Colour Discrimination Functions of the Humming Bird Hawkmoth (Macroglosum Stellatarum) Modelled From Behavioural Choices

doi:10.21203/rs.3.rs-1706041/v1

Download PDF

Research Article

Colour Discrimination Functions of the Humming Bird Hawkmoth (Macroglosum Stellatarum) Modelled From Behavioural Choices

https://doi.org/10.21203/rs.3.rs-1706041/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The hummingbird hawkmoth (Macroglossum stellatarum) is a diurnal, visually oriented insect that has become an important animal model for comparative studies. Hawkmoths use colour information to find and identify profitable flowers, and they are major pollinators in various regions of the world. Recent research on other flower visiting animals including bees, hoverflies and birds shows that their respective colour choices for target-distractor stimuli of varying similarity can be reliably modelled using sigmoidal-type continuous functions, as postulated by von Helversen 50 years ago. We formally test if colour choices by M. stellatarum can also be reliably explained by these sigmoidal functions that are important for modelling plant-pollinator interactions in other animals. We fit multiple formulations of sigmoidal functions selecting those which best fit the observed data. We show that colour vision in hawkmoths varies across various regions of colour space. Our results are consistent with theories that colour choices are principally mediated by higher level neural processes in the animal brain, since it is at this level physiological signal modulation varies with stimulus modulation in a probabilistic fashion. These functions thus enable important insights into how to model how animals interact with colour stimuli like flowers in complex environments.

Humans typically sense the colour visual environment with three spectrally different photoreceptor classes, whose maximum sensitivity is respectively located at about 420 (blue-sensitive), 534 (green) and 564 (red) nm (Bowmaker and Dartnall 1980). Each receptor class pools all available sensed photons as described by the principle of univariance, and subsequent signals are projected to multiple higher level processing areas in the brain which enable the generation of colour perception via opponent neural mechanisms (Rushton 1972; Hurvich 1981; Lee 2005).

Flowers are amongst the most colourful objects in the natural environment, and from the earliest days of visual ecology seminal researchers appreciated that flowers typically evolved signals to suit the visual capabilities of the most important pollinators, which have very different colour visual capabilities compared to humans (von Frisch 1914; Kühn 1927; Daumer 1958; Dyer et al. 2021). Nevertheless, studies on human colour vision set the main foundation for animal colour vision studies (Kelber et al. 2003b; Kemp et al. 2015).

Our trichromatic visual system allows us to differentiate between saliently different colours like ripe fruit from green foliage (Mollon 1989). However, as colours become increasingly similar, discrimination accuracy drops until reaching a limit after which it is no longer possible to discern between the two stimuli, even if these differ in their spectral properties (MacAdam 1985). This relationship between colour (dis)similarity and accuracy of discrimination will take the characteristic shape of a letter “S” and often referred to as being sigmoidal. The sigmoidal shape of the colour discrimination function arises from the presence of both lower and upper asymptotes where discrimination accuracy remains constant in spite of changes in colour difference. The upper limit of a colour discrimination function is imposed by the impossibility of significantly increasing accuracy in colour discrimination after approaching approximately 100% of correct choices in spite of an increment in colour difference. On the other hand, an observer will continuously choose at random, i.e. an accuracy around 50%, when discriminating between colour stimuli that are so similar to each other as to be below a perceptual threshold. Only when the difference between stimuli is large enough to become apparent to the observer the function will start to rise, indicating the end of the lower plateau region, until reaching the upper asymptote of the function (Garcia et al. 2017).

Von Helversen (1972a), inspired by the experimental results of colour discrimination by the trichromatic European honeybee (Apis mellifera) (von Frisch 1914; Daumer 1956; von Helversen 1972b), proposed that the relationship between colour dissimilarity and discrimination accuracy by an animal observer should also be described by a sigmoidal curve. However, he did not provide a mathematical formulation for a function which could match his theoretical model, nor explicitly examine how the theory was supported by empirical data for different bee species or other animals.

The first empirical evidence of sigmoidal functions in hymenopteran insects was reported by Dyer and Chittka (2004a) for bumblebee colour choices, who noted that such a probabilistic way of processing colour information was consistent with evidence showing that the discharge rates of single neurons in macaque monkeys (Macaca fuscata) correlates with the relative position of colour similarity of stimuli on a colour map for primate vision (Komatsu and Ideura 1993). This fits with other evidence that whilst there are some differences in colour processing mechanisms between primates and insects, overall there are highly convergent visual strategies for chromatic discrimination (Pichaud et al. 1999, Hempel de Ibarra et al. 2014). Other research on bumblebees (Dyer and Chittka 2004c) and honeybees (Giurfa 2004; Avarguès-Weber et al. 2010; Dyer and Garcia 2014) shows that conditioning experience has on effect on the ability of an hymenopteran observer to discriminate colours and lead to changes in the brain and long-term memory (Sommerlandt et al. 2016), suggesting that colour discrimination is influenced by neural tuning processes in these animals.

Hawkmoths include various diurnal and nocturnal visually driven moth species which feed on flower nectar and are considered to be major pollinators in various ecosystems across their distribution ranges. For example, hawkmoths are responsible for pollinating 10% of all tree species in tropical dry forests in addition to various species of shrubs, herbs and epyphites (Haber and Frankie 1989), and are the main pollinators of orchid species in Madagascar (Nilsson et al. 1992) and North America (Fox et al. 2013). In Costa Rica, long tongued hawkmoths are reported as being generalist flower visitors, feeding on many different species of flowers, including those adapted to other animal pollinators such as bats and hummingbirds, which display a variety of different colours (Haber and Frankie 1989).

The hummingbird hawkmoth (Macroglossum stellatarum) is a diurnal, visually oriented insect which uses colour information to find and identify profitable flowers (Kelber et al. 2003a), and can override innate preferences following training to discriminate between distinct colour stimuli (Kelber 1996). M. stellatarum demonstrates a capacity to be a food generalist and forage on a variety of different spectrally different colour stimuli to obtained preferred sugar nutrition (Kelber 2003). M. stellatarum has functional trichromatic vision (Kelber and Henique 1999) which facilitates a capacity to discriminate between spectral colours at a level which is finer than honeybees (Telles et al. 2016). It is also known that hawkmoths show evidence of improved capacity to discriminate between colours following differential conditioning to stimuli (Kelber 2010). However, in spite of the current knowledge on the colour vision of M. stellatarum, it remains untested if discrimination for (dis)similar colour differences in this species can be reliably described by a continuous, sigmoidal function as has recently been demonstrated for other major pollinators including insects like bees and flies (Garcia et al. 2017, 2018; Hannah et al. 2019; Garcia et al. 2022), and violet sensitive birds (Garcia et al. 2021).

Here we specifically test a hypothesis that sigmoidal functions may describe colour choices in hawkmoths by using previously recorded behavioural data on colour discrimination of (dis)similar colour differences by M. stellatarum (Telles et al. 2016). If this hypothesis proves true, our model allows predicting accuracy of colour discrimination without resorting to assumptions of currently unknown physiological parameters of M. stellatarum colour vision. We discuss the outcomes of our modelling with respect to underlying mechanistic accounts of whether biologically relevant colour discrimination in these animals is principally mediated due to receptor level, and/or higher level neural processes.

Colour modelling

We modelled colour discrimination functions for the hummingbird hawkmoth (Macroglosum stellatarum) corresponding to the six different reference quasi-monochromatic stimuli reported by (Telles et al. 2016). To control for potential differences in discrimination accuracy across the spectrum, we fitted functions to the respective colour differences corresponding to stimuli of either longer- or shorter-wavelength to each training reference stimulus..

Functions were fitted following the protocol detailed in Garcia et al. (2017). Firstly, the spectral profile of the stimuli used for the behavioural experiment by Telles et al. (2016) were expressed as a loci in the Maxwell colour space (Fig. 1, panel a) implementing standard colorimetric calculations (Renoult et al. 2017). The selection of the colour space was based on the fact that this model is suitable for trichromatic vision and makes no assumptions of the currently unknown potential opponent mechanisms in the vision of M. stellatarum (Kelber et al. 2003a).

We modelled each monochromatic using a Gaussian function of form:

$$f\left(x\right)=\text{e}\text{x}\text{p}-\left(\frac{{\left(x-b\right)}^{2}}{{2c}^{2}}\right)$$

where b denotes the wavelength corresponding to the peak transmission of the stimuli and c = 5.24. The constant c ensures that the half bandwidth of each stimulus is equal to about 15 nm, as reported by Telles et al. (2016). Eq. 1 was evaluated within 300 to 700 nm at 5 nm intervals (Fig. 1 panels b and c).

We used the exponential functions proposed by Stavenga (1993) as template to model the spectral sensitivity functions for the three photoreceptors present in the visual system of M. stellatarum. Peak absorption of each photoreceptwas located at 349, 440 and 521nm (Telles et al. 2014), with a secondary peak of absorption, β-band, located at 320 nm.

For our calculations we employed the illumination profile available for the behavioural experiments consisting on a combination of white light emitting diodes and florescent light tubes (Telles et al. 2014, 2016). We assumed a spectrum uniformly reflecting 30% of all incident radiation across the 300 to 700 nm interval to match the experimental conditions reported by Telles et al (2016) as adaptation background for colour modelling. Finally, we modelled chromatic adaptation by M. stellatarum using a standard von Kries type scaling function (Garcia et al. 2017). Colour difference between reference and each tested stimuli pair is here expressed as the Euclidean distance between their respective loci in the Maxwell colour space, although the formulation of colour discrimination function as a continuous function can accommodates other colour distance metrics; see for example (Garcia et al. 2017, 2018, 2022). All colour modelling was done using custom written scripts for Matlab release 2017b.

Fitting of the colour discrimination function

The response variable of the colour discrimination functions for M. stellatarum consists of the proportion of correct choices corresponding to each colour difference tested. We thus calculated the mean proportion of correct choices for each pair discrimination experiment by dividing the number of correct choices over the total number of choices made by all tested hawkmoths as reported in the supplementary material of Telles et al. (2016).

We fitted two sigmoidal functions (Equations 2 and 3) of different complexity to each dataset, as the shape of these functions has been shown to successfully describe colour discrimination by hymenopterans and dipteran pollinators (Garcia et al. 2017, 2018, 2022).

$$\pi \left(\varDelta C\right)=\frac{{M}_{o}K}{{M}_{o}+\left(K-{M}_{o}\right)\text{e}\text{x}\text{p}\left(-r\bullet \varDelta C\right)}$$

$\pi \left(\varDelta C\right)=\frac{{M}_{o}+(K-{M}_{o})}{1+\text{exp}\left(\frac{xmid-\varDelta C}{r}\right)}$ s(3)

In Equations 2 and 3, the variable π denotes the probability of accurate discrimination of the colour difference (ΔC). We expressed ΔC as the Euclidean distance of two loci in the Maxwell triangle, although other metrics can also be implemented (Garcia et al. 2017, 2018, 2022), and the constants define the unique shape of the function.

Values for the constants K, r, Mo, xmid and scal were found by implementing the non-linear least square routine (nls) available in the nlme package (Pinheiro et al. 2013) for R release 4.0.2. If for any given colour discrimination function both Equations fitted the data, a loglikelihood test was performed to identify the model which better described the dataset. If the null hypothesis of equality failed to be rejected, the simpler model was selected in favour of the more complex alternative following standard model selection practices (Crawley 2013).

Once we identified the best model fitting each data set, we used simulations to recover the uncertainty around model parameters as proposed by (Garcia et al. 2017). Briefly, for each function we generated 100,000 realisations of each discrimination experiment by randomly sampling from a binary distribution whose defining parameters are given by results reported by Telles et at (Telles et al. 2016) and fitted the selected equation to each simulated response set. Median and 95% confidence intervals (CI) for each coefficient were then calculated from the distribution of values obtained from the simulation.

We obtained eight different colour discrimination functions out of the twelve data sets analysed (Table 1). Functions were obtained for the long wavelength stimuli tested relative to a reference of 380 nm, and for both the short and long wavelength stimuli of references centred at 400, 440 and 480 nm.

Three and four parameter functions fitted the datasets corresponding to the longer wavelength stimuli tested for references centred at 400 and 440 nm. In both cases the more complex function provided a better fit than the simpler, three parameters alternative (χ²_{400 nm} = 11.1, df = 1, P = 0.001, (χ²_{440 nm} = 5.54, df = 1, P = 0.018), thus the more complex function (Eq. 3) was selected for these two cases (Table 1). All remaining discrimination functions were fitted by a three parameter logistic model (Eq. 2). Figure 1 depicts the colour discrimination function obtained for those reference wavelengths showing a significant increase in probability of accurate discrimination as a function of colour difference.

Table 1

Reference and tested stimuli used to derive the seven different colour discrimination functions for the hummingbird hawkmoth (*Macroglossum stellatarum*), values for their defining coefficients and colour difference (ΔC) predictions for accuracy thresholds (π) of 60 and 0 %. Colour differences are expressed as units in the Maxwell triangle and obtained by inverting Equations 2 or 3 and solving for ΔC for each accuracy value. * Correct choices for this reference did not significantly increase from chance level so it is not possible to predict colour distances from this function.
Reference wavelength	Tested stimuli	Discrimination function	Parameter values (95% Confidence interval).	ΔC_{π 0 %}	ΔC_{π 0 %}
380	Shorter n = 4	No fit
380	Longer n = 6	3-log (Eq. 2)	K = 1.02 (0.870, 1.33) r = 20.9 (3.33, 111) Mo = 0.448 (0.083, 0.705)	0.028	0.049
400	Shorter n = 5	3-log (Eq. 2	K = 0.934 (0.697, 1.92) r = 15.5 (1.65, 304) Mo = 0.42 (0.02, 0.758)	0.051	0.084
400	Longer n = 6	4-log (Eq. 3)	K = 0.323 (-0.456, 0.591) Mo = 0.892 (0.667, 2.93) xmid = 0.329 (0.213, 0.497) scal = 0.003 (0.002, 0.01)	0.329	0.331
440	Shorter n = 6	3-log (Eq. 2)	K = 0.928 (0.725, 2.38) r = 57 (8.57, 389) Mo = 0.387 (0.004, 0.677)	0.026	0.016
440	Longer n = 5	4-log (Eq. 3)	K = 0.330 (-1.04, 0.634) Mo = 1.00 (0.821, 1.21) xmid = 0.143 (0.004, 0.190) scal = 0.015 (0.001, 0.076)	0.146	0.137
480	Shorter n = 5	3-log (Eq. 2)	K = 0.932 (0.723, 2.73) r = 16.1 (-53.9, 75.7) Mo = 0.079 (0.00, 0.887)	0.216	0.184
480	Longer n = 5	3-log (Eq. 2)	K = 0.941 (0.771, 1.71) r = 18.0 (2.21, 140) Mo = 0.5 (0.035, 0.772)	0.052	0.024
520	Shorter n = 6	No fit
520	Longer n = 4	3-log (Eq. 2)	K = 0.725 (0.449, 1.68) r = -348 (-6810, 462) Mo = 0.661 (0.000, 1.45)	NaN*	NaN*
560	Shorter n = 5	No fit
560	Longer n = 4	No fit

In the current manuscript we used mathematical descriptions of the sigmoidal functions originally proposed by von Helversen (1972a) to predict accuracy of colour discrimination by hummingbird hawkmoths based on perceived colour difference to different reference stimuli (Fig. 2). These functions provide a prediction to the question of the effect of colour dissimilarity on colour discrimination by hawkmoths based on behavioural observations. In bees (Dyer and Chittka 2004c; Sommerlandt et al. 2016) several studies show that conditioning can enable improved colour discrimination, and enable long-term memory formation that results in fundamental changes in the brain. These changes likely define the “inner conditions” responsible for the behavioural responses observed, as proposed by von Helversen (1972a). It has also been reported that hawkmoths improve colour discrimination with conditioning (Kelber 2010), although evidence of how this may influence the physiology of the brain remains outstanding. The sigmoidal shaped functions have now been demonstrated to reliably predict colour choices in hawkmoths (Fig. 2), bees (Garcia et al. 2017) and a variety of other animal pollinators (Garcia et al. 2021, 2022), and physiological evidence suggests that the way neurons respond to perceptually (dis)similar colour stimuli (Komatsu and Ideura 1993; Dyer and Chittka 2004c) is a plausible explanation for such observations.

Initial estimates of colour discrimination thresholds for M. stellatarum (Telles et al., 2016) were based on the receptor-noise threshold model (Vorobyev and Osorio 1998) which assumes that colour discrimination is limited by purely physiological aspects of vision namely, the fixed signal to noise ratio imposed by the density and number of photoreceptor available in the visual system of the observer (Vorobyev et al. 2001). This assumption, however, does not seem to hold entirely when considering the data obtained from discrimination experiments of similar colours by this species. More specifically, Telles et al. (2016), noted that whilst the receptor-noise model provides a partial fit to the observed hawkmoth choices, an overall fit was not observed. The colour discrimination functions presented (Fig. 2) account for the entire processing of the colour signals by visual system of the animal without making assumptions of currently unknown physiological aspects of the system as these are modelled directly on the animal’s response. Thus with quality psychophysics and modelling, the problem of unknown physiological data can be avoided, and improved predicted power is possible (Table 1).

Modelling of the behavioural data by Telles et al., (2016) revealed evidence of considerable variability in animal choices for using colour information (Fig. 2). Such evidence is consistent with data for bee foragers where other ecological factors impact on how and why flower visitors use colour information to make decisions. For example, both honeybees (Ng et al. 2021) and bumblebees (Chittka et al. 2003, 2009; Dyer and Chittka 2004b) show that bees trade of speed for accuracy, and Evans and Raine (2014) show that such a variability in accuracy may be optimal in a foraging environment with many hive mates engaging in different exploratory behaviour. An important component of sigmoidal functions describing animal colour choices is that these functions also enable capturing such complexity of animal behaviour through the variability around their defining coefficients (Table 1).

Our results also confirm differences in sensitivity to colour differences across the sampled colour space of M. stellatarum (Figs. 1 and 2), as was reported by Telles et al. (2016). This aspect of colour perception has been well appreciated in humans (Hurvich 1981), bees (von Helversen 1972b; Dyer and Neumeyer 2005; Garcia et al. 2017), flies (Hannah et al. 2019) and pigeons (Wright 1972; Garcia et al. 2021) suggesting that across different regions of the visible spectrum, not all colour are discriminated in the same way. Such differences are likely to be due to both physiological mechanisms including how photoreceptors overlap and opponent neural processing of photoreceptor input (Rushton 1972; Hurvich 1981; Kelber et al. 2003b; Lee 2005; Kemp et al. 2015).

To model colour discrimination by M. stellatarum using the proposed method it is first necessary to identify the more likely colour to be discriminated and choose one of the functions in Fig. 2a-g whose reference wavelength is closest to that of the problem stimuli. Then the respective coefficients in Table 1 are used in either Eq. 1 or 2 to predict accuracy of discrimination of a colour distance input by user. Alternatively Equations 1–2 can be inverted to solve for the colour difference required for attaining a given accuracy level as shown in Table 1.

An important caveat of the model predictions are that these are based on quasi-monochromatic, narrow band stimuli which are not often observed on natural stimuli. This limitation implies that the accuracy of the predicted discrimination around threshold colour differences may be different to those potentially observed for natural stimuli. However we expect that discrimination accuracy for small colour differences between natural stimuli are within the uncertainty bounds of our functions (Fig. 2). Likewise it is also important to point out that none of the tested quasi-monochromatic stimuli used for our modelling correspond to those of lavender flowers often visited by M. stellatarum in natural environments and that any potential effect of this preference in colour discrimination is not accounted for by the proposed models. Finally another important point to consider is that our model assumes a flat, achromatic background which is not representative of most naturally occurring scenarios. Indeed, colour discrimination in heterogenous background, as those typically occurring in nature, tend to be more difficult than when considering homogenous backgrounds and hymenopteran vision (Bukovac et al. 2017). Therefore, under most naturally occurring viewing conditions, accuracy of discrimination by M. stellatarum may be lower than the one predicted by our models.

Plant-pollinator interactions are one of the most conspicuous and well-studied fields of visual ecology (Kemp et al. 2019; Shrestha et al. 2019), dating back to seminal formulations of evolutionary theory proposed by Darwin (1876). Whilst it has been established for some time that colour is the main signal used by different pollinators as their respective colour discrimination capabilities match the colours that most frequently evolve (Chittka and Menzel 1992; Dyer et al. 2012), until recently it has remained difficult to interpret how individual pollinators of different species may process (dis)similar colours (Dyer and Chittka 2004a, Shrestha et al., 2019). The methods here proposed provide a valuable new solution to approach questions about complex plant-pollinator interactions involving the model hawkmoth M. stellatarum, based on empirical behavioural data reflecting the outcome of colour signal processing by the entire visual system of an observer.

Acknowledgment

AGD was supported by the Australian Research Council Discovery Project (DP160100161).

Autrum H, Zwehl V v (1964) Die spektrale Empfindlichkeit einzelner Sehzellen des Bienenauges. Z Vgl Physiol 48:357–384
Avarguès-Weber A, de Brito Sanchez MG, Giurfa M, Dyer AG (2010) Aversive Reinforcement Improves Visual Discrimination Learning in Free-Flying Honeybees. PLoS One 5:e15370. https://doi.org/10.1371/journal.pone.0015370
Bowmaker JK, Dartnall HJ (1980) Visual pigments of rods and cones in a human retina. J Physiol 298:501–511
Bukovac Z, Shrestha M, Garcia JE, Burd M, Dorin A, Dyer AG (2017) Why background matters to bees and flowers. J Comp Physiol A 203:369–380
Chittka L, Dyer AG, Bock F, Dornhaus A (2003) Bees trade-off foraging speed for accuracy. Nature 424:388
Chittka L, Menzel R (1992) The evolutionary adaptation of flower colours and the insect pollinators’ colour vision. J Comp Physiol A 171:171–181. https://doi.org/10.1007/BF00188925
Chittka L, Skorupski P, Raine NE (2009) Speed–accuracy tradeoffs in animal decision making. Trends Ecol Evol 24:400–407. https://doi.org/10.1016/j.tree.2009.02.010
Crawley MJ (2013) The R book, Second. John Wiley & Sons, West Sussex
Darwin C (1876) The effects of cross and self-fertilisation in the vegetable kingdom. John Murray, London
Daumer K (1958) Blumenfarben, wie sie die Bienen sehen. Z Vgl Physiol 41:49–110
Daumer K (1956) Reizmetrische Untersuchung des Farbensehens der Bienen. Z Vgl Physiol 38:413–478. https://doi.org/10.1007/BF00340456
Dyer AG, Boyd-Gerny S, McLoughlin S, et al (2012) Parallel evolution of angiosperm colour signals: common evolutionary pressures linked to hymenopteran vision. Proc R Soc B Biol Sci 279:3606–3615. https://doi.org/10.1098/rspb.2012.0827
Dyer AG, Chittka L (2004a) Biological significance of distinguishing between similar colours in spectrally variable illumination: bumblebees (Bombus terrestris) as a case study. J Comp Physiol 190:105–114. https://doi.org/http://dx.doi.org/10.1007/s00359-003-0475-2
Dyer AG, Chittka L (2004b) Bumblebees (Bombus terrestris) sacrifice foraging speed to solve difficult colour discrimination tasks. J Comp Physiol A 190:759–763. https://doi.org/10.1007/s00359-004-0547-y
Dyer AG, Chittka L (2004c) Fine colour discrimination requires differential conditioning in bumblebees. Naturwissenschaften 91:224–227
Dyer AG, Garcia JE (2014) Color Difference and Memory Recall in Free-Flying Honeybees: Forget the Hard Problem. Insects 5:629–638. https://doi.org/10.3390/insects5030629
Dyer AG, Jentsch A, Burd M, et al (2021) Fragmentary Blue: Resolving the Rarity Paradox in Flower Colors. Front Plant Sci 11:. https://doi.org/10.3389/fpls.2020.618203
Dyer AG, Neumeyer C (2005) Simultaneous and successive colour discrimination in the honeybee (Apis mellifera). J Comp Physiol 191:547–557. https://doi.org/http://dx.doi.org/10.1007/s00359-005-0622-z
Dyer AG, Paulk AC, Reser DH (2011) Colour processing in complex environments: insights from the visual system of bees. Proc R Soc B Biol Sci 278:952–959. https://doi.org/10.1098/rspb.2010.2412
Evans LJ, Raine NE (2014) Foraging errors play a role in resource exploration by bumble bees (Bombus terrrestris). J Comp Physiol A 200:475–484
Fox K, Vitt P, Anderson K, et al (2013) Pollination of a threatened orchid by an introduced hawk moth species in the tallgrass prairie of North America. Biol Conserv 167:316–324. https://doi.org/10.1016/j.biocon.2013.08.026
Garcia JE, Hannah L, Shrestha M, et al (2022) Fly pollination drives convergence of flower coloration. New Phytol 233:52–61. https://doi.org/10.1111/nph.17696
Garcia JE, Rohr DH, Dyer AG (2021) Colour Discrimination From Perceived Differences by Birds. Front Ecol Evol 9:1–13. https://doi.org/10.3389/fevo.2021.639513
Garcia JE, Shrestha M, Dyer AG (2018) Flower signal variability overwhelms receptor-noise and requires plastic color learning in bees. Behav Ecol 29:1286–1297. https://doi.org/10.1093/beheco/ary127
Garcia JE, Spaethe J, Dyer AG (2017) The path to colour discrimination is S-shaped: behaviour determines the interpretation of colour models. J Comp Physiol A 203:983–997. https://doi.org/10.1007/s00359-017-1208-2
Giurfa M (2004) Conditioning procedure and color discrimination in the honeybee Apis mellifera. Naturwissenschaften 91:228–231
Haber WA, Frankie GW (1989) A Tropical Hawkmoth Community: Costa Rican Dry Forest Sphingidae. Biotropica 21:155–172. https://doi.org/10.2307/2388706
Hannah L, Dyer AG, Garcia JE, et al (2019) Psychophysics of the hoverfly: categorical or continuous color discrimination? Curr Zool 65:483–492. https://doi.org/10.1093/cz/zoz008
Hempel de Ibarra N, Vorobyev M, Menzel R (2014) Mechanisms, functions and ecology of colour vision in the honeybee. J Comp Physiol A 200:411–433 http://doi.org/ï»¿10.1007/s00359-014-0915-1
Hurvich LM (1981) Color Vision. Sinauer Associates Inc., Sunderland
Kelber A (2003) Sugar preferences and feeding strategies in the hawkmoth Macroglossum stellatarum. J Comp Physiol A 189:661–666. https://doi.org/10.1007/s00359-003-0440-0
Kelber A (2010) What a hawkmoth remembers after hibernation depends on innate preferences and conditioning situation. Behav Ecol 21:1093–1097. https://doi.org/10.1093/beheco/arq115
Kelber A (1996) Colour learning in the hawkmoth Macroglossum stellatarum. J Exp Biol 199:1127–1131. https://doi.org/10.1242/jeb.199.5.1127
Kelber A, Balkenius A, Warrant EJ (2003a) Colour vision in diurnal and nocturnal hawkmoths. Integr Comp Biol 43:571–579. https://doi.org/10.1093/icb/43.4.571
Kelber A, Henique U (1999) Trichromatic colour vision in the hummingbird hawkmoth, Macroglossum stellatarum L. J Comp Physiol - A Sensory, Neural, Behav Physiol 184:535–541. https://doi.org/10.1007/s003590050353
Kelber A, Vorobyev M, Osorio D (2003b) Animal colour vision - behavioural test and physiological concepts. Biol Rev 78:81–118. https://doi.org/10.1017/S1464793102005985
Kemp DJ, Herberstein ME, Fleishman LJ, et al (2015) An Integrative Framework for the Appraisal of Coloration in Nature. Am Nat 185:705–724. https://doi.org/10.1086/681021
Kemp JE, Bergh NG, Soares M, Ellis AG (2019) Dominant pollinators drive non-random community assembly and shared flower colour patterns in daisy communities. Ann Bot 123:277–288. https://doi.org/10.1093/aob/mcy126
Kien J, Menzel R (1977) Chromatic properties of interneurons in the optic lobes of the bee. II. Narrow band and colour opponent neurons. J Comp Physiol 113:35–53. https://doi.org/10.1007/BF00610452
Komatsu H, Ideura Y (1993) Relationships between color, shape, and pattern selectivities of neurons in the inferior temporal cortex of the monkey. J Neurophysiol 70:677–694. https://doi.org/10.1152/jn.1993.70.2.677
Kühn A (1927) Über den Farbensinn der Bienen. Z Vgl Physiol 5:762–800
Lee H-C (2005) Introduction to Color Imaging Science. Cambridge University Press, Cambridge
MacAdam DL (1985) Color Measurement: Theme and Variations. Springer-Verlag, Berlin, Germany
Mollon JD (1989) “Tho’she kneel’d in that place where they grew… The uses and origins of primate colour vision. J Exp Biol 146:21–38
Ng L, Garcia JE, Dyer AG (2021) Mission impossible: honey bees adjust time allocation when facing an unsolvable task. Anim Behav 182:59–66. https://doi.org/10.1016/j.anbehav.2021.10.001
Nilsson LA, Rabakonandrianina E, Razananaivo R, Randriamanindry J (1992) Long pollinia on eyes: hawk-moth pollination of Cynorkis uniflora Lindley (Orchidaceae) in Madagascar. Bot J Linn Soc 109:145–160. https://doi.org/10.1111/j.1095-8339.1992.tb00263.x
Pichaud F, Briscoe A, Desplan C (1999) Evolution of colour vision. Curr Opin Neurobiol 9:622–627. ï»¿https://doi.org/10.1016/S0959-4388(99)00014-8
Pinheiro J, Bates D, DebRoy S, Sarkar D (2013) nlme: linear and nonlinear mixed effects models.
Renoult JP, Kelber A, Schaefer HM (2017) Colour spaces in ecology and evolutionary biology. Biol Rev 92:292–315. https://doi.org/10.1111/brv.12230
Rushton WAH (1972) Review lecture. Pigments and signals in colour vision. J Physiol 220:1–31
Shrestha M, Dyer AG, Garcia JE, Burd M (2019) Floral colour structure in two Australian herbaceous communities: it depends on who is looking. Ann Bot 124:221–232. https://doi.org/10.1093/aob/mcz043
Sommerlandt FMJ, Spaethe J, Rössler W, Dyer AG (2016) Does fine color discrimination learning in free-flying honeybees change mushroom-body calyx neuroarchitecture? PLoS One 11:e0164386. https://doi.org/10.1371/journal.pone.0164386
Stavenga DG, Smits RP, Hoenders BJ (1993) Simple exponential functions describing the absorbance bands of visual pigment spectra. Vision Res 33:1011–1017. https://doi.org/10.1016/0042-6989(93)90237-Q
Telles FJ, Kelber A, Rodríguez-Gironés MA (2016) Wavelength discrimination in the hummingbird hawkmoth Macroglossum stellatarum. J Exp Biol 219:553–560. https://doi.org/10.1242/jeb.130484
Telles FJ, Lind O, Henze MJ, et al (2014) Out of the blue: the spectral sensitivity of hummingbird hawkmoths. J Comp Physiol A 200:537–546
von Frisch K (1914) Der Farbensinn und Formensinn der Biene. Zool Jahrbücher Abteilung für Allg Zool und Physiol 37:1–187
von Helversen O (1972a) The relationship between difference in stimuli and choice frequency in training experiments with the honeybee. In: Wehner R (ed) Information Processing in the Visual Systems of Anthropods. Springer Berlin Heidelberg, pp 323–334
von Helversen O (1972b) Zur spektralen Unterschiedsempfindlichkeit der honigbiene. J Comp Physiol 80:439–472
Vorobyev M, Brandt R, Peitsch D, et al (2001) Colour thresholds and receptor noise: behaviour and physiology compared. Vision Res 41:639–653. https://doi.org/10.1016/s0042-6989(00)00288-1
Vorobyev M, Osorio D (1998) Receptor noise as a determinant of colour thresholds. Proc R Soc London Ser B Biol Sci 265:351–358. https://doi.org/10.1098/rspb.1998.0302
Wright AA (1972) Psychometric and psychophysical hue discrimination functions for the pigeon. Vision Res 12:1447–1464. https://doi.org/http://dx.doi.org/10.1016/0042-6989(72)90171-X
Yang E-C, Lin H-C, Hung Y-S (2004) Patterns of chromatic information processing in the lobula of the honeybee, Apis mellifera L. J Insect Physiol 50:913–925. https://doi.org/10.1016/j.jinsphys.2004.06.010

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Colour Discrimination Functions of the Humming Bird Hawkmoth (Macroglosum Stellatarum) Modelled From Behavioural Choices

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Colour modelling

Fitting of the colour discrimination function

Results

Discussion

Declarations

References

Additional Declarations

Status:

Version 1