), and combining with the L. japonicus presence points in the Global Biodiversity Information Platform (GBIF, https://www.gbif.org/), the existing distribution positions of L. japonicus were preliminarily obtained, and then the corresponding latitude and longitude coordinates of each distribution point were obtained by Baidu coordinate system. All were used for modeling.
environmental variables：Species’ ecoloical niches are affected by climate, topography, biology, and other factors. In consideration of the comprehensiveness and complexity of ecological factors, 34 environmental variables which could reflect species’ ecoloical niches were selected. The list included 19 bioclimatic factors, 14 soil factors and a topographic factor (altitude).The current (1970–2000), 2050s (2041–2060), and 2090s (2081–2100) bioclimatic factor data used in this research were derived from the world climate database Worldclim (http://www.worldclim.Org), and the pixel size of the data was 2.5 arc-minutes (-5 km). The climate data of the 2050s and 2090s were obtained from the Beijing Climate Center-Climate System Model-Medium Resolution (BCC-CSM2-MR), one of the Coupled Model Inter-Comparison Project Phase 6 (CMIP6) datasets, which included three scenarios: sustainable development (SSP126), intermediate development (SSP245) and conventional development (SSP585). SSP scenarios have a high accuracy and separation rate and can integrate local development factors, and so are more convincing than CMIP5 data. The data of soil factors and topographic factors were obtained form the World Soil Database (HWSD) of the FAO (http://www.fao.org/faostat/en/#data), and the provincial national vector map were from China’s Ministry of Natural Resources (http://www.mnr.gov.cn/).
The environmental variables is in ASCii format. ASCii can be viewed using standard GIS software such as:
environmental variables\climate\50126\bio1.asc
Naming convention:
Type Variables Description UNITS
Bio1 Annual Mean Temperature ℃×10
Bioclimatic Bio2 Mean Diurnal Range ℃×10
Variables Bio3 Isothermality 1
Bio4 Temperature Seasonality 1
Bio5 Max Temperatur ℃×10
Bio6 Min Temperature of Coldest Month ℃×10
Bio7 Temperature Annual Range ℃×10
Bio8 Mean Temperature of Wettest Quarter ℃×10
Bio9 Mean Temperature of Driest Quarter ℃×10
Bio10 Mean Temperature of Warmest Quarter ℃×10
Bio11 Mean Temperature of Coldest Quarter ℃×10
Bio12 Annual Precipitation mm
Bio13 Precipitation of Wettest Month mm
Bio14 Precipitation of Driest Month mm
Bio15 Precipitation Seasonality 1
Bio16 Precipitation of Wettest Quarter mm
Bio17 Precipitation of Driest Quarter mm
Bio18 Precipitation of Warmest Quarter mm
Bio19 Precipitation of Coldest Quarter mm
T_GRAVEL Topsoil Gravel Content %vol.
Top Soil Variable T_SAND Topsoil Sand Fraction % wt.
T_SILT Topsoil Silt Fraction % wt.
T_CLAY Topsoil Clay Fraction % wt.
T_USDA_TEX_CLASS Topsoil USDA Texture Classification name
T_REF_BULK_DENSITY Topsoil Reference Bulk Density kg/dm3
T_OC Topsoil Organic Carbon % weight
T_PH_H2O Topsoil pH (H2O) -log(H+)
T_CEC_CLAY Topsoil CEC (clay) cmol/kg
T_CEC_SOIL Topsoil CEC (soil) cmol/kg
T_BS Topsoil Base Saturation %
T_TEB Topsoil TEB cmol/kg
T_ESP Topsoil Sodicity (ESP) %
T_ECE Topsoil Salinity (Elco) dS/m
Terrain ELEV Elevation m
ENMeval package: To avoid overfitting due to the high complexity of the model constructed with the default parameters, which may cause the predicted distribution of the potential habitat of L. japonicus to deviate too much from the actual situation, this study used the ENMeval package in R 4.3.1, and adjusted the two most important parameters, namely, regularization multiplier (RM) and feature combination (FC), to improve the prediction accuracy of the model.
CoordinateCleaner：The R software package ‘CoordinateCleaner’ was used to removing records without coordinate precision and suspected outliers. Based on the ‘subset’ ‘clean_coordinates’ operation in CoordinateCleaner, we obtained the results of bias corrections on the datasets.
SpThin package: Spatial thinning of species occurrence records can help address problems associated with spatial sampling biases. Ideally, thinning removes the fewest records necessary to substantially reduce the effects of sampling bias, while simultaneously retaining the greatest amount of useful information. Spatial thinning can be done manually; however, this is prohibitively time consuming for large datasets. Using a randomization approach, the ‘thin’ function in the spThin R package returns a dataset with the maximum number of records for a given thinning distance, when run for sufficient iterations.
geosphere package:The geosphere package in the R environment was used to calculate the centroid range shift distance of L. japonicus under different climate change scenarios.
SDMTools：The package in R language was used to calculate the location of centroid in the suitable area of Leonurus japonicus under 6 different economic paths in the current and future periods.
VIF package: The usdm package provides a set of functions to support dealing with problematic situations in species distribution modelling (e.g., multicollinearity, positional uncertainty).To detect whether predictor variables are subjected to multicollinearity, you may use vif (variance inflation factor) metric, and some methods implemeted in this package including vifstep or vifcor (a stepwise procedure to identify collinear variables).