Research On Morphological Variations And Genetic Structure Of Three Macromitrium Species

Macromitrium cavaleirei Cardot & Thér., M. gymnostomum Sull.& Lesq.and Macromitrium japonicum Dozy & Molk. are three widely spread species in eastern China. In order to understand their morphological and genetic variations and their relationships with geographical background, thus to know the characteristics of their intraspecific variation and diversity,we collected 192 samples of these three species from 24 geographical locations in Zhejiang, Guangxi, Fujian and Taiwan provinces, and determined their leaf and cytological traits. After our establishing the optimized systems of ISSR-PCR for plant materials of Macromitrium by using orthogonal design,we also obtained their ISSR data. Finally, we analyzed the characteristics of their morphological and genetic variations and genetic diversity, and the correlation among morphological variation, genetic variation and geographical background. The main results are as follows:(1)Thirteen morphological traits of 106 samples in eleven populations of M. gymnostomum were determined. Distinct variations existed in the leaf length, the ratios of length to width of the leaves, and lamina cells. The lower lamina cells have the highest variation coefficient. The 106 samples could be clustered into three groups based on thirteen morphological traits. Leaf apex shape, leaf length, papillosity in medial lamina cells, basal lamina cells, width and length of costae are obviously different among these three groups(P< 0.001).The 106 samples could be further divided into eight groups. Except costa length, other twelve traits vary much among these eight groups, each with P< 0.001. Principal Component Analysis on the morphological traits showed that the ratio of leaf length to leaf width has the largest effects on the differentiation of M. gymnostomum, while the papillosity on the medial cells has the least effects. In order to discuss the rationality of the intraspecific grouping, we firstly performed Detrended Correspondence Analysis(DCA) on the 106 samples based on 13 morphological traits. With their information of the first three dimensional axes, we conducted Fuzzy C-Mean Clustering Analysis on these 106 samples,revealing the rationality of the grouping. However, by using the latter method, we are able to know whether it is rational to include a focal sample into given group.(2)For M. gymnostomum, a total of thirteen ISSR primers with stable experimental results were selected. 150 bands were amplified by using these selected primers, of which 148 were polymorphic,with a polymorphic percentage of 98.67%. The Nei index was 0.3094 for all populations,the intra-population gene diversity accounting for 62.22%. Based on the information of 148 loci,the 106 samples could be clustered into eight groups;The morphological variations of M. gymnostomum were genetically controlled to an extent(r = 0.159, P<0.2). There also existed distinctive geographical influences on morphological differentiation(r=0.309, P< 0.01)and genetic differentiation(r = 0.251, P< 0.01).(3)Thirteen morphological traits of 40 samples in six populations of M. cavaleriei were determined. Leaf length/leaf width at upper part have the highest variation coefficientt(29.691%) and the leaf length have the lowest variation coefficient(12.36%). The 40 samples could be preliminarily clustered into three groups based on thirteen morphological traits.For M.cavaleriei,a total of thirteen ISSR primers with stable experimental results were selected. 135 bands were amplified by using these selected primers, of which 128 were polymorphic, witha polymorphic percentage of 94.81%. The Nei index was 0.3186 for all populations, the intra-population gene diversity accounting for 65.44%. Based on the information of 128 loci,the 40 samples could be clustered into three groups. There existed slight genetic influences on morphological variations(r = 0.14, n = 40, P> 0.2),while the genetic differentiation(r = 0.213, n=40, P < 0.1)and morphological variations(r = 0.224, n=40, P < 0.1)were influenced by geographical locations to an extent.(4)Seven morphological traits of 46 samples in eleven populations of M. japonicum were determined. Leaf length/leaf width at upper part have the highest variation coefficientt(20.68%), while the leaf length have the lowest variation coefficient(11.25%). The 40 samples could be clustered into four groups based on their seven morphological traits. For M.japonicum, a total of thirteen ISSR primers with stable experimental results were selected. 192 bands were amplified by using these selected primers, of which 188 were polymorphic,with a polymorphic percentage of 97.92%. The Nei gene diversity index(H)was 0.2467 for all populations,the intra-population gene diversity accounting for 61.25% of the total diversity. The 40 samples could also be clustered into four groups. There existed slight relationship between morphological variations and genetic differentiation(r = 0.204, n=46, P < 0.2), between morphological varations and geographical locations(r = 0.209, n=46, P < 0.2). However, the genetic differentiation of M. japonicum populations was not influenced by their geographical locations(r =-0.025, n=46, P > 0.5).(5)Based on 13 morphological traits and the information of 249 loci, 192 were analyzed by using the common clustering, Fuzzy C-Mean Clustering based on DCA, to produce clustering dendrogram and two-dimensional ordination plot, respectively. The results agree well each other. In the clustering dendrogram based on thirteen morphological traits, only six samples were clustered into wrong groups in relation to M. cavaleriei and M. gymnostomum, and Macromitrium cavaleriei is closer to M. gymnostomum than M. japonicum in morphological clustering dendrogram, however, in the genetic dendrogram, Macromitrium cavaleriei is closer to M. japonicum than M. gymnostomum, which agree with the traditional classification that M. cavaleriei and M. japonicum belong to Sect. Goniostoma, while M. gymnostomum belong to Sect. Leiostoma.(6)In the present paper, based on the morphological and ISSR data, Fuzzy C-Mean Clustering Analysis were performed on 192 samples to get Silhouette width values for each groups, and member values for each samples. By comparison of these Silhouette width values and member values, we are able to know the rationality of the grouping, which also could be intuitively revealed on two-dimensional ordination by using Principal Axes Ordination.