Deciduous Wings in Crickets:a New Basis for Wing Dimorphism

Dimorphism in wing length occurs in species o at least seven insect orders (Richards, 96 ). It is conspicuous in many species crickets. Crickets with the o,lded metathoracic wings extending beyond the tegrnina are termed macropterous, and those with the olded metathoracic wings covered by the tegmina are termed micropterou (Alexander, 96, 968). Since ying cricketssuch as those ing to light traps are always ma.cropterous, it seems s.ae to co,nclude that micropterous crickets cannot fly. It seems probable, though the evidence is sparse, that macropterous individuals generally do fly and that wing dimorphism in crickets has the same behavioral correlation with emigration that it has in aphids (Kennedy and Stroyan, x959). In studying three species o short-tailed cricket (Anurogryllus), I noted dimorphism in wing length. Most specimens had no visible metathoracic wings (i.e. were micropterous), but 5% .o those the U. S. species and about 20,% each o two VCest Indian species had conspicuously protruding metathoracic wings (i.e. were macropterous). At least 5 o the 6 macropterous West Indian specimens were collected a light, but none o the 4 macropterous specimens o the U. S. species were. Three o these 4 were recently molted specimens rom a laboratory colony and the other was a teneral specimen dug rom its burrow in the ield. Obviously none had ever flown. In studies o the same species Weaver and Sommers (969, p. 338) also noted macroptreous teneral specimens, and in addition described behavior that accounts or the absence o macropterous individuals among non-teneral specimens.: "When the cricket first transforms into the adult stage, the wrinkled, whitish hindwings extend or some distance beyond the posterior edge o the orewings Usually within 24 hr the hindwings are broken off at the base a.nd eaten."

relation with emigration that it has in aphids (Kennedy and Stroyan,x959).
In studying three species o short-tailed cricket (Anurogryllus) , I noted dimorphism in wing length. Most specimens had no visible metathoracic wings (i.e. were micropterous), but 5% .o those the U. S. species and about 20,% each o two VCest Indian species had conspicuously protruding metathoracic wings (i.e. were macropterous). At least 5 o the 6 macropterous West Indian specimens were collected a light, but none o the 4 macropterous specimens o the U. S. species were. Three o these 4 were recently molted specimens rom a laboratory colony and the other was a teneral specimen dug rom its burrow in the ield. Obviously none had ever flown. In studies o the same species Weaver and Sommers (969, p. 338) also noted macroptreous teneral specimens, and in addition described behavior that accounts or the absence o macropterous individuals among non-teneral specimens.: "When the cricket first transforms into the adult stage, the wrinkled, whitish hindwings extend or some distance beyond the posterior edge o the orewings Usually within 24 hr the hindwings are broken off at the base a.nd eaten." "The three species are presently known as .4. muticus (De Geer), but they will be given distinctive binomials in a paper now in press (Walker 1973). Manuscript received by the editor October 9, 1972 Psyche [ December Prompted by these clues, I examined more than 2o "micropterous" specimens o each o the three species o Anurogryllus. All had wing stumps rather than short wings. I pulled with tweezers on one o the wings o. an alcohol-preserved, non-teneral macropterous specimen. The wing tore loose with difficulty but let a stump like those o the "micropterous" specimens. Apparently all the specimens had once been macropterous, and the dimorphism in wing length in each o the three species is based on a dichotomy in wing deciduousness rather than a dichotomy in wing length in the newly ormed adult.

DISCUSSION
The term micropterous is inappropriate o.r crickets with the stumps o deciduous wings. The term dealated has been used or similar cases in other insects and seems appropriate here.
So xCar as now known, dealated A nurogryllus are originally macropterous rather than micropterous, but since wing shedding has not been observed in either o the West Indian species, the length the shed wings is unknown. Furthermore the timing o shedding is not known or the West Indian species. One possibility is that it occurs only in teneral adults, as in the U. S. species. I this be the case, shedding the wings could depend on either the wings being weakly attached to the stumps or on wing-removing behavior or on both. I some or all teneral adults are competent both to retain and to shed their wings, their remaining macropterous or becoming dealates would most likely be an adaptive response to some aspect their environment. For example a stimulus associated with dense population might inhibit wing shedding and promote emigration.
A second and contrasting possibility concerning the timing o.
wing shedding in West Indian Anurogryllus is that none sheds its wings while teneral and all or nearly all individuals disperse by flight before losing their wings. Situations analogous to this possibility occur in termites, ants, and perhaps certain zorapterans (Imms, 1957). I know o no case analogous to the first possibility (i.e. dimorphism with nonmigratory individuals shedding potentially unctional wings). Certain Australian roaches o the genus Panethia are apparently like the U. S. Anurogryllus--i.e. all individuals shed well-developed wings shortly ater the final molt (Mackerras,197o;p. 273).
The extent to which crickets other than Anurogryllus have deciduous wings is unknown. I such wings were characteristic ox any o( the species in which wing dimorphism has been carefully studied, 1972] tIalker--Deciduo.us Wings 313 they would have been repo.rted. The only instances known to me ot wing shedding in crickets other than Anurogryllus are in Gryllus. R. D. Alexander (personal communication ca. I96O, I972) told me he had seen a Gryllus bimaculatus De Geer emale pull off and eat the wings o a courting male. In handling living macropterous G. rubens Scudder, I have no.ted that the wings occasionally detach with only a slight pull. Mo.re recently I have tried pulling the wings rom alcoho.l-preserved specimens o G. rubens. In both macropterous and micropterous individuals the wings were oten easily detached. They tore just distal to. the axillary sclerites and let stumps like those in A nurogryllus. These specimens had been reshly killed within the first week o. adult lie and had never flown. Ot more than oo such specimens examined none was already dealated.
Two attributes o wing shedding that have probably contributed to its evolution are (I) it sometimes aids escape rom predators (cL G. rubens escaping rom my grasp) and (2) it allows unctionless or no-longer-unctional wings to be eaten (eL many insects, includin crickets, eating their exuviae, apparently to nutritional advantage).

UMMARY
In three species oi: 4nurogryllu, that are superficially dimorphic in wing length all "micropterous" individuals have the stumps o longer wings. The dimorphism is in the occurrence o wing shedding and is not known to correlate with a dimorphism in wing length.