Investigating environmental variables that facilitate changes in flower colour morphs in the Drakensberg Rhodohypoxis baurii

Abstract

The maintenance of diversity in floral traits and flower colour polymorphism (FCP) have been of interest in evolutionary biology. As such, it is becoming increasingly necessary to examine the ecological and evolutionary explanations of FCPs in populations or species, especially since a growing body of work on flower colour continues to form the basis of important discoveries in biology. Specifically, environmental variables affect FCPs within populations or species, thereby facilitating speciation and subsequent floral evolution. The near-endemic Drakensberg plant species, Rhodohypoxis baurii (Baker) Nel. var. confecta Hilliard & Burtt is a non-model system that occurs in a vulnerable mountain grasslands habitat and comprises populations with FCPs. Investigating how intraspecific FCP (within taxon and population) relates to environmental variables and subsequent fitness across the flowering season in R. baurii var. confecta may enable us to make accurate predictions about the evolutionary trajectories of naturally polymorphic populations and provide insight into how this species may respond to environmental change. I attempted to replicate field conditions where soil moisture and solar radiation differ at three time points (early, mid- and late flowering season) at a polymorphic population of R. baurii var. confecta that occurs in Sentinel Peak car park (SCP), Free State, South Africa, in a growth chamber experiment, to investigate whether soil moisture or solar radiation, or their interactive effect underpin the shift in flower colour morphs (from a majority of white flowers to a majority of pink flowers as the flowering season progresses). My findings indicated that solar radiation, but not soil moisture best explained the shift in white and pink flowers over the flowering season. Specifically, there were more pink than white flowers when both soil moisture and solar radiation were relatively higher towards the end of the 8 flowering season than in the early and mid-time points over the flowering season, when flowers are predominantly white. In the growth chamber experiment, I used UV exposure as a proxy for increased solar radiation. Only combinations of well-watered and UV exposure had a significant overall effect on the percentage of pigmented flowers produced during the experiment. Further, I performed an amplified fragment length polymorphism (AFLP) scan in three populations of R. baurii var. confecta with different extents of FCPs (the SCP population and two monomorphic populations) at a single locality in Free State, South Africa, to compare among-population genetic differentiation. Interestingly, only one outlier locus was detected in the R. baurii var. confecta genome scan but there was little support that it was effective. The outlier locus was found in 80% and 30% of individuals from the flat catchment and Witsieshoek lodge populations (the monomorphic populations), respectively, while only 8% of individuals in the polymorphic SCP population possessed the outlier locus. However, no genetic differences among populations were significant, with little molecular variance among populations (3%). Together, my findings highlight that FCP in R. baurii var. confecta may be due to population response to environmental change, but not population genetic differences. Consequently, FCPs in this study system are likely population-specific. Nevertheless, my research contributes to the growing body of work that investigates the relationship between environmental variables and FCPs, and offers insights into the ecological and genetic relationship with regards to FCPs and how they are maintained in a mountain endemic plant.