RESULTS
Sequence characteristics and variation study of the mtCOI gene fragment in M. graminicola populations: The COI gene fragments of 54 M. graminicola populations isolated in 10 provinces of China were PCR-amplified and sequenced. The sequence fragment lengths of 787 bp were obtained by sequence splicing and multiple sequence alignment. The accession numbers of the generated sequences are shown in Table 1. Thirty-nine polymorphic loci were found (4.9% of the total number of bases examine), with 12 S-singleton sites and 27 parsimony-informative sites, accounting for 30.8% and 69.2% of the total polymorphisms, respectively. The S-singleton sites on the mtCOI gene fragment were found at positions of 205, 260, 285, 301, 305, 312, 315, 379, 418, 498, 546and 582, whereas the parsimony-informative sites were found at positions of 51, 114, 182, 188, 195, 266, 289, 330, 333, 353, 381, 414, 441,446, 447, 468, 519, 523, 565, 577, 586, 609, 628, 634, 642, 651 and 681, respectively. The contents of a, t, c, and g were 28.1%, 46.3%, 7.4%, and 18.2%, respectively, and the content of a+t was 74.4%, indicating a significant a/t bias. The conversion/transversion rate R was 0.6.
Phylogenetic analysis of M. graminicola populations based on COI genes: Phylogenetic suite software was used to perform Bayesian interference analysis and construct a phylogenetic tree of COI genes inM. graminicola populations (Fig.1). M.enterolobii strains from NCBI with the same mtCOI gene were selected as out groups for the phylogenetic tree. Fifty-four populations of M. graminicola were divided into three clades; the population from Henan province, Jiangsu and Anhui Province in Yangtze valley (YP) were grouped in Cluster 1; Cluster 2 comprises the populations from central region of China (CR) (Hunan, Sichuan, Jiangxi provinces) and eight French strains of M. graminicola selected from NCBI, Cluster 3 includes all populations in southern China (SC) (Guangdong, Guangxi, Fujian and Hainan provinces).
Nucleotide and haplotype diversity analysis of M. graminicola populations: The haplotype diversity (Hd) and nucleotide diversity (π) of mtCOI genes in M. graminicola populations (Table 2) showed that the Hd was 0.646, indicating that the total population was higher in haplotypes. The nucleotide diversity (π) and nucleotide mean difference number (k) were 0.00682 and 5.370, respectively. The Tajima’s D (-1.252) and Fu’s Fs values (-3.06764) of the total population were less than zero (0), which indicated that the entire population conformedto the law of neutrality.
Among the mtCOI gene sequences of the three groups, the highest point of variation (31) was observed in the YP group (Table 2), and the lowest mutation sites were observed in the CR group. The haplotype diversity (Hd) of the three groups ranged from 0.143 to 0.772, while the lowest (0.143) and the highest (0.804) Hd value was in CR group and YP group, respectively. The π values of the three groups ranged from 0.00018-0.01127, the smallest π value was detected in the CR groups (0.00018) and the largest π value was detected in the YP group (0.01127). The highest mean number of nucleotide differences (k) among groups was found in the YP group whilethe lowest was found in the CR group. The neutral test results for the three groups were shown to be less than zero (0) for Tajima’s D and Fu’s Fs values and were not significantly different.
Haplotype frequency analysis of M. graminicola: There are 15 different haplotypes (Hap 1-15) that have been discovered in 54M. graminicola populations (Table 3). Hap8 appeared significantly more frequent among all individuals tested, accounting for 59.3% (32/54). Hap10 was detected in four populations, accounting for 7.4% (4/54). Hap1, Hap6, Hap7, Hap12, and Hap15 accounted for 3.7% (2/54) of the populations examined. Among these 15 haplotypes identified, 8 haplotypes occurred only once and were found to be endemic in M. graminicola populations. The haplotype results (Fig. 2) of the clusters showed that only one haplotype (Hap8) was shared among the three clusters. Nine haplotypes species were in the YP group, of which were identified as endemic haplotypes and the highest haplotype frequency was identified (60%). There were five haplotypes in the CR group, with a haplotype frequency of 33.3%. Four haplotype species was found in SC group with the lowest haplotype frequency of 26.7%.
Haplotype mediation network map of M. graminicola populations: The M. graminicola populations’ mtCOI genes were then used to construct the haplotype-mediated network map (Fig.3). The haplotype-mediated network formed a circular topological distribution pattern, indicating that the M. graminicolapopulations have historically undergone expansion. Hap8 occurred most frequently and had the largest area distribution. The haplotype of the three clusters in common atHap8. Therefore, Hap8 may be the original haplotype of M. graminicola . However, the haplotypes in Henan province were grouped separately. Hap5 (HENXX) was a transitional haplotype linking the Henan haplotype to other haplotypes. Hence, this network may clarify the evolutionary relationships between each haplotype and the geographical distribution of each group, bolstering the phylogenetic tree.
Genetic differentiation and gene flow analysis of the M. graminicola populations: The genetic differentiation and gene flow of the 54 populations of M. graminicola were analyzed based onmtCOI gene sequence data (Table 4). The overall genetic differentiation coefficient Gst value, fixation coefficient Fst value, and gene flow Nm value among the populations of M. graminicolawere 0.431, 0.593, and 0.33, respectively. These suggested that there was a large genetic differentiation (Fst> 0.25) and a low gene exchange (Nm < 1) among the studied populations. The minimum Fst value observed between the population of Hainan (HAN) and Fujian (FUJ)was 0.047, while the highest gene flow was 10.125, which indicated that the gene exchange between these two populations was frequent with low genetic differences. The gene flow between the Guangdong (GUD) population and the Jiangsu (JIS) population was higher (1.75), indicating that there was some gene exchange between them.Similarly, the gene flow between populations of GUD, JIX, ANH, HAN, and JIS were all 1,suggesting that there was low gene exchange between these populations.
Genetic and geographic distance analysis of M. graminicola populations: Using the Kimura2-Parameter model, the genetic distances between different M. Graminicola populations were calculated based on mtCOI gene sequences (Table 5). Results showed that the genetic distances between different populations varied from0 to0.013. Among them, the lowest genetic distances (0.000) was observed between the populations of HAI and SIC, FUJ; GUD and JIX, ANH, JIS; JIX and ANH, JIS; ANH and JIS; SIC and FUJ, while the highest genetic distance (0.013) was observed between Henan, GUX, and Hunan.
Based on mtCOI gene, the relationship between genetic distance and geographic distance of different M. graminicola populations were examined using SPSS software (Fig.4). The findings revealed that among the collected samples, there was no significant correlation between genetic distance and the natural logarithm (LN km) matrix of geographical distance (Table 5) (R<0.2), P>0.05), suggesting that geographic distance was not the major factor contributing to M. graminicola populations differentiation.
Analysis of molecular variance (AMOVA) of the M.graminicola groups: This analysis results of AMOVA based on different M. graminicola populations were shown in Table 6. The genetic differentiation (FST) among different populations was low (FST= 0.17847 P < 0.0001), 96.3% of total variation was mainly occurred within populations whereas only 3.7% of the total variation was within groups. These findings suggested that the M. graminicola populations’ genetic variation was basically from within the population.