Phenotypic variations of days to flowering, maturity and other associated agronomic traits
Significant differences (p <0.05) were observed among the genotypes for all the phenotypic traits measured (Table 3). There was also a significant difference between years and the genotype by year interactions for these traits, except for VE-R7. This shows a high genetic variability of the tropical genotypes used in the study. Days to from emergence to beginning bloom (VE-R1) ranged from 21.67 to 41.45 days among the 159 genotypes. A box plot of VE-RI (Fig 1) showed two outliers (TGX 1987-62F and IC192) with flowering days of 35.4 and 41.8, respectively. Days from emergence to physiological maturity (VE-R7) ranged from 63.89 to 85.08 days. The plant height of the genotypes ranged from 15.3 to 68.4 cm, a box plot of plant height of genotypes showed two outliers (IC192 and CIGRAS-06). The plant heights of the two genotypes were 68.5 and 64.3 cm, respectively and the number of nodes and pods per plant ranged from 5.43 to 13.92 and 4 to 27.59 respectively.
Significant positive correlations were observed among all phenotypic traits studied (Table 4). A significant (
p<0.05) and positive correlation was observed between VE-R1 and VE-R7. The correlations of plant height at maturity with the numbers of nodes per plant and pods per plant were highly positive. Plant height is an important trait that affects soybean adaptation and yield. A positive significant correlation was observed among the phenotypic traits indicating that delayed flowering resulted in an increase in the other important agronomic traits such as plant height, which sometimes correlates with high yields in tropical regions (Cober and Morrison, 2010).
Selection of the genotypes that delayed flowering based on the phenotypic data
Sanya-Hainan Islands, the experimental site used for phenotyping of the long juvenile trait, is the southernmost region of China with a typical tropical climate. The day length and temperatures provided an ideal environment for screening for the long juvenile trait in China. Two long juvenile varieties Huaxia-3 and Huaxia-10 reported by
Yue et al., (2017) and
Lu et al., (2017) were included in the study to serve as checks. The two check varieties flowered about 32-33 days after emergence. Since the determination of juvenility of a genotype is determined by the photoperiod of the given environment, genotypes that have same or similar flowering days as the check varieties or flowered later than the check varieties were considered to be long juvenile. Based on the phenotypic data we have identified 53 genotypes that flowered 30 days and above (Table 5).
SNP haplotyping to distinguish different alleles of the J gene
Based on the phenotypic differences observed, we have genotyped the 159 soybean genotypes and carried out haplotype analysis to determine the genotypes carrying the functional or non-functional alleles of the gene. Sequence comparison identified 10 polymorphisms comprising of seven SNPs and three deletions in the coding sequence, three of which resulted in the loss-of-function of the
J gene. Fig 2 shows the position of the base substitution in the coding sequence and the number of haplotypes identified. Seven haplotypes were defined of which three haplotypes (
j1,
j2 and
j4) resulted in the loss-of-function of the
J gene. The other functional haplotypes include Hap1, Hap2, Hap3 and Hap 4. Only one soybean germplasm from Brazil MG/BR-48 had the loss-of-function allele
j2. Huaxia 3 and Huaxia 10, the two Chinese tropical check varieties, had the
j4 allele. Two genotypes TGx 1835-10E and TGx 1987-62F from IITA-Nigeria had the
j2 allele. I.C.192 from the US had the
j1 allele (Table 5, Fig 3). The identification of these alleles confirms studies by
Lu et al., (2017) that the multiple loss-of-function
J alleles existing in soybean germplasm worldwide may be responsible for delayed flowering and maturity.
Association of the haplotypes with the phenotypic traits
Fig 3 shows the variability of the phenotypic traits among the haplotypes defined. Haplotypes Hap1 and Hap3 are the most common among the genotypes with high variability in all the six phenotypic traits. Hap 4 is the least common in only two genotypes. Some genotypes with functional
J haplotypes had similar flowering days and, in some cases, even higher (Table 3) than those harboring the loss-of-function alleles indicating that the
J gene might not be the only gene responsible for the long juvenile trait. The same observations were recorded on the other associated phenotypic traits such as the plant height, number of nodes per plant and number of pods per plant. This confirms the polygenic nature of the long juvenile trait by previous studies
(Ray et al., 1995, Carpentieri-Pipolo et al., 2000, Cober and Morrison 2010,
Lu et al., 2017, Yue et al., 2017 and
Fang et al., 2020).
The present study mainly focused on identifying new long juvenile genotypes and verifying whether the
J mutations are associated with delayed flowering in soybean. The long juvenile genetic mechanism in most tropical genotypes especially Sub-Sahara Africa is not known
(Miranda et al., 2020). Huaxia 3, the check variety, was introduced to some African countries and resulted in an increased productivity
(Yue et al., 2017). Introgression of the loss-of-function alleles Hap
j2 and Hap
j4 to temperate cultivars has also resulted in the development of new long juvenile varieties in Brazil and southern China
(Lu et al., 2017), indicating that the loss-of-function
J alleles will be a good strategy to enhance the adaptation and yield increases in the tropics. Therefore, there exists a possibility that incorporation of the identified long juvenile genotypes with or without the loss of function alleles for the
J gene in tropical soybean breeding programs will lead to yield increases. However, proper yield trials have to be conducted first, as past studies reported background effects could influence delayed flowering in soybean
(Ray et al., 1995). The lack of Selection of the genotypes that delayed flowering based on the phenotypic data Sanya-Hainan Islands, the experimental site used for phenotyping of the long juvenile trait, is the southernmost region of China with a typical tropical climate. The day length and temperatures provided an ideal environment for screening for the long juvenile trait in China. Two long juvenile varieties Huaxia-3 and Huaxia-10 reported by
Yue et al., (2017) and
Lu et al., (2017) were included in the study to serve as checks. The two check varieties flowered about 32-33 days after emergence. Since the determination of juvenility of a genotype is determined by the photoperiod of the given environment, genotypes that have same or similar flowering days as the check varieties or flowered later than the check varieties were considered to be long juvenile. Based on the phenotypic data we have identified 53 genotypes that flowered 30 days and above (Table 5).
SNP haplotyping to distinguish different alleles of the J gene
Based on the phenotypic differences observed, we have genotyped the 159 soybean genotypes and carried out haplotype analysis to determine the genotypes carrying the functional or non-functional alleles of the gene. Sequence comparison identified 10 polymorphisms comprising of seven SNPs and three deletions in the coding sequence, three of which resulted in the loss-of-function of the
J gene. Fig 2 shows the position of the base substitution in the coding sequence and the number of haplotypes identified. Seven haplotypes were defined of which three haplotypes (
j1,
j2 and
j4) resulted in the loss-of-function of the
J gene. The other functional haplotypes include Hap1, Hap2, Hap3 and Hap 4. Only one soybean germplasm from Brazil MG/BR-48 had the loss-of-function allele
j2. Huaxia 3 and Huaxia 10, the two Chinese tropical check varieties, had the
j4 allele. Two genotypes TGx 1835-10E and TGx 1987-62F from IITA-Nigeria had the
j2 allele. I.C.192 from the US had the
j1 allele (Table 5, Fig 3). The identification of these alleles confirms studies by
Lu et al., (2017) that the multiple loss-of-function
J alleles existing in soybean germplasm worldwide may be responsible for delayed flowering and maturity.
Association of the haplotypes with the phenotypic traits
Fig 3 shows the variability of the phenotypic traits among the haplotypes defined. Haplotypes Hap1 and Hap3 are the most common among the genotypes with high variability in all the six phenotypic traits. Hap 4 is the least common in only two genotypes. Some genotypes with functional
J haplotypes had similar flowering days and, in some cases, even higher (Table 3) than those harboring the loss-of-function alleles indicating that the
J gene might not be the only gene responsible for the long juvenile trait. The same observations were recorded on the other associated phenotypic traits such as the plant height, number of nodes per plant and number of pods per plant. This confirms the polygenic nature of the long juvenile trait by previous studies
(Ray et al., 1995, Carpentieri-Pipolo et al., 2000, Cober and Morrison 2010,
Lu et al., 2017, Yue et al., 2017 and
Fang et al., 2020).
The present study mainly focused on identifying new long juvenile genotypes and verifying whether the
J mutations are associated with delayed flowering in soybean. The long juvenile genetic mechanism in most tropical genotypes especially Sub-Sahara Africa is not known
(Miranda et al., 2020). Huaxia 3, the check variety, was introduced to some African countries and resulted in an increased productivity
(Yue et al., 2017). Introgression of the loss-of-function alleles Hap
j2 and Hap
j4 to temperate cultivars has also resulted in the development of new long juvenile varieties in Brazil and southern China
(Lu et al., 2017), indicating that the loss-of-function
J alleles will be a good strategy to enhance the adaptation and yield increases in the tropics. Therefore, there exists a possibility that incorporation of the identified long juvenile genotypes with or without the loss of function alleles for the
J gene in tropical soybean breeding programs will lead to yield increases. However, proper yield trials have to be conducted first, as past studies reported background effects could influence delayed flowering in soybean
(Ray et al., 1995). The lack of proper yield data is a key limitation to the study. Future studies will focus on comprehensive yield trials on the identified long juvenile genotypes.
