Genome-wide and haplotype mapping reveals genetic drivers of biomass productivity per unit water and nitrogen in sorghum

Background Improving crop productivity per unit of water and nitrogen is central to climate resilient sorghum improvement, yet the genetic architecture of biomass-based resource productivity remains poorly resolved across contrasting production environments. This study integrated multi-environment phenotyping with 1.75 million SNPs derived from whole genome resequencing of 172 cultivated Sorghum bicolor and S. bicolor × S. halepense derivative lines evaluated across temperate Mediterranean and semi-arid tropical environments. Results Biomass-based water productivity and partial factor productivity of applied nitrogen displayed extensive genetic variation and high broad-sense heritability of 0.75 to 0.81, whereas grain-based indices were more strongly influenced by genotype × environment interaction. Multi-model genome wide association analyses identified 27 significant SNPs defining 22 genomic regions for biomass-based productivity. A recurrent chromosome 9 signal at S09_55037491 was detected across three analytical models and biomass-derived indices and colocalized with Sobic.009G200500, a nitrate-responsive HAT3.1 homeobox transcription factor. Haplotype analysis independently prioritized two additional loci: a chromosome 9 haplotype associated with a 2.7-fold difference in biomass-based water productivity and containing an AHP-like cytokinin-signaling gene, and a chromosome 5 haplotype associated with a 1.86-fold difference and encompassing DAD1, a phospholipase implicated in jasmonate biosynthesis. Candidate gene prioritization also implicated aquaporin-mediated transport, trehalose-6-phosphate metabolism, oxidative signaling and, provisionally, C4 carbon assimilation. Most favorable common haplotypes occurred within cultivated germplasm rather than the introgressed lines. Conclusions Together, these findings reveal genetically tractable loci and biological pathways underlying sorghum biomass productivity per unit of supplied water and nitrogen. Following independent validation, the identified SNPs and haplotypes could support diagnostic marker development and haplotype informed selection, while genomic selection remains more appropriate for environmentally sensitive grain-based productivity traits.