
This repository contains the data and code required to reproduce the analyses presented in the paper:
Sode, A. I., Fandohan, A. B., Krainski, E. T., Assogbadjo, A. E., & Glèlè Kakaï, R. (2026). Integrating Presence-only and Abundance Data to Predict Baobab (Adansonia digitata L.) Distribution: A Bayesian Data Fusion Framework. Environmental and Ecological Statistics. https://doi.org/10.1007/s10….
Overview
The present study employs a Bayesian spatial data fusion framework to integrate presence-only and structured abundance data of the African baobab. The objective of this study is to comprehend and map the spatial variation of this multipurpose agroforestry tree species, which provides a range of nutritional and medicinal products that rural inhabitants in sub-Saharan Africa extensively utilise.
The analysis is conducted in Benin (West Africa), which is characterised by a marked environmental gradient across three distinct climatic zones. Its ecological diversity, combined with limited structured survey data, provides an effective testing ground for heterogeneous spatial data fusion to enhance species distribution models in data-scarce scenarios.
Archive Contents
To ensure full computational reproducibility, this research compendium includes:
Data: Structured abundance and presence-only data with ethically compliant, 1-km jittered geographic coordinates and high-resolution environmental covariates.
Code Pipelines: Complete R scripts detailing data preprocessing, INLA-SPDE modelling, evaluation and validation steps.
Software Snapshot: The source code and binaries for the underlying R package isdmtools (v0.4.0).
Outputs: Pre-computed exploratory data analysis objects and the final manuscript figures.