Rice (Oryza sativa L) has demonstrated the highest progresses in efficient genomics from the all crops in last recent decades. It has a small genome and diploid in nature comparison to any other cultivated cereals crop (Moin et al., 2017). However rice genotypes par-ades light genetic variability due to repeated use of similar genotypes. Appreciative the genetic basis and the function of a particular gene can help for breeder to develop new and more productive with submergence tolerance genotypes. Extreme levels of flash flood were announced in the river basins of South and Southeast Asia. The estimated annual economic loss is more than US$ 600 million. Submergence especially flash flood affected 15 million ha in South and South-East Asia region (Neeraja et al., 2007). 

Rice is only aquatic crops that are well developed aerenchyma tissues for that facilitate oxygen dif-fusion through continuous air spaces from shoot to root. However complete submergence due to frequent flash flood can adversely affected yield and other yield component. Rice is a main food in half of the world population and hence, is referred to as “Global Grain” (Prasad et al., 2018; Kumar et al., 2020) but its pro-duction chronologically decline because of submer-gence and other climate change. About more than 2.0 million ha land areas affected by flash flood and average yield about 5% loss in Bangladesh (Iftekharuddaula et al., 2011; Ray et al., 2018). Marker assisted selection (MAS) obviously proves the superiority by using MAB compared to conventional breeding because a small donor region only a few backcross population would be impossible using conventional methods (Ray et al., 2014). For this reason we have to develop high yielding submergence tolerant genotypes and Marker assisted selection (MAS) can play a vital role to dev-elop those genotypes. These researches will signify-cantly assist to introgression of Sub1 QTL into HYV rice cultivars through marker assisted breeding and evaluated high yield with highly tolerance to subm-ergence. Submergence is the most challenging abiotic stress in the Rainfed Lowland Rice (RLR) ecosystem among the rice growing ecosystem. The Sub1 QTL was about 70% on chromosome 9 of the phenotypic variation for survival under submergence has fine mapped on chromosome 9 and the cluster of genes underlying the QTL cloned (Karim et al., 2021; Xu et al., 1996; 2000; 2006). 

These QTL has introgressed into few HYV at the International Rice Research Institute (IRRI) (Iftek-haruddaula et al., 2011; Neeraja et al., 2007; Septin-ingsih et al., 2009; 2013; 2014) and Bangladesh Institute of Nuclear Agriculture (BINA). So far Bang-ladesh Rice Research Institute (BRRI) has been rele-ased two submergence tolerant varieties viz. BRRI dhan51 (Swarna-Sub1) and BRRI dhan52 (BR11-Sub1) and BINA has released another two submer-gence tolerant varieties viz. BINA dhan11 (Ciherang-Sub1) and BINA dhan12 (Samba Mahsuri-Sub1). These varieties can tolerate around two to three weeks of flash flood submergence. But Northern part of Bangladesh such as Rajshahi division and Rangpur division, duration of flash flood remain about 3 to 4 weeks. Now a days, already released varieties in Bang-ladesh cannot tolerate such type of stress level (3-4 weeks). Nonetheless, flash flood stress depends on duration submergence, turbidity, water temperature, pH, depth of water etc. With that case Mohanty et al. (2000) specified that submergence is polygenic trait and Sub1 QTL does not alone completely represent this trait. Even with 3-4 weeks of submergence tole-rance, current submerged varieties often damaged whenever environmental factors go beyond the toler-ance of the current varieties except for the immersion period. Thats why new submergence tolerant QTLs other than Sub1 with the representation of agron-omically suitable phenotypic variation need to be pyra-mided along with Sub1 QTL. So far, very limited sources of submergence tolerance accessions were detected, and one of them coming from a moderately tolerant variety, IR72 (Septiningsih et al., 2012). Many research studies have been recognized quantitative trait loci (QTLs) for submergence tolerance obtained from several populations (Nandi et al., 1997; Siangliw et al., 2003; Toojinda et al., 2003; Xu and Mackill, 1996; Septiningsih et al., 2012; Gonzaga et al., 2016; Ifte-kharuddaula et al., 2016). Among the discovered QTLs, the Sub1 QTL from chromosome 9 derived from FR13A is known to be the most significant, contributing up to 2–3 weeks of submergence toler-ance depending on the genetic background and the environmental conditions at the time of flooding. Sub1 was previously fine mapped, the FR13Aderivedline was sequenced for the Sub1 region, and a cluster of three ethylene response factor (ERF) genes, namely Sub1a, Sub1b, and Sub1c, were unveiled (Xu et al., 2000; Xu et al., 2006). Sub1A was found in a subset of indica and as accessions, while Sub1b and Sub1c were detected in all indica and japonica accessions screened for these genes (Xu et al., 2006; Li et al., 2010). More-over, Sub1A was confirmed as the causal gene-pro-viding submergence tolerance at the Sub1 QTL (Septiningsih et al., 2009; Singh et al., 2010; Xu et al., 2006). The activation of Sub1A under complete sub-mergence stimulates the expression of the slend-errice-1 (SLR1) and SLR like-1 (SLRL1) genes through the suppression of ethylene consequently inhibiting gib-berellic acid (GA)-mediated shoot elongation (Fukao and Bailey-Serres, 2008). Thereby starch and sugar of leaf consumption are intentionally slower compared to the intolerant lines lacking Sub1that quickly consume carbohydrate energy reserve for shoot elongation due to the activation of ethylene that signal GA accu-mulation. Hence, intolerant lines die after the water subsides because all the carbohydrate reserves were spent during submergence, while lines containing the tolerant Sub1A allele retain enough carbohydrate serves to recover after the water subsides (Fukao and Bailey-Serres, 2008) in Fig. 1. 

Rice materials

This experiment consists of 16 genotypes among them two submergence tolerant genotypes as BRRI dhan52, FR13A that have sub1QTL and 3 susceptible HYV as Binadhan-7, Binadhan-17, BRRI dhan49 varieties and 4 swarna genotypes such as Guti swarna, Mamun swarna, Bilati swarna and Panpata swarna and 7 cross combinations as Binadhan-7 x BRRI dhan 52 (F2), Bindhan-17 x BRRI dhan52 (F2),Guti swarna x BRRI dhan52 (F2), Panpata swarna x BRRI dhan52 (F2), Bilati swarnax BRRI dhan52 (F2), Mamun swarna x BRRI dhan52 (F2), BRRI dhan49 x FR13A(F2). Data on dm= Days to maturity, ph= Plant height (cm), tt = Total tillers/plant, et= Effective tillers/plant, pl= Panicle length (cm), fg= Filled grains/plant, ug= Unfi-lled grains/plant, sw=Thousand seed weight (g), Yld= Yield/plant (g) were recorded. 

Genotypic Analysis

The leaves segment (2-3 cm pieces) of every genotype for DNA extraction  completed by using the Cetyl Trimethyl Ammonium Bromide (CTAB) mini-prep method (IRRI. 1997) at Biotechnology laboratory, Bangladesh Institute of Nuclear Agriculture (BINA), Mymensingh Bangladesh. CTAB method described by Zheng et al. (1995) is modest and fast compared to the any other methods and no need liquid nitrogen (Ray et al., 2016). DNA qualities of this method were enough for PCR. PCR were accomplished by using the procedure of Chen et al. (1997) and the PCR product were visualized on 1.5% agarose gels. After initial deformation for 2 min at 94°C, each cycle has a 30 second denaturation at 94° C temperature, 30 second annealing at 55° C, and final minute extension for 5 minutes at 72° C at the end of 34 cycles. The PCR pro-duct were mixed with bromophenol blue gel loading dye and were analyzed by electrophoresis on 6% poly-acrylamide gel (PAGE) for the all SSR primers using mini vertical polyacrylamide gels for high throughout manual genotypes. The gels were stained in 0.5 mg/ml ethidium bromide and gel picture were taken using Molecular Imager gel documentation unit. SSR profi-ling were performed to tag the submergence tolerant QTL by using sub1 flanking markers (Table 7)

Data analysis

SSR primers each allele size were measured according to the molecular weight by using Alpha-Ease 5.5 soft-ware. Allele size of the SSR markers were exam-ined using the program Power Marker version 3.25 (Liu and Muse, 2005). The statistic of the primers measured with the counting the numbers of alleles per locus, major allele frequency, gene diversity, polymorphism information content (PIC) values were analyzed by using this software. Genetic distant coefficient  and a dendogram demonstrating the genetic relation bet-ween genotypes, based on the unweighted pair group method with arithmetic averages (UPGMA) were created by using same program and viewing in Tree View (MEGA software). The value of this data on dm= Days to maturity, ph= Plant height (cm), tt = Total tillers/plant, et= Effective tillers/ plant, pl= Panicle length (cm), fg= Filled grains/plant, ug= Un-filled grains/plant, sw= Thousand seed weight (g), Yld= Yield/plant (g) were recorded for each popu-lation of trial. Data analyses were done to calculate Analysis of variance (ANOVA), mean performance, correlation coefficients, Principal Components Ana-lysis (PCA), Dendogram by using the software ige-tintopc.com_Minitab_18.1 for yield and yields contri-buting characters of rice genotypes in Aman season, 2019. Data calculated on significance difference (P<0.001) that is *** = Significant at 0.01% level of probability, ** = Significant at 1% level of probability (P<0.01), * = Significant at 5% level of probability (P<0.05).

The analysis of variance supported for nine characters existed in Table 1 which showed that the genotypes differed significantly for that all same characters. Among the all parent genotypes the top mean value yield plant-1 (32.00gm) was produced by Panpata swarna and the bottom (21.00g) by Binadhan-17 and FR13A. On the other hand the group of Binadhan-7 x BRRI dhan52 (F2) crosses was higher yielding (38.00 g) than Guti swarna x BRRI dhan52 (F2) cross comb-ination (22.00g). Binadhan-7 x BRRI dhan52 (F2) crosses were possessed the highest grain yieldplant-1(38.00 g) which was significantly higher than its both parents (Table 2). Other high yielding cross com-bination of Bindhan-17 x BRRI dhan52 (F2) were 31.00 g and those were also significantly higher than their related parents. So, desired grain yeild plant-1were found from those high yielding cross combination Binadhan-7 x BRRI dhan52 (F2). Binadhan-7 x BRRI dhan52 (F2) and Bindhan-17 x BRRI dhan52 (F2) cross combinations were early maturing also as 124 day and 125 day correspond (Fig. 2). It was an ex-cellent observation that most of the high yielding combinations than their both parents. So, yield contributing gene(s) from that parent were transferred to the cross combinations successfully also. It was importantly noticed that the Binadhan-7 x BRRI dhan52 (F2) cross combination, highly significant heterosis over mid parent and better parent were also positively heterosis and significant in most case for the other yield contributing character or traits  such as number of effective tillers plant-1, filled grains panicle-1. Among the all crosses, Binadhan-7 x BRRI dhan52 (F2) were excellent considering overall performance. Binadhan-7 x BRRI dhan52 (F2) expressed preferable performance in case of days to maturity and showed significant in some important yield contributing characters such as for crosses the highest effective tillers plant-1 (14.93) was observed in Binadhan-7 x BRRI dhan52 (F2) combination, where the value was in desired direction and more than its both parents. Effective tillers plant-1, filled grains panicle-1, thousand grain weight (g) also. Therefore, these cross com-bination may be recommended for further research with their control varieties to desirable lines with high yield capabilities, high sub-mergence tolerance with earliness.

Dendrogram

UPGMA (Unweighted Pair Group Method of Arith-metic Mean) 

A dendrogram constructed based on Neis, (1983) genetic distance using UPGMA indicated segregation of 16 rice lines into two main groups. One of the main groups consists of two cluster, cluster4 and cluster5. Cluster4 contain one genotypes viz. FR13A and clus-ter5 contain BRRI dhan49 x FR13A (F2). Another main group consists of three clusters, cluster1, cluster 2 and cluster3. cluster3 consists of two sub clusters, sub cluster I consist of Mamun swarna, Bilati swarnax BR-RI dhan52 (F2), Mamun swarna x BRRI dhan52 (F2) and  sub cluster II consists of Bindhan-17 x BRRI dhan52 (F2) and Sub cluster IV consists of Cluster2 consists of two sub clusters, sub cluster III and sub cluster IV. Sub cluster III consists of BRRI dhan49, BRRI dhan52, Guti swarna x BRRI dhan52 (F2), Sub cluster IV consists of Guti swarna, Panpata swarna, Bilati swarna, Panpata swarna x BRRI dhan52 (F2) (Fig. 3). Binadhan-7, Binadhan-17, Binadhan-7 x BRRI dhan52 (F2) genotypes are in the same cluster1. Based on above result, it may be concluded that, the maxi-mum rice genotypes grouped in same cluster where the control genotype BRRI dhan52 is present and this may occur due to crosses  and higher genetic distance (Table 3). According to the rang of diversity, five cluster were grouped for 16 genotypes. The distri-bution pattern revealed maximum number of genotypes (7 genotypes) in cluster 2 while cluster 4 and cluster 5 included minimum numbers of genotypes (1genotype). Maximum intra-cluster degree of diversity was obser-ved in cluster4 (79.93) and minimum in cluster3 (31. 44). Maximum inter-cluster distance (74.56) was showed from cluster5 Minimum distance was found between the genotypes of the cluster3 (41.68) in Table 4.

Correlations of Phenotypic Parameters

The correlation co-efficient between yield and yield contributing characters in rice are presented in Table 5. The correlations between the phenotypes of the par-ental lines and the crossing populations were also analyzed (p < 0.05 and 0.01). Highly significant and positive correlations were found among the grain yield (GY) and Total tillers/plant, effective tillers/plant, pan-icle length and filled grains/panicle. In addition, non-significantly but negative correlations were observed the grain width (GW) and days to maturity, plant height, unfilled grains/panicle. Grain yield was posi-tively and non-significantly correlated with 1000 grains weight.

Here, dm= Days to maturity, ph= Plant height (cm), tt = Total tillers/plant, et= Effective tillers/plant, pl= Panicle length (cm), fg= Filled grains/plant, ug= Unfilled grains/plant, sw=Thousand seed weight (g), Yld= Yield/plant (g).

Trait association

Principal component analysis

The PCA of nine yield and yield related traits and their other contrition characters with the total genetic diver-gence were presented in Table 6. According to the suggestion of Brejda et al. (2000), data were con-sidered in each component with Eigen values more than 1, as it determines a minimum 10% of the vari-ation. Superior Eigen values are considered as best attributes in principle components. A loading plot graphically showed the important PCs that explained major variability obtained by drawing a graph between Eigen values and principle component numbers (Fig. 4). The result of the study showed that three com-ponents indicated Eigen values of greater than one. The first three principal component with Eigen values explained 73.7 % of the total variation among 16 rice genotypes for the 9 quantitative traits studied. These first three principle components PC1, PC2 and PC3 explained 37.0 %, 22.3 % and 14.5 % of data variation, respectively (Table 7).The first principal component (PC1) accounted for more than 39.4% of total variance. The results showed that number of effective tillers/plant had the highest positive loadings (0.506), followed by number of total tillers/plant (0.504), panicle length (0.426), Grain yield per plant (0.409) plant height (0.310), Thousand seed weight (0.150), days to maturity (0.121) and number of unfilled grains/plant (0.088). Only one characters showed negative loading in the first principal component (PC1) that is number of filled grains/plant (-0.001) (Fig. 4). The second principal component (PC2) accounted for more than 22.3 % of total variance whereby plant height (0.514), number of unfilled grains/plant (0.433), days to maturity (0.360), panicle length 0.056) ) exhibited positive loadings. On the contrary, yield/ plant (-0.415) total tillers/plant (-0.049), effective tillers/plant (-0.080), and. and thousand seed weight -0.201), number of filled grains/plant (-0.441), showed negative loading (Table 6). 

The third principal component (PC3) accounted for more than 14.5 % of total variance and days to maturity 0.627) panicle length (0.354), number of filled grains/plant 0.285), yield/-plant (0.076) exhibited positive loading. Besides, these showed negative loading Plant height (-0.138). Unfilled grains per panicle (-0.308) thousand seed weight (-0.502) number of total tillers/plant (-0.063) and number of effective tillers/plant (-0.158).

Marker analysis

Totally 56 simple sequence repeat (SSR) primer were hand-me-down for parental survey (IRGSP, 2005; Ma-Couch et al., 2002; Temnkh et al., 2001). All infor-mation of the SSR primers collected from the Gramene database (www.gramene.org). For the primer survey DNA band of  the all genotypes as FR13A,BRRI dhan52,Guti swarana, Mamun swarna, Bilati swar-na,Binadha-17, Binadhan-7 etc polymorphism were score according to their molecular weight on polya-crylamide gel electrophoresis. The band of the primer amplified at many level were measured as polymer-phic marker but those primers which produced same level band that is called monomorphic marker. On the other hand, those primers which is not produced any types of band were considered as not amplified. Notably, water was used as negative control for every primer. This strategy is very helpful to identify fake bands produced from any type of pollution or unexpected nucleotides in primer or master mix.

Confirmation of F1

For producing BC1F1 seeds F1 seeds were disinfected by treating with 0.1% Bavistin 50SP solution for one hour. Confirmation of crosses was done by comparing F1s with the parents as well as by marker assisted selection. At the same time backcrossing of F1 plants with the recipient parent was done. Leaves collection of F1 plants (1-14 plants) and their parents were completed for DNA extraction with marker assay. The   F1population confirmation of Binadhan-7 x BRRI dhan52 (F1) were completed by using the primer RM 1115. Plants number 1, 4, 13, and 14 were not hete-rozygous band but plants number 2,3,5,6,7,8,9,10, 11,12 were heterozygous band that is F1 confirmation were finished (Fig. 5). This confirmed F1 goneto BC1F1 & F2 generations. The amplicon size of all genotypes for RM1115 marker allele was measured and shown in Table 7. The number of alleles was 6 per locus. The gene diversity value was 0.7610 and the polymorphism information content (PIC) values were 0.7432 & allele frequencies (%) were0.3328.

Ahangar et al. (2008) found significant mean sum square for all the characters studied except 1000 grain weight. Average performance of the genotypes with two standard check varieties such as BRRI dhan51, BRRI dhan52 presented that lowest growth dur-ation was obtained from BRRI dhan52 (148 days). Plant height from 94 to 120cm. The highest grain yield (4.85 t/ha) was obtained from BRRI dhan52 followed by BRRI dhan51 (4.6 t/ha). The survival percentage were more than 90% from all geno-types showed that there was not much stress of flash flood of this experiments (Ray et al., 2013). Based on the above result, it may be concluded that, the highly tolerant and moderately tolerant germ-plasms grouped in different cluster due to higher genetic distance. The dendrogram also revealed that the genotypes that are derivative of genetically similar type clustered together. Tehrim et al. (2012) conducted a study with 35 rice cultivars and resolved a significant cluster analysis where most of the cultivars were close to each other showing a high level of genetic relate-dness. Chakravarti et al. (2006) classified the rice genotypes into 11 distinct groups. Cluster analysis of the 193 accessions parental lines of rice showed 3 major groups and nine subgroups (Ren et al., 2003). Lower levels of polymorphism in rice germplasms indicated that there is a basic similarity among the rice cultivars, which is to be expected due to their same ancestors and selection for similar characteristics. Vange et al. (2009), Yadav et al. (2010), Akinwale et al. (2011), Selvaraj et al. (2011) described also posi-tive and significant correlation of yield with other yield contributing traits. Prasad et al. (2001) and Nand-eswar et al. (2010) establish significant nut negative correlation of plant height with yield whereas Rangare et al. (2012) initiate non-significant and positive cor-relation of the traits with yield. But the findings of the experiments contradict with the results of Selvaraj et al. (2011) presented positive and significant corre-lation of yield with filled grain/panicle. Thousand grains weight (TGW) was negatively and significantly correlated with days to maturity and yield per plant. Similar result also obtained by Karad et al. (2008). These principal components have been seen contri-buting to 78.2632 percent per total variety (Kumar et al., 2020). Similar results were reported earlier in rice by Gana et al. (2013) and Nachimuthu et al. (2014). Prasad et al. (2001) and Ali et al. (2010) were reported that the number of alleles per locus ranged from 2 to 8 with an average 3.8 using BRRI released varieties. The comparable studies is to three previous estimates of microsatellite analysis in rice viz., 0.26 to 0.65 with an average of 0.47 (Singh et al., 2015), 0.28-0.50 with a mean of 0.45 (Umadevi et al. 2014) and 0.239 to 0.765 with an average of 0.508 (Hossain et al., 2012). 

More than 2.0 million ha land area are affected by different types of flash flood and reduce average 5% reduces yield in Bangladesh. This is to introgression of Sub1QTL into the genetic HYV background rice by through marker assisted selection and to evaluate high yielding rice varieties with tolerant to submergence. SSR profiling were performed to tag the submergence tolerant QTL by using sub1 flanking markers and F1 confirmation of Binadhan-7 x BRRI dhan52 (F1) by using the primer RM1115. Plants number 1, 4, 13, 14 were not heterozygous band but plants number 2,3,5,6,7,8,9,10,11,12 were in heterozygous band that is F1 confirmation were finished. Binadhan-7 x BRRI dhan52 (F2) crosses were possessed the highest grain yieldplant-1(38.00 g) which was significantly higher than its both parents  and also early maturing as 124 day from  all crosses. Therefore these cross combinations may be suggested for further research with check varieties to select desirable genotypes or segregates with higher yield potentiality, earliness and other important characteristics.

I want to give very special thanks are due to my col-leagues in the workplace who were there to help me in every moment of this huge work. Chemicals and instruments are supported by BINA and GPB, BAU.

The authors declare that there is no conflict of interest