This blog is part of our colourful countdown to the holiday season where we’re celebrating the diversity and beauty of the natural world. In this post, Caroline Dong of Tulane University unearths the diverse world of floral colouration and what we have yet to discover in these decorative but mysterious plants.

Floral colouration can be a useful and predictable trait. The colouration of flowers is a signal that has evolved for efficient and effective visual communication between them and their pollinators. Plants aim to produce flowers that are most conspicuous towards a particular pollinator by producing colours and patterns that are easily perceived by that pollinator’s visual system. Maintaining specificity is vital because it is advantageous to minimise visits from other groups of pollinators that do not contribute to reproduction. In addition to colour, traits such as odour, shape, size, and phenology also contribute to pollinator discrimination.

Due to its integral role in plant-pollinator dynamics, floral colour is a trait that has broad ecological contributions to reproductive fitness and the maintenance of species boundaries by regulating gene flow. Beyond the rainbow of human-visible wavelengths (400–700 nm), there are significant differences in floral colouration and patterning in the ultraviolet (UV; 300–400 nm) and near-infrared wavelengths (NIR; 700–2500 nm). These are relatively lesser-explored aspects of colouration in plants and other organisms but have the potential to influence plant-pollinator dynamics as much as human-visible colouration.

Differences in absorbance and reflectance in the UV wavelengths form central elements of many floral colour patterns. These striking UV patterns are widely established to play a role in attracting pollinators. Insects and birds, common pollinators, both have well-developed colour vision extending into UV wavelengths. Floral UV patterning is evolutionary labile and there are general patterns that emerge in pollinator groups. For example, flowers that are pollinated by bees often have UV-absorbing centres and UV-reflecting margins whereas those pollinated by birds are often entirely UV-absorbing. Additionally, variation in UV patterning has been found to be associated with bioclimatic variation in temperature, altitude, humidity, and UV irradiance.

However, environmental selection on UV patterning is not yet fully understood and contrasting patterns have been found in different systems. Further, there are changes in pollinator assemblages and their perception of colour (differing light environments) along environmental gradients (e.g. altitude) which may make it difficult to disentangle selective forces at play.

Sessile organisms, such as plants, still require adaptations to regulate their internal temperature. Colour is one method of thermoregulation, theoretically accomplished by having darker colouration and/or altering reflectance in the NIR wavelengths, which comprises approximately 50% of total incident radiation from sunlight. However, despite the intuitive hypothesis that darker colours absorb more radiation and increase temperatures, studies of flowers have not identified a broad-scale relationship between temperature and colour. It has also been shown in butterflies and beetles that visible colour and temperature are not necessarily linked. This is likely because visible colour has competing functions aside from thermoregulation (e.g. camouflage, signalling, physical and physiological protection).

On the other hand, differences in NIR are expected to fill thermoregulatory functions because animals, including insects, cannot see NIR wavelengths. In plants, it has been suggested that NIR may modulate the thermal profile of a flower to benefit plant reproduction directly (e.g. optimal temperature for metabolic and fertilisation processes) or indirectly by attracting pollinators with a thermal reward (i.e. heat gain) and thereby influencing plant-pollinator dynamics.

I am developing research projects to explore whether cryptic (beyond human-visible) differences in floral colouration and patterning may contribute to reproductive isolation and fitness within the Mimulus guttatus species complex, a group of closely related wildflowers that occur in diverse environments across Western North America. In particular, I am interested in the seep monkeyflower (M. guttatus) and cut-leaf monkeyflower (M. laciniatus), which are locally adapted to contrasting microhabitats and diverge in several phenotypes such as leaf shape, flowering time, and mating system.

Mimulus guttatus is outcrossing, primarily pollinated by bees, whereas M. laciniatus is highly self-fertilising. The visible colourations of Mimulus guttatus and M. laciniatus are similar where both have yellow flowers with variable red spots. Flowers of M. guttatus have defined regions of UV reflectance and absorption, which has been shown to be important for attracting pollinators. UV patterning is also present in M. laciniatus flowers but variation is not well-characterised and the function is unclear, given that it is self-fertilising and does not require pollinators.

There is evidence of occasional hybridisation and gene flow between Mimulus guttatus and M. laciniatus when their microhabitats are in close proximity in sympatric populations. Although divergence in visible colour patterns is minimal, theoretically selection could favour the emergence of more pronounced cryptic differences in sympatric populations to increase pollinator discrimination and reduce gene flow. This process is termed ‘reinforcement’ and may occur if postzygotic reproductive isolation (e.g. hybrid sterility or inviability) is present between the two species. Further experiments may be to assess pollinator preference and behaviour, and interpret floral patterns using models for pollinator vision, in order to form inferences on the role of colour traits on gene flow.

Additionally, UV and NIR reflectance and absorbance along elevational gradients may be examined for correlations with abiotic factors. The existence of both allopatric and sympatric populations between M. guttatus and M. laciniatus provides us with a natural comparative study in signal evolution and reproductive isolation.

Many broad themes are available to be explored by researchers regarding the ecological and evolutionary drivers of floral colouration. For example, in highly self-fertilising species, is their floral colouration an artefact of phylogenetic signal from an outcrossing ancestor or is it influenced by abiotic selection? What is the degree of phenotypic plasticity in floral colouration and is this adaptive? Is there local adaptation of NIR reflectance in populations that differ in bioclimatic factors such as elevation? How does the strength of selection on a population compare between biotic (i.e. pollinator) and abiotic (i.e. environmental) drivers?

In summary, floral colouration is a compelling and a dynamic lens through which to view the lives of plants.

Discover more posts as part of our Colour Countdown series here.