Farmer-managed trials in developing countries often show great variability in yield and treatment effects. The use of statistical techniques is demonstrated to explain this variability in a maize + cassava intercropping trial in southwestern Nigeria. Concomitant variables, measured in the trial fields, are used as covariates in the ANOVA and as regressors in the analysis of mean site yield. It was found that a new maize variety (TZSR-W) performed better than the lcoal irrespective of farmers' yield level. Fertilizer effect increased as farmers' overall yield level (reflected in their mean site yield) increased. At current fertilizer and maize prices, fertilizer was profitable for all except the poorest-yielding fields. Variation in yield level was partly explained by degree of shade and stand losses during the season. Part of the stand losses was due to termites. The majority of female farmers in the sample had lower yields than males because of a combination of factors such as shade, late planting, poor weeding and longer occupation of the land.

https://doi.org/10.1016/0378-4290(90)90059-K

Multi standard citation

APAHarvardMLAVancouverChicagoIEEE

https://doi.org/10.1016/0378-4290(90)90059-K

Maize; Cassava; Crop Systems

Maize; Crop Yield; Nigeria; Varieties; Fertilizers; On-Farm Research; Intercropping

Africa; West Africa

Nigeria