Crop simulation models are widely used as research tools to explore the impact of various technologies and compliment field experimentation. Machine learning (ML) approaches have emerged as promising artificial intelligence alternative and complimentary tools to the commonly used crop production models. The study was designed to answer the following questions: (a) Can machine learning techniques predict maize grain yields under conservation agriculture (CA)? (b) How close can ML algorithms predict maize grain yields under CA-based cropping systems in the highlands and lowlands of Eastern and Southern Africa (ESA)? Machine learning algorithms could predict maize grain yields from conventional and CA-based cropping systems under low and high potential conditions of the ESA region. Linear algorithms (LDA and LR) predicted maize yield more closely to the observed yields compared with nonlinear tools (NB, KNN, CART and SVM) under the conditions of the reported study. However, the KNN algorithm was comparable in its yield prediction to the linear tools tested in this study. Overall, the LDA algorithm was the best tool, and SVM was the worst algorithm in maize yield prediction. Evaluating the performance of different ML algorithms using different criteria is critical in order to get a more robust assessment of the tools before their application in the agriculture sector.

Agriculture is threatened by ever increasing temperatures and this trend is predicted to continue for the near and distant future. The negative impact of rising temperatures on agri-food systems is also compounded by the erratic and highly variable rainfall in most parts of southern Africa. Minimum and maximum temperatures’ variability and trend analysis were undertaken using daily time series data derived from 23 meteorological stations spread across Malawi, Mozambique, South Africa and Zimbabwe. The modified Mann–Kendall and Theil–Sen slope models were used to assess temperature trends and their magnitudes. Temperature varied with location and minimum temperature was more variable than maximum temperature. Semi-arid regions had higher variation in minimum temperature compared to humid and coastal environments. The results showed an upward trend in minimum (0.01–0.83 °C over a 33–38 year period) and maximum (0.01–0.09 °C over a 38–57 year period) temperatures at 9 and15 locations, respectively. A downward trend in minimum temperature (0.03–0.20 °C over 38–41 years) occurred in South Africa at two locations and Dedza (Malawi), while a non-significant decline in maximum temperature (0.01 °C over 54 years) occurred at one location in coastal dry sub-humid Mozambique. The results confirm the increase in temperature over 33–79 years, and highlight the importance of including temperature when designing climate change adaption and mitigation strategies in southern Africa and similar environments.

Mechanisation is back among top development policy priorities for transforming African smallholder agriculture. Yet previous and ongoing efforts ubiquitously suffer from lack of scientific information on end‐user effective demand for different types of mechanical innovations to inform public investment or business development programmes. We assess smallholder farmers' willingness to pay (WTP) for two‐wheel tractor (2WT)‐based ripping, direct seeding and transportation using a random sample of 2800 smallholder households in Zambia and Zimbabwe. Applying the Becker–DeGroot–Marschak Mechanism (BDM) experimental auctions, we find that at least 50% of sample households in Zambia and Zimbabwe were willing to pay more than the prevailing market prices for ripping. In nominal terms, sample households in Zimbabwe were willing to pay more than those in Zambia for the different services. Empirical results suggest that wealth is the strongest driver of WTP for tillage and seeding 2WT services while labour availability and using animal draft power reduce it. These findings imply a need to (i) raise awareness and create demand for 2WT‐based services in an inclusive business manner that does not create perverse incentives and (ii) better target mechanisation to operations with comparative advantage, using approaches that bundle 2WT‐based and other mechanisation services with asset‐agnostic credit schemes or other interventions meant to overcome asset‐mediated barriers.