Colour Selection and Olfactory Responses of Papilio demoleus during Foraging and Courtship

Simple Summary In this work, we analysed the visual and olfactory sensory responses of the butterfly Papilio demoleus during foraging and courtship. We found that P. demoleus has a preference for red and orange. There is a significant difference between the visual and olfactory behaviour of both males and females; males rely more on visual cues, and females respond more to olfactory signals during foraging and courtship. Colour is the dominant factor during foraging and courtship; members of P. demoleus find the target by colour. Butterfly movement is an important role during the courtship of P.demoleus, and odour can increase accurate recognition of males and females. Opsin genes of P. demoleus were analysed, and we found that P. demoleus have genes that recognise the long wavelength, blue and ultraviolet regions of the spectrum. These pieces of evidence are beneficial to understand the coevolve of the butterfly and plant. Abstract Colours and odours are the most important cues for butterflies to localise food and mating partners. We studied the visual and olfactory responses of the widely distributed butterfly Papilio demoleus Linnaeus during foraging and courtship. P. demoleus visited odourless flowers with six colours except green and black, with red as the favourite colour (650–780 nm). Males and females differed in behaviour while visiting flowers. Males were more active than females during foraging. The application of honey water resulted in a significant increase in flower visits by both females and males, and they scarcely visited the apetalous branches with odours. Under natural conditions, four patterns were observed: males chasing males (42.28%), males chasing females (30.56%), females chasing females (13.73%), and females chasing males (13.43%). Males chasing males was the most frequent, probably because males drive away competing con-specific males. When butterflies visited odourless mimics, males chasing females (70.73%) and males chasing males (29.27%) also occurred, indicating that males could accurately distinguish mates using colours only without any chemical cues, and females need chemical cues. The behavioural responses of P. demoleus to floral visits and courtship suggest that colour is the dominant factor during foraging and courtship. We verified the presence of P. demoleus rhodopsin genes, including Rh2, Rh3, Rh4, and Rh5, for long wavelength, blue, and ultraviolet (UV) spectrum recognition, which is consistent with the colour recognition of flowers and wings during visiting flowers and courtship.


Introduction
Flowering plants provide floral information, such as odour, colour, size and shape, that is important to butterfly's foraging and mating [1][2][3][4], and butterflies use floral information to recognise flowers and rewards [5][6][7]. The colour and odour of flowers seem to be the most important cues during foraging. Butterflies use colour cues to find food via colour vision [5,[8][9][10][11][12]. Odours also serve as important cues for butterflies [1,13]. Specific regions in China and has important ornamental and ecological value because it is the main pollinator of some economically important plants [35]. The mechanisms of visual and olfactory functions used by P. demoleus remain unclear. In this study, we observed colour recognition behaviours during foraging and courtship, compared the effects of colour, odour, and the combined signals of colour and odour, and analysed the influence of flight movement on butterfly courtship. We also analysed the visual and olfactory patterns used by P. demoleus, and the types of opsin genes in the compound eye of P. demoleus were detected. Finally, we discuss the ecological significance of visual and olfactory cues during foraging and courtship. These pieces of evidence are beneficial to understand the coevolve of the butterfly and plant.

Experimental Facility
The observations were carried out at the Middle Yunnan Plateau Experimental Station, The Research Institute of Resources Insects, Chinese Academy of Forestry, Kunming Yunnan, China (25 • 3 48 N, 102 • 3 48 E). The elevation of the location is 1808 metres above sea level, with an annual average temperature of 16.2 • C and average rainfall of 930~950 mm. The location has a subtropical plateau monsoon climate. The experiment was carried out in a transparent mesh net house with dimensions of 8 m (long) × 8 m (width) × 4 m (high).

Butterflies
P. demoleus butterflies were maintained in a greenhouse. Butterfly eggs were collected from the host plants of P. demoleus in the Butterfly Garden of the Chinese Academy of Forestry and incubated into larvae in a breeding cage, and larvae were fed into pupae with the leaves of the host plants. When larvae of P. demoleus became pupae, the pupae were placed in a growth chamber for emergence (temperature: 28 • C; humidity: 70%; light period: 12 h/24 h). Three days after the eclosion, the butterflies were used for observation.

Artificial Flowers and Apetalous Branch
Artificial flowers. Artificial flowers were produced from cotton cloth with 8 colours, namely, red, orange, yellow, green, blue, purple, white ( Figure 1A) and black. The flowers had similar corolla dimensions, with a diameter of 8.56 ± 0.18 cm and a depth of 2.73 ± 0.04 cm. The artificial flowers were washed, the spectral reflectance of the corolla of artificial flowers ( Figure 1B) of the flowers was measured using a spectrograph, and the artificial flowers were positioned within the experimental arena, as described in more detail in a previous study [36]. The spectral reflectance of the black flowers was 0.
The apetalous branches. The corolla was removed from the artificial flowers of 8 colours, leaving only apetalous branches ( Figure 1A).

Butterfly Mimics
Butterfly mimics. Artificial butterfly wing mimics were made from plastic materials with colours similar to those of real butterflies. Empty mimics. Colourless plastic mimics of butterflies were also similarly made and used as controls. After ensuring that the wing mimics did not release any volatiles, they were used for subsequent observations of butterfly courtship behaviour.
Odours of females or males. Fresh dead female and male butterflies were immersed in a solution of n-hexane100 µg/µL for more than 12 h to extract insect sex pheromones according to the conventional method [37][38][39]. The extract was filtered through filter paper and then sprayed extract on mimics for observation of the butterflies' chase behaviour. The spectral reflectance of butterfly wings. Spectral reflectance measurements of male and female butterfly wings were performed according to previous research methods [36]. The apetalous branches. The corolla was removed from the artificial flowers of 8 ours, leaving only apetalous branches ( Figure 1A).

Butterfly Mimics
Butterfly mimics. Artificial butterfly wing mimics were made from plastic mater with colours similar to those of real butterflies. Empty mimics. Colourless plastic mim of butterflies were also similarly made and used as controls. After ensuring that the w mimics did not release any volatiles, they were used for subsequent observations of b terfly courtship behaviour.
Odours of females or males. Fresh dead female and male butterflies were immer in a solution of n-hexane100 μg/μL for more than 12 h to extract insect sex pheromo according to the conventional method [37][38][39]. The extract was filtered through fi

Observation of the Behaviour of Butterflies Visiting Flowers
Three days after eclosion, healthy butterflies were collected, and 20 males and 20 females were selected for each experiment per day, using a different butterfly group every day. Each experiment was repeated for 3 days from 8:30 a.m. to 18:30 p.m.
The female and male butterflies labelled with Arabic numerals on the wings were released to the net house to observe the frequency of visits to flowers of different colours. Artificial flowers with either red, orange, yellow, green, blue, purple or white colours were put in a net house. In each assay, 5 flowers of the same colour were bundled together to form a bunch, which was inserted into a transparent plastic bottle. Flowers were placed in a net house at a height of 100 cm. Different bunches of flowers were placed 200 cm apart. The positions of different flowers were rotated every 30 min. Experiment 1. Odourless flowers with 8 colours were placed in a net house. Female and male butterflies (20 males and 20 females) were put in the net house to observe the frequency of visits to flowers with different colours. Experiment 2. Butterfly responses to colour plus honey water during foraging. We used honey water represents the volatile organic chemicals (VOCs) of flowers because it has common VOCs in the nectar of flowers [40,41]. A 10 mL aliquot of 10% honey solution was uniformly sprayed onto each flower bunch every 30 min. Female and male butterflies (20 males and 20 females) were placed in the net house for observation. Experiment 3. The corolla was removed from all artificial flowers, leaving apetalous branches. Female and male butterflies (20 males and 20 females) were placed in the net house for observation, and the responses of female and male butterflies to apetalous branches were recorded. Experiment 4. The corolla was removed from all artificial flowers, leaving apetalous branches. A 10 mL 10% honey solution was uniformly sprayed onto each apetalous branch bunch every 30 min. Female and male butterflies (20 males and 20 females) were put in the net house for observation, and the response of female and male butterflies to the apetalous branches with added honeydew was recorded.

Observation of Courtship with Natural Butterflies and Mimics
Three days after eclosion, healthy butterflies were collected, and 10 males and 10 females were selected for each experiment per day, using different butterfly individuals for the experiment every day. Each experiment was repeated for 10 days.
The female and male butterflies were labelled with Arabic numerals on their wings. Five female wing mimics and 5 male wing mimics were placed into the net house, and 2 adjacent mimics were 100 cm apart. Colourless plastic butterfly mimics (without wings), Danaus chrysippus and D. chrysippus mimics were used as the control for the experiment (each mimic placed 10 cm apart). Experiment 1. Natural population. Twenty sexually mature, unmated adults (10 males and 10 females) were released into a net house for observation. The numbers of chase and courtship behaviour were observed and recorded. Mated butterflies were removed immediately when we observed mating, and the same number of new individuals was added at the same time. Experiment 2. Ten wing mimics of P. demoleus (5 female mimics and 5 male mimics), ten D. chrysippus mimics (5 female mimics and 5 male mimics) and 5 colourless plastic butterfly mimics were placed into a net house. Males or females of P. demoleus (10 males or 10 females) were released into the net house for observation, respectively. Female or male butterflies that chased motionless mimics and motional mimics were recorded. Experiment 3. The materials and methods were the same as in Experiment 2; we used 10% butterfly extract by n-hexane solution and n-hexane solution as control, 2 solutions (10 µL/each solution) sprayed on mimics, sprayed them on each mimic every 30 min. 10 Males and 10 females of P. demoleus were released into the net house for courtship observation, and female and male butterflies chasing odourant mimics were recorded for comparing and analysing butterfly's behaviour response to seek sex partners based on colour and odour information.
Motionless mimics remained immobile. For the motional mimics, the observer pulled the line suspending each mimic every 5 min during the experiment.
Each experiment of butterfly courtship or chase of mimics was performed for 10 days from 8:30 a.m. to 18:30 p.m. A response was recorded when butterflies chased butterflies, and butterflies chased or touched the mimics.

Opsin Cloning
DNA extraction: Genomic DNA was extracted from the retina of the compound eye of P. demoleus by DNAiso Reagent (TaKaRa, Beijing, China) and stored at −20 • C for later use.
Primer design: Referring to the gene sequences of P. xuthus opsins in the NCBI database (Table 1), Primer 5.0 software was used to analyse the gene sequences of opsins and design primers for PCR amplification. The PCR amplification procedure was as follows: predenaturation at 94 • C for 5 min; 35 cycles of 94 • C for 30 s, 55 • C for 30 s, and 72 • C for 10 s; and extension at 72 • C for 10 min.
Agarose gel electrophoresis: PCR products were detected by 4% agarose gel electrophoresis. Cloning sequencing: The MiniBEST Agarose Gel DNA Extraction Kit Ver. 4.0 (TaKaRa) was used to cut and collect the amplification product bands. PCR products were subcloned into the PMD19-T vector (TaKaRa) and transformed into TOP10 competent cells. Positive clones were screened by an LB plate containing ampicillin (Amp+). The M13 primer (upstream primer sequence GTTTTCCCAGTCACGAC; downstream primer sequence CAGGAAACAGCTAT-GAC) was used to identify the recombinant plasmid by liquid bacterial PCR, and the identified plasmid was submitted to Shenggong Bioengineering (Shanghai) Co., Ltd., for sequencing. The DNA sequences of Rh2, Rh3, Rh4 and Rh5 were obtained by splicing the correct sequences, and then blast analysis was performed with the NCBI database to determine sequence accuracy.

Data Analysis
The data were analysed using SPSS 18.0. Fisher's test was used to analyse the difference between visiting different colour flowers between artificial flowers sprayed with a honey solution and odourless artificial flowers, and the difference between the numbers of male and female butterflies chasing motionless mimics and motional mimics.
A 1-way ANOVA was used to test the differences in chasing models between male and female butterflies in the natural population and in mimics during the courtship.
A t-test was used to test the difference between the numbers of butterflies chasing male mimics and female mimics, and the difference between the numbers of butterflies chasing odourant mimics and odourless mimics, the difference in the numbers of flower visits between female and male butterflies.
The spectral reflectance of female and male butterfly wings was measured. A chisquare test (L-R X 2 degrees of freedom ) was used to detect significant differences in the wing reflectance spectra of males and females.

Olfactory and Visual Responses of P. demoleus during Foraging
In the observations with artificial flowers, males and females visited all artificial flowers except for the green and black flowers. The percentages of visits to different flowers by males were 52.59% to red, 17.04% to blue, 14.07% to orange, 6.67% to yellow, 8.89% to purple and 0.74% to white flowers. For females, the visit percentages were 41.67% to red, 29.17% to orange, 25% to blue, and 4.16% to purple flowers ( Figure 2A); P.demoleus did not visit green and black flowers. The number of visits by male butterflies was significantly higher than that of female butterflies (F = 10.434, t = −2.559, df = 54, p < 0.05; Figure 2B).
Fisher's test was used to analyse the difference between visiting different colour flowers between artificial flowers sprayed with a honey solution and odourless artificial flowers, and the difference between the numbers of male and female butterflies chasing motionless mimics and motional mimics.
A 1-way ANOVA was used to test the differences in chasing models between male and female butterflies in the natural population and in mimics during the courtship.
A t-test was used to test the difference between the numbers of butterflies chasing male mimics and female mimics, and the difference between the numbers of butterflies chasing odourant mimics and odourless mimics, the difference in the numbers of flower visits between female and male butterflies.
The spectral reflectance of female and male butterfly wings was measured. A chisquare test (L-R X 2 degrees of freedom) was used to detect significant differences in the wing reflectance spectra of males and females.

Olfactory and Visual Responses of P. demoleus during Foraging
In the observations with artificial flowers, males and females visited all artificial flowers except for the green and black flowers. The percentages of visits to different flowers by males were 52.59% to red, 17.04% to blue, 14.07% to orange, 6.67% to yellow, 8.89% to purple and 0.74% to white flowers. For females, the visit percentages were 41.67% to red, 29.17% to orange, 25% to blue, and 4.16% to purple flowers ( Figure 2A); P.demoleus did not visit green and black flowers. The number of visits by male butterflies was significantly higher than that of female butterflies (F = 10.434, t = −2.559, df = 54, p < 0.05; Figure  2B). The pattern of butterflies visiting artificial flowers sprayed with 10% honey solution was consistent with that of butterflies visiting odourless artificial flowers, and the number of visiting flowers significantly increased (Fisher, p < 0.01; Figure 2C). The number of butterflies visiting flowers with 10% honey solution was increased by 0.73 times, female visits were increased by 1.42 times, and male visits were increased by 0.61 times. The number of male butterflies visiting red flowers was increased by 0.56 times (n = 6, r = 0.98, t = 3.374, p < 0.05), and that visiting orange flowers was increased by 1.21 times. Male butterflies begin to visit the white flowers. The number of female butterflies visiting red flowers in- The pattern of butterflies visiting artificial flowers sprayed with 10% honey solution was consistent with that of butterflies visiting odourless artificial flowers, and the number of visiting flowers significantly increased (Fisher, p < 0.01; Figure 2C). The number of butterflies visiting flowers with 10% honey solution was increased by 0.73 times, female Butterflies rarely visited the apetalous branches, regardless of whether they had been sprayed with the 10% honey solution ( Figure 2D), indicating that P. demoleus mainly used colours for visual sensory recognition during foraging.

Visual and Olfactory Behaviour of P. demoleus during Courtship
Natural population. Among our observations, events of males chasing males accounted for 42.28%, males chasing females accounted for 30.56%, females chasing females accounted for 13.73%, and females chasing males accounted for 13.43% (F = 44.078, df = 3, p < 0.01). Males showed significant differences in their courtship of different sexes, but females showed no significant differences, chasing males and females equally ( Figure 3). During the courtship of natural butterflies of P. demoleus, males always took the initiative, males and females rarely chased interspecific butterflies, and individuals could distinguish between interspecific and intraspecific butterflies.  Behavioural response of butterflies to mimics. During the courtship of butterf displayed towards the mimics, only males chased females (70.73%), males chased ma (29.27%), and females did not chase mimics. Males were able to distinguish females a males by wing colours (F = 3.64, t = −3.327, df = 18, p < 0.01; Figure 3), and males a females did not chase interspecific mimics or blank mimics, indicating that wing colo were the main cues used during P. demoleus courtship.
Behavioural response of butterflies to mimics plus odour. After the extract of males or males and the control solution were sprayed onto the mimics, butterflies har chased mimics that were sprayed with the control solution, only chased mimics t sprayed extract; there are two ways: males chased males, and males chased females. Ma that chased motionless and mobile mimics were significantly increased compared w those that chased the odourless mimics (n = 40, r = 0.857, t = −4.125, p < 0.01). Males chas females increased by 60.34%, and males chasing males increased by 29.17% (F = 13.318 = 3, p < 0.01) during chasing of the motionless mimics, males chasing females increased 39.71%, and males chasing males increased by 39.29% (F = 43.698, df = 3, p < 0.01), and behaviour of males chasing empty mimics appeared (Figure 3). The combination of col and odour provides stronger cues than either alone for butterfly courtship. Courtship behaviour of P. demoleus in a natural population and mimics. Above each bar, we provide the number of chase ±1 standard error. ANOVA test was applied within each group (natural population or mimics). Fisher's test was used to analyse the difference between the number of butterflies chasing motionless mimics and motional mimics. Different letters indicate statistically significant differences between groups (p < 0.05).
Behavioural response of butterflies to mimics. During the courtship of butterflies displayed towards the mimics, only males chased females (70.73%), males chased males (29.27%), and females did not chase mimics. Males were able to distinguish females and males by wing colours (F = 3.64, t = −3.327, df = 18, p < 0.01; Figure 3), and males and females did not chase interspecific mimics or blank mimics, indicating that wing colours were the main cues used during P. demoleus courtship.
Behavioural response of butterflies to mimics plus odour. After the extract of females or males and the control solution were sprayed onto the mimics, butterflies hardly chased mimics that were sprayed with the control solution, only chased mimics that sprayed extract; there are two ways: males chased males, and males chased females. Males that chased motionless and mobile mimics were significantly increased compared with those that chased the odourless mimics (n = 40, r = 0.857, t = −4.125, p < 0.01). Males chasing Insects 2023, 14, 249 9 of 14 females increased by 60.34%, and males chasing males increased by 29.17% (F = 13.318, df = 3, p < 0.01) during chasing of the motionless mimics, males chasing females increased by 39.71%, and males chasing males increased by 39.29% (F = 43.698, df = 3, p < 0.01), and the behaviour of males chasing empty mimics appeared (Figure 3). The combination of colour and odour provides stronger cues than either alone for butterfly courtship.
Behavioural response of butterflies to motional mimics. The number of male butterflies chasing the motional mimics was significantly greater than that chasing the motionless mimics (Fisher, p < 0.05), and the male butterfly response to the odourant-treated mimics was stronger than that to the odourless mimics (Figure 3). The pattern of mimic recognition by female and male butterflies was similar to that observed for natural population courtship; individuals could accurately recognise female and male individuals, and the flight movement of butterflies could enhance courtship.

Colour and Reflectance Spectra of Wings
Under natural sunlight, the wings of P. demoleus showed a range of colours. The overall common background colour accounted for more than 50% of the wing area, and light yellow was a supplementary colour covering approximately 30% of the wing area. The dorsal front wing was mainly black (≥85%), and many yellow dots covered the surface of the base area. A stripe of circular spots covered the edge of the wing surface. There were many irregular patches of different sizes in the central region of the front wings. The dorsal surface of the hind wings was mainly black (≥60%), with yellow-brown patches, blue-purple patches, and red and black circles with blue-purple semicircles all over the leading edge of the hind wing. The ventral surface of the hind wings was mainly light yellow (≥60%), and there were several light-yellow spots at the base. Large red patches were distributed at the end of the inner edge, and the overall patch patterns were similar to those on the dorsal side ( Figure 4A).
The dorsal and ventral sides of the wings of both females and males also showed rich patterns under UV light, indicating that the wings have a strong ability to reflect UV light ( Figure 4B). On the integument of the butterflies, black was mixed with light-yellow colours on the dorsal side, and light-yellow colour was distributed all over the ventral surface. The ventral surface of the integument of butterflies showed strong UV reflectance. Under natural sunlight and UV light, the pattern of wing colour variation of female and male butterfly mimics was similar to that of female and male butterflies.
Spectrum analyses showed that the spectral reflectance of the wings of females and males at 380~780 nm increased as the wavelength gradually increased ( Figure 5). Both female and male wings showed marked differences between the dorsal side and ventral side (Pearson L-R X 2 = 20,575.368 a , df = 1, p < 0.01). The spectral reflectance of the dorsal wing of males (Pearson L-R X 2 = 24.253A, df = 1, p < 0.01) was significantly higher than that of females, and the spectral reflectance of the ventral wing of females (Pearson L-R X 2 = 2.133A, df = 1, p > 0.05) was significantly higher than that of males. The images of females and males were significantly different.

Opsin Gene Analysis of P. demoleus
Five opsin genes of P. xuthus have been found. We detected the opsin genes in our butterflies. Four opsin genes (excluding Rh1) of P. demoleus were detected ( Figure 6). Genes Rh2, Rh3, Rh4 and Rh5 were highly similar to opsin gene sequences in the NCBI database (Table 3). P. demoleus has four opsin genes, Rh2, Rh3, Rh4 and Rh5, indicating that P. demoleus has a wide range of spectrum recognition, can recognise a variety of colours, and can also recognise intraspecific and interspecific individuals by UV reflectance. many irregular patches of different sizes in the central region of the front wings. The dorsal surface of the hind wings was mainly black (≥60%), with yellow-brown patches, bluepurple patches, and red and black circles with blue-purple semicircles all over the leading edge of the hind wing. The ventral surface of the hind wings was mainly light yellow (≥60%), and there were several light-yellow spots at the base. Large red patches were distributed at the end of the inner edge, and the overall patch patterns were similar to those on the dorsal side ( Figure 4A).  The dorsal and ventral sides of the wings of both females and males also showed rich patterns under UV light, indicating that the wings have a strong ability to reflect UV light ( Figure 4B). On the integument of the butterflies, black was mixed with light-yellow colours on the dorsal side, and light-yellow colour was distributed all over the ventral surface. The ventral surface of the integument of butterflies showed strong UV reflectance. Under natural sunlight and UV light, the pattern of wing colour variation of female and male butterfly mimics was similar to that of female and male butterflies.
Spectrum analyses showed that the spectral reflectance of the wings of females and males at 380~780 nm increased as the wavelength gradually increased ( Figure 5). Both female and male wings showed marked differences between the dorsal side and ventral side (Pearson L-R X 2 = 20,575.368 a , df = 1, p < 0.01). The spectral reflectance of the dorsal wing of males (Pearson L-R X 2 = 24.253A, df = 1, p < 0.01) was significantly higher than that of females, and the spectral reflectance of the ventral wing of females (Pearson L-R X 2 = 2.133A, df = 1, p > 0.05) was significantly higher than that of males. The images of females and males were significantly different.

Opsin Gene Analysis of P. demoleus
Five opsin genes of P. xuthus have been found. We detected the opsin genes in our butterflies. Four opsin genes (excluding Rh1) of P. demoleus were detected ( Figure 6). Genes Rh2, Rh3, Rh4 and Rh5 were highly similar to opsin gene sequences in the NCBI database (Table 3). P. demoleus has four opsin genes, Rh2, Rh3, Rh4 and Rh5, indicating that P. demoleus has a wide range of spectrum recognition, can recognise a variety of col-

Opsin Gene Analysis of P. demoleus
Five opsin genes of P. xuthus have been found. We detected the opsin genes in ou butterflies. Four opsin genes (excluding Rh1) of P. demoleus were detected ( Figure 6 Genes Rh2, Rh3, Rh4 and Rh5 were highly similar to opsin gene sequences in the NCB database (Table 3). P. demoleus has four opsin genes, Rh2, Rh3, Rh4 and Rh5, indicatin that P. demoleus has a wide range of spectrum recognition, can recognise a variety of co ours, and can also recognise intraspecific and interspecific individuals by UV reflectance

Discussion
P. demoleus butterflies have significant colour selectivity during foraging. Males preferred visiting red (52.59%), blue (17.04%) and orange flowers (14.07%), and females visited red (41.67%), orange (29.17%) and blue (25%) flowers. These behaviours are highly consistent with the four opsin genes in P. demoleus. When honey water was added to the artificial flowers, the flower visits of P. demoleus increased, indicating that odours can promote butterfly visits to flowers during foraging. In addition, the gustatory stimulus of butterflies could intensify their impression of food upon contact with the honey, causing the number of flower visits to increase. P. demoleus rarely visited apetalous branches, indicating that colour was the main factor affecting foraging. Male butterflies were more active than female butterflies during foraging, which may be because males need more energy during foraging.
During courtship of the natural population of P. demoleus, butterflies exhibited behaviours of males chasing females, males chasing males, females chasing males and females chasing females. The frequency of males chasing males was the highest, the total number of males chasing females was the second highest (≥30), and males took the initiative. During the courtship of butterflies from the natural population by P. demoleus, males always took the initiative. Males chasing males is driving other males and competing for mating rights.
During the courtship of mimics, the frequency of males chasing females (70.73%) was greater than that of the other categories, and males could recognise both males and females according to their wing colours. In natural populations, males chasing males was the most frequent type of behaviour. The combination of colours and male odours may be what causes males to chase. Males chasing males represents intraspecific mating competition, and males chasing males may compete for mating resources. Males will actively reject other males and compete for mating rights during courtship. Males actively reject individuals of the same sex and compete for mating opportunities. The decreased frequency of males chasing males may be due to the lack of odour cues during butterfly chasing of mimics. During butterfly chasing mimics with odours, the frequency of males chasing males dropped dramatically, while the number of males chasing females was higher. Suggested that odour plays an important role in individual recognition during butterfly courtship.
During the courtship of odourless mimics by P. demoleus butterflies, the behaviour of the butterflies was similar to that of C. cyane [26], and males were able to accurately distinguish between females and males. The frequencies of butterflies chasing butterflies and butterflies chasing odourant mimics were significantly higher than that of butterflies chasing odourless mimics, indicating that butterflies can preliminarily recognise females and males using wing colours and then accurately distinguish them by volatile organic compounds. The role of odours is also very important during butterfly courtship, and males can more accurately recognise females through odours than through colours, thus promoting successful courtship. P. demoleus mainly used visual sensory cues during foraging, and olfactory sensory cues promoted accurate recognition. In this experiment, the frequency of males chasing motional mimics was significantly higher than that of males chasing motionless mimics. During courtship in natural butterfly populations, butterflies usually chase males or females in flight, and the number of P. demoleus chasing motional mimics was greater than that chasing motionless mimics.
The courtship behaviour of P. demoleus and previous studies have shown that males actively recognise females and initiate courtship through multiple cues [22,24,31,[42][43][44][45], and wing colours [24,25,37,46] are the main visual cues. The process by which butterflies recognise individuals of the opposite sex is quite complex. Both the colour and odour of butterflies impact courtship and different butterfly species have different visual and olfactory preferences [2,21,28,47]. The spectral reflectances of male and female wings were significantly different, and the colour of the wings also showed brighter patterns under UV light. It is suggested that the spectral reflection and UV colour of butterflies play important roles during courtship.
The Rh2, Rh3, Rh4 and Rh5 genes are expressed in the compound eyes of P. demoleus contain. The Rh2 and Rh3 genes may mainly target receptors of long wavelengths, such as red, orange and yellow. Rh4 recognises the blue spectrum, and Rh5 is associated with receptors of the UV spectrum [9,12,42], which is consistent with the recognition of colours and the UV spectrum of P. demoleus. P. demoleus recognised red, orange, yellow, blue, purple and white flowers during foraging, but P. demoleus did not visit green and black flowers or the green apetalous branches, indicating that P. demoleus does not use green and black as a cue for food. Red, orange, yellow, blue, purple, white, and black are the colours on the wings of each male and female of P. demoleus. Thus, P. demoleus can recognise colours similar to the colours of the wings. The results suggested that the diversity of visual gene expression underlies the recognition of diverse colours in P. demoleus, and the colour recognition ability of P. demoleus may be related to the richness of wing colours.
During evolution, butterflies can find food by the colour of flowers and identify male and female individuals by wing colours. Odours can promote butterfly foraging and mating. Butterflies possess visual genes that function in accordance with their colour-recognition behaviour. Colour may promote visual development and wing colour evolution. Butterflies finding food through flower colour promotes their visual development, and their wings become colourful. Colour vision and wing colour are coevolving and may be the main reasons for the colourful wings of butterflies.
Author Contributions: X.C. designed the study. S.C., M.L., J.L., J.Y. and L.S. performed the experiments; S.C., Y.F. and X.C. analysed the data; S.C., M.L., Y.F. and X.C. wrote the paper. L.S. and J.Y. provided administrative, technical, or material support. All authors have read and agreed to the published version of the manuscript.