A Study of the Sensory Biology of Labellar Gustation by the Hale Queensland Fruit Fly, Dacus Tryoni (Froggatt) (Diptera: Tephritidae) With Respect To Cue-Lure

Abstract

Hales of the Queensland fruit fly Dacus tryoni (Froggatt) are attracted to and readily ingest 4-(pacetoxyphenyl)- 2-butanone (cue-lure). The development of responsiveness of these flies follows a sigmoid curve: almost no flies respond in the first two days after eclosion, the proportion of responsive flies rises gradually to a peak (70- 80%) 7 days post-eclosion and remains there for at least 21 days . Light, scanning and transmission electron microscopy were used to study the morphology, dis t ribution, and the innervation of the labellar sensilla of male Dacus tryoni.These were classified into six morphological types on the basis of their lengths: longest CLH), long ClH), medium long CMH> ,short straight (ssh), bristle Cb), and very short bristle (sb) sensilla. Because of a dimensional overlap between lH and MH and the pattern of their distribution, these have been classified as subtypes of one group, the intermediate sensilla CiH). The total number of fringe sensilla had a positive correlation to the size of the labellum and to that of the fly. This has been suggested to occur within upper and lower limits, about a modal size characteristic of members of the species . Longest sensilla on larger labella were longer than corresponding sensilla on small label la. Labellar fringe sensilla are in five rows, almost parallel to the marginal strip of the oral disc, the first or lowermost row being the marginal row. Bristle and short bristle sensilla occur on the marginal and fifth (uppermost) rows, but more of the "sb" are on the marginal row. The LH and iH are predominantly in the 3rd and 4th rows and, whilst ssh occur in all rows, the large proportion of them are in the marginal row. The marginal ssh are curved inwards over the marginal strip, towards the oral surface, such that when the labellum is everted their tips are in direct contact with food. Except for the shorter sensillar types on the upper rows, a large proportion of all the sensilla contact food substances when the proboscis is extended prior to and during ingestion. The papillae on the oral surface numbering 56-66 were classified into two morphological types: the peripheral papillae CPP) and the actual interpseudotracheal papillae CIP>. The PP occur, one at the origin of each pseudotrachea, all along and close to the marginal strip. The peg of these papillae is stout with either a pore or slit on the blunt tip and each has a wide ring-like bulging socket that may allow for multidirectional mechanosensitivity. The IP are in deep, narrow sockets located at the bases of the pseudotracheae, in the spaces between these and the interpseudotracheal plates. These papillae have sharp tips and are closely apposed to the pseudotracheae such that, the positioning of the pore allows monitoring of chemicals in food flowing in the tubes during suction. Their positioning seems to limit mechanical displacement. Histological methylene blue staining showed that papillae and fringe sensilla are innervated by two or more neurones. Ultrastructure of the innervation of the labellar sensilla revealed that, except for the short bristle sensilla which are innervated by two gustatory neurones only, all the other fringe sensillar types have 4 gustatory neurones and a putative mechanoreceptor. The papillae, PP and IP have 3 gustatory neurones in the single lumen of the peg and a putative mechanoreceptor. Action potential frequencies were recorded from certain long and/or longest labellar sensilla when stimulated with a series of LiCl concentrations, and with cue-lure and five othe r closely related compounds. Dose/response plots and probit analysis showed that two neurones, labelled as Pl and P2, with medium-large and large spike amplitudes are positively sensitive to increase in LiCl concentration. A third cell, the n-neurone, ha d a smaller spike and a firing rate that negatively correlated with Li Cl concentration; this corresponds to the classical "water cell" of Evans and Mellon (1962) and Rees < 1970a). A fourth neurone (Lp), with a larger spike amplitude than P2, was also evident in a small number of the recordings with molar LiCl. Response characteristics of Pl in longest and long sensilla are similar for all LiCl concentrations; as also are the threshold and firing of P2, at concentrations less than 100 mM LiCl. However, at O.S H and molar LiCl, P2 had a significantly higher firing rate in long than in longest sensilla. SO mH LiCl was chosen as the carrier electrolyte for cue-lure in e lectrophysiological tests, because of the relatively low firing rate of the n-neurone, very low or no response f r om the Pneurones, and adequate conductivity. A gustatory neurone, which was rapidly adapting to millimolar cue-lure and dose-sensitive to 4-phenyl-2-butanone and two of its derivatives {4-(p-acetoxyphenyl)-2-butanone and 4-Cp - hydroxyphenyl)-2-butanone}, was recognised and denoted as the c-neurone. The electrophysiology of the c-neurone when stimulated with cue-lure, correlated well with the feeding behaviour data over the range of concentrations tested. Further recording, using changing stimuli produced by a flow system, showed that at least one of the P neurones responding to 0.5 M LiCl has a different spike shape to that of the c-neurone. Differences in responsiveness of the c-neurone to a range of molecules: 4-decanone (4-DCN), toluene CTLN), acetophenone CACP), 4-phenyl-2-butanone (4-PBN), and 4-(phydroxyphenyl)- 2-butanone (p-HPBN) relative to cue-lure Cp- APBN), were used to identify some of the molecular characteristics that may be complementary to a hypothetical receptor site for the cue-lure molecule. ACP and TLN, with shorter molecules than 4-PBN, and the aliphatic 4-DCN were not stimulatory to the c-neurone. The results are discussed on the basis of the stereochemistry, hydrophobic and electronic charge characteristics of the functional groups of the substituents of the molecules tested. Possible application of these results to studies on the modification of lure potency by structureactivity modelling is considered.