Antennal sensilla and compound eyes of fruit flies, Drosophila melanogaster (Meigen), Drosophila suzukii (Matsumura), Drosophila immigrans (Sturtevant), and Drosophila hydei (Sturtevant) (all Diptera: Drosophilidae), were observed with stereoscopic microscopy and scanning electron microscopy. The results showed that the antenna consists of 3 segments: a proximal scape, a pedicel, and a flagellum composed of a funiculus and a dendritic arista. Six morphologically distinct types of sensilla were observed on the antenna: chaetica, microtrichia, trichoid (Tr I, Tr II), basiconic, clavate, and coeloconic (Co I, Co II). Basiconic sensilla were significantly shorter than other sensilla in the funiculus of D. melanogaster. The number of clavate sensillae was less than other sensilla in the funiculus of D. melanogaster and D. immigrans, but was greater in D. suzukii and D. hydei. Moreover, coeloconic sensilla were absent in D. suzukii and D. hydei. The length and abundance of the chaetica sensilla on the compound eyes were different significantly among the 4 species. Drosophila hydei had chaotic sensilla with the greatest length and abundance; they were lowest for D. suzukii. Based on the previous literature, the possible functions of these sensilla are discussed. We inferred that fruit flies may regulate their behaviors according to the information detected by these sensilla.
There are about 3,700 described species of Drosophila fruit flies in the world, though most of them are not recognized as pests because their larvae mostly develop in damaged or rotting fruit (Bolda et al. 2010). However, Drosophila suzukii (Matsumura) (Diptera: Drosophilidae) can feed on ripening cherries, berries, grapes, and > 60 other kinds of fruits (Mitsui et al. 2006), which has attracted the attention of fruit growers and researchers around the world (Cini et al. 2012). Therefore, among these Drosophila species, D. suzukii were found to be the most harmful species because they oviposit in fresh fruits with their serrated ovipositor (Mitsui et al. 2006; Hauser et al. 2009). Subsequently, secondary infection by Drosophila or microorganisms may contribute to further fruit deterioration, causing considerable damage (e.g., monilinia brown rots, botrytis rots, and sour rot) (Cini et al. 2012). Fruit flies are very difficult to control due to their short generation time, high reproductive capacity, and a wide range of fruit hosts (Bolda et al. 2010; Cini et al. 2012). Moreover, other Drosophila species also were reported to be fruit pests. Katoh et al. (2007) reported that Drosophila immigrans (Sturtevant) (Diptera: Drosophilidae) is one of the widespread fruit flies in the Japanese and East Asia regions. Ren et al. (2014) reported that D. immigrans and Drosophilidae hydei (Sturtevant) (Diptera: Drosophilidae) also can damage cherries in different regions of China. Drosophila melanogaster (Meigen) (Diptera: Drosophilidae) and D. immigrans were described as the principal Drosophila pests attacking cherry in Gansu Province, China (Guo 2007). These 4 Drosophila species are the dominant Drosophila pests in China, but have different ecological niches and biological characteristics.
Female D. melanogaster preferentially oviposit on food substrates containing a high concentration of ethanol (Azanchia et al. 2013). Gao et al. (2018) found that D. melanogaster and D. suzukii occupy different ecological niches due to their differences in sensitivity and tolerance to ethanol from the host plant. Studies on the interaction between insects and host plant volatiles have shown that the olfactory system of insects plays a critical role in finding host plants, mating, and reproduction (Krieger et al. 1999). Antennae are the main olfactory organs in the olfactory system of insects, and play an important role in the survival of insects and a wide range of behaviors, such as habitat selection, host location, and sexual communication (Chapman 1998; Isidoro et al. 2001). The functions of antennae are facilitated by specialized parts of the antennal epidermis, called sensilla, which are the important neuronal receptors (chemoreceptors) for receiving the signal of volatiles (Bin et al. 1989), but also can serve as mechanoreceptors, thermoreceptors, hygroreceptors, and CO2 receptors (Keil 1999; Stange & Stowe 1999). Moreover, compound eyes of insects are important visual organs. Different from antennae, the photoreceptor in compound eyes can help insects to find plants by receiving light signals from the environment (Chapman 1998). However, whether other receptors exist in the compound eyes needs to be studied. Therefore, analyzing the morphological structure and distribution of sensilla from different Drosophila species is important to understand the differences in olfactory behavior and host identification mechanisms of insects.
Much research has been reported on the distribution, external morphology, and ultrastructure of various sensilla located on the antennae of different fruit flies. Sensilla are borne directly on the antennae of insects in the form of hairs, pegs, pits, or cones. The antennae of D. melanogaster can have about 200 basiconic, 150 trichoid, and 60 coeloconic sensilla, about 75% of all sensilla, covering the surface of the funiculus (Stocker 1994; Shanbhag et al. 1999; Gao et al. 2007). However, the distribution or ultrastructure of various sensilla has not been studied in D. suzukii, D. immigrans, or D. hydei.
To improve our understanding of the peripheral sensory structures involved in the perception of pheromones, the external structure of the antennae, and the type and distribution of the antennal sensilla and compound eyes sensilla in the adults of D. melanogaster, D. suzukii, D. immigrans, and D. hydei species were observed using stereoscopic microscopy and scanning electron microscopy. This study can provide guidance for trapping insects using attractants such as plant-derived volatiles and synthetic sex pheromones (Howse et al. 1988), which will contribute to the management of fruit flies.
Materials and Methods
COLLECTION AND REARING OF FLIES
Drosophila melanogaster, D. suzukii, D. immigrans, and D. hydei were obtained from fields in Yantai (37.2316°N, 121.6000°E), People's Republic of China, in Jun 2016 and reared in the laboratory with an artificial diet. This diet was composed of mashed banana and apple, corn flour, sucrose, yeast extract, sorbitol and agar (Zhai et al. 2014). The colony was maintained in a climate-controlled growth chamber at 25 ± 0.5 °C, 70 ± 0.5% RH, and a photoperiod of 16:8 h (L:D). Flies were used in this study 6 d post-emergence.
For general morphology, the heads together with antennae were removed from 30 individuals of each species and were cleaned 3 times in a phosphate-buffered saline (pH 7.0). Then the antennae were removed from antennal fossa with a dissecting needle and placed on a glass slide in a drop of water. After being covered by a cover slip, the samples were observed using an Olympus BX53 stereoscopic microscope (Olympus Corp., Hamburg, Germany).
SCANNING ELECTION MICROSCOPY (SEM)
For each species, 30 individuals were transferred individually into 1.5 mL centrifuge tubes and rinsed 3 times in phosphate-buffered saline pH 7.0 for 15 min each, and placed in 2.5% glutaraldehyde at 4 °C for 24 h. After fixation, the specimens were washed 3 times in 0.1 M pH 7.0 phosphate buffer for 15 min each. Ethanol was used as the dehydration agent. Specimens were dehydrated in a series of ascending alcohols (30%, 50%, 70%, 90% [2 washes], 100% [3 or 4 washes], each for 2 h). Finally, excess alcohol was removed, and the specimen was submerged in amylacetate for 1 to 2 d. Specimens were air dried for 1 to 3 h at 35 °C. Then the specimen was fixed on an SEM specimen stub with sticky tapes. The specimen then was coated with gold film with 150 Å thickness using a JEOL sputtering device (JEOL Ltd., Tokyo, Japan) for 2 to 3 min.
SEM and Analysis
The specimen then was examined using a SUPRA55 SEM (Carl Zeiss AG, Oberkochen, Germany). The types of different sensilla were identified and classified according to the descriptions of Zacharuk (1985). The images of the sensilla on the surfaces of the antennae from different species were taken at magnifications of 1,500× to 3,500×.
The mean length of body, antennae, and each type of sensilla from different Drosophila species was calculated and analyzed using a 1-way ANOVA (α = 0.05) and Student-Newman-Keuls multiple comparisons using the SPSS 19.0 statistical analysis package (IBM Corp., New York, USA; www.ibm.com). All micrographs were processed in Adobe Photo-shop CS5 (Adobe Systems, Inc., San Jose, California, USA).
MORPHOLOGY OF FLY ANTENNAE
The antennae of fruit flies were situated in the antennal fossa between 2 compound eyes. Each antenna was characterized by 3 segments, a proximal scape, a pedicel, and a flagellum composed of a funiculus and a dendritic arista (Fig. 1). The movable scape was attached to the pedicel, allowing the movement of antenna. Both the scape and the pedicel were heavily covered with microtrichia and possessbristles. The arista was found on the dorso-proximal end of the funiculus, which is an unsegmented flagellum.
For the 4 Drosophila species, the funiculus was about twice as long as the pedicle and 4 times longer than the scape (Table 1). The lengths of each part of the antennae in different species were significantly different. The length of the scape in D. hydei was greater than that of D. immigrans and D. melanogaster, and that in D. suzukii was shortest among the 4 species (F = 14.83; df = 3; P = 0.001). Drosophila immigrans had the longest funiculus and arista, which were shortest in D. melanogaster (funiculus: F = 21.80; df = 3; P < 0.01; arista: F= 101.93; df = 3; P < 0.01) (Table 1).
SENSILLA OBSERVED ON THE SCAPE AND PEDICEL
Figure 2 shows the sensilla on the scape and pedicel segments of 4 Drosophila species. There were no differences in sensilla types among the 4 species. Two distinct morphological types of sensilla, chaetica, and microtrichia sensilla were observed in these segments. These 2 types of sensilla were mainly distributed on the forward side of the scape and pedicel. A small number of microtrichia sensilla were distributed at the base of the funiculus. Chaetica sensilla, spread over the surface of the pedicel, were bristle-like structures with some in a single row, running along the periphery of the scape. The length of chaetica sensilla varied from 27 µm to 64 µm (Table 2). Numerous microtrichia sensilla distributed on the surface of the scape and pedicel were hairlike, with the length from 5.6 µm to 9.1 µm. The number and lengths of chaetica and microtrichia sensilla were not significantly different among the 4 species (F = 1.325; df = 3; P = 0.277 for number; F = 1.643; df = 3; P = 0.197 for length).
SENSILLA OBSERVED ON THE FUNICULUS
Four types of sensilla were observed on the funiculus surface: trichoid, basiconic, clavate, and coeloconic sensilla. All sensilla were distributed mostly on the windward side of the funiculus surface, oriented toward the tip of the antenna (Fig. 3).
Trichoid sensilla were most abundant and had the widest distribution on the funiculus of these species. Trichoid sensilla tapered from the base, terminating in a pointed, needle-like structure. Two types of trichoid sensilla were observed, both with a length of about 12 to 14 µm. Trichoid sensilla I (Tr I) is straight and has ridges, whereas Trichoid sensilla II (Tr II) is a crooked structure. The abundance, distribution, and morphology of the sensilla on the funiculus of these Drosophila species were not significantly different (Table 3)
Basiconic sensilla were distributed at the surface of the funiculus, arising from a raised cuticular collar at the surface of the cuticle and slightly curved or not. They were characterized as digitiform (finger-like) structures with a rounded point and a smooth surface. The number of basiconic sensilla was less than that of trichoid sensilla. These basiconic sensilla on D. melanogaster (9.4–11 µm) were significantly shorter than the others (18–24 µm) (F = 187.106; df = 3; P = 0.000).
The antennal length of the 4 Drosophila species.
Clavate sensilla were shorter than trichoid sensilla, and less widespread on the surface of the funiculus. The clavate sensilla had abruptly blunt tips. They were similar to, but shorter and thicker than, the basiconic sensilla. The length of clavate sensilla was about 9.5 to 12.6 µm. There were few clavate sensilla observed on D. melanogaster and D. immigrans, but they were more abundant on D. suzukii and D. hydei.
Coeloconica sensilla were scattered irregularly on the surface of the funiculus, and were the shortest (3.1–4.1 µm) and least abundant sensilla. They arose from a sunken cavity that had a single opening with an irregular rounded margin at the cuticle surface. Two types of coeloconica sensilla can be observed on the funiculus surface. Coeloconica sensilla I (Co I) was characterized by a short peg with grooves over the sensilla, whereas Coeloconica sensilla II (Co II) was usually smooth, curved and approximately triangular in shape. Coeloconica sensilla were observed in D. melanogaster and D. immigrans, but absent in D. suzukii and D. hydei.
The dendritic arista, with many bifurcations, was located proximally near the base of the funiculus. The hair of the arista in females was significantly longer than that of males in the 4 species (F = 101.983; df = 7; P < 0.001) (Table 1).
SENSILLA OBSERVED ON THE COMPOUND EYES
Only chaetica sensilla were observed on the surface of the compound eyes (Fig. 4). The length and abundance of the chaetica sensilla were significantly different among the 4 species (F = 126.287; df = 3; P < 0.001) (Table 2). The length and abundance of the chaetica sensilla were greatest for D. hydei, followed by D. melanogaster or D. immigrans, and were the least for D. suzukii.
Length and amount of sensilla observed on the scape, pedicel and compound eyes.
The antennal morphology of Drosophila in this study is similar to that in other cyclorrhapha species such as Bactrocera dorsalis Hendel, Bactrocera tau Walker, Bactrocera zonata Saunders (all Diptera: Tephritidae), Eristalis (Diptera: Syrphidae), and Liriomyza huidobrensis Blanchard (Diptera: Agromyzidae) (Cai et al. 2014; Awad et al. 2015), which consists of 3 segments and a long arista. Eight morphologically distinct types of sensillae were observed on the antenna of 4 Drosophila species, such as chaetica, microtrichia, trichoid (Tr I, Tr II), basiconic, clavate, and coeloconica sensillae (Co I, Co II). In profile, all of the Drosophila species had similar antennal sensilla, except that coeloconica sensilla were absent in D. suzukii and D. hydei. Unlike the Drosophila species, 7 types of sensilla, including placodea sensilla, were found in Lysiphlebus fabarum (Marshall) (Hymenoptera: Braconidae) but clavate sensillae were not observed (Xi et al. 2010). In 6 species of Noctuidae, there were 9 antennal sensilla on the antennal segments, such as trichoid, chaetica (I, II), coeloconica, basiconic, and 4 other sensilla types (Wei 2015). Therefore, trichoid, chaetica, coeloconica, and basiconic sensilla were widely distributed on the different types of insect antenna.
The sensillae in insects occur in different numbers, morphologies, and have different functions (Romani et al. 2009). The chaetica and microtrichia sensilla were observed on the scape and pedicel in the Drosophila species in this study. Compared to the other sensilla, chaetica sensilla were larger, thicker, and higher on the antenna, and initially could contact signals in the environment. Hu et al. (2010) and Awad et al. (2015) reported that the possible functions of the chaetica and microtrichia sensilla on the surface of scape and pedicel in B. tau, Bactrocera minax Enderlein (Diptera: Tephritidae), B. zonata, and 3 other Bactrocera species might have a function of mechanoreception. Moreover, chaetica and microtrichia sensilla have been shown to have olfactory functions (Seada 2015). The abundance and length of chaetica and microtrichia sensilla of antennae were not significantly different in D. melanogaster, D. immigrans, D. suzukii, and D. hydei. Interestingly, chaetica sensilla observed on the surface of complex eyes were significantly different among the 4 species. As is well known, the function of complex eyes is attributed to the photoreceptor in compound eyes, which can receive light signals (Chapman 1998). There is no research about sensilla on the surface of the complex eye. In this study, the D. hydei, D. melanogaster, or D. immigrans had more numerous chaetica sensilla than D.suzukii. Possibly, mechanoreception is similar for the antenna of Drosophila species, but not the reason for differences in ecological niches among them. However, the function of compound eyes also may be involved in recognizing the softer rotted fruits in addition to mechanoreception by chaetica sensilla, though this needs to be assessed in future research.
Types and amount of sensilla observed on the funiculus.
The pyriform funiculus is the most conspicuous segment of the antenna, housing a variety of sensilla. Six types of antennal sensilla, including trichoid (Tr I, Tr II), basiconic, clavate, and coeloconica (Co I, Co II) sensilla were observed on the funiculus of Drosophila species. Generally, trichoid sensilla were the most abundant, and function of mechanoreceptors or proprioceptors (Ochieng et al. 2000; Fernandes et al. 2002). Sukontason et al. (2004) reported that the thrichoid sensilla had both chemoreceptor and mechanoreceptor functions in Calliphoridae, Sarcophagidae, and Muscidae. Basiconic and clavate sensilla were considered to be olfactory receptors to a wide range of simple molecules such as carbon dioxide, ammonia, esters, amines, or volatile n-alcohols (Stocker 1994; Lopes et al. 2002; Onagbola & Fadamiro 2008). In this study, basiconic sensilla were significantly shorter in D. melanogaster than the other 3 Drosophila species studied. The number of clavate sensillae of D. melanogaster and D. immigrans was lower than that of D. suzukii and D. hydei. Possibly basiconic and clavate sensilla are involved in host location and selection for different Drosophila species. The coeloconica sensilla in Drosophila antennae had highly specialized neurons and performed chemosensory function (Yao et al. 2005). However, coeloconica sensilla were sensitive to temperature and humidity in Manduca sexta L. (Lepidoptera: Sphingidae) (Shields & Hildebrand 1999) and had olfactory functions in wasps (Van Baaren et al. 2007). Therefore, the function of coeloconica sensilla varies significantly in different insects. In this study, coeloconica sensillae were absent in D. suzukii and D. hydei. Therefore, in Drosophila species, coeloconica sensillae may not be the principal sensilla for location of hosts and identification of signals in the environment. The function of all the sensilla in Drosophila species, especially D. immigrans, D. suzukii, and D. hydei, need to be verified by single-cell recording and electroantennagraphy studies.
The results presented here could be valuable for further investigation of insect olfactory behavior and host identification mechanisms, and provide the basis to study the relationships between morphology, and insect behavior and taxonomy, in Drosophila species.
This work was supported financially through a grant from Natural Science Foundation of China (31801750), Agricultural scientific and technological innovation project of Shandong Academy of Agricultural Sciences (CXGC2018E17), and The National Key Research and Development Program of China (SQ2018YFD020082). The authors wish to thank Jianlong Bi for the manuscript language revision.