The evolution of wing pattern in Lepidoptera is a popular area of inquiry but few studies have examined microlepidoptera, with fewer still focusing on intraspecific variation. The tineid genus Moerarchis Durrant, 1914 includes two species with high intraspecific variation of wing pattern. A subset of the specimens examined here provide, to my knowledge, the first examples of wing patterns that follow both the ‘alternating wing-margin’ and ‘uniform wing-margin’ models in different regions along the costa. These models can also be evaluated along the dorsum of Moerarchis, where a similar transition between the two models can be seen. Fusion of veins is shown not to effect wing pattern, in agreement with previous inferences that the plesiomorphic location of wing veins constrains the development of colour pattern. The significant correlation between wing length and number of wing pattern elements in Moerarchis australasiella shows that wing size can act as a major determinant of wing pattern complexity. Lastly, some M. australasiella specimens have wing patterns that conform entirely to the ‘uniform wing-margin’ model and contain more than six bands, providing new empirical insight into the century-old question of how wing venation constrains wing patterns with seven or more bands.
During recent decades, the longstanding scientific interest in the evolution of lepidopteran wing pattern has experienced a revival . Though the vast majority of studies have focused on butterflies, various contributions have recognized the importance of broader evolutionary context, including the wing patterns of microlepidoptera [2,3]—a paraphyletic grade of early-diverging Lepidoptera often known as ‘tiny brown moths.’ Attempts to designate homologous wing pattern elements between butterflies and microlepidoptera began over a hundred years ago , but this issue remains unresolved—largely because homologies within microlepidoptera are still not fully understood.
One century ago, two studies examined the relationship between wing pattern and wing venation in individual genera and families of microlepidoptera, but did not propose any predictive models that could be tested explicitly in the context of ordinal-level homologies [5,6]. The nymphalid ground plan, a model for wing pattern in butterflies, was proposed soon thereafter, in the 1920s [7,8]. A decade later, in 1935, the first predictive model for wing pattern in microlepidoptera was proposed . This model, now called the ‘vein-fork’ model, predicts that the basal edge of each dark band lies along the points where veins bifurcate; recent studies have found no support for this model [10,11]. A second predictive model for microlepidopteran wing patterns—previously known simply as ‘wing-margin’ model, and called the ‘alternating wing-margin’ model here—was proposed much more recently [12,13]. According to this model, dark and light bands straddle/abut alternating veins along the costal margin of the forewing (figure 1a); two recent studies strongly support this model [10,11].
The ‘vein-fork’ model predicts a primitive wing pattern ground plan with seven dark bands, whereas the ‘alternating wing-margin’ model predicts either five or six dark bands, depending on whether Rs4 terminates along the costa or the termen. Though Lemche’s ‘vein-fork’ model appears to have little predictive power, he was correct in recognizing the need to understand wing patterns with more than six dark bands, as such wing patterns do occur in various taxa [5,15]. A recent examination of the micropterigid genus Sabatinca suggested a mechanism through which patterns with more than six bands could arise from ancestral patterns that follow the ‘wing-margin’ model: dark spots accumulate around all veins at the costa—rather than alternating veins—as seen in derived Sabatinca species such as Sabatinca demissa (figure 1b), and these spots can eventually extend down towards the dorsum, forming bands (figure 1c). This newest model is now called the ‘uniform wing-margin’ model, and the wing patterns of various Psychidae conform to its predictions .
Because so few lineages of microlepidoptera have been examined in this context, many questions remain. The relative prevalence of the ‘alternating wing-margin,’ ‘uniform wing-margin,’ and other potential models are unknown; the transition between the two variants of the ‘wing-margin’ model is still not understood; and the relationship between wing pattern and wing venation along the dorsal margin of the wing has not yet been explored.
Psychidae and Tineidae are the two most early-diverging moth families that contain over 1000 described species , and these two families occupy an intermediate phylogenetic position between the Micropterigidae that have been examined very recently [10,11] and the apoditrysian moths in the family Tortricidae that originally inspired the ‘alternating wing-margin’ model [12,13]. The wing patterns of various Psychidae were described in a recent publication .
The tineid genus Moerarchis contains seven species . Of these, the two Australian species Moerarchis australasiella (Donovan, 1805) and Moerarchis clathrata (Felder and Rogenhofer, 1875) have highly variable wing patterns comprising many spots and bands. Each wing pattern always consists of two colours—very light brown markings against a very dark brown background—but the number, shape and size of the individual pattern elements varies greatly between individuals. The species-level taxonomy of Moerarchis requires extensive further study , but recent systematic work has confirmed that Moerarchis occurs within the ‘tineine lineage’ .
Tineidae are not generally considered to be colourful, but Moerarchis is very charismatic (figure 2) and has been noted for its ‘striking’ wing patterns . Moerarchis australasiella was even featured as adornment on the cover of Tineid Genera of Australia (Lepidoptera)  and is one of three insect species included in the logo of the Australian National Insect Collection. Moerarchis australasiella was also the first tineid species described from Australia . The high intraspecific variation of wing pattern in Moerarchis provides an opportunity to explore the evolution of this character in microlepidoptera.
2. Material and methods
All specimens examined here belong to the species M. clathrata and M. australasiella and are held in the Australian National Insect Collection in Canberra, Australia. All ‘pattern elements’ discussed here are the very light brown markings that occur on the wing. Only forewings were examined because hindwings do not have any colour pattern. All pinned specimens with spread wings were initially examined. Moerarchis clathrata and M. australasiella, and the two sexes within each species, are represented in different proportions in the Australian National Insect Collection, and so different numbers of specimens were examined for each of these groups.
The only specimens that were not examined were those whose right and left forewings are both badly damaged, owing to either wear or breakage. For each specimen that was suitable for examination, one forewing was photographed. If both wings were in equally suitable condition, the right forewing was photographed. Only unique wing patterns were illustrated; if a specimen was found to have the same wing pattern as another specimen that had already been figured, then its wing pattern was not illustrated. One hundred and eighty two specimens were examined and 159 unique wing patterns were illustrated.
To observe the relationship between wing pattern and wing venation, the clearing agent Histolene was applied to individual forewings. Because Histolene leaves a residue that can only be removed by very careful application of acetone, this method may be destructive in some cases and, therefore, could only be carried out on a subset of the specimens examined. Twenty individuals, five representatives of each sex and species, were chosen with the goal of examining the widest possible variety of wing patterns. When Histolene was applied, both the right and left forewings were examined, in order to illustrate variation within individuals.
To produce all illustrations, of wing pattern alone and of pattern and venation together, initial photographs were taken with a Leica DSC 500 microscope/camera and the Leica Application Suite software. Illustrations were created by tracing the light and dark pattern elements in each photo using Affinity Designer graphics software. Each specimen was assigned a unique number that begins with the letter ‘S.’
To explore the relationship of wing pattern with wing length, the wing patterns of each sex and species were divided into two categories (‘joined’ and ‘spotted’ for M. clathrata, and wing patterns with ‘few’ or ‘many’ discrete pattern elements for M. australasiella). Ten specimens representing each category were randomly selected for both sexes of M. clathrata, and the 10 specimens with the most and fewest discrete pattern elements were selected for both sexes of M. australasiella (electronic supplementary material, table S1). One wing of each specimen was measured in its longest dimension. Potential differences in wing length for each pair of wing pattern categories were tested with a Mann–Whitney–Wilcoxon test in R  using the base function wilcox.test.
The wing patterns of M. australasiella consist of common pattern elements that can usually be recognized on wings belonging to different individuals. Twenty-three specimens belonging to M. australasiella have wing patterns that are not unique, and had already been illustrated from other specimens. The wing patterns of M. clathrata were more variable and were assigned to one of two categories: ‘joined’ wing patterns consist of many light spots and bands that are individuated at the costa and/or dorsum, but often become confluent in the interior of the wing, and ‘spotted’ wing patterns primarily contain spots. ‘Joined’ and ‘spotted’ wing patterns often appear very similar, causing the distinction between the two to be unclear in many cases; these categories can therefore be considered end points of a single continuum. All wing patterns of M. clathrata were found to be unique.
The 159 unique wing patterns are as follows: 40 wing patterns from female M. australasiella (electronic supplementary material, figures S1–S4), 61 from male M. australasiella (electronic supplementary material, figures S5–S9), 27 from female M. clathrata (electronic supplementary material, figures S10–S12) and 31 from male M. clathrata (electronic supplementary material, figures S13–S15).
Wing pattern elements are as finely differentiated along the dorsum as they are along the costa. All wings of both species have a large spot of light scales at the base of the wing that nears or reaches the costa and reaches the dorsum. Immediately distal to this large spot is an unbroken light transverse band that reaches the dorsum and either approaches or reaches the costa, with very few exceptions: this band is broken in one specimen (electronic supplementary material, figure S9l) and does not reach the dorsum in two specimens (electronic supplementary material, figures S10a and S15j). In M. clathrata, this band can become partially confluent with other pattern elements. The wings of M. australasiella have between four and eight differentiated patches of light colour (separated by patches of dark colour) along the costa, and between six and 10 along the dorsum, including the spot at the base of the wing. Many pattern elements are transverse bands that extend from the costa to the and dorsum, and many pattern elements that reach the same wing margin become confluent in the interior of the wing.
The differences in wing length between M. clathrata specimens with ‘spotted’ and ‘joined’ colour patterns are not significant; however, M. australasiella specimens with longer wings tend to have significantly more pattern elements on the wing in both sexes (electronic supplementary material, table S2; figure 3). The wings of males tend to be smaller than those of females (figure 3).
3.1. Wing pattern and wing venation
In the species examined here, six veins reach the costa: Sc, R, Rs1, Rs2, Rs3 and Rs4. Rs4 often terminates near the apex of the wing.
In M. australasiella, Rs4 occasionally terminates past the apex (figure 4e) and Rs1+Rs2 can become fused (figure 5d). The relationship between wing pattern and wing venation occasionally follows the ‘uniform wing-margin’ model (figure 4d). However, a light pattern element usually surrounds Rs3 (figures 4a,b,c,e and 5a,d,e). In other specimens, a single light pattern element surrounds Rs3 and Rs4 at the costa (figure 5b), or light pattern elements surround both Rs1 and Rs3 (figure 5c).
In both species examined here, CuP is nearly always straddled by a light pattern element. There are only two exceptions: on the right wing of one specimen, CuP is clearly surrounded by dark scales (figure 4d), and on the left wing of another specimen, CuP is surrounded by a small patch of dark scales that is only noticeable at high magnification (figure 6c). Light wing pattern elements occur between nearly all pairs of adjacent wing veins along the dorsum, as is the case along the costa. In female M. clathrata, CuP is the only vein that is ever straddled by a light pattern element. But in males of both M. clathrata and M. australasiella, CuA1+M3 is usually also straddled by a light pattern element (figures 7a–d and 5b–e, respectively). A light pattern element also straddles M3 in one female M. australasiella specimen (figure 4a).
Along the costa of some specimens belonging to M. clathrata, all veins are surrounded by dark scales and each pair of adjacent veins is separated by a light pattern element, as predicted by the ‘uniform wing-margin’ model (figures 6a,c,d and 7d). In one female M. clathrata specimen examined, the dark pattern element surrounding Rs4 is greatly reduced, such that it appears to be absent if the specimen or illustration is not viewed at high magnification (figure 6e), and in another, Rs3 is surrounded by a broad light pattern element on only one wing (figure 6b). In three of the five male M. clathrata specimens examined, Rs4 is surrounded by a light pattern element on at least one wing (figure 7a–c,e).
3.2. Variation within individuals
A certain amount of variation is always seen between the right and left forewing patterns of the same individual. As noted above, the relationship between pattern and venation can vary slightly, e.g. a vein such as Rs4 can be surrounded by light scales on one wing and dark scales on the other (figure 7b). In addition, the shape of pattern elements can also vary.
In M. australasiella, there is typically little or no variation in the number of pattern elements that reach the wing margin; in one specimen, an apical light pattern element is of small size on the right wing and is absent on the left wing (figure 5c). Wing pattern typically varies through the way in which pattern elements become confluent in the interior of the wing, especially in the case of joined bands (figures 4a,d,e and 5b,d,e). Even when the pattern elements of both wings are identical in terms of confluence, the shape of pattern elements varies noticeably (figures 4b,c and 5a).
In M. clathrata, the number of pattern elements reaching the wing margin can vary along the costa (figure 6d) or, more commonly, along the dorsum (figures 6b–e and 7). In all specimens examined, there is notable variation between the right and left forewings in the shape of wing pattern elements. However, the two forewings of a single individual are always broadly similar, whether comprised mainly of spots, thin joined bands or wide joined bands.
The results presented here provide insight into transitions between the ‘alternating’ and ‘uniform wing-margin’ models, the effect of vein fusion on wing pattern, the relationship between pattern and venation along the dorsum, the importance of wing size, and the number of bands along the wing. Moerarchis demonstrates that highly variable wing patterns can share a common relationship with venation.
4.1. Transitions between models
The ‘uniform wing-margin’ model was proposed following observations of the micropterigid species S. demissa, S. sp. 6 and S. sp. 12 . Because the species-level phylogeny for Sabatinca is characterized by low support values , because wing pattern is highly variable in this genus  and because many species remain undescribed and many more probably remain undiscovered , the transition from the ‘alternating’ to ‘uniform wing-margin’ models in Sabatinca is not understood. Moerarchis clarifies one possible mechanism through which this transition can occur.
In all female M. australasiella examined here, and in two of the five male specimens, wing pattern always appears to follow the ‘uniform wing-margin’ model from Sc to Rs2: each vein is surrounded by dark scales, and light pattern elements reach the costa between each pair of adjacent veins (figures 4 and 5a,b). In male M. australasiella, two specimens lack the light pattern element between Rs1 and Rs2. In the first specimen, this light pattern element is lacking on both wings, and on one wing, Rs1 and Rs2 are fused (figure 5d). In the second specimen, this light pattern element is present on the left wing but absent on the right wing (figure 5e).
Finally, in one male M. australasiella specimen, Rs1 and Rs3 are both straddled by light pattern elements while R, Rs2 and Rs4 are not (figure 5c); the wing pattern follows the ‘uniform wing-margin’ model from the base to the median area, and then follows the ‘alternating wing-margin’ model from the median area to the apex. While this is the only specimen in which a light pattern element straddles Rs1, light pattern elements straddle Rs3 on both wings of all other male M. australasiella specimens (figure 5), on both wings of three female M. australasiella specimens (figure 4b,c,e), and on one wing of an additional female specimen (figure 4a). In the few remaining specimens of female M. australasiella, Rs3 is surrounded by a small dark spot that occurs within a wide light band on one wing (figure 4a) or on both wings (figure 4d).
A similar transition between the two models can be seen on the wings of M. clathrata, but Rs3 is surrounded by a light pattern element on only one wing of a single specimen (figure 6b); Rs4 is straddled by a light pattern element (figure 7a–c,e) or nearly abutted by a light pattern element (figure 7d) in all specimens examined. The areas of the wing that follow the ‘alternating wing-margin’ model in M. australasiella and M. clathrata do so in opposite ways, in the sense that Rs1 and Rs3 can be surrounded by light pattern elements in M. australasiella, whereas Rs4 is far more commonly surrounded by a light pattern element in M. clathrata; in the region surrounding Rs3 and Rs4, what is light in M. australasiella is dark in M. clathrata and vice versa. This ‘flipping’ of light and dark pattern elements between species has also been found in Micropterigidae  and Psychidae .
The generalization can be made that Moerarchis wing patterns tend to follow the ‘uniform wing-margin’ model toward the base of the wing and the ‘alternating wing-margin’ model towards the apex, with the boundary between these two models shifting within the region where the Rs veins terminate along the costa. Because different pattern elements on the very same wing can follow both variants of the ‘wing-margin’ model, these two variants should be conceptualized as endpoints along a single continuum rather than mutually exclusive character states for entire wings or species (figure 8).
4.2. Fusion of veins
Various plesiomorphic lepidopteran wing veins appear to be ‘lost’ in derived lineages. This can occur via one of two mechanisms: fusion—whereby two veins continue to be expressed in the adult wing, but cannot be visually distinguished from each other—or true lack of expression. Because partial fusion of veins has been observed in countless lineages, and because veins that appear to be ‘lost’ are occasionally regained [25–28], fusion may be the most parsimonious explanation for vein ‘loss’ . However, the true prevalence of fusion versus lack of expression is unknown, as is the effect of vein fusion on wing pattern.
In a specimen examined here, Rs1 and Rs2 are partially fused on the left wing and are completely fused on the right wing, with the fused vein terminating along the costa at the position associated with Rs2 (figure 5d). Rs1, therefore, appears to be ‘lost’ on this wing. The two wings have identical patterns along the costa despite the difference in venation. This observation is consistent with previous findings which have indicated that the ancestral location of wing veins determines pattern development, regardless of whether veins are present in a differentiated form in the adult wing [10,11]. The ‘loss’ of a vein via fusion does not affect pattern development in Moerarchis.
4.3. The dorsal wing margin
At present, both variants of the ‘wing-margin’ model have been evaluated only for the costal margin of the wing. This model was originally based on Olethreutinae and is considered to apply to all Tortricidae ; moths in this family have up to five bands that run from the costa to the dorsum, but venation is so reduced that few veins reach the dorsum . In Micropterigidae, which have a relatively complete suite of plesiomorphic lepidopteran veins, confluence of pattern elements along the dorsum obscures any possible relationship between venation and pattern along this margin of the wing [10,11].
But in Moerarchis, wing pattern elements are as finely differentiated along the dorsum as they are along the costa. Along the dorsum, the relationship between pattern and venation is similar to that along the costa, as follows. The ‘uniform wing margin’ model usually holds, with veins typically surrounded by dark scales, and light pattern elements usually occurring between each pair of adjacent veins. One vein is typically straddled by a light pattern element: Rs3 in M. australasiella and Rs4 in M. clathrata, along the costa; CuP along the dorsum. And one additional vein is sometimes straddled by a light pattern element: Rs1 along the costa in M. australasiella (figure 5c), and M3 along the dorsum (figures 4a, 5, 7a–d).
The same veins are straddled by light pattern elements along the dorsum of both species—M3 and CuP—but different veins are straddled by light pattern elements along the costa in the two species—Rs3 in M. australasiella and Rs4 in M. clathrata. This suggests that the relationship between wing pattern and wing venation along the costa is decoupled from that along the dorsum, though far more data are needed to fully address this issue.
Moerarchis is generally considered to have typical tineid wing venation , though one vein branch is not visible on the adult wing of males belonging to M. australasiella and M. clathrata. One branch of the CuA vein is considered to have been lost in many tineid genera, but these are all narrow-winged . In two genera, Monopis and Setomorpha, M3 and CuA1 are stalked, or partially fused . Here it is assumed that M3 and CuA1 are fused in the males of Moerarchis; this assumption is very tenuous, and does not have a major impact on the interpretation of wing patterns.
Unknown vein homologies are another unresolved issue. A four-branched M vein has been known from Agathiphaga queenslandensis for decades . A more recent examination of Agathiphaga vitiensis from Vanuatu found up to 10 M, Cu and A veins reaching the wing margin: 5 M veins, 2 CuA veins, 2 CuP veins and a fused A vein . A branched CuP vein and a fourth or fifth branch of the M vein are unknown outside of Agathiphaga , such that the wing venation ground plan for Lepidoptera is usually reconstructed with a three-branched M vein and an unbranched CuP vein . The three M veins visible on the wings of Moerarchis are referred to here as M1, M2 and M3; this terminology is strictly positional as homologies remain unknown between the three M veins in Moerarchis—and all Lepidoptera other than Agathiphaga—and the five M veins in A. vitiensis. Because plesiomorphic wing veins continue to influence the development of wing pattern even when not expressed in the adult wing [10,11,14], homologies between M veins must be understood in order for findings from taxa such Moerarchis to be placed in broader evolutionary context.
4.4. Wing size
In general, smaller, microlepidopteran moths are considered to have simpler wing patterns than larger moths such as the Papilionoidea, Bombycoidea and Geometridae [31,32]. This observation might be stem from the fact that larger wings have more space for complex patterns, or it could be owing to advanced developmental mechanisms that originated within certain derived lineages of moths. The intricacy of wing patterns in the Hepialidae , a very early-diverged, but large-bodied, family belonging to the Homoneura , demonstrates that complex wing patterns can originate in early-diverged moth lineages, given sufficient wing surface area. However, the relative contributions of phylogeny and wing size to wing pattern complexity have not been tested explicitly. This question is beyond the scope of the present contribution; however, Moerarchis provides preliminary intraspecific data that can inform future studies.
In M. clathrata, the number of differentiated pattern elements—or classification into the ‘joined’ or ‘spotted’ categories—can be deceiving in that these criteria do not necessarily correspond with the number of discrete light and dark patches of colour that reach the wing margin. In M. australasiella, however, the number of discrete pattern elements does correspond to the number of light or dark patches of colour that reach the wing margin, making the characterization of wing pattern far more straightforward. The significant correlation between wing size and number of pattern elements in M. australasiella shows that wing size can underlie wing pattern complexity in the absence of any phylogenetic influence. Of course, number of pattern elements is just one measure of wing pattern complexity. In other moth lineages, additional measures of wing pattern complexity could include the number of colours, the number of types of pattern elements—bands, ripple patterns, etc.—or the complexity of individual pattern elements such as symmetry systems.
4.5. Patterns with more than six bands
Some moths have wing patterns with more than six transverse bands. Lemche’s ‘vein-fork’ model for the relationship between wing pattern and wing venation predicts a primitive ground plan with seven dark bands , but no empirical support has been found for his predictive model, and recent examinations of Micropterigidae have recovered a primitive wing pattern ground plan for Lepidoptera with only five or six bands [10,11].
The ‘uniform wing-margin’ model was originally proposed based on Sabatinca species with small spots surrounding each vein at the costa ; no wing pattern had yet been shown to consist of transverse bands running from the dorsum to the costa, with each band surrounding only one vein at the costa. Various Australian psychids have wing patterns with transverse bands that follow the ‘uniform wing-margin’ model, but fewer than seven bands are present . Certain specimens of M. australasiella figured here have more than six dark bands on the wing; when viewed simultaneously with wing venation, one such specimen provides, to my knowledge, the first empirical confirmation of the ‘uniform wing-margin’ model with more than six bands (figure 4a).
4.6. Future directions
In general, more data are needed in order to determine the relevance of the two ‘wing-margin’ models to various lineages of microlepidoptera; additional studies would allow the analysis of wing pattern in a robust phylogenetic context. This particular study raises two specific issues that should be addressed.
The first issue regards transitions between the ‘alternating’ and ‘uniform wing-margin’ models. The frequency with which a transition can be seen on a single wing is completely unknown outside of Moerarchis. Furthermore, constraints on the directionality of such transitions remain unknown: it is possible, but certainly not confirmed, that wing patterns could follow the ‘alternating wing-margin’ model at the base of the wing, and the ‘uniform wing-margin’ model towards the apex.
The second issue regards wing pattern elements that extend from the costal to the dorsal margin. The present study has found that wing pattern elements can have a uniform relationship with venation along the entirety of the wing margin. Wing pattern in Moerarchis could be described as consisting of individual spots that are associated with only one margin of the wing, which can become confluent with any number of spots along the opposite margin. In other words, in Moerarchis many pattern elements have inconsistent relationships with the costal versus dorsal wing margins. This may not have been the case in ancestral moths, but plesiomorphic wing pattern along the dorsum remains unknown because of the high degree of confluence of pattern elements along the dorsum in the Micropterigidae [10,11]. Now that a consistent venation-pattern relationship has been established along the entirety of the wing margin in certain microlepidoptera, the next step is to determine whether bands of colour on the wings of ancestral moths may have had consistent relationships with particular veins along both the costa and the dorsum.
The wings of Moerarchis figured here address various long-standing questions regarding the evolution and development of microlepidopteran wing pattern. Firstly, Moerarchis demonstrates that a transition can occur between the ‘uniform’ and ‘alternating wing-margin’ models on a single wing: pattern can follow one model towards the base of the wing, and another towards the apex. Second, Moerarchis demonstrates that fusion of veins does not affect the development of wing pattern; when veins fuse, pattern elements remain in alignment with the plesiomorphic locations of wing veins. Third, Moerarchis demonstrates that the relationship between venation and pattern along the costal wing margin can also be observed along the dorsal margin. Fourth, intraspecific variation in wing size is shown to influence the number of discrete pattern elements that occur on the wings of M. australasiella. And lastly, M. australasiella provides strong empirical support for previous speculation that wing patterns with more than six dark bands can develop in accordance with the ‘uniform wing-margin’ model.
This study is based on observations of insect wings, all of which are figured in either the main body or the electronic supplementary material of this paper.
I have no competing interests.
Funding was provided by National Science Foundation Graduate Research Fellowship grant no. DGE-1125191 and by Graduate Research Opportunities Worldwide, a joint initiative of Universities Australia and the National Science Foundation.
This study was made possible by David Rowell, who hosted S.R.S. at the Australian National University, and by Ted Edwards, Marianne Horak, You Ning Su and Andreas Zwick of the Australian National Insect Collection. Richard L. Brown provided feedback at various stages of this project and Maria Heikkilä assisted with discussion of wing venation. Celia L. Curtis painted watercolour illustrations of Moerarchis australasiella. Lauri Kaila and one anonymous reviewer provided insightful feedback on this manuscript.
Electronic supplementary material is available online at https://dx.doi.org/10.6084/m9.figshare.c.3695515.
- Received December 7, 2016.
- Accepted February 2, 2017.
- © 2017 The Authors.
Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited.