Identifying contrasting drivers of maize and legume yield gaps in Zimbabwean smallholder farms using a locally parameterised crop model
Abstract
Problem: In sub-Saharan Africa, cereal and legume yields on smallholder farms typically fall far below both potential and water-limited levels. Understanding the drivers of yield variability in these farms is critical to design relevant pathways for sustainable intensification. Objective: Our objective was to assess the relative contribution of water, nitrogen (N) and reducing factors (weeds, pests and diseases) on the yield of maize (Zea mays L.), and grain legumes (groundnut - Arachis hypogaea L., cowpea - Vigna unguiculata (L.) Walp. and pigeon pea - Cajanus cajan (L.) Huth) in smallholder cropping systems. Methods: Crop growth variables and yields were monitored in 99 cropping situations in on-farm trials and farmers' fields, during two growing seasons in Murehwa district, sub-humid Zimbabwe. We identified 'best-managed situations' where yield reducing factors were minimized, yet with still pest and disease issues for legumes. The STICS soil-crop model was parameterised and evaluated for its ability to reproduce the observed variability in crop growth in these best-managed cropping situations that were assumed to be closest to model assumptions. The model was then used to compute the share of the yield gap caused by stress from water, N and other reducing factors. Results: The model reproduced the variations in observed grain yield and aboveground biomass in the best-managed cropping situations with R2 of 0.86 and 0.74, respectively. The computed yield gap between observed and potential yield due to water stress was large for pigeon pea, small for maize, and negligible for groundnut and cowpea. For maize, observed yield was close to simulated water- and N-limited yield (Ywn) when 160 kg N ha−1 and best management were applied. For legumes this gap between actual yield and Ywn was large across all cropping situations, even with best management. Conclusion: Under the sub-humid climate of this study, maize yield was much more limited by N than water stress. Our study suggests a larger impact of pests and diseases on legumes than on maize, in the smallholder context of the two-year on-farm monitoring. Our findings highlight that reducing pests and diseases is critical to close legumes yield gap.
Related
- Socioeconomic and biophysical drivers of cocoa yield variability in smallholder farms in Côte d’Ivoire
- Species distribution modelling of three Aacia species in the Negev Desert: predicting future range shifts and identifying climatic drivers
- City-specific or generalisable resistances? An urban animal connectivity model performs better when parameterised from two cities
- Burden and determinants of hyperglycaemia and dyslipidaemia in Zimbabwean adults living with HIV on dolutegravir-based first-line antiretroviral therapy: A cross-sectional study
- Smallholder farmers' preferences for public expenditure priorities to support maize productivity: evidence from a best–worst scaling experiment in western Kenya
BibTeX
@article{pret2026identifying,
title = {{Identifying contrasting drivers of maize and legume yield gaps in Zimbabwean smallholder farms using a locally parameterised crop model}},
author = {Valentin Pret and Gatien Falconnier and Katrien Descheemaeker and Marc Corbeels and Régis Chikowo and Rémi Cardinael and Antoine Couedel and Souleymane Diop and Juliet Murimwa and Hardlife Chinwa and Abderrahim Bouhenache and Illiana W. Kwenda and Mathilde de Freitas and François Affholder},
journal = {Field Crops Research},
year = {2026},
month = sep,
doi = {10.1016/j.fcr.2026.110716},
url = {https://hal.science/hal-05763612v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Identifying contrasting drivers of maize and legume yield gaps in Zimbabwean smallholder farms using a locally parameterised crop model
AU - Valentin Pret
AU - Gatien Falconnier
AU - Katrien Descheemaeker
AU - Marc Corbeels
AU - Régis Chikowo
AU - Rémi Cardinael
AU - Antoine Couedel
AU - Souleymane Diop
AU - Juliet Murimwa
AU - Hardlife Chinwa
AU - Abderrahim Bouhenache
AU - Illiana W. Kwenda
AU - Mathilde de Freitas
AU - François Affholder
PY - 2026
DA - 2026/09/17
JO - Field Crops Research
DO - 10.1016/j.fcr.2026.110716
UR - https://hal.science/hal-05763612v1
AB - Problem: In sub-Saharan Africa, cereal and legume yields on smallholder farms typically fall far below both potential and water-limited levels. Understanding the drivers of yield variability in these farms is critical to design relevant pathways for sustainable intensification. Objective: Our objective was to assess the relative contribution of water, nitrogen (N) and reducing factors (weeds, pests and diseases) on the yield of maize (Zea mays L.), and grain legumes (groundnut - Arachis hypogaea L., cowpea - Vigna unguiculata (L.) Walp. and pigeon pea - Cajanus cajan (L.) Huth) in smallholder cropping systems. Methods: Crop growth variables and yields were monitored in 99 cropping situations in on-farm trials and farmers' fields, during two growing seasons in Murehwa district, sub-humid Zimbabwe. We identified 'best-managed situations' where yield reducing factors were minimized, yet with still pest and disease issues for legumes. The STICS soil-crop model was parameterised and evaluated for its ability to reproduce the observed variability in crop growth in these best-managed cropping situations that were assumed to be closest to model assumptions. The model was then used to compute the share of the yield gap caused by stress from water, N and other reducing factors. Results: The model reproduced the variations in observed grain yield and aboveground biomass in the best-managed cropping situations with R2 of 0.86 and 0.74, respectively. The computed yield gap between observed and potential yield due to water stress was large for pigeon pea, small for maize, and negligible for groundnut and cowpea. For maize, observed yield was close to simulated water- and N-limited yield (Ywn) when 160 kg N ha−1 and best management were applied. For legumes this gap between actual yield and Ywn was large across all cropping situations, even with best management. Conclusion: Under the sub-humid climate of this study, maize yield was much more limited by N than water stress. Our study suggests a larger impact of pests and diseases on legumes than on maize, in the smallholder context of the two-year on-farm monitoring. Our findings highlight that reducing pests and diseases is critical to close legumes yield gap.
SN - 0378-4290
ER -