The Evolution of Wing Dimorphism in Insects

Derek A. Roff

Evolution, Vol. 40, No. 5. (Sep., 1986), pp. 1009-1020.

Stable URL:

http://links.jstor.org/sici?sici=0014-3820%28198609%2940%3A5%3C1009%3ATEOWDI%3E2.0.CO%3B2-L

Evolution is currently published by Society for the Study of Evolution.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtainedprior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content inthe JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/journals/ssevol.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academicjournals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community takeadvantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org.

http://www.jstor.orgSun Nov 11 22:18:17 2007

Evolutron, 40(5), 1986. pp. 1009-1020

THE EVOLUTION OF WING DIMORPHISM IN INSECTS

DEREKA. ROFF McGill University, Department of Biology, 1205 Avenue Dr. PenJield, Montr6a1, Quebec H3A IBI , Canada

Abstract. -Wing-dimorphic insects are excellent subjects for a study of the evolution of dispersal since the nondispersing brachypterous morph is easily recognized. The purpose of this paper is to develop a framework within which the evolution ofwing dimorphism can be understood. A review of the literature indicates that the presence or absence of wings may be controlled by a single locus, two-allele genetic system or a polygenic system. Both types of inheritance can be subsumed within a general threshold model.

An increase in the frequency of a brachypterous morph in a population may result from an increased relative fitness of this morph or the emigration of the macropterous type. The abundance of wing-polymorphic species argues for an increased fitness of the brachypterous form. An analysis of the life-history characteristics of 22 species of insects indicates that the brachypterous morph is both more fecund and reproduces earlier that the macropterous morph. Unfortunately, data on males are generally lacking.

It is suggested that suppression of wing production results when some hormone, perhaps juvenile hormone, exceeds a threshold value during a critical stage of development. Further, it is known that in the monomorphically winged species Oncopeltus fasciatus both flight and oviposition are regulated by the titer of juvenile hormone. These observations are used to construct a possible pathway for the evolution of wing dimorphism. This suggests that evolution to a dimorphic species requires both an increase in the rate of production of the wing suppressing hormone and a change in the threshold level at which wing and wing-muscle production are suppressed. The stage in this evolutionary sequence that an organism will reach depends on the stability of the habitat.

Received December 5, 1984. Accepted May 7 , 1986

In a constant environment there may be circumstances in which dispersal will be a selectively favorable character (Hamilton and May, 1977). However, in a spatially and temporally heterogeneous environment dis- persal becomes a virtual necessity, without which extinction may be very rapid (Roff, 1974a, 19743; Leigh, 1981). The impor- tance of spatial and temporal heterogeneity in determining population fluctuations and life-history characteristics has long been recognized (Elton, 1930; Andrewartha and Birch, 1953; Southwood, 1962), but it has been only comparatively recently that de- tailed theoretical studies have been under- taken. Apart from the not unexpected find- ing that dispersal can have a strong stabilizing influence on population fluctua- tions (Reddingius and den Boer, 1970; Roff, 1974a, 1974b; Vance, 1980; den Boer, 198 1; Kuno, 198 1 ; Hanson and Tuckwell, 198 1; Hastings, 1982), simulation studies suggest that dispersal strategies depend upon both the quantitative and qualitative aspects of environmental heterogeneity (Van Valen, 1971; Roff, 19743, 1975,1986~; Palmer and Strathmann, 1981). The evolution of dis- persal strategies will be a function of both the habitat characteristics and the trade-offs

involved in dispersal. For example, in D. melanogaster both flight and egg production depend on the same energy reserves, and hence dispersal reduces fecundity, with ob- vious consequences for the fitness of the dis- perser (Roff, 1977).

A fundamental assumption of any model or speculation concerning the evolution of dispersal traits is that these be determined, at least in part, by the genetic constitution of the parents of the disperser. This as-sumption is also of importance with respect to population dynamics since the numerical response of a population to changes in en- vironmental parameters depends upon the mode of inheritance of dispersal tendency (Roff, 1975). A study of the genetic basis of dispersal must be able to distinguish be- tween dispersers and nondispersers. In this regard wing dimorphic insect species are ideal organisms, since the inability to dis- perse is clearly recognizable by the absence of wings (aptery) or wings too small to per- mit flight (microptery or brachyptery). The presence of wings does not inevitably mean that an insect can fly, since in some species flight muscles may be absent and wings present (Jackson, 1956a, 19563; Larskn, 1966). Generally, however, this condition

DEREK A. ROFF

morphism? Second, what are the benefits MICROPTEROUS

of being wingless? Finally, what is the likely evolutionary sequence from a monomor-

HORMONE phic winged population to a dimorphic pop-

LEVEL THRESHOLD..-,-.-,-.-,-,-,-,~.~*..-.-~-.-.-.-.--.-.-~-.-.-.-.-.-.-.-ulation? 88

BbMACROPTEROUS

b b

G E N O T Y P E

THRESHOLD

G E N O T Y P E (HORMONE L E V E L 1

FIG.1. A schematic representation of a threshold model for the determination of pterygomorphism. a) A single locus, two-allele system in which the bra- chypterous allele appears dominant because the ad- ditive effect of Bb exceeds the threshold level of the regulatory compound for wing production. b) A poly-genic system for wing determination. In both models the switch from one morph to another is assumed for convenience to be controlled by the level of a hormone, although some other substance(s) may be ~nvolved. In the polygenic model, the level of the hormone can be equated with the genotype though not with a unlque genotype.

probably results from the histolysis of the wing muscles at the onset of reproduction and after the dispersal episode (Johnson, 1976; Dingle, 1979, 1982). It is relatively easy to determine whether winged individ- uals are capable of flight and, hence, wheth- er wing dimorphism is a character by which a population or family can be divided into potential dispersers and nondispersers. Thus. for example, it has been shown that in Homoptera and Gerridae the degree of brachyptery is correlated with the stability of the habitat as is predicted from theoret- ical considerations (for Homoptera see Denno, 1978, 1979; Denno et al., 1980; McCoy and Rey, 198 1; for the Gerridae, see Vepsalainen, 1973, 1974a, 1978; for a gen- eral review see Harrison, 1980).

In this paper, I address three problems. First, what is the genetic basis of wing di-

The Genetic Basis of Pterygornorphisrn The simplest model for the genetic de-

termination of pterygomorphism is a single locus with two alleles, with macroptery being either dominant or recessive (for conve- nience I shall refer throughout this discus- sion to the fully winged condition as macroptery and the short-winged or wing- less conditions as either brachyptery or mi- croptery). At the other extreme the char- acter may be inherited in a polygenic manner.

Modes of inheritance can be postulated to be the result of the additive effects of alleles at a single locus and/or across many loci. In the case of single-locus inheritance, either macroptery or brachyptery may ap- pear dominant depending on the values of the two alleles (Fig. la). A polygenic mode of inheritance can be understood within the framework proposed for threshold charac- ters in general (Falconer, 198 1; Roff, 1986b; Fig. lb).

Of the 22 studies in which it is possible to suggest a genetic basis for the trait, eight indicate a simple Mendelian mechanism (Table 1). I have omitted from Table 1 stud- ies in which environmental factors such as photoperiod or temperature may be re-sponsible for the observed polymorphism (Poisson, 1924; Ekblom, 1928, 194 1, 1949) and those cases in which reduction or loss of wings is due to a spontaneous mutation as in Drosophila melanogaster (Eker, 1935), Pieris napi(Bowden, 1963), or Bombyx mori (Tazima, 1964).

A polygenic system appears to be the more general situation, although the number of loci involved is unknown. It has been shown that genetic models involving only two or three loci produce distributions that can only be distinguished from normal distributions with very large sample sizes (Thoday and Thompson, 1976). In general, the sample sizes from the published analyses on wing inheritance comprise relatively few families and individuals (hundreds of individuals and at best only a dozen or so families).

TABLE1 . The probable genetic basis of pterygomorphism in a variety of insect species. Methods of analysis: 1 ) Data from individual crosses; 2) grouped data from known parent morphs; 3) selection for wing morphs; 4) comparison of different geographic strains; 5) comparison of clones.

Order Family

Coleoptera Curculionidae Coleoptera Curculionidae Coleoptera Carabidae Coleoptera Carabidae Coleoptera Carabidae Coleoptera Carabidae Coleoptera Ptiliidae Coleoptera Bruchidae Diptera Sphaeroceridae Hymenoptera Bethylidae Hymenoptera Formicidae Hemiptera Pyrrhocoridae

Hemiptera Gemdae

Hemiptera Gemdae Orthoptera Gryllidae Orthoptera Gryllidae Orthoptera Gryllidae

Orthoptera Acrididae Homoptera Delphacidae Homoptera Delphacidae Homoptera Delphacidae Homoptera Aphididae

Homoptera Aphididae

Species

Sitona hispidula Apion virens Pterostichus anthracinus Bembidion lampros Calathus erythroderus Calathus melanocephalus Ptinella apterae Callosobruchus maculatus Apterina pedestris Cephalonomia gallicola Harpagoxenus sublaevis Pyrrhocoris apterous

Gerris lacustris

Limnoporus caniculatus Gryllus pennsylvanicus Gryllus firmus Gryllodes sigillatus

Melanoplus lakinus Laodelphax striatellus Nilaparvata lugens Javesella pellucida Schizaphis gramimum

Acyrthosiphon pisum

Mode of inheritance

Single locus, two alleles, brachyptery dominant Single locus, two alleles, brachyptery dominant Single locus, two alleles, brachyptery dominant Single locus, two alleles, macroptery dominant (?) Single locus, two alleles, brachyptery dominant Polygenic Polygenic Polygenic Single locus, two alleles, brachyptery dominant Diploid, apterous; haploid, 2 alleles Diploid, two alleles, brachyptery dominant Polygenic

Polygenic

Polygenic Polygenic Polygenic Polygenic

Insufficient data Polygenic Polygenic Polygenic Polygenic

Polygenic

Method of

analysis Reference

Jackson, 1928 Stein, 1973a Lindroth, 1 946b Langor and Larson, 1 983C Aukema, unpubl. Aukema, unpubl. Taylor, 198 l d Utida, 1972 Guib6, 1939e Kearns, 1934 Buschinger, 1978 HonEk 1976a, 19766,

1979 Vepsalainen, 1974b, pers.

c0mm.f Zera et al., 1983 Hanison, 1979 Roff, 1984, 1986b Ghouri and McFarlane,

1958 Bland and Nutting, 1969 Mahmud, 1980 Mochida, 1975 Ammar, 1973 Kvenberg and Jones,

1974 Lamb and McKay, 1979

a No statisttcal analysts ts presented by Stem However, it is possible to estimate the expected frequencies from his data; these do not differ significantly from the observed. Given the large number of offspring (855) the proposed genetlc bass can be accepted.

No stat~stlcal analvs~s is resented bv Lindroth. The observed and exwcted frequencies do not differ significantly, but the sample size is small (52 offspring among seven crosses); therefore, any conclusion must . . . -

be tentatlve. Contrary to the contention of Langor and Larsen, a one-locus, two-allele model is consistent with their breeding data. However, the number of offspring from each cross is very low (42 offspring among 16 crosses);

hence the above model serves principally as a basis for further analysis. ~ i t e m a leffects could not be discounted.

'No macropterous ~ndtvtduals have been found in the wild. However, the presence of well-developed wing muscles In the macropterous individuals (Gu~td., 1939) suggests that the genetlc switch from brachyptery to macroptery 1s well Integrated wlth the developmental process, and therefore is probably not a spontaneous mutation.

Vepsiila~nen (1 9746) hypothesized a super gene with a temperature switch. Data on G e m s lacusrris were not entirely consistent with this model (Vepsilainen, 1974b). and in another gerrid, L~mnoporuscanrculafus. a polygen~c basis is evldent (Zera et al., 1983).

c-.L

0 -c-.L

1012 DEREK A. ROFF

Other studies have involved only mass crosses of macropterous x macropterous and brachypterous x brachypterous. For these reasons, no general conclusion can be reached on the number of loci involved or the heritability of the trait. The analysis of wing inheritance in the cricket Gryllus jr- mus (Roff, 1986b) based on over 100 fam- ilies and 10,000 individuals gives an esti- mate of heritability of approximately 0.65. Only further analyses on other species can establish whether this is a usual value.

The Fitness of the Two Wing Morphs For a vestigal or brachypterous mutation

to spread in a population other than by ge- netic drift it must either possess a higher fitness within that population or be pre- served simply because the winged individ- uals disperse, leaving the flightless individ- uals as an isolated population. The latter circumstance may account for the high fre- quency of flightless D. melanogaster found in a deep pit with decomposing fruit (Du- binin et al., 1937; cited by Dobzhansky, 195 1 p. 64). Wing polymorphism is com- mon in the Orthoptera, Hemiptera, Ho- moptera, Plecoptera, and Coleoptera. In ad- dition, wing reduction in one or both sexes is found in most insect orders (for examples see Kalmus, 1945; Brues et al., 1954; Pois- son, 1946; C.S.I.R.O., 1970; in the subclass Pterygota I can find examples of wing re- duction in all orders except the Ephemer- optera, Odonata, and Megaloptera). The prevalence of wing polymorphism and the more extreme case of complete wing reduc- tion argues against preservation by isolation of vestigial-winged individuals as being the general explanation.

What are the advantages of being flight- less? Flight is energetically expensive (Weis- Fogh, 1952; Hocking, 1953; Sotavalta and Laulajainen, 196 1; Yurkiewicz, 1965), and in Drosophila species it reduces egg pro- duction (Roff, 1977; Inglesfield and Begon, 1983). Dispersal may also be risky in that it may increase the chance of being preyed upon (unfortunately we do not have ade- quate estimates of the relative mortality risks of dispersing and non-dispersing insects), and the disperser may fail to locate a fa- vorable habitat. The energy cost and mor-

tality cost of dispersal might mitigate against flight but need not in themselves select for flightlessness since a fully winged individual need not fly. The evolution of flightlessness in so large an array of insect orders, families, and species suggests that there is a cost to possessing the capability of flight whether or not flight ever occurs. Table 2 summa- rizes data on comparisons between life his- tory characteristics in macropterous and micropterous (including brachypterous or apterous) morphs of a wide range of insect species. I have omitted from this table sev- eral studies in which the experimental de- sign or data base are clearly inadequate (for Homoptera the study by Watson and Sinha [I9591 and for gerrids the studies of Poisson [1924], Brinkhurst [l959], Anderson [1973]; see Zera [I9841 for a detailed discussion on the genid data). Statistically nonsignificant differences between morphs may not be very informative, since they may arise as a result of small sample size and/or high variation.

There are no consistent trends in devel- opment time between macropterous and micropterous morphs; in most instances there are no differences. A similar pattern is found for adult longevity, at least for fe- males. There are clear trends in preovipo- sitional period and fecundity. In the former case the brachypterous morph either begins reproduction before or at the same time as the macropterous morph. Brachypterous fe- males consistently produce more eggs than macropterous. Notable exceptions are Pti-nella apterae, P. errabunda, and Orgyia thyellina in which the winged morph pro- duces significantly more eggs than the wing- less morph (Taylor, 1978; Sato, 1977).

Brachypterous Orgyia thyellina usually lay diapause eggs, while macropterous fe- males lay nondiapause eggs (Kimura and Masaki, 1977). These eggs differ greatly in size, those from the brachypterous females weighing 0.267 mg and from macropterous only 0.180 mg (Sato, 1977). Although bra- chypterous females lay on average 390.4 eggs compared to 476.9 eggs laid by macropter- ous females, the total reproductive capacity (egg weight x fecundity) of the latter is greater than the former (1 04.2 mg compared to 85.8 mg). If brachypterous females laid eggs the same size as macropterous females, their predicted fecundity would be 579 eggs.

1013 WING DIMORPHISM IN INSECTS

TABLE A comparison of life-history parameters between macropterous and brachypterous morphs of various 2. insect species. 0: no difference between morphs; +: brachypterous morph > macropterous morph; -: brachyp-terous morph < macropterous morph.

Develop- Pre-ovi- Adult ment positional lon-

Order Species time period Fecundity gevlty Reference

Homoptera Javesella pellucida 0 0 Ammar, 1973 Javesella pellucida 0 - +, Oa 0 Mochida, 1973 Javesella pellucida -b,c Waloff, 1973 Doratura stylata -b,c Waloff, 1973 Laodelphax striatellus OC OC OC Mitsuhashi and Koyama,

1974 Stenocranus minutus -c +c + C May, 1975 Nilaparavata lugens -c,d +c Nasu, 1969 Nilaparavata lugens 0 0 0 Manjunath, 1977e Nilaparavata lugens - + 0 Kisimoto, 1965e Sogata furcifera - + 0 Kisimoto, 1965e Delphacodes striatella - + 0 Kisimoto, 1965e Delphacodes striatella + Tsai et a]., 1964 Sitobion avenae 0 + + Wratten, 1977 Metopolophium dirhodum 0 + + Wratten, 1977 Aphis fabae 0 + Dixon and Wratten, 197 1 Drepanosiphum dixoni - + Dixon, 1972 Macrosiphum granarium -J Noda, 1960 Rhopalosiphum prunifoliae -J Noda, 1960 Aphis maidis -J Noda, 1960

Hemiptera Sigara dorsalis of +g Young, 1965 Sigara fallen; of + Young, 1965 Sigara scotti of +g + Young, 1965 Limnoporus canaliculatus - + - Zera, 1984

Coleoptera Ptinella apterae - Taylor, 1978 Ptinella errabunda - Taylor, 1978 Sitona hispidula +" Jackson, 1928 Callosobruchus maculatus - - + - Caswell, 1960; Utida,

1972 Orthoptera Concocephalus discolor - + Ando and Hartley, 1982

Zonocerus variegatus 0 McCaffery and Page, 1978 Pteronemobius taprobanensis -c,d +c -g Tanaka, 1976 Pteronemobius nitidus + Tanaka, 1978 Gryllodes sigillatus Ok +" Ghouri and McFarlane,

1958 Gryllus jirrnus + I + 0 Roff, 1984 Allonemobius fasciatus 0 +, Oa 0 Roff, 1984

Lepidoptera Orgyia thyellina -I Sato, 1977 Psocoptera Graphopsocus cruciatus Oh New, 1969

Ectopsocus briggsi Oh New, 1969 a Fecundity of brachypterous morph significantly greater than that of the macropterous In the early stage of reproduction but no signlficant difference

over the total lifespan. Based on appearance of mature adults in the held and therefore may include differences in development tlme of nymphs. No statistical analysis and ~nsufficient data presented to undertake such an analysis. However, the differences between the groups are sufficient

to conclude that a signlficant difference probably exists. Likewise, In those Instances where no difference is indicated, the difference 1s very small. * The delay In reproduction of the macropterous morph is supported by examination of ovariole development. Unpublished statistical analysis of data presented in paper.

'Analys~s of field populations. Polymorphism is in flight muscles rather than wings, but morphs can be distinguished externally. g Under conditions of starvation. The method of statistical analysis by Young (1965) 1s incorrect. Insufficient information Bven to perform a correct

test. No data are presented to support this contention.

'Eggs produced by the brachypterous morph are larger than those of the macropterous (see text). J The difference in development time is in part due to a higher threshold temperature for development in the macropterous form.

Arai (1978 Japanese with English summary) states that the macropterous morph takes longer to develop at 35'C than the micropterous. ' ~npublisheh analysis of larger data set than analyzed in Roff (1984) where "no differences" are reported.

Thus the negative sign in the fecundity col- females are more fecund than macropters. umn of Table 2 is somewhat misleading. Data for males are generally lacking. Mochi-

Despite the exceptional case of Ptinella, da (1973) found no difference in the devel- the data strongly suggest that brachypterous opment of the male reproductive structures

1014 DEREK A. ROFF

of the two wing morphs of Javesella pellu- cida. Alate males of the hymenopteran Tri-chogrammatoidea armigera are able to in- seminate more females than apterous males (Manjunath, 1972). But in this species ap- terous males appear to be developmental aberrations: these males are almost invari- ably associated with a female, the two in- dividuals developing from a single fertilized egg. It is hypothesized "that one of the cleavage nuclei or polar bodies becomes separated off along with a little cytoplasm from the parent cell and develops indepen- dently" (Manjunath, 1972 p. 146). In both male and female macropterous Callosobru-chus maculatus the abdominal cavity is filled by fat body at adult emergence whereas in the micropterous form both ovaries and testes are well developed (Utida, 1972). The relative percentage of micropterous males differs between species and in some cases, as with Gryllus firmus (Roff, 1984, 1986b) there may be a higher frequency of microp- terous males than females. This suggests that being winged carries some cost, even in males.

The Evolution of Wing Dimorphism The sequence of events in the average

macropterous insect is:

wing wing dis- ovi-muscle pro- - persal - position+ pro-duction duction L, muscle _S histolysis

with overlap in events, particularly wing and muscle development.

It has been suggested that brachyptery re- sults when the level of some hormone, pos- sibly juvenile hormone, exceeds a threshold value during a critical period in develop- ment (Southwood, 196 1; Wigglesworth, 196 1). Both flight and oviposition are reg- ulated by the titer of juvenile hormone in Oncopeltus fasciatus (Rankin, 1978; Ran- kin and Riddiford, 1978; Slansky, 1980) and it is likely that the genetic variation in flight duration (Dingle, 1968, 1980) is actually due to genetic variation in the rate of production of juvenile hormone. Polygenic control of flight duration has also been demonstrated in Lygaeus kalmii (Caldwell and Hegmann, 1969) and various Cicadulina species (Rose,

1972). Assuming the dispersal syndrome to be controlled in both its morphological and behavioral aspects by the titer of juvenile hormone we can construct a possible evo- lutionary sequence from monomorphic macroptery (as in 0.fasciatus) to wing di- morphism as follows (Fig. 2). If genetic vari- ation exists in the amount of flight required to initiate reproduction, some individuals with wings and wing muscles will reproduce without dispersing (Fig. 2a). Differences in habitat stability will, therefore, select for dif- ferences in the readiness to disperse and amount of flight required to initiate repro- duction. Such differences have been found in different geographic populations of On-copeltus fasciatus (Dingle, 1978, 1980) and between species comparisons of Dysdercus (Dingle and Arora, 1973; Derr et al., 198 1).

As habitat stability increases, selection will probably favor an increasing proportion of nondispersers, and the distribution of ju- venile hormone will shift as shown in Figure 2b. At this point the population contains individuals that produce wings but not wing muscles. These individuals reproduce with- out dispersing and have a selective advan- tage over those that do not disperse but still produce wing muscles, since energy can be immediately channelled into reproduction. An increase in the frequency of individuals that do not produce wing muscles will be opposed by the decrease in the proportion of dispersing types produced (Fig. 2b). Se- lection then favors a change in the threshold level at which dispersal is initiated, even- tually making the thresholds ofwing-muscle production and dispersal equivalent (Fig. 2c). Such a change may involve a shift in only one threshold level, as shown in Figure 2c, or a change in both levels to make them coincident.

Although the production of wings per se may not be energetically expensive, there may be selection against wing production simply as an indirect result of selection for earlier oogenesis, which requires selection of those individuals with yet higher rates of juvenile hormone production and a com- mon threshold for flight capability (Fig. 2d). Furthermore, there will be selection of mod- ifications that enhance reproduction, such as an enlargement of body parts concerned with egg production and a reduction of those

WING DIMORPHISM IN INSECTS

W ~ N G S W I N G M U S C L E S A N D D I S P E R S A L I N H I B I T E D

> !A 0 Z W C3 I 0 W I N G M U S C L E S A N D W(0

D I S P E R S A L I N H I B I T E D

> >= Z<

0 3

0

W

a

I 1 D I S P E R S A L W l N G W I N G S

N H i B T E D M U S C L E S N H l B i T E D I N H I B I T E D

H O R M O N E L E V E L

FIG.2. A possible evolutionary sequence of wing polymorphism from a macropterous population. Fol- lowing the sequence from bottom to top: a) The dis- tribution of juvenile hormone (JH) production is such that all members of the population are macropterous. However a small percentage have a sufficiently high production of JH that egg production is initiated with- out dispersal. b) A higher frequency of nondispersal being favored, the proportion of individuals producing high levels of JH increases in the population. The shift in the distribution produces a small proportion of in- dividuals that abort wing muscle production. The in- termediate type (nondisperser with wing muscles) is selected out of the population by selection of individ- uals with a threshold level for dispersal that is coin- cident with the threshold for wing muscle production. c) A higher frequency of nondispersers can now be produced without decreasing the proportion of dispers- ers. The shift in the frequency distribution of JH pro- duction produces some individuals that abort wing production. d) Selection for earlier oogenesis and/or for structural modifications of the micropterous morph produces a strictly dimorphic population.

concerned with flight. There is a complete spectrum of degrees of external modifica- tions, from practically none to radical changes that make the morphs entirely dif- ferent in structure (Fig. 3).

Concluding Comments It seems likely that the flight and repro-

ductive systems are coupled in both males and females. In winged individuals energy is channelled into the production and main- tenance of the wing muscles, which in many species are not fully developed at the time

FIG.3. Some examples of micropterous (right side) and macropterous (left side) morphs showing the vari- ation in differences in morphology between the two types. The two morphs are drawn to the same scale for each species except (c), in which the micropterous morph is eleven-tenths the size of the macropterous morph. a) Pterostichus anthracinus: no obvious external mor- phological differences between the morphs (redrawn from Lindroth, 1946). b) Gelis corruptor: the thorax of the macropterous morph is more robust than that of the micropterous (redrawn from Salt, 1952). c) Hal-ticus chrysolepis: Morphological differences between the two morphs involve the entire body shape (redrawn from Zimmerman, 1948). d) Plastosciara perniciosa: the differences in the two morphs are so great that they might be classified as separate species (redrawn from Steffan, 1975).

of adult emergence (Williams et al., 1943; -Balboni, 1967; De Kort, 1969; Anderson and Finlayson, 1973; Ready and Josephson, 1982). In micropterous individuals, this en-

1016 DEREK A. ROFF

ergy can be channelled immediately into the development of the reproductive organs. As the energetic cost of producing and main- taining the wing muscles is probably much greater than actually producing the wings per se the "decision" whether to become capable of flight may be made after the full development of the wings. Wings develop at an earlier stage than flight muscles; thus, it is not unexpected to find some species in which all individuals are fully winged but which are polymorphic with respect to wing- muscle development. It is not clear how common this phenomenon is, since it re- quires considerable labor to detect, and the appropriate studies have been undertaken for relatively few species (Jackson, 1956a, 1956b; LarsCn, 1966). It is also not clear how frequently this polymorphism is sim- ply a result of histolysis of the wing muscles immediately prior to reproduction. There are also instances in which muscle regen- eration may occur (Chapman, 1956; Mor- gan et al., 1984). There are frequently sig- nificant morphological differences between brachypterous and macropterous morphs, attaining an extreme form in the dipteran PIastosciara perniciosa in which the mi- cropterous morph more closely resembles a maggot than the typical fly-like macropter- ous morph (Steffan, 1973, 1975). The mul- tiplicity of changes that may accompany the loss of flight suggests that, in many if not most cases, the developmental changes that must occur may well disrupt the production of wings.

Any fecundity advantage that accrues to a flightless morph will be offset by the dis- advantage of the inability to disperse. If the habitat is ephemeral and patchy, a poly- morphism may be established provided the winged morph mates prior to dispersal and/ or the trait has a polygenic basis. Mating prior to dispersal is a necessity for those species in which macroptery is a recessive character. The increase in the brachypter- ous morph of Apion virens and Sitona his- pidulus in newly colonized fields (Stein, 1977) suggests that mating does occur prior to dispersal in these species.

If a habitat is stable for long periods or distributed so that an insect can move from one patch to another by walking, hopping or swimming, the polymorphism will shift

towards the micropterous morph. An ex- treme result of this shift may be the com- plete loss of the macropterous morph and an entirely wingless population or species. Although habitat stability and/or small in- terpatch distances are essential for the per- sistence of a completely flightless morph, other factors may accelerate or retard the evolution from polymorphism to mono-morphism. For example, life under bark or d e e ~within leaf litter or caves mav favor modifications that will lead to loss okwings (Dybas, 1960; Barr, 1967; Hackman, 1964; Hamilton, 1978). The coleopteran species AgIyptinus dimorphicus is particularly in- teresting as it is found in both caves and forest litter. The cave populations are all short-winged whereas the forest litter pop- ulations are polymorphic (Peck, 1972). An increased understanding of the genetic and physiological basis and evolution of wing polymorphism may shed considerable light both on the evolution of dispersal strategies in general and specifically on the evolution of monomorphically flightless species, as has occurred in many insects.

My considerable thanks go to Ms. Janice Joyce of the McGill Botany-Genetics li-brary for her efficiency and speed at tracking down and procuring the references cited in this paper. This study was supported by NSERC Operating Grant A7764.

AMMAR,E. D. 1973. Factors related to the two wing forms in Javesella pellucida (FAB) (Homoptera: Delphacidae). Z. Ang. Ent. 74:211-216.

ANDERSON, 1973. Seasonal polymorphism and N. M. developmental changes in organs of flight and re- production in bivoltine pondskaters (Hem. Gem- dae). Entomol. Scand. 4: 1-20.

ANDERSON, 1973. Ultra- M., AND L. H. FINLAYSON. structural changes during growth of the flight mus- cles in the adult tsetse fly, Glossina austeni. J. Insect Physiol. 19: 1989-1 997.

ANDO,Y., AND J. C. HARTLEY.1982. Occurrence and biology of a long winged form of Conocephalus dis- color. Entomol. Exp. Appl. 32:238-24 1.

ANDREWARTHA,H. G., AND L. C. BIRCH. 1953. The Distribution and Abundance of Animals. Univ. Chicago Press, Chicago, IL.

ARAI,T. 1978. Effects of environmental conditions on the wing form and growth in Gryllodes sigillatus Walker (Orthoptera: Gryllidae). Jap. J. Ecol. 28: 135-142.

BALBONI, 1967. The respiratory metabolism of E. R.

- -

WING DIMORPHISM IN INSECTS 1017

insect flight muscle during adult maturation. J. In-sect Physiol. 13: 1849-1856.

BARR, T. C. 1967. Observations on the ecology of caves. Amer. Natur. 101:475-492.

BLAND,R. G., AND W. L. NUTTING. 1969. A histo- logical and biometrical study of wing development in the grasshopper Melanoplus lakinus. Ann. Ento- mol. Soc. Amer. 62:419-436.

BOER, P. J. DEN. 198 1. On the survival of populations in a heterogeneous and variable environment. Oecologia 50: 39-53.

BOWDEN,S. R. 1963. A recessive lethal, "wingless" in Pieris napi L. (Lep., Pieridae). Entomologist 96: 52.

BRINKHURST,R. 0 . 1959. Alary polymorphism in the Gerroidea (Hemiptera-Heteroptera). J. Anim. Ecol. 28:211-230.

G. Kennedy (eds.), Movement of Highly Mobile Insects: Concepts and Methodologies in Research. North Carolina State Univ., Raleigh.

-. 1980. Ecology of evolution and migration, pp. 1-101. In S. A. Gauthreaux, Jr. (ed.), Animal Migration, Orientation, and Navigation. Academic Press, N.Y.

-. 1982. Function of migration in the seasonal synchronization of insects. Entomol. Exp. Appl. 3 1 : 36-48.

DINGLE, H., AND G. ARORA. 1973. Experimental studies of migration in bugs of the genus Dysdercus. Oecologia 12: 1 1 9-1 40.

DIXON, A. F. G. 1972. Fecundity of brachypterous and macropterous alatae in Drepanosiphurn dixoni (Callaphididae, Aphididae). Entomol. Exp. Appl. 15:335-340.

AND F. M. CARPENTER. BRUES,C. T., A. L. MELANDER, DIXON,A. F. G., AND S. D. WRATTEN. 1971. Labo- 1954. Classification of insects. Harvard Univ. Mus. Comp. Zool. Bull. 108: 1-97.

BUSCHINGER, 1978. Genetish bendigte entstehung A. gefugelter weibchen bei der sklavenhaltenden ameise Harpagoxenus sublaevis (Nyl.) (genetically induced origin of alate females in the slavemaker ant Har-pagoxenus sublaevis Hymenoptera formicidae). In- sectes Soc. 25(2): 163-172.

CALDWELL,R. L., AND J. P. HEGMANN. 1969. Her-itability of flight duration in the milkweed bug Ly-gaeus kalrnii. Nature 223:9 1-92.

CASWELL, 1960. Observations on an abnormal G. H. form of Callosobruchus rnaculatus. Bull. Entomol. Res. 50:671-680.

CHAPMAN, 1956. Right-muscle changes during J. A. adult life in a scolytid beetle. Nature 177:1183.

C.S.I.R.O. 1970. The Insects of Australia. Melbourne Univ. Press, Melbourne, Australia.

DE KORT, C. A. D. 1969. Hormones and the struc- tural biochemical properties of the flight muscles in the Colorado beetle. Meded. Land. Wageningen 69:l-63.

DENNO,R. F. 1978. The optimum population strat- egy for planthoppers (Homoptera: Delphacidae) in stable marsh habitats. Can. Entomol. 110: 135-142.

-. 1979. The relation between habitat stability and the migratory tactics of planthoppers (Homop- tera: Delphacidae). Misc. Publ. Entomol. Soc. Amer. 11:41-49.

DENNO,R. F., M. J. RAUPP, D. W. TALLAMY, AND C. F. REICHELDERFER.1980. Migration in heteroge- neous environments: Differences in habitat selec- tion between the wing forms ofthe dimorphic plant- hopper, Prokelisia rnarginata (Homoptera: Delphacidae). Ecology 6 123594367.

DERR, J. A., B. ALDEN, AND H. DINGLE. 198 1. Insect life histories in relation to migration, body size and host plant array: A comparative study of Dysdercus. J. Anim. Ecol. 50:181-183.

DINGLE,H. 1968. The influence of environment and heredity on flight activity in the milkweed bug On-copeltus. J. Exp. Biol. 48:175-184.

-. 1978. Migration and diapause in tropical, temperate, and island milkweed bugs, pp. 254-276. In H. Dingle (ed.), Evolution of Insect Migration and Diapause. Springer-Verlag, N.Y.

-. 1979. Adaptive variation in the evolution of insect migration, pp. 64-87. In R. L. Rabb and G.

ratory studies on aggregation, size and fecundity in the black bean aphid, Aphis fabae Scop. Bull. Ento- mol. Res. 6 1:97-111.

DOBZHANSKY, 195 1. Genetics and the Origin of TH. Species, 3rd Ed. Columbia Univ. Press, N.Y.

DUBININ, AND G. G. TINIAKOV. N. P., N. N. SOKOLOV, 1937. Intraspecific chromosome variability. B. Zh. 6:1007-1054.

DYBAS,H. 1960. A new genus of blind beetles from a cave in South Africa. Fieldiana, Zool. 39:339- 405.

EKBLOM, 1928. Vererbungsbiologische Studien iiber T. Hemiptera-Heteroptera. Hereditas 10:333-359.

-. 194 1. Untersuchungen iiber den Rugeldi- morphismus bei Gerris asper L. Notulae Ent. 2 1: 49-64.

-. 1949. Neue untersuchungen iiber den fliigel- polymorphismus bei Gerris asper. Fieb. Notulae Ent. 29:l-15.

EKER, R. 1935. The short-wing gene in Drosophila rnelanogaster and the effect of temperature on its manifestation. J. Genetics 30:357-368.

ELTON, C. 1930. Animal Ecology and Evolution. Clarendon Press, Oxford, U.K.

FALCONER, 198 1. Introduction to Quantitative D. S. Genetics, 2nd Ed. Longman, N.Y.

GHOURI,A. S. K., AND J. E. MCFARLANE. 1958. Oc- currence of a macropterous form of Gryllodes sig- illatus (Walker) (Orthoptera: Gryllidae) in labora- tory culture. Can. J. Zool. 362337-838.

G U I B ~ , 1939. Contribution A I'etude d'une espece: J. Apterina pedestris Meigen (Diptera) Bull. Biol. France et Belg. Suppl. 26: 1-1 12.

HACKMAN, 1964. On reduction and loss of wings W. in Diptera. Notul. Ent. Helsinki 44:73-93.

HAMILTON, 1978. Evolution and diversity un- W. D. der bark. Symp. Roy. Entomol. Soc. Lond. 9: 154- 175.

HAMILTON,W. D., AND R. M. MAY. 1977. Dispersal in stable habitats. Nature 269:578-581.

HANSON,F. B., AND H. C. TUCKWELL. 1981. Logistic growth with random density independent disasters. Theoret. Popul. Biol. 19: 1-18.

HARRISON, 1979. Flight polymorphism in the R. G. field cricket, Gryllus pennsylvanicus. Oecologia 40: 125-132.

-. 1980. Dispersal polymorphism in insects. Ann. Rev. Ecol. Syst. 11:95-118.

1019 WING DIMORPHISM IN INSECTS

POISSON,R. 1924. Contribution A l't'tude des HCm- -. 1975. Morphological and behavioural poly- iptbres aquatiques. Bull. Biol. 58:49-305. morphism in Plastosciara perniciosa (Diptera:

-. 1946. L'apterisme chez les insectes. Rev. Sci. Sciaridae). Proc. Entomol. Soc. Wash. 77: 1-14. Paris 84:605-6 17. STEIN, W. 1973. Zur Vererbung des Flugeldimor-

RANKIN,M. A. 1978. Hormonal control of insect phismus bei Apion virens Herbst (Col., Curculioni- migratory behavior, pp. 5-32. In H. Dingle (ed.), dae). Z. Ang. Ent. 74:62-63. Evolution of Insect Migration and Diapause. -. 1977. Die Beziehung swischen Biotop-Alter Springer-Verlag, N.Y. und Auftreten der Kurzflugeligkeit bei population-

RANKIN,M. A,, AND L. M. RIDDIFORD.1978. Sig- en dimorpher Riisselkafer-Arten (Col., Curculion- nificance of haemolymph juvenile hormone titer idae). Z. Ang. Ent. 83:37-39. changes in timing of migration and reproduction. TANAKA,S. 1976. Wing polymorphism, egg produc- J. Insect. Physiol. 24:31-38. tion and adult longevity in Pteronemobius tapro-

READY,N. E., AND R. K. JOSEPHSON. 1982. Right banensis Walker (Orthoptera: Gryllidae). Kontyu muscle development in a hemimetabolous insect. 44:327-333. J. Exp. Zool. 220:49-56. -. 1978. Photoperiodic determination of wing

REDDINGIUS,J., AND P. J. DEN BOER. 1970. Simula- form in Pteronemobius nitidus. Bolivar (Orthop- tion experiments illustrating stabilization of animal tera, Gryllidae). Kontyu 46:207-2 17. numbers by spreading of risk. Oecologia 5:240- TAYLOR,V. A. 1978. A winged t'lite in a subcortical 284. beetle as a model for a prototermite. Nature 276:

ROFF,D. A. 1974a. Spatial heterogeneity and the 73-75. persistence of populations. Oecologia 15:245-258. -. 198 1. The adaptive and evolutionary signif-

-. 19746. The analysis of a population model icance of wing polymorphism and parthenogenesis demonstrating the importance of dispersal in a het- in Ptinella Motschulsky (Coleoptera: Ptiliidae). Ecol. erogeneous environment. Oecologia 15:259-275. Entomol. 6239-98.

-. 1975. Population stability and the evolution TAZIMA,Y. 1964. The Genetics of the Silk Worm. of dispersal in a heterogeneous environment. Oeco- Logos, London, U.K. logia 19:2 17-237. THODAY, JR.J. M., AND J. N. THOMPSON, 1976. The

-. 1977. Dispersal in dipterans: Its costs and number of segregating genes implied by continuous consequences. J. Anim. Ecol. 46:443-456. variation. Genetica 46:335-344.

-. 1984. The cost of being able to fly: A study TSAI, P., F. HWANG, W. FENG, Y. Fu, AND Q. DONG. of wing polyrnorphism in two species of crickets. 1964. Study on Delphacodes striatella Fallen (Ho- Oecologia 63:30-37. moptera: Delphacidae) in north China. Acta. Ent.

-. 1986a. The evolution of wing polymorphism Sinica 13552-571. and its impact on life cycle adaption in insects. In UTIDA, S. 1972. Density dependent polymorphism F. Taylor and R. Karban (eds.), The Evolution of in the adult of Callosobruchus maculatus (Coleop- Insect Life Cycles, Springer-Verlag, N.Y. In press. tera, Bruchidae). J. Stored Prod. Res. 8: 11 1-126.

-. 1986b. The genetic basis ofwing dimorphism VANCE,R. R. 1980. The effect of dispersal on pop- in the sand cricket, Gryllusfirmus and its relevance ulation size in a temporally varying environment. to the evolution of wing dimorphisms in insects. Theoret. Popul. Biol. 18:343-362. Heredity. In press. VAN VALEN, L. 197 1. Group selection and the evo-

ROSE, D. J. W. 1972. Dispersal and quality in pop- lution of dispersal. Evolution 2559 1-598. ulations of Cicadulina spp. (Cicadellidae). J. Anim. VEPSLLLINEN, 1973. The distribution and habitats K. Ecol. 41:589-609. of Gerris Fabr. species (Heteroptera, Gerridae) in

SALT, G. 1952. Trimorphism in the Ichneumonid Finland. Ann. Zool. Fenn. 10:419-444. parasite Gelis corruptor. Quart. J. Microsc. Sci. 93: -. 1974a. The life cycles and wing lengths of 453-474. Finnish Gerris Fabr. species (Heteroptera, Gem-

SATO,T. 1977. Life history and diapause ofthe white- dae). Acta. Zool. Fenn. 14 1: 1-73. spotted tussock moth Orgyia thellina Butler (Lep- -. 19746. Determination of wing length and idoptera: Lymantriidae). Jap. J. Appl. Entomol. diapause in water striders (Gerris Fabr., Heterop- Zool. 21:6-14. tera). Hereditas 77: 163-1 75.

SLANSKY,F., JR. 1980. Food consumption and re- -. 1978. Wing dimorphism and diapause in production as affected by tethered flight in female Gerris: Determination and adaptive significance, milkweed bugs (Oncopeltus fasciatus). Entomol. Exp. pp. 2 18-253. In H. Dingle (ed.), Evolution of Insect Appl. 28:277-286. Migration and Diapause. Springer-Verlag, N.Y.

SOTAVALTA,O., AND E. LAULAJAINEN. 196 1. On the WALOFF,N. 1973. Dispersal by flight of leafhoppers sugar consumption of the drone fly (Eristalis tenax (Auchenorrhyncha: Hornoptera).-J. Appl. ~ c o l . 10: L.) in flight experiments. Ann. Acad. Sci. Fenn. A4 705-730. 53:l-25. WATSON,M. A,, AND R. C. SINHA. 1959. Studies on

SOUTHWOOD,T. R. E. 1961. A hormonal theory of the transmission of European wheat striated mosaic the mechanism of wing polymorphism in heter- virus by Delphacodes pellucida. Virology 8: 139- optera. Proc. Roy. Entomol. Soc. Lond. (A) 36:63- 163. 66. WEIS-FOGH,T. 1952. Weight economy of flying in-

-. 1962. Migration of terrestrial arthropods in sects, pp. 341-347. In Transactions of the IXth relation to habitat. Biol. Rev. 37: 17 1-214. International Congress of Entomology, Vol. 1. In-

STEFFAN,W. A. 1973. Polymorphism in Plastosciara ternat. Congr. Entomol., Amsterdam, Neth. perniciosa. Science 182:1265-1266. WIGGLESWORTH, 196 1. Insect polymorphism- V. B.

1020 DEREK A. ROFF

a tentative synthesis. Symp. Roy. Entomol. Soc. Lond. 1:103-113.

WILLIAMS, AND W. H. SAWYER. C. M., L. A. BARNES, 1943. The utilization of glycogen by flies during flight and some aspects of the physiological ageing of Drosophila. Biol. Bull. 84:263-272.

WRATTEN,S. D. 1977. Reproductive strategy of winged and wingless morphs of the aphids Setobion avenue and Metopolophium dirhodum. Ann. Appl. 13iol. 85:319-331.

YOUNG, E. C. 1965. Flight muscle polymorphism in British Corixidae: Ecological observations. J. Anim. Eco~. 34:353-390.

YURKIEWIC~,W. J. 1965. Right exhaustion studies on the blowfly Phaenicia sericata using I4Clabelled sucrose. J. Insect Physiol. 14:335-339.

ZERA, A. J. 1984. Differences in survivorship, de- velopment rate and fertility between the long- winged and wingless morphs of the waterstrider, Limnoporus canaliculatus. Evolution 38:1023- 1032.

ZERA,A. J., D. T. INNES, AND M. E. SAKS. 1983. Ge- netic and environmental determinants of wing polymorphism in the waterstrider, Limnoporus canaliculatus. Evolution 3 7 3 13-522.

ZIMMERMAN, 1948. Insects of Hawaii, Vol. 111. E. C. Heteroptera. Univ. Hawaii Press, Honolulu.

Corresponding Editor: J. A. Endler