Microstructure and defense function of the acoustic organ ofPapilio xuthus (Lepidoptera: Papilionidae) pupae

YANG Shuai, LIU Fu, ZHAO Yu-Fei, SUN Qi,HAO Xiang-Yu1,, YUAN Xiang-Qun

(1. Key Laboratory of Plant Protection Resources and Pest Management, Ministry of Education, College of Plant Protection,Northwest A&F University, Yangling, Shaanxi 712100, China; 2. College of Life Sciences, Northwest A&F University,Yangling, Shaanxi 712100, China)

Abstract: 【Aim】 There are a variety of defense mechanisms used against predators in the animal kingdom. Some lepidopteran pupae can make a wriggling sound when they are mechanically stimulated. 【Methods】 We observed the morphology characteristics of acoustic organs of Papilio xuthus pupae under scanning electron microscope and analyzed the acoustic characteristics of captured sounds using Audacity software. 【Results】The pupae of P. xuthus produce a regular hissing sound by their acoustic organs located in the intersegmental membranes between their 4th-5th and 5th-6th abdominal segments. The sound-producing organs are composed of scrapers and plates, both of which are composed of multiple layers of chitin. There are totally 50-90 protuberances on scrapers and plates. When the abdomen of the pupae is stimulated by the antennae of the parasitoids for more than 30 s, the scrapers and plates will quickly rub against each other repeatedly wiggling from one side to another side of the pupal abdomen to make sounds. The sounds are detected to be composed of a series of short pulses that occur three times every 2 000-3 000 ms. The frequency band is very wide, mainly distributed in the 5-20 kHz range. The activity of fresh pupae is different from that of overwintering ones, resulting in different sound intensities. 【Conclusion】 We firstly described the structure of sound-producing organs of P. xuthus pupae with mechanical stimulation and the results support the hypothesis that the pupae of some butterflies have evolved a special defense mechanism (acoustic defense) against parasitoids. In addition, the sound-producing organs of the same species at different geological areas can generate dialectal phenomena by comparing the characteristics of the acoustic waves of P. xuthus pupae from two different habitats.

Key words: Papilionidae; Papilio xuthus; ultrasound; acoustic defense; morphology; dialect; adaptive evolution

1 INTRODUCTION

In the animal kingdom, sound is a frequent way of transmitting information to its companions or natural enemies. Since the time of Darwin, many investigations have already been conducted on the sounds of nearly the entire class of insects such as Orthoptera, Hemiptera, Coleoptera, and so on (Alvarezetal., 2014). And nowadays it is widely accepted that most lepidopteran insects, including the majority of butterflies, such as nymphalids and lycaenids can also make sounds, which are probably associated with various kinds of defense systems against predators, including birds, bats, invertebrates,etc. Additionally, the sound emission was found to mediate the butterfly’s interaction with ants (DeVries, 1991; Travassos and Pierce, 2000), to promote territorial interaction (Yacketal., 2001; Fletcheretal., 2006; Bowenetal., 2008) and to detect predators (Meyh?feretal., 1997; Castellanos and Barbosa, 2006), and so on. For example, lycaenid larvae can emit sounds to imitate the queen ant, which can enhance the interaction between lycaenid larvae and ants for their protection by ants (Travassos and Pierce, 2000; Dancietal., 2014; Rivaetal., 2017).

For lepidopteran insects, the most studied sound-producing insects are also the adults or larvae of moths. For example, the adult moths ofSaturniapyricommonly emit warning sounds to detect and interfere communications with their predators: the bats (Spangler, 1986; Conner, 1999; Minet and Surlykke, 2003). A common North American silk-moth caterpillar also makes warning sounds by using its mandibles accompanied by chemical defenses (Brownetal., 2007; Buraetal., 2009). In addition, the sounds made by some insects are thought as alarm to the predators, telling the predators that the prey is unprofitable or mimicking another dangerous species (Barber and Conner, 2007).

As for butterflies, more than 200 studies have been conducted in recent decades (Minet and Surlykke, 2003). For example, Hinton (1948) had found that the sound-producing organs of the pupae of butterflies and moths were located between the abdominal segments, but their exact locations were not given in details. Alvarezetal. (2014) and Dolleetal. (2018) studied the pupae of 36 lycaenid species and 35 other butterfly species and found that there existed three types of acoustic organs, and they also gave a brief description of the sounds detected.

In this study, we firstly observed the microstructure ofPapilioxuthuspupae in details, and meanwhile, recorded the different sounds of the pupae produced when mechanically stimulated by the antennae of their natural enemies (parasitic wasps) and by a brush. The photos of pupae ofP.xuthusare shown in Fig. 1.

2 MATERIALS AND METHODS

2.1 Experimental insects

Twenty individuals of overwintering pupae ofP.xuthus, as well as twenty individuals of fresh pupae, were collected from the Entomological Museum of Northwest A&F University, Yangling, Shaanxi Province and Taiyuan, Shanxi Province, China, respectively. The antennae taken from live parasitoid wasps (Vespaducalis, Hymenoptera: Vespoidea), as a stimulation, can keep the pupae active. The more active the pupae are, the easier it is for us to collect the sound. As a comparison, a brush is used to eliminate the bias caused by mechanical stimulation. It is more indicative of the relationship between its sound and defense mechanisms.

2.2 Sound recording and analysis

A bat detector with a sound probe (Batbox Duet HFD6526) was used to record the sounds of butterfly pupae. The frequency range of the detector is set to be 8-120 kHz with the filter frequency set to be 44 kHz. The pupae were placed on a hard table, and the probe was placed on the abdomen of the pupae at a distance of 2 cm between the 4th and 5th, 5th and 6th segments to collect the sound waves. The sound was recorded using audio playback equipment (PCM-D100, SONY) through which we are able to clearly detect the collected ultrasound waves by using a frequency division method. The device can play out the filtered sound for real-time identification and store the sound on an SD card in WAV file format. After the repeated recordings, the basic acoustic characteristics of the pupal sound (including duration, time interval, arrangement of multiple short waves,etc.) were analyzed by Audacity 2.3.1.

2.3 Devices for observing morphology

The morphological structures of the pupae, not only the living pupae, but also the pupae shells were observed using a stereoscope (SteREOS Discovery V20) with an enlargement of 2-8.3-fold. We fixed the position of the acoustic organs and used ZEN 2 professional software to analyze the macroscopical pictures of the acoustic organs. The samples were predisposed by an ultrasonic cleaning apparatus at 100 kHz in a centrifuge tube with anhydrous ethanol for 30 s, and this operation was repeated 5 times. The 4th-5th and 5th-6th segments of the pupae were dissected out and attached onto the aluminum specimen holder with scanning electron microscope (SEM) carbon conductive tape, and then stuck to a flat surface and placed in a vacuum for a few days to reduce sample moisture. The microstructure of the acoustic organs was observed by using the Hitachi S-4800 through a lens of 300-1 800-fold enlargement, with the lengths and widths of acoustic organs being measured with tpsDig2.

3 RESULTS

The activity of the fresh pupae, whose scratching sounds could be heard directly from the pupae at a small distance, was more intense than that of the overwintering pupae. When the abdomen of the pupae is stimulated with the parasitoid’s antennae or brushes, the stomata between the 4th-5th and 5th-6th abdominal segments were detected to emit sound waves; meanwhile, the pupae stroked the pupal shell violently to produce rhythmic squeaking when stimulators reached close enough to the pupae. Under microscope, the sound-producing organs in the membrane between the abdominal segments are invisible in the pupa state. But only when the chest is held tightly with tweezers and the abdomen was bent to the left or right forcefully, we could observe the structure of opposite organs (Fig.2).

Sounds of the pupae are strongly associated with acoustic defense, supported by the different sounds perceived when the pupae are mechanically stimulated by the antennae of their natural enemies (parasitic wasps) and by a brush, respectively. The microstructure of acoustic organs was observed and described in depth. It was confirmed that the microstructure of the pupae also consists of scrapers and plates, and that their relative movement generates ultrasound. Moreover, based on the different characteristics of the acoustic waves of pupae collected from two locations, we hypothesized that dialects have evolved in the two areas.

3.1 Microstructure of organs

The acoustic organs ofP.xuthuspupae are located on the intersegmental membrane between the 4th and 5th, as well as 5th and 6th stomata on both sides of the pupal abdomen. The abdomen of the pupae swings back and forth from left to right to let the sound-producing organs on alternative side be exposed to emit sounds. These four organs are all composed of 2 intersegmental membranes: plate and scraper (the upper section of each intersegmental membrane is fixed as a plate, and the lower section moves as a scraper), whose surface is regularly decorated with transversely striated protuberances. The plate and the scraper totally have 50-90 protuberances, which are 10-11.5 and 8-10 μm in height, and 9.5-9.6 and 8-8.2 μm in width for those between the 4th and 5th segments and the 5th and 6th segments, respectively (Fig. 3). The thickness of the plate and the scraper is 32-38 μm for the 4th and 5th segments and 34-37 μm for the 5th and 6th segments (Fig. 3: A and B). The protuberances of the two differently positioned acoustic organs are remarkably different in height and width, while those between left and right sides at the same location forming a mirror image, are not significantly different.

We found that the acoustic organs are composed of many longitudinal layers of chitin when observing the cross section of the acoustic organs under scanning electron microscope (Fig. 3: C).

3.2 Characteristics of the sounds

The sound is composed of a series of short pulses lasting approximately 15-27 ms. Most of the sound frequencies are between 5-20 kHz (Figs. 4 and 5), with different intensities (Fig. 4). The time intervals between short pulses are irregular and the waveform of each short pulse is composed of 7-15 fluctuations.

The overwintering pupae from Yangling, Shaanxi Province emit 3 short bursts of sound every 2 000-3 000 ms. The sound burst is very short, lasting only 16-20 ms, with the time interval of 707-817 ms. The first and second sound frequencies both range from 6 to 20 kHz, however, the frequency of the third sound is 6 to 11 kHz. In addition, every short pulse waveform consists of 7 fluctuations undulations. The overwintering pupae of Taiyuan, Shanxi Province, also emit 3 short-sharp sounds per 2 000-3 000 ms, each of which lasts 18-27 ms with their intervals ranging from 105 to 1 314 ms; their sound frequencies mainly range from 10 to 15 kHz. In addition, their most short pulses consist of 12-15 fluctuations, only a few of short pulses consist of 7 fluctuations.

4 DISCUSSION AND CONCLUSIONS

4.1 Pupal sounds of P. xuthus

We here report a new phenomenon of sound-producing of pupae inP.xuthus. From a morphological perspective, the pupal acoustic organs ofP.xuthusthat we collected from two regions, are located between the 4th-5th and 5th-6th segments of the abdomen and aligned with the stomata, which is consistent with previous descriptions of the locations in Lepidoptera (Hinton, 1948; Downey, 1966; Dolleetal., 2018). The microscopical morphology of the organ basically accords with the type I (ventro-laterally between abdominal 4th-5th and 5th-6th segments) organ described by Dolleetal. (2018). However, the sound-producing organs they reported to exist between the 6th and 7th abdominal segments in some insects are not detected inP.xuthuspupae of this study. When the samples were prepared for SEM, the comparison between the living pupae and their exuviae showed that there was no obvious structural difference between them. But in practice, when both parts were softened with 75% alcohol, the exuviae were more brittle, easier to break than the living pupae. Dolleetal. (2018) chose to use exuviae as the samples, and thus it is difficult to describe the fine structure of the acoustic organs for their possible transverse sections. Therefore, it is suggested that living pupae used as samples can produce a more detailed morphological characteristic of the acoustic organs.

Analysis of the pupal sounds ofP.xuthus(Figs. 4 and 5) showed that the sounds produced from two different areas were basically the same: short pulses per 2 000-3 000 ms with a main energy distribution of 5-20 kHz. However, there were some minor differences in duration of each short pulse and the time interval between each short pulse. The differential sounds in the same species have been reported previously. For example the phenomenon of the bats with regional calls (Pratetal., 2017). Therefore, we speculate that the difference in sound may also be a dialect. However, we only used 20 pupal individuals from each region, and could not rule out the possibility of small sample size error. In addition, we have not distinguished the sexes of the pupae and cannot rule out the sex differences of the pupal sounds, though Pratetal.(2017) proposed that the gender difference is irrelevant to the sounds in bats.

Compared with the new pupae, the overwintering pupae created significantly weaker sounds. For new and overwintering pupae, the different amplitude of the sound is likely to be caused by different activities. Dolleetal. (2018) proposed that the degree to which each pupa is stimulated cannot be controlled even in the same place at the same time, however, it also cannot be ruled out that the sound amplitude varies with the different degrees of stimulation. Thus the result of this study is that the difference in activity between the two kinds of pupae may contribute to the difference of their sound amplitude.

4.2 Structure of the sound-producing organs

The composition of the insect acoustic organs differs in different life stages. In the larval stage, sound-producing organs only consist of plate and file, while in the pupal stage, acoustic organs usually consist of plate and scraper (Hinton, 1969), which rub against each other to make sounds. Some moths, such asSaturniapyri(Bombycoidea) make this type of sound by the friction between the edge of one mandible and the inner surface of the other (Buraetal., 2009, 2011). Dolleetal. (2018) proposed a new sound mechanism, a stick-slip mechanismsensuAkay (2002), but the sound-producing organs under the stick-slip mechanism emit additional vibrations, which are uncontrollable and unquantifiable (Baumetal., 2014). Our research supports the view of sound friction-production, namely, the two parts of the acoustic organs have regular protrusion strips rubbing against each other to produce the same sounds each time, that is to say, in the absence of sound, the protrusions and the depressions close together, while in making sound, the abdomen begins to pull the internode, allowing their rapid movements of friction to produce a sound with relatively high frequency.

4.3 Function of the pupal sounds

At all stages of their lives, the vulnerability of insects to natural enemies has made a wide variety of adaptive evolutions that protect individuals from different types of predators (Critchlowetal., 2019). Traditionally, adult anti-predator defense mechanisms have been relatively well studied (Lindstedtetal., 2019). For example, moths (Forrestetal., 1995), lacewings (Hoy, 1989), locusts (Moiseff and Hoy, 1983), mantises (Yager and Hoy, 1986), beetles katydids, crickets (Orima) (Rosenetal., 2009) and others can respond quickly to a bat’s location-specific ultrasound using an escape behavior called acoustic startle response (ASR) (Rosenetal., 2009). In contrast, there are few studies on anti-predation strategies at the egg or pupal stage by physical methods. The pupal stage is a necessary stage in the life history of holometabolous insects such as butterflies, whose pupae might be attacked by their natural enemies such as the parasitoid ovipositors. Even though their pupal shells are very hard, the parasitoid ovipositor is able to penetrate to the pupae (Gross, 1993; Dancietal., 2014). In addition, their internode tympanic membrane outside the sound-producing organs is relatively soft, so it is more likely to be the target of the attack. Like theAmorphajuglandislarva (Buraetal., 2011), the velvet ant (Hymenoptera: Mutillidae) (Masters, 1979; Polidorietal., 2013) andXylophanesfalco(Lepidoptera: Sphingidae) (Buraetal., 2016) have acoustic anti-predator devices in order to reduce the attack rate of predators and protect themselves, we also hypothesized that the ultrasound emitted by the pupal intersegmental sound-producing organs of butterflies might be related to defense against parasitoids. Our results of this study suggested that although the pupal movement is relatively weak when we record the sound of the pupae in an intact state, the pupal response becomes more intense when the pupal sound-producing organs were stimulated with bristles, and the pupal movement becomes most intense when stimulated by the antennae of a parasitoid.

4.4 Conclusion

Insect vulnerability to their predators leads to a wide variety of adaptive evolution, protecting themselves from different types of predators, especially the immobile pupae. There are many ways to ward off predators, including secreting toxic chemicals, changing body color, acoustic startle response, and so on. The results of this study showed that the pupae ofP.xuthusstimulated by the antennae of their natural enemy (parasitoid wasps) could produce regular hissing sounds by rubbing their sound-producing organs at the 4th-5th and 5th-6th segments of their abdomen for response. This provides strong evidence that the pupal ultrasounds are strongly associated with the pupal stage’s defense against natural enemies, which has been called acoustic defense. Based on the analysis and comparison of the sound characteristics of the two areas, the pupae ofP.xuthusmay have dialect phenomenon.

ACKNOWLEDGEMENTSWe would like to express our sincere gratitude to Prof. John Richard SCHROCK (Emporia State University, USA) for revising the manuscript and the Entomological Museum of Northwest A&F University for providing insect samples and its staff for their assistance, in particular Dr. HUANG Wei-Jian and Dr. ZHANG Guo-Yun for providing technical support for our use of stereoscope and scanning electron microscope.