Physiological, biochemical and genetic studies reveal differing responses of West Africa rice genotypes under induced upland field and greenhouse drought stresses

Background : Drought is the single most important factor limiting rainfed upland rice production in Africa and is aggravated by climate change leading to a serious food security threat in the continent. This study aimed to assess the impact of drought stress on morpho-physiological and biochemical traits among rice genotypes and quantify the genetic variability among the genotypes under both reproductive drought stress (RS) and non-stress (NS) conditions. Further, this study aimed to select drought-tolerant genotypes and identify traits to monitor during screening for drought under field and greenhouse conditions. Methods: To achieve these objectives, we first screened 100 genotypes under upland reproductive stage drought field conditions as well as nonstress conditions using alpha-lattice design with three replications. Out of the 100, we selected 14 genotypes based on yield and yield-related traits under both conditions (RS and NS) for further screening under drought in the greenhouse condition using completely randomized block design (RCBD) with three replicates. Data on yield-related traits, relative water content (RWC), malondialdehyde (MDA), proline and genetic diversity parameters were collected. Results : Under the field screening, for all the investigated traits, a general reduction was observed among the genotypes under drought stress compared to non-stress conditions. Traits such as spikelet fertility (SFP), grain yield per plant (GYP), hundred grain weight (HGW), leaf rolling score, and leaf drying score recorded high broad-sense heritability (>60%) combined with high genetic advance (>20%) implying that selection could be rewarding for these traits under drought stress. Grain yield was positively associated with most yield-related traits under both water regimes. Under the greenhouse screening, RWC of all genotypes was reduced by >14%, except G11 (1.58%), G99 (8.43%), and G100 (5.43%), while increased MDA of >45% for all genotypes was observed, except G11 (25.15%), G99 (39.37%), and G100 (43.65%) under drought. Conclusions : Based on the correlation analysis, principal component analysis, and the multi-trait genotype-ideotype distance index (MGIDI) using relative trait values, grain yield, spikelet fertility, grain length, MDA and RWC are recommended in descending order for monitoring during screening for drought tolerance. Overall, G60, G100, G99, and G11 were selected in chronological order as genotypes with enhanced tolerance to drought. Among these genotypes, G60 (Viwornor short) is glaberrima ecotype, while G100 (UPLR-17), G99 (APO), and G11 (CRI-Enapa) are indica ecotypes.