The degree of genetic variation within and between populations of the freshwater angiosperm Myriophyllum alterniflorum DC. was determined using gel electrophoresis on plants grown under standard conditions. Two hundred and seventy-three individuals were examined from 13 populations, of which ten were from a limited area in the northwest of Scotland, and one each from central Scotland, the south of England and Norway. Of the 12 enzymes studied, eight were polymorphic across the 13 populations: acid phosphatase, aspartate aminotransferase, α-esterase, fluorescent esterase, malic enzyme, peroxidase, phosphoglucose isomerase and tetrazolium oxidase (equivalent to superoxide dismutase). The plants studied showed a high degree of variation. Fifty-eight multi-enzyme phenotypes, based on five enzymes, were distinguished, distributed across the 13 populations and the 263 individuals which could be analysed. Twenty (35%) of these phenotypes were represented by one individual and 36 (62%) phenotypes were restricted to one site. The two most frequent phenotypes were found in five sites and one of these was found in 21 individuals (8% of the plants studied). No population was encountered with a single multi-enzyme phenotype. Simpson's index of diversity ( D) varied between 0.536 at the least variable site and 0.978 at the most variable site, with a mean of 0.792. Even for sites within one watershed, the variation was great; thus, for tetrazolium oxidase, plants at the head, middle and foot of the watershed comprised 96% Phenotype A, 100% Phenotype C and 100% Phenotype A, respectively. A cluster analysis of the populations based on the presence or absence of single- and multi-enzyme phenotypes revealed four clusters of varying composition separated at the 53% and 17% level, respectively, but these were not correlated with presence in a watershed, geographic distance or site alkalinity. The majority of the populations appeared to be in Hardy-Weinberg equilibrium for two loci (α-esterase-2 and tetrazolium oxidase-2), where analysis was possible, which suggests that sexual reproduction is common in these populations and that it occurs via random mating. Calculations for these loci using Wright's F statistics revealed that complete panmixia was not occurring ( F IT = 0.498), largely because of genetic subdivisions between populations ( F ST = 0.469) rather than inbreeding within populations ( F IS = 0.054). The possible reason for the high genetic variation uncovered here is discussed in terms of the stability of populations, frequency of the species in nearby sites, flowering frequency, pollination mechanism and the distribution of the species in different types of water body. The mitotic chromosome number was 2 n = 14 for all the material studied.
Read full abstract