Comparative Analysis of Drought Tolerance in Maize Landraces

Thesis Abstract

Drought stress is a principal constraint to maize production in semi-arid regions, undermining yield stability and smallholder livelihoods, and highlighting the need to understand genetic and phenotypic mechanisms underpinning drought tolerance. This study addresses the gap in comparative assessments of drought tolerance among diverse maize landraces, aiming to identify reliable phenotypic and genotypic indicators that differentiate tolerant from sensitive lines under water-limited conditions. The primary objective is to (i) quantify drought-induced variations in agronomic performance across a representative panel of 120 maize landraces sourced from drought-prone agroecologies, (ii) associate morpho-physiological traits with grain yield under controlled drought and well-watered environments, (iii) evaluate heritability and genetic correlations of key drought tolerance traits, and (iv) develop a predictive selection index integrating stable indicators for rapid screening in breeding programs. The study adopts a comparative cross-sectional design conducted over two growing seasons at the Maize Research Farm of the National Institute of Agricultural Research, with a split-plot arrangement water regime (drought-stressed vs. irrigated) as the main plot and landrace genotype as the sub-plot, replicated thrice to ensure robustness. The population comprises landraces collected from arid and semi-arid zones with wide genetic diversity, while the sample includes 120 distinct lines plus three standard checks. Data collection instruments encompass standardized agronomic measurement protocols, high-throughput phenotyping for leaf relative water content, stomatal conductance, chlorophyll fluorescence (Fv/Fm), grain yield, kernel number, kernel weight, root length density, and anthesis-silking synchrony; complemented by molecular data from simple sequence repeat (SSR) markers to support genetic association analyses. Validity and reliability are ensured through calibration of instruments, inter-observer checks, and pilot testing; data quality is maintained via double data entry and random audit of measurements. Descriptive statistics, followed by analysis of variance (ANOVA) under both water regimes, will test genotype-by-environment interactions. Genotypic and phenotypic variances will be partitioned to estimate heritability on a broad-sense scale; Pearson correlation and path analysis will elucidate direct and indirect effects of drought-related traits on grain yield. Multivariate analyses, including principal component analysis (PCA) and cluster analysis, will reveal trait syndromes associated with tolerance. A mixed-model approach, incorporating genotype as random effects and environment as fixed effects, will be employed to refine estimates of genetic parameters. Marker-trait associations will be explored using a genome-wide association study (GWAS) framework with a binomial mixed model to account for population structure and kinship, leveraging SSR-derived genetic relatedness to corroborate phenotypic findings. The theoretical basis draws on drought adaptation theories, including the leaf-rolling and stay-green paradigms, and aligns with the ideotype concept of efficient water use and resource partitioning. Expected findings include identification of landraces exhibiting stable grain yield under drought with parallel maintenance of root system architecture, osmotic adjustment, and delayed leaf senescence; several traits (e.g., high root length density, high harvest index, and favorable stay-green score) are anticipated to show strong direct effects on yield under water limitation. The study contributes to knowledge by (i) providing a comparative, empirically derived ranking of maize landraces for drought tolerance, (ii) clarifying trait combinations that underpin tolerance across diverse environments, and (iii) delivering a practical, multi-trait selection index for accelerated breeding of drought-resilient maize. The main conclusion is that drought tolerance in maize landraces results from coordinated trait syndromes rather than single-trait effects, and that reliable screening requires integrating physiological measurements with genetic markers. Recommendations include adopting the proposed selection index in breeding pipelines, validating identified tolerant landraces across additional agro-ecologies, and expanding genomic resources with high-density SNP data to enhance GWAS resolution and genomic selection for drought resilience.