Introduction

For a majority of the people in India rice is the predominant food and the crop plays a crucial role in the Indian economy (FAO 2015). Among the Indian rice varieties, Samba Mahsuri (also known as BPT 5204), developed by Acharya N.G. Ranga Agricultural University, Guntur, Andhra Pradesh India is grown in more than 3 Mha across the country, mainly in South Indian states of Andhra Pradesh, Telangana, Tamilnadu and Karnataka and occupies an important position among the mega-varieties of rice grown in India. It is a high-yielding variety with excellent cooking and eating qualities and it possess a highly desirable medium slender grain type and thus commands a high market price (Sundaram et al. 2008). Despite these features, the yield of Samba Mahsuri is severely limited by several diseases and insect pests. In 2008, ICAR-Indian Institute of Rice Research (ICAR-IIRR), Hyderabad, India and CSIR-Centre for Cellular and Molecular Biology (CSIR-CCMB), Hyderabad, India jointly developed and released a novel bacterial blight disease resistant derivative of Samba Mahsuri, named Improved Samba Mahsuri (ISM; also known as RP Bio 226) through marker-assisted backcross breeding (MABB), possessing three major bacterial blight (BB) resistance genes, viz., xa1, Xa21 and xa5 (Sundaram et al. 2008). ISM is presently estimated to grown by rice farmers across the Indian states of Andhra Pradesh, Telangana, Tamilnadu, Karnataka and Chhattisgarh, particularly in bacterial blight endemic regions, in an area of ~ 1, 00,000 ha. However, in the recent past, the spread of the new variety has been observed to be limited in those rice cultivation areas, which are endemic to another major disease of rice, called blast as both the original variety, Samba Mahsuri and its derivative, Improved Samba Mahsuri are highly susceptible to the dreaded disease.

Magnaporthe oryzae, the causative agent of rice blast lead to the significant yield loss in the major rice growing areas in India and as well as in the world. (Khush and Jena 2009). The most appropriate strategy for blast disease management is to deploy host-plant resistance. Till now, at least 100 rice blast resistance genes and several QTLs conferring resistance to blast have been identified (Sharma et al. 2005; Sharma et al. 2012). Among them, two blast resistance genes viz., Pi2 and Pi54 are known to be very effective, confer broad-spectrum resistance and hence are considered to be suitable for deployment in India. Further, gene-specific markers are available for both Pi2 and Pi54 for deployment in marker-assisted breeding. Considering the imminent need to improve blast disease resistance of ISM in the present study, we attempted to deploy the strategy of marker-assisted backcross breeding (MABB).to transfer Pi2 and Pi54 into the genetic background of the bacterial blight resistant version of Samba Mahsuri.

Materials and methods

Plant materials

Two near-isogenic lines of Samba Mahsuri, viz., RPBio Patho-1 and RPBio Patho-2 (Prasad et al. 2011; Madhavi et al. 2011), possessing the blast resistance genes, Pi2 and Pi54y, developed from the crosses, Samba Mahsuri X C101A51 and Samba Mahsuri X Tetep, respectively were used as donor parents. The bacterial blight resistant variety, Improved Samba Mahsuri (ISM) harboring three bacterial blight resistance genes (Xa21, xa13 and xa5) was used as the recipient parent. In addition to these, Samba Mahsuri was used as susceptible check while ISM was used as resistant check, respectively in bacterial blight screening experiments. HR12 was used as the susceptible check in blast screening experiments along with C101A51 and Tetep as resistant checks.

Strategy for marker-assisted backcross breeding

Two separate crosses i.e., ISM X RPBio Patho-1 (Cross I) and ISM X RPBio Patho-2 (Cross II) were carried out to combine Pi54 and Pi2 into the genetic background of ISM. The F1s developed from both the crosses were confirmed for their heterozygosity using target resistance gene-specific markers [viz., AP5659-5 (specific for Pi2, Fjellstorm et al. 2006) and Pi54MAS (specific for Pi54, Ramkumar et al. 2011), pTA248 (specific for Xa21; Ronald et al. 1992), xa13prom (specific for xa13; Hajira et al. 2016), xa5FM (specific for xa5; Hajira et al. 2016)] to identify heterozygous plants in both the crosses and such plants were backcrossed with ISM to generate BC1F1s. They were subjected for foreground selection using the gene-specific markers mentioned above and positive plants (i.e. heterozygous BC1F1 plants) were then analysed for the recurrent parent genome recovery through background selection using a set of 144 SSR parental polymorphic markers evenly distributed across the 12 rice chromosomes to identify those plants with maximum recurrent parent genome recovery in both the crosses. A single such BC1F1 plant was then selected from both the crosses (i.e. cross I and II) and backcrossed with ISM to generate BC2F1s. Marker-assisted foreground and background selection (as explained above) was repeated among BC2F1 plants and a single heterozygous BC2F1 plant from both the crosses, possessing the target resistance genes and also with maximum ISM genome recovery were selfed to develop BC2F2s. Homozygous positive BC2F2 plants with respect to desired target resistance genes (i.e. homozygous for Xa21, xa13, xa5 and Pi2 in Cross I and Xa21, xa13, xa5 and Pi54 in Cross II) were identified with the help of markers and a single homozygous BC2F2 plant each from both the crosses, which closely resembled ISM (based on morphological features) were then crossed to develop intercross F1s (ICF1s). The ICF1s were selfed to generate intercross F2s (ICF2s). Among them, homozygous positive plants for the target BB and blast resistance genes (viz., Xa21, xa13, xa5, Pi2 and Pi54) were identified with the help of markers and advanced through pedigree method of selection till ICF5-. Three promising homozygous ICF5 lines closely resembling ISM were advanced for further evaluation along with plants of ISM homozygous for either Pi2 + Xa21 + xa13 + xa5 or Pi54 + Xa21 + xa13 + xa5 at BC2F6 generation.

For marker-assisted selection at each generation of backcrossing or intercrossing, mini-scale, rapid DNA isolation was done as described in Zheng et al. (1995). PCR protocols as described in Ronald et al. (1992; for Xa21), Hajira et al. (2016; for xa13 and xa5), Fjellstorm et al. (2006; for Pi2) and Ramkumar et al. (2011; for Pi54) were adopted for foreground selection of the target resistance genes. A set of 144 parental polymorphic SSR markers (listed in Supplementary Table 1) were used for background selection as per the protocol explained in Sundaram et al. (2009). The extent of donor parent segment introgression among the selected backcross derived plants was assessed utilizing the software tool, Graphical Genotype V 2.0 (Van Berloo 1999) following the protocol described in Abhilash et al. (2016c).

Assessment of BB and blast resistance in the backcross derived lines of ISM

Screening for blast resistance During Kharif 2016 (i.e. wet season 2016), gene-pyramid lines of ISM, possessing either Pi2 (at BC2F6) or Pi54 (at BC2F6) or Pi2 + Pi54 (at ICF5) were subjected to stress against blast in the UBN beds at ICAR-IIRR (Prasad et al. 2011) along with the susceptible (HR12) and resistant (C101A51 and Tetep) checks using a local isolate, SPI-40 of the blast pathogen, Magnaporthae oryzae as per the protocol described in Mohan et al. (2011). Fifteen days after inoculation, the lines were scored for their resistance/susceptibility to the disease based on IRRI-SES scale, 1996 (IRRI 1996).

Screening for bacterial blight resistance: The blast resistant gene-pyramid lines of ISM, possessing either Pi2 (at BC2F6) or Pi54 (at BC2F6) or Pi2 + Pi54 (at ICF5) were screened along with the susceptible (i.e. Samba Mahsuri) and resistant (i.e. ISM) checks for their resistance/susceptibility to bacterial blight during Kharif 2016 (i.e. wet season 2016) using a virulent isolate of the pathogen, Xanthomonas oryzae pv. oryzae, viz., DX-020 (Laha et al. 2009) through clip-inoculation procedure described by Kauffman et al. (1973). The plants were scored 15 days after inoculation for their resistance/susceptibility as per IRRI-SES scale, 1996 (IRRI 1996).

Evaluation of agro-morphological characters of the improved lines

During wet season 2016 (i.e. June–November), three promising five-gene pyramided lines at ICF5 (possessing Xa21, xa13, xa5, Pi2 and Pi54) generation, two each of selected lines of ISM possessing Pi2 (at BC2F6) and Pi54 (at BC2F6), the recurrent parent ISM were transplanted (twenty-seven day-old seedlings) into the experimental farm of ICAR-IIRR in 2 m2 plots in three replications with a 15 cm × 20 cm spacing. The seedlings were raised till maturity by adopting standard package of practices. Data was collected for selected agro-morphological traits, viz., plant height (cm), days to 50% flowering (DFF), panicle length (cm), number of productive tillers per plant, grain yield per 33 plants (i.e. per m2; gms), 1000-grain weight (g) among 33 plants for each replication (n = 3) as explained in Abhilash et al. (2016c). The data was statistically analyzed following the procedures described in Freeman et al. (1978). Least Significance Difference (LSD) values at 5 to 7 percent level of significance and Coefficient of variation (CV) were calculated using standard errors of mean (S. Em. ±) using Microsoft Excel package. The software package, SAS Version 9.2 (SAS Institute Inc., Cary, NC, USA) was used for statistical analysis. PROC GLM procedure of SAS was utilized for analysis of variance (ANOVA) for determination of significant variation among the improved breeding lines of ISM.

Evaluation for grain physic-chemical quality traits

The selected advanced breeding lines of ISM possessing BB and blast resistance were evaluated for key grain and cooking quality related traits. After harvest, the rice grains were dried directly under sunlight followed by drying to below 14% moisture content, and preserved at room temperature for 90 days. The rice grains were dehusked, polished and evaluated for physical characters like grain length, grain width and L/B ratio. The data was also recorded for the cooking qualities traits, amylose content (%) in which the rice grains were classified into low (9–20%), intermediate (20–25%) and high (25–33%) and Gel consistency which classifies the quality of cooked rice as soft (61–100 mm), medium (41–60 mm) and hard (26–40 mm) following method of Juliano et al. (1973), Cagampang et al. (1973) and Melissa et al. (2009).

Results

Introgression of Pi2 into ISM

True F1 plants (i.e. heterozygous ones for the target resistance genes) derived from the cross ISM/RPBio-Patho-1 were backcrossed to the recurrent parent ISM. Among the 270 BC1F1s, 130 were found to be heterozygous for Pi2 and 15 of them also possessed Xa21, xa13 and xa5 in homozygous condition. Screening with 75 parental polymorphic SSR markers revealed that a single BC1F1 plant (# ISM-9) had the highest recurrent parent genome (RPG) recovery (i.e. 77.3%; Supplementary Table 1) and it was then backcrossed to ISM to generate 126 BC2F1s. A total of 60 heterozygous positive plants possessing Pi2 were identified and subjected for RPG recovery analysis. A solitary BC2F1 plant (# ISM-9-18) was identified with maximum RPG (~ 90.6%) and selfed to develop 496 BC2F2s. Out of these, 24 plants were observed to be possess the gene combination Xa21 + xa13 + xa5 + Pi2 in homozygous condition and a single four-gene homozygous plant (# ISM-9-18-217) was identified using maximum recurrent parent genome recovery (i.e. 94.6%) using background analysis. This plant was further analyzed to assess the extent of linkage drag around the target blast resistance gene, Pi2. At the proximal end of the gene, a segment of 0.5 Mb was introgressed from the donor parent, while at the other end, a 0.2 Mb segment was introgressed from the donor parent genome. Thus, in total, a segment of 0.7 Mb (Fig. 1) was introgressed in the vicinity of Pi2 from RPBio-Patho 1 in the selected BC2F2 plant (i.e. plant # ISM-9-18-217), which was later used for intercrossing.

Fig. 1
figure 1

Analysis of donor genome introgression associated with blast resistance genes in the selected BC2F2 plant: a At Pi2 locus, a segment of 0.5 Mb was introgressed at the proximal end, while a 0.2 Mb segment was observed to be introgressed at the distal end from the donor parent genome in the best BC2F2 plant (i.e.ISM-9-18-217), Thus, in total, a segment of 0.7 Mb was introgressed from the donor parent with respect to the genomic region in the vicinity of Pi2. b At Pi54 locus, a segment of 0.2 and 0.4 Mb was introgressed at proximal and distal ends, respectively from the donor parent genome (totaling to 0.6 Mb) in the best BC2F2 plant (i.e.ISM-22-102-28). The position of the polymorphic SSR markers in Mb on Chr. 6 and 11 is given in parenthesis adjacent to each marker

Full size image

Introgression of Pi54 into ISM

Among the 321 BC1F1s generated by backcrossing the true F1s derived from the cross ISM/RPBio-Patho-2 with ISM, 158 were heterozygous for the target gene Pi54. Screening of these plants using gene-specific markers for the three bacterial blight resistance genes revealed that 18 were homozygous. Background selection of these BC1F1 plants with SSR markers polymorphic between RPBio-Patho 2 and ISM (i.e. n = 69) helped in identification of a single plant (# ISM-22), possessing maximum RPG of 75.3% (Supplementary Table 1). This BC1F1 plant was backcrossed with ISM to produce 108 BC2F1s. Foreground selection of these plants resulted in identification of 48 heterozygous plants carrying Pi54. A single BC2F1 plant (# ISM-22-102) possessing a maximum RPG (i.e. ~ 86.9%) was selected using background analysis and advanced to develop 346 BC2F2s. Through analysis with gene-specific markers, 84 four-gene homozygous plants (for Xa21, xa13, xa5 and Pi54) were identified. Out of these, a solitary plant (# ISM-22-102-179) was identified with recovery of maximum RPG (~ 93%). This plant was also observed to be introgressed with 0.2 and 0.4 Mb segments from RPBio-Patho 2 at proximal and distal ends of Pi54, totaling to ~ 0.6 Mb (Fig. 1). It was then utilized for intecrossing.

Pyramiding of Pi2 and Pi54 into ISM

The best BC2F2 plants developed from the two sets of backcrosses (i.e. plant # ISM-9-18-217 possessing Pi2 in homozygous condition from the first cross and # ISM-22-102-179 possessing Pi54 in homozygous condition from the second cross) were intercrossed to develop 51 intercross F1 (i.e. ICF1) plants. Among them, 46 were identified as true F1s through analysis with gene-specific markers (Fig. 2), which were selfed to generate an ICF2 population comprising of 850 plants. Marker-assisted selection, coupled with phenotyping selection resulted in the identified of 50 plants which closely resembled ISM and possessing Pi2 and Pi54 in homozygous condition. They were advanced using pedigree-method and three best improved ICF4 lines (IC-12-9-56, IC-12-199-83, and IC-12-305-106) were identified (Supplementary Fig. 2) to be identical to ISM with respect to morphological traits and they were advanced for evaluation of their disease resistance, agromorphological features and grain quality at ICF5 generation.

Fig. 2
figure 2

Phenotypic screening of the selected five-gene pyramid ICF4 lines against rice blast and bacterial blight diseases. a Screening of the selected five-gene pyramid ICF4 plants, viz., plant # IC-12-9-56 (IL-1), IC-12-199-83 (IL-2), IC-12-305-106 (IL-3) against blast disease under uniform blast nursery. T- Tetep (donor parent and resistant check); A51-C101A51 (donor parent and resistant check); I- Improved Samba Mahsuri (recurrent parent and susceptible check); H- HR12 (susceptible check). b Screening of the selected five-gene pyramid ICF4 plants, viz., plant # IC-12-9-56 (IL-1), IC-12-199-83 (IL-2), IC-12-305-106 (IL-3) against bacterial blight disease through clip-inoculation method. I- Improved Samba Mahsuri (recurrent parent and resistant check); SM- Samba Mahsuri (susceptible check). All the three selected five-gene homozygous ICF4 lines were resistant to both bacterial blight and blast diseases

Full size image

Evaluation of the backcross derived lines of ISM for resistance against blast and BB

HR12 and the recurrent parent ISM (susceptible checks), were found to be highly susceptible to blast (Score 9), while the resistant checks, RPBio Patho-1 (Score 2), RPBio Patho-2 (score 1) were highly resistant to the disease (score; 1–2; Fig. 2a; Table 1). The backcross and intercrossed derived lines displayed excellent resistance against blast similar to the donor parents. With respect to screening for BB resistance (Fig. 2b; Table 1), the recurrent parent (ISM) and donors (RPBio Patho-1 and RPBio Patho-2) were highly resistant to the disease with a lesion length ranging from 0.0 to 1.7 ± 0.3 (Score 1). All the backcross and intercrossed lines displayed high level of resistance as of donor and recurrent parent (Table 1).

Table 1 Reaction of selected pyramided lines of ISM after inoculation with bacterial blight and blast pathogen
Full size table

Evaluation of agromorphological traits for improved lines

Among the improved lines, BC2F5 plant, # ISM-9-18-217-15 and ICF5 plant # IC-12-9-56 displayed grain yield per plant (17. 1 ± 0.2 and 17.0 ± 1.2 g, respectively) equivalent to ISM (i.e., 17.1 ± 1.7 g), while other backcross and intercross derived lines displayed grain yield higher than that of the recurrent parent (21.6 ± 1.0–26.0 ± 0.8; Table 2). All the introgressed lines were observed to be equivalent to or slightly taller than the recurrent parent ISM (83.0 ± 4.0–93.0 ± 0.5), except a single line, ISM 22-102-179-8 (80.6 ± 4.7) (Fig. 3; Table 2), which was found to be considerably shorter than ISM. The gene pyramid lines were found to perform equivalent to or better than ISM in terms of number of productive tillers per plant (12.3 ± 0.3–17.7 ± 1.3), panicle length (19.7 ± 0.9–22.0 ± 0.3 cm) and 1000 grain weight (14.5 ± 0.4–18.5 ± 2.0 g; Table 2). Four improved lines were observed to flower 9–11 days earlier than ISM along with better panicle exsertion (Table 2). The five-gene pyramid line, # IC-12-305-106 was significantly better in terms of number of tillers, plant height, panicle length, yield per plant, 1000 grain weight and also better exsertion of panicles with respect to the recurrent parent, ISM (Table 2).

Table 2 Analysis of agro-morphological characters of improved ISM lines along with parents under field conditions
Full size table
Fig. 3
figure 3

Single plants of selected five-gene pyramid ICF4: ISM- Improved Samba Mahsuri (recurrent parent), selected five-gene pyramid ICF4 plants, viz., plant # IC-12-9-56 (IL-1), IC-12-199-83 (IL-2), IC-12-305-106 (IL-3). While IL-1 and IL-2 was observed to be equivalent to Improved Samba Mahsuri, IL-3 was observed to be superior to the recurrent parent

Full size image

Quality traits of the improved lines

All the improved lines were observed to be of medium-slender grain type, possessed intermediate values with respect to amylose content (20–25%) and had hard gel consistency (Table 3), equivalent to the values noted for ISM.

Table 3 Analysis of grain and cooking qualities of ISM improved lines
Full size table

Discussion

Samba Mahsuri (BPT5204), a very popular high yielding Indian rice variety fetches premium price for farmers due to its excellent grain and cooking quality attribues. However, it is highly susceptible to many biotic and abiotic stresses. We had earlier developed a near isogenic line (NIL) of Samba Mahsuri, possessing resistance against a dreaded disease of rice called bacterial blight (BB; conferred by the genes, Xa21, xa13 and xa5) in collaboration with CSIR-CCMB. The NIL was released as a new variety in 2008 for cultivation with the name Improved Samba Mahsuri (ISM; Sundaram et al. 2008). The newly released bacterial blight resistant variety was observed to give 15–30% more yield than Samba Mahsuri and other bacterial blight susceptible varieties in areas endemic to bacterial blight and hence it is being enthusiastically cultivated in Indian states of Telangana, Andhra Pradesh, Tamilnadu, Karnataka and Chhattisgarh in an estimated area of ~ 1, 00,000 ha. This is one of the first examples of a rice variety developed in India through molecular marker-assisted selection and successfully released and commercialized. As ISM was observed to recently show low to moderate blast susceptibility at a few locations across India, we planned and executed the current study, aimed to improve ISM for durable blast resistance by utilising an adoptable strategy, marker assisted incorporation of two, broad spectrum blast resistance genes, Pi54 and Pi2.

MABB is a fastidious strategy, which involves the introgression of one or more target genes into elite crop variety/hybrid to improve one or few traits lacking in the original variety/hybrid (Wen and Gao 2012). The purpose of backcross breeding is to achieve a progressive reduction in the donor genome content to retain the characteristics of the recurrent parent in the improved varieties apart from introgression of one or more target genes (for e.g. genes conferring resistance against diseases). As single gene conferred resistance is known to be short lived (Victoria and Zeigler 1993; Shanti et al. 2010; Hua et al. 2015), rice breeders are striving to breed durable resistant varieties by pyramiding two or more genes into elite varieties and hybrids. There are a few earlier reports, where rice breeders have successfully adopted marker-assisted breeding strategy for improvement of rice varieties for resistance against blast (Das and Rao 2015; Tanweer et al. 2015; Usatov et al. 2016), bacterial blight (BB) (Pradhan et al. 2015; Sundaram et al. 2008, 2009), BB and blast (Singh et al. 2011; Hari et al. 2013; Balachiranjeevi et al. 2015; Ellur et al. 2016; Abhilash et al. 2016a, b, c). Taking cues from these reports, the present study was formulated and executed.

Pi54, a major blast resistance gene derived from Tetep (a Vietnamese rice line) is known to be very efficient under Indian circumstances (Rai et al. 2011; Sharma et al. 2010). NILs of Swarna and Samba Mahsuri possessing Pi54 displayed excellent resistance in multi-location trials of All India Coordinated Rice Improvement Project (AICRIP) trials (DRR Progress report, Vol. 2, 2008–2013). Along with Pi54, we also considered another major blast rsistance gene, Pi2, a gene known to confer broad-spectrum resistance (originally derived from the indica rice line, 5173; Mackill and Bonman 1992) for introgression into ISM. The efficacy of Pi2 gene in conferring excellent resistance against blast in the several parts of India has been reported (Variar et al. 2009; Imam et al. 2014; Khanna et al. 2015; Indian Council of Agricultural Report 2009–2013). Further, a recent study (Ellur et al. 2016), revealed that the gene pyramid combination, Pi2 + Pi54 is very effective in the southern states of India such as Kerala, Tamilnadu apart from northern and eastern parts of India. Hence based on these reports, we narrowed down on the two genes, Pi2 and Pi54 for introgression into ISM.

MABB strategy is known to be very helpful in limiting the number of backcrosses required for near-complete recovery of the recurrent parent genome (RPG). There are some earlier reports wherein the recurrent parent genome (RPG) has been successfully recovered by using just two or three back crosses. (Abhilash et al. 2016c; Balachiranjeevi et al. 2015; Singh et al. 2011). A similar approach was followed in the current study, wherein the number of backcrosses was limited to just two. In addition to the deployment of marker-assisted selection for accelerate recovery of the RPG at each stage of backcrossing, the donor parents used in the present study are actually pre-breeding lines with Samba Mahsuri genetic background possessing medium-slender grain type (Prasad et al. 2011). These factors helped to restrict to two rounds of backcrossing. A single plant (# ISM-9-18-217) at BC2F2 (derived from Cross I), possessing Pi2 gene with maximum recurrent parent genome recovery (94.6%) was observed to have very limited linkage drag of only 0.7 Mb around Pi2. Similarly, another BC2F2 plant (# ISM-22-102-179 l; derived from cross II) possessing Pi54 gene with maximum recurrent parent genome recovery (~ 93%) had a linkage drag of only ~ 0.6 Mb in the vicinity of Pi54. Thus when these plants were intercrossed, as expected, the derived lines closely resembled Samba Mahsuri and ISM in terms of most of the agro-morphological characters and grain type, while possessing excellent resistance against BB and blast.

Phenotypic screening of the improved versions of ISM possessing either Pi2 or Pi54 or both the genes revealed that all the lines are highly resistant to both blast and BB (Table 1; Fig. 3a, b). Among these lines, ICF4 lines possessing the complete complement of BB and blast resistance genes i.e. combinations of Pi2, Pi54 and xa5, xa13, Xa21 were observed to possess excellent resistance against blast as compared to be backcross derived lines incorporated with single blast resistant genes (Pi2 or Pi54). As the recurrent parent used in the present study is a BB resistant variety having three BB resistant genes (xa5, xa13 and Xa21), all the improved lines of ISM at BC2F5 and ICF4 generation were observed to be highly resistant to BB (0.3 ± 0.3–1.7 ± 0.3), similar to ISM (Table 1; Fig. 2b). The results obtained in the present study with respect to resistance against BB and blast are in congruence with the results obtained in our previous reports (Abhilash et al. 2016a, b, c; Hari et al. 2013, Balachiranjeevi et al. 2015), wherein no negative interaction was observed between blast and BB resistance genes.

Interestingly, most of the ISM derived lines in the present study were found to be equivalent to ISM in terms of grain yield, some of them were observed to be superior with respect to yield (Table 2) and panicle exsertion. Further all the selected lines, possessing five resistance genes were observed to be retaining the fine-grain type and other grain quality attributes of the ISM (Table 2; Supplementary Fig. 2). It can be expected that the blast resistant lines of ISM which have been developed through the study may command the same premium price for farmers in the market similar to Samba Mahsuri (BPT5204) or ISM as they are identical in most of the key attributes related to grain and cooking quality. The approach adopted in the present study, i.e. combining phenotype-based selection with MAS was helpful in not only recovering good features of the recurrent parent, but also helpful in selection of superior lines (with better yield and panicle exsertion) as compared to ISM in the later generations of backcrossing and intercrossing.

The significant contribution of this study is the achievement of complete recovery of yield, yield related traits along with grain yield and cooking quality related characters of ISM in the derived lines of ISM possesing, Pi2 and Pi54 along with achievement of a marginal improvement in its panicle exsertion and yield in some of the improved lines (Tables 2, 3). Even though in a recent study, improved breeding lines of Samba Mahsuri possessing blast and BB resistance have been developed (Madhavi et al. 2016), the improved breeding lines developed in the reported study were confirmed to be possessing only Xa21 and xa13 and not xa5. However, in our study, we have ensured transfer of the complete complement of the bacterial blight resistance genes, viz., xa5, xa13 and Xa21 into Samba Mahsuri along with transfer of the two blast resistance genes and hence it can be considered as an improvement over the report of Madhavi et al. (2016). In addition, we have also analyzed the key grain and cooking quality attributes of the pyramided lines of ISM possessing Pi2 and Pi54 and demonstrated their equivalence to ISM.

The improved versions of Samba Mahsuri possessing BB and blast resistance, developed in the present study may confer a distinct advantage to Samba Mahsuri farmers whose fields are affected by both BB and blast. Further, the improved lines of ISM developed in this study can also be used as donors to transfer BB and blast resistance into other genetic backgrounds as they possess high yield and medium-slender grain type. Among the improved lines of ISM, the intercross derived line # IC-12-305-106 (possessing Xa21 + xa13 + xa5 +Pi2 + Pi54) possessed very good phenotype and better panicle exerstion has been identified as one of the best line (Table 2) and this line will shortly be nominated for multi-location trials across India.