Abstract Meeting current rice demand in sub-Saharan Africa (SSA) requires narrowing yield gaps on currently available agricultural land. The objectives of this study were to decompose rice yield gaps into efficiency, resource and technology yield gaps and to identify priority areas for research and development in the major rice production systems (irrigated lowland, rainfed lowland, and rainfed upland) in SSA. Data were collected during the 2012–2015 wet seasons on soil properties, field operations and yields in 1529 fields at 34 sites in 20 countries using a standardized protocol. Stochastic frontier analysis using data on biophysical environment and fertilizer management practices together with a crop simulation model (ORYZA2000) was used to quantify the yield gap, and efficiency, resource, and technology yield gaps. Cluster analysis was performed to classify the site-production system combinations into yield gap groups. Actual rice yields were on average 3.8, 2.6 and 1.7 t/ha in irrigated lowland, rainfed lowland, and rainfed upland, respectively. The yield yap ranged from 2.0–10.0 t/ha across site-production system combinations while the efficiency, resource, and technology yield gaps varied between 0.9 to 5.7, 0.1 to 2.3 and 0 to 7.5 t/ha, respectively. On average, efficiency, resource, and technology yield gaps accounted for 23, 5 and 37 % of the benchmark yield (potential yield in irrigated lowland or water-limited potential yield in rainfed lowland and upland). Four yield gaps groups were identified and were related to the production systems, soil properties, and fertilizer application. Narrowing yield gaps requires the dissemination of integrated crop management practices in yield gaps groups with a large efficiency yield gap, whereas, in yield gaps groups with a large technology yield gap, the development of technologies to improve soil properties and fertilizer use should be given priority.

中文翻译：

---

将撒哈拉以南非洲水稻产量差距分解为效率、资源和技术产量差距

摘要 满足撒哈拉以南非洲 (SSA) 当前的大米需求需要缩小当前可用农田的产量差距。本研究的目的是将水稻产量差距分解为效率、资源和技术产量差距，并确定 SSA 主要水稻生产系统（灌溉低地、雨养低地和雨养高地）的优先研发领域。在 2012-2015 年雨季期间，使用标准化协议收集了 20 个国家 34 个地点的 1529 个田地的土壤特性、田间作业和产量数据。使用生物物理环境和肥料管理实践数据以及作物模拟模型 (ORYZA2000) 的随机前沿分析被用于量化产量差距以及效率、资源和技术产量差距。进行聚类分析以将场地-生产系统组合分类为产量差距组。灌溉低地、雨养低地和雨养高地的实际水稻产量平均分别为 3.8、2.6 和 1.7 吨/公顷。整个场地-生产系统组合的产量 yap 范围为 2.0-10.0 t/ha，而效率、资源和技术产量差距分别在 0.9 到 5.7、0.1 到 2.3 和 0 到 7.5 t/ha 之间变化。平均而言，效率、资源和技术产量差距占基准产量的 23%、5% 和 37%（灌溉低地的潜在产量或雨育低地和高地的限水潜在产量）。确定了四个产量差距组，它们与生产系统、土壤特性和施肥有关。