Shiga T., 2007. A systematic study of Nuphar (Nymphaeaceae) in Japan with special reference to the role of hybridization. Thesis of doctoral degree, Kobe University.
Summary
Nuphar Sm. (Nymphaeaceae), yellow water-lily, is a genus of perennial freshwater macrophytes and is distributed in the temperate zone of the Northern Hemisphere. The genus is one of the most taxonomically problematic aquatic plants. I investigated the morphological and genetic variations of Nuphar in Japan and adjacent countories and, based on the results, I tried the taxonomical revision of Nuphar in Japan.
In Chapter 1, I gave an outline of the taxonomical, morphological, genetical
and ecological characterstics of the genus Nuphar, with special reference
to the possibility of homoploid hybrid speciation.
In Chapter 2, I investigated the morphological variations and genetic relationships of a total of 56 populations of Nuphar japonica DC., N. subintegerrima (Casp.) Makino, N. oguraensis Miki and unidentified intermediate plants in central to western Japan.
The phenogram of cluster analysis based on 15 morphological characters
revealed five cluster groups (Groups 1-5). Groups 1, 3 and 5 corresponded
well to the description of N. japonica, N. oguraensis and N. subintegerrima
sensu stricto, respectively. On the other hand, Groups 2 and 4 showed intermediate
values in most of their characters. Morphological relationships suggested
that the two intermediate groups were of hybrid origin between N. japonica
and N. oguraensis and between N. japonica and N. subintegerrima, respectively.
Allozyme analyses (analysis of allele frequencies and principal coordinate analysis based on shared allele distance among the 162 multi-locus genotypes) showed the three Nuphar species to be well distinguished. The intermediate plant groups were also shown genetically to be of hybrid origin. However, many of the intermediate plants did not show additive combinations of parental allozymes, which were expected in F1 hybrids. The absence of characteristic parental bands suggested that sexual reproduction had occurred within the hybrids. Crossability analysis supported that reproductive isolation might be weak in Nuphar species and promote introgressive hybridization.
In Chapter 3, I investigated morphological and allozyme variation and pollen
viability in 22 populations of Nuphar japonica, N. pumila (Timm) DC., and
unidentified intermediate plants in Hokkaido to assess hybridization and
introgression between the two species and the potential existence of homoploid
hybrid speciation.
A phenogram based on cluster analysis using 15 morphological characters
revealed three cluster groups. Groups 1 and 3 had distinctive morphological
characteristics and corresponded to the published descriptions of N. japonica
and N. pumila, respectively, whereas Group 2 showed intermediate values
in most characteristics. In the allozyme study, many morphologically intermediate
plants showed additive combinations of species-specific alleles from the
two species. In the DNA analysis, Pollen viability was significantly lower
in intermediate populations than in the two species. Intermediate populations
were found within the area of overlap between the ranges of the two species.
On these evidences, I concluded that the intermediate plants represent
hybrids between the two species. Relationships among species-specific AFLP
bands, cpDNA haplotypes and morphology indicated that the hybrids have
backcrossed to N. japonica repeatedly and genomes of N. pumila have been
introduced to N. japonica as nuclear genes.
Four hybrid populations showed different relationships between morphology and pollen viability. Hybrids in artificially disturbed populations had discontinuously intermediate morphology between their parental species and were completely fertile in some cases. It is probable that ecological selection has contributed to the discontinuous morphological patterns. Homoploid hybrid speciation may occurr in these Nuphar hybrids.
In Chapter 4, I investigated morphology, allozyme variation and pollen
viability in 9 populations of Nuphar japonica, N. submersa Shiga &
Kadono, and unidentified intermediate plants in central to eastern Japan.
A phenogram based on cluster analysis using 10 morphological characters
revealed three cluster groups. Groups 1 and 3 corresponded to N. japonica
and N. submersa, respectively, whereas Group 2 showed intermediate values
in most characteristics between the two species. In allozyme study, many
morphologically intermediate plants showed additive combinations of species-specific
alleles in two loci (lap1 and mdh3) from the two species. Pollen viability
was significantly lower in intermediate plants than in the two species.
Based on these evidences, I concluded that the intermediate plant was of
hybrid origin between the two species.
In Chapter 5, I tried to solve phylogenetic relationships and geographical patterns of Nuphar species, using AFLPs and allozyme markers.
In phylogenetic analysis, the neighbor-joining topology based on AFLP data supported interspecific relationships which were proposed by previous phylogenetic study, except for N. subintegerrima. Asian Nuphar plants were monophyletic group with 83% bootstrap values in the NJ tree. Within Asian Nuphar taxa, four strongly supported monophyletic groups were recognized; the first cluster consisted of N. oguraensis, N. shimadae Hayata and N. submersa; the other three clusters corresponded to N. pumila, N. japonica and N. subintegerrima, respectively. The phylogenetic tree showed both N. pumila and N. oguraensis were sister taxa but well separated phylogenetically. In N. oguraensis group, two subclusters were recognized. The first subcluster consisted of N. submersa and the second subcluster consisted of N. oguraensis and N. shimadae. Nuphar oguraensis and N. shimadae should be considered conspecific.
In the AFLP analysis, the two clades of N. japonica showed significant
correlation each other between geographic distance and genetic distance.
The geographical distribution ranges of the two AFLP groups did not overlap
and the two cpDNA haplotypes indicated almost the same distributional patterns.
Furthermore, the plants of the two groups had significantly different morphological
traits. Allopatric speciation may be in progress in N. japonica.
Nucleotide diversity, genetic differentiation and genotypic diversities within and among populations lower in the populations of all the species distributed in east to north regions than those of central to south regional ones. It is possible that the eastern to northern populations of Nuphar had experienced bottleneck effect. Present distributional patterns and geographical genetic variation of Japanese Nuphar may result from the distribution of refugia in glacial periods.
In Chapter 6, I proposed a taxonomic revision of Japanese Nuphar taxa based on the results of previous chapters and ca. 800 sheets of herbarium specimens, including some European specimens. Here I revised the Japanese Nuphar into six species and three hybrid taxa, as follows:
Nuphar japonica DC.,
N. pumila (Timm) DC.,
N. saikokuensis Shiga & Kadono, sp. nov.,
N. shimadae Hayata,
N. subintegerrima (Casp.) Makino,
N. submersa Shiga & Kadono, sp. nov.,
N. x fluminalis Shiga & Kadono, hybr. nov.,
N. x hokkaiensis Shiga & Kadono, hybr. nov.,
N. x saijoensis (Shimoda) Shiga & Kadono, comb. nov..
I reduced taxonomic rank of the taxa characterized by the red stigmatic disc to form.
Natural hybridization and introgression frequently occur in Nuphar. Aquatic species with high morphological variability may consist of a complex of homoploid hybrid species and parental species. In this study, I conclude that repeated hybridization and introgression caused high morphological variability of Japanese Nuphar and geographic isolation is an important factor for divergence and speciation in the genus.