Constraints and options to sustainably intensifying smallholder maize farming systems in southern Africa

In southern Africa, sustainable intensification (SI) of low input farming is promoted as a key strategy to improve the livelihoods and food security of smallholder farmers. It has been argued, however, that due to the severity and frequency of droughts, irrigation is a prerequisite for sustainable yield improvement and stability, and less crop failures. Restricted access to water for such farmers in the study region necessitates the investigation of alternative adaptive management options suited to smallholder systems. Using the Limpopo province South Africa as a case study, we use a combination of survey data (140 households) and detailed quantitative agronomic measurements and observations (116 georeferenced on-farm plots) to understand yield limitations in maize-based smallholder systems. Data was collected from five villages in the Mopani district representing a distinct climate gradient. Agronomic measurements included soil characteristics such as CN ratio, texture, rooting depth and management aspects such as weed type and soil cover, as well as maize planting density, biomass and yield. Combined insights from the interviews and detailed on-farm observations were used to benchmark the agro-ecosystem model APSIM, which was then setup for different technology levels. These were defined through combinations of advanced crop and soil management practices plus the status quo as observed through the ground-truthing campaign with no irrigation, zero to low fertilisation, little weeding, no pest management, and low planting density. Advanced practices involved higher input levels including irrigation and fertiliser, as well as management aspects such as increased planting density and intense weeding.

Survey results showed that farmers adjusted sowing time and planting density according to rainfall availability and perceived risk. Overall, input intensity levels were low (fertiliser and density) and all villages expressed similar challenges to adapt to climate variability. It appeared most farmers lacked knowledge about drought avoidance measures, and only very few had access to water for crop irrigation.

Our simulation results showed that irrigation could increase maize grain yields by around two tons ha^-1 over a three-year average for a moderately wet site under current management practices. For the driest site, this led to an increase of just over one ton ha^-1. If irrigation is applied it necessitates an increase in biotic stress management, as failing to do so can compromise potential yield gains. Higher labour input, increased input costs and possibly associated increased economic risks make such intensification strategies unattractive for some farmers depending on their age and household economic security.

For this case study, we outlined and implemented a novel method of linking survey and agro-ecosystem modelling data to assess ex-ante potential impacts of SI in smallholder cropping systems vulnerable to climate-induced risk.