Comparative species distribution modeling of Ricania speculum: Assessing global invasion risks and bioclimatic vulnerability

The potential global distribution of the Black planthopper (Ricania speculum) was assessed under current and projected future climate change scenarios. 

Occurrence records for R. speculum were obtained from the Global Biodiversity Information Facility (GBIF, https://www.gbif.org), and spatial filtering at a 10 km resolution was applied to address issues of observational and sampling bias.

Climate data: Monthly climate data — including mean temperature, maximum temperature, minimum temperature, and precipitation — at a spatial resolution of 10 arcmin were acquired from WorldClim (https://www.worldclim.org) for a 35-year period (1990–2024). From these data, bioclimatic variables (BIO1–BIO19) and elevation (DEM) were derived, and a final set of 10 variables was selected for analysis after removing multicollinearity.

Future climate change scenarios: Changes in habitat suitability were evaluated across 18 scenarios, constructed by combining two Shared Socioeconomic Pathways (SSP2-4.5 and SSP5-8.5), three General Circulation Models (MIROC6, IPSL-CM6A-LR, and MPI-ESM1-2-HR), and three future time periods (2041–2060, 2061–2080, and 2081–2100).

Machine learning models: Three algorithms were employed: MaxEnt and Random Forest (RF), which are widely used in species distribution modeling, and XGBoost (XGB), which has seen comparatively limited application in this context. Model performance was evaluated using accuracy, AUC, and TSS, with 10-fold block cross-validation applied throughout this process.

Dataset Overview

Rsp_codes.zip Scripts for running RF and XGB models (MaxEnt was executed using dedicated software)Rsp_occurrences.zip Raw occurrence data and processed data with spatial rarefaction appliedRsp_plots.zip Visualizations of the results for each modelRsp_results.zip Output results from the MaxEnt, RF, and XGB modelsRsp_scenarios.zip Climate change scenarios used in the analysisRsp_variables.zip Bioclimatic variables and elevation data used in the analysis