The influence of pollinators and herbivores on daisy flower colour community assembly patterns, plant species distributions and flower colour evolution in Namaqualand, South Africa

[摘要] ENGLISH ABSTRACT: Plant species with flowers that are not phenotypically specialized are usually considered tointeract randomly with the available pollinator species. Because such plants are consideredgeneralist in their pollination interactions, they have received much less attention than thosewith complex specialised morphologies. However, plants with conserved generalist floralmorphologies may still strongly filter the community of flower visitors through traits such ascomplex flower colour signals.In my dissertation, I explored how various pollinators and herbivores interact with flowercolour, and what the outcomes of these interactions are in terms of community assembly,plant species distributions, and floral trait evolution. Particularly, I focus on the diversecommunities of daisies (Asteraceae) that annually flower en-masse in Namaqualand, SouthAfrica. Because these daisies are highly reliant on pollination for persistence, we mightexpect selection for increased pollination efficiency and decreased floral herbivory.In chapter 2, I assessed how flower colour patterns are assembled into communities. I foundthat daisy species within communities tend to share the same bulls-eye colour pattern, whichsuggests that evolutionary convergence, ecological filtering or facilitation is important instructuring communities. Further, I used bipartite interaction networks to confirm the role ofpollinators in community assembly processes, and I showed that communities are dominatedby a single pollinator species that interacted most strongly with the overrepresented flowercolour pattern.In chapter 3, I built on the findings from chapter 2, and asked whether pollinators havedivergent colour preferences. I identified two dominant fly pollinator species that frequently visited different dominant flower colours. Colour preference experiments showed thatMegapalpus capensis had a strong preference for orange flowers, whereas Corsomyzanigripes had a strong preference for white flowers. Next, I quantified the densities of thesetwo pollinator species independent from focal plant species in 100 communities, and Ishowed that pollinator densities are associated with the dominant flower colour incommunities.In chapter 4, I investigated the influence of herbivores on flower colour evolution. Manydaisy capitula are only open for a few hours a day, and I modelled the reflectance spectra inmultiple herbivore visual spaces of the petal surfaces that are exposed when daisies areclosed. I showed that closing daisies tend to have cryptic lower petal surfaces, and I confirmmy findings by conducting a series of experiments with herbivores.In chapter 5, I assessed whether selection by various agents has resulted in the evolution ofnarrow pollination niche breadths. I calculated pollination niche breadths for all the daisyspecies in the networks from chapter 2, and showed that most daisy species interact withfewer pollinator species than expected under random visitation. Further, ecologicalspecialization was associated with various visual signalling and reward-related traits. Iconducted transplant experiments across a pollinator diversity gradient whilst altering plantcompetition, and showed that one Gorteria diffusa morphotype had a narrow fundamentalniche that was not influenced by community context.My findings challenge the prevailing perception of the generalised pollination systems ofdaisies, and I highlight the importance of flower colour in pollinator foraging choices andpollination niche breadth evolution.

[发布机构] Stellenbosch University