NATURAL HISTORY OF THE WORKERLESS INQUILINE ANT POGONOMYRMEX COLEI ( HYMENOPTERA : FORMICIDAE ) * BY

At least 10 workerless inquiline ant species are known from North America (Francoeur 1968, 1981; Wilson 1971, 1976; Talbot 1976; Buschinger 1979; DuBois 1981; Snelling 1981), most only from original collections. In this paper I present field and laboratory observations of Pogonomyrmex colei Snelling a new, apparently workerless, inquiline ant inhabiting a colony of Pogonomyrmex rugosus. P. colei appears to be a very rare species" extensive searching of the type locality for 4 yr has resulted in discovery of only a single colony. Nonetheless, observations on this colony provide insight into several important aspects of inquiline ant biology. P. colei is also of interest since it is the second apparently workerless congeneric inquiline inhabiting colonies of P. rugosus. Cole discovered the first inquiline species, Pogonomyrmex anergismus, near Silver City, New Mexico apparently prior to any major flight since he exposed "more than one hundred" inquiline reproductives upon opening the host nest (Cole 1954, 1968). Since host species mating flights occur soon after rain during mid to late summer (Htlldobler 1976; Rissing personal observation), it seemed reasonable to suspect P. anergismus responds to the same environmental cues for mating as does its host. Accordingly, in an effort to rediscover P. anergismus, I routinely checked most P. rugosus nests on a 25 ha study area in Boulder City, Nevada for flight activities and possible presence of inquilines during late summer 1978 and 1979 (study area described in Rissing 1981). P. colei was discovered during this effort.

P. colei appears to be a very rare species" extensive searching of the type locality for 4 yr has resulted in discovery of only a single colony. Nonetheless, observations on this colony provide insight into several important aspects of inquiline ant biology. P. colei is also of interest since it is the second apparently workerless congeneric inquiline inhabiting colonies of P. rugosus. Cole discovered the first inquiline species, Pogonomyrmex anergismus, near Silver City, New Mexico apparently prior to any major flight since he exposed "more than one hundred" inquiline reproductives upon opening the host nest (Cole 1954(Cole , 1968. Since host species mating flights occur soon after rain during mid to late summer (Htlldobler 1976; Rissing personal observation), it seemed reasonable to suspect P. anergismus responds to the same environmental cues for mating as does its host. Accordingly, in an effort to rediscover P. anergismus, I routinely checked most P. rugosus nests on a 25 ha study area in Boulder City, Nevada for flight activities and possible presence of inquilines during late summer 1978 and 1979 (study area described in Rissing 1981). P. colei was discovered during this effort. *Manuscript received by the editor June 7,1983 Psyche [VOW. 90 OBSERVATIONS Mating Activities and Season. Five P. colei males were collected at a single P. rugosus nest during the morning of 13 August 1978; a series of thunderstorms and rain had occurred 12 hr earlier. Frenzied host worker activity suggested a mating flight or similar activity occurred immediately prior to my arrival. No flights of either species occurred at any nearby P. rugosus nests observed simultaneously.
I observed a complete inquiline and host flight at this same nest on 15 September 1978 following an extensive rain storm the preceding day. Flights were occurring at 2 of 23 nearby P. rugosus nests; P. colei was not found at any other nest. Mating activities began w.ith accumulation of several hundred host workers in and around the nest crater. These workers pugnaciously defended the area throughout both flights as is typical during P. rugosus flights (Rissing, personal observation). As ground and air temperatures increased male P. colei climbed to the crater and were soon joined by much larger females. While both sexes of P. colei are winged, mating occurred at the nest entrance followed by females flying from the area and males re-entering the nest. Such in situ mating is common in rare ant species apparently due to very low probability of reproductives finding individuals from other nests with which to mate (Wilson 1963). Following copulation and departure of P. colei females, male and female P. rugosus flew from the crater as the temperature continued to climb. Reproductive forms of P. rugosus fly to a site away from the nest and copulate there (Htlldobler 1976). Mating activities of host and inquiline were separated by at least 30 min and, perhaps more importantly, 3 C ground temperature (Table 1). Reproductive forms of each species were seen occasionally in the nest entrance during the mating activity of the other. On at least one occasion, P. colei males tried unsuccessfully to mount a P. rugosus female. During this flight I observed no differences in behavior of host workers to host or inquiline reproductives. P. rugosus workers frequently encircled copulating pairs of P. colei and frantically ran around them, although they never interfered. During 1979 routine  Colony foundation. Ten newly mated P. colei females from the 15 September 1978 flight were placed into a 7.5 m high flight enclosure made of plastic sheeting and permitted to fly. Subsequent to this all females removed their wings but did not dig burrows when placed into laboratory nest boxes containing moist sand. Five of these dealate inquilines were transferred to 5 laboratory nests containing only newly mated P. rugosus queens. These P. rugosus queens had been collected one week earlier at a mating site 3.2 km from the host nest making it unlikely that they were related to the host colony.
Four of these laboratory nests contained a single, mated dealate P. rogusus queen; the fifth contained two P. rugosus queens. The P. colei queen added to the nest with two P. rugosus queens was immediately attacked and removed from the glass tube occupied by the P. rugosus queens. Of the P. colei queens added to the single queen P. rugosus colonies, one was found dead within several hours (decapitated), and the other was found dead (entire) 5 d later. The other two P. colei queens lived peacefully along side the P. rugosus queens for at least a month. During this time frequently observed the P. colei queens grooming the P. rugosus queens; P. rugosus queens did not reciprocate. These last two colonies ultimately failed during (or possibly in response to) transportation from Boulder City to Seattle.
Five other newly mated, dealate P. colei queens were released in the field at the entrance of large, active P. rugosus colonies near the host nest. Inquilines were always removed immediately from the nest by one or more workers and dropped several meters from the crater. The P. colei queens made no attempt to re-enter these nests following removal.
Psyche [Vo. 90 DISCUSSION Repeated (and continuing) attempts to find P. colei or P. anergismus around Boulder City, NV, or Globe, AZ, where a single P. colei male has been collected (Snelling 1981) have yet to be successful. Nonetheless, observations of P. colei from the type nest in Boulder City provide insight into several questions of general inquiline biology including possible method of inquiline entry into host colonies and fate of host queen.
Inquiline entry into host colonies. Newly mated P. colei queens are accepted into week old workerless host nests in the laboratory, while they appear incapable of entering established host nests in the field (see above). Similar observations have been made in laboratory experiments with the inquiline Plagiolepis xene and its host, Plagiolepis pygmaea (Passera 1964). This suggests that at least some inquiline species enter a host colony at the founding stage prior to production of any workers. That this may occur in the field is supported by discovery of a workerless inquiline queen (Strumigenys xenos) in an incipient host colony containing one queen, brood and a single worker of Strumigenys perplexa (Brown 1955).
If entry into host colony commonly occurs at host colony foundation in some species of inquilines, overlap with host species flight season would be advantageous. Since all nests of a given species in a locality tend to have a longer "flight season" than any single nest (e.g. for P. rugosus see Htilldobler 1976), the inquiline might further be expected to lengthen its flight season relative to that of its host colony to take advantage of the entire flight season and availability of founding nests in its locality. The extended flight season of P. colei relative to that of P. rugosus may occur for these reasons. Similarly, occurrence of P. anergismus reproductives during mid September in the type nest reported by Cole (1954,1968) may also indicate inquiline-host reproductive overlap.
Fate of host queens. Simultaneous production of host and inquiline reproductives during the 1978 flight (Table 1) strongly suggests coexistence of host and inquiline queen(s) at that time. Continuing existence of the host colony until at least September 1982 further substantiates this. Estimates of maximum longevity of worker ants is 1-2 yr (Rosengren 1971, Brian 1972, Nielsen 1972. Further, there has never been a reported case of queen adoption in any Pogonomyrmex species. For the host colony to have a normal foraging group size in 1982, the host queen must have been alive during the 1978 and 1979 inquiline flights. Although inquiline-host coexistence has been regarded as a "primitive" inquiline trait (Wheeler 1933, Haskins andHaskins 1964), it offers the obviously adaptive advantage of a continuously renewed host worker force for the inquiline. Coexistence occurred in the type nest of P. colei and appears common in other workerless inquiline species where information regarding fate of host queen(s) is available (Table 2). Host queen elimination does occur in at least two well documented cases (Table 2). Wilson (1971) suggests such behavior may develop in short-lived inquiline species; inquiline longevity, however, may be more of an effect than a cause of this behavior. Host queen elimination may be adaptive only when inquiline entry is gained by a queen after development of a host worker force. Host workers appear to be the primary defense against inquiline entry in many colonies. In order to be accepted by host workers, it may be necessary for the prospective inquiline queen to first render the prospective host colony queenless. In those cases where host queens are known or highly suspected of being eliminated (Table 2), the inquiline queen enters an established colony containing workers. In at least one of these cases, Epimyrma vandeli, the inquiline must fight with host workers until she is able to kill the host queen.
Recent discovery that E. vandeli is a degenerate slave-maker Psyche [Vol. 90 Brown 1959Haskins and Haskins 1964Elmes 1974a, 1978Brian 1972 this study Wheeler 1910, Crawley 1912, Donisthorpe 1915, Creighton 1950Stumper 1950+, Kutter 1969Buschinger 1965Smith 1942, Buschinger 198 Vandel 1927 Stumper and Kutter 1951Kutter 1945+, 1969+ Creighton 1950Buschinger 1979 Monomorium  Passera 1964Passera , 1966Passera , 1972 Aporomyrmex Plagiolepis ampeloni vindobonensis survives Faber 1969+ *Presence of host queen(s) determined by presence of host reproductives +Cited in Wilson 1971) ( Buschinger 1981, Buschinger andWinter 1982) may explain this behavior which is rather unusual among most other inquilines (Table 2). Only the extreme inquiline Teleutomyrmex schneideri is known to enter established host nests without having to eliminate host queens; these inquilines may produce a substance highly attractive to host workers (reviewed in Wilson 1971). Comparison with P. anergismus and other workerless inquilines. P. colei may represent an intermediate form between its host P. rugosus and the closely related workerless inquiline P. anergismus (for a complete discussion of morphological differences see Snelling 1981). Discovery of P. colei adds the genus Pogonomyrmex to a growing list of ant genera with more than one workerless inquiline species (Table 2). Such "concentration" of inquilines into a few genera may occur either due to non-random search by myrmecologists (P. colei was discovered during an intentional search for Pogonomyrmex inquilines) or because certain genera are more likely to give rise to inquilines. The basic biology of the inquilinerich genera, however, is quite variable suggesting several evolutionary routes may lead to workerless inquilinism. The genus Leptothorax, for example, has small, ephemeral colonies subject to slave raids from numerous species and has given rise to several closely Psyche [Vol. 90 related Epimyrma inquiline species, themselves degenerate slavemakers (Buschinger 1981, Buschinger andWinter 1982). Myrmica, on the other hand, has larger colonies and many species that are highly polygynous (Brian 1972;Elmes 1974a,b); this genus has given rise to at least 7 workerless inquiline species: Myrmicafaniensis (van Boven 1970), Myrmica hirsuta (Elmes 1974a(Elmes , 1978, Myrmica lampra (Francoeur 1968(Francoeur , 1981, Myrmica myrmecophila (Bernard 1968), Myrmica quebecensis (Francoeur 1981), Sifolinia karavajevi (Kutter 1969) and Sifolinia laurae (Brian 1972), the Sifolinia species likely being congeneric with the other Myrmica species (Elmes 1978). Monomorium is similar with polygynous species (Dennis 1938, Cole 1940, Gregg 1945) and a number of congeneric inquilines (reviewed in Wilson 1971, see also Talbot 1979 andDuBois 1981). These inquiline species may have evolved through a process of some polygynous host queens acquiring the trait of laying only reproductive eggs (Buschinger 1970, Elmes 1978. To this list must be added the genus Pogonomyrmex whose basic biology is unlike any of the above three host genera. Colonies are substantially larger than Leptothorax, Myrmica or Monomorium (Lavigne 1969, Rogers et al. 1972, Whitford et al. 1976, MacKay 1981, strictly monogynous (Lavigne 1969, H/511dobler and Wilson 1977, MacKay 1981, with no slave-making or similar behavior in any species. Evolutionary processes giving rise to P. colei and P. anergismus are likely different from those that have given rise to the Leptothorax, Myrmica or Monomorium inquilines. Certainly, the idea of multiple evolutionary pathways leading to workerless inquilinism is not new (see Wheeler 1919, Buschinger 1970, Wilson 1971. Continued study and search for workerless inquilines can only serve to clarify this challenging evolutionary process.