Antennal and palpal sensilla of three predatory Lispe species (Diptera: Muscidae): an ultrastructural investigation

Antennae and maxillary palps are the most important chemical reception organs of flies. So far, the morphology of antennae and maxillary palps of flies of most feeding habits have been well described, except for that of relatively rare aquatic predatory species. This study describes sensilla on antennae and maxillary palps of three aquatic predatory Lispe species: Lispe longicollis, L. orientalis and L. pygmaea. Types, distribution, and density of sensilla are characterised via light and scanning electron microscopy. One type of mechanoreceptors is found on antennal scape. Mechanoreceptors (two subtypes) and one single pedicellar button (in L. pygmaea) are located on antennal pedicel. Four types of sensilla are discovered on antennal postpedicel: trichoid sensilla, basiconic sensilla (three subtypes), coeloconic sensilla and clavate sensilla. A unique character of these Lispe species is that the coeloconic sensilla are distributed sparsely on antennal postpedicel. Mechanoreceptors and basiconic sensilla are observed on the surface of maxillary palps in all three species. We demonstrated clear sexual dimorphism of the maxillary palps in some of the Lispe species, unlike most other Muscidae species, are larger in males than females. This, along with their courtship dance behaviour, suggest their function as both chemical signal receiver and visual signal conveyer, which is among the few records of a chemical reception organ act as a signal conveyer in insects.

Antennae and maxillary palps are the main chemical reception organs of flies on which numerous sensilla of various types can be found 1,2 . These organs play indispensable roles in the lives of flies in searching for food sources, mates, oviposition sites as well as other key life history stages [2][3][4][5][6][7][8][9][10][11] . Flies are under high selection pressure for receiving sufficient chemical signals and/or cues that are associated with their life history, such as searching for mates 12 , foods 13 or hosts 14 , and this could influence the morphology of the antennae 15 . Flies have a wide range of feeding habits including saprophagy, phytophagy, parasitism, hematophagy and predatory 16,17 , making them ideal models for studying the adaptation of insect olfactory organs according to different olfactory requirements. It is well documented that flies of different feeding habits have different antennal shape and sensillar types [7][8][9][10][11] . Structure of antennae and maxillary palps, especially the distribution and morphology of sensilla have been documented in detail in saprophagy, phytophagy and parasitismflies 7,8,18,19 , but few researches have focused on the predatory flies.
The species of genus Lispe Latreille (Diptera: Muscidae) are among the relatively rare predatory flies, closely associated to aquatic and subaquatic habitats 20 . Adult Lispe flies are commonly found around the margin of ponds, lakes, streams or seashore and prey on various insects including several mosquito species, such as anopheline and chironomid 21,22 . Visual perception is comparatively more important for these flies in hunting for their flying preys than chemical cues, yet they should still rely on their antennae and maxillary palps for olfactory cues and signals. Therefore, it is expected that the olfactory perception requirements of Lispe flies are largely different from that of saprophytic and parasitic flies, and this presumably results in specific antennal morphology adaptations. For example, Lispe neimongola Tian et Ma 9 has two conspicuous distinctions: the absence of coeloconic sensilla (Co) and enlarged spoon-like maxillary palps. It is unclear whether these morphological characteristics are common among other Lispe flies. The second segment of the antenna is the antennal pedicel, also covered with microtrichia. Two subtypes of mechanoreceptors can be distinguished by their shape and size on the antennal pedicel ( Fig. 2a,b). Usually there are one or two longer mechanoreceptors (Mr I) located on the antennal pedicel. Shorter mechanoreceptors (Mr II) are morphologically like those found on antennal scape, but are straighter in shape and more variable in length.
One pedicellar button (PB) is found in pedicellar recess and near the pedicellar cleft after separated antennal pedicel from antennal postpedicel in L. pygmaea. Pedicellar button consists of a circular central dome and a slightly convex peripheral ring with a small bunch of peripheral microtrichia (Fig. 2d).
Maxillary palp. Maxillary palps of males are swollen in the three Lispe species, and can be regarded as a representative character of Lispe. The ladle-shaped maxillary palps of L. orientalis with near right-angled edge have the highest degree of swelling among the three species (Figs. 1c, 6a). Comparatively, spoon-shaped maxillary palps of L. longicollis with a nearly round edge have a lower degree of swelling (Fig. 1a), and that of L. pygmaea are slightly swollen (Figs. 1e, 6b). The swelling degree of the maxillary palp are significantly different among the three species and between sexes ( Our results show that compared to their body length, the relative maxillary palp length and the relative width are different between sex and among these Lispe species (Fig. 7b,c). There are significant differences in the ratio of maxillary palp length to body length (LMP/BL) (    Table 2). Densities of Tr increase from the proximal region towards distal region on both anterior surface and posterior surface of antennal postpedicel (Table 3).

Discussion
The present study describes the antennal sensilla of three aquatic predators, L. longicollis, L. orientalis and L. pygmaea using scanning electron microscopy. The morphology and distribution of mechanoreceptors, pedicellar button, trichoid sensilla, basiconic sensilla, and clavate sensilla of three Lispe species resemble to previous results on L. neimongola 9 and other muscoid species, such as Delia radicum L. 18  The swollen maxillary palps in Lispe may increase their chemosensory functions. Different from the clublike maxillary palps in most of other fly species, maxillary palps of Lispe species are significantly swollen and flattened, and swelling degrees of maxillary palps in the three Lispe species are generally larger than typical Table 3. Average density of sensilla (10 −3 μm −2 ± SD, n = 10) on antennal postpedicel of three Lispe species. Ba basiconic sensilla, Cl clavate sensilla, Co coeloconic sensilla, Tr trichoid sensilla.  Table 4. Length, basal diameter (μm ± SD) and average density (10 −3 μm −2 ± SD) of coeloconic sensilla on antennal postpedicel of muscoid species. M male, F female, − unidentified; × no data.  (Table 1). Maxillary palps mainly acted as gustatory sensory organ 37 that react to molecules with low or zero vapor pressure, involved in contact or short-distance chemosensory functions 38 , compared to that of antennae, which typically perceive more volatile olfactory signals or chemical cues 39 . Shiraiwa 40 pointed out that maxillary palps of fruit flies can improve their sensitivity to food odours, others suggested that maxillary palps perceive olfactory signals at shorter distance, and can be integrated with the signals perceived in antennae to allow better manoeuvring when approaching lures 41 . Larger maxillary palps of L. neimongola were suggested to provide larger surface area for basiconic sensilla and enhance the perception of gustatory odours or signals 9 . For predators flying rapidly to chase prey like Lispe 22,42 , more sensilla can increase their behavioural responsiveness 43 . The swollen maxillary palps of the three Lispe species may function similarly to improve their gustatory and/or olfactory sensitivity. Maxillary palps could also be a signaller as well as a signal receiver. In respond to higher selection pressure of searching for food, hosts, or oviposition sites by chemosensory, female insects usually have larger antennae and maxillary palps or more sensilla attached to them 9,44,45 . In L. orientalis and L. neimongola 9 , maxillary palps are significantly more swollen in males than in females. Light microscopy photos (Figs. 1c, Supplementary Fig. S1c) and field observations show that maxillary palps of male L. orientalis are more conspicuous than other species when observed from a distance. Empiracle evidence shows that during courtship dance, some male Lispe flies circle around the female and flash their maxillary palps 21,42 . This could be important in correct species recognition and successful copulation, as many Lispe flies have highly overlapped habitats and ecological niches 42,46 . Thus, the maxillary palps could be dual-functional for male Lispe flies, and this is among some rare cases that an olfactory organ also plays a role as chemical signal receiver and as visual signal conveyer, which also indicates maxillary pales of L. orientalis are under different levels of sexual selection pressure.
Coeloconic sensilla are common on antennal postpedicel in most other fly species 8,19,25,26,47,48 , but are relatively sparse on antennal postpedicel of these three Lispe species, even completely missing in L. neimongola 9 . Beside chemosensory function, Coeloconic sensilla have been also proved to be sensitive to temperature and/ or humidity signals 49,50 . Compared with olfactory, temperature or humidity, predators rely more on acute vision which enhance their ability of colour vision, movement awareness and depth perception, especially on fast moving preys 19,51 . Lower number of coeloconic sensilla on antennal postpedicel reflect the adaptation to predatory lifestyle and could be regarded as a character of the genus Lispe.

Methods
Adult L. longicollis, L. orientalis and L. pygmaea were captured from Kalamaili Ungulate Nature Reserve, Xinjiang, China, in August 2013. All specimens were pinned as museum samples and air dried on site. Morphology of antennae and maxillary palps were examined under Olympus SZX16 stereoscopic microscope (Olympus Corp., Tokyo, Japan), morphological photographs were taken by a Canon 500D digital camera (Canon, Inc., Tokyo, Japan) coupled with stereoscopic microscope. Continuous images on different focal lengths were composed by Helicon Focus for Windows (Helicon Soft Ltd., Kharkov, Ukraine). Five specimens for both sexes of each species were used for measuring body length as well as length and width of maxillary palp. Three male specimens for each species were used for scanning electron microscopy. Heads of all specimens were cut off, then surface debris was removed by rinsing in phosphate buffered saline buffer (pH 7.4). Subsequently, antennae and maxillary palps were dissected respectively, cleaned with detergent by ultrasonic cleaner. After dehydration in graded ethanol series (twice 15 min each with 60%, 70%, 80%, 90%, 95%, 100% ethanol), antennae and maxillary palps were mounted on aluminium stubs with double-sided adhesive tape, then left in a desiccator for 24 h to dry thoroughly. Samples were coated with gold and observed with a HITACHI S34Q scanning electron microscopy (Hitachi Corp., Tokyo, Japan) at the Microscopy Core Facility, Biological Technology Centre, Beijing Forestry University (Beijing, China).
Length, width of maxillary palps and body length of five specimens for each sex were measured. Then the swelling degree (width to length) of maxillary palps and the ratio of maxillary palp measures to body length (length of maxillary palp to body length and width of maxillary palp to body length) were calculated and compared by two-way ANOVA in SPSS 22.0 (IBM Corp., Armonk, New York) between different species and different sexes of each species. Results of two-way ANOVA were visualized by Sigmaplot 12.5 (Systat Software, Inc., Chicago, Illinois). Length, basal diameter, tip diameter (clavate sensilla only), density and distribution of sensilla were measured using micrographs taken under different magnifications. The length of each single sensillum was measured (ten repeats of each type of sensilla) from the proximal rim to the tip. Density and distribution of various types of sensilla were measured by measuring square areas (each representing 576 µm 2 ) from proximal, median, and distal part (each consists one third of the antenna in length) of the antenna on both sides 2 , and ten quadrates were measured for each part. In this study, the terminology applied to describe antennal morphology and classification of types of sensilla followed those used by Cumming and Wood 52 .