Flooding afflicts some 20 million hectares in Asia. As much as one-third of the rainfed lowland areas in sub-Saharan Africa are thought to be affected by submergence. Submergence stress is estimated to cause annual losses of up to US$1 billion in Asia. Modern rice varieties are not adapted to these conditions. Farmers suffer from either regular yield loss when they grow these varieties or low yield when they continue to cultivate local landraces.


New major QTLs associated with stagnant flooding (SF) fine-mapped, validated in two backgrounds

QTL analysis using the MAGIC population was finalized for traits related to SF and control conditions. A total of four QTLs for grain yield under SF were identified, three of which showed effect in DS2017 and across seasons. One QTL showed effect across seasons as well as in both WS2016 and DS2017. This QTL, however, was not specific to SF and also led to yield advantage under non-stress conditions. A QTL with effect on tiller number was identified in DS2017 which also showed effect on mean tiller number across both seasons. Similarly, one QTL for percentage yield reduction as compared to non-stress was identified. This QTL also showed effect in DS2017 and on the mean across both seasons. QTL positive lines from this MAGIC population are available for inclusion in breeding programs. Genome-wide association studies using 3K lines have allowed the identification of QTLs related to SF tolerance. QTLs were identified for traits such as stem elongation, plant height, elongation rate, tiller number, yield, biomass, and harvest index. Only a few QTLs showed effect across seasons, yet these still make valuable candidates for allele mining and identification of SNPs associated with these traits. Previously, the QTL qGY5.1 was identified using a mapping population derived from a cross of Ciherang-Sub1 with IR10F365. This QTL showed effect on multiple traits related to SF including grain yield. These are already included in the breeding program.

Physiological traits necessary for SF tolerance identified

In a study to characterize physiological response to SF, four genotypes were selected from earlier screenings – two tolerant (IRRI 119, inherently tall and with SUB1 locus, and IRRI 154, intermediate stature) and two sensitive (PSBRc18-Sub1 and Swarna-Sub1, both short stature) – were screened under SF conditions in which flood depth was increased at 1.5 cm per day up to 50 cm at day 28. SF significantly increased shoot growth and decreased tillering, root growth, biomass, leaf area, survival, yield, and carbohydrate concentration in shoot but to a lesser extent in tolerant genotypes. Decline in tillering commenced on day 14 of SF (~30 cm water depth). Biomass relative to the control shifted towards the stem (7%** increase), consequently reducing the proportions for root biomass (16%** decline). Sugar partitioning moved toward the stem (20%** increase), with corresponding reductions in leaves and roots. Though stem aerenchyma is constitutively formed in tolerant genotypes under control conditions, SF treatment induced profound aerenchyma formation in the first internodes. The SUB1 gene alone was a disadvantage under SF. In the screening of 189 MAGIC-Plus lines (AC 2.1.2), the most significant GWAS peak was for panicle number (on chr 8).

Gene discovery and Marker-Assisted Selection (MAS)

Forty-eight (48) Sub-1 introgression lines were developed in the background of three varieties from Madagascar (B453-F-SR-6-3, FOFIFA160, and IR15579-24-52). Three promising submergence tolerance lines derived from the B453-F-SR-6-3*4/IRRI 119 cross with high level of fertility and flood tolerance traits were identified in flood-prone areas in Antsirabe and Mahitsy in Madagascar. During the trial, six separate flooding events occurred, with continuous flooding ranging from 3 to 7 days. Overall, the plants were under water for a total of 30 days. Twenty-nine (29) new Sub-1 introgression lines in two other backgrounds (FOFIFA162 and BOTRY) were developed for evaluation in Madagascar. The QTL qCEA1 on chromosome 1, with beneficial allele from NERICA L 52, has been introgressed into NERICA L 19 which resulted in the development of 12 BC4F2 and BC3F3 lines. Three mapping populations were developed for stagnant flooding (SF) tolerance and one population for submergence tolerance.

Breeding lines for AG tolerance evaluated in the field

The qAG7-3 fine mapping results from 2017 have been confirmed and 15 near-isogenic lines (NILs) with the fine mapped region of the QTL were identified and are currently being tested in the field for AG and dry direct seeding (DDS). Meanwhile, 102 lines segregating for the identified segment of the QTL from the previous season are being screened in table set-up to further reduce the region. In a separate study, three QTLs for germination under anaerobic conditions and one for seedling height were identified across two different mapping populations. These two populations were derived by crossing the tolerant parent Kalaratta with two elite varieties NSIC Rc238 and NSIC Rc222. The three QTLs identified for AG showed effect under controlled table screening as well as in field-like condition in the screenhouse. The seedling height QTL located at a different position from the sd1 gene as previously seen in several studies. This QTL provides an opportunity to increase seedling height without disturbing the sd1 region for better adaptation to not only AG but also to SF.

At least 100 new landraces and genebank accessions screened every year to identify new sources of tolerance for submergence, stagnant flooding (SF), and anaerobic germination (AG)

More than 1,500 lines from the 3K Rice Genome Panel were screened for AG in 2017 and a total of 22 lines with high tolerance were identified. Genome-wide association study (GWAS) was conducted and two major peaks were identified. Under SF, a subset (400 lines) of the 3K lines was screened to identify donor lines. Traits related to yield under SF such as tiller number, panicle number, plant height, flowering, and survival were recorded along with grain yield per se. A set of 13 donor lines with high survival and tiller number was identified and GWAS analyses are currently being finalized. The lines identified through both studies are valuable donors in the development of multi-parent mapping populations for confirmation of GWAS results and identification of new QTLs.

Selected lines under prolonged submergence evaluated at NARES sites experiencing longer submergence durations

A total of 1,440 lines identified from extended submergence screening (30 days submergence + 7 days slow draining) in DS2017 were tested in replicated trials in WS2018. The lines were kept under water for 27 days followed by slow draining of 5 days which kept them submerged for 30 days. A total of 49 lines were identified as survivors from this experiment. The seeds of these lines were increased in WS2018 and 49 families comprising a total of 694 F6 lines were developed.

At present, these families are being checked for performance under normal submergence and the speed of their recovery is being tested. The selected lines will be tested for yield potential under non-stress in the current season and shared to the partners. These lines will also be used to develop mapping populations to identify QTLs underlying extended submergence tolerance and fast recovery.

A set of high-value diagnostic candidate gene-based markers for use in marker-assisted breeding and fine-mapping schemes for one AG QTL developed

Diagnostic markers are available for the SUB1, AG1, and AG2 QTLs and are being used for NIL development and breeding on a regular basis. Further, fine-mapping of the AG3 region has allowed a more precise localization of the QTL region for which SNPs have been shortlisted. Furthermore, SNPs associated with the two new QTLs identified through GWAS are available for validation and marker design.