Many smallholder farmers in Africa face unpredictable climate conditions, pests, and resource constraints—yet conventional breeding trials often fail to reflect their realities. Early-stage On-Farm Sparse Testing (OFST) offers a game-changing approach by shifting crop trials directly to smallholder farms, capturing real-world conditions and accelerating the development of resilient, high-yielding varieties. Could this be the key to better crop breeding and food security?
Over 80 per cent of the world’s 570 million farms are smallholder farms under 2 hectares, supporting rural livelihoods in impoverished regions. Smallholder farmers, who form a significant portion of the 690 million people experiencing hunger, need improved crop varieties to thrive under challenging conditions like low inputs, climate change stresses and pests.
Challenges of breeding for smallholder farmers
Particularly at early stages, breeding programs face difficulties replicating the diverse and resource-constrained environments of smallholder farms, referred to as the Target Population of Environments (TPE). The TPE encompasses all locations where new crop varieties will be grown, characterised by varied biophysical conditions, environmental stresses and farming practices.
Conventional research stations, where new selection candidates are tested, don’t fully replicate smallholder conditions. Practices like manual labor for weed management or intercropping are common among smallholders but rarely modelled on research stations. This mismatch can lead to inaccurate predictions of crop performance on farms and discarding potentially successful candidate varieties.
Early-stage on-farm sparse testing (OFST)
Early-stage OFST shifts testing to hundreds of smallholder farms at early stages, addressing two major issues. First, it evaluates crops under real-world, farmer-managed conditions. Second, it captures the diversity within the TPE by conducting trials on numerous farms.
Using farm-as-incomplete-block (FAIB) designs, small farms test 3 to 5 candidate varieties, aligning with their plot size and resource constraints. A genomic relationship matrix connects trials across farms, ensuring comparability between farms and enhancing selection accuracy through the sharing of information. This approach reduce replication and enable testing more candidates, thereby improving breeding efficiency and providing a basis for accelerated parent recycling. Smallholder farmers are highly diverse, and careful sampling ensures a wide range of farmers can participate. The small land requirements of this approach allow small, poor and/or women farmers to equally participate.
On-farm testing and on-station testing: complementary yet demanding approaches
Early-stage on-farm testing complements, rather than replaces, on-station trials. Controlled assessments for traits like disease resistance and managed abiotic stresses remain essential at research stations. Conversely, insights from early-stage OFST can help refine on-station testing to better represent farming realities.
Scaling early-stage OFST demands strong partnerships between CGIAR, NARES, and farmers for decentralised trial management. Farmers must consent to participate and be compensated for risks. Additionally, significant resources and coordination are required to ensure trials are representative and reliable.
Despite these challenges, early-stage OFST holds transformative potential. By aligning breeding programs with smallholder realities, it can deliver improved crop varieties faster and more effectively, enhancing food security for those who need it most.
This piece was originally published by CIMMYT and has been edited to suit Farming First’s editorial guidelines.
