Georeferenced dataset on agroecological practices (2011–2020), NDVI, and rainfall in the Maradi Region, Niger (2001–2025)

This dataset compiles georeferenced information on agroecological practices implemented in the Maradi Region of Niger between 2011 and 2020. It includes the geographic coordinates of intervention sites, types of agroecological practices, and implementation years. It also contains maximum Normalized Difference Vegetation Index (NDVImax) and Rainfall mean values calculated for the three years before and the three years after interventions for intervention sites and control sites.
The dataset also includes annual regional time series of NDVImax and rainfall for the Maradi Region, covering the period 2001–2025.

Dataset objectives
This dataset was compiled to:
- Assess annual rainfall and NDVImax trends in the Maradi Region between 2001 and 2025;
- Evaluate the contribution of agroecological practices to vegetation recovery using a Before–After Control–Impact (BACI) approach;
- Quantify the relative and interactive contributions of rainfall variability and agroecological interventions to NDVImax using regression-based attribution analysis.
Main variables
- Geographic coordinates of intervention sites.
- Year of implementation of agroecological interventions.
- Type of agroecological practice (Farmer Managed Natural Regeneration (FMNR), Zaï, half-moons, benches, dune fixation, development of grazing land, forage seeding)
- NDVImax mean values calculated for the three years before and the three years after interventions for intervention sites and control sites.
- Rainfall mean values calculated for the three years before and the three years after interventions for intervention sites and control sites.
- Annual regional NDVImax time series for the Maradi Region (2001–2025).
- Annual regional rainfall time series for the Maradi Region (2001–2025).

Data collection
Data on agroecological practices were obtained from records and databases provided by the Regional Directorate of Environment and the Regional Directorate of Agriculture of Maradi (Niger). These data were complemented with remote sensing data from Moderate-resolution Imaging Spectroradiometer (MODIS) and Landsat satellites for NDVImax derivation, as well as rainfall data from the Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) product extracted using Google Earth Engine.

Data analysis
Analyses were conducted in the R environment to assess vegetation and rainfall trends. The Before–After Control–Impact (BACI) approach was applied to evaluate the effects of agroecological practices on vegetation dynamics. Multiple regression analyses were subsequently performed to quantify the relationships between vegetation dynamics, rainfall variability, and agroecological interventions.

Data notes and limitations
The technical characteristics associated with agroecological practices (e.g., structure dimensions, spacing, or other implementation parameters) correspond to general specifications reported in the scientific literature. They do not necessarily represent site-specific measurements, as detailed implementation parameters were not systematically recorded for all intervention sites.
Repository content
The repository includes:
- Georeferenced data on agroecological practices;
- NDVI max mean and rainfall mean datasets for intervention and control sites, as well as annual regional NDVI max and rainfall time series;
- Datasets prepared for trend analyses and multiple regression analyses, BACI results (including NDVI before and after intervention, NDVI, relative NDVI increase, BACI index, and p-value)
- Files required for understanding the agroecological practices and analysis codes.

Data availability on request: The agroecological intervention data used in this study were compiled from information provided by the Regional Directorate of Environment and the Regional Directorate of Agriculture of Maradi (Niger) for scientific research purposes, complemented by data from the IRD/DataSuds dataset on degraded land reclamation actions (Sadda et al., 2024; DOI: 10.23708/ACJTAW). The compiled dataset is not publicly available. Access to the data may be possible upon reasonable request to Djibo Amadou Abdoulkader, subject to the applicable conditions established by the Regional Directorate of Environment and the Regional Directorate of Agriculture of Maradi (Niger), as well as the terms of use applicable to the IRD/DataSuds dataset.