aJ-13 29 April 1988 PROC. ENTOMOL. SOC. WASH. 90(2), 1988, pp. 204-206

NOTES ON THE BIOLOGY OF CAENOCEPHUS ALDRICHI BRADLEY (: )

PAUL E. HANSON AND JEFFREY C. MILLER

Department of Entomology, Oregon State University, Corvallis, Oregon 97331-2907 (PEH present address: Escuela de Biologia, Universitad de Costa Rica, Ciudad Univer- sitaria, Costa Rica).

Abstract. —The larva of Caenocephus aldrichi Bradley was found boring in stems of Holodiscus discolor (Pursh) (Rosaceae). This represents the first host plant record for the genus. The rarity of C. aldrichi is demonstrated by the finding of 12 infested stems in 30 hours of searching patches of H. discolor. In Oregon, C. aldrichi appears to be univoltine, with adults present from early May to early July. Observed mortality factors include parasitism by Pteromalus sp. (Hymenoptera: ) and possible predation by birds. Key Words: Caenocephus, Cephidae, Holodiscus, host plant.

The host plants of stem (Hyme- dlekauff 1952, 1969), but specimens are very noptera: Cephidae) are fairly well known at rare in collections (D. R. Smith, pers. the generic level. In North America there comm.). The purpose of this paper is to re- are six genera in the Cephidae, all in the port the discovery of the host plant of C. Cephinae. Species with known hosts in the aldrichi and present brief notes on its bi- Cephini (, 1 sp.; , 2 spp.; ology. and , 1 sp.) are stem borers in grasses as larvae. Species with known hosts MATERIALS AND METHODS in fl-le (Thenocephus, 1 sp.; Ha.- - A sea:ch for the larval host 1.-L..nt w--3 be- tigia, 6 spp.; and , 4 spp.) may bore gun by limiting field observations to species in the stems of various angiosperms (Smith of Rosaceae, Fagaceae, and Salicaceae where 1979, 1986). Larvae of Hartigia species ap- geographic distributions were within the re- pear to be restricted to the Rosaceae (roses corded distribution of adult C. aldrichi. Be- and blackberries) while species of Janus uti- cause species in the closely related genus lize various woody dicots (Hanson 1986, Hartigia are confined to Rosaceae, special Middlekauff 1969). However, the host plant attention was devoted to members of this of Caenocephus aldrichi Bradley was here- family. tofore unknown. The only other species in Field sampling was conducted by looking this genus is the Palearctic C. lunulatus for severed stems for a set period of time (Strobl) and its host plant is also unknown that varied depending on plant density. (Schedl 1981). Typically, several thousand stems were ob- Limited data are available on the biology served in an hour of searching. Stems with of adults. The species has been recorded severed tips were clipped and set aside until from British Columbia, Washington, Idaho, the sampling period was over. The severed Oregon, and California (Bradley 1905, Mid- stems were then partially dissected to verify VOLUME 90, NUMBER 2 205 the presence of an larva. The "oc- H. discolor is more than an incidental host cupied" stems were placed individually in record. glass tubes (20 cm x 2.5 cm) plugged with Species of Cephidae either sever the stem cotton. Tubes containing infested stems were as an adult at the time of oviposition or as kept outside for three months (November— a larva when stem boring is initiated. The January) and were then brought into the stems harboring larvae of C. aldrichi were laboratory (22 ± 2°C) for observation of severed in a manner which resulted in a line adult emergence. of girdling that was very smooth and with- When cephid larvae were discovered in out jagged marks. In comparison, the girdle stems of Holodiscus discolor (Pursh), addi- line on severed shoots containing Janus ru- tional field observations and collections were fiventris (Cresson), in which the adult female concentrated on this species. Infested stems girdles the stem with her ovipositor, is rough were collected July—October of 1986 from and exhibits jagged marks around the cir- the following localities in Oregon: Benton cumference of the stem (Hanson 1986). In County (Corvallis and Marys Peak); Curry the absence of direct observations this sug- County (15 miles east of Port Orford); and gests that in C. aldrichi the shoot is probably Wasco County (13-16 miles west of Dufur). girdled by the larva. Adult stem sawflies, a larva, and parasit- Only the wider stems of H. discolor were oids obtained from this study were depos- infested. Stem width of all growing shoots ited in the Systematic Entomology Labo- ranged from 0.9 mm to 7.0 mm in diameter. ratory, Department of Entomology, Oregon Infested stems ranged from 2.8-7.00 mm in State University. diameter when measured at mid-length of the infested portion. RESULTS AND DISCUSSION The pattern of the tunnel in the stem il- Adult cephids reared from stems of H. lustrates the behavior of the larva. We noted discolor, ocean spray, were identified as C. patterns in tunnelling that suggest the fol- aldrichi using keys published by Middle- lowing: the newly hatched larva after gir- kauff (1969) and comparison with speci- dling the shoot tunnels toward the base in mens identified by D. R. Smith in the Or- the cambial zone for a short distance (1-3 egon State University insect collection. mm), the larva then turns back (upwards) The number of infested shoots was ex- to the girdled apex, tunnels to the midsec- tremely low at all sites where C. aldrichi tion of the shoot, and then bores down the were found. Several stands of H. discolor center toward the base of the stem. At the were searched for a total of 16 hours without initiation of the downward tunnelling the yielding any infested stems. However, in larva is consuming almost the entire interior Wasco County (13-16 miles west of Dufur), of the shoot and packing frass in the vacated 8 infested stems were found after a total of tunnel. 6 hours of searching. Only 4 infested stems Patterns in the length of infested stems were found at the other sites after 10 hours suggested that a second severing of the shoot of searching. Thus, a total of 12 infested occurs. Stems collected in July had tunnels stems was found from all sites after 30 hours averaging 5 cm in length (severed apex to of searching. The difficulty in finding stems location of larval head), whereas stems col- of H. discolor infested by C. aldrichi could lected in October had tunnels less than 1 be a result of not observing the primary cm in length (newly severed apex to the base host, which then would remain to be dis- of the prepupal chamber). These observa- covered. However, the relative abundance tions suggested that the last instar severs the of infested stems at the Wasco Co. site sug- stem again, before forming a pupal cham- gests that the association of C. aldrichi with ber. Thus, the portion of the stem contain- 206 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON

ing the initial tunnelling falls off the plant. Cephidae, Middleton (1917) noted that in Similar behavior has been documented in Hartigia the lateral area of the basal anal other cephids (Middlekauff 1969). lobe is setose, whereas in Janus this area is Observations of emergence in the labo- bare. In the single preserved specimen of C. ratory indicated that adults exit the shoot aldrichi the lateral area of the basal anal lobe by chewing through the apical frass plug, is setose as in Hartigia. Additional collec- rather than through the stem as in other tions of larval C. aldrichi are needed for Hartigiini. • studies on variation in larval morphology. We observed some larval-pupal mortality Based on our limited observations of this factors. A larva in one of the 12 infested rarely collected species it appears that stems contained larvae of a gregarious Pter- Caenocephus is more similar to Hartigia omalus sp. (Hymenoptera: Pteromalidae; than to Janus. The following observations identification by P.E.H.). These parasitoid suggest this relationship: host plant is in the larvae were reared and produced six females Rosaceae, initial stem-girdling is likely per- and two males (on July 8) one week after formed by the larva, and the basal anal lobe collection. Two of the stems with tunnels is setose in the larva. had irregular holes in the region of the pupal ACKNOWLEDGMENTS chamber. The holes exhibited peeled edges and the chamber lacked a cephid larva, sug- We are grateful to D. R. Smith, System- gesting possible mortality by bird predation. atic Entomology Laboratory, USDA, In three of the infested stems the cephid Washington, D.C., for sharing his personal larva had already died from unknown caus- records on C. aldrichi with us. This is tech- es and in four of the infested stems larvae nical paper no. 8365, Oregon State Univer- died after collection (one was preserved), sity Agricultural Experiment Station. probably because they were collected too LITERATURE CITED early in the season. Thus, we obtained only Bradley, J. C. 1905. Caenocephus in America. Can. two adults (both female) out of 12 infested Entomol. 37: 363-364. stems. Hanson, P. E. 1986. Biology of Janus rufiventris (Hy- Data from specimens in the Oregon State menoptera: Cephidae). Ann. Entomol. Soc. Am. University insect collection indicate that 79: 488-490. adult G. aldrichi are active from early May Middlelcauff, W. W. l 2. Notes or. two species of California stem borers. Pan. Pac. Entomol. 23: at lower elevations in the Willamette Valley 108-109. (about 100 m), to early July at higher ele- . 1969. The cephid stem borers of California vations on Marys Peak (about 1000 m). (Hymenoptera: Cephidae). Bull. Calif. Insect Surv., Our field observations on the state of larval vol. 11, 19 pp. development in the field indicated that this Middleton, W. 1917. Notes on the larvae of some Cephidae. Proc. Entomol. Soc. Wash. 19: 174- species is univoltine in Oregon. 179. The only other species presently placed Schedl, W. 1981. Zur Nomenklatur, Morphologie and in Caenocephus is the Palearctic C. lunu- Verbreitung der Halmwespe Caenocephus lunu- latus (Strobl), formerly known as C. jakow- latus (Strobl, 1895) comb. nov. (Hymenoptera, lejfi Knonow (Schedl 1981). Because Hol- Cephidae). Nachrichtenblatt der Bayerischen En- tomologen 30: 28-31. odiscus is absent from the Old World, the Smith, D. R. 1979. Superfamily , Family host plant of this species must be different Cephidae, pp. 133-137. In K. V. Krombein, P. D. from that of C. aldrichi. Hurd, Jr., D. R. Smith, and B. D. Burks, eds., With the discovery of a host plant con- Catalog of Hymenoptera in America North of taining the larva of C. aldrichi, it will be Mexico, vol. 1. Smithson. Inst. Pr. Wash. D.C. 1986. The berry and rose stem-borers of the possible to compare the morphology of im- genus Hartigia in North America (Hymenoptera: mature Caenocephus with that of other gen- Cephidae). Trans. Am. Entomol. Soc. 112: 129- era of Hartigiini. In a key to the larvae of 145.