The First Biological Portrait of Stalk-Eyed Fruit Flies: Life History, Reproductive Biology and Host Use Patterns in Pelmatops spp. (Diptera: Tephritidae)

Abstract

The stalk-eyed fruit flies, with their eyes borne at the ends of long stalks, are distinctly different from all other members of the family Tephritidae (Diptera). They resemble stalk-eyed flies (Diptera, Diopsidae) but they are much larger and their antennae are located in the middle of the head instead of on the eye stalks. The stalk-eyed fruit flies are represented by two genera (Pelmatops Enderlein and Pseudopelmatops Shiraki) mainly found in the Oriental tropics and subtropics, but their basic biology remains poorly documented. Here, we describe the life history, reproductive biology, and host use patterns of Pelmatops spp. (mainly P. ichneumoneus (Westwood)). These flies used two local brambles, Rubus setchuenensis and R. multibracteatus (Rosales, Rosaceae), as hosts, with females laying eggs below the epidermal tissue. The larvae bore into the stem, where they feed, eventually dropping to the ground to pupate in the soil. We describe the pupal morphology and eclosion, including the elongation of their eye stalks, feeding, mating, and agonistic behavior in adults. We observed mating between female P. ichneumoneus and male P. tangliangi and tentatively suggest that the two species could be conspecific. Our work presents the first detailed report on the biology of stalk-eyed fruit flies and it lays a significant foundation for future studies on the ecology and evolution of this group.

Some of the most unusual flies in the world are those with secondary paired processes of the head capsule, often developed into elongate structures referred to as antlers or eye stalks. Within the order Diptera, eye stalks occur in eight families, including Diopsidae, Drosophilidae, Micropezidae, Ulidiidae, Periscelididae, Platystomatidae, Richardiidae, and Tephritidae (Wilkinson and Dodson 1997, Vasconcelos et al. 2019a). Flies with these unusual structures have long fascinated biologists, both for the structures themselves and the behavior of the flies (Burkhardt and de la Motte 1983, Grimaldi and Fenster 1989, McAlpine 1994, Vasconcelos et al. 2019b, Constantino 2021).

The subtribe Pelmatopina of tribe Adramini, first proposed by Wang (1996), includes two genera, Pelmatops Enderlein and Pseudopelmatops Shiraki. The group has been called ‘the stalk-eyed fruit flies’ because of their unusual head structures, with the eyes borne at the ends of long stalks (Fig. 1). Pelmatopina is mainly found in the Oriental tropics and subtropics, with eight documented species (3 Pelmatops spp. and 5 Pseudopelmatops spp.). So far, Pelmatops fukienensis Zia & Chen has been found in China (Fujian, Sichuan, Shaanxi, Taiwan) and Burma, while Pelmatops ichneumoneus (Westwood) occurs in China (Sichuan, Yunnan, Xizang, Hainan), India, Nepal, Burma, and Thailand. A third species, Pelmatops tangliangi Chen has been recorded in China (Yunnan, Sichuan), India, and Vietnam (Chen et al. 2015). Both sexes of the first two species have been collected and are well documented, but only three males of P. tangliangi have been recorded, and one female has been recorded although we could not confirm it (Chen et al. 2010, 2015). P. ichneumoneus and P. tangliangi can be distinguished by the length of their eye stalks and compound eye shape, P. tangliangi with longer eye stalks and their compound eye with a sharp projection in the male (Chen et al. 2010, 2015).

Fig. 1.

Pelmatops flies on the leaves of Rubus setchuenensis in the field. (A) Pelmatops ichneumoneus (female); (B) P. ichneumoneus (male); (C) P. tangliangi. (male).

Pelmatops spp. show obvious sexual dimorphism just like the more commonly known stalk-eyed flies (Diopsidae) (Hingle et al. 2001; Cotton et al. 2006; Rogers et al. 2006, 2008; Bath et al. 2015). Basically, the females have much shorter eye stalks than the males. However, detailed studies on the sexual dimorphism of Pelmatops have not been reported until now. Similarly, other aspects of their biology are poorly known, including their mating behavior and host use.

Flies in the family Tephritidae are usually phytophagous, and they are commonly called fruit flies because many of them breed in the fruits or flower-heads of plants or induce galls in various tissues; fewer are leaf-miners, shoot-, stem- or root-borers (Wang 1996, Dohm et al. 2014). Some tephritids are pests of commercially-grown fruit and vegetable crops, thus, many tephritid groups are well-known (Christenson and Foote 1960, Fitt 1984, Drew and Lloyd 1987, Fletcher and Prokopy 1991, Uchôa 2012, Lin et al. 2014, Abduwahap et al. 2017). The biology of stalk-eyed flies (Diopsidae) has also been intensively studied, including their oviposition, larval development and adult emergence (Burkhardt and La Motte 1983), sexual selection (Hingle et al. 2001, Cotton et al. 2006, Bath et al. 2015), eye span length (Rogers et al. 2006, 2008; Small et al. 2009; Vasconcelos et al. 2019a), host plants (Alghali and Osisanya 1982) and aggressive interactions (Egge et al. 2011). However, the biology of the stalk-eyed fruit flies (Tephritidae) remains poorly documented, apart from the basic information regarding collection sites from Chen and Wang (2006) and Chen et al. (2010, 2015). Chen et al. (2015) described the habitat where P. ichneumoneus was collected. The habitat of P. tangliangi in Vietnam is very similar to the site Chen et al. (2015) observed in Yunnan (E. Jendek, pers. comm.).

Here, we report our comprehensive research results on the biology of P. ichneumoneus and P. tangliangi, covering their mating behavior, the development of larvae and pupae and eclosion, as well as host use patterns and various lab observations. This represents the first biological portrait of the stalk-eyed fruit flies and provides a significant foundation for future studies on the ecology, behavior, and evolution of this group.

Materials and Methods

Field Observation

Fieldwork was mainly carried out from 2017 to 2020 at Zhougong Mountain (Ya’an city, Sichuan Province) and Yingjiang County (Yunnan Province) in China, where both P. ichneumoneus and P. tangliangi are found (monthly timing and other details in Table 1). At Zhougong Mountain, our study was carried out along a trail with abundant Rubus setchuenensis Bureau & Franch in an area covered by secondary forests of conifers, shrubs, bamboo, and kiwifruit crops (Fig. 2A and B). In Yingjiang County, near the village Tongbiguan, we observed P. ichneumoneus and P. tangliangi on both sides of a road covered with Rubus multibracteatus Levl. et Vant (Fig. 2C and D).

Table 1.

Occurrence of Pelmatops ichneumoneus and P. tangliangi in the field

	Month
Location	Jan	Feb	Mar	April	May	June	July	Aug	Sep	Oct	Nov	Dec
Zhougong Mountain (2017–2020)	Pupa	X	Pupa	X	None	None	Adult	Adult	Adult Larva	Larva	Larva	Larva
Yingjiang (2019–2020)	X	Larva	X	None	X	None	X	X	X	Adult Larva	Adult Larva	Adult Larva Pupa

None: no adult or larva was observed; X: No field work conducted.

Fig. 2.

The habitats and host plants of Pelmatops in the field. (A) Rubus setchuenensis; (B) The environment of Zhougong Mountain; (C) R. multibracteatus; (D) the environment of Yingjiang County.

Insect Collection and Captive Breeding

To collect adult flies, we first sprinkled sugar water on the leaves of host plants along the roadside and then waited for adults for 1–4 h. The flies were captured with a sweep net or specimen tubes and kept in plastic cups containing a piece of paper dampened with a yeast-sugar solution (yeast:sugar ratio 3 g:3.5 g, added to 220 ml of water). In the laboratory, the flies were transferred to a cage with gauze sides and a host plant. They were also fed daily with a yeast-sugar solution placed on paper. They were incubated in a growth chamber under conditions of 25°C ± 1, 60% ± 10 humidity, and a 9 h–15 h light-dark cycle.

To obtain larvae of P. ichneumoneus and P. tangliangi, we collected branches or whole plants of R. setchuenensis and R. multibracteatus with oviposition scars and ‘air holes,’ then cut the stems open to transfer the larvae into the healthy host plants which had been cultured in the laboratory. When transferring larvae, we used forceps to dig out an artificial larva gallery in the cultivated host plant’s stem, and then transferred the larva into the artificial gallery with a paper funnel under it to prevent the larva from falling. We periodically pinched the stem of the plant to confirm the stage of larval development. If the stem seemed hollow, it meant that the larva had developed into its older stage. Alternatively, we dissected the stem to check the larval development directly. To study the feeding behavior of larvae within the gallery, we dissected galleries from about 50 stems.

For the collection of pupae, we first searched for R. setchuenensis and R. multibracteatus branches with ‘exit holes’ and then examined the nearby soil for pupae by digging around the plant roots, usually within an area covered by a radius of 40–50 cm.

Preservation and Photography of Specimens

All specimens, including pinned adults and preserved larvae, were deposited at the Institute of Zoology, Chinese Academy of Sciences, Beijing, China (IZCAS). Photographs were taken with either an OLYMPUS E-M5 Mark II or a Canon 600D camera. Stacked images were taken using a Nikon D7000 camera mounted on a Nikon SMZ1500 stereo microscope, then generated a single, highly-focused image by using Helicon Focus.

Abbreviations and Terms

The description of male agonistic behaviors follows Egge et al. (2011):

Approach opponent: movement toward opponent; Line up eye stalks: oriented with eye stalks parallel to those of the opponent; Rear up: use mid- and hindlegs to rise in posture; Flex and extend forelegs: bends and flicks forelegs near own eye stalks, directed at the opponent.

Results

Adult Biology

Occurrence, Habitat, Food, and Daily Behaviors

Our study at Zhougong Mountain suggests that there is one generation per year in both species (Table 1). Adults were observed or collected from July to October, and the immature stages were observed from October of the same year through March of the following year. The flies overwintered as pupae and larvae at Zhougong Mountain; however, at Yingjiang, the flies overwintered as adults, larvae, and pupae, which could suggest that there may be more than one generation per year in this relatively more tropical environment.

The flies were sometimes seen on both sides of leaves or branches deep in the bushes of Rubus spp, although they were more often observed on the stem of R. setchuenensis and R. multibracteatus. The adults were attracted to sugar water on the leaves during the daytime from 10:00 to noon and from 16:00 to 19:00 in the field on clear days. During the hottest period of the daytime, they usually stayed on stems deep inside the bush. They were observed in the field to feed on dew or rotten stems and the leaf surface of Rubus. In the laboratory, they would start searching for food at about 8:00 and imbibe the yeast–sugar solution many times a day. After feeding on the sugar water, they would regurgitate a small droplet of the food content on the leaf surface and imbibe it back in a few seconds.

The flies typically performed short-distance flights within 0.5 m but would fly longer distances occasionally in the field. Pelmatops ichneumoneus often brushed the eye stalks from the base to the end using its forelegs, while P. tangliangi used the midlegs to comb the eye stalks and wings. The hindlegs were usually used to comb the abdomen and wings by both species.

Agonistic behaviors were observed between individuals of the same sex in both field and laboratory settings, but we have not observed any males attacking females. When startled by other insects or spiders, the fly would wave its wings, either symmetrically or asymmetrically, and use the forelegs to attack the intruder. In male-male combats (Fig. 3), two males confronted each other in a head-to-head position, with a rapid lateral flicking of wings and up-and-down movement of the abdomen. This was observed twice and one recording lasted 1 min 37s, with the larger male winning (Fig. 3, the left one). The males also struck at their opponent and fought with their fore- and midlegs.

Fig. 3.

A sequence of combat behaviors between two males of Pelmatops tangliangi in captivity. The combat behavior usually included two males approaching each other, lining up their eye stalks, rearing up, flexing and extending the forelegs.

Pre-mating and Mating Behaviors

The pre-mating and mating behaviors of P. ichneumoneus and P. tangliangi were observed both in the field and in the laboratory (Fig. 4). In the field, mating was observed only between a male P. tangliangi and a female P. ichneumoneus. However, in the laboratory, matings between male P. ichneumoneus and female P. ichneumoneus and between male P. tangliangi and female P. ichneumoneus were observed. Mating between male P. tangliangi and female P. tangliangi was not observed either in the field or laboratory (the authors have never personally observed a female P. tangliangi, alive or dead). Courtship typically started with a male pursuing a female for a few minutes.

Fig. 4.

Mating between Pelmatops ichneumoneus females and P. tangliangi males. (A) In the field; (B–D) in the laboratory.

In the field, we observed that male P. ichneumoneus and P. tangliangi often stayed on young shoots or the nearby stems for a long time, apparently waiting until a female arrived. After some courtship behavior, the flies eventually mated on the stem or shoot. During mating, the female often used the foreleg to comb its head (including the eye stalks and the proboscis), and the hindlegs to comb the wings.

In the laboratory, matings were observed between female P. ichneumoneus and male P. tangliangi (Fig. 4B–D). We have recorded three such matings in the laboratory. The mating behaviors lasted at least 56 min (Table 2) and occurred with no apparent time-of-day preference. The male leaped forward several times before it successfully mounted the female, which was imbibing the sugar solution on the paper. The male perched on the back of the female, grasping the female with its fore-tarsi. Their eye stalks were held erect at an obtuse angle. The wings of the male were slightly overlapped on the back, while the wings of the female were placed at the sides of the body. Both the male and female occasionally combed their wings. The female fed during mating and combed its eye stalks and proboscis.

Table 2.

Female Pelmatops ichneumoneus mating time and duration observed in the laboratory

Number	Time			Duration, min
1	9:51–10:04	10:05–10:14	10:16–10:50	56
2	15:10–16:14	16:15–16:36		85
3	13:09–13:50	13:52–14:23		71

Time is not continuous because there are short inactive intervals (<1 min).

Oviposition

Females of P. ichneumoneus were observed laying eggs in young stems or shoots of R. setchuenensis and R. multibracteatus both in the field (Fig. 5A and B) and in the laboratory (Fig. 5C and D). Oviposition in the laboratory typically included the following three stages. (1) Pre-oviposition. The female probed the selected stem or shoot with the ovipositor. (2) Oviposition. The female occasionally flicked its wings slightly and moved its abdomen left and right to insert the ovipositor into the stem. (3) Post-oviposition. The female withdrew the ovipositor, walked away while flicking its wings, and then sought food, typically imbibing the sugar water placed there. Oviposition was observed for two females in the lab. One made five egg-laying scars on the stem, and the other made four. The oviposition behaviors lasted 1–13 min (Table 3) and occurred in the morning or afternoon, with no apparent time preference. Oviposition activity did not necessarily result in eggs being laid, and a female may also lay in a stem which has other egg-laying scars.

Table 3.

Females Pelmatops ichneumoneus oviposition attempts, time, and duration observed in the laboratory

Date	Time	Duration/min
12/24/2019	9:45–9:58	13
	10:01–10:04	3
	10:43	1
	16:17–16:25	8
12/25/2019	16:28–16:34	6
12/30/2019	16:41–16:48	7
1/9/2020	9:46–9:49	3
1/11/2020	12:04–12:07	3

Fig. 5.

Females of Pelmatops ichneumoneus laying eggs on young stems or shoots. (A and B) In the field; (C and D) in the laboratory.

In the field, we observed oviposition behavior twice. At the Zhougong Mountain site, a female was found extending its ovipositor (not touching the stem) on a young stem about 15 cm tall, and after a few minutes inserted its ovipositor into the stem; the oviposition process lasted 8 min. Plants produced fluid when they were injured by oviposition. In the field, we found that newly-laid eggs were usually completely covered by the plant epidermis, and two narrow longitudinal depressions were visible from the outside (Fig. 6A and B). We found egg-laying scars on 30 plants, with 1–6 scars on the same stem. The young stems with egg-laying scars were 0.4–2.0 m tall and typically about 1.5 cm in diameter. As the plants grew, the stem bark sometimes bulged or split at the egg-laying scars (Fig. 6C and D).

Fig. 6.

Egg-laying scars on Rubus setchuenensis in the field (Pelmatops ichneumoneus). (A–B) A newly formed scar in anterior and lateral views; (C–D) an older egg-laying scar.

Immature Biology

Larva and Host Use Pattern

After hatching (Fig. 7), the larvae bored further into the stem to feed. The larval galleries measured from 0.5–5 mm in diameter, increasing as the larvae grew. There were some small holes (‘air holes’) on the stem wall (Fig. 8G) made by the larva, which connected the gallery with the outside and allowed air to flow through. A larger round exit hole in the stem wall was present in some, which was necessary for the larvae to exit and pupate. (Figs. 8C, 9C–D). The appearance of air holes on the apical part of a host plant often indicated the presence of a late stage larva. These larvae could be safely transferred to another healthy host plant and reared until pupation. The mature larva was generally pale yellow with the mesothorax, metathorax, and first abdominal segment a darker yellow (Fig. 10A). Measurements for the last-instar larvae are given in Table 4.

Table 4.

Last instar larval measurement

Number	Length, mm	Width, cm	Species	Sex
1	9.0	1.5	X	X
2	13.0	2.0	P. ichneumoneus	Male
3	13.0	2.0	P. ichneumoneus	Male
4	16.0	3.0	X	X
5	18.0	2.5	P. tangliangi	Male
6	18.0	3.0	P. tangliangi	Male

X: Eclosion failed or other use.

Fig. 7.

A sequence of a Pelmatops larva hatching from an egg laid on Rubus multibracteatus.

Fig. 8.

A schematic representation of the Pelmatops larva gallery in the stem of Rubus setchuenensis. (A) Late-instar larvae gallery; (B) early-instar larva gallery; (C) exit hole; (D) early-instar larva gallery transection; (E) late-instar larvae gallery transection; (F) oviposition site with egg shell; (G) air hole; (H) schematic diagram of marks in the stem.

Fig. 9.

The larva and exit hole of Pelmatops ichneumoneus. (A–B) A larva emerging from an exit hole, as observed in the laboratory. (C–D) The exit hole on R. multibracteatus made by a larva before dropping onto soil to pupate, as observed in the field.

Fig. 10.

Pelmatops ichneumoneus pupation and pupal development process. (A) Larva; (B) a pupa found in soil at Zhougong Mountain; (C) bubble shown in the middle (0 day old); (D) mouth hook appeared (10 day old); (E) π-shaped pattern appeared (25 day old); (F) wings formed (43 day old); (G) black part under the π-shaped pattern, possibly the Malpighian tube (44 day old); (H) appendage (46 day old); (I) the whitish line from which the adult would break out later (48 day old); (J) puparium (49 day old); (K) adult fly (49 day old). (B–K) are from the same individual.

Pupal Development

At Zhougong Mountain, the pupae of Pelmatops spp. were found about 4.5 cm below the ground surface and typically within 30 cm of R. setchuenensis roots. Pupae averaged 10.81 mm long (9.0–13.5 mm) and 3.53 mm wide (2.5–5.0 mm) (Table 5), oriented vertically or obliquely (Fig. 10B) in the soil.

Table 5.

Size of Pelmatops pupae

					Mean value
Number	Species	Length/mm	Width/mm	Length-Width ratio	Length/mm	Width/mm	Eclosion
1	P. tangliangi	11.0	3.5	3.14	Sex	3.75	Succeeded	Male
2	P. tangliangi	12.0	4.0	3.00			Succeeded	Male
3	P. tangliangi	12.0	4.0	3.00			Succeeded	Male
4	P. tangliangi	12.0	3.5	3.43			Succeeded	Male
5	P. tangliangi	12.0	3.5	3.43			Succeeded	Male
6	P. tangliangi	13.5	4.0	3.38			Succeeded	Male
7	P. ichneumoneus	9.0	3.0	3.00	10.38	3.62	Succeeded	Female
8	P. ichneumoneus	10.5	3.5	3.00			Succeeded	Female
9	P. ichneumoneus	11.0	3.0	3.67			Succeeded	Female
10	P. ichneumoneus	11.0	5.0	2.20			Succeeded	Female
11	P. ichneumoneus	9.0	2.5	3.60	10.26	3.25	Succeeded	Male
12	P. ichneumoneus	9.0	2.5	3.60			Succeeded	Male
13	P. ichneumoneus	9.0	3.0	3.00			Succeeded	Male
14	P. ichneumoneus	9.0	3.0	3.00			Succeeded	Male
15	P. ichneumoneus	10.0	3.0	3.33			Succeeded	Male
16	P. ichneumoneus	11.0	4.0	2.75			Succeeded	Male
17	P. ichneumoneus	12.0	4.0	3.00			Succeeded	Male
18	P. ichneumoneus	13.1	4.0	3.28			Succeeded	Male
19	Pelmatops sp.	9.5	3.0	3.17	X	X	Failed	X
20	Pelmatops sp.	9.5	3.2	2.97	X	X	Failed	X
21	Pelmatops sp.	11.0	4.0	2.75	X	X	Failed	X
22	Pelmatops sp.	11.0	4.0	2.75	X	X	Failed	X
23	Pelmatops sp.	11.5	4.0	2.88	X	X	Failed	X
	Mean value	10.81	3.53	3.10

Pupal development progressed as follows, with the day that the pupa of P. ichneumoneus was collected at Zhougong Mountain being day 0. On the pupa, a remnant of the larval mouth hook was apparent (Fig. 10D) at 10 days old, and a special π-shaped pattern was present (Fig. 10E) at 25 days old. The wing (Fig. 10F) first appeared at 43 days old and the Malpighian tubules appeared the next day (44 days; Fig. 10G, also the π-shaped pattern is visible in the lower section). The pupal color had gradually darkened during the process of pupation and finally turned from yellow to orange-brown (Fig. 10). The adult emerged at the longitudinal white line (Fig. 10I) at 48 days old.

Although the pupal development of P. ichneumoneus was recorded, its emergence and eye stalk elongation were not photographed due to technical difficulties.

Eclosion

The eclosion of P. tangliangi was observed in the laboratory from a larva collected at Zhougong Mountain. Before emergence, a process of separating the appendages from the pupal shell (Fig. 11A–C) was recorded. The imago eclosed 27 min after the appendages separated from the puparium (Fig. 11C and D).

Fig. 11.

The pre-eclosion and eclosion process of a Pelmatops tangliangi male (same individual as Fig. 12). (A) The pupa 3 days before eclosion, with appendages clear; (B) the pupa on the morning of emergence day, with appendages clearly seen; (C) the pupa at the emergence time, with appendages separated from the pupal shell; (D) the imago broke out of its pupal shell (27 min after appendages separated); (E–I) the eclosion process, which took ~1 min.

The imago broke out of its pupal shell (Fig. 11D) using its ptilinum to push through the operculum. First, the ptilinum and antennae were exposed, and then the compound eyes appeared (Fig. 11E). The adult initially emerged from the pupal shell via the bulging ptilinum, until the forelegs touched the ground and could be used to pull the rest of the body out (Fig. 11E–K).

It took 2 min from the opening of the operculum until the adult fully climbed out of the pupal shell, leaving behind a membranous layer. The adult walked away quickly when it emerged from the pupal shell and the ptilinum collapsed. The whole body of the newly-emerged fly was pale yellow to orange in color, and the eye stalks were very short, about 0.3 mm long, and dark brown in color (Fig. 12A). Newly-emerged adults with eyes close to the head resembled diopsids (Burkhardt and La Motte 1983). Wings were wrinkled together and unexpanded; the costal vein and apical area of the wing were black and gradually became light to orange at the base. At this stage, the sex of the adult could be determined from the abdomen. The imago kept walking around in the plastic cup. When it stopped at a certain place, it started to rub its head (including proboscis, antenna, ptilinum, and eye stalk) with its forelegs, and the rubbing of the ptilinum was most frequent.

Fig. 12.

Eye stalks elongation in a Pelmatops tangliangi male (same individual as Fig. 11). (A) Newly Newly-emerged adult with eyes close to the head (16 min-old); (B–G) The gradual elongation of eye stalks (B: 23 min; C: 24 min; D: 25 min; E: 29 min; F: 31 min; G: 50 min). (H) The eye stalks sclerotized and darkened (56 min); I: The fly being active (the next day).

Sixteen minutes after crawling out of the pupal shell, the fly became less active. While rubbing the foreleg and the eye stalks, the latter began to elongate (Fig. 12B), and the body undulated slightly. At first, the eye stalk was curved or irregular in shape and asymmetric, but at the final stage, the eye stalk straightened out in the horizontal direction (Fig. 12B–G). Since the body color had not darkened, the eye stalk was translucent, leaving the movement of body fluids visible. During the elongation process, the thread-like nerves in the eye stalk could be seen clearly.

When the eye stalk was straightened and fixed, the adult began to extend its wings. After the wings were unfolded, they stopped at one place and barely moved as the eye stalks began to darken. During this process, the body became sclerotized and also darkened. Yellow liquid excrement mixed with black was also produced under the abdomen. Two hours and 7 min after exiting the puparium, the fly started to walk around with its permanent body color.

Discussion

Mating and Oviposition

Our observations give potential insights into the underlying bases of these behaviors. Before courtship, Pelmatops males tended to vibrate their wings and wait on young stems or shoots for a female to come; this may be a type of resource defense for oviposition sites, similar to Phytalmia (Tephritidae), where males search the trunk surface of downed trees for limited oviposition sites (Dodson 1997). Several males may be present on the same stem, exhibiting rendezvous behavior (Zwölfer 1974) or lek behavior. In many tephritids, precopulatory behavior includes intraspecific communication among individuals via visual, chemical, and acoustic signals (Díaz-Fleischer and Aluja 1999). Headrick and Goeden (1994) described 48 species of native California fruit flies’ reproductive behaviors, and found male courtship displays included at least wing displays and one or more of the following: abdominal pleural distension, appendage displays (other than the wings), swaying, or trophallaxis. The tendency of males to vibrate their wings during courtship may serve as visual or acoustic communication between the male and female, and the continuous raising of the abdomen may indicate the release of sex pheromones. This would be similar to many other tephritids such as Rioxa sexmaculata (van der Wulp), the male of which attract females by using pheromonal calling (Headrick and Goeden 1994, 1996; Kovac et al. 2010). Some studies show trophallaxis is an important part of the courtship and female choice both in non-frugivorous (Freidberg 1981) and frugivorous species (Aluja et al. 1993, Benelli and Romano 2018), but there is no such behavior recorded in male Pelmatops. In tephritid flies, copulation duration varies across different species. It lasts only about five minutes in Rioxa sexmaculata (Kovac et al. 2010) and many other tephritids, but for Euaresta stigmatica Coquillett it can last for an hour (Headrick et al. 1995), and the longest average copulation duration found in Tephritidae was 24 hours in Euarestoides acutangulus (Thomson) (Headrick and Goeden 1999). As such, although longer than some other species, copulation in this group cannot be considered unusually long for the Tephritidae. It should also be noted that, although we did not take detailed records of these behaviors, we observed males mating with both virgin flies and flies that had already mated with other males.

The males exhibited territorial behaviors as they waited for females on young stems or shoots. Consequentially, the same stem sometimes attracted multiple males, but only the winners stayed according to our observations. We speculate that both females and males were attracted by chemicals emitted from the young stems or shoots of the host plants. The prolonged stay on host plants by Pelmatops males may characterize a ‘rendezvous-behavior’, where the males occupy the larval host plants for courtship and mating, as is known for more than 45 species of tephritids (Zwölfer 1974). Males of P. tangliangi and female P. ichneumoneus mated in both field and laboratory settings, thus, these two species may conspecific. It may be that differences in male eye stalk length and the shape of the compound eyes between these two species simply represent different size morphs, but more evidence from morphological, molecular, and behavioral studies is needed to test this hypothesis.

In Phytalmia, males guard ovipositing females at the oviposition site (Moulds 1977), but this is not the case with Pelmatops males. We observed that males had a strong tendency to stay near stems where females had not already oviposited. However, similar to Phytalmia, P. ichneumoneus, females would sometimes return to and oviposit further in a site that they previously utilized or those other females utilized (Moulds 1977).

Other Behaviors in Adults

The agonistic behaviors of Pelmatops we observed were somewhat similar to those of Teleopsis dalmanni (Wiedemann) (Diopsidae) which possesses elongate eye stalks, and Phytalmia mouldsi McAlpine & Schneider (Tephritidae), the males of which possess lateral projections from the head resembling antlers (Moulds 1977, Dodson 1997). Low-intensity behaviors included approaching the opponent, lining up eye stalks, rearing up, flexing, and extending their forelegs (Egge et al. 2011). These behaviors were occasionally observed when flies were reared in the same cage and when on the same shoot of the host plant in the field. Agonistic behavior between P. ichneumoneus and P. tangliangi males was uncommon overall. But in the P. tangliangi male combat (period after Fig. 3), the larger individual (the male on the left) won and no apparent injuries were incurred. This is similar to combat in Phytalmia mouldsi, in that it appears to be purely a display of strength (Dodson 1997).

We also observed flies regurgitating after drinking sugar solution, although regurgitation was rarely observed in the field. Drew and Romig (1999) stated that protein in the female fruit fly diet is essential to reproduction, and the main candidate sources of protein and amino acid in nature are honeydew, extrafloral plant exudates, nectar, pollen, bird feces, and bacteria. We speculate that Pelmatops feeding on leaves of the host plant in the field and laboratory may be ingesting bacteria growing on the plant surface leachates and extrafloral exudate, such that re-ingesting food deposited on plant surfaces would pick up some portion of these nutrients.

Immature Biology

The young stems pierced by the female ovipositor produced drops of fluid on the egg-laying scars that might help to moisturize the eggs and thus be beneficial for the egg’s hatching, although, alternatively, it could also simply be plant defensive compounds. Although host plants may contain one to six egg-laying scars on the same stem, only one larva survived to maturity in each plant in almost all cases (about 70 stems). More research is required to determine the reason why.

In the laboratory, the large round hole in the stem containing a larva proved to be an exit hole (Fig. 9B). One plant that was transferred from the field may have lacked water for some time, causing the lignification and hardening of the stem and probably making the exit hole less flexible. Consequently, the larva in that stem got stuck while trying to crawl out to pupate (Fig. 9A and B). This behavior contrasts with the majority of flower head feeding Tephritinae and all gall-inducing Tephritinae that stay in the host until the adult emerges from its pupa, such as Paracantha gentilis Hering, Acinia reticulata Aczél, and Dracontomyia footei Aczél (Headrick et al. 1990, Norrbom et al. 2013). Another example is Ichneumonopsis burmensis Hardy (Tephritidae: Dacinae: Gastrozonini), the larvae of which develop in bamboo shoots. The larvae of Ichneumonopsis burmensis also produce exit holes when they become mature, but they then remain within to pupate in the bamboo and wait to emerge as adults (Kovac et al. 2013). It remains unclear why these different strategies are used, but in some environments the plant host may be safer for pupae.

In the field, it was occasionally observed that older larvae remained in stems with exit holes. We assume that the larvae had chewed out a small visible exit hole in advance and waited for suitable conditions to escape (such as good weather). However, when an older larva was transferred to a young stem in the laboratory, the larva would use an artificial hole to exit to pupate without chewing an exit hole. More research is needed to explore why this happens, but it could simply be an energy-saving strategy.

Eclosion

The mechanisms by which the eye stalks develop are of great interest. Buschbeck et al. (2001) proposed that the cuticular folding of the eye stalk and the coiling of the optic nerve prepare the pupa for the rapid and dramatic eye-stalk inflation after eclosion in the stalk-eyed fly Teleopsis whitei (Curran). The imago of Pelmatops likely uses a pumping process to inflate the eye stalks to their full length during eclosion, as seen in Diopsidae (Burkhardt and La Motte 1983). Further research is needed to confirm this for the stalk-eyed fruit fly, although our observations of what appears to be rapid fluid movement in the eye stalks seem supportive of this. Interestingly, the males of P. tangliangi initially elongate their eye stalks mostly in irregular directions, unlike Diopsidae, which elongate their eye stalk more linearly (Buschbeck et al. 2001). However, the male and female of P. ichneumoneus elongate their eye stalks linearly in a more similar fashion to the Diopsidae.

The environmental conditions for successful emergence seem quite stringent. In one case, an adult Pelmatops failed to eclose successfully in the laboratory (Fig. 13). Four hours after the adult broke the pupal shell, its body was still in the pupal shell. We discovered that the membrane inside the pupa shell was stuck to the fly (Fig. 13G and H). It seems likely that insufficient humidity contributed to the failure of this eclosion. When we increased the humidity by adding wet cotton to the plastic box for other pupae, adult P. tangliangi (Fig. 12I) subsequently emerged successfully.

Fig. 13.

A case of eclosion failure eclosion in a Pelmatops fly. (A) 0 min; (B) 2 min; (C) 3 min; (D) 3 h 20 min; (E) 4 h 33 min; (F) 4 h 48 min; (G) 7 h 13 min; (H) 7 h 34 min; (I) 8h 3 min; (A–D) the imago broke the pupal shell; (E–G) the imago stuck to the membrane inside the pupa; (H–I) the imago crawled on the soil.

Conclusions and Future Work

This is the first in-depth study on the life history of stalk-eyed fruit flies. Our research lays an important foundation for future work on this rare group of fruit flies. Much remains to be studied, including aspects of sexual selection and related evolutionary selection pressures, as well as the underlying developmental pathways and genetic mechanisms involved in eye stalk development. We hope that our work will spur more research on these rare and unique fruit flies.

Acknowledgments

This project was supported by the National Natural Science Foundation of China (32170444 and 31672325) and the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA19050203). We are very grateful to Dr. Michael Orr for his help in improving the English during revisions.