Management adaptations for water-limited pearl millet systems in Senegal

Abstract

Pearl millet (Pennisetum glaucum L.) is a staple food crop for rural households in Senegal. However, yields remain low due to reduced precipitation and low adoption of agricultural technologies. This study aims to (i) calibrate and evaluate the performance of the Agricultural-Production-System-sIMulator (APSIM)-Millet model to simulate crop phenology and grain yield, and (ii) develop regional-specific recommendations based on management scenarios for pearl millet in Senegal. The APSIM-Millet model was adjusted for early- and late-flowering landraces, with field-data from 351 locations during 2016–2021. Independent datasets were used to evaluate the agreement of crop model outcomes with (i) crop phenology (with a root mean square error, RMSE, of 2.4 days for early- flowering and 3.7 days for late-flowering landraces), and ii) grain yield (with average RMSE of 439 kg ha-1), across locations. For all Senegal millet producing districts, crop management scenarios were simulated by combining: six levels of nitrogen (N) fertilization, three plant-densities, and three planting-dates. These scenarios were compared against a literature-reported standard management. The implementation of the recommended management practices can increase farmer yields by two-fold in Senegal, mainly by increasing plant-density, N fertility, and by adjusting planting date to better match precipitation with the period of greater crop water demand.