The evolution and persistence of polyploidy in Oxalis obliquifolia Steud. ex A.Rich. populations in Gauteng

Abstract

Background: Polyploidy is a major factor in the adaptation and speciation of many plant lineages. Many evolutionary factors may contribute to autopolyploid frequency within plant populations, including rates of new polyploid formation, the ability of new polyploids to establish successfully, long-term persistence of polyploids in the environment, and the ability of new polyploids to expand their range. Despite recent progress, there are still many questions regarding polyploid success, in spite of the challenges posed by minority cytotype exclusion, and relatively little is known about ploidy variation in the South African Flora. Aims and objectives: Oxalis obliquifolia Steud. ex A.Rich. is notable for its large distribution range (from the Cape to Ethiopia) and high degree of morphological variability. The aims of this investigation were to document the occurrence of different cytotypes of O. obliquifolia across Gauteng Province, South Africa, and assess the impact of empirical data on theories that attempt to explain polyploid persistence in populations. The objectives of the study were: firstly, to sample individuals of O. obliquifolia across Gauteng and assess their cytotype using flow cytometry and chromosome squashes; secondly, to determine if there were differences in abiotic niches occupied by different cytotypes; thirdly, to determine the morphological and phenological traits associated with polyploidy; fourthly, to assess the degree of reproductive isolation between different cytotypes; and finally, to assess the degree of relatedness between individuals of different ploidy-levels across mixed-ploidy sites. Methods: Over 320 samples from 25 sites were collected and cytotyped, using standard flow cytometric and ploidy confirmed using meiotic chromosome squashes. Individuals were mapped and abiotic variables assessed for correlations with cytotype distribution using GIS, climate data, field observations, soil data, and ordinations and PerMANOVAs. Different cytotypes (100 individuals, including diploids, tetraploids and hexaploids) were grown under identical conditions to assess the associations between polyploidy and morphology and phenology, and results were analysed using linear models and discriminant analyses. Reproductive isolation and frequency of polyploidisation were assessed using crossing experiments (1140 crosses, with different maternal cytotypes), as well as AMOVA analyses based on Internal Transcribed Spacer DNA sequences. Results: Remarkably, six distinct cytotypes were identified, with over 50% of sites comprising multiple ploidies. Abiotic variables were not associated with cytotype distribution possibly due to scale. The common garden experiment demonstrated a clear Gigas effect, which may confer a competitive advantage for polyploids over their smaller diploid progenitors. Larger flowers and differences in flowering phenology suggest pollinator interactions may play a role in enabling polyploid persistence. Crosses between cytotypes are possible under artificial settings, however DNA analysis suggests diploids and polyploids are reproductively isolated in the wild, and that polyploidisation is not a frequent enough event to explain the high levels of cytotype sympatry observed. Diploids and polyploids are behaving as separate species, despite high sympatry and non-zero potential inter-cytotype seed set. Tests on biotic interactions may provide insights into how polyploids have flourished in this system.