Behavioural evidence for marsupial trichromacy
2006; Elsevier BV; Volume: 16; Issue: 6 Linguagem: Inglês
10.1016/j.cub.2006.02.036
ISSN1879-0445
AutoresCatherine A. Arrese, L.D. Beazley, Christa Neumeyer,
Tópico(s)Primate Behavior and Ecology
ResumoThe ability to discriminate red–green colours was thought to be unique among mammals to trichromatic primates [1Jacobs G.H. The distribution and nature of colour vision among the mammals.Biol. Rev. 1993; 68: 413-471Crossref PubMed Google Scholar, 2Regan B.C. Julliot C. Simmen B. Viénot F. Charles-Dominique P. Mollon J.D. Fruit, foliage and the evolution of primate colour vision.Phil. Trans. R. Soc. B. 2001; 356: 229-283Crossref PubMed Scopus (359) Google Scholar], until recent microspectrophotometric studies revealed that marsupials also have the potential for trichromatic colour vision [3Arrese C.A. Hart N.S. Thomas N. Beazley L.D. Shand J. Trichromacy in marsupials.Curr. Biol. 2002; 12: 657-660Abstract Full Text Full Text PDF PubMed Scopus (116) Google Scholar, 4Arrese C.A. Oddy A.Y. Runham P.B. Hart N.S. Shand J. Hunt D.M. Beazley L.D. Cone topography and spectral sensitivity in two potentially trichromatic marsupials, the quokka (Setonix brachyurus) and quenda (Isoodon obesulus).Proc. R. Soc. B. 2005; 272: 791-796Crossref PubMed Scopus (51) Google Scholar]. Functional colour vision cannot be inferred from physiological studies alone [5Kelber A. Vorobyev M. Osorio D. Animal colour vision – behavioural tests and physiological concepts.Biol. Rev. 2003; 78: 81-118Crossref PubMed Scopus (634) Google Scholar, 6Goldsmith T.H. Optimization, constraint, and history in the evolution of eyes.Quart. Rev. Biol. 1990; 65: 281-322Crossref PubMed Scopus (402) Google Scholar, 7Neumeyer C. Wavelength discrimination in the goldfish.J. Comp. Physiol. A. 1986; 158: 203-213Crossref Scopus (87) Google Scholar, 8Neumeyer C. Tetrachromatic colour vision in goldfish: evidence from colour mixture experiments.J. Comp. Physiol. A. 1992; 171: 639-649Crossref Scopus (120) Google Scholar], however, a point of particular importance in this case as molecular analyses have failed to identify the third marsupial cone photoreceptor type [9Strachan J. Chang L.Y. Wakefield M.J. Graves J.A. Deeb S. Cone visual pigments of the Australian marsupials, the stripe-faced and fat-tailed dunnarts: sequence and inferred spectral properties.Vis. Neurosci. 2004; 21: 223-229Crossref PubMed Scopus (27) Google Scholar]. Here we report that an Australian marsupial, the fat-tailed dunnart (Sminthopsis crassicaudata), has trichromatic colour vision that encompasses ultraviolet sensitivity. We previously reported maximum spectral absorbances (λmax) of the cone photoreceptors in the fat-tailed dunnart at 363, 509 and 535 nm, representing sensitivity to ultraviolet (UVS), medium (MWS) and long (LWS) wavelengths, respectively [3Arrese C.A. Hart N.S. Thomas N. Beazley L.D. Shand J. Trichromacy in marsupials.Curr. Biol. 2002; 12: 657-660Abstract Full Text Full Text PDF PubMed Scopus (116) Google Scholar]. Although spectral absorbance characteristics were determined for the three cone types, only behavioural experiments can establish unequivocally their contribution to colour discrimination. We therefore investigated the dimensionality of colour vision in the fat-tailed dunnart using additive colour mixture experiments, where choice between a coloured light (training wavelength) and an additive mixture of two different coloured lights (primary wavelengths) is based on differences in chromatic content and occurs regardless of relative brightness [5Kelber A. Vorobyev M. Osorio D. Animal colour vision – behavioural tests and physiological concepts.Biol. Rev. 2003; 78: 81-118Crossref PubMed Scopus (634) Google Scholar, 6Goldsmith T.H. Optimization, constraint, and history in the evolution of eyes.Quart. Rev. Biol. 1990; 65: 281-322Crossref PubMed Scopus (402) Google Scholar, 7Neumeyer C. Wavelength discrimination in the goldfish.J. Comp. Physiol. A. 1986; 158: 203-213Crossref Scopus (87) Google Scholar, 8Neumeyer C. Tetrachromatic colour vision in goldfish: evidence from colour mixture experiments.J. Comp. Physiol. A. 1992; 171: 639-649Crossref Scopus (120) Google Scholar] (see Supplemental Data available with this paper online for experimental procedures). Using the normalised absorbance data from our previous measurements, we constructed a colour triangle representing the colour space of the fat-tailed dunnart (Figure 1A) and determining the training and primary wavelengths to be tested. We trained three dunnarts (two males and one female) to recognise training wavelengths and tested their choice against another coloured light of a different wavelength. To ensure that discrimination was based on chromatic cues, the intensity of the two stimuli was varied [5Kelber A. Vorobyev M. Osorio D. Animal colour vision – behavioural tests and physiological concepts.Biol. Rev. 2003; 78: 81-118Crossref PubMed Scopus (634) Google Scholar, 6Goldsmith T.H. Optimization, constraint, and history in the evolution of eyes.Quart. Rev. Biol. 1990; 65: 281-322Crossref PubMed Scopus (402) Google Scholar, 7Neumeyer C. Wavelength discrimination in the goldfish.J. Comp. Physiol. A. 1986; 158: 203-213Crossref Scopus (87) Google Scholar, 8Neumeyer C. Tetrachromatic colour vision in goldfish: evidence from colour mixture experiments.J. Comp. Physiol. A. 1992; 171: 639-649Crossref Scopus (120) Google Scholar]. Once choice frequency exceeded 80%, training wavelengths were tested against mixtures of two primary wavelengths, using an experimental set-up modified from Neumeyer [8Neumeyer C. Tetrachromatic colour vision in goldfish: evidence from colour mixture experiments.J. Comp. Physiol. A. 1992; 171: 639-649Crossref Scopus (120) Google Scholar] (Figure 1B). To assess discrimination between green, yellow and orange coloured lights, the dunnarts were trained on wavelengths of 535, 555 and 565 nm, and their choice tested against additive mixtures of 525 and 580 nm. Discrimination towards the red region of the spectrum was assessed with 580 nm as training wavelength and 525 and 620 nm as mixed primaries. The three dunnarts discriminated all training wavelengths with 80–100% accuracy until they were matched by primary mixtures, at which choice frequency decreased to 50% (Table 1). For the training wavelengths 535, 555 and 565 nm, the dunnarts could no longer differentiate the two test fields at mixture ratios of 80:20%, 40:60% and 30:70% of 525 and 580 nm, respectively. A match was found at 50:50% of 525 and 620 nm when dunnarts were trained on 580 nm. Their ability to discriminate between various wavelengths ranging from 525–620 nm provided an initial indication of trichromacy.Table 1Results of additive colour mixture trialsTrialsTraining wavelength (nm) or training 'white' lightPrimary mixture (nm)Matching primary mixture ratios1535525 + 58080%(525):20%(580)2555525 + 58040%(525):60%(580)3565525 + 58030%(525):70%(580)4580525 + 62050%(525):50%(620)5atungsten-'white'360 + 52520%(360):80%(525)5b525450 + 62060%(450):40%(620)6atungsten-'white'410 + 62030%(410):70%(620)6b410360 + 45080%(360):20%(450)Results of additive colour mixture trials selected from the colour triangle of the fat-tailed dunnart, showing ratios at which primary mixtures match training wavelengths or tungsten-'white' light. For the eight trials, each ratio of primary mixture was tested ten times. Depending on the mixture ratio at which matching occurred, the dunnarts undertook 50–100 experimental sessions per trial. As no significant differences were found between the performances of the three dunnarts (P > 0.05), results for each trial were pooled (n = 30). For a critical value of 0.5 (or 50%), the p-value obtained from the Binomial Distribution Table (p < 0.05), indicated a significant difference between performance due to chance, and that due to discrimination. Open table in a new tab Results of additive colour mixture trials selected from the colour triangle of the fat-tailed dunnart, showing ratios at which primary mixtures match training wavelengths or tungsten-'white' light. For the eight trials, each ratio of primary mixture was tested ten times. Depending on the mixture ratio at which matching occurred, the dunnarts undertook 50–100 experimental sessions per trial. As no significant differences were found between the performances of the three dunnarts (P > 0.05), results for each trial were pooled (n = 30). For a critical value of 0.5 (or 50%), the p-value obtained from the Binomial Distribution Table (p < 0.05), indicated a significant difference between performance due to chance, and that due to discrimination. Following the rationale that where three cone signals contribute to colour vision, any spectral stimulus can be matched with a specific mixture of three primary wavelengths [5Kelber A. Vorobyev M. Osorio D. Animal colour vision – behavioural tests and physiological concepts.Biol. Rev. 2003; 78: 81-118Crossref PubMed Scopus (634) Google Scholar, 6Goldsmith T.H. Optimization, constraint, and history in the evolution of eyes.Quart. Rev. Biol. 1990; 65: 281-322Crossref PubMed Scopus (402) Google Scholar, 8Neumeyer C. Tetrachromatic colour vision in goldfish: evidence from colour mixture experiments.J. Comp. Physiol. A. 1992; 171: 639-649Crossref Scopus (120) Google Scholar], we further assessed trichromacy by investigating a match between tungsten-"white" light and a mixture of 360, 450 and 620 nm; wavelengths defining the dunnart colour space (Figure 1B, and see Table S1 in Supplemental Data). We adopted a two-step process to reduce trial time associated with multiple mixtures. We first determined the spectral locus (position) of tungsten-'white' light within the colour triangle and established intermediate primary wavelengths, which when mixed with the shortest and longest wavelengths, namely 360 and 620 nm, generate a shift towards tungsten-'white' light. The second step consisted of assessing a match between the derived intermediate wavelengths against the remaining two primaries. Our results revealed that 525 and 410 nm would match tungsten-'white' light when mixed with 360 and 620 nm respectively. We subsequently tested the two decisive equations, after training the dunnarts on tungsten-'white' light:(1)a + d = tungsten-'white', b + c = d, then a + b + c = tungsten-'white'(2)e + c = tungsten-'white', a + b = e, then a + b + c = tungsten-'white',where a = 360 nm, b = 450 nm, c = 620 nm, d = 525 nm, e = 410 nm Trichromacy was confirmed by a match between tungsten-'white' and a mixture of 360 and 525 nm (20:80% respectively), and between 525 nm and a mixture of 450 and 620 nm (60:40%), as well as by a match between tungsten-'white' and a mixture of 410 and 620 nm (30:70%) and between 410 nm and a mixture of 360 and 450 nm (80:20%, Table 1). We conclude that the fat-tailed dunnart has functional trichromacy, as suggested by our previous findings [3Arrese C.A. Hart N.S. Thomas N. Beazley L.D. Shand J. Trichromacy in marsupials.Curr. Biol. 2002; 12: 657-660Abstract Full Text Full Text PDF PubMed Scopus (116) Google Scholar], but that it differs from that of primates with the contribution of a UVS cone. As our findings contrast with those for the tammar wallaby, reported to be dichromatic [10Hemmi J.M. Behavioural colour vision in an Australian marsupial, the tammar wallaby (Macropus eugenii).J. Comp. Physiol. A. 1999; 185: 509-515Crossref PubMed Scopus (38) Google Scholar], subsequent behavioural studies with a wider range of species will determine whether trichromacy is the general colour system of Australian marsupials. Download .pdf (.02 MB) Help with pdf files Document S1. Supplemental Experimental Procedures
Referência(s)