Open Access
Add to BibSonomyAbstract
The structural origin of the coloration mechanisms and related extraordinary optical properties of the wing scales of two breeds of Papilio butterflies, namely, Papilio ulysses and Papilio blumei, are explored. The precise ordered biophotonic nanostructures of the wing scales are characterized by scanning electron microscopy (SEM). Despite their structural similarities, the two breeds of Papilio butterflies do not exhibit any analogy in their optical performances. When illuminated with UV-Vis light, P. ulysses gives rise to two reflection peaks: one is from concavities, and the other is from ridges. These two spectral peaks shift their positions under different illumination angles (normal and 45° incident light). In contrast, the spectra for the green scales of P. blumei give one broad reflection peak, and the peak remains the same under normal and 45° incident light. The optical microscopy images indicate that the cap-shaped concavities on P. blumei’s wing scales generate an abnormal bicolor reflection with a strong polarization effect. Both of these two breeds of butterflies take advantage of color mixing strategy: the blue color of P. ulysses is mixed by the colors reflected from concavities and ridges; the green color of P. blumei is produced by the biocolor reflection from concavities. The differences of their coloration mixing mechanisms and optical performances are due to the variations of their nanostructures. The investigation of the color mixing mechanisms of these biologically photonic nanostructures may offer a convenient way for fabricating optical devices based on biomimicry.
©2011 Optical Society of America
1. Introduction
The coloration in the animal kingdom, as seen in birds’ feathers and butterflies’ wings, is often an additive mixture of structural colors [1–4]. The structural coloration is produced by physical interactions of light with biomaterials which have nanostructural variation in the refractive index on the order of the visible wavelength. The structural colors can be divided into two classes: iridescent and non-iridescent. The iridescent color normally is generated by reflection or scattering of light from an ordered array of scatters [5]. On the other hand, the non-iridescent color from the feathers of many birds is produced by quasi-ordered array of air vacuoles in the medullary keratin [6]. Due to its wide applications in photonic crystals, cosmetics, and display technology [7,8], structural color has become the subject of extensive studies recently. During the last two decades, much effort has been devoted to mimicking natural structural color [9–16]. However, to obtain such dedicated structures with structural color as seen in animal kingdom remains to be a big challenge. One way to acquire structural color was to adopt the nature structural color materials as templates to replicate the nanostructures so as to obtain the optical properties [9,10]. To fabricate colloid crystals with photonic band gap (PBG) locating inside the visible part turns out to be another option [11–16]. Although great accomplishment has been made, the structural color produced by the animal kingdom are much richer and more effective than what we can produce so far. Furthermore, it is especially difficult to mimic some unique optical properties of natural structural color, such as polarization, colors mixing etc. We notice that some optical properties and visual ecology of butterflies are correlated with their behaviors, such as mating and aposematic communication [17,18]. Thus, investigating the correlation between the optical properties and the structures may help to explain the innate behaviors of the butterfly kingdom. Investigating the coloration mechanisms of these optical properties and the corresponding structures also has crucial implications for biomimicry, including color-stimulus synthesis, display technologies, various polarization applications [19,20].
Certain breeds of butterflies exhibit typical blue and green structural colors, which are brighter and more deeply saturated than those typically arising from pigments. The structural coloration is known to be produced by the nanostructures in wing scales of these butterflies [21,22]. These structures show some variations based on two different central design principles [23]. The first, named as class 1 or Morpho type, comprises of a multi-layer structure within the discrete ridged structures on the surface of scales that cover the wings. The second, referred to class 2 or Urania type, comprises of continues multi-layers within the body of scales. One particular characteristic of class 2 type is the modulation of the profile of the multi-layer structure, which introduces concave structure into the scale. This specific concave structure can be found in many Papilio butterflies, and gives rise to some particular optical properties, such as polarization and color mixing [24]. Papilio ulysses and Papilio blumei are two breeds of butterflies exhibiting colorful scales with typical properties of structural colors. P. Vukusic et al. discussed the colors mixing and polarization effect of Papilio palinurus butterfly [24,25]. However, some particular optical properties and their structural originations are not fully understood yet. For instance, the structural colors exhibited by Papilio butterflies are generally thought to be iridescent as they are originated from ordered structures. To the best of our knowledge, no further efforts have been made to discuss the effects of Papilio butterflies’ concavity structure on the angular reflectance property of their structural colors. Also, the ridges on the surface of Papilio butterflies’ wing scales are rare discussed.
In this paper, the biophotonic nanostructures and optical properties for two breeds of Papilio butterflies, Papilio ulysses and Papilio blumei, will be investigated. We will explore their coloration mechanisms, especially the correlation between their nanostructures and optical properties (e.g. polarization, colors mixing). The role of the structures of the ridges and the concavities, as well as the profile of the concavities on the colors mixing and polarization effect will be examined. The reflectance spectra under normal and 45° illumination of these two breeds of butterflies will be discussed as well. A deep and comprehensive understanding on this matter will help us not only to study some animal’s behaviors, but also to mimic the structural colors in nature.
2. Materials and Methods
The two breeds of Papilio butterflies, P. ulysses and P. blumei, were bought from Butterfly Park and Insect Kingdom of Singapore.
The reflectance spectra for butterfly wings were measured with an Ocean Optics USB2000 spectrometer attached to a PX-2 pulsed xenon light source (200-850nm). All measurements were taken with R200-7 UV-VIS optic fiber reflection probe. The R200-7 consists of a tight bundle of 7 optical fibers in a stainless ferrule—1 illumination fiber surrounded by 6 read fibers. By placing the probe into a holder with vertical and 45° inclined holes, which was above the sample with a certain distance, the reflected lights in different directions for normal and 45° illumination were recorded and analyzed by OOIBase32 Spectrometer Operating Software.
Bright field and polarized optical images of the colorful scales were taken by a conventional optical microscope (Olympus BX61). The digital images were captured and analyzed by the acquired image processing software (analysis 5).
The surface morphologies, transverse cross section structures and the inter-layer structures of the colorful scales on butterfly wings were characterized by a field emission scanning electron microscope (FE-SEM; JEOL JSM-6700F), previous an Pt sputtering treatment of 5-10nm.
Additionally, a theoretical analysis for the two butterflies was carried out to better understand their coloration mechanisms.
3. Results
3.1 Observation and spectral analysis of Papilio Butterflies
To a human observer, P. ulysses and P. blumei are distinct by their respective bright blue and green coloration (Fig. 1 ). To quantify the structural coloration of the butterflies’ wings, the UV-Visible (λ = 200-850nm) reflectance spectra for the two breeds of butterflies were measured and given in Fig. 2 . In Fig. 2a, the reflectance spectra of P. ulysses for both normal (black curve) and 45° incident light (red curve) are shown. Two separated peaks can be identified from the spectra of the two incident angles. Under normal incident light, a main peak locates at ~550nm and a small peak at ~380nm. However, the positions of the spectral peaks shift under 45° incident light: the main peak shifts to ~350nm and the small peak to ~550nm. These apparent shifts of the spectral peaks indicate that the coloration in the blue scales of P. ulysses is iridescent. And the two distinct spectral peaks of P. ulysses suggest the possibility of color mixing mechanism, which leads to the blue coloration in P. ulysses wings. Unlike P. ulysses, the reflectance spectra of P. blumei for normal incident (black curve) and 45° incident (red curve) light almost overlap with a broad peak ranges from 480nm to 620nm (Fig. 2b).
Fig. 1 Photograph of (a) P. ulysses and (b) P. blumei butterflies showing their blue and green colors.
Fig. 2 Measured reflectance spectra for (a) P. Ullysses and (b) P. blumei, black lines indicate spectra for normal incident light while red lines are spectra for 45° incident light
Here two questions need to be addressed. Firstly, which corresponding parts of the iridescent scales of P. ulysses are responsible for the two distinct spectral peaks? Secondly, why does the spectrum remain the same at different incident angles for P. blumei? In the following session, we will examine the microstructures and the related optical properties of the colorful scales for these two breeds of butterflies to answer these questions.
3.2 Structural characterization and optical properties of Papilio butterflies
The microstructures of colorful wings of P. ulysses and P. blumei were characterized by field emission scanning electron microscope (FE-SEM) (Fig. 3 ). The surface of their wings is composed of millions of scales. The scales of P. ulysses are of a size around 150 × 90 µm2, and consist of a fairly regular array of concavities (Appendix A, Fig. 7 , left). The profile of the concavities is almost flat (Fig. 3a). The ridges run through the full length of the scales with a periodicity of 4-5µm. The details of the period structure of a ridge are displayed in Fig. 3b, and inset of Fig. 4b illustrate the structure schematically. The main ribs are of a thickness (d1) of ~70nm with the inter-distance (D1) of ~70nm. Two adjacent ribs are bound together by a row of smaller sub-ribs with a thickness of ~60nm (d 2) and an inter-distance of ~100nm (D 2). This configuration constructs a 2D array of 70nm × 100nm rectangular air squares surrounded by organic cuticle layers (the main and sub-ribs) with a periodicity of ~140nm (D1 + d1) along its length direction and ~160nm (D2 + d2) along the main ribs. This long-range ordered structure with a very small periodicity can be considered to be a 2D photonic crystal slab tilt about 30° with respect to the surface of the scales [10]. Therefore, the ridges exhibit typical structures of class 1 or Morpho type. The scales of P. blumei have a size of ~180 × 100µm2, and the surface also consists of a fairly regular array of concavities (Fig. 7, right). Different from P. ulysses, the concavities of P. blumei are cap shaped, with 4-6μm in diameter, and their profile is much deeper than P. ulysses (Fig. 3d). The inclined sides of each concavity tilt ~45° with respect to the horizontal surface, and the opposites of each concavity are perpendicular to each other. The ridges run through the full length of each scale with a
Fig. 3 SEM images of the concavities (a), ridges (b) and transverse cross section (c) structures for P. ulysses, SEM images of the concavities (d), ridges (e) and transverse cross section (f) structures for P. blumei.
Fig. 4 Optical microscopy images of the butterflies. (a) Bright field image and (b) taken under crossed polarizers for P. ulysses. (c) Bright field image and (d) taken under crossed polarizers light for P. blumei. Scale bar: 20μm
periodicity of 7-8μm. Compared with that of P. ulysses, the ridges are smaller and have no significant multi-layered structures (Fig. 3e).
To obtain the detailed structure of the scales, the transverse cross section structure was examined for the two breeds of butterflies. The transverse cross section for the scales of P. ulysses consists of 21 alternative cuticle and air layers (Fig. 3c), which share almost the same thickness (~95nm). The section structure for the colorful scales of P. blumei is similar with P. ulysses, except the concavities’ profile and the thickness of each layer (~110nm) (Fig. 3f). Therefore, the concavities for these two breeds of butterflies possess structural properties of class 2 type.
The optical microscopy images of the two breeds of butterflies under normal incident light and linear polarized light were carried out in order to acquire their optical properties (Fig. 4). When illuminated and observed at normal incident light, the concavities in P. ulysses appear to be green and the flat portions between and in the concavities of P. blumei reflect yellow. Furthermore, in P. blumei, the inclined sides around concavities reflect blue light. However, upon crossing an input linear polarizer with an exit analyzer, the green reflected light in P. ulysses and the yellow color in P. blumei almost disappear while the deep purple (near-UV) color reflected by ridges in P. ulysses and the blue color in P. blumei reflect back. This implies that the purple and the blue reflected light are not altered by the polarizers.
As mentioned, there are two peaks in the reflectance spectra of P. ulysses. Therefore, the blue coloration perceived by human eyes is a mixture of these two peaks. Comparing with the optical microscopy image (Fig. 4a and Fig. 4b), under normal incident light, the main peak (550nm) is produced by the green colored concavities and the small peak (380nm) is produced by the deep purple colored ridges. In addition, the reflectance spectra for normal and 45° incident light differ greatly, suggesting the coloration of the colorful scales in P. ulysses is angle-dependent. In contrast, the reflectance spectra of P. blumei are almost the same under normal and 45° illumination light. The green color of P. blumei is produced by mixing the yellow and blue colors reflected from the concavities. In this process, the blue color undergoes a polarization conversion, meaning that a double reflection occurs from the inclined sides of the concavities.
3.3 Coloration mechanisms of the Papilio butterflies
To obtain an in-depth understanding on the color mixing strategies of these two breeds of butterflies, the coloration mechanisms were explored theoretically. Based on the plane-wave expansion method [26,27], the reflectance spectrum by a multi-layer structure can be predicted by the transfer matrix method. The characteristic matrix of an assembly of n layers of films is calculated by multiplying the matrices of each individual film.
where , , and , Nr denotes the refractive index of the film, λ is the wavelength of incident light, is the thickness of the r th film, is the external angle of incidence. The reflectance (R) of the assembly can readily be expressed as,
To find the structural origin of the two characteristic peaks in the measured reflectance spectra for P. ulysses, we first analyze the structure given in Fig. 3. The transverse cross section of the concavities reveals a multi-layer structure composed of 21 alternative cuticle and air layers. These cuticle and air layers share almost the same thickness (~95 nm). The refractive index of the cuticle layer is 1.56 [28]. For the air layer, it consists of disordered cuticle particles with a density (C) of ~20% (Appendix A, Fig. 8a ). Thus, the effective refractive index of the air layer can be acquired based on the density of the cuticle particles: n eff = 1.56 × C + (1-C) ≈1.11.
Fig. 8 (a) The air layer structures for the multi-layered scales, some cuticle particles distribute on the surface of the air layer. Inset is 2D Fast Fourier Transform of the image, which proves that the cuticle particles are randomly distributed. (b) The densities of the cuticle particles on different layers are different, as well as the refractive indexes of the air layers. The value of density and refractive index increases from the first layer to fifth layer, and then decreases from the fifth layer to tenth layer.
The reflection peak produced by the concavities can be calculated based on Eq. (1): ~550nm for normal incident light, and ~350nm for 45° incident light (Appendix A, Fig. 9a ). However, the situation for the ridges is totally different. As described in section 3.2, the ridges can be considered as 2D photonic crystal slabs, and tilt ~30° with the horizontal surface. The normal incident light interacts with the main ribs at 60° (Fig. 5a ), and produces a reflectance peak at ~380nm, while the 45° incident light interacts with the main ribs at 15° (Fig. 5b), and gives rise to a reflection peak at ~550nm. Therefore, the two spectra peaks originating from the concavities and the ridges could be observed for both normal and 45° incident light. Under normal incident light, the main peak is at ~550nm (produced by concavities) and small peak at ~380nm (produced by ridges). Under 45° incident light, the main peak shifts to ~350nm (concavities) and the small peak to ~550nm (ridges). These theoretical results are in excellent agreement with the experimental measurements (Fig. 2a). The spectral peaks of P. ulysses for two illumination cases are summarized in Table 1 .
Fig. 9 Theory calculated reflective spectra for (a) P. ulysses and (b) P. blumei according to the multi-layer structures of their concavities, black lines are spectra for normal incident light and red lines are for 45° incident light. For P. ulysses, the spectral peak reflected by concavities is 550nm under normal incident light and 380nm under 45° incident light. For P. blumei, the spectral peak reflected by the flat portions of concavities is 600nm under normal incident and 450nm under 45° incident light.
Fig. 5 Illustration of the coloration mechanism of the ridge on the scale surface of P. ulysses. (a) The normal incident light interacts with the main ribs at an angle ~60°, and (b) ~15° for the 45° incident light.
Table 1. Illustration for the two reflection peaks of P. ulysses
The cap shaped concavities of P. blumei are also composed of 21 alternative cuticle and air layers with a thickness of ~110nm. For normal incident light, the calculation predicts that the reflection peak resulted from the flat portions locates at 600nm, in agreement with the yellow color observed under optical microscopy. The inclined sides of each concavity tilt ~45° with respect to the horizontal surface, and the opposites of each concavity are perpendicular to each other. As illustrated in Fig. 6a : the normal incident light, reflected from one 45° side, travels across the concavity to the opposite orthogonal side, and then reflects backward in parallel to the original incident direction. The reflection peak originated from the inclined sides of the concavity is 480nm, in accordance with the blue color observed under optical microscope. Through this double reflection process, the blue reflected light undergoes a polarization conversion. As a result, it survives upon the crossed polarizers (Fig. 4d). Under 45° light, the reflection peak arising from the flat portions is 480nm (blue color); while the inclined sides is incident normally by the light, and gives rise to a reflection peak locating at 600nm (yellow color) (Fig. 6b). Therefore, for normal and 45° incident light, the cap shaped concavities both produce yellow and blue colors. These two colors mix up the green coloration caught by human eyes. This gives answer to the question why the reflectance spectra are almost overlapped for these two cases illumination. In Table 2 , we summarize the reflectance peaks for P. blumei in two cases illumination.
Fig. 6 Illustration of the coloration mechanism of P. blumei’s concavities under normal incident light (a) and 45° incident light (b).
Table 2. Illustration for the color mixing mechanism of P. blumei
4. Discussion
Our experiment and calculation analysis reveal that the two breeds of butterflies wings take advantage of the color mixing strategy. The blue color of P. ulysses is mixed by the green and deep purple colors reflected by concavities and ridges respectively. The green color seen from P. blumei is a mixture of yellow and blue colors reflected by the flat portions and inclined sides of concavities. By varying the nanostructures and the profile of their concavities, the two breeds of butterflies exhibit totally different optical properties. Under optical microscope, the flat concavities in P. ulysses’ wings create single color with no polarization effect. The concavities in P. blumei generate a biocolor reflection. Through a double reflection process, the blue color reflected from the inclined sides of P. blumei’s concavities undergoes a polarization conversion. When illuminated with UV-Vis light, P. ulysses gives rise to two reflection peaks. One originates from the concavities, and the other from the ridges. These two peaks shift their positions under different illumination conditions (normal and 45° incident light). On the other hand, P. blumei reflects similar reflectance spectra under these two cases illumination due to the special profile of the concavities.
We notice that many studies show that the eyes of the butterflies have a duplicated gene, allowing them to see ultraviolet colors and distinguish the spectral properties and spatial distribution of the visual colors [29]. And also, they are sensitive to the polarized light. Therefore, the knowledge on the structural origination of the two reflection peaks and polarization property of butterfly wings would have broad biological implications [30,31]. Furthermore, with the understanding of the correlation between the optical properties and the corresponding structures, researchers could be able to find a way to mimic natural structural colors with designated properties. Evidently, the ability to mimic the structural color with its spectacular function will broaden the biomimicry field in the design of structural color materials targeted for ultra and smart performance.
Appendix A: Supporting information
References and links
1. S. Banerjee and Z. Dong, “Optical characterization of iridescent wings of Morpho butterflies using a high accuracy nonstandard finite-difference time-domain algorithm,” Opt. Rev. 14(6), 359–361 (2007). [CrossRef]
2. S. Kinoshita and S. Yoshioka, “Structural colors in nature: the role of regularity and irregularity in the structure,” ChemPhysChem 6(8), 1442–1459 (2005). [CrossRef] [PubMed]
3. J. Zi, X. D. Yu, Y. Z. Li, X. H. Hu, C. Xu, X. J. Wang, X. H. Liu, and R. T. Fu, “Coloration strategies in peacock feathers,” Proc. Natl. Acad. Sci. U.S.A. 100(22), 12576–12578 (2003). [CrossRef] [PubMed]
4. S. Kinoshita, S. Yoshioka, and J. Miyazaki, “Physics of structural colors,” Rep. Prog. Phys. 71(7), 076401 (2008). [CrossRef]
5. S. Berthier, Iridescences: the Physical Colors of Insects (Springer, 2007).
6. H. Noh, S. F. Liew, V. Saranathan, S. G. J. Mochrie, R. O. Prum, E. R. Dufresne, and H. Cao, “How noniridescent colors are generated by quasi-ordered structures of bird feathers,” Adv. Mater. (Deerfield Beach Fla.) 22(26-27), 2871–2880 (2010). [CrossRef]
7. S. S. Choi, S. M. Morris, W. T. S. Huck, and H. J. Coles, “Simultaneous red-green-blue reflection and wavelength tuning from an achiral liquid crystal and a polymer template,” Adv. Mater. (Deerfield Beach Fla.) 22(1), 53–56 (2010). [CrossRef]
8. A. C. Arsenault, D. P. Puzzo, I. Manners, and G. A. Ozin, “Photonic-crystal full-colour displays,” Nat. Photonics 1(8), 468–472 (2007). [CrossRef]
9. Y. Chen, J. J. Gu, S. M. Zhu, T. X. Fan, D. Zhang, and Q. X. Guo, “Iridescent large-area ZrO2 photonic crystals using butterfly as templates,” Appl. Phys. Lett. 94(5), 053901–053903 (2009). [CrossRef]
10. J. Y. Huang, X. D. Wang, and Z. L. Wang, “Controlled replication of butterfly wings for achieving tunable photonic properties,” Nano Lett. 6(10), 2325–2331 (2006). [CrossRef] [PubMed]
11. Y. Liu, R. G. Xie, and X. Y. Liu, “Fine tuning of equilibrium distance of two-dimensional colloidal assembly under an alternating electric field,” Appl. Phys. Lett. 91(6), 063105–063107 (2007). [CrossRef]
12. R. G. Xie and X. Y. Liu, “Electrically directed on-chip reversible patterning of two-dimensional tunable colloidal structures,” Adv. Funct. Mater. 18(5), 802–809 (2008). [CrossRef]
13. R. G. Xie and X. Y. Liu, “Controllable epitaxial crystallization and reversible oriented patterning of two-dimensional colloidal crystals,” J. Am. Chem. Soc. 131(13), 4976–4982 (2009). [CrossRef] [PubMed]
14. K. Q. Zhang and X. Y. Liu, “In situ observation of colloidal monolayer nucleation driven by an alternating electric field,” Nature 429(6993), 739–743 (2004). [CrossRef] [PubMed]
15. K. Q. Zhang and X. Y. Liu, “Two scenarios for colloidal phase transitions,” Phys. Rev. Lett. 96(10), 105701 (2006). [CrossRef] [PubMed]
16. K. Q. Zhang and X. Y. Liu, “Controlled formation of colloidal structures by an alternating electric field and its mechanisms,” J. Chem. Phys. 130(18), 184901 (2009). [CrossRef] [PubMed]
17. P. Pirih, K. Arikawa, and D. G. Stavenga, “An expanded set of photoreceptors in the Eastern Pale Clouded Yellow butterfly, Colias erate,” J. Comp. Physiol. A Neuroethol. Sens. Neural Behav. Physiol. 196(7), 501–517 (2010). [CrossRef] [PubMed]
18. A. Sweeney, C. Jiggins, and S. Johnsen, “Insect communication: polarized light as a butterfly mating signal,” Nature 423(6935), 31–32 (2003). [CrossRef] [PubMed]
19. A. C. Arsenault, H. Miguez, V. Kitaev, G. A. Ozin, and I. Manners, “A polychromic, fast response metallopolymer gel photonic crystal with solvent and redox tunability: a step towards photonic ink (P-Ink),” Adv. Mater. (Deerfield Beach Fla.) 15(6), 503–507 (2003). [CrossRef]
20. T. Labhart, F. Baumann, and G. D. Bernard, “Specialized ommatidia of the polarization-sensitive dorsal rim area in the eye of monarch butterflies have non-functional reflecting tapeta,” Cell Tissue Res. 338(3), 391–400 (2009). [CrossRef] [PubMed]
21. S. Kinoshita, Structural Colors in the Realm of Nature (World Scientific, Singapore, 2008).
22. H. Tabata, K. Kumazawa, M. Funakawa, J. Takimoto, and M. Akimoto, “Microstructures and optical properties of scales of butterfly wings,” Opt. Rev. 3(2), 139–145 (1996). [CrossRef]
23. P. Vukusic and J. R. Sambles, “Optical classification of microstructure in butterfly wing scales,” Photonics Sci. News 6, 61–66 (2000).
24. P. Vukusic, R. Sambles, C. Lawrence, and G. Wakely, “Sculpted-multilayer optical effects in two species of Papilio butterfly,” Appl. Opt. 40(7), 1116–1125 (2001). [CrossRef]
25. P. Vukusic, J. R. Sambles, and C. R. Lawrence, “Structural colour: colour mixing in wing scales of a butterfly,” Nature 404(6777), 457–458 (2000). [CrossRef] [PubMed]
26. H. A. Macleod, Thin-Film Optical Filters (IOP, Bristol, 2001).
27. A. Vasicek, Optics of Thin Films (Interscience New York, 1960).
28. P. Vukusic, J. R. Sambles, C. R. Lawrence, and R. J. Wootton, “Quantified interference and diffraction in single Morpho butterfly scales,” Proc. Biol. Sci. 266(1427), 1403–1411 (1999). [CrossRef]
29. M. L. M. Lim, M. F. Land, and D. Q. Li, “Sex-specific UV and fluorescence signals in jumping spiders,” Science 315(5811), 481 (2007). [CrossRef] [PubMed]
30. A. D. Briscoe, S. M. Bybee, G. D. Bernard, F. R. Yuan, M. P. Sison-Mangus, R. D. Reed, A. D. Warren, J. Llorente-Bousquets, and C. C. Chiao, “Positive selection of a duplicated UV-sensitive visual pigment coincides with wing pigment evolution in Heliconius butterflies,” Proc. Natl. Acad. Sci. U.S.A. 107(8), 3628–3633 (2010). [CrossRef] [PubMed]
31. Y. Takeuchi, K. Arikawa, and M. Kinoshita, “Color discrimination at the spatial resolution limit in a swallowtail butterfly, Papilio xuthus,” J. Exp. Biol. 209(15), 2873–2879 (2006). [CrossRef] [PubMed]
