Humidity-dependent colour change in the green forester moth, Adscita statices

Abstract

The colours of insects serve important visual functions in aiding mate recognition, camouflage and warning. The display of insects is usually static, as cuticle coloration does not (or hardly) change during the lifespan of a mature imago form. Here, we describe a case of humidity-dependent, brilliant coloration in the green forester moth, Adscita statices. We show, by employing spectroscopic and ultrastructural methods, that the moth's colour results from the interference of incident light with an unusual hydrophilic melanized-chitin multilayer present in the wing scales. Humidity changes in the environment affect the multilayer properties, causing a significant shift of the green-peaking reflectance in the dry state to a rusty colour when damp, resulting in the strong colour change between day and dusk or dawn.

1. Background

Coloration in insects can serve a wide range of functions, including signalling to mates, warning of being poisonous and camouflage [1,2]. Animal colours are generated by two different mechanisms, chemical and physical, or a combination of both, which has attracted the intense interest of scientists and collectors for more than a century [3–5]. A chemical colour results from tissue-containing pigments that absorb in a restricted wavelength range. A physical colour emerges from periodic nanostructures, combining materials with different refractive indices, in insects usually chitin and air [5]. Not unusually, structurally coloured insect scales often contain a pigment that tunes and alters the coloration. For example, certain female lycaenid butterflies feature a brown pigmentary colour, whereas this is overlaid by a blue structural colour in males [6]. Specifically, in birdwing butterflies pigment tunes the wing reflectance spectra, presumably to the spectral sensitivity of the photoreceptors of their mates [7].

Structural coloured media often reflect light very directionally and can thus create highly dynamic signals, as displayed, for instance, by Morpho butterflies [8,9]. Day-flying moths can feature equally bright coloration as their butterfly counterparts. Brilliant examples of colourful displays in moths include the Madagascan sunset moth, Chrysiridia rhipheus (Uraniidae), where a whole rainbow of colours can be seen [10] and many arctiid moths [11–14]. Arctiidae are particularly well studied in regard to their dominant yellow and red (pigmentary) coloration, which has been attributed to an aposematic function [11]. Another clade of colourful moths is the burnet moths or foresters (Zygaenidae). The typically day-flying zygaenids feature metallic sheens (e.g. in the subfamily Procridinae) and/or prominent red or yellow spots (e.g. genus Zygaena) that have been attributed to be aposematic, similar to many tiger moths (Arctiidae) [15]. While the general appearance of many zygaenid moths has been described and the phylogenetic connections been determined [16], a thorough investigation of the coloration mechanisms of these moths is lacking.

A specifically interesting member of the zygaenids is the green forester moth, Adscita statices (Linnaeus, 1758), which can be observed during the daytime, on dry grassland and wet meadows with sorrel (Rumex sp.), its larval food plant, across Europe and northern Asia [16–19]. Like other species of the genus, this species has a green sheen on body and forewings when the sun is out. However, in cool evenings and until the morning dew, this moth features a rusty-red colour (figure 1a–c). This colour change has repeatedly been observed and occasionally been described by lepidopterologists (see for instance [18]), but the morphology of the wing scales has barely been described (but see an image of scales of A. albanica in [16]), and the mechanism by which the colour change happens has remained unestablished. Here, we investigate the coloration of the green forester moth using light and electron microscopy and optical modelling and show that colour change results from a unique combination of pigmentary and structural coloration in an unusual hydrophilic scale structure.

Figure 1.

The colour of the green forester moth, A. statices, differs between the damp and dry state. (a–c) Habitat photographs (taken by Markus Enekvist) of the green forester moth in various states from damp (a) to dry (c). (d,e) Epi-illumination light micrographs of the scale lattice on the dorsal side of the forewing when dry (d) or damp (e). Note the strong colour change of the scales from green in its dry state to rusty-red when damp. (f) Reflectance spectra of the wing scales in the dry (green, solid line) and states of increasing humidity (orange to red, dotted lines).

2. Material and methods

Specimens of A. statices were kindly provided by the Biological Museum of Lund University. We used scanning and transmission electron microscopy to elucidate the wing scale ultrastructure, and light microscopy and microspectrophotometry to characterize the photonic nanostructure and pigment presence in the wings of A. statices. Optical modelling, based on analytic multilayer algorithms and FDTD simulation, was performed to understand the photonic response of the complete structure. For more detailed methods, see the extended Material and Methods section in the electronic supplementary material.

3. Results

Adult A. statices have a wingspan of 2.5–3.0 cm, with antennae, head, thorax, legs and abdomen shiny green; the thorax is occasionally blue-green and iridescent (figure 1; electronic supplementary material, figure S1). The forewings are usually less glossy than the body, and the colour varies between individuals, from blue-green, to green and yellow-green (see electronic supplementary material, figure S1 for a description of the spectral variation). When at rest, the forewings overlay hardly visible, greyish-coloured hindwings.

(a). Colour change

Figure 1a–c shows habitat photographs of a forester moth at different times of the day, resting on a flower stem with its forewings overlapping the body. While the colour of the wings is rusty-red in the early morning (figure 1a), the colour gradually changes (within less than an hour; Markus Enekvist & Tor Dvärv 2010, personal communication) to a vivid green (figure 1b,c). In the evening, the colour changes back to rusty-red.

The colour change can also be observed on the scale lattice level, existing of brightly coloured cover scales and blackish ground scales (figure 1d,e). Figure 1d features the bright green colour normally seen during the daytime. Applying a drop of water creates a pronounced gradient of colours in the cover scales (figure 1e); the black ground scales do not change. While the dry scales (figure 1e, top) are bright green, the cover scales in or close to the water droplet show a deep-red colour (figure 1e, bottom). The scales in-between show a gradient between green and red.

To describe the colour shift quantitatively, we measured reflectance spectra of single scales using a microspectrophotometer. Figure 1f shows reflectance spectra of different scales in the dry and damp state. The dry scales feature a broad reflectance band in the green, with a peak reflectance of approximately 0.4 at 550 nm. This peak shows variations among adjacent scales on a wing and between animals and can range from 510 (cyan-green) to 620 nm (yellow, see electronic supplementary material, figure S1). When damp, the reflectance red-shifts and decreases in intensity, reaching a reduced reflectance that peaks at approximately 700 nm when fully immersed in water.

(b). Ultrastructure of the scales

To identify the structural cause for the colour change, we performed scanning and transmission electron microscopy on wing scales (figure 2). Figure 2a–c show that two distinct types of scales can be identified. The black ground scales follow the typical Bauplan of lepidopteran scales with cross-ribs and ridges [20,21] and a coarse network of irregular trabeculae (figure 2d,f). The coloured cover scales, by contrast, have a more elaborate patterning with a filled lumen that takes the shape of depressions featuring small irregularly sized pores with pore sizes ranging between 50 and 300 nm (figure 2c). Transmission electron micrographs of scale cross-sections show that the cover scales feature a multilayer system composed of alternating layers of porous chitin and air, with occasional trabeculae connecting the layers (figure 2d,e). The multilayer has four to five layers of chitin, with a first (light staining) chitinous layer of approximately 90 nm thickness above three to four (darker staining) chitinous layers of approximately 75–80 nm, each spaced approximately 110 nm apart. By contrast, the ground scales are unstructured (figure 2d,f).

Figure 2.

Ultrastructure of the pigment-containing wing scales. (a–c) Scanning electron micrographs. (a) The scale lattice shows two different types of wing scales. (b) Zoom-in on the different scale types shows a porous cover scale (c) and an open ground scale (g). (c) The cover scale shows a porous top surface. (d–f) Transmission electron micrographs show a multilayered arrangement in the cover scale lumen (e), which is absent in ground scales (f). (g) Simulated humidity-dependent reflectance spectra for a melanin-containing chitin structure. The background refractive index n was varied between air (n = 1, solid green line) and water (n = 1.33, dashed-dotted red line) with various humidity levels in-between. (Online version in colour.)

To investigate whether the staining is due to a pigment present in the scales, we measured the absorbance of single scales immersed in a refractive-index matching oil with the same refractive index of 1.55 (at 546 nm) as lepidopteran cuticular chitin [22]. Indeed, a strong broadband absorbing pigment is found in the coloured scales (electronic supplementary material, figure S2), with a shape that is strongly reminiscent of melanin [23–25].

(c). Changing colours: modelling colour change

To understand the interplay of optical structure with increasing water content, we calculated reflectance spectra of a chitin–air multilayer in the presence of different filling fractions of the air spaces with water using multilayer theory (figure 2g) and FDTD modelling (electronic supplementary material, figure S3). For the multilayer model, we assumed that the chitin layers contain melanin and are porous with a volume fraction of approximately 20%, and the air layers are completely accessible to the environment through the pores. For the dry state, the simulations confirm the green colour of the scales with a broad reflectance band centred at approximately 535 nm with a peak reflectivity of approximately 0.8. For increasing degrees of surrounding humidity, the reflectance peak is reduced and strongly red-shifted reaching 620 nm with a peak reflectance of approximately 0.4 in the fully wet state (figure 2g). Scaling the multilayer structures (electronic supplementary material, figure S4) results in a blue/red shift of the spectra and thus easily explains the colour variation observed between scales on a single wing and across different colour morphs (figure 1d,e; electronic supplementary material, figure S1).

4. Discussion

The wings of the green forester moth are structurally coloured, due to constructive interference of light from melanin-pigmented, porous multilayer reflectors in the scale lumen (figure 2), completely different from the scales found in moths of the subfamily Zygaeninae [16]. Similar forms of this type of scale have been observed in lycaenid [5,26–28] and papilionid butterflies [29]. As in these butterflies, the wing scales of A. statices are melanin-pigmented, likely to suppress stray-light and increase colour brilliance similar to the function of melanin in damselflies [23]. However, the humidity-dependent colour change is highly unusual, as lepidopteran wing scales are well known for their superhydrophobic properties [30,31]. The porous structure investigated here, however, readily takes up water vapour, resulting in the strong, visible colour shift from green to red. Whether the capillary condensation of water vapour into the porous wing scale structure is enough to allow such a water uptake or whether these moths have adapted their chitin composition to take up water remains to be investigated. The water uptake will not only have an effect on the colour but will also lead to a non-negligible increase in wing mass, likely impeding flight abilities of the moth.

Porous photonic structures that can change colours, combining pigmentary coloration beneath a structural coloured layer, have been previously observed in beetles [32,33], butterflies [34–36] and algae [37]. Bioinspired vapour sensors, based on natural structures from butterfly wing scales, have even surpassed the performance of conventional sensors [34].

What is the biological significance of the colour changes? Water uptake leads to a red-coloured moth, which might be advantageous as camouflage in the native habitat of the moths around brown and reddish stems of (figure 1) meadow plants, and specifically on the larval food plant of the species, sorrel. In green vegetation, the green form is likely well camouflaged, but the animals are often sitting on the pink flowers of thistles and scabiouses, where they are easily visible for diurnal predators such as birds, which will learn to avoid the green shine associated with the poison that all Zygaenids contain [38]. In addition, the green colour might simultaneously act as a sexual signal; males of Theresimima ampellophaga, another species of Procridinae indeed use colour in addition to pheromones for mate choice [39]. In a previous investigation of the forester moth, Büchi [17] concluded for the concave reflectors of the wing scales a strong polarization-dependent reflectance due to the difference between the reflectance of the central part and the sides of the concave troughs (see also [29]). Although we found the polarization-effect to be rather minor (not shown), combined with the strong and shiny colour, it might make the moth highly visible to conspecifics, as a potentially visible, yet faint, polarization signal [10,13].

We conclude that the humidity-dependent, showy coloration of forester moths results from the interference of incident light with a melanized-chitin multilayer where absorption of water vapour in a porous structure changes the multilayer's refractive index contrast. Humidity changes in the environment cause a red shift of the reflectance and completely change the colour of the moth at dusk and dawn, making these moths living water vapour sensors—or in other words, allowing a dynamic colour signalling.

Data accessibility

All data used for the analysis can be found in this article and in the electronic supplementary material.

Authors' contributions

B.D.W. designed the study, performed and analysed experiments and wrote the manuscript. K.M. performed and analysed experiments and helped draft the manuscript. A.K. designed the study, performed experiments and wrote the manuscript. All authors gave final approval of the final version and agree to be held accountable for the content therein.

Competing interests

We declare we have no competing interests.

Funding

This research was supported by the Swedish Research Council (2009-5683 to A.K.), the National Centre of Competence in Research ‘Bio-inspired Materials' and the Ambizione programme of the Swiss National Science Foundation (168223 to B.D.W.).