Genetic Studies and High Throughput Phenotyping for Nutritional Quality Traits in Urochloa and Its Implications for Forage Breeding

Abstract

Urochloa species, widely cultivated in the tropics as animal feed, present strong potential for integration into Uganda’s pasture-based livestock systems. Despite their importance, improvement and selection efforts in Sub-Saharan Africa remain constrained by limited and poorly characterized genetic diversity, inadequate high-throughput phenotyping tools, and insufficient genomic information resulting from the species’ reproductive complexities, including apomixis and variable polyploidy. These challenges have slowed the development of improved Urochloa varieties with enhanced productivity and forage quality. To address these gaps, the study aimed to: (i) assess genetic diversity and population structure of Ugandan Urochloa collections, (ii) develop and validate Near-Infrared Spectroscopy (NIRS) predictive models for rapid forage quality phenotyping, and (iii) identify genome-wide single nucleotide polymorphisms (SNPs) associated with key forage quality traits. Results from objective one unraveled significant morphological variability among genotypes, displaying diverse growth habits (decumbent to prostrate), leaf textures (hairy or hairless), and stem colors (purple or green). Ploidy levels ranged from diploid to hexaploidy. Genotyping 188 accessions using the DArTseq platform produced 19,668 quality SNPs with moderate polymorphism (PIC = 0.01–0.38). Population structure analysis identified six genetically distinct subpopulations, driven mainly by genetic rather than geographic differences, showing high gene flow and low differentiation (PhiPT = 0.00108), with 99% of variation occurring within subpopulations. Results from objective two demonstrated that near infrared spectrometry as an effective, high throughput phenotyping tool for forage quality evaluation. Using a 350–2500 nm spectral range and partial least squares regression (PLSR), the Acid Detergent Fiber (ADF) model showed the best predictive accuracy (R²cv = 0.93, RMSEP = 1.34, RPD = 3.6), followed by the Metabolizable Energy model (R²cv = 0.91, RMSEP = 0.12, RPD = 3.29). Other models such ash (R²cv = 0.92, RPD = 2.94), in vitro organic matter digestibility (R²cv = 0.87, RPD = 2.73), and Crude Protein (R²cv = 0.85, RPD = 2.57) all exceeded an RPD value of 2.5, confirming their reliability for routine phenotyping. Objective three generated genome-wide SNP data (19,668 DArTseq SNPs) with phenotypic data from 188 accessions across three environments. Using multi-locus models (Bayesian Information and Linkage Disequilibrium Iteratively Nested Keyway and Multiple Loci Mixed Linear Model) in GAPIT software, 32 SNPs associated with forage nutritional quality traits were detected. Five marker-trait associations (MTAs) were consistent across models, including one on chromosome 5 linked to in vitro digestibility, Neutral Detergent Fiber (NDF), and ADF, suggesting pleiotropic effects. Candidate genes identified included translation initiation factor 2 (protein synthesis regulation), ATP-binding protein kinases (stress response regulation), and domains such as IBR and kinase inducible domains. In conclusion, substantial genetic diversity exists within Ugandan Urochloa germplasm, and forage quality traits are shaped by both genetic and environmental factors. The identified SNPs, once converted into Kompetitive Allele-Specific PCR (KASP) markers, can support marker assisted selection for enhanced forage quality in Urochloa and related tropical grasses.