Mapping QTL hotspots associated with weed competitive traits in backcross population derived from Oryza sativa L. and O. glaberrima Steud.

To improve grain yield under direct seeded and aerobic conditions, weed competitive ability of a rice genotype is a key desirable trait. Hence, understanding and dissecting weed competitive associated traits at both morphological and molecular level is important in developing weed competitive varieties. In the present investigation, the QTLs associated with weed competitive traits were identified in BC1F2:3 population derived from weed competitive accession of O. glaberrima (IRGC105187) and O. sativa cultivar IR64. The mapping population consisting of 144 segregating lines were phenotyped for 33 weed competitive associated traits under direct seeded condition. Genetic analysis of weed competitive traits carried out in BC1F2:3 population showed significant variation for the weed competitive traits and predominance of additive gene action. The population was genotyped with 81 genome wide SSR markers and a linkage map covering 1423 cM was constructed. Composite interval mapping analysis identified 72 QTLs linked to 33 weed competitive traits which were spread on the 11 chromosomes. Among 72 QTLs, 59 were found to be major QTLs (> 10% PVE). Of the 59 major QTLs, 38 had favourable allele contributed from the O. glaberrima parent. We also observed nine QTL hotspots for weed competitive traits (qWCA2a, qWCA2b, qWCA2c, qWCA3, qWCA5, qWCA7, qWCA8, qWCA9, and qWCA10) wherein several QTLs co-localised. Our study demonstrates O. glaberrima species as potential source for improvement for weed competitive traits in rice and identified QTLs hotspots associated with weed competitive traits.

Phenotyping of mapping population for weed competitive traits under direct seeded rice condition. The experiment was laid out in augmented block design with four blocks, wherein, each block consists of 36 BC 1 F 2:3 families along with five checks. Each family was sown in two rows of two meter length with spacing of 20 × 15 cm. The population was phenotyped for six weed competitive traits such as seedling height (cm), number of tillers, number of leaves, leaf area (cm 2 ), shoot fresh weight (g) and shoot dry weight (g). The each observation was made on average of three plants at 15,30 and 45 days after sowing. The leaf area (cm 2 ) was measured using leaf are meter (LI-COR, LI-3100C). Where, fresh leaves of three seedlings were passed through the leaf area meter and average was obtained for each genotype. Similarly, six physiological parameters such as absolute growth rate 36 , relative growth rate 36 , crop growth rate 36 , specific leaf area 47 , leaf area index 47 and leaf area ratio 47 were determined at each sampling interval using following formulas.
where as, "t" is number of days after sowing at which observation was recorded; t 1 and t 2 are first and second interval time (e.g., 15 DAS and 30 DAS) of observation respectively. Where, h 1 and h 2 are seedling height at t 1 and t 2 respectively. The difference between seedling heights of two intervals was divided by difference in days for two sampling intervals and expressed as cm day -1 .
Absolute growth rate cm day −1 = Specific leaf area cm 2 g −1 = Leaf area per plant cm 2 Leaf dry weigh per plant g , Leaf area ratio cm 2  www.nature.com/scientificreports/ where, w1 and w2 are plant dry weight at times t1 and t2, P = spacing (m 2 ).
Genotyping of mapping population. Total genomic DNA of 30 days old seedlings was extracted using 2% Cetyl Trimethyl Ammonium Bromide (CTAB) method 49 . The DNA quantity and quality was analyzed by running on 0.8% agarose gel (Biorose agarose) and quantity of DNA in each sample was estimated by comparing the band intensity with known quantity of DNA (Lamda DNA ladder, Takara). The isolated genomic DNA was diluted with 1X TE buffer to get required concentration of DNA (~ 50 ng/μl) in each sample.
A total of 428 SSR markers spanning all over 12 rice chromosomes were used for identification of polymorphic markers between two parents IR64 and O. glaberrima (IRGC105187). The SSR marker found polymorphic between parents were used for BC 1 F 2 population genotyping. The PCR amplification was carried out in Effendorf Vapo.protect PCR cyclers in 96 well plates using following PCR cycling conditions. The 10 μl volume for each reaction containing 3.5 μl of 2X PCR Taq mastermix (ABM with dye), 0.5 μl of each forward and reverse SSR primer (5 pmol), 2 μl of diluted genomic DNA (~ 50 ng/μl) and 3.5 μl of nuclease free water. The PCR cycling involves, initial denaturation (94 °C for 3 min), denaturation (94 °C for 30 s), annealing (50-58 °C for 30 s), extension (72 °C for 40 s) and final extension (72 °C for 5 min) with 35 cycles (step 2, 3 and 4). Finally samples were stored in 4 °C in cyclers. PCR amplified products were resolved in 3-4 per cent agarose and sizes of amplified fragments were determined by comparing with 100 bp ladder (Genei). The documented gels with amplified products were scored visually and allele score "A" was assigned to recurrent parent IR64, allele score "B" was assigned to donor parent O. glaberrima and heterozygotes were assigned with allele code of "H". Whereas, missing alleles were scored as "NA" and non parental alleles as "C".
Linkage map construction and QTL analysis. The linkage map was constructed using QTL ICI mapping software v 4.2 (CIMMYT) with MAP function as procedure described by 50 . Recombination frequency of 30 cM was threshold value for grouping, ordering within group was based on K-optimally with 2-optMAP and rippling by recombination frequency with window size of 5. The information regarding marker segregation based on chi-square goodness of fit based also obtained.
The QTL mapping was carried out with Windows QTL cartographer v 2.5 (N.C. state university, Bioinformatics Research Centre).The composite interval mapping methods (CIM) was performed with 1000 permutations and significance level of 0.05 along with the standard model (model 6) of composite interval mapping with forward and backward regression method. The QTLs with threshold of > 2.5 LOD was used as criteria for declaring the QTL. The graphics showing QTL location were obtained from Windows QTL cartographer v 2.5. The standard procedure for QTL nomenclature was outlined by "The committee on Gene Symbolization, Nomenclature and Linkage (CGSNL) of the Rice Genetic Cooperative was followed 51 . Comparison of QTLs with previously reported QTLs were carried out using Q-TARO, Gramene QTL database and research publication (using physical position).

Development of interspecific mapping population.
The O. glaberrima accession, IRGC105187 was crossed to O. sativa cv. IR64, a widely adapted mega variety, but poor in weed competitive ability. The interspecific F 1 developed by crossing IR64 (♀) with O. glaberrima (♂ ) showed complete pollen sterility, therefore, the BC 1 F 1 seeds were generated by backcrossing F 1 with recurrent parent IR64, wherein, F 1 s served as female parent and IR64 as pollen donor. However, very low seed set during backcross was observed as clipping the florets of the F 1 plants lead to shattering of florets within 24 h of pollination. The BC 1 F 1 plants showed partial spikelet fertility (data not shown) and set seeds upon selfing. Each selfed seed obtained from the BC 1 F 1 plants were raised as individual BC 1 F 2 plants. Among the BC 1 F 2 plants the spikelet fertility varied significantly and set seeds upon selfing. The seeds from each BC 1 F 2 plants were constituted to develop BC 1 F 2:3 families. In total, 144 BC 1 F 2:3 families were generated from the cross IR64*1/IRGC105187. Phenotyping of BC 1 F 2:3 mapping population for WCA traits under DSR condition. One hundred forty four BC 1 F 2:3 families along with five checks grown under direct seeded condition were phenotyped for 12 WCA traits at 15, 30 and 45 DAS. The following abbreviations such as, SH (seedling height), NT (number of tillers), NL (number of leaves), LA (leaf area), SFW (shoot fresh weight), SDW (Shoot dry weight), AGR (absolute growth rate), SLA (specific leaf area), LAI (leaf area index), LAR (leaf area ratio), CGR (crop growth rate), RGR (relative growth rate) with corresponding sampling interval (i.e., SH15, SH30 and SH45) were used hereafter in  Table. S2 and graphically represented in box plots (Fig. 1). Although, seedling height in mapping populations was lower at SH15 (12.29 cm) as compared parents, it increased rapidly at SH30 (20.56 cm) and at SH45 (30.30 cm) and recorded higher values than both parents. Number of tiller at NT15 was confined to single tiller per plant. However, average number of tillers were higher than both the parents in mapping population at NT30 (3.99) and NT45 (9.24). Similar findings were observed for number of leaves at NL15, as mapping population including parents had three leaves per plant and average number of leaves in mapping population at NL30 (14.28) and NL45 (32.46) were higher than both the parents. Mean leaf area at LA15 was 3.58 cm 2 , which is lower than both parents. However, significant increase in the leaf area was observed at LA30 and LA45 with mean leaf area of 32.68 cm 2 and 116.82 cm 2 respectively. At LA45 average leaf area in the mapping population was higher than the donor parent O. glaberrima (105.8 cm 2 ). The mean shoot fresh weight (g) of the mapping population at SFW15 was 0.090 g which was lower than both parents. However, significant increase in average shoot fresh weight (g) was observed at SFW30 with 1.04 g and at SFW45 with 5.09 g, which is higher than donor parent O. glaberrima (0.91 g and 2.66 g) at both sampling intervals. Similar findings were observed for shoot dry weight (g) at SDW15 with the mapping population recording a mean value of 0.021 g which is lower than both parents. As observed for the other traits, significant increase in average shoot dry weight was observed at SDW30 with 0.210 g and at SDW45 with 0.990 g which is higher than donor parent O. glaberrima (0.170 g and 0.497 g) at both sampling intervals. The heritability estimates (Supplementary Table S3) were found high (60.57-98.40%) for all above traits except that it was moderate for SH30, SDW15 (39.66-59.46%) and found low for LA15 (19.45%).
The mean absolute growth rate of the mapping population was observed to decrease from AGR15 (0.81 cm day −1 ) to AGR30 (0.55 cm day −1 ). However, it increased at AGR45 (0.64 cm day −1 ), while IR64 (0.92-0.13 cm day −1 ) and O. glaberrima parent (0.88-0.41 cm day −1 ) showed decreasing trends from AGR15-AGR45. The mean specific leaf area of the mapping population was observed to decrease in successive sampling intervals as it reduced from 350.29 cm 2 g −1 at SLA15 to 251.79 cm 2 g −1 at SLA45. Similarly, O. glaberrima parent showed decreasing trend from 618.99 cm 2 g −1 at SLA15 to 342.49 cm 2 g −1 at SLA45. The leaf area index found increasing from 0.11 at LAI15 to 3.51 at LAI45. Similarly, O. glaberrima parent shown increase in leaf are index from 0.41 at LAI15 to 3.16 at LAI45. The leaf area ratio shown similar trends as of specific leaf area, where mean leaf area ratio was decreased at each sampling interval as it was reduced from 173.27 cm 2 g −1 at LAR15 to 124.98 cm 2 g −1 at LAR45. Similarly, O. glaberrima parent shown decreasing trend from 381.82 cm 2 g −1 at LAR15 to 211.26 cm 2 g −1 at LAR45. The mean crop growth rate in mapping population at CGR15 was 0.05 g m 2 day −1 which is www.nature.com/scientificreports/ lower than both parents. However, crop growth rate was increased at CGR30 with mean of 0.46 g m 2 day −1 and at CGR with mean of 1.73 g m 2 day −1 , which is higher than the donor parent O. glaberrima. Relative growth rate in mapping population at RGR30 was ranging from 0.07-0.26 g g −1 day −1 with average of 0.15 g g −1 day −1 . Whereas, at RGR45 it was ranging from 0.01-0.17 g g −1 day −1 with average of 0.10 g g −1 day −1 . The heritability estimates (Supplementary Table. S3) were found high for AGR15, AGR30, AGR45, LAI30, LAI45, CGR30, CGR45, RGR30 and RGR45 (60.57-98.06%). The traits such as SLA15, SLA30, SLA45, LAR45 and CGR15 had moderate heritability (39.66-59.98%), while heritability was found to be low for LAI15, LAR15 and LAR30 (7.1-20.95%). All weed competitive traits exhibited positively skewed distribution, except RGR30 and RGR45 which were negatively skewed ( Supplementary Fig. S1). Transgressive segregants were observed for all weed competitive traits while the number of families performing better than donor parent increased with later sampling interval for all the traits.
Correlation among weed competitive ability traits. The traits such as seedling height, number of tillers, number of leaves, leaf area, shoot fresh weight, shoot dry weight, absolute growth rate, leaf area index and crop growth rate had significant positive association among them at all stages of sampling. However, specific leaf area and leaf area ratio had significant negative association with shoot dry weight, crop growth rate and relative growth rate (Fig. 2).

Discussion
Even though scarcity of labour and water is forcing shifting of rice cultivation from irrigated to direct seeded, it is besotted with major production constraints like weed infestation. Hence, development of weed competitive cultivars through plant breeding has become imperative for sustainable production under direct seeded and aerobic rice conditions. However, weed competitive ability is a quantitative trait, determined by interaction of associated traits such as plant height, tiller number, leaf canopy traits and root traits. Hence, insight into the genetic and molecular mechanisms of weed competitive ability will help in rapidly developing weed competitive cultivars. Therefore, the present study was designed and carried out to evaluate weed competitive ability of a mapping population derived from O. sativa × O. glaberrima cross and to dissect their association with chromosomal region using QTLs identification approach.  www.nature.com/scientificreports/ our observations in the this study, a recent study 53 reported higher dry biomass accumulation in introgression lines of O. glaberrima at 28 DAS. Rapid biomass accumulation in mapping population was evidenced by higher means of crop growth rate and relative growth rate in mapping population at 30 DAS and 45 DAS. The leaf area at 15 DAS did not show any significant variability in the mapping population and none of introgression lines were found better than the O. glaberrima parent. However, range of phenotypic variability significantly widened at later stages and mapping population had higher mean leaf area than IR64 at 30 DAS and both parents at 45 DAS.
These results indicate introgression lines/progenies have the ability to put up more rapid leaf area. In support to our findings, a study 43 suggested the rapid initial growth of O. glaberrima interspecific progenies was associated with faster leaf growth. Another study 54 also reported large variability for leaf area in introgression lines. Our findings showed similar trends with respect to leaf area index. Specific leaf area at 15 DAS showed that, most of the introgression lines had higher specific leaf area than the recurrent parent IR64 but lesser than the O. glaberrima parent. However, at later sampling stages (30 and 45 DAS), mean specific leaf area of mapping population shown decreasing trends. These results indicate leaves will get thicker as growth progresses and found to be having intermediary values of both parents. As opined by Dingkhun 45 , cultivars that have large specific leaf area during early developmental stages (more ground coverage) and small specific leaf area (for yield benefit) during advance stage is desirable for weed competitiveness. Several earlier studies 43,44 have reported O. glaberrima interspecific progenies had intermediate specific leaf area during early growth stages, followed by a decrease as that of O. sativa parent. Similar results were also reported 52,54 in O. glaberrima interspecific progenies. In our study, the mapping population had lower leaf area ratio than both the parents at all sampling stages. It indicates mapping population produced less leaf area per unit dry weight and same dry matter might be diverted to development of seedling height, tillers and leaves. There is lack of reports on leaf area ratio in O. glaberrima and their progenies.
Interspecific linkage map. The result found that, 137 out of 428 (32%) SSR markers were polymorphic.
The polymorphism% in the present study was found to be low when compared to other studies involving interspecific cross between O. sativa × O. glaberrima parents. For instance a study 55 reported 100 out of 140 (71.40%) SSRs were polymorphic, while other 56,57 reported higher an even higher 79.3% (130 out of 164) and 77.9% (109 out of 140 microsatellites) polymorphism, respectively. However, there are studies 58,59 , which reported lower level of polymorphism (27-38% and ~ 40% respectively). Another study 60 , reported polymorphism % ranging from 23.87 to 50.66%. Based on these reports, there is a large variation for polymorphism % and results are confined to specific particular studies and they cannot be generalized. The factors such as, extent of genetic diversity between the parents, gene pool they belong to and distribution of chosen markers on chromosome affects results of polymorphism. In our study, linkage map with length of 1423 cM was constructed using 81 polymorphic SSR markers. The length of linkage map and marker interval varies with number of markers used, recombination between the markers and population type. Hence, there is no common agreement between linkage maps of different studies which O. glaberrima as source of weed competitiveness ability traits. In the present study, 59 major QTLs were detected with phenotypic variance of 10.47-58.99%. Among the 59 major QTLs, 38 QTLs were derived from O. glaberrima. These 38 major QTLs from O. glaberrima were identified for weed competitive traits such as seedling height, number of tillers, number of leaves, leaf area, shoot fresh weight, shoot dry weight, specific leaf area, leaf area index, leaf area ratio and crop growth rate. Whereas, QTLs from O. glaberrima for traits such as absolute growth rate and relative growth rate were observed to be minor. The QTLs such as qSH30-7.1 (qLTG7 66 , qSL7 67 , qSH12-7.1 31 , qSV7a 38 ), qSH45-3.1 (qTAA3-1 25 , RZ313, RG369 69 , qSHL3.1 24 , qSV3b 38 ), qNT30-5.2(qSV5b 38 ), qNT45-10.1 (qSV10b 38 ), qLA30-9.1 & qLA45-9.1 (qLA-9 31 ), qSFW30-6.1 (qRS6-1 25 , qSFWd6 70 ) and qSDW30-9.1 & qSDW45-9.1 (qSDW9 77 , rdq9 76 , SUB1 75 , qSV9c 38 ) were had beneficial alleles from O. glaberrima with major QTL effects. Hence, these QTLs can be employed in marker-assisted selection for developing weed competitive cultivars. These results indicates, O. glaberrima species inherits genes/allele essential for development of weed competitive traits and could be used as potential donor in breeding for weed competitiveness. The QTLs identified in present study will pave for development of weed competitive rice genotypes suitable for DSR.
Dynamic expression of weed competitive QTLs. In rice, several studies have reported QTL associated with WCA traits present on 12 chromosomes [24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40] . Each study used different determinants to assess and identify genomic regions associated with WCA viz., shoot length, shoot weight, coleoptile length, shoot dry weight, leaf area and specific leaf area. Most of the QTLs identified for WCA traits have been carried out under controlled conditions using petri dishes, slant plates, paper rolls, hydroponics, sand and hydroponics and soil filled pipes 2 , while the major QTLs associated with WCA traits were identified through a thorough field screening experiment. It was also observed that, traits such as seedling/plant height and shoot/total dry weight were used as major determinants of WCA in most the studies and large number of QTLs were reported for these two traits. The traits such as fresh biomass, number of leaves and leaf area were also given relative importance. However,

Scientific Reports
| (2020) 10:22103 | https://doi.org/10.1038/s41598-020-78675-7 www.nature.com/scientificreports/ trait such as number of tillers and physiological parameters such as specific leaf area, growth rate estimates were sparsely used and relatively less number of QTLs were reported. In our study, relatively equal number of QTLs was identified at each sampling stage. Among the 72 QTLs, 28 QTLs were identified at 15 DAS, 21 QTLs at 30 DAS and 23 QTLs at 45 DAS. It was also observed that, QTLs shown stage specific or dynamic expression as none of QTL region commonly found across three sampling stages. However, QTL region on chromosome 9 (RM23901-RM566) found common at two sampling intervals for several traits such as LA30, LA45, SDW30, SDW45, LAI30, LAI45, LAR15, LAR30, CGR30 and CGR45. Similarly, QTL region on chromosome 2 (RM13155-RM13962) found common SLA15 and SLA45. These stable QTLs can be further analysed and characterized for future studies. Similar to our findings, stage specific or dynamic expression of WCA QTLs has been well documented in previous studies. Recently a study 38 reported that, 33.3% (plant height), 10.7% (tiller number) and 3.4% (above ground dry weight) of the total QTLs were detected in all three sampling stages. Our study indicates that, QTL identified are specific to either 15 DAS or 30 or 45 DAS and we could not find common QTLs across all three stages. Therefore, it can be concluded that QTLs expression is highly dynamic and stage specific. The consistent QTLs and stage specific QTLs associated with weed competitive in our study plays significant role in understanding the genetic architecture of weed competitive traits.

Comparison of QTLs identified in the present study with those from previous reports. The
QTLs identified in the present study were compared with previously reported QTLs (using QTARO, Gramene QTL database and research reports with physical positions). The QTLs identified for cold or low temperature tolerance/submergence tolerance/drought tolerance in the previous studies were taken into consideration as they are directly correlated with vigour 24,31,33,36,38,[63][64][65][66][67][68][69][70][71][72][73][74][75][76][77][78][79][80] . Based on these findings, QTLs identified in the present study for traits such as seedling height, number of leaves, number of tillers, leaf area, shoot fresh weight, shoot dry weight at all sampling co-localized with those identified in previous studies except qNL30-5.1, qNT30-5.1 and qSFW45-2.1. These three minor QTLs were derived from O. glaberrima parent and found to be novel. As most of the identified QTLs were co-localized with previous reports, which are based on different mapping population and diverse environments, the QTLs identified in the present study can be utilized in developing WCA varieties with greater confidence. However, QTLs for the traits such as absolute growth rate, specific leaf area, leaf area index, leaf area ratio crop growth rate and relative growth rate have been sparsely studied so far and only few QTLs have been reported with respect to WCA in rice.
QTL hotspots for weed competitive ability traits. The QTL "hotspots" for WCA and related traits were reported by many studies 23,24,29,30,36,67,70 on different chromosomes for various co-localized traits. The possible reasons for co-localization may be due to linkage or pleiotropy. Similarly, significant correlations among co-localized traits provide possible explanation for QTL "hotspots". The study 25 also pointed out that, co-localization may have occurred by chance, whenever large numbers of QTLs were detected. In the present study, QTL hotspot qWCA9 (7.71 Mb) harboured 12 major QTLs with average phenotypic variance of 39.65%. Whereas, QTL hotspot qWCA5 (19.19 Mb) had eight co-localized QTLs with average phenotypic variance of 18.66%. The QTL hotspots on chromosome 2, qWCA2a (5.03 Mb), qWCA2b (16.0 Mb), and qWCA2c (4.70 Mb) found to harbour four QTLs each with average phenotypic variance of 6.0%, 26.59% and 40.40%. However, all the QTLs in qWCA2a and qWCA2c had positive alleles from O. glaberrima and in qWCA2b, recurrent parent IR64 contributed positive alleles. Similarly, the QTL hotspot qWCA10 (7.61 Mb) co-localized with four major QTLs contributed from O. glaberrima with average phenotypic variance of 31.03%. The QTL hotspot qWCA3 (2.65 Mb) and qWCA5 (13.17 Mb) had three QTLs each with average phenotypic variance of 4.92% and 9.78% respectively with positive alleles contribution from O. glaberrima. While, QTL hotspots qWCA7 (19.32 Mb) and qWCA8 (3.97 Mb) had average phenotypic variance of 27.62% and 23.94% respectively with positive alleles contribution from both parents. These results indicate both parents associated with positive alleles of WCA associated traits and presence wide range of molecular and phenotypic diversity for WCA associated traits among mapping population 23 . These genomic regions need to be further characterized and used as potential target in marker assisted breeding for improvement of rice varieties for WCA. However, QTL hotspots identified in multiple environments could be ideal for understanding its regulatory role and further use in molecular breeding.

Segregation distortion in O. sativa × O. glaberrima mapping population. Segregation distortion
(deviation of observation of marker ratio from the expected ratio) is the strong evolutionary force, commonly encountered in mapping populations derived from diverse genotypes 81,82 . In QTL mapping, the segregation distortion known to affect precision of QTL mapping by altering the genetic distance between the markers and the order of the markers on the linkage group. However, Zhang et al., 83 concluded that, in general, segregation distortion will not produce more false QTL, nor will it have significant impact on the estimation of QTL position and effect. As Zhang et al. 83 and Xu 84 suggested, dense linkage map with large-size mapping population would minimize power loss even in the presence of segregation distortion. In the present study, mapping population showed presence of segregation distortion for all the markers used. Presence of hybrid sterility genes and gametophyte competition gene 55,59,61,85 in O. sativa × O. glaberrima interspecific population were found to cause segregation distortion. A recent study 85 reported 10 hybrid sterility loci (S1-glab, S19-glab, S20-glab, S37-glab, S38glab, S39-glab) as gamete eliminator or pollen killer between O. sativa × O. glaberrima interspecific population. We also observed high amount of spikelet sterility in the backcross progenies. Similar to our findings, several studied in have reported presence of sterility and segregation distortion in O. sativa × O. glaberrima interspecific populations 55,59,61,62 . Very recently, Neelam et al. 62 , reported segregation distortion in SNP genotyping while mapping QTL for bacterial leaf blight from O. glaberrima derived population. As Xu and Hu 86 opined, for a long period of time distorted markers were simply discarded from QTL mapping for the reason of precaution www.nature.com/scientificreports/ and they also found that if distorted markers handled properly, they can be beneficial to QTL mapping with no detrimental effects. Previous studies used statistical package Proc QTL 86 (SAS) and MapDisto 2.0 55 for handling of segregation distortion in F 2 and BC 1 F 1 , RIL and DH population only. However, there is a need develop suitable package for BC 1 F 2 population. The QTLs identified in the present study can be considered as least affected by segregation distortion as most of the QTLs identified were co-localized with previously identified QTLs. In support of our observation, Xian-Liang 81 also opined that, effect of segregation distortion is minimal with the use of co-dominant markers like SSRs and also in backcross population.

Conclusions
Shift towards DSR cultivation has been gaining importance in recent years, which offer considerable saving of water and labour resources. However, it inherited the threat of weed infestation. Hence, breeding for weed competitive rice cultivars is need of the hour to tackle problem of weed infestation under DSR. The African endemic rice species O. glaberrima has the weed competitive ability due to their early vigorous growth and high specific leaf area. However, there are no systemic studies till date to identify the genomic regions (QTLs/genes) associated with weed competitiveness in O. glaberrima. The present study is, first of its kind, in which interspecific mapping population developed from IR64 × O.