Modeling Climate Risk and Adaptation Strategies for Pearl Millet (Pennisetum glaucum (L.) R. Br.) Production in Senegal Using DSSAT

Abstract

This thesis explores how climate-smart technologies (CSTs) can enhance the resilience of dryland agricultural systems in Senegal, improving sustainability, productivity, and smallholder farmer incomes. It aims to generate science-based knowledge on climate-smart management practices for pearl millet production in Sub-Saharan Africa through experimentation, modeling, and participatory extension approaches (PREA). The study focuses on four key areas: first, examining the role of socioeconomic and institutional support in shaping community responses across three climate zones in Senegal; second, assessing the performance of micro-dosing fertilization strategies and plant populations on pearl millet production; third, evaluating the DSSAT model’s ability to simulate the response of millet varieties to these strategies; and fourth, predicting millet yields under different fertilization and planting densities scenarios in Senegal’s agroecological zones (AEZs) under current and future climate. Results show that most farmers recognize climate threats, with significant impacts from seasonal rainfall deficits (72%), delayed rainfall onset (88%), frequent dry spells (68%), and prolonged droughts (76%). Key drivers of CST adoption include access to credit, extension services, government subsidies, and advisor interactions, providing valuable insights for resilience-building strategies. The findings highlight that organo-mineral micro-dosing fertilizer (combining Senegal’s recommended mineral rate with 2.5 tons of organic manure), improved plant populations (25,000 hills/ha), and dual-purpose millet varieties consistently enhance yields across study areas. Long-term simulations confirmed these parameters’ effectiveness under both current and future climates. The CERES-Millet model accurately simulated four millet cultivars (Souna 3, SL 28, SL 423, and Thialack-2) with acceptable statistical index values. Future climate projection showed a decrease in rainfall across AEZs (up to 25%) and an increase in minimum and maximum temperatures for the mid-century. The results emphasized that using Senegal's recommended mineral fertilizer rate alone is more sensitive to future climate change, with yield declines ranging from 4% to 16% under SSP2–4.5 and from 8% to 33% under SSP5–8.5 by mid-century compared to organo-mineral doses. Higher planting densities were less sensitive to adverse climate impacts, demonstrating improved resource use efficiency and water retention compared to low-density practices, which exhibited a higher yield decline. Simulations indicate 6–17% yield increases compared to local varieties under various climate projections across AEZs. The effectiveness of interventions varies spatially, with northern regions showing higher vulnerability. Dual-purpose varieties, while effective under current conditions, are expected to decline by 1% to 13% under SSP2–4.5 and 6% to 27% under SSP5–8.5 in the future. These findings underscore the role of integrated soil fertility management (ISFM) and optimal plant density in mitigating climate effects. Smallholder farmers could adopt ISFM to enhance organic manure application combined with mineral fertilizers for sustainable productivity. This study highlights the importance of integrating high-resolution climate projections with agronomic modeling to guide adaptation strategies.