Sound production and cannibalism in larvae of the pine-sawyer beetle Monochamus sutor L. (Coleoptera: Cerambycidae)

JONAS VICTORSSON & LARS-OVE WIKARS

Victorsson, J. & Wikars, L.-O.: Sound production and cannibalism in larvae of the pine- sawyer beetle Monochamus sutor L. (Coleoptera: Cerambycidae). ILjudalstring och kanni' balism hos larver av tallbock Motrochamus sutorL. (Coleoptera: Cerambycidae).1 - Ent. Tidskr. ll7 (l-2):29-33. Uppsala, Sweden 1996. ISSN 0013-886x.

e-sawyer beelle, Monochamus sulor L. (Coleoptera: Ceram- rping sounds, easily heard up to l0 meters away. Larvae were tion tunnel in sapwood of recently fire-killed spruce and pine ed outwards, and the sound was probably produced by the scratching of their mandibles against the bark. Since the larvae produce this sound in spite of the risk of attracting parasites and predators such as wood-peckers, this behaviour most likely has an adaptive value. We suggest that sound production helps a larva to secure resources around its hibernation tunnel for its own development by keeping away other potentially competitive larvae. That the larvae are cannibalistic was demonstrated during a trial perfor- med in petri dishes filled with sawdust. Any larva ignoring the sound signal could therefore face a real threat of being killed.

victorsson, J. & Wikars, L.-o., Department of Zoology, IJppsala university, villav. 9, s-752 36 Uppsala, Sweden. (Correspondence: Wikars' L.-O.)

Strange chirps in burned forests During visits in some forest areas burned over their heads projecting outwards against the bark the past year in Central Sweden (Fig. 1), a (Fig. 2). They presumably produced the sound by regular, chirping sound could be heard from fire- scratching their mandibles against the bark. As killed spruce (Picea abies) and pine (Pinus the sub-cortical tissue of a tree is more or less silvestris) trees. The sound was heard during consumed after attack the space beneath the bark sunny weather at the end of June 1995, and is air-filled and can act as a resonance chamber. consisted of three to four chirps (sounding like a No feeding was observed in connection with finger-nail being scratched repeatedly over a sound production. Hence we can exclude the comb), together lasting for about 2 sec. The possibility that it is a by-product of feeding. sound was repeated at intervals of five to ten In central and northern Sweden, the pine- seconds or longer. Several sources of the sound sawyer beetle has a two year life-cycle. During the were heard in different trees simultaneously, and first summer of development, larvae live entirely the sound could be detected at a distance of more sub-cortically whereas in their second summer than l0 meters. By exactly locating the sound larvae make a hibernation tunnel into the sapwood and removing the bark, larvae of the pine-sawyer (Fig. 2). However, extensive feeding excursions beetle, Monochamus sutor L. (Coleoptera: are still made under bark adjacent to the hiberna- Cerambycidae), were invariably found. These tion tunnel. After passing a second winter, larvae were most likely one year old, as colonisation by pupate early in the following summer. The pupa is pine-sawyer beetles take place immediately after formed at the innermost end of the now U-shaped' fires, at least when fires occur during July and 5-10 cm deep and 20-25 cm long tunnel. August (Forsslund 1934 and personal observa- Emergence of adult beetles take place in July to tions). The larvae were positioned in the sap- August the same year, the third summer of wood. each inside its hibernation tunnel, with development. 29 Jonas Victorsson & Lars-Ove Wikars Ent. Tidskr. I l7 ( I996)

Fig. l. Burned pine forest I 5 months after a fire. Pines showing signs offoraging by black woodpeckers Drlocopus martius and three-loed woodpeckers Picoides tridactylus. The woodpeckers have been searching for larvae of the pine'sawyer beetle Monochamus sutor and olher beetles. Brat(orsheden, Vcirmland, south-central Sweden, October :,993. Photo: Sven-Ake Bergtind.

Brdnd tallskog I 5 mdnader efter brand. Tallar med spdr efter fcidosdk av spillkrdka och tretdig hackspett. Hack- spetiarna har sdkt efter larver ov tallbocken och andra skalbaggar Brattfors brandfcilt, Viirmland, oktober 1993.

What is known about larval sound ..'.:.'. -..'...... '.... production in beetles j.' :-.'.: ':';r j-'-.'::'--'-'.':.: :'.'." Sound production in beetle larvae, excluding that produced by feeding, is very rare (Crowson I 98 I ). The first known example is of the gregariously Ii- ving chrysomelid beetle Paropis sp., whose larvae produce sound by tapping their abdomens against the leaves of Eucalyptus on which they live (Meyer-Rochow 1972, cited in Crowson l98l). The behaviour is assumed to synchronise a defen- Fig. 2. Iarva the pine-sawyer of beetle Monochamus sive reaction against predators. sutor in the some position as when sound is made. A) In Japan, Izumi et al. (1990) reported larval bark, B) air-filled space, C) wood, D) hibernation tun- nel. sounds in the close relative Monochamus alterna- rrzs Hope and gave a detailed description of the Larv av tallbocken i samma ldge som ncir ljudalstringen sound. They were similar in strength to those we sker. A) bark, B) luftfyllt mellanrum, C) ved, D) dver- heard (audible up to 15 meter away), and were vintringsgdng. recorded as having a sound pressure of40 dB. The

30 Ent. Tidskr. I 17 (1996) Sound production in larvae of Monochamus sutor sound of this species consisted of sound-units, or Tab. l. Results of predation erperiment with pairs o{ Monochamus sutor in petri dishes chirps, that had a length of 0.04 seconds and were pine-sowyer larvae with sawdust. The weight of larvae al the start of the repeated with an intersound interval of 2 seconds. experiment and the outcome are presented (* that the sound was produced by They established cannibalised larvae). the scratching of mandibles against the beetles gallery walls, but gave no explanation why larvae Resultat av predationsexperiment med par av tallbock- produced sound. slarver i petriskdlar med sdgspdn. Larvernas vikt vid Letler (1992) described the sounds of two other fiirsdkets bdrjan saml utfallet presenteras (* iiten larv). cerambycid larvae, Niphona pecticornis Muls' and Ceroplesis aetuans 01. These bored inside Outcome of wood and presumably made their sound by Weight of larger Weight of smaller (g) larvae (g) interaction knocking their heavily sclerotised mandibles larvae downward against the wall of the air-filled larval 0.0905 0.0295 * cannibalism burrows. Both these species belong to the 0.0699 0.0444* cannibalism subfamily Lamiinae, as does Monochamus. Leiler 0,3414 0.0420* cannibalism cannibalism (1992) suggested that the vertical type of head of 0.0750 0.0334* 0.0t t4* cannibalism lamiinae larvae could be especially well suited for 0.0'712 0.0'726 0.0365 no cannibalism production. He speculated that the function sound 0.1242 0.0451 no cannibalism synchronise hatching. of the sound would be to 0.0951 0.0426 no cannibalism However, in Monochamus this is not very prob- able as hatching does not occur until a year after sound production. Izumi et al. (1990) found that sound production by individuals decreases during M. alternatus and two meters above ground. The lower limit of larval development. During later larval stages, their distribution on trees normally coincided with when the sub-cortical tissue has been consumed. burrows of the buprestid Melanophila acuminata larvae live progressively more inside the wood DeGeer, which prefers to burrow into the basal than under bark. part of trees. Larvae were then kept in plastic jars If sound is produced and has some functional at 8oC for five days. To prevent dessication, moist, significance. it ought to be percieved. Many green moss was put in the jars. The larvae remai- wood-living beetle larvae (e.g. Buprestidae and ned active and mobile until the onset of the trial. Cerambycidae) have chordotonal organs attached They ranged in weight between 0.01 1 and 0.341 g to the pleural discs in the abdomen (Saliba 1972). with an average of 0. I 18 g (Tab. l). The difference The hearing function of these organs has been in weight between the two larvae in trials ranged revealed through dissections and comparative between 0.025 and 0.299 g. The large variation in anatomical studies tn Monochamus confusor weight could be due to the fact that the time for Kirby, a North-American species closely related larval development is variable in this species to M. sutor (Hess 1917). Thus sound production in (Forsslund 1934). To prevent larvae from esca- M. sutor may serve some intraspecific function. ping, a plastic lid with holes for ventilation was fitted on each dish. The larvae were then left to interact for six to eleven days at room temperature and in complete darkness. Monochamus larvae can be cannibalistic In five out of eight petri dishes, the larger larva To test if larvae of Monochamus sutor were can- predated the smaller (Tab. 1). Remains of larvae, nibalistic. a trial was conducted in which two such as mandibles or whole head-capsules found larvae were placed in wetted sawdust in petri dis- in the sawdust, was taken as evidence of hes (57 mm diameter). Larvae were collected predation. from a burned forest stand at Prlistvallen in the Unfortunately the larvae were not determined province of Dalarna (61"22' N, 14"35' E, 520 m to species before the experiment. After the experi- a.s.l.) at the end of June. They were found under ment it was possible to determine l2 out of 16 spe- the bark of burned spruce trees at between a half cimens to species, all of which were M. sutor. Gi- 3r Jonas Victorsson & Lars-Ove Wikars Ent. Tidskr. l17 \1996)

ven the habitat and position on the trees where the larva that ignores the territorial behaviour of a larvae were found, we think it is a safe assumption signalling larva faces a potential danger of being that the other four larvae were also M. sutor. seriously injured or killed. Hellrigl (1971) also noted that M. sutor has a Sound can be used as a cue for larvae to avoid a cannibalistic tendency. By examining logs for certain part of the tree-trunk. Hellrigl (1971) eggs, larval burrows and adult exit-holes, Rose speculated that larvae of M. sutor could locate (1957) estimated that cannibalism was the largest their conspecifics by their chewing sounds mortality factor in M. scutellatus, a North-Ameri- produced during feeding, and thereby avoid being can species with very similar biology to that of M. cannibalised. Saliba (1972) showed in laboratory sutor;both among first-, second- and third-instar experiments that cerambycid larvae of several larvae of M. scutellatus, cannibalism was the main species were able to avoid each other by moving mortality factor (together more than 80 Vo were away from chewing sounds. Larvae approaching killed). Woodpeckers caused 12 Eo mortality, each other turned away at sharp angles well before mainly in the third and fourth larval instars (when meeting, and this behaviour could be manipulated the larvae had bored into the wood). Mortality by mimicking the sounds artificially. caused by desiccation and parasites as well as by It is probably safer for M. sutor larvae to invertebrate parasitoids and predators was con- produce sound while in a hibernation tunnel rather sidered less important. than under bark. In the tunnel, they can quickly move backwards and protect the tunnel entrance with their sclerotised head and well-developed mandibles. This is probably a sufficient defence Why sound production? against other larvae, though probably not against For a beetle larva living under bark, it seems woodpeckers. hazardous to emit any sound. That hymenopteran Whether sound production definitely functions parasitoids use sound to find their host is doubtful as a territorial defence signal remains to be shown (Vinson 1976, Alphen & Vet I 986), even if vibra- experimentelly. A possible next step would be to tions through the substrate probably can guide mimic the sound and observe behavioural respon- them (Matthews & Matthews 1978). However, it ses directly. Another approach would be to see if is likely that sounds can guide other enemies, like sound production differs between small and large woodpeckers and other insect parasitoids, to the larvae as well as to see how the density of larvae larva. Therefore this recorded sound production and amount of resources under bark affects this must have some other, strong, adaptive value. behaviour. Forsslund (1934) showed that M. sulor needs sub-cortical tissue to complete its development. Sapwood is not enough. As the sub-cortical tissue Acknowledgement normally becomes totally consumedby Monocha- Bengt Ehnstrtim, Robert Paxton and Stefan As gave mus larvae one year after attack (Triigirdh 1929, comments and valuable criticism. R. Paxton made Rose 1957), increasingly severe competition for linguistic corrections. Susanne Godow and Gunilla Ols- the remaining resource arises. We observed that son assisted in the field. Thanks also to Sven-Erik M. sutor larval burrows under bark become very Green, Viisteris stifts skogsfdrvaltning, for initiating long (30-50 cm), which suggests that one year old and performing the prescribed burnings at Siirnstugan and Priistvallen, and thereby for the creation good larvae have to move considerable distances to find of habitats for M. sutor and other insects benefitting from enough food. forest fires. Perhaps sound production by Monochamuslar- vae is some kind of territory defence mechanism. Thereby a sound producing larva could secure resources for itself and increase the possibility of References acquiring enough energy for pupation. we Since Alphen, van J.J.M. & Vet, L.E.M. 1986. An evolutionary showed in the petri dish experiment that large approach to host finding and selection. - In: Waage, J. larvae can indeed cannibalise smaller ones, sound & Greathead, D. (eds). Insect parasitoids. p. 23-61. production would be a case ofhonest signalling. A London (Academic Press)

5Z Ent. Tidskr. ll7 (1996) Sound production in larvae of Monochamus sutor

Crowson, R. A. 1981. The biology of the Coleoptera. Sammanfattning London (Academic Press). Vid besiik pi tvi brandfiilt i slutet av juni 1995 vid Forsslund, K.-H. 1934. Tallbockens (Monochamus sutor Siirna och Orsa i norra Dalarna (briinda i juli L.) upptr2idande pi brandfiilt i norra Sverige somma- 1994) hdrdes regelbundna skrapliiten frin brand- ren 1933. - Svenska skogsvirdsf6rbundets Tidskrift dridade granar och tallar. Det liit ungefiir som nAr 1934(t-2): 23-38. man drar fingernageln flera ginger

Rf,ttelser och fiirklaringar till Sveriges myror

Redaktciren har med berdmviird ihiirdighet civerty- gorlunda siiker pA artbestamningen. Gingen vid gat mig att narmare fijrklara hur en myra kan ha ett bestiimningsarbetet er sAledes att man riiknar anta- halvt borst, vilket pistis i punkt 16 i bestiim- let borst pA lAt siiga fem individer (givetvis frin ningstabellen for Formica pfl sidan 96 i min arti- samma bo/samhiille), reknar ut medelvardet och kel Sveriges myror i ET nr 3 1995. Det skulle anviinder detta niir man nycklar i tabellerna. naturligtvis ha stett "I snitt fiirre iin 0,5 utstiende Redaktdren har ocksfl pipekat fel i tabellen pA borst ..." respektive "I snitt minst 0,5 utstaende sidan 96. Punkt I I fcirsta alternativet ska frira vi- borst ..." Ndr man bestdmmer myror dr det en stor dare till punkt 12 (ej 13 som det stir) och punkt 12 fdrdel att ha tillging till flera individer frin sam- andra alternativet ska h[nvisa till l3 (ej l2). ma samhAlle och fcir vissa Formica och Lcsius iir det alldeles ncidviindigt om man ska kiinna sig ni- Per Douwes

55