• Published on

  • View

  • Download


<ul><li><p>Ent. exp. &amp; appl. 21 (1977) 99--11 I. North-Holland Publ. Co. Amsterdam </p><p>ADAPTATIONS AND RESPONSES OF DASYMUTILLA OCCIDENTALIS (HYMENOPTERA: MUTILLIDAE) TO </p><p>PREDATORS </p><p>BY </p><p>JUSTIN O. SCHMIDT and MURRAY S. BLUM </p><p>Department of Entomology, University of Georgia, Athens, Georgia, U.S.A. </p><p>The mutillid wasp Dasymutilla occidentalis possesses several adaptations and exhibits a number of responses which appear to be of defensive value: a long mobile sting with powerful venom; a strong, rounded and slippery cuticle; an ability to run very rapidly and evasively; an aposematic warning coloration pattern; and the ability to respond to an attack by making stridulatory sounds and by releasing a chemical secretion or both. The effectiveness of these .defenses is supported by tests utilizing various vertebrate and arthropod predators. The raison d'&amp;re of the multiple lines of defense possessed by D. occidentalis and the relative value of each line of defense are discussed. It is postulated that aposematic coloration, audible stridulation, and a volatile defensive exudate all function primarily as part of an early warning system enabling a predator to recognize this wasp - with its very algogenic venom - as unpalatable and potentially dangerous. </p><p>Dasymutiila occidentalis (L.) is a large, conspicuous mutillid wasp potentially susceptible to the predatory advances of a variety of vertebrates and invertebrates. The apterous diurnally active female wasps, parasites of larger Hymenoptera, engage in extensive searching activities near the ground surface for potential hosts. Such long daily exposures over the lifespan of at least several months add up to an extended period of time in which these wasps are susceptible to predation. In spite of the female's inability to fly and its low reproductive potential, this mutillid is a fairly common species, suggesting that it possesses an effective array of defensive adaptation~ and behaviors. The paucity of more than anecdotal reportings of the defensive mechanisms of mutillid wasps prompted us to look in depth at the various defensive capabilities of this species. The results of this investigation are reported in the present paper. </p><p>METHODS AND MATERIALS </p><p>Females and males of D. occidentalis were captured during July and August of 1974 and 1975 in Clarke Co., Georgia. Female wasps were maintained in the laboratory in plastic tubs containing sand and were provided with water and honey. Male wasps were utilized only for the preparation of extracts for chemical investigations; in all other studies only female wasps were used. </p></li><li><p>100 JUSTIN O. SCHMIDT AND MURRAY S. BLUM </p><p>Chemical analysis Methylene chloride extracts of both heads and thorax-abdomens of 74 male D. </p><p>occidentalis were prepared and stored at -20 ~ until analyzed. Similarly prepared extracts from female wasps were analyzed immediately. </p><p>Gas-liquid chromatographic analyses were performed isothermally on a Tracor MT - 220 gas chromatograph equipped with flame ionization detectors and 183 0.65 cm glass columns. Columns used were: 3% OV-17 on Chromosorb G AW- DMCS 100/120 mesh (65~ 10% SP-1000 on Supelcoport 80/100 mesh (80~ 10% Carbowax 20 M on Chromosorb W AW-DMCS 100/120 mesh (80~ and 5% SE-30 on Chromosorb W AW-DMCS 60/80 mesh (65~ Combined GLC-mass spectro- scopic analyses were performed using an LKB 9000 instrument with a 183-cm column of 1% OV-17 on Supelcoport 80/100 mesh, programmed from 50 ~ to 300 ~ at a rate of 10~ </p><p>Quantitative analyses of the volatile components were determined by cutting out the area under each chromatographic peak and weighing it. Ten replicate injections on the OV-17 column maintained at 65 ~ were analyzed. </p><p>Sodium borohydride reduction was carried out by adding 20 I~1 of extract to 200 lai of ethanol containing 25 mg of NaBH4. The solution was centrifuged and the supernatant injected onto the OV-17 column at 65~ Standard compounds included 4-methyl-3-heptanone (Aldrich Chemical Co.) and 4-methyl-3-heptanol (Chem. Samples Co.). </p><p>Acoustical recordings Recordings of female D. occidentalis were made in the soundproof chamber </p><p>maintained by the USDA laboratory in Gainesville, Florida. Recordings of wasp stridulations were made both when a wasp was held with forceps and when it was free but harassed. Amplitude versus time osciliograms and frequency profiles were prepared. </p><p>Analysis of cuticule strength Various insects which are considered hard shelled and/or are related taxonomic- </p><p>ally to mutiUids, were tested for resistance to a crushing force. A Hanson Cook-O- Meter (Model No. 1310) was adapted to measure the force necessary to crush various body parts of the dried insect specimens. Analysis consisted of the measurement of the force necessary to induce skeletal collapse. Each analysis was replicated ten times. </p><p>Cursorial speed Running speeds of D. occidentalis females were determined at a temperature of </p><p>25 ~ on ground lightly covered with pine needles. Wasps were repeatedly released and "chased" with fingers, during which time distances and stopwatch times were recorded. A total of ten uninterrupted runs was analyzed. </p><p>Predator responses The following potential predators of D. occidentalis were maintained under </p></li><li><p>ADAPTATIONS AND RESPONSES OF DASYMUTILLA TO PREDATORS 101 </p><p>laboratory conditions: the red imported fire ant (Solenopsis invicta Buren), the Western harvester ant (Pogonomyrmex owyheei Cole), the Chinese mantid (Tenodera aridifolia sinensis Saussure), wolf spiders (Lycosa spp., Geolycosa sp., L ycosa carolinensis Walck.), tarantulas (Dug esiella hentzi (Girard) and Aphonopelma sp.), the Cuban anole (Anolis sagrei sagrei Duneril &amp; Dibron), the Carolina anole (Anolis carolinensis Voight), the six-lined race runner (Cnemidophorus sexlineatus (L.)), the Florida scrub lizard (Scelophorus woodi Stejneger), the common starling (Sturnus vulgaris L.), and the gerbil (Meriones shawi Duvernoy). Predation responses were measured by observing the responses of both the wasp and the predator when the former was introduced into the environment of the latter. After each encounter with a wasp, the predator was given a German cockroach (Blattella germanica (L.)) or another previously demonstrated palatable insect in order to determine if the predator was hungry. With the possible exception of the lizards, which were field captured, all predators were naive. Results of all experimental encounters, in which the subsequently introduced palatable prey were not eaten, were discarded. In the case of fire ants, laboratory experiments were supple- mented by field tests. </p><p>The deterrent value of the major component of mutillid mandibular gland secretion was determined by placing filter paper flags or palatable prey treated with I lal of the compound in the immediate vicinity of potential predators and recording their subsequent reactions. </p><p>RESULTS </p><p>Chemical analysis Preliminary gas chromatography (GC) of cephalic extracts of D. occidentalis on </p><p>various columns including SE-30, OV-17, Carbowax 20 M, and SP-1000 demon- strated the presence of one major volatile component, two components in moderate amounts and several trace components. GC-MS analysis of the major component yielded the following prominent peaks: m/e (relative intensity) mol. ion 128 (3), 99 (8), 86 (34), 71 (61), 57 (100), 43 (92), and 29 (52). This spectrum was identical to that of 4-methyl-3-heptanone (McGurk et al., 1966). The identity was further confirmed by simultaneous co-injection of 4-methyl-3-heptanone and the wasp extract. Peak 1 remained sharp and symmetrical while additively increasing in size. </p><p>Head extracts of males gave GC patterns identical to those given by female extracts, but contained a much smaller amount of material per wasp. Dissections of the individual sexes revealed the presence of well-developed mandibular gland reservoirs which contained yellowish fluid. Crushing a gland yielded the same characteristic sweet ketonic odor as was detected when a wasp was restrained. GC analysis showed extracts of the mandibular gland to be identical with whole head extracts. GC analysis of the extract of the rest of the body revealed no low-boiling constituents. </p><p>Reaction of D. occidentalis extract with sodium borohydride resulted in a loss of the peak corresponding to 4-methyl-3-heptanone and the appearance of a peak </p></li><li><p>102 JUSTIN O. SCHMIDT AND MURRAY S. BLUM </p><p>with the same retention time as 4-methyl-3-heptanol. The largest minor constituent did not react with the reagent while the smaller component disappeared. The chemistry of the unidentified compounds is currently under investigation. </p><p>Quantitative analysis of the three reproduceable peaks of D. occidentalis gave the following approximate values for percent of total low-boiling volatiles: peak 1 (4-methyl-3-heptanone) 72%; peak 2, 20%; peak 3, 8%. </p><p>Sound production When disturbed, both males and females of D. occidentalis produce a sound </p><p>audible to the human ear at a distance greater than a meter. Although stridulitra, series of transverse striations, and ridge-like plectra are present on the second through sixth metasomal (= apparent abdominal) tergites (Hermann &amp; Mullen, 1974), our observations indicate the sound is produced mainly by the movement of the stridulitrum on the third metasomal tergite across the plectrum on the posterior margin of the second tergite. Because acoustical recordings of a female wasp held with forceps yielded frequency and amplitude profiles essentially identical to those obtained with a wasp allowed to run free while being harassed, only results obtained with the latter are presented. </p><p>Fig. 1 represents an amplitude versus time oscillogram recorded at a slow speed. The variable nature of the signal amplitudes, the irregular time durations of the pulse trains (the closely spaced series of pulses which combine to form each individual "chirp"), and the uneven temporal spacings of the individual pulse trains, is clearly seen. Two single pulse trains are expanded in Fig. 2. Again variability in amplitudes and temporal spacings are evident. Furthermore, the number of individual pulses per pulse train derived from sixteen consecutive trains, ranged from 31 to 44 with an average of 36.1 + 3.27. Fig. 3 represents four individual pulses taken from different parts of pulse trains and recorded on a greatly expanded time scale. Fig. 4 represents a frequency profile from 0 to 32 kilohertz (kHz) for several pulse trains. Fig. 4A represents one direction of the metasomal movement necessary to produce sound and Fig. 4B represents the other direction. As can be seen, no narrow frequency ranges predominate: the intensity gradually increases to a maximum from 4 to 8 kHz and trails off from there to 45 kHz, the sensitivity limit of the apparatus. Background noise was uniform across the profile and was considerably lower than signal level. </p><p>Cuticular resistance to crushing The results of forces applied across the mesothorax of various dried insect </p><p>specimens, are recorded in Table I. The forces required to crush the head and abdomen were less than those required to crush the thorax. For D. occidentalis, for example, the crushing force applied antero-posteriorly by the bar, placed horizontally across the frons, was 19.5 + 8.3 Newtons and the dorso-ventrally applied force to crush the second metasomal segment, was 14.5 + 3.0 Newtons. </p><p>During the crushing tests, it became evident that the actual strength of the cuticle was only part of the protective value of the integument. Excluding appendages, all external integumentry parts of D. occidentalis were extremely </p></li><li><p>ADAPTAT IONS AND RESPONSES OF DASYMUTILLA TO PREDATORS 103 </p><p>rounded and slippery. These factors frequently resulted in the specimens of this species slipping from within the apparatus during the testing process. No other species exhibited this "slipping" tendency. </p><p>Running speed From ten trials using female D. occidentalis at 25 ~ the mean ground speed was </p><p>13.8 + 3.0 cm/sec, or roughly 0.5 km/hr. </p><p>i i 9 i . L I II I i , t . , I ~ t,m b* Ind Ldd l ,b l~ ,A Id l J ,&amp;a lhd i . Jn l~ i ~ a _ _ I I I d lLI Idi l l i l l I d l t l l l l l l n i i l i i B l i i l _ _ N i n i _ </p><p>.l See Fig. 1. Oscillogram of stridulatory series of pulse trains produced by D. occidentalis. </p><p>i i i l n i l i i i i i i l l l l l l l l l l l l l l ummnl l l lUU l lnmml l l i i l m In </p><p>N u </p><p>m </p><p>im, lmu~,m-~q </p><p>H n i l ~ ~ 1 n m l </p><p>Fig. 2. (A) Oscillogram representing the expansion of the pulse train produced by movement of the metasoma (=apparent abdomen) in one direction; (B) oscillogram of the pulse train produced by </p><p>movement of the metasoma in the opposite direction. </p></li><li><p>104 JUSTIN O. SCHMIDT AND MURRAY S. BLUM </p><p>Predator responses The results of encounters of D. occidentalis and hungry predators are shown in </p><p>Table II. Because many of the observational results obtained from these predator studies are difficult to quantify, brief descriptions of the encounters are presented. </p><p>When attacked by small numbers of ants, D. occidentalis removes the ants by </p><p>I 1 m sec </p><p>I </p><p>Fig. 3 (A--D). Oscillograms representing expansions of four separate pulses. </p><p>I I I </p><p>2i </p><p>4 8 12 16 20 24 28 FREQUENCY (KHZ) </p><p>Fig. 4. Frequency profiles of D. occidentalis stridulation. (A) The average of several pulse trains produced by movement of the metasoma in one direction; (B) the average of several pulse trains </p><p>produced by movement of the metasoma in the opposite direction. </p></li><li><p>ADAPTATIONS AND RESPONSES OF DASYMUTILLA TO PREDATORS 105 </p><p>Insect </p><p>TABLE 1 </p><p>Force in Newtons necessary to crush the thorax of selected dried insects </p><p>Common name Approx. Force to dried weight crush thorax2, 3 </p><p>in g' </p><p>Force per gm dry weight to crush thorax 2 </p><p>Dasvmutilla occidentalis (females) velvet ant 0.112 27.8 _+5.3 Scolia dubia (Say) scoliid wasp 0.083 11.5 _+ 3.2 Apis mellifera L. (workers) honey bee 0.020 2.44 _+ 0.36 Vespula maculata (L.) (queens and workers) bald-faced hornet 0.137 5.74 _+ 1.3 PachylobiuspicivorusGerm.pitcheatingweevil 0.041 11.6 _+3.04 Pseudlucanus capreolus (L.) stag beetle 0.550 14.6 _+ 2.85 26.6 _+ 5.1 Phyllophaga spp. brown June beetles 0.169 5.25 _+ 1.1 31.1 _+ 6.5 </p><p>5.30_+0.735 31.4_+ 4.3 6.83 -+ 2.14 40.4_+ 12.0 7.50+0.52 ~,5 44.3-+ 3.1 </p><p>' Average weight often dried specimens Mean force _+ one unit standard deviation </p><p>3 Force directed laterally across mesothorax unless otherwise specified 4 Force applied to prothorax 5 Force applied dorso-ventrally </p><p>247.0-+ 47.0 139.0_+38.0 123.0_+ 18.0 </p><p>32.8+ 9.7 284.0 + 73.0 </p><p>means of rapid scraping movements of the legs. During these attacks the wasp continues walking over the surface in its usual manner. However, when attacked en masse by ants, the wasp simultaneously increases locomotor rate and ant removal activities. In natural surroundings, these two responses are sufficient to allow the wasp to escape from attack by large numbers of formicids, but in the confinement of laboratory containers which housed ant colonies, the wasp was rapid...</p></li></ul>


View more >