DURATION OF COPULATION IN POANES HOBOMOK ( LEPIDOPTERA HESPERIIDAE ) AND SOME BROADER SPECULATIONS * By

Many aspects of diurnal lepidopteran reproductive biology are still poorly known. Duration of copulation--an awkward phrase which, for convenience, is here symbolized T1u can readily be determined in various species but rarely has been. It is of interest not only as a behavioral dement of possible taxonomic value but also as a highly critical time in the life cycle: copulation is, of course, required for insemination; but copulating individuals, being mutually occupied and encumbered, must often be more vulnerable to predation than separate ones are. Since the act of copulation is vital for contributing genetic material to the succeeding generation but is not performed without risk, one may ask, What fraction of adult life is, on an average, spent copulating? Answers depend on knowing such attributes as TI and mating frequency, as well as adult longevity for each sex. Progress has recently been made in gathering comparative data on mating frequency by counting spermatophores dissected from reproductive tracts of wild females and in interpreting these data (Burns 1966, I968 Shields I968 Pliske, in prep.). On the other hand, T has received scant attention. Scattered observations include the following. An interspecific copulation involving pierid butterflies, Colias interior c X C. eurytheme , lasted 67 minutes (Ae I956). Among crosses of C. eurytheme carried out to study the genetics of an intricate enzyme polymorphism (Burns and Johnson I967), the two that were timed gave Tts of 55 and 75 minutes. In Danaus plexiDpus, a nymphalid butterfly, "It is not known for what length of time the male and female remain united, but on one occasion such a pair was found an hour and a half later on the same tree and in the same position" (Urquhart I96O). _A related species, .D. giliM)us, copulates for a period of about one to (usually) several hours (Brower, Browerand Cranston I965). Indeed, timed copula-

tions in this species have ranged from a low of OO minutes to a high of 12 (--3) hours (T.E. Pliske, personal communication).
A pyrgine skipper butterfly, Erynnis tristis, copulated for a little less than one hour (Shields I968).Data reported below were obtained in the course of genetically analyzing sex-limited wing-color dimorphism in a hesperiine skipper, Poanes hobomok (Burns, unpublished) Middletown, Connecticut, reared virgins were placed in outdoor screen cages, large enough (60" long )< :8"' wide ) 39"' high) to permit flight, and were continuously watched.Copulations were timed from beginning to end, with the result shown in Table I.
The T1s are normally distributed around a mean of 38 I/4 min- utes (fig. ).In view of their considerable length, the T1s are remarkably consistent.
Males of P. hobomok are monomorphic but females are dimorphic" ene female morph (light) is similar in facies to the male whereas the other (dark) is not.Seven experimental crosses involved light females and six, dark ones.Female color-pattern did not significantly affect T. Nor did the time of summer at which crosses occurred, later crosses not being significantly longer than earlier ones.
In general, T, like so many behavioral phenomena, is best ap- proached statistically, with due regard, however, for prevailing weather conditions.Casual observations suggest that cloudiness and lower temperatures tend to prolong T, which is not surprising.
Presumably it cannot be shortened indefinitely because of the logistics of spermatophore production.Taken together, the meager data assembled here from four un- related genera suggest, first, that T will tend to be a normally dis- tributed variable (in any particular species population and under similar environmental .conditions);and, ,second, that it will vary widely from some groups of species to others.In the series Poanes Erynnis Colias Danaus, mean T1s run a gamut from 38 minutes to nearly one hour to roughly 66 minut,es to several hours.
The excessively long T of D. yilippus may relate to the fact that individuals of this species are often distasteful to vertebrate predators that can learn to leave them alone.Similarly, it may be on this account that danaines can afford to mate so very many times (see Burns 1968;Pliske, in prep.).But inedibility does not explain why they mate so long or so much.Although it has been suggested that the high number of matings may partly derive from increased longevity conferred by distastefulness (Pliske, in prep.), far more than this must be involved because danaines will mate several times in what, for a butterfly, can only be considered rapid succession.For example, three pairs of D. blexilpus in separate small outdoor cages at San Antonio, Texas, were seen to mate once or (usually) twice each day over a four-day period (R. O. Kendall, personal commu- nication).The question remains open.A correlation noted, in D.
gililpus, between high mating frequency and a low population density associated with great mobility (Burns 1968) hints at directions for future inquiry.
Until we have hard longevity data for these relatively long-lived danaine butterflies, we .cannotbe precise about the proportion of
. All material used in this work came from southern New England: Rockfall and Portland, Hampden County, Massachusetts; and Jacksonville, Wind- ham County, Vermont.Although, in nature, P. hobomok is univoltine and .spring-flying, a second generation was forced in late summer by laboratory rearing.On sunny days in August and September at Middlesex County, Connecticut; vicinity of Mt.Tom, north of Holyoke,