The global population is projected to increase significantly in the coming decades, necessitating a substantial rise in crop production to meet the growing food demand. Rice, a staple food for a large portion of the world's population, has seen yield increases in the past, primarily through the use of semi-dwarf varieties and heterosis. However, recent years have witnessed a stagnation in global rice yields. Asian rice cultivars are mainly categorized into two subspecies: *O. sativa* L. ssp. *japonica* and *indica*. These subspecies display distinct morphological and physiological characteristics, most notably a significant difference in lifespan, with *indica* exhibiting considerably earlier senescence. This early senescence in *indica* rice, an adaptation potentially driven by r-selection in tropical environments favoring rapid reproduction, is detrimental in temperate and sub-tropical regions with high population densities. Early senescence limits grain-filling time, resulting in reduced productivity. While these subspecies are extensively used in breeding programs, the impact of senescence has largely been overlooked. Senescence, the final developmental stage of plants, is a complex process involving the orderly degradation and remobilization of cellular components. It directly affects grain yield and quality in crops. Delaying senescence extends the photosynthetic period, offering a significant opportunity to enhance crop yields. Stay-green traits, characterized by impaired chlorophyll catabolism, are positively correlated with increased grain yield in certain crops. However, previous studies of rice *ossgr* mutants showed no yield advantage, suggesting that they are non-functional stay-green mutants. While loss-of-function mutants have helped characterize chlorophyll breakdown pathways, the impact of natural variations within these pathways remains to be fully explored. This research aims to identify the genetic factors responsible for the different senescence patterns and lifespans between *indica* and *japonica* rice cultivars using QTL mapping, focusing on the role of promoter polymorphisms in the *OsSGR* gene. The findings are expected to provide insights for the development of improved rice cultivars with enhanced yields.

Previous research has established the significance of senescence in determining crop yield. Studies on maize and sorghum have demonstrated a positive correlation between delayed senescence (stay-green traits) and increased grain yield. Genetic studies have identified stay-green genes in various species, such as *sgr* mutants in rice and *Arabidopsis*. However, early studies on rice *ossgr* mutants revealed a lack of yield increase, indicating that these mutants exhibit a non-functional stay-green phenotype. Understanding the mechanism behind this non-functional stay-green phenotype is crucial. Work on chlorophyll degradation pathways using loss-of-function mutants has illuminated the biochemical processes involved. Moreover, genome-wide association studies have recently uncovered new alleles related to chlorophyll content and stay-green traits in both japonica and indica rice cultivars. These studies provide a foundation for this research focusing on uncovering the genetic basis for the different senescence patterns between indica and japonica rice cultivars and identifying potential targets for breeding programs.

This study employed a multifaceted approach combining QTL mapping, map-based cloning, gene expression analysis, and evolutionary analyses to investigate the genetic basis of senescence differences between rice subspecies. For QTL mapping, F2:3 populations derived from a cross between *indica* cultivar IR72 and *japonica* cultivar JN were used to identify quantitative trait loci (QTLs) associated with leaf senescence. Senescence phenotypes were assessed through visual observation of leaf color changes and quantitative measurements of chlorophyll levels using a chlorophyll meter and an Automated Colourimetric Assay (ACA). The QTLs identified were then fine-mapped to pinpoint the responsible genes. Map-based cloning focused on a significant QTL located on chromosome 9, leading to the identification of *OsSGR* (LOC_Os09g36200) as a candidate gene. To confirm the role of *OsSGR*, *OsSGR* knockout mutants were generated in the *indica* cultivar Kasalath using CRISPR/Cas9 genome editing, and activation-tagging lines of OsSGR and OsPRR95 were isolated from a japonica cultivar. The senescence patterns of these mutants were then compared with wild-type plants. To investigate the molecular mechanism underlying differential senescence, genomic polymorphisms between the OsSGR alleles of the parental cultivars were analyzed, including both coding and promoter regions. The expression levels of OsSGR during senescence were examined using qRT-PCR in both indica and japonica cultivars. Transient expression assays in rice protoplasts were conducted to compare the activity of OsSGR promoter segments from indica and japonica cultivars. To investigate the broader role of OsSGR promoter polymorphisms in natural populations, an association analysis was performed between OsSGR promoter region polymorphisms and leaf senescence phenotypes in 105 rice accessions belonging to different subgroups. The phylogenetic relationship between OsSGR promoter sequences from these accessions and wild rice species was determined. Near isogenic lines (NILs) were created to evaluate the effect of replacing indica *OsSGR* alleles with *japonica* alleles in elite indica cultivars. The photosynthetic competence of the NILs was evaluated by measuring net CO2 assimilation rates and Fv/Fm ratios. Relative growth rates (RGR) were calculated to quantify plant growth speed. Finally, grain yield and grain-filling rate were determined in the NILs to assess the impact of *OsSGR* alleles on productivity.

The study identified promoter variations in the *OsSGR* gene as a major determinant of differential senescence patterns between indica and japonica rice cultivars. QTL mapping revealed a strong association between a locus on chromosome 9 and leaf senescence. Fine-mapping of this locus pinpointed *OsSGR* as the key gene. *Indica* cultivars showed higher and earlier induction of *OsSGR* compared to *japonica* cultivars due to promoter variations. CRISPR/Cas9-generated *ossgr* knockout mutants in *indica* rice exhibited delayed senescence, confirming the role of *OsSGR* in accelerated senescence. Transient expression assays revealed higher promoter activity in the *indica* *OsSGR* promoter compared to the *japonica* promoter. Association analysis across 105 rice accessions confirmed that *OsSGR* promoter polymorphisms broadly influence senescence, lifespan, and life cycle. Phylogenetic analysis indicated that the *indica*-type promoter was acquired early during the evolution of rapid cycling in rice subspecies from its progenitor *O. nivara*. Near isogenic lines (NILs) with the *japonica* OsSGR allele showed delayed senescence, increased photosynthetic competence, and significantly higher grain yield compared to their parental indica cultivars. Conversely, introducing the *indica* allele into a japonica cultivar resulted in earlier senescence and lower grain yield. Analysis of the OsSGR promoter revealed differences in Dof-binding motifs between indica and japonica promoters, suggesting a potential regulatory mechanism.

The findings highlight that naturally occurring variations in the *OsSGR* promoter, not the coding sequence, are primarily responsible for the differences in senescence and lifespan between indica and japonica rice. The early and high induction of OsSGR in indica rice, driven by promoter variations, accelerates leaf and panicle senescence, leading to a shorter lifespan but faster grain filling, which may have been advantageous in its original environment. The introgression of the japonica OsSGR allele into high-yielding indica cultivars resulted in a significant yield improvement, indicating that delaying senescence can be a highly effective strategy to increase grain yield. The results suggest that the japonica OsSGR allele allows for extended photosynthetic competence during senescence, which contributes to higher grain yields. This is in contrast to non-functional stay-green mutants that do not show yield advantage, and highlights the importance of the specific timing and level of OsSGR expression. The contrasting effects of indica and japonica *OsSGR* alleles on senescence and yield are likely related to different cultivation practices and cropping systems of indica and japonica rice, suggesting that selection pressures on senescence related traits are influenced by cropping systems. The findings have significant implications for rice breeding programs, offering a novel strategy to improve yields through manipulation of *OsSGR* promoter polymorphisms.

This research identifies promoter variations in the *OsSGR* gene as a key factor in controlling senescence and lifespan in rice. The *indica* allele's promoter leads to early and strong OsSGR expression, resulting in accelerated senescence, while the *japonica* allele's promoter delays senescence, leading to higher yield. Introgression of the *japonica* allele into elite indica cultivars significantly increased grain yield. These findings suggest a promising strategy for improving rice productivity through marker-assisted selection and gene editing targeting the OsSGR promoter.

The study focused primarily on a limited set of rice cultivars and accessions. Further investigation is needed to assess the generalizability of the findings across a wider range of genetic backgrounds and environmental conditions. While the association analysis across a diverse panel of rice accessions supported the role of OsSGR promoter polymorphisms, the specific regulatory mechanisms mediating the differential expression of OsSGR remain to be fully elucidated. Additional studies are needed to explore other potential QTLs that might contribute to the differences in senescence and to investigate the interactions of OsSGR with other genes and environmental factors influencing senescence. Finally, the field experiments were conducted in a specific location. More tests in a wider range of conditions would be helpful to fully establish the universality of the results.