Aquatic plants are a polyphyletic plant group that depend on the aquatic environment for part of their life history. Each of these plants is considered to have adapted and speciated into a waterside environment; however, evolutionary and taxonomic research has been stagnant because of the difficulty in investigation and observation compared with that in general terrestrial plants. Lemna L. (Araceae, Lemnoideae) is a good model plant for aquatic botanical research. It is the largest genus in the subfamily Lemnoideae and is a tiny floating aquatic plant that is widely distributed throughout the world, except in the polar areas. Because Lemna species have simple organs and their flowers are rare, identification is difficult, and taxonomic challenges remain. Therefore, in the present study, the genetic diversity, morphology, floral morphology, interspecific hybridization, and distribution of Lemna and their relationship with evolution and speciation are discussed in five chapters. In Chapter 1, I summarize the taxonomic challenges, relationships between evolution and species differentiation, genetic diversity, morphological characteristics, flower morphology, and interspecies hybridization of Lemna. In Chapter 2, I examined whether the physiological and morphological characteristics were phenotypic plasticity or a reflection of species differences in Le. aequinoctialis Welw. and Le. aoukikusa Beppu et Murata. Physiological analysis of 11 strains revealed two types of flowers: protogynous and homogamous. Physiological group A, which was protogynous, was self-incompatible, and mature seeds could not be obtained. In contrast, mature seeds were obtained from group B, which was homogamous. These two groups were morphologically different, consistent with the characteristics of Le. aequinoctialis and Le. aoukikusa, respectively. Chloroplast DNA-based phylogenetic analyses also supported the division of these two groups, and it was concluded that Le. aequinoctialis and Le. aoukikusa are independent taxa. In Chapter 3, I estimated the clonality of the Japanese Le. trisulca L. based on genome-wide Single-nucleotide polymorphisms (SNPs) data, clarified the population structure, and evaluated the relationship between morphology (root presence and number of frond veins) and genetic features. Based on the genetic distance calculated for three relationships, which are "within individual", "among individuals in the same population", and "between population", there was no significant difference in genetic distance between "within individual" and "among individuals in the same population" relationships. This suggests that each population is usually maintained by the clonal reproduction of a single plant, that each population does not share the same individual, and that migration among populations is likely caused by seeds. This seems to be because submerged plants have a low tolerance to drying conditions, making it difficult for leaflets to disperse long distances. No relationship was detected between the morphological diversity and genetic characteristics in the presence frond veins or their number. In Chapter 4, to clarify the relationship between interspecific hybridization, differentiation of morphological traits, and speciation in Lemna, I investigated physiological and morphological variation, pollen germination rate, and phylogenetic relationships for Japanese Le. minor L., Le. japonica Landolt, and Le. turionifera Landolt in section Lemna. Chloroplast DNA analysis yielded three haplotypes and clustering analysis based on genome-wide SNP data revealed three genetic clusters in the three Japanese species. As flower characteristics of Le. japonica have not been sufficiently reported, I could not identify the actual Le. japonica. Based on Le. japonica type material (7182) observation and comparison, cluster B (haplotype J), which has a light reddish spathe and never an open anther, was classified as Le. japonica. Subcluster a (haplotype M), with a transparent spathe, was identified as Le. minor. Cluster C (haplotype T), with strong reddish spathe and dehiscent anthers, was identified as Le. turionifera. There was only one substitution between haplotypes J and M; however, there were 16 and 15 substitutions within haplotypes J and T, M and T, respectively. The three species were also morphologically distinguished in Principal Component Analysis (PCA), and the variances in frond number, frond thickness, ovary diameter, and stigma width were high. In the estimation of population demographic history, the scenario in which Le. minor and Le. turionifera hybridized approximately 6,850 (2,100-13,100) generations ago was most highly supported, and the possibility that Le. japonica originating from hybridization, was supported, as previously reported. In the pollen germination rate test, putative parental species Le. turionifera had a significantly higher germination rate than Le. minor, whereas pollen from Le. japonica seeds did not germinate. These results suggested that the combination of the putative parental species Le. minor as the female seed parent and Le. turionifera as a male pollen donor, as deduced from chloroplast DNA analysis. Additionally, the distribution of Le. minor and Le. japonica was divided into Northern and Southern Japan, which may reflect the seed-formation ability of the two species. In Chapter 5, I report two new distributions of the Lemnoideae plants (Landoltia punctata and Le. turionifera) and their habitat information to date based on field observations.
新大院博(理)第485号