Abstract
The presence of ultraviolet (UV, wavelengths between 300-400 nm) reflectance in insect-pollinated flowers has been linked to pollination efficiency and pollination shifts, but little is known about its prevalence and function in other pollination systems and African species. We chose the genus Erica for studying the prevalence of UV because of its extreme radiation (c. 680 species) in the Cape, South Africa, with a diversity of pollination syndromes. This study quantified the prevalence and brightness of UV reflectance for five Erica pollination syndromes and tested pollinator preferences for UV reflectance in the two groups with the highest prevalence: sunbirds and long-proboscid flies. Our results show that UV colouration is absent or rare in Erica species pollinated by unclassified insects, rodents or wind. About 17 % of bird-pollinated species reflected UV but choice experiments revealed that free-ranging sunbirds showed no preference for UV signals. All sampled long-proboscid fly-pollinated species reflected UV and its experimental removal decreased seed set drastically, suggesting that long-proboscid flies in the Cape strongly prefer or depend on UV and thereby contributed to selecting for the evolution of this signal.
Highlights
Colour is one of the most important flower advertisements for pollinators
Pollinators have contributed greatly to the large diversity of flower colours in angiosperms, which evoke specific behavioural responses in different flower visitors due to the differences in their colour vision systems and neural processing (Junker et al 2013)
This study aims to (a) quantify the prevalence and brightness of UV reflectance in Erica across different pollination syndromes, and (b) test pollinator preference in the groups with the highest prevalence since this may have been the mechanism that selected for the signal
Summary
Colour is one of the most important flower advertisements for pollinators. pollinators have contributed greatly to the large diversity of flower colours in angiosperms, which evoke specific behavioural responses in different flower visitors due to the differences in their colour vision systems and neural processing (Junker et al 2013). 680 species in the Cape, South Africa) is highly suitable for studying this because of its diversity of flower colours and pollinators (Rebelo et al 1985). The species can be grouped into five pollination syndromes: insect-, bird-, long-proboscid fly(LPF), rodent- and wind-pollinated (Rebelo et al 1985; Turner et al 2011; Lombardi et al 2017). Since these species share an evolutionary history, a comparison of the prevalence of UV reflectance between pollination syndromes could indicate the role of pollinators in selecting for UV signals. Differences in colour vision and behavioural responses between the pollinators might have selected for different signals amongst the pollination syndromes and thereby contributed to reproductive isolation and species divergence (e.g. Streisfeld et al 2013)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
