Is maize-cowpea intercropping a viable option for smallholder farms in the risky environments of semi-arid southern Africa?

•Maize-cowpea intercropping had either no or a positive effect on maize yields.•Intercropping reduced cowpea yields by 5–35% when compared with the sole cowpea.•Maize had a larger root length density compared with cowpea allowing it to outcompete cowpea when intercropped.•Intercropping improved land-use productivity regardless of season quality and soil type. Intercropping cereals with legumes can potentially enhance productivity and soil fertility. There is limited experimental evidence on the mechanisms underlying benefits or risks in intercropping systems and below-ground interactions in intercrops remain largely unstudied. Such understanding can inform strategies towards maximising returns to investments, particularly in poor fertility soils on smallholder farms in semi-arid areas of sub-Saharan Africa. Additive intercropping experiments were established covering several seasons (2010/11–2014/15) and different conditions (on-station and on-farm) to determine effects on soil chemical variables, root dynamics and yield of intercrops. Maize was planted with the first effective rains and received either no fertiliser or 40kgNha−1. Cowpea was planted on the same date as maize or three weeks after planting maize in intercrops or sole stands and received no fertiliser. End-of-season available N was highest (P1. Intercropping, however, resulted in compromised cowpea yields especially under the relay intercrop compared with the sole cowpea stands whilst maize yield was either not affected or improved. We attributed this to the lack of below-ground niche differentiation in root distribution between maize and cowpea. Maize–cowpea intercropping with low doses of N fertiliser resulted in over-yielding compared with monocropping. Intercropping proved to be a robust option across seasons and soil types, confirming that it is a promising option for resource-poor smallholders.