A COMPARATIVE-STUDY OF THE SYMBIOTIC RELATIONSHIPS BETWEEN BEETLES OFTHE GENUS CREMASTOCHEILUS (COLEOPTERA, SCARABAEIDAE) AND THEIR HOST ANTS (HYMENOPTERA, FORMICIDAE)

The North American species of the genus Cremastocheilus are reviewed. These belong to 5 subgenera, Macropodina, Trinodea, Anatrinodia, Myrme cotonus, and Cremastocheilus. Taxonomic changes are: The inclusion of Cremastocheilus nitens and C. chapini in the subgenus Cremastocheilus rather than Myrmecotonus. Also Anatrinodia is elevated to subgeneric s tatus. A key to the subgenera is provided, as is a key to the species of the 5 subgenera, recognizing that the 35 species in the subgenus Cr emastocheilus are in need of revision. A critical review of the host r ecords, geographic distribution, and ecology of the Tribe Cremastochei lini (Family Scarabaeidae, subfamily Cetoniinae) is provided. This con tains enormous numbers of new records for both the genera Genuchinus a nd Cremastocheilus both from the literature and from the extensive fie ld work that is reported here for the first time. A summary of the hos t records is presented in tabular form. This table shows the associati on of all species of Cremastocheilus with ants as adults and the larva e either associated with the vegetable material of the ant nests or wi th vegetable material in rodent burrows. Genuchinus is shown to be a g eneral predator on soft bodied insects while the other genera of the C remastocheilini are associated with plants, particularly bromeliads. A detailed study of the external morphology and sexual dimorphism of th e genera Genuchinus and Cremastocheilus is presented. All species of C remastocheilus can be sexed with the naked eye by the difference in th e shapes of the abdominal terminal segments, wherein males have the po sterior border of the last ventral abdominal segment either straight o r slightly bowed, while females have this border broadly rounded. Ther e are other microscopic sexual differences in the structure of the leg s. The rest of the external morphology is also presented, particularly from the point of view of adaptations to either a predaceous or myrme cophilous existence. Particularly adapted for predation are the pointe d maxillae which are used for piercing prey. Particularly adapted for myrmecophily are the mentum, the maxillae, the generally thick exo-ske leton, trichomes on both the anterior and posterior angles of the pron otum, the elytra, and the legs (which are adpated to the nest substrat e of the host and nests. Exocrine glands are described for Genuchinus ineptus and at least 1 species of each of the 5 subgenera of Cremastoc heilus. In general, there are no gland cells nor glandular areas in Ge nuchinus that are comparable to those of Cremastocheilus. The gland ce lls and glandular areas are quite extensive and variable among species of Cremastocheilus. The frontal gland of some Cremastocheilus (strong ly developed in C. castaneus and the C. canaliculatus species group, b ut weakly developed in the C. wheeleri species group) is described for the first time. Because these glands are not found in Genuchinus inep tus, a species with general predatory habits, it is thought that these play a role, as yet unknown, in interactions with ants. The life cycl es of the subgenera of Cremastocheilus are described. The general life cycle entails adult beetles eclosing in ant nests during the summer a nd then undertaking dispersal flights. The adults then enter ant nests and overwinter there, eating ant larvae during the winter. Another di spersal flight occurs in the spring during which the adults mate and e nter ant nests again. The females then lay eggs and the adults die. Th e eggs hatch and the larvae spend 3 instars feeding upon vegetable mat erial in the nests. The larvae then pupate in typical scarabaeine eart hen cells made of fecal material and soil. These eclose in the summer and the cycle is repeated. Variation from species to species is largel y in the timing. Leaving the nest in late summer, mating seems to be t riggered by rainfall in all the species studied. Mating of C. (Macropo dina) beameri takes place in rodent burrows. Males seem attracted to f emales from a distance but the mechanism of this remains obscure. In t he subgenus Trinodia, mating takes place on sandy washes or roadsides where females land. In the subgenus Myrmecotonus, mating also takes pl ace in sandy areas. In C. (Cremastocheilus) mating takes place on sand bars along rivers in the southeastern U.S. and in sand dunes in north eastern U.S. The females dig down into the sand. Males locate these pl aces by some unknown mechanism and then dig down to copulate with the females. Field experiments showed unequivocally that males dig only in to areas occupied by females. No sex-specific sex attractant glands ha ve been located in females so far. Dispersal to ant nests occurs after mating except for C. (Macropodina) beameri which lays its eggs in the rodent burrows and then probably disperses to ant nests. Beetle activ ity going in and out of nests was studied using wire hardware cloth sc reens over entrances to Myrmecocystus nests. The mesh size was such th at the ants could move freely in or out but the beetles got stuck by t heir thoraces. The direction then could be interpreted by the directio n in which they got stuck. By this method, C. stathamae was shown to l eave nests from 23 June to 1 September with a peak on 6 July, just aft er the beginning of the summer rains. Beetles entered nests from June 23 to August 3, however 39% entered on July 16, probably pulsed by the leaving time which was correlated with the rains. Life cycle timing: C. (Macropodina) develop in the nests of Wood rats (Neotoma sp.). Fema les lay about 40 eggs each. The 3 larval instars to pupation take abou t 1 month. Pupae are found from late August to well into September. In other subgenera as well, larvae are found in parts of the nest devoid of ants. The timing is similar in all the subgenera found with ants. Mortality factors: While ants attack Cremastocheilus adults, there is no evidence that they are ever killed by ants nor is there evidence th at ants kill larvae nor hard earthen pupae cases which protect the pup ae. During dispersal flights and mating, the adults are exposed to pre dation and evidence is presented that shows predation by horned toads, spiders, magpies, and tiger beetles. Probably most mortality occurs i n the larval and pupal stages where the beetles are attacked by intern al parasites and fungus. Further mortality is caused by limitation of the food supply during the larval stage. Reentering nests: Females of C. (Macropodina) beameri select specific rodent and other bu