Meta-Analysis of Crop Yield Response to Fertiliser Application in Nigerian Smallholder Systems: Patterns and Determinants

This study conducts a systematic meta-analysis of crop yield response to fertiliser application in Nigerian smallholder farming systems, identifying patterns of response magnitude and variability and determining the factors that explain yield response variation. Smallholder yield response to fertiliser is highly variable in sub-Saharan Africa, and Nigeria is no exception, yet the meta-analytical synthesis needed to extract reliable patterns from the accumulated experimental literature is absent. This study systematically searches and screens Nigerian agronomic fertiliser trial literature from 2000 to 2023, extracting standardised yield response data from 320 published experiments across six agro-ecological zones covering maize, rice, sorghum, millet, cassava, yam, and legumes. A random-effects meta-analysis framework is applied to compute mean yield response and response variability by crop, agro-ecological zone, fertiliser type, and management context. Meta-regression identifies determinants of response variation. Findings reveal a mean yield response to fertiliser of 68 percent above unfertilised control across all crops and zones, with very high variability (coefficient of variation of 84 percent) indicating that aggregate means mask highly site-specific responses. Soil organic carbon below 1.2 percent, absence of liming on acid soils, and single basal application timing are the strongest negative modifiers of fertiliser response. Response is consistently higher in on-station compared to on-farm conditions, revealing a systematic 22 percent research-practice yield gap. The study contributes the first comprehensive meta-analysis of fertiliser response in Nigeria and recommends that extension recommendations incorporate zone-specific response modifiers rather than single-rate recommendations.

Keywords: meta-analysis, fertiliser yield response, Nigerian smallholder, agro-ecological zones, research-practice gap.