Degree
Doctor of Philosophy (PhD)
Department
Plant, Environmental Management & Soil Sciences
Document Type
Dissertation
Abstract
Genetic improvement depends on separating favorable breeding targets from unfavorable linked variation and characterizing traits whose phenotypic expression is strongly influenced by environmental conditions. This dissertation addressed these challenges through three studies focused on rice grain yield and milling quality.
The first study integrated genome-wide association analysis, haplotype characterization, pedigree reconstruction, and local ancestry inference to examine variation surrounding the Green Revolution gene sd1. A Presidio-derived haplotype within the 34.22–38.11 Mb interval upstream of sd1 was associated with an average grain-yield reduction of 0.92 t ha⁻¹ in a multiparent population. This effect was independently validated in a Presidio × Catahoula population, in which the haplotype reduced yield by 0.6 t ha⁻¹. The absence of a significant yield effect in the subregion containing sd1 demonstrated that the yield penalty was associated with linked variation rather than the semi-dwarf allele itself, providing evidence of linkage drag surrounding this historically important gene.
The second study evaluated whether fitting the chromosome 1 yield-associated haplotype as a fixed effect improved genomic prediction. Haplotype-fixed RR-BLUP models produced modest improvements during within-population cross-validation, but these gains were not consistent across populations and years. The unfavorable Hap2 haplotype occurred at low frequency and was absent among the independent prediction candidates. Thus, its primary breeding value may be for parent and germplasm selection rather than for improving genomic prediction.
The third study developed a controlled-stress framework for evaluating fissure-related milling stability. A sequential heat, humidity, and drying treatment reduced mean head rice yield from 52.04% to 35.95% and differentiated genotypic responses. Silica desiccation caused greater measured bulk moisture loss than prolonged heat but produced less milling loss, demonstrating that bulk moisture loss alone did not explain milling response. Although the treatment differentiated genotypes under standardized stress, correspondence with field responses differed between years and did not establish field predictive ability.
Together, these studies identify yield-reducing linked variation surrounding sd1, define its practical role in breeding decisions, and establish a controlled framework for characterizing fissure-related milling response.
Date
10-23-2026
Recommended Citation
Ferreira Furlan, Flavia Juliana, "Uncovering Hidden Constraints on Rice Yield and Grain Quality: Linkage Drag Associated with sd1 and Assessment of Kernel Fissure Susceptibility" (2026). LSU Doctoral Dissertations. 7218.
https://repository.lsu.edu/gradschool_dissertations/7218
Committee Chair
Famoso, Adam
LSU Acknowledgement
1
LSU Accessibility Acknowledgment
1