When considering the genetic implications of immigrant gene flow, it is important to evaluate both the proportions of immigrant gametes and their genetic composition. We simultaneously investigated paternal and maternal gene flow in dispersed seeds in a natural population of Pinus densiflora located along a ridge. The paternity and maternity of a total of 454 dispersed seeds (in 2004 and 2005) were accurately and separately assigned to 454 candidate adult trees, by analyzing the nuclear DNA of both diploid biparentally derived embryos and haploid maternally derived megagametophytes of the seeds. The relative genetic diversities and differences between within-population and immigrant groups of both paternally and maternally derived gametes (4 groups) that formed the genotypes of the seeds were evaluated. Using 8 microsatellite markers, we found that 64.0-72.6% of paternally derived gametes, and 17.8-20.2% of maternally derived gametes, were from other populations. Principal coordinate analysis showed that the 4 gamete groups tended to be plotted at different locations on the scattergram, indicating that they each have different genetic compositions. Substantial paternal and maternal immigrant gene flow occurred in this population, and therefore, the overall genetic variation of dispersed seeds is enhanced by both paternally and maternally derived immigrant gametes.

Although most of the previously detected variation in flower biomass was among populations that were far apart from each other, differences in population characteristics (population size, area, and plant density) may bring the variation in flower biomass to a more local scale. To examine the variation in flower biomass among nearby populations of Impatiens textori Miq. (Balsaminaceae), field studies were conducted on six natural populations located along a stream in Japan. We also examined the dependence of flower biomass and outcrossing rate on population characteristics, as well as the differences in plant size and pollinator behavior among populations. We conducted pollination experiments with potted plants, in which plant density and flower size were independently manipulated. Mean flower mass varied among populations, being negatively dependent on plant density. One-factor ANCOVA showed that both plant size and the other population-level factor affected flower biomass variation. Experiments with potted plant arrays showed that geitonogamous pollination more likely occurred in sparse populations, but in field studies, the population outcrossing rate was not significantly dependent on plant density of the population. Thus, the variation in flower biomass cannot be fully explained by these commonly considered factors. Our results show that the flower biomass of populations may evolve locally in response to plant density or other population characteristics.Key words: variation, flower biomass, population, Impatiens textori, plant density, local scale.

No abstract

No abstract

To investigate how accurately biparentage assignments for coniferous seeds can be improved by using different kinds of seed tissues, we assigned biparentage to dispersed seeds in a natural stand of Pinus densiflora Siebold & Zucc. (288 mature trees) using two procedures: with or without megagametophyte haplotype data (“MH” and “ordinary” procedures, respectively). Using archived clones, we confirmed the conifer-specific modes of inheritance of three kinds of seed tissues from certain maternal trees using six microsatellite markers. In the natural stand, under the MH procedure, 39.2% of male parents and 77.0% of female parents for a total of 204 seeds analyzed were assigned to at least one mature tree within the study area. At that time, the proportion of exactly matching seeds out of seeds with one matching parent under the MH procedure was significantly larger than that of the seeds under the ordinary procedure. The biparentage assignments under the ordinary procedure corresponded to only 53.7% of the accurate separate assignments under the MH procedure. It is suggested that analyzing different seed tissues is effective for exact and accurate biparentage assignments in investigations of biparentally mediated gene flow in coniferous populations, particularly at the seed-dispersal stage.