Unexpected cocoa yield responses to fertilizer application in existing smallholder farms in South-West Nigeria
Аннотация
Cocoa yield responses to fertilizers in smallholder farms are poorly understood. We evaluated yield responses across ten on-farm trials in South-West Nigeria from 2020 to 2023. Experimental plots were established in smallholder fields with 15–20 year-old cocoa trees that had never received mineral or organic fertilizers. Treatments included a control without fertilizer and combinations of 22 kg ha −1 P and 42 kg ha −1 K with 0, 50, and 100 kg N ha −1 applied annually. Soil properties were measured at the start, while management practices (pruning, weeding, pest and disease management) were assessed monthly. Cocoa yields were determined in 441 m 2 plots where 25 trees were monitored for healthy and diseased harvested pod numbers and pod mass. Yield responses were compared and treatment effects were estimated using linear and generalized linear mixed regression models. Average cocoa yield responses were + 30.9 ± 34.8, − 0.5 ± 37.9, and − 65.6 ± 37.8 kg ha −1 representing 9.5, − 0.2, and − 20% for treatments with 0, 50, and 100 kg N ha −1 , respectively. The small positive response without N and negative responses with N resulted from poor pod set and high pod losses to diseases. Yield variability within fields was large, with many poor yielding trees and few higher-yielding trees. Weather conditions, initial productivity, and cocoa tree density explained part of the spatial and temporal variation. Farmer practices improved within a year and are therefore unlikely to explain the poor yield responses. We conclude that N fertilizers used in existing cocoa plantations in South-West Nigeria did not increase yields under smallholder management, requiring a re-think of entry-points for sustainable intensification.