Introgression of Two Drought QTLs into FUNAABOR-2 Early Generation Backcross Progenies Under Drought Stress at Reproductive Stage

Christian Okechukwu Anyaoha, Mamadou Fofana, Vernon Gracen, Pangirayi Tongoona, Semon Mande

?

Introgression of Two Drought QTLs into FUNAABOR-2 Early Generation Backcross Progenies Under Drought Stress at Reproductive Stage

Christian Okechukwu Anyaoha1, 4, Mamadou Fofana2, Vernon Gracen1, 3, Pangirayi Tongoona1, Semon Mande2

(West Africa Centre for Crop Improvement, University of Ghana, PMB 30, Legon, Ghana; Africa Rice Nigeria Station, PMB 5320, Oyo Road Ibadan, Nigeria; Department of Plant Breeding and Genetics, Cornell University, Ithaca, New York 14853, USA; National Horticultural Research Institute, PMB 5432, Idi-Ishin, Jericho GRA, Ibadan, Oyo state, Nigeria)

FUNAABOR-2 is a popular Ofada rice variety grown in a large area under rainfed upland condition across western states of Nigeria. We used the combination of phenotypic and marker-assisted selection (MAS) to improve grain yield of FUNAABOR-2 under drought stress (DS) at the reproductive stage via introgression of two drought quantitative trait loci (QTLs),and. Foreground selection was carried out using peak markers RM511 and RM250, associated withand, respectively, followed by recombinant selection with RM28099 and RM1261 distally flanking. Furthermore, BC1F2-derived introgressed lines and their parents were evaluated under DS and non-stress (NS) conditions during the 2015–2016 dry season. Overall reduction of grain yield under DS compared to NS was recorded. Introgressed lines withandcombinations showed higher yield potential compared to lines with single or no QTL under DS, indicating significant positive interactions between the two QTLs under the FUNAABOR-2 genetic background. Pyramiding ofandin the FUNAABOR-2 genetic background led to higher grain yield production under DS and NS.

drought; foreground selection; introgressed line; peak marker; yield potential; rice; quantitative trait locus

Rice is one of the most important food crops in the world consumed largely across Asia and Sub-Sahara Africa. It is a water loving crop, and its production requires regular availability of water (Kumar et al, 2014). Rainfed system constitutes over 45% of world area dedicated to rice production and it is also the prevalent production system in Sub-Sahara Africa and Southeast Asia (Dixit et al, 2014). Rice yields under rainfed upland ecosystems are yet to equal what is obtainable in irrigated or lowland system. This might be credited to low farm input, high level of variation in terms of soil features and water availability resulting from unpredictable rainfall patterns leading to drought (Kumar et al, 2008). Furthermore, the major focus for most rice breeding programmers in the past was selection in favor of high yield potential and resistance to blast with little or no effort towards selecting for drought tolerance (Imolehin, 1991; Bernier et al, 2008). This led to development of high-yielding cultivars often extremely prone to yield reduction even under mild drought stress conditions (Kumar et al, 2014).

Drought is one of the important abiotic stresses that affects rainfed upland rice production and can occur anytime within the rice growing stages. It is most damaging during flowering and grain filling stages (Lafitte et al, 2004; Atlin et al, 2006; Bernier et al, 2008). To ensure food security and mitigate the negative impacts associated with cultivation of drought susceptible rice varieties, there is need to develop resilient and high-yielding drought-tolerant versions of popular upland rice cultivars in a given upland rice growing region (Kumar et al, 2008). Upland rice cultivars, such as IRAT109, NERICA4, WAB56-109, Moroberekan and Azucena, are reported to carry drought tolerance genes that can be harnessed for drought breeding programmes. Unfortunately, most of the genes, conferring drought tolerance to these cultivars, are yet to be fully characterized and might be associated with unfavorable loci resulting to linkage drag (Dixit et al, 2014; Kumar et al, 2014). To overcome linkage drag related challenges and also improve the precision of introgression, several large effect quantitative trait loci (QTLs) and associated pre-breeding lines (QTL donors) have been reported (Bernier et al, 2008; Dixit et al, 2012; Mishra et al, 2013; Palanog et al, 2014). Some of these pre-breeding lines have already been deployed to effectively develop improved drought tolerant versions of popular farmers’ preferred mega rice varieties across Southeast Asia (Swamy and Kumar, 2013; Shamsudin et al, 2016). For instance,,located on rice chromosome 12, is the first large effect QTL for upland rice ecosystem reported to be associated with grain yield increased under drought stress at the reproductive stage, whileis identified from the same population on rice chromosome 2 contributes to more panicles per unit area under severe stress (Bernier et al, 2007). Vandana introgressed pre-breeding line with, IR84984-83-15-481-B, shows a yield advantage of 0.5 t/hm2compared to itself as the drought tolerant parent under drought while similar to Vandana under non-stress conditions (Swamy and Kumar, 2011). Furthermore, pyramiding of two or more of these drought QTLs might also lead to appreciable yield increase under drought stress. Dixit et al (2012) showed thathaving interaction withunder severe stress always has a positive effect on grain yield irrespective of stress levels. Shamsudin et al (2016) recorded a yield of 1.5 t/hm2in newly developed drought-tolerant MR219 pyramided lines, carrying different combinations of the qDTYs.

Popular upland rice landraces (the Ofada rice cultivars in Nigeria) are widely grown under rainfed drought-prone areas primarily for their unique grain quality rather than yield (Dalton, 2004; Saka and Lawal, 2009). These cultivars are characterized by moderate to high grain yield (2.0–5.0 t/hm2) under optimal water conditions (non-stress, NS) and less than 0.5 t/hm2under drought stress (DS), thus we need to improve these heirloom cultivars to ensure high yield in non-drought years and an acceptable yield in drought years (Saka and Lawal, 2009; Ologbon et al, 2012; Anyaoha et al, 2018). However, limited studies have been reported on the genetic improvement of these important upland rice landraces (Ofada rice cultivars) in Nigeria for drought stress tolerance using marker-assisted breeding techniques and the effects of introgressed QTLs on early generation progenies under field conditions.

This study reported the successful introgression of two drought QTLs (and) into the genetic background of FUNAABOR-2, a popular Ofada rice cultivar in western Nigeria using stepwise phenotypic and marker-assisted selection (MAS), techniques and the effects of interactions between introgressed drought-tolerant QTLs (and) in early generation populations of BC1F2progenies under field conditions.

MATERIALS AND METHODS

Rice materials

A total of 70 backcross progenies under the FUNAABOR-2 genetic background were used, which were derived from backcross breeding programme with two upland rice cultivars IR84984-83-15-481-B and FUNAABOR-2 initiated in 2014 (Fig. 1). The drought tolerant pre-breeding line IR84984-83-15-481-B developed at International Rice Research Insitute was used as the donor parent for QTLsand(Bernier et al, 2007; Palanog et al, 2014), whereas the recipient parent FUNAABOR-2 is a widely grown purified Ofada rice landraces assembled from farmers’ fields in southwestern Nigeria (Showemimo et al, 2011). For the development of breeding population, a stepwise combination of phenotypic and marker-assisted selection was used. Crosses were carried out between the donor and recipient parents to obtain F1seeds. Heterozygous F1s were backcrossed with FUNAABOR-2 to obtain a good number of BC1F1seeds.

Fig. 1.FUNAABOR-2 introgressed BC1F3progenies by marker- assisted selection scheme.

Phenotypic selection and marker-assisted introgression of drought QTLs

Three hundred seeds from 375 BC1F1lines and their two parents were progeny tested in the field during the 2015 planting season under non-stress condition with spacing of 0.25 m within and between rows. Phenotypic selection was imposed on the segregating population to identify backcross progenies with similar grain characteristics to the recipient parent (FUNAABOR-2) and also expressing the agronomic traits of interest to Ofada rice farmers based on information obtained from Participatory Rural Appraisal with upland rice farmers in southwestern Nigeria (unpublished data). Some of the important agronomic traits to the Ofada rice farmers considered during field selections were medium plant height and early to medium maturity in combination with high tillering ability, while round, bold and short grains are the preferred grain shapes and sizes. Seventy BC1F1lines to be evaluated under replicated drought stress were harvested separately from unique individual BC1F1lines with desired agronomic traits and similar grain characteristics to the recipient parent. Also leaf samples collected from the selected 70 BC1F1unique individual lines for evaluation under drought stress were also genotyped for foreground and recombinant selection.

Individual lines that were heterozygous at thelocus were identified at the BC1F1generations using peak marker RM511 (foreground selection). From individual plants heterozygous atlocus, recombinant selections were also carried out by identifying backcross progenies from foreground selection homozygous for the recipient allele (FUNAABOR-2) at two or more marker loci (RM28076 and RM28099) distally flanking thelocus (Neeraja et al, 2007; Collard and Mackill, 2008). Use of flanking markers for recombinant selection facilitates the recovering of important traits of the recipient parent and also minimizes the effects of linkage drag from the qDTY donors (Hospital and Charcosset, 1997; Shamsudin et al, 2016). Phenotypic selection as described earlier was used to identify and select introgression lines (ILs) that looked more like the recurrent parent and also passed the foreground and recombinant selections for advancement according to Shamsudin et al (2016). The progenies carryingidentified from foreground selection were further genotyped with peak marker RM250 associated withon chromosome 2 to identify individual lines harboring a combination of the two drought QTLsandsimultaneously (Fig. 1).

Molecular characterization of rice genotypes

DNA was extracted from freeze-dried leaf samples of each progeny using the modified cetyl trimethyl ammonium bromide (CTAB) mini-prep method at the Biotechnology Laboratory, International Institutes of Tropical Agriculture, Ibadan, Nigeria. Young fresh leaf samples were collected from three-week-old seedlings to extract genomic DNA. The leaves were cut wearing gloves to avoid contamination of the samples and preserved in ice to slow down degradation of leaf DNA. The harvested leaf samples were transferred in polythene bags embedded in an ice box without any damage to the leaves. Finally, the samples were stored in -80 oC and later lyophilized for DNA extraction and PCR.

Table 1.Details of markers associated with drought quantitative trait loci (QTLs)and.

RT, Repeat motif; Chr, Chromosome; AT, Annealing temperature.

The quality of the isolated DNA in the protocol was confirmed using Thermo Scientific NanoDropTM1000 spectrophotometer (Thermo Fisher Scientific, USA). The PCR amplification was done with 15 μL reaction mixture having 2 μL (10–20 ng) of DNA, 1.5 μL of 10× PCR buffer, 1.5 mL of 1 mmol/L dNTPs, 1.0 μL of 5 μmol/L each of forward and reverse primers, 1.0 μl of 1:10polymerase and 7.0 μL of sterile water. The PCR reactions were: initial denaturation at 94 oC for 5 min, then final denaturation at 94 oC for 1 minute and annealing at 55 oC for one minute. Polymerization was carried out at 72 oC for 2 min to complete a cycle and cycle was repeated for 34 times. The final extension was at 72 oC for 7 min. Polymorphisms in the PCR products were detected by ethidium bromide staining after electrophoresis on 1.5% agarose gel using UV transilluminator.

Field establishment and management of drought stress

Seventy BC1F2progenies similar in seed characteristics with the recurrent parent (FUNAABOR-2) were evaluated under field-based upland drought stress conditions at Africa rice drought experimental station located in International Institute for Tropical Agriculture Ikenne, Nigeria during the 2015–2016 dry season. Ikenne (altitude of 60 m, 6o54′ N, 3o42′ E) is located at the rainforest agro-ecological zone of southwest Nigeria. The annual rainfall is about 1 421 mm. Ikenne experiences a long drought spell with little to no rainfall, from November to March every year, followed by a short break in rainfall (2–3 weeks) around August, which is suitable and conducive for dry season drought tolerance screening conditions (Badu-Apraku et al, 2010).

The 70 BC1F4progenies with the two parents FUNAABOR-2 and IR84984-83-15-481-B, a susceptible check (IR64) and two tolerant checks (IRAT109 and NERICA4) were planted in an alpha lattice with three replicates for each treatment. Seeds were sown in single row plots, 2.0-m long with a spacing of 0.2 m between plants row. NERICA4, an early maturing semi-dwarf upland rice cultivar, was used to plant the border rows at the beginning and the end of the rows. For each drought-stress experiment, an irrigated control (non-stress) condition was planted out simultaneously. Irrigation was provided by sprinkler twice a week for both the stress and non-stress plots to maintain the soil at field capacity for seven weeks after sowing (7 WAS). Drought stress was imposed in the stress plots by reducing the frequency of irrigation to once every 10–12 d at 7 WAS when about 50% of most lines had initiated booting. Tensiometer was used to determine the soil-water tension around the root zones at 30 cm soil depth. Stress was maintained until the tensiometer reading was between -50 to -60 kPa and about 75% of the lines expressed severe leaf rolling. Life-saving irrigation was supplied through overhead irrigation to restore the soil at field capacity before initiating the second cycle of drought stress. According to the soil water data and yield difference between the stress plots and the control, imposed drought stress was considered severe after harvest since the yield reduction compared to the irrigated control condition was more than 65%.

The relative yield reduction between the stress and non-stress was estimated according to Kumar et al (2008). Days to flowering (DAF) for each plot was recorded when 50% plants in the plot were headed according to the IRRI standard evaluation system (2014). Six uniform plants in each plot were selected and measured for plant height (PH) using the meter rule at maturity and sampled by cutting the plants from the bottom (at the soil surface). The plants sampled from each plot were put into paper bag and air-dried inside the glass house for 5–7 d. The dried samples were then threshed and measured for total grain yield (GY) per plot.

Fig. 2.Banding pattern of foreground selection for 70 FUNAABOR-2 BC1F1progenies carryingusing peak marker RM511 (A) and 20 selected FUNAABOR-2 introgresed lines carrying a combination ofandusing peak marker RM250 (B).

M, Marker; P1, FUNAABOR-2; P2, IR84984-83-15-481-B; ‘1’, Susceptible allele; ‘2’, Resistant allele; ‘3’, Heterozygote.

Data analysis

Data analysis was performed using the Breeding Management System (2015). Mean values of entries were determined using a model, in which lines were treated as a fixed effects and replications and blocks within replications as random. Breeding lines mean-based broad-sense heritability (H) was computed as(%) =g/ (g+e/) × 100, Wheregis genetic variance,eis error variance, andis the number of replications. Associations between yield and other traits were also estimated.

RESULTS

Foreground and recombinant selection

About 96% of the total F1seeds from the crosses amplified alleles of both parents (heterozygous), indicating their true hybrid nature. Out of 70 BC1F1lines, 20 were heterozygotes (‘3’), 36 showed fixed recipient allele (susceptible allele, score ‘1’) while the remaining showed fixed donor allele (score ‘2’). The gel picture of the foreground selection with RM511 of BC1F1generation (plant number 1 to 70) is presented in Fig. 2-A. Eight promising progenies (lines 30, 33, 42, 48, 59, 62, 64 and 67) out of the twenty introgressed lines carryingalso had(Fig. 2-B).

Phenotypic performance of introgressed lines and heritability estimates

The non-stress condition had a mean grain yield of 192.37 g/m2compared to 29.89 g/m2under drought stress. Mean DAFs among the ILs ranged from 53 to 78 d under non-stress condition and 57 to 89 d under drought stress (Table 2), while plant height varied from 113.33 to 163.33 cm under NS and 84.33 to 149.33 cm under DS. Reductions in yield and plant height followed by increased days to flowering were observed in the DS plots compared to the NS plots. Furthermore, DAFs was delayed by 6 d under DS compared to the NS, while about 80% yield reduction was observed in the DS plots compared to the NS plots. Mean plant height at maturity was also reduced by 19 cm under the DS plots compared to NS plots. Table 2 shows the heritability estimates of the studied traits under the drought stress and non-stress conditions. Heritability was higher in the non-stress condition except for grain yield. High heritability in the broad sense was recorded for GY and DAF under both water regimes while PH had moderate (58%) and high (82%) heritability under stress and non-stress conditions, respectively.

Table 2.Traits for parents andprogenies under drought stress and non-stress conditions.

GY,Grain yield; DAF, Days to flowering; PH, Plant height at maturity.

Yield performance of FUNAABOR-2 progenies

The average performance of the parents and 18 highest yielding FUNAABOR-2 BC1F2progenies under drought stress and non-stress conditions respectively is presented in Table 3. Backcross progenies, BC103 (417.73 g/m2) and BC27-1 (398.48 g/m2) were the top two highest yielding lines under NS, whereas BC91-1 (79.38 g/m2) and BC50 (76.12 g/m2) gave the highest grain yield under DS (Table 3). The two highest yielding lines (BC103 and BC27-1) had about 58% and 52% yield advantage over the recurrent parent FUNNABOR-2 under the NS, respectively, while close to 600% yield advantage was recorded for the two best progenies (BC91-1 and BC50) under the DS conditions. The recurrent parent had a mean reduction in yield of 96% and the drought tolerant parent IR84984-83-15-481-B had 84% mean reduction in grain yield under DS trail compared to the NS.

Effects of QTL combinations on grain yield and days to flowering

The average yield performance of ILs withwas lower compared to ILs having two qDTYs under DS (Table 4). However, under DS condition, the mean grain yield of ILs with combination ofandwas 40.12 g/m2, while ILs withgave a yield of 31.13 g/m2. Progenies with no qDTYs had the least mean grain yield (29.2 g/m2) under DS condition. For the NS conditions, ILs with/combination also had the highest yield followed by ILs with onlywhile the least grain yield was observed in ILs having no qDTY. Mean grain yields of 206.73, 197.92 and 190.01 g/m2were recorded for ILs with/,and no qDTY, respectively.

The effects of QTL combinations on mean DAF of the ILs with single qDTY, combination of qDTYs and no qDTY are shown in Table 4. In the NS condition, ILs withflowered the earliest (61 d), followed by ILs without qDTY (65 d), while ILs with the two qDTYs (and) flowered last at 74 d. However, under DS conditions, lines without qDTY were the earliest, flowering at 69 d followed by ILs with(72 d). Lines with the combination of the two drought QTLs (and) also flowered last at 73 d under DS condition.

Table 3. Quantitative trait locus and grain yield of 18 best yielding backcross progenies at BC1F2generation.

GY, Grain yield; DAF, Days to flowering; PH, Plant height at maturity; Y,Presence of qDTY; N,Absence of qDTY., Probability level at 0.05 level.

Table 4. Effects of quantitative trait locus combinations on FUNAABOR-2 introgressed lines at BC1F2generation.

GY, Grain yield; DAF, Days to flowering; PH, Plant height; EF,Yield of early maturing lines; MF,Yield of medium flowering lines; LF, Yield of late flowering lines.

To compare mean yield across the progeny lines based on early, medium and late flowering lines without considering presence of qDTYs, the late flowering lines produced the highest grain yield (219.00 g/m2) under NS conditions, followed by medium flowering lines (178.94 g/m2) whereas the early lines gave the least yield of 153.09 g/m2(Table 4). However, comparing mean yield based on effects of QTLs across flowering dates in the stress condition, it was observed that for late flowering, ILs carrying single QTL () gave the highest mean yield of 38.67 g/m2, followed by medium and early flowering lines with mean yield of 17.3 and 5.3 g/m2, respectively (Table 4). For ILs with combination of two drought QTLs,andunder DS, the late flowering lines also had the highest yield of 43.38 g/m2while none of the ILs in this category was in the early flowering group (Table 4). Considering lines with onlyunder the NS condition, late flowering lines were also the best performers with mean yield of 239.17 g/m2, followed by early and medium flowering lines that produced 174.58 and 154.7 g/m2, respectively (Table 4). Meanwhile, for progenies with qDTY combinations, late and medium flowering backcross progenies produced yields of 237.5 and 175.96 g/m2, respectively, whereas none of the early flowering lines had combination of the two qDTYs.

DISCUSSION

Drought is the most important abiotic stress to rainfed upland rice production across the Sub-Sahara Africa. Developing new rice varieties with high grain yield and improved tolerance to various abiotic stresses using farmer-preferred cultivars is a sustainable path to enhance adoption of new improved varieties and ensure food security in drought-prone rice producing areas across Nigeria.

Selection of appropriate QTL donor and receptor coupled with a combination of phenotypic and marker- assisted selection led to precise introgression of two drought QTLsandinto the genetic background of Ofada rice cultivar FUNAABOR-2 in this study. The step-wise breeding technique led to selection of backcross progenies with different QTL combinations at the early stages of the breeding progamme, thereby reducing the number of individuals to be genotyped and cost of genotyping. Shamsudin et al (2016) also employed step-wise screening technique that effectively combines marker-assisted backcrossing and good field phenotype to identify backcross introgressed lines with desirable QTL combinations. This technique also leads to significant reduction in time and cost of genotyping due to reduced number of individual plants screened in each step (Sreewongchai et al, 2010). Selection of appropriate parents, good field phenotyping complemented by foreground, and recombinant selection also facilitate the reduction of the size of donor chromosome in the promising backcross progenies. Furthermore, identification of ILs harboring different combinations of drought QTLs with similar grain shapes to the recurrent parent at early generation (BC1F2) also ensures that only progenies with the desired agronomic traits are advanced in the population. This is one of the main advantages of systematic combination of conventional breeding techniques (good phenotyping) and marker-assisted selection since this approach might lead to reduced breeding cycle with increased efficiency (Swamy and Kumar, 2013).

The backcross progenies with their parents were subjected to severe DS, and this was established by yield reduction of more than 65% compared to NS. This level of stress (severe stress) aids to separate high grain yield potential under stress from drought tolerance (Kumar et al, 2008). Significant levels of variations were observed for the agronomic traits considered among ILs and the two parents. The presence of varying performances recorded among evaluated progenies compared to their two parents suggests that selection to identify transgressive segregants with higher grain yield potentials over the recurrent parents under DS and NS conditions might be effective. Mean grain yield were higher under the NS compared to the DS, further confirming yield reduction due to drought stress in rice. Similar results have been reported on the effects of drought stress on rice yield (Atlin et al, 2006; Venuprasad et al, 2008; Sohrabi et al, 2012; Shamsudin et al, 2016). However, mean DAF among ILs was longer in the DS condition compared to the NS condition, which indicates that occurrence of drought stress at the reproductive stage in this population might lead to flowering delayed (Ghimire et al, 2012; Dixit et al, 2014; Shamsudin et al, 2016). Most of the progenies identified in this study flowered earlier under both DS and NS conditions compared to the recurrent parent FUNAABOR-2. This can be attributed to the linkage of most reproductive stage qDTYs with early flowering as reported by Kumar et al (2014).

Grain yield under drought has been reported to be a complex trait with low heritability under the influence of many genes. However, recent findings from IRRI have shown otherwise by reporting moderate to high heritability for grain yield under RS condition (Kumar et al, 2008; Swamy and Kumar, 2013; Dixit et al, 2014). The high estimates of heritability in the broad sense recorded for grain yield under DS and NS conditions in this study further confirms earlier reports about the effectiveness of direct selection for high grain yield under upland drought stress (Venuprasad et al, 2007; Kumar et al, 2008). High heritability for yield and other traits under DS and NS conditions observed in this study are in line with the earlier reports (Mishra et al, 2013; Palanog et al, 2014; Shamsudin et al, 2016).

Eighteen BC1F2FUNAABOR-2 ILs having yield advantage of 31.48 to 68.12 g/m2under DS were identified as promising lines for advancement and further evaluation under DS. BC1F2progenies with a combination of drought QTLs (and) were better performers compared to lines with only. Some of the promising FUNAABOR-2 ILs among the top performers under DS also showed high yield potentials under NS condition, confirming the reports of Venuprasad et al (2007) on the feasibility of developing drought tolerant varieties with high yield under NS conditions. Three ILs BC91-1, BC69 and BC27-1 all carrying the two qDTYs (and) were among the top seven lines that had the highest grain yields under the reproductive stage drought stress condition. Backcross progenies BC27-1 and BC57 with eitheror in positive significant combination withalso produced high grain yield under both DS and NS conditions.

This study should be the first reports on the effects of introgreesed drought QTLs (and) and their combined effects on yield and DAF in the genetic background of popular farmers’ preferred Ofada rice variety FUNABOR-2 in Nigeria. FUNAABOR-2 ILs carrying onlyflowered 10 d earlier compared to the recurrent parent under NS condition. ILs carrying a single or combination of the two qDTYs also flowered earlier than the recurrent parent under DS conditions, indicating positive effect of qDTYs on DAF under drought stress. It can also be inferred that introgression of singlein the genetic background of FUNAABOR-2 resulted in earlier flowering lines than that of combining the two qDTYs. Earlier reports showed positive effects of qDTYs on DAF under DS and NS conditions (Bernier et al, 2007; Dixit et al, 2012). Furthermore, selection in favour of tall, late flowering lines carryingandmight also lead to high- yielding drought tolerant lines as BC91-1 under DS in this study. Vickram et al (2015) also reported tight linkage between high grain yield and tallness under drought stress.

As already mentioned, ILs carrying onlyor with two qDTYs produced higher grain yield under DS compared to the recurrent parent, thus indicating positive effects of these QTLs. The consistent positive effects of these drought QTLs on high grain yield production under drought stress have been reported (Bernier et al, 2008; Dixit et al, 2012; Mishra et al, 2013; Palanog et al, 2014). Under the NS conditions, ILs positive for single qDTY or combination of the two drought qDTYs also produced higher grain yield compared to ILs negative for the qDTYs. This indicates that the qDTYs probably do not only express their effects on high grain yield in the FUNAABOR-2 genetic background under drought stress alone but also under non-stress conditions. The results also agree with the findings of Bernier et al (2007) and Palonog et al (2014) thatalso shows positive effects on grain yield under NS. Swamy and Kumar (2012) reported that similar QTLs (and) for drought tolerance identified in rice also increase grain yield under DS and NS conditions. Recently, Shamsudin et al (2016) reported the positive effects of pyramiding three drought QTLs (,and) on high grain yield pyramided lines developed from high-yielding, popular but drought susceptible Malaysian rice cultivar MRQ74.

CONCLUSIONS

Step-wise phenotypic screening and marker-assisted backcross were successfully employed in this study to develop promising backcross FUNAABOR-2 ILs that produced a yield advantage over the recurrent parent FUNAABOR-2 under DS while also giving higher appreciable grain yield under NS conditions. This establishes the feasibility of improving local upland rice landraces in Nigeria for high grain yield under optimal water condition and acceptable grain yield under drought stress at reproductive stage simultaneously. Presence of two drought QTLs in the FUNAABOR-2 background gave better grain yield under DS in this study whereas introgression of a single QTL seemed to be more promising when earliness is the trait of interest. From the results of this study, we recommended step-wise combination of molecular techniques and field based phenotyping of breeding populations at early generation stages. This will help to reduce cost and also ensure that selected lines for advancement possess the desired traits of interest.

Acknowledgements

We are grateful to the alliance for a Green Revolution in Africa for providing funds for this study through the West Africa Centre for Crop Improvement, University of Ghana Ph. D fellowship. We also appreciate Dr. A. Kumar of International Rice Research Institute and Dr. S. O. Adigbo of the Federal University of Agriculture Abeokuta, Nigeria for making available the QTL donors (IR84984-83-15-481-B) and Ofada rice cultivar (FUNAABOR-2).

Anyaoha C, Fofana, M, Gracen V, Tongoona P, Blay E, Semon M, Popoola B. 2018. Yield potential of upland rice varieties under reproductive-stage drought and optimal water regimes in Nigeria., 16(4): 378–385.

Atlin G N, Cooper M, Bj?rnstad A. 2006. A comparison of formal and participatory breeding approaches using selection theory., 122(3): 463–475.

Badu-Apraku B, Menkir A, Ajala S O, Akinwale R O, Oyekunle M, Obeng-Antwi K. 2010. Performance of tropical early-maturing maize cultivars in multiple stress environments., 90(6): 831–852.

Bernier J, Kumar A, Ramaiah V, Spaner D, Atlin G. 2007. A large-effect QTL for grain yield under reproductive-stage drought stress in upland rice., 47(2): 507–516.

Bernier J, Atlin G N, Serraj R, Kumar A, Spaner D. 2008. Breeding upland rice for drought resistance., 88(6): 927–939.

Collard B C Y, Jahufer M Z Z, Brouwer J B, Pang E C K. 2005. An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: The basic concepts., 142: 169–196.

Collard B C Y, Mackill D J. 2008. Marker-assisted selection: An approach for precision plant breeding in the twenty-first century., 363: 557–572.

Dalton T J. 2004. A household hedonic model of rice traits: Economic values from farmers in West Africa., 31: 149–159.

Dixit S, Swamy B P M, Vikram P, Ahmed H U, Sta Cruz M T, Amante M, Atri D, Leung H, Kumar A. 2012. Fine mapping of QTLs for rice grain yield under drought reveals sub-QTLs conferring a response to variable drought severities., 125: 155–169.

Dixit S, Singh A, Kumar A. 2014. Rice breeding for high grain yield under drought: A strategic solution to a complex problem., 2014: 1–15.

Ghimire K H, Quiatchon L A, Vikram P, Swamy B P M, Dixit S, Ahmed H, Hemandez J E, Borromeo T H, Kumar A. 2012. Identification and mapping of a QTL () with a consistent effect on grain yield under drought., 131: 88–96.

Hospital F, Charcosset A. 1997. Marker-assisted introgression of quantitative trait loci., 147(3): 1469–1485.

Imolehin E D. 1991. Rice improvement and production in Nigeria. WARDA Upland Breeding Task Force Workshop, Bouake.

Kumar A, Bernier J, Verulkar S, Lafitte H R, Atlin G N. 2008. Breeding for drought tolerance: Direct selection for yield, response to selection and use of drought-tolerant donors in upland and lowland-adapted populations., 107(3): 221–231.

Kumar A, Dixit S, Ram T, Yadaw R B, Mishra K K, Mandal N P. 2014. Breeding high-yielding drought-tolerant rice: Genetic variations and conventional and molecular approaches., 65(21): 6265–6278.

Lafitte R 2003. Managing water for controlled drought in breeding plots.: Fischer K S, Lafitte R, Fukai S, Atlin G, Hardy B. Breeding Rice for Drought-Prone Environments. Los Banos, the Philippines: International Rice Research Institute: 23–26.

Lafitte H R, Price A H, Courtois B. 2004. Yield response to water deficit in an upland rice mapping population: Associations among traits and genetic markers., 109(6): 1237–1246.

Mishra K K, Vikram P, Yadaw R B, Swamy B P M, Dixit S, Sta Cruz M T G, Marker S, Kumar A. 2013.: A locus with a consistent effect on grain yield under drought in rice., 14: 12.

Neeraja C N, Maghirang-Rodriguez R, Pamplona A, Heuer S, Collard B C Y, Septiningsih E M, Vergara G, Sanchez D, Xu K, Ismail A M, Mackill D J. 2007. A marker-assisted backcross approach for developing submergence-tolerant rice cultivars., 115(6): 767–776.

Ologbon O A C, Ikheloa E E, Akerele E O. 2012. Adoption of ‘Ofada’ rice variety and technical efficiency of rice-based production systems in Ogun State, Nigeria., 8(6): 624–631.

Palanog A D, Swamy B P M, Shamsudin N A A, Dixit S, Hernandez J E, Boromeo T H, Sta Cruz P C, Kumar A. 2014. Grain yield QTLs with consistent-effect under reproductive- stage drought stress in rice., 161: 46–54.

Saka J O, Lawal B O. 2009. Determinants of adoption and productivity of improved rice varieties in southwestern Nigeria., 8(19): 4923–4932.

Shamsudin N A A, Swamy B P M, Ratnam W, Sta Cruz M T, Raman A, Kumar A. 2016. Marker assisted pyramiding of drought yield QTLs into a popular Malaysian rice cultivar, MR219., 17(1): 30.

Showemimo F A, Gregorio G, Olowe V I O, Ukwungwu M N, Maji A, Adigbo S O, Olaoye O J, Akintokun P O, Bodunde J G, Idowu O T H, Awe C A. 2011. Varietal release: Release of two dual purpose ofada rice varieties (Funaabor-1 and Funaabor-2) by federal university of agriculture, abeokuta ()., 11(2): 122–123.

Sreewongchai T, Toojinda T, Thanintorn N, Kosawang C, Vanavichit A, Tharreau D, Sirithunya P. 2010. Development of eliterice lines with wide spectrum of resistance to Thai blast isolates by pyramiding multiple resistance QTLs., 129(2): 176–180.

Swamy B P M, Kumar A. 2011. Sustainable rice yield in watershort drought-prone environments: Conventional and molecular approaches.: Lee T S. Irrigation Systems and Practices in Challenging Environments. Rijeka, Croatia: InTech: 149–168.

Swamy B P M, Kumar A. 2012. Sustainable rice yield in water-short drought-prone environments: Conventional and molecular approaches. INTECH Open Access Publisher.

Swamy B P M, Kumar A. 2013. Genomics-based precision breeding approaches to improve drought tolerance in rice., 31(8): 1308–1318.

Venuprasad R, Lafitte, H R, Atlin G N. 2007. Response to direct selection for grain yield under drought stress in rice., 47(1): 285–293.

Vikram P, Swamy B P M, Dixit S, Singh R, Singh B P, Miro B, Kohli A, Henry A, Singh N K, Kumar A. 2015. Drought susceptibility of modern rice varieties: An effect of linkage of drought tolerance with undesirable traits., 5: 14799.

28 February 2018;

27 April 2018

Anyaoha Christian Okechukwu (kriskoty@yahoo.com)

Copyright ? 2019, China National Rice Research Institute. Hosting by Elsevier B V

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Peer review under responsibility of China National Rice Research Institute

http://dx.doi.org/10.1016/j.rsci.2018.04.006

(Managing Editor: Wang Caihong)