Direct selection for yield under drought has resulted in the release of a number of drought-tolerant rice varieties across Asia. In this study, we characterized the physiological traits that have been affected by this strategy in breeding trials across sites in Bangladesh, India, and Nepal. Drought- breeding lines and drought-tolerant varieties showed consistently longer flag leaves and lower stomatal density than our drought-susceptible check variety, IR64. The influence of environmental parameters other than drought treatments on leaf traits was evidenced by close grouping of treatments within a site. Flag-leaf length and width appeared to be regulated by different environmental parameters. In separate trials in the Philippines, the same breeding lines studied in South Asia showed that canopy temperature under drought and harvest index across treatments were most correlated with grain yield. Both atmospheric and soil stress strengthened the relationships between leaf traits and yield. The stable expression of leaf traits among genotypes and the identification of the environmental conditions in which they contribute to yield, as well as the observation that some breeding lines showed longer time to flowering and higher canopy temperature than IR64, suggest that selection for additional physiological traits may result in further improvements of this breeding pool.

Low light intensity is a great limitation for grain yield and quality in rice. However, yield is not significantly reduced in low light tolerant rice varieties. The work therefore planned for comparative transcriptome profiling under low light stress to decipher the genes involved and molecular mechanism of low light tolerance in rice. At active tillering stage, 50% low light exposure for 1 day, 3 days and 5 days were given to Swarnaprabha (low light tolerant) and IR8 (low light sensitive) rice varieties. Illumina (HiSeq) platform was used for transcriptome sequencing. A total of 6,652 and 12,042 genes were differentially expressed due to low light intensity in Swarnaprabha and IR8, respectively as compared to control. CAB, LRP, SBPase, MT15, TF PCL1 and Photosystem I & II complex related gene expressions were mostly increased in Swarnaprabha upon longer duration of low light exposure which was not found in IR8 as compared to control. Their expressions were validated by qRT-PCR. Overall study suggested that the maintenance of grain yield in the tolerant variety under low light might be results of accelerated expression of the genes which enable the plant to keep the photosynthetic processes moving at the same pace even under low light.

A panel of 60 genotypes comprising New Plant Types (NPTs) along with indica, tropical and temperate japonica genotypes was phenotypically evaluated for four seasons in irrigated situation for grain yield per se and component traits. Twenty NPT genotypes were found promising with an average grain yield varying from 5.45 to 8.8 t/ha. A total of 85 SSR markers were used in the study to identify QTLs associated with grain yield per se and related traits. Sixty-six (77.65%) markers were found to be polymorphic. The PIC values varied from 0.516 to 0.92 with an average of 0.704. A moderate level of genetic diversity (0.39) was detected among genotypes. Variation to the tune of 8% within genotypes, 68% among the genotypes within the population and 24% among the populations were observed (AMOVA). This information may help in identification of potential parents for development of transgressive segregants with very high yield. The association analysis using GLM and MLM models led to the identification of 30 and 10 SSR markers associated with 70 and 16 QTLs, respectively. Thirty novel QTLs linked with 16 SSRs were identified to be associated with eleven traits, namely tiller number (qTL-6.1, qTL-11.1, qTL-4.1), panicle length (qPL-1.1, qPL-5.1, qPL-7.1, qPL-8.1), flag leaf length (qFLL-8.1, qFLL-9.1), flag leaf width (qFLW-6.2, qFLW-5.1, qFLW-8.1, qFLW-7.1), total no. of grains (qTG-2.2, qTG-a7.1), thousand-grain weight (qTGW-a1.1, qTGW-a9.2, qTGW-5.1, qTGW-8.1), fertile grains (qFG-7.1), seed length-breadth ratio (qSlb-3.1), plant height (qPHT-6.1, qPHT-9.1), days to 50% flowering (qFD-1.1) and grain yield per se (qYLD-5.1, qYLD-6.1a, qYLD-11.1).Some of the SSRs were co-localized with more than two traits. The highest co-localization was identified with RM5709 linked to nine traits, followed by RM297 with five traits. Similarly, RM5575, RM204, RM168, RM112, RM26499 and RM22899 were also recorded to be co-localized with more than one trait and could be rated as important for marker-assisted backcross breeding programs, for pyramiding of these QTLs for important yield traits, to produce new-generation rice for prospective increment in yield potentiality and breaking yield ceiling.

Phenomics technologies have advanced rapidly in the recent past for precision phenotyping of diverse crop plants. High-throughput phenotyping using imaging sensors has been proven to fetch more informative data from a large population of genotypes than the traditional destructive phenotyping methodologies. It provides accurate, high-dimensional phenome-wide big data at an ultra-super spatial and temporal resolution. Biomass is an important plant phenotypic trait that can reflect the agronomic performance of crop plants in terms of growth and yield. Several image-derived features such as area, projected shoot area, projected shoot area with height constant, estimated bio-volume, etc., and machine learning models (single or multivariate analysis) are reported in the literature for use in the non-invasive prediction of biomass in diverse crop plants. However, no studies have reported the best suitable image-derived features for accurate biomass prediction, particularly for fully grown rice plants (70DAS). In this present study, we analyzed a subset of rice recombinant inbred lines (RILs) which were developed from a cross between rice varieties BVD109 × IR20 and grown in sufficient (control) and deficient soil nitrogen (N stress) conditions. Images of plants were acquired using three different sensors (RGB, IR, and NIR) just before destructive plant sampling for the quantitative estimation of fresh (FW) and dry weight (DW). A total of 67 image-derived traits were extracted and classified into four groups, viz., geometric-, color-, IR- and NIR-related traits. We identified a multimodal trait feature, the ratio of PSA and NIR grey intensity as estimated from RGB and NIR sensors, as a novel trait for predicting biomass in rice. Among the 16 machine learning models tested for predicting biomass, the Bayesian regularized neural network (BRNN) model showed the maximum predictive power (R² = 0.96 and 0.95 for FW and DW of biomass, respectively) with the lowest prediction error (RMSE and bias value) in both control and N stress environments. Thus, biomass can be accurately predicted by measuring novel image-based parameters and neural network-based machine learning models in rice.