Home > Research > Understanding Biodiversity: Evolutionary and Behavioural Ecology > Bird pollination in the Cape Floristic Region
Current Research Programmes

Bird pollination in the Cape Floristic Region

Why do flowers come in so many different colours, both within and between species? This deceptively simple question is surprisingly difficult to answer. The Cape Floristic Region’s bird pollination systems provide an ideal opportunity to address it in the context of natural communities of co-flowering species. Postdoctoral Fellow Anina Coetzee is tackling this question in collaboration with Claire Spottiswoode and Colleen Seymour. She is testing whether flower colour variation within species arises from selection for convergence in flower colour with other species in the local community, to benefit from shared signalling to pollinators. Such ecological processes may now be threatened by habitat fragmentation in many parts of the Cape Floristic Region, and understanding these effects is the goal of PhD student Daniël Cloete’s research.

The genus Erica is one of the most diverse in the fynbos biome, and its many bird-pollinated species are striking for the high levels of colour polymorphism in their flowers. Some Erica species have up to five different colour morphs both within and between species. These species are pollinated predominantly by just one bird species, the Orange-breasted Sunbird Anthobaphes violacea, so the prevalence of these polymorphisms is a mystery. This project is investigating the origin and maintenance of these flower colour polymorphisms: what role does plant community context and sunbird foraging behaviour play in generating the intra-specific colour diversity of this group of plants, and how might habitat transformation influence these ecological and evolutionary processes?

Flower colour polymorphisms may evolve for two opposing reasons. Firstly, if co-existing species experience competition for the attention of pollinators, then colour divergence may promote assortative foraging by pollinators. Secondly, when a plant species receives low pollinator visitation rates, it may benefit from mimicking the signals of co-existing species in order to attract more pollinators. Evidence of such honest signal mimicry, as opposed to deceptive mimicry, is very rare in nature. A cause for concern is that habitat fragmentation may change sunbird behaviour in such a way that the evolutionary process maintaining this diversity in ericas is changed completely.

Fitz PhD student, Daniël Cloete, is working in and around the Tsitsikamma section of the Garden Route National Park to address the effect of habitat fragmentation on the fitness of bird-pollinated plants. Daniël’s research aims to investigate how bird pollination relationships are affected by habitat fragmentation, and to test whether certain thresholds of patch size and isolation exist where pollination services by birds start to break down. To do so, he is measuring pollination by sunbirds and sugarbirds of Protea and Erica species across 17 fynbos patches, natural and fragmented, in areas located in the vicinity of Nature’s Valley. This is a good area to address this question because it naturally comprises a matrix of forest, fynbos and coastal thicket, now further fragmented by agriculture, plantations, alien plant infestations, farmland and urban areas. Insights from Daniël’s research will hopefully shed light on how threats, including land-use change, alien invasive vegetation and climate change are affecting, and will further affect, ecosystem function and services in the Cape Floristic Region.

Activities in 2017

  • Anina Coetzee completed surveys of ten communities of bird-pollinated Erica species in the south-western Cape, collecting data on sunbird visitation rates, pollinator-attracting traits and reproductive morphology. Three of these communities were sampled monthly to establish the flowering phenology patterns of co-existing species.
  • Hybrid pollination experiments were conducted on three Erica species at two different sites.
  • The study has found that flower colour diversity seems to be driven by two opposing processes: less attractive Erica species mimic the colour of favoured species to increase their visitation rate, but they also compete for pollen placement sites on sunbirds which causes colour divergence.
  • Daniël Cloete completed a year of focused fieldwork, collecting data for each of the 17 fynbos patches. Bird abundance ws recorded in relation to bird-pollinated Protea flowering phenology and bird-pollinated Protea and Erica nectar traits and seed set.
  • Aerial photos taken by a drone were used to estimate the distribution and density of bird-pollinated Proteas on each of the 17 sites.


  • Anina presented the results of her study at the annual Evolution Meeting in Oregon, USA, and at the conference of the South African Association of Botanists in Pretoria.
  • Anina was awarded a highly competitive Postdoctoral Fellowship from the Claude Leon Foundation. She also helped to train an intern from the rural Eastern Cape as part of her fieldwork.

Impact of the project:

The unique sunbird-Erica mutualism will allow us to gain key insights into the mechanisms by which individual bird behaviour affects community ecology. It provides an opportunity to address knowledge gaps, particularly because human disturbance may directly interact with evolutionary processes in this system. Insights into the effects of habitat transformation on pollination systems such as this will inform the development of guidelines for maintaining biodiversity and ecosystem functioning.

Key co-sponsors

DST-NRF CoE grant; The Botanical Education Trust; Biotechnology and Biological Sciences Research Council; South African National Botanical Institute; Harry Crossley Green Matter Scholarship.

Research team

Dr Anina Coetzee (FIAO, UCT)
Prof. Claire Spottiswoode (FIAO, UCT and U. Cambridge)
Dr Phoebe Barnard (FIAO, UCT and SANBI)
Dr Mark Brown (Nature’s Valley Trust)
Prof. Peter Ryan (FIAO, UCT)
Dr Colleen Seymour (SANBI)

Student: Daniël Cloete (PhD, UCT)