Small-scale agriculture, not plantations, fueling deforestation in Central Africa

Ssmall-scale agriculture is the dominant driver of forest loss in Central Africa, far outweighing large-scale plantation agriculture in land impact, reveals a major remote sensing study published in Remote Sensing. The study, titled "Characterizing Crop Distribution and the Impact on Forest Conservation in Central Africa", utilized deep learning algorithms and high-resolution satellite data from Sentinel-1 and Sentinel-2 to map crop types and cultivation scales across six countries, exposing the spatial relationship between agriculture and deforestation from 2000 to 2022.

The countries studied include the Democratic Republic of Congo (DRC), Republic of Congo, Central African Republic (CAR), Cameroon, Equatorial Guinea, and Gabon. Researchers used a U-Net semantic segmentation model to distinguish between small-scale and large-scale production of tree crops, non-tree crops, and oil palm. The results offer unprecedented clarity on land-use dynamics in a region critical to global biodiversity and carbon storage.

How extensive is agricultural land use in central Africa?

The study found that small-scale agriculture overwhelmingly dominates Central Africa's agricultural landscape, with small-scale tree crops (SSTC) and small-scale non-tree crops (SSNTC) covering approximately 293,249 km² and 164,824 km² respectively. In contrast, large-scale non-tree crops (LSNTC) and large-scale tree crops (LSTC) accounted for just 20,153 km² and 7436 km², respectively. Large-scale oil palm (LSOP) and small-scale oil palm (SSOP) plantations covered 2812 km² and 1386 km², respectively.

Other land types, including built-up areas, waterbodies, grasslands, and shrublands, still represent the largest share of land use, estimated at over 3.55 million km². However, small-scale agriculture now occupies a significant 12 percent of total land cover, showing a clear and growing encroachment into previously forested regions.

Accuracy assessments for the deep learning model revealed strong performance in detecting large-scale oil palm (91.2% user accuracy) and large-scale tree crops (60.5%). However, the user accuracy for small-scale crops was lower, especially for SSTC at just 5.2%, indicating classification challenges due to field heterogeneity and mixed farming practices.

What are the impacts on forests and biodiversity?

The conversion of forests to cropland, particularly for small-scale agriculture, has steadily increased over the past two decades, especially in DRC, Congo, and Cameroon. DRC experienced the highest forest-to-SSTC conversion, reaching 60,000 km² by 2015. Although a modest decline occurred by 2022, the cumulative effect remains substantial. Congo followed with a peak SSTC forest conversion of 1237 km² by 2015, while Cameroon reached 405 km² by 2020.

Forest conversion to SSNTC was also notable, with DRC again leading, starting at 374 km² between 2000–2005 and reducing to 19.6 km² by 2022. Similar, though smaller, trends were observed in Congo and Cameroon. Gabon, CAR, and Equatorial Guinea saw minimal forest-to-cropland conversion due to greater forest protection and smaller agricultural footprints.

Large-scale plantations contributed less to forest loss than previously assumed. Even in high-coverage countries like DRC and Cameroon, LSTC-induced forest loss rarely exceeded 100 km² in any five-year interval. These findings counter the perception that industrial plantations are the primary culprits of regional deforestation.

What policy actions are recommended to balance agriculture and conservation?

The study calls for a major policy reorientation to address smallholder-driven deforestation. It recommends integrating biodiversity conservation goals with agricultural policies, particularly in rural regions where subsistence farming dominates. The authors stress that while large-scale agriculture is often scrutinized, smallholder farming, driven by shifting cultivation, population growth, and export crop demand, plays a far larger role in landscape transformation.

Key recommendations include:

• Supporting sustainable smallholder practices, such as agroforestry and climate-smart farming.
• Investing in rural market infrastructure to reduce land pressure by improving yields and commercial viability.
• Enforcing spatial planning and land-use zoning to prevent expansion into biodiversity hotspots.
• Expanding forest restoration initiatives, particularly through programs like AFR100, which targets millions of hectares for rehabilitation.

The authors underscore that conserving the Congo Basin's forest requires strengthening local governance, enhancing conservation incentives, and building capacity in remote sensing and land monitoring. Without these actions, small-scale agriculture will continue to outpace conservation efforts, undermining both climate goals and food security.

This study provides Central African policymakers and international conservation partners with critical data to reshape regional land-use strategies. As deforestation accelerates, understanding its granular drivers becomes essential for sustainable development and climate resilience.