Detection of Quantitative Trait Loci (QTL) associated with the spring regrowth vigor trait in alfalfa (Medicago sativa L.)


 Background: Alfalfa ( Medicago sativa L.) is a perennial forage legume with a reputation as being the “queen of forage”. Spring regrowth vigor refers to the process of perennial alfalfa returning to growth after winter survival. The objective of this research was to identify candidate genes significantly associated with spring regrowth vigor.Results: We used a tetraploid alfalfa F1 population comprised of 392 progenies to identify quantitative trait loci (QTL) that control this trait. The F1 population phenotypic data were collected using a total of three environmental phenotypic data. The mapping population was genotyped using Genotyping-by-sequencing (GBS), and linkage maps were developed based on single nucleotide polymorphism (SNP) markers. Fifteen significant QTL for spring regrowth vigor were detected in both parents. Five QTL were identified in the male genetic map, while ten QTL were identified in the female. Four QTL were located on homolog 7D of the male parent, and two QTL were colocalized. Five QTL were mapped on homolog 6D of the female parent, and two QTL were colocalized.Conclusions: The QTL presented in this study can be used to improve the efficiency of alfalfa breeding programs and provide valuable resources for the genetic improvement of alfalfa using marker-assisted selection (MAS).β These authors contributed equally to this research. *Correspondence:Junmei Kangkangjunmei@caas.cn

4 cycle and has low efficiency. Modern programs are turning to breeding techniques based on genotyping, including QTL mapping, genome wide association studies (GWAS) and genomic selection (GS), which offer the promise of more rapid breeding cycles and fewer necessary phenotypic evaluations [5].
Spring regrowth vigor refers to the process of perennial alfalfa returning to growth after winter survival. For perennial forage, spring regrowth is an important index to measure its ability to recover growth after fall dormancy and winter hardiness. Moreover, plants with an excellent spring regrowth vigor trait can reduce the pressure of weeds in the spring, which is conducive to growth and has an important impact on the yield and quality of the first cut of alfalfa. Perennial forage yields are believed to be strongly influenced by the amount of organic reserves developed during the last growing season [6]. These reserves are predominantly nonstructural carbohydrates, including quantities of nitrogenous compounds and accumulate in the roots and crowns of plants [7]. Dhont et al. (2006) reported that an untimely autumn defoliation of alfalfa reduces root accumulation of specific N reserves such as proline, arginine, histidine and vegetative storage proteins that are positively related to the vigor of spring regrowth but poorly related to winter survival [8]. At the same time, spring regrowth is also affected by the last harvest in the previous year. Harvests at full bloom allow for greater spring regrowth than cutting at the late bud stage, possibly because of the accumulation of higher root reserves [9,10]. For perennial grass, carbohydrate and nitrogen reserves and their remobilization to active growth sites provide substrates for spring regrowth [11].
Major QTL related to yield [12,13], flowing time [14], water use efficiency [15], fall dormancy and winter-hardiness [16,17] and leaf rust resistance [18] have been mapped in tetraploid alfalfa. Mccord et al. (2014) identified QTL and markers associated with forage yield, resistance to lodging, and spring vigor in alfalfa [19]. Four QTL for spring vigor were identified, and some of the QTLs were located at the same or similar positions as those related to forage yield, possibly explaining the significant correlation between these traits. Genetic mechanisms of spring regrowth in perennial alfalfa are not well understood due to the quantitative trait of spring regrowth and the complex genetic nature and genomes of most popular perennial plants. Exploration of genetic mechanisms underlying spring regrowth is useful for breeding programs aimed at improving the winter hardiness of perennial forage grasses.
Genetic maps are very important for QTL mapping [20], and single nucleotide polymorphism (SNP) markers can be used to represent the most abundant sources of variation in genomes. SNP markers are ideal for constructing genetic maps because of their large quantity and low cost [21], and SNPbased genetic maps have been developed in several corp species, such as wheat [22], rice [23], and alfalfa [18]. A large number of SNP markers can be obtained by sequencing methods such as genotyping-by-sequencing (GBS), RNA-seq and restriction associated DNA sequencing (RAD-seq). GBS [24], as described by Elshire et al. (2012), is one of the most popular reduced-representation approaches for genomic selection in crop species [5] and utilizes selected restriction enzymes such as methylation-sensitive enzymes to cut the genome at fixed points and then sequences the restriction fragment ends for genotyping [24].
In the present study, we established an F1 population consisting of 392 progeny lines. Genotyping was done using GBS, and linkage maps were constructed by GBS-SNP markers. The objective of this study was to map QTL associated with alfalfa spring regrowth vigor. The identified QTL significantly correlated with the alfalfa spring regrowth vigor trait and will provide important reference information to understand the genetic basis of spring regrowth vigor, and it can also be used in MAS breeding programs of alfalfa.

Results
Phenotypic data Significant phenotypic differences among the 392 F1 individuals were found in each environment.
Additionally, the genotype × year interaction was significant. The F1 population exhibited transgressive segregation (    Five QTL for male parents and ten QTL for female parents were mapped on genetic linkage maps for the phenotypic datasets. Left markers, marks at the left of the LOD peak; right markers, marks at the right of the LOD peak; LG, linkage group; interval (cM), 1-LOD support interval; LOD, the logarithm of the odds; PVE, the percentage of the phenotypic variation explained by QTL; Add: the additive effects of the QTL.

Discussion
In alfalfa, no consensus DNA map exists, despite that a limited number of genetic maps were published, and genomic resources are very scarce. Before the development of third-generation sequencing, obtaining a large number of markers was very difficult. Genetic maps were constructed by SSR, RFLP, and other marker types not based on large-scale genome sequencing [5]. An early reported tetraploid alfalfa linkage map only had seven linkage groups with 443 cM [28]. With the development of sequencing technology, SNP markers are widely used in genetic maps. SNP markers can be generated using many sequencing methods such as RNA-seq [29] and RAD-Seq [13], and we used GBS to sequence the 392 alfalfa individuals. Using the GBS sequencing method, genetic linkage maps have been constructed for rice [30], barley and wheat [20], olive [31,32], alfalfa [26] and other plant species.
Spring regrowth vigor is a complex trait that is controlled by multiple genes. While it has been studied in perennial ryegrass [29]. However, genetic studies of spring regrowth vigor are limited, and genes controlling spring regrowth vigor have not been elucidated. Since spring regrowth is complex and involves a large number of genes, some key genes likely have only a small phenotypic variation. This notion indicates that QTL mapping analysis of complex quantitative traits in alfalfa could be challenging and that large populations will be needed [33]. Compared with other QTL mapping populations in alfalfa [12,17] and other species [34,35], the population in our study was sufficiently large for the QTL mapping of spring regrowth vigor traits.
Among the 15 QTL identified in our study, 9 QTL were located on 7D of the male parent and on 6D of the female parent. These nine QTL explained the high phenotypic variation (PVE > 10% A previous study by Mccord et al. (2014) reported QTL for alfalfa spring regrowth vigor [19].
Unfortunately, no QTL in our study were located at the same or similar positions as that reported in their study [19]. They reported a significant correlation between forage yield and spring vigor because of colocalized QTL that were identified for these traits. In the present study, two QTL (qLF2019 ~ 5, qCP2019 ~ 4) were located at similar positions to the QTL associated with yield in the study [19]. At the same time, QTL detected on chromosome 6 (qLF2019 ~ 4, qCP2019 ~ 1) in this study were located at similar genomic locations as reported in previous studies [12]. The QTL were used to BLAST search against the M. truncatula genome [18,36], and flanking markers were used for comparative analysis of the genomic region. However, M. truncatula is an annual plant that has little regrowth ability [37]. Sakiroglu and Brummer (2017) found that no SNPs associated with spring regrowth were mapped to the M. truncatula genome [38]. In our study, we only found some genes (MTR_7g118350, MTR_7g118120, and MTR_6g069870) that were associated with forage yield in our QTL region (QTL name). These results further indicate that there is a significant correlation between alfalfa spring vigor and forage yield.
The QTL analysis performed in this study revealed that some of the QTL associated with spring vigor were located at a similar region as those associated with fall dormancy and winter injury in previous studies [16,17]. Adhikari et al. (2018) mapped four QTL associated with winter injury on chromosome Unfortunately, there was no QTL located in the same or similar interval to the QTL associated with fall dormancy in their study [17].  identified six QTL associated with fall dormancy on chromosomes 4, 6 and 7, and five QTL associated with winter injury on chromosomes 3 and 7 [16].
These QTL were located in the same or similar intervals as the QTL (qLF2019 ~ 5, qCP2018 ~ 3, qLF2019 ~ 1, qLF2019 ~ 4, qCP2019 ~ 1) associated with spring regrowth vigor in our study. However, due to differences in mapping populations and methods used in the research process, further investigation of the association between these traits in alfalfa is warranted.
The traits related to QTL and SNPs that were stably present in the different environments and had a higher percentage of the phenotypic variation can be used to accelerate alfalfa breeding programs.
However, the QTL identified in this paper were not sufficient for use in direct molecular breeding because a single quantitative trait locus may contain hundreds of genes [39], and some QTL may only exist in a specific population [40]. Therefore, it is necessary to verify the reported QTL in different genetic backgrounds and multiple environments.

Conclusion
We have established an F1 population of 392 individuals, and QTL mapping analysis was performed using the genetic linkage maps that were constructed in a previous study. We identified 15 QTL associated with spring regrowth vigor in the F1 population. Four QTL were located on LG 7D of the male parent, and five QTL were located on LG 6D of the female parent. Both male and female parents had one colocalized quantitative trait locus on LG 7D and LG 6D, respectively. A correlation was observed between spring regrowth vigor and forage yield. These results suggest that there is a significant correlation between alfalfa spring vigor and forage yield. Our results will provide useful information for molecular breeding of alfalfa. These QTL will be a valuable genomic resource for the development of alfalfa spring regenerative vigor by MAS.

Mapping Population
An F1 population consisting of 392 progeny lines was described in our previous study (Zhang et [26]. QTL IciMapping was used to identify QTL with the functionality of BIP by inclusive composite interval mapping with an additive effect (ICIM-ADD) [27]. A LOD threshold of 3.0 was used for declaring significant QTL and other parameters remained at the default value. The QTL detected for each parental map were indicated on linkage maps using MapChart.
Alfalfa phenotypic data were collected as regrowth height and plant density after fall dormancy and winter hardiness in the spring. To ensure consistency between the plants at the two locations, the last clipping was performed simultaneously. We collected phenotypic data at LF (2018) and CP (2018, 2019). We have comprehensively considered the plants height and cover density. Specifically, spring regrowth vigor was rated on April 25 at LF and on April 27 at CP using a 0-3 scale in which 0, 1, 2, and heritability (H2) was calculated using the function AOV of QTL IciMapping [27]. The mean data for each environment (CP2018, CP2019, LF2018) were used for QTL analysis.