Colour constancy in insects

Colour constancy is the perceptual phenomenon that the colour of an object appears largely unchanged, even if the spectral composition of the illuminating light changes. Colour constancy has been found in all insect species so far tested. Especially the pollinating insects offer a remarkable opportunity to study the ecological significance of colour constancy since they spend much of their adult lives identifying and choosing between colour targets (flowers) under continuously changing ambient lighting conditions. In bees, whose colour vision is best studied among the insects, the compensation provided by colour constancy is only partial and its efficiency depends on the area of colour space. There is no evidence for complete ‘discounting’ of the illuminant in bees, and the spectral composition of the light can itself be used as adaptive information. In patchy illumination, bees adjust their spatial foraging to minimise transitions between variously illuminated zones. Modelling allows the quantification of the adaptive benefits of various colour constancy mechanisms in the economy of nature. We also discuss the neural mechanisms and cognitive operations that might underpin colour constancy in insects.     

Pollination of Schisandra henryi (Schisandraceae) by female, pollen-eating Megommata species (Cecidomyiidae, Diptera) in south-central China

BACKGROUND AND AIMS: The mutualistic interaction between insects and flowers is considered to be a major factor in the early evolution of flowering plants. The Schisandraceae were, until now, the only family in the ANITA group lacking information on pollination biology in natural ecosystems. Thus, the objective of this research was to document the pollination biology and breeding system of Schisandra henryi. METHODS: Field observations were conducted in three populations of S. henryi and the floral phenology, floral characters and insect activities were recorded. Floral fragrances were sampled in the field and analysed using TCT-GC-MS. Floral thermogenesis was measured with a TR-71U Thermo Recorder. Pollen loads and location of pollen grains on insect bodies (including the gut) were checked with a scanning electron microscope and under a light microscope. KEY RESULTS: Schisandra henryi is strictly dioecious. Male flowers are similar to female flowers in colour, shape, and size, but more abundant than female flowers. The distance between tepals and the androecium or gynoecium is narrow. Neither male nor female flowers are fragrant or thermogenic. Schisandra henryi is pollinated only by adult female Megommata sp. (Cecidomyiidae, Diptera) that eat the pollen grains as extra nutrition for ovary maturation and ovipositing. Both male and female flowers attract the pollinators using similar visual cues and thus the female flowers use deceit as they offer no food. CONCLUSIONS: Schisandra henryi exhibits a specialized pollination system, which differs from the generalized pollination system documented in other ANITA members. Pollen is the sole food resource for Megommata sp. and the female flowers of S. henryi attract pollinators by deceit. This is the first report of predacious gall midges utilizing pollen grains as a food source. The lack of floral thermogenesis and floral odours further enforces the visual cues by reducing attractants for other potential pollinators.     

Bee boys and fly girls: Do pollinators prefer male or female umbels in protandrous parsnip,Trachymene incisa (Apiaceae)?

Abstract Complete dichogamy occurs when temporal separation prevents any overlap in male and female function within and among flowers of one or more inflorescences. Although dichogamy may increase outcrossing and prevent inbreeding, it also results in the presentation of inflorescences with different floral resources. Pollinators may prefer one gender over the other based on these differences, which can reduce the transfer of pollen to conspecific stigmas and reduce floral resources for effective pollinators. We investigated whether the insect visitors of Trachymene incisa (Apiaceae), an Australian herb demonstrating complete protandry at the umbel level, show a preference for male or female umbels. The male phase umbels present pollen and nectar, whereas the female phase umbels offer nectar only. Therefore, we expect pollen‐collecting insects to favour male umbels, whereas insects that forage only for nectar will favour female umbels. In natural patches that exhibited a male umbel bias, insects showed a preference for male phase umbels at Agnes Banks in 2003 and at Myall Lakes in both 2003 and 2004. By contrast, insects showed no preference for umbel phases, visiting umbels at a similar frequency to which they occurred, at Agnes Banks in 2004 and at Tomago in both 2003 and 2004. This suggests spatial and temporal variation in insect preferences for umbel phases that differ in floral rewards. In experimental arrays where the umbel gender ratio was equal, there were no significant differences between male and female umbels in terms of insect visitation during a foraging trip and mean foraging time per visit. The differing patterns of preference may be due to a differential response by insects when the umbel ratios vary, where a male bias in umbel genders leads to a preference for male umbels, whereas an equal umbel gender ratio leads to equal visitation to male and female phase umbels.     

Foraging strategies of insects for gathering nectar and pollen, and implications for plant ecology and evolution

The majority of species of flowering plants rely on pollination by insects, so that their reproductive success and in part their population structure are determined by insect behaviour. The foraging behaviour of insect pollinators is flexible and complex, because efficient collection of nectar or pollen is no simple matter. Each flower provides a variable but generally small reward that is often hidden, flowers are patchily distributed in time and space, and are erratically depleted of rewards by other foragers. Insects that specialise in visiting flowers have evolved an array of foraging strategies that act to improve their efficiency, which in turn determine the reproductive success of the plants that they visit. This review attempts a synthesis of the recent literature on selectivity in pollinator foraging behaviour, in terms of the species, patch and individual flowers that they choose to visit.

The variable nature of floral resources necessitate foraging behaviour based upon flexible learning, so that foragers can respond to the pattern of rewards that they encounter. Fidelity to particular species allows foragers to learn appropriate handling skills and so reduce handling times, but may also be favoured by use of a search image to detect flowers. The rewards received are also used to determine the spatial patterns of searches; distance and direction of flights are adjusted so that foragers tend to remain within rewarding patches and depart swiftly from unrewarding ones. The distribution of foragers among patchy resources generally conforms to the expectations of two simple optimal foraging models, the ideal free distribution and the marginal value theorem.

Insects are able to learn to discriminate among flowers of their preferred species on the basis of subtle differences in floral morphology. They may discriminate upon the basis of flower size, age, sex or symmetry and so choose the more rewarding flowers. Some insects are also able to distinguish and reject depleted flowers on the basis of ephemeral odours left by previous visitors. These odours have recently been implicated as a mechanism involved in interspecific interactions between foragers.

From the point of view of a plant reliant upon insect pollination, the behaviour of its pollinators (and hence its reproductive success) is likely to vary according to the rewards offered, the size and complexity of floral displays used to advertise their location, the distribution of conspecific and of rewards offered by other plant species, and the abundance and behaviour of other flower visitors.     

Chromatic signals control proboscis movements during hovering flight in the hummingbird hawkmoth Macroglossum stellatarum

Most visual systems are more sensitive to luminance than to colour signals. Animals resolve finer spatial detail and temporal changes through achromatic signals than through chromatic ones. Probably, this explains that detection of small, distant, or moving objects is typically mediated through achromatic signals. Macroglossum stellatarum are fast flying nectarivorous hawkmoths that inspect flowers with their long proboscis while hovering. They can visually control this behaviour using floral markings known as nectar guides. Here, we investigate whether this is mediated by chromatic or achromatic cues. We evaluated proboscis placement, foraging efficiency, and inspection learning of naïve moths foraging on flower models with coloured markings that offered either chromatic, achromatic or both contrasts. Hummingbird hawkmoths could use either achromatic or chromatic signals to inspect models while hovering. We identified three, apparently independent, components controlling proboscis placement: After initial contact, 1) moths directed their probing towards the yellow colour irrespectively of luminance signals, suggesting a dominant role of chromatic signals; and 2) moths tended to probe mainly on the brighter areas of models that offered only achromatic signals. 3) During the establishment of the first contact, naïve moths showed a tendency to direct their proboscis towards the small floral marks independent of their colour or luminance. Moths learned to find nectar faster, but their foraging efficiency depended on the flower model they foraged on. Our results imply that M. stellatarum can perceive small patterns through colour vision. We discuss how the different informational contents of chromatic and luminance signals can be significant for the control of flower inspection, and visually guided behaviours in general.     

Getting to the start line: how bumblebees and honeybees are visually guided towards their first floral contact

Much of the literature on foraging behaviour in bees focuses on what they learn after they have had rewarded experience with flowers. This review focuses on how honeybees and bumblebees are drawn to candidate food sources in the first place: the foundation on which learning is built. Prior to rewarded foraging experience, flower-naïve bumblebees and honeybees rely heavily on visual cues to discover their first flower. This review lists methodological issues that surround the study of flower-naïve behaviour and describes technological advances. The role of distinct visual properties of flowers in attracting bees is considered: colour, floral size, patterning and social cues. The research reviewed is multi-disciplinary and takes the perspectives of both the bees and the plants they visit. Several avenues for future research are proposed.     

A bee’s eye view of remarkable floral colour patterns in the south-west Australian biodiversity hotspot revealed by false colour photography

Background and AimsColour pattern is a key cue of bee attraction selectively driving the appeal of pollinators. It comprises the main colour of the flower with extra fine patterns, indicating a reward focal point such as nectar, nectaries, pollen, stamens and floral guides. Such advertising of floral traits guides visitation by the insects, ensuring precision in pollen gathering and deposition. The study, focused in the Southwest Australian Floristic Region, aimed to spot bee colour patterns that are usual and unusual, missing, accomplished by mimicry of pollen and anthers, and overlapping between mimic-model species in floral mimicry cases.MethodsFloral colour patterns were examined by false colour photography in 55 flower species of multiple highly diverse natural plant communities in south-west Australia. False colour photography is a method to transform a UV photograph and a colour photograph into a false colour photograph based on the trichromatic vision of bees. This method is particularly effective for rapid screening of large numbers of flowers for the presence of fine-scale bee-sensitive structures and surface roughness that are not detectable using standard spectrophotometry.Key ResultsBee- and bird-pollinated flowers showed the expected but also some remarkable and unusual previously undetected floral colour pattern syndromes. Typical colour patterns include cases of pollen and flower mimicry and UV-absorbing targets. Among the atypical floral colour patterns are unusual white and UV-reflecting flowers of bee-pollinated plants, bicoloured floral guides, consistently occurring in Fabaceae spp., and flowers displaying a selective attractiveness to birds only. In the orchid genera (Diuris and Thelymitra) that employ floral mimicry of model species, we revealed a surprising mimicry phenomenon of anthers mimicked in turn by model species.ConclusionThe study demonstrates the applicability of ‘bee view’ colour imaging for deciphering pollinator cues in a biodiverse flora with potential to be applied to other eco regions. The technique provides an exciting opportunity for indexing floral traits on a biome scale to establish pollination drivers of ecological and evolutionary relevance.     

Use of habitat odour by host‐seeking insects

Locating suitable feeding or oviposition sites is essential for insect survival. Understanding how insects achieve this is crucial, not only for understanding the ecology and evolution of insect–host interactions, but also for the development of sustainable pest‐control strategies that exploit insects' host‐seeking behaviours. Volatile chemical cues are used by foraging insects to locate and recognise potential hosts but in nature these resources usually are patchily distributed, making chance encounters with host odour plumes rare over distances greater than tens of metres. The majority of studies on insect host‐seeking have focussed on short‐range orientation to easily detectable cues and it is only recently that we have begun to understand how insects overcome this challenge. Recent advances show that insects from a wide range of feeding guilds make use of ‘habitat cues’, volatile chemical cues released over a relatively large area that indicate a locale where more specific host cues are most likely to be found. Habitat cues differ from host cues in that they tend to be released in larger quantities, are more easily detectable over longer distances, and may lack specificity, yet provide an effective way for insects to maximise their chances of subsequently encountering specific host cues. This review brings together recent advances in this area, discussing key examples and similarities in strategies used by haematophagous insects, soil‐dwelling insects and insects that forage around plants. We also propose and provide evidence for a new theory that general and non‐host plant volatiles can be used by foraging herbivores to locate patches of vegetation at a distance in the absence of more specific host cues, explaining some of the many discrepancies between laboratory and field trials that attempt to make use of plant‐derived repellents for controlling insect pests.     

Visual cues and foraging choices: bee visits to floral colour phases in Alkanna orientalis (Boraginaceae)

When a pollination vector is required, any mechanism that contributes to floral visitation will potentially benefit the reproductive fitness of a plant. We studied the effect of floral colour change in the desert perennial Alkanna orientalis on the foraging behaviour of the solitary bee Anthophora pauperata . Flowers changed colour over time from bright yellow (with moderate nectar reward) to pale yellow/white (with significantly lower nectar reward). Bee visitation was non-random with respect to colour phase availability within the flower population and was biased towards the more rewarding flowers. At plants where the availability of colour phases had been manipulated experimentally to produce 'bright' or 'pale' plants, bees visited significantly more flowers (and for longer periods) on the bright plants. The change of flower colour was not simply age-related; we observed variation in the temporal course of colour change and our data suggest that visitation, leading to deposition of cross-pollen, can accelerate the process. In subpopulations with limited pollinators, Alkanna can influence bees by using their colour-related foraging preferences to alter visitation patterns.  © 2006 The Linnean Society of London, Biological Journal of the Linnean Society , 2006, 87 , 427–435.     

'Floral' scent production by Puccinia rust fungi that mimic flowers

Crucifers (Brassicaceae) in 11 genera are often infected by rust fungi in the Puccinia monoica complex. Infection causes a 'pseudoflower' to form that is important for attracting insect visitors that sexually outcross the fungus. 'Pollinator' attraction is accomplished through visual floral mimicry, the presence of a nectar reward and floral fragrances. Here we used gas chromatography and mass spectrometry to identify and quantify fragrance production by these rust fungi on several Arabis hosts, and by co-occurring true flowers that share insect visitors. Fungal pseudoflowers produced distinctive floral fragrances composed primarily of aromatic alcohols, aldehydes and esters. Pseudoflower fragrances were chemically similar to noctuid-moth-pollinated flowers, such as Cestrum nocturnum and Abelia grandiflora , but were very different from host flowers, host vegetation and the flowers of coblooming, nonhost angiosperms. There was variation in the quantity and composition of fragrance profiles from different fungal species as well as within and among hosts. The evolution of scent chemistry is relatively conservative in these fungi and can be most parsimoniously explained in three steps by combining chemical data with a previously determined rDNA ITS sequence-based phylogeny. Pseudoflower scent does not appear to represent a simple modification of host floral or vegetative emissions, nor does it mimic the scent of coblooming flowers. Instead, we suspect that the unique fragrances, beyond their function as pollinator attractants, may be important in reducing gamete loss by reinforcing constancy among foraging insects.     

Colour association influences honey bee choice between sucrose concentrations

Certain colours associated with floral food resources are more quickly learned by honey bees (Apis mellifera) than are other colours. But the impact of colour, and other floral cues, on bee choice behaviour has not yet been determined. In these experiments, colour association and sugar concentration of reward were varied to assess how they interact to affect bee choice behaviour. Thirty-five bees were individually given binary choices between blue and yellow artificial flowers that contained either the same rewards or rewards of different sucrose concentrations. Honey bee choice between sucrose concentrations was affected by colour association and this effect was greatest when absolute difference between rewards was the lowest. The honey bee's ability to maximize energetic profitability during foraging is constrained by floral cue effectiveness.     

Effects of recent experience on foraging decisions by bumble bees

The temporal and spatial scales employed by foraging bees in sampling their environment and making foraging decisions should depend both on the limits of bumble bee memory and on the spatial and temporal pattern of rewards in the habitat. We analyzed data from previous experiments to determine how recent foraging experience by bumble bees affects their flight distances to subsequent flowers. A single visit to a flower as sufficient to affect the flight distance to the next flower. However, longer sequences of two or three visits had an additional effect on the subsequent flight distance of individual foragers. This suggests that bumble bees can integrate information from at least three flowers for making a subsequent foraging decision. The existence of memory for floral characteristics at least at this scale may have significance for floral selection in natural environments.     

The Complexity of Background Clutter Affects Nectar Bat Use of Flower Odor and Shape Cues

Given their small size and high metabolism, nectar bats need to be able to quickly locate flowers during foraging bouts. Chiropterophilous plants depend on these bats for their reproduction, thus they also benefit if their flowers can be easily located, and we would expect that floral traits such as odor and shape have evolved to maximize detection by bats. However, relatively little is known about the importance of different floral cues during foraging bouts. In the present study, we undertook a set of flight cage experiments with two species of nectar bats (Anoura caudifer and A. geoffroyi) and artificial flowers to compare the importance of shape and scent cues in locating flowers. In a training phase, a bat was presented an artificial flower with a given shape and scent, whose position was constantly shifted to prevent reliance on spatial memory. In the experimental phase, two flowers were presented, one with the training-flower scent and one with the training-flower shape. For each experimental repetition, we recorded which flower was located first, and then shifted flower positions. Additionally, experiments were repeated in a simple environment, without background clutter, or a complex environment, with a background of leaves and branches. Results demonstrate that bats visit either flower indiscriminately with simple backgrounds, with no significant difference in terms of whether they visit the training-flower odor or training-flower shape first. However, in a complex background olfaction was the most important cue; scented flowers were consistently located first. This suggests that for well-exposed flowers, without obstruction from clutter, vision and/or echolocation are sufficient in locating them. In more complex backgrounds, nectar bats depend more heavily on olfaction during foraging bouts.     

A generalised mimicry system involving angiosperm flower colour, pollen and bumblebees’ innate colour preferences

Flower colour is a major advertisement signal of zoophilous plants for pollinators. Bees, the main pollinators, exhibit innate colour preferences, which have often been attributed to only one single floral colour, though most flowers display a pattern of two or several colours. The existing studies of floral colour patterns are mostly qualitative studies. Using a model of bee colour vision we quantitatively investigate two questions: whether or not component colours of floral colour patterns may mimic pollen signals, and whether or not bumblebees exhibit innate preferences for distinct parameters of naturally existing floral colour patterns. We analysed the spectral reflectances of 162 plant species with multicoloured flowers and inflorescences, distiniguishing between inner and outer colours of floral colour patterns irrespective of the particular structures so coloured.We found that:– The inner colour of radially symmetrical flowers and inflorescences and of zygomorphic flowers appears less diverse to bees than the peripheral colour.– The inner colour of most radial flowers and inflorescences as well as the inner colour of a large number of non-related zygomorphic flowers appears to bees to be very similar to that of pollen.– Bumblebees (Bombus terrestris) exhibit innate preferences for two-coloured over single-coloured dummy flowers in a spontaneous choice test.– Bumblebees exhibit innate preferences for dummy flowers with a large over those with a small centre area.– Bumblebees exhibit innate preferences for dummy flowers with a centre colour similar to that of pollen over those with another centre colour.Our findings support the hypotheses that the inner component of floral colour patterns could be interpreted as a generalised and little recognised form of mimicry of the colour of visually displayed pollen, that bumblebees exhibit innate preferences regarding colour and size parameters of floral colour patterns, and that these correspond to visually displayed pollen. These findings together suggest a prominent role of floral colour patterns in advertisement to and guidance of naive flower visitors.     

The exploitation of floral nectar in Eucalyptus incrassata by honeyeaters and honeybees

Summary During October and November, 1977, a study of nectar production and nectarivore foraging in Eucalyptus incrassata was conducted at Wyperfeld National Park in south-eastern Australia in order to evaluate the extent to which introduced honeybees (Apis mellifera) compete with native honeyeaters for floral nectar. Data on nectar production, nectar availability, ambient air temperature and the numbers of visiting honeyeaters and honeybees were collected. Most of the daily nectar production in E. incrassata occurs early in the morning when temperatures are too low for insects to forage. In addition, insects, particularly honeybees, are unable to exploit nectar in the youngest flowers because the stamens are clustered tightly around the style. As a result of these temporal and structural characteristics of the flowers, honeyeaters are able to harvest most of the nectar. Honeybees potentially have access to 35–47% of the average daily production of floral nectar in E. incrassata and actually harvest considerably less. These data show that E. incrassata flowers are adapted to restrict insect foragers despite their superficially unspecialized appearance. Eight forest and woodland eucalypts do not have a flower stage which excludes insects and the significance of this difference is discussed.     

论昆虫与植物的相互作用和进化的关系

昆虫与植物是陆地生物群落中最为重要的组成部分,二者间的相互作用是多方面的,其中最为重要的是昆虫选择植物作为食物和生长场所、昆虫为植物传授花粉两方面。该文集中讨论这两方面的相互作用有哪些因素与进化有密切的关系。植食性昆虫根据其寄主植物范围,通常分为专食性（寄主范围窄）和广食性（寄主范围广）。从生态关系来看,广食性的取食行为比专食性的更为有利,但实际情况却与此相反,统观植食性昆虫的取食行为,有向专食性演化更为普遍的倾向。专食性发展有利于提高昆虫对寄主植物的选择效率,还可缓和天敌作用所造成的压力。根据昆虫与植物相互作用的特点,目前已提出很多昆虫与植物的进化理论,包括成对的协同进化、弥散的协同进化、群落的协同进化以及顺序进化。在昆虫对寄主植物的选择中,以植物对昆虫的影响较昆虫对植物的影响更为重要,称为顺序进化是适宜的;昆虫为被子植物传授花粉造成互惠共生,其中的进化关系应称为协同进化。     

Floral sesquiterpenes and their synthesis in dioecious kiwifruit

Kiwifruit species are vigorously growing dioecious vines that rely on bees and other insects for pollen transfer between spatially separated male and female individuals. Floral volatile terpene cues for insect pollinator attraction were characterized from flowers of the most widely grown and economically important kiwifruit cultivar Actinidia deliciosa ‘Hayward’ and its male pollinator ‘Chieftain’. The sesquiterpenes α-farnesene and germacrene D dominated in all floral tissues and the emission of these compounds was detected throughout the day, with lower levels at night. Two terpene synthase (TPS) genes were isolated from A. deliciosa petals that produced (+)-germacrene D and (E,E)-α-farnesene respectively. Both TPS genes were expressed in the same tissues and at the same times as their corresponding floral volatiles. Here we discuss these results with respect to plant and insect ecology and the evolution and structure of sesquiterpene synthases.Key words: terpene, dioecy, kiwifruit, volatile, ecology, evolution, flower     

Divergent responses of two cereal aphids to previous infestation of their host plant

The role of pollen odour in resource location by the pollen beetle, Meligethes aeneus (Fabricius) (Coleoptera: Nitidulidae), a pollen-feeding insect regarded as a pest of oilseed rape, Brassica napus L., (Brassicaceae) crops, was investigated in a linear track olfactometer. Both male and female beetles were attracted to the odour of whole oilseed rape flowers, indicating that these insects can locate their host plants using floral odours as cues. The attractive odour of flowers was found to emanate from all floral parts tested: the petals/sepals, the anthers, and from pollen itself. Therefore, at least part of the attractive odour of oilseed rape flowers emanates from pollen. Beetles were more attracted to floral samples containing anthers than those without anthers when these odours were directly compared in a choice-test, and this indicates that there were detectable differences between them. Anthers and pollen may therefore release distinctive odours that are quantitatively and/or qualitatively different from the odour of the rest of the flower. These experiments support the hypothesis that pollen-seeking insects use pollen odour cues to locate this food source.     

Unidirectionality of floral colour changes

Many angiosperms have arranged their flowers in inflorescences forming a distinct signalling unit to flower visitors. In some species, the flowers of inflorescences undergo a temporal colour change corresponding exactly to a change in the reward status. Based on information obtained from the spectral reflection curves of pre-change and postchage colours of flower corollas and/or floral guides, it was possible to demonstrate that the colour phase associated with reward closely corresponds to the visual stimuli which trigger behavioural responses of inexperienced flower visitors, and that the colour phase associated with less reward corresponds to visual stimuli less attractive to naïve flower visitors. Reciprocal colour changes were not observed. It is to be assumed that the unidirectionality of floral colour changes is an adaptation of angiosperms aimed at the guidance of first-time flower visitors. Signalling reward to inexperienced flower visitors is an additional function of floral colour changes. The main function of floral colour changes, however, is to provide cues with which the flower visitors can learn to associate one colour phase with reward.     

Colour similarity to rewarding model plants affects pollination in a food deceptive orchid, Orchis boryi

We studied the pollination of Orchis boryi at five different locations on the Greek mainland. Orchis boryi is food deceptive and obligatorily insect pollinated. Primary pollinators were Apis mellifera and Bombus spp., which foraged on rewarding plant species nearby and visited O. boryi in between. To analyse floral colour similarity among rewarding plants and O. boryi as perceived by bees, a model of bee colour vision was employed. For each food plant an index was calculated that described the probability of a bee foraging on it to subsequently choose an orchid flower. This choice probability correlated to colour distance according to the model of bee colour vision, indicating that bees chose the deceptive orchid more frequently if they foraged on more similarly coloured species. At different sites different plant species served as models. Bees foraging on food plants from which a high choice rate to the orchid was observed visited the orchid less often after approaching it than other bees, which is likely to reflect avoidance learning. In general, the pollination syndrome appears to be a generalized form of Batesian mimicry, in which similarity to rewarding plants determines reproductive success. As expected by negative density-dependent selection, individual fruit set and pollinia export rate correlated negatively with orchid density, but were unaffected by food plant density, orchid frequency, individual variation of labellum colour, labellum size, or mouth width of the flowers.