Genetica (2010) 138:5–18

DOI 10.1007/s10709-009-9358-y

Evolution of genitalia: theories, evidence, and new directions

William G. Eberhard

Received: 23 October 2008 / Accepted: 26 February 2009 / Published online: 24 March 2009
Ó Springer Science+Business Media B.V. 2009

Abstract Many hypotheses have been proposed to distinguishing closely related species. Taxonomists have
explain why male intromittent genitalia consistently tend to compiled a huge accumulation of data on genital evolution
diverge more rapidly than other body traits of the same that is unparalleled in scope; in many groups, male genital
individuals in a wide range of animal taxa. Currently the morphology is better known than any other trait (reviewed
two most popular involve sexual selection: sexually in Eberhard 1985). What could be responsible for this
antagonistic coevolution (SAC) and cryptic female choice evolutionary pattern? The major objective of this paper
(CFC). A review of the most extensive attempts to dis- (which is condensed from a more extensive treatment in
criminate between these two hypotheses indicates that SAC Eberhard 2009, and was presented as part of a symposium
is not likely to have played a major role in explaining this on genital evolution in the 20th International Congress of
pattern of genital evolution. Promising lines for future, Zoology in Paris) is to briefly review data and ideas that
more direct tests of CFC include experimental modification can help answer this question for a general audience. A few
of male genital form and female sensory abilities, analysis possible new directions to test theory more directly are also
of possible male–female dialogues during copulation, and discussed critically.
direct observations of genital behavior.

Keywords Sexual conflict
Sexual selection
Non-genital contact devices

Sexually antagonistic coevolution
Cryptic female choice
Before beginning, it is necessary to explain that I will
discuss both primary and secondary genitalia, and also non-
Hypotheses concerning the evolutionary forces responsible genital species-specific male structures that are specialized
for the evolution of animal genitalia have a long history of to make contact with females prior to or during copulation
controversy. Genital evolution requires special explana- (non-genital contact devices). I lump these structures with
tions because genitalia are often species-specific, and their genitalia because they all show the same evolutionary
forms are often more divergent among closely related pattern of consistent, relatively rapid divergent evolution,
species than are the forms of other traits such as legs, and thus probably require a similar explanation (Eberhard
antennae, and eyes. This relatively rapid divergence of 1985; Robson and Richards 1936). The ‘‘secondary’’ male
genitalia is extremely widespread taxonomically (Eberhard genitalia (which are distant from the primary genitalia,
1985), and has made them especially useful in such as the pedipalps of spiders, but which house sperm
received from the male’s gonopore and then introduce
them into the female), show the same pattern of relatively
W. G. Eberhard (&) rapid divergent evolution as primary genitalia, and are also
Smithsonian Tropical Research Institute, and Escuela de generally included in discussions of genital evolution.
Biologı́a, Universidad de Costa Rica, Ciudad Universitaria,
Other structures that are not physically close to the geni-
San Pedro, Costa Rica
e-mail: [Link]@[Link]; talia, and that are specialized to contact non-genital parts of
archisepsis@[Link] the female in sexual contexts also show the same pattern

123
6 Genetica (2010) 138:5–18

(see list of 63 groups in Eberhard 2004b; also Darwin Shapiro and Porter 1989). Arguably the most important
1871; Robson and Richards 1936; Eberhard 1985). In fact, evidence against lock and key is that in many groups the
some entomologists have included as ‘‘genitalia’’ some supposed female locks simply do not exist (Eberhard 1985;
non-genital structures (sensu strictu) that are not associated Scudder 1971; Shapiro and Porter 1989). There also other
with the segment on which the genital opening occurs; serious problems, including an almost complete lack of the
these include cerci and sternites near the ‘‘true’’ genitalia in predicted pattern of character displacement in males in
groups in which these structures contact females directly zones of sympatry of closely related species (McAlpine
during copulation, and they also show a pattern of rapid 1988; Ware and Opell 1989), and clear evidence of genital
divergent evolution along with elaborate forms that is species-specificity in species (such as those of species
typical of more strictly genital structures (e.g., Tuxen 1970; endemic to caves and oceanic islands, and of parasites
Wood 1991). Almost any part of the male can be modified isolated from all close relatives in their different hosts) that
in this way, from the sucker-like ‘‘bursa’’ of male nema- have probably evolved in complete or nearly complete
todes to the cephalothorax, the chelicerae and anterior legs physical isolation from all close relatives, and should thus
of spiders, the antennae and telson in crustaceans, and the not have evolved species-specific genitalia (Eberhard 1985,
head, mandibles, antennae, pronotum, cerci, and wings of 1996; Hedin 1997; Shapiro and Porter 1989). While lock
insects. As pointed out by Robson and Richards (1936), the and key is very probably not a general explanation for
mechanical function of many (though not all) of these genital evolution, and the evidence that has been adduced
structures is to grasp the female during copulation, the in its favor in particular studies (Mikkola 2008) has other
same function that is performed by a large fraction of the possible interpretations (Eberhard 2009), the possibility
male genital structures that are species-specific in form that it applies in any particular cases cannot be discarded
(summaries in Eberhard 1985, 2004a; Scudder 1971). The without careful study.
line between ‘‘true’’ genital claspers and non-genital
claspers is, in the end, arbitrary (Chapman 1969; Darwin
1871; Eberhard 1985; see also Ghiselin 2009; Leonard and Sexually antagonistic coevolution and cryptic female
Cordoba-Aguilar 2009). choice
Inclusion of non-genital organs specialized to contact
females is especially useful because they have two Recent developments in several fields facilitated forging
advantages over ‘‘true’’ genitalia: the details of their new links between genital evolution and sexual selection
physical interactions with the female are often better by female choice (Birkhead 1996), and the more recent
understood; and the female structures that they contact are emphases on male–female conflicts during copulation
often more easily manipulated and studied, because contact (Arnqvist and Rowe 2005; Parker 1979, 2005) have
is made with the female’s outer surface rather than her sparked new hypotheses. I will focus below on the two
internal genitalia. I will use the phrase ‘‘non-genital contact currently most popular sexual selection hypotheses to
structures’’ for species-specific male structures that are not explain consistent rapid genital divergence: sexually
near his genitalia and that are specialized to contact antagonistic coevolution (SAC), and sexual selection by
females in sexual contexts. cryptic female choice (CFC). SAC proposes that male and
female interests with respect to events associated with
copulation, insemination, and fertilization are not identical,
Hypotheses to explain genital evolution and that males and females are engaged in coevolutionary
races for control over these processes (Alexander et al.
Lock and key 1997; Arnqvist and Rowe 2005; Chapman et al. 2003;
Holland and Rice 1998). Male adaptations that increase
The classic explanation of rapid divergent genital evolution male control are thought to entail damage to females, such
is species isolation by lock and key: females evolve genital as reduced longevity, reduced ability to produce optimum
structures into which only the male genitalia of their own numbers of offspring at optimum times, etc. that are dis-
species can fit, and thus avoid losing their large investment advantageous under natural selection on females. They are
in eggs due to fertilization by heterospecific sperm; males expected to lead to female adaptations to reduce such
also benefit, though to a lesser extent due to their less damage, and these adaptations in turn are expected to result
costly gametes. Although this theory can explain why in further selection on males for additional adaptations to
genitalia are often species-specific in form and may be control reproduction.
correct in particular cases, it has gradually lost favor as a CFC also proposes that male and female interests are not
general explanation for several reasons, and is probably in identical, but supposes that there are important benefits to
the process of being discarded definitively (Eberhard 1985; females from allowing some but not other males to fertilize

123
Genetica (2010) 138:5–18 7

their eggs, derived from the production of superior off- difficult (if not impossible in some cases) (Cordero and
spring. Male genitalia are thought to be courtship devices, Eberhard 2003).
often stimulating the female inside her reproductive tract.
Sexual selection by female choice occurs after copulation
has begun, with females favoring some male genital Support for CFC and SAC
designs over others via biases in post-copulatory processes
such as sperm transport, oviposition, remating, etc. (Eber- One clear prediction made by CFC is that the frequency of
hard 1996). Male genital designs can be favored because female remating in different groups should tend to correlate
they result in more effective stimulation of the female, or positively with the rate of genital divergence in that group
adjust mechanically more appropriately to her genital (Arnqvist 1998; Eberhard 1985). In particular, if females
morphology. Sexual selection often leads to relatively rapid consistently make genital contact with only a single male,
divergence in many other traits (e.g., structures and then CFC among males on the basis of genital traits is not
behavior involved in visual displays, songs; Darwin 1871, possible. SAC is less categorical, but also predicts a negative
West-Eberhard 1984) in evolutionary independent lines, correlation with monandry (Arnqvist 1998). Conflict
because of the intense selective pressures, the multiple between male and female may be reduced or eliminated by
possible mechanisms by which males can improve their female monandry, especially if the male is also monoga-
competitive abilities, and also because of the shifting set- mous (in which case conflict should not occur and male and
tings in which competition is played out (e.g., female female genitalia should not coevolve). Selection favoring
preferences, the abilities of other males). many possible male manipulations that might be disadvan-
The controversy between CFC and SAC explanations tageous to the female under natural selection (induce earlier
over genital evolution is part of a more general contro- oviposition, more massive ovulation, more resistance to
versy swirling in the recent literature on sexual selection subsequent mating, etc.) is eliminated in monandrous
regarding phenomena formerly attributed to female choice females. If females can benefit from polyandry but males
(e.g., Arnqvist 2004; Arnqvist and Rowe 2005; Chapman ‘‘impose’’ monandry via use of their genitalia, then conflict
et al. 2003; Cordero and Eberhard 2003, 2005; Kokko could arise even in a species in which nearly all females are
et al. 2003; Pizarri and Snook 2003, 2004). While there strictly monandrous. Even if monandry is not imposed by the
have been some misunderstandings (often due to the male, conflict could occur if the male provides the female
mistaken idea that conflict between the reproductive with some resource that is in short supply (e.g., a large
interests of males and females implies that SAC rather nutritious ejaculate), and if the male is polygamous and
than CFC is operating—Cordero and Eberhard 2003), sometimes provides the monandrous female with less than
there is a major contrast between the two hypotheses: she wants. Whether this sort of conflict could ever play out in
under SAC the payoffs to a female from resisting the genital morphology (e.g., the female attempting to induce
sexual attentions of some of the males that she encounters greater male contributions) is not clear.
are to avoid reduction in the number of offspring she Thus CFC clearly predicts that groups with strictly
produces (direct reproduction); under CFC, her payoff monandrous females (or at least females that make genital
is to obtain superior offspring (Chapman et al. 2003; contact with only a single male) should have male genitalia
Arnqvist and Rowe 2005, p. 36). SAC presumes that that are not species-specific in form, while the SAC
resistance to males results in direct gains to the female in expectation is for a mix of unspecified proportions, but
the number of her offspring; CFC supposes that the gains probably mostly a lack of species-specificity. Possible
are from improved quality of her offspring. If genitalia correlation between female monogamy and genital diver-
are typical of other sexually selected traits, then resolution gence has been tested in 22 different groups, including
of the controversy for genitalia will point the way toward termites (males also monogamous) and Heliconius butter-
a more general resolution. flies (Arnqvist 1998; Eberhard 1985), bees (Roig-Alsina
Discriminating between CFC and SAC explanations for 1993), primates (some males polygynous; Dixson 1987,
genital evolution is difficult. The two hypotheses are not 1998; Verrell 1992), Ischnura dragonflies (Robinson and
mutually exclusive. Theoretically they can act simulta- Novak 1997; Simmons 2001), mole rats (Parag et al. 2006),
neously and can either reinforce or counteract each other and in 16 other groups of insects (Arnqvist 1998).The
(Arnqvist and Rowe 2005; Cordero and Eberhard 2003, prediction of reduced genital divergence was fulfilled in all
2005; Eberhard 2004b; Hosken and Stockley 2004); they cases, despite the fact that the prediction involves rates of
could also act one after the other on the same trait (Eber- genital divergence, while the data in most cases involved
hard 2004b). Some predictions are the same for both, and amounts of divergence, and in some groups the behavioral
direct measurements of some crucial variables involved in data were incomplete (e.g., Arnqvist 1998). There were
balancing potential costs and benefits is technically generally no controls for the amount of time since

123
8 Genetica (2010) 138:5–18

divergence, although Arnqvist’s (1998) finding that geni- (Rodriguez 1994; Rodriguez et al. 2004), suggest that
talia but not other structures correlated with the frequency mechanical fit in the female’s rigid, tortuous spermathecal
of female remating suggests this was not a problem in his duct, rather than stimulation, may be an important deter-
study. In general, this evidence favors both CFC and SAC. minant of sperm precedence. Sperm is deposited in this
Differences in male genital morphology have been duct, but it is also deposited outside the duct in the female’s
found to correlate with paternity when a female mates with bursa, and the significance of bursal sperm remains
more than a single male in five insect species: two in the unclear. No male-inflicted damage to female reproduction
water strider genus Gerris (Arnqvist and Danielsson 1999; (as predicted under SAC) is known, and the highly scleri-
Danielsson and Askenmo 1999); two distantly related tized spermathecal duct seems unlikely to be damaged by
scarab beetles, Onthophagus taurus and Anomala orientalis the male. Nevertheless, damage has never been searched
(House and Simmons 2003; Wenninger and Averill 2006); for, and might occur deeper in the female (e.g., the sper-
and the chrysomelid beetle Chelymorpha alternans mathecal valve; D. Windsor pers. comm.). In the oriental
(Rodriguez et al. 2004). In addition, experimental modifi- beetle, A. orientalis, the male sclerite whose size correlates
cations in the sepsid fly Archisepsis diversiformis of both with paternity hooks the female just inside her vagina,
the morphology of a non-genital clamping structure on the where it probably stimulates her and may also provide
male’s front leg, and of the female’s ability to sense this purchase for deeper thrusting by other, inflatable portions
organ reduced the likelihood of female acceptance of of his genitalia (Wenninger and Averill 2006). Possible
copulation (Eberhard 2002a). These cases support CFC, but damage to the female was not checked. Very little is known
some are also compatible with SAC. about how the male genital structures that correlate with
SAC is unlikely in probably the best understood species, paternity are used in Gerris. A failure to even consider
A. diversiformis. The male’s clamp fits very precisely with stimulation of the female as a possible function in the study
the female’s wing (Eberhard 2001a), but experimental of genital function in O. taurus (Werner and Simmons
modification of the male’s clamp did not impair his ability 2008) makes functional interpretations difficult to evaluate
to hold onto the female with his front legs (on the basis of in this species.
the durations of riding times; Eberhard 2002a), arguing The data from four of these species have another pos-
against a SAC interpretation. Female receptors that sense sibly important limitation. Except for the chrysomelid and
stress in the cuticle occur in the area contacted by the sepsid studies, only correlations, rather than cause and
male’s front leg in this species (Eberhard 2001a; as well as effect relationships were demonstrated. In addition, no
in other sepsid species with species-specific male front control was devised in the chrysomelid study for the effects
legs—Ingram et al. 2008), and could thus enable her to of the operation itself (it was not feasible to cut the male’s
sense his grip, supporting a CFC interpretation. The genitalia and then glue them back together). It should be
female’s wing base is quite sturdy, and there were no signs noted, however, that the tests in all six species were con-
of damage (a possible prediction of SAC). Parenthetically, servative, in that they did not take into account the possible
male damage to female wings by clasping them was effects of male genitalia on many other female reproduc-
claimed in another sepsid with similar male grasping tive processes, such as decreased remating or increased
devices (Mühlhäuser and Blanckenhorn 2002), but the oviposition.
wing damage that was observed was in other parts of the One type of evidence that supports CFC and some but
wing, and likely occurred when flies beat their wings not other versions of SAC comes from a growing number
against the walls of their small glass containers (see Baena of observations of genitalia used in ways that are appro-
and Eberhard 2007). In addition, in only one of the [10 priate to stimulate the female, but not to physically coerce
sepsid species (in the genera Archisepsis, Microsepsis, her. Probably the genital behavior in which stimulation is
Palaeosepsis, Sepsis, and Themira) is there any even least controversial is stridulation, which has been observed
potentially defensive modification of the female’s wing in directly in the tipulid fly Bellardina sp. (Eberhard and
the area where the species-specific modifications of the Gelhaus 2009), and inferred from male morphology in
male’s front legs grasp her (Eberhard 2001a, 2005, unpub.; moths (Forbes 1941; Gwynne and Edwards 1986) and
Ingram et al. 2008). wasps (Richards 1978). Copulation in various mammals
None of the other species allow such clear discrimina- also involves behavior apparently designed to stimulate the
tion between CFC and SAC. In C. alternans, the increased female with the genitalia (summary Eberhard 1996; see
paternity associated with greater length of one male genital also Dixson 1998), and some aspects of copulation
structure, the effects of experimental shortening this behavior in rodents correlate with indicators of increased
structure, morphological studies of how male genitalia probability of competition with sperm from other males
engage the female during copulation, and the dramatic (Stockley and Preston 2004). In addition, the male genitalia
variation in the ducts of females of different species of several insects and spiders perform long, highly

123
Genetica (2010) 138:5–18 9

rhythmic series of taps, or squeezes on membranous por- and the relative development of female defensive struc-
tions of the female, that also suggest that stimulation of the tures. An independent contrasts analysis based on a robust
female is important; these include a dryomyzid fly (Otro- phylogeny showed that changes in male and female mor-
nen 1990), a buprestid beetle (Eberhard 1990), a sciarid fly phological traits (both genital and non-genital) probably
(Eberhard 2001c), several sepsid flies in different genera coevolved. Another type of possible female counter-adap-
(Eberhard and Pereira 1996; Eberhard 2001b, 2003, 2005), tation would be species-specific resistance behavior that is
a pholcid spider (Huber and Eberhard 1997; Peretti et al. appropriate to free her from species-specific male grasping
2006), some scathophagid flies (Hosken et al. 2005), sev- structures; but such specialized behavior has never been to
eral species of tsetse flies (Briceño et al. 2007; Briceño and my knowledge observed for any species.
Eberhard 2009), and the hesperiid butterfly Urbanus Even in Gerris it is difficult to rule out CFC, however.
dorantes and the katydid Idiathron sp. (W. G. Eberhard Possible additional, stimulatory effects on females from
unpub.). In those groups in which the genital behavior of male genitalia have never been checked (e.g., by covering
more than one congeneric species is known (the spider, or inactivating sense organs at the tip of her abdomen). In
tsetse flies, and the sepsid flies Microsepsis and Archisep- addition, the expectation that such a clear case of SAC
sis), the temporal patterning of squeezes differs among might lead to morphological diversity in males and females
species (Briceño and Eberhard 2009; Eberhard 2001b; is not clearly fulfilled. The morphological designs of both
Eberhard and Pereira 1996; A. Peretti, pers. comm.), as sexes of Gerris do differ somewhat among species, but
would be expected to often be the case if this behavior is both male and female structures are relatively simple and
under sexual selection by CFC. Alternative SAC interpre- practical. A morphologically similar abdominal spine that
tations based on physical coercion can be ruled out in some can fend off males also occurs in female Aquarius palu-
of these cases. Direct male effects on internal female drum, but spines also occur in males (where they are
genital structures with squeezing behavior are not possible proportionally longer); female fertility in captivity is
on morphological grounds in the sepsids and the crane fly. increased rather than decreased (as predicted by SAC) by
Possible external physical damage to the female resulting additional matings in this species (Ronkainen et al. 2005).
from male movements may occur in some but not all of the Several other species provide possible support for SAC.
tsetse flies and the pholcid spider, but not in the sepsids or In Lucilia blowflies, complex, species-specific male genital
the katydid (data are not sufficient to judge in the others asperities (Aubertin 1933) rub holes in apparently defen-
groups). sive thickenings in the lining of the female’s reproductive
SAC cannot be ruled out, however, because it is possible tract (Lewis and Pollock 1975; Merrett 1989). Species-
that male stimulation of the female sometimes leads to specificity in female defensive morphology, and the ques-
losses of offspring for the female (Arnqvist 2006) when tion of whether female reproduction is actually reduced by
males use sensory traps. Such traps may have played a copulatory damage both remain to be checked, however.
large role in the early stages of cryptic female choice on The possibility also remains that stimulation (which seems
genitalia (Eberhard 1996). Female countermeasures could likely to occur) induces female responses favoring the
be in her sense organs and nervous system, and thus male, so CFC can not be ruled out.
invisible externally. If such a coevolutionary struggle Summarizing, very few species give direct evidence that
between males and females did not ‘‘spill over’’ into compellingly discriminates between the CFC and SAC
struggles involving physical coercion, it would not be hypotheses. I think the clearest data favoring CFC over
visible in studies of external morphology such as those SAC come from the front leg grasping organs of sepsid
reviewed in this paper (Eberhard 2004a). flies, and from some species with male genitalia that are
The strongest support for SAC in genitalia comes from designed to stimulate the female. The strongest support for
water striders in the genus Gerris. Dorsally projecting SAC comes from Gerris water striders, but here no
spines near the female’s genitalia are elongated to different attempts have been made to rule out CFC.
degrees in different species, and have independently
become especially elongate in Gerris incognitus and Gerris
odontogaster. Longer female spines impede male attempts Discriminating between SAC and CFC
to clamp the tip of the female’s abdomen with his genitalia
(Arnqvist and Rowe 2002a, b; Rowe and Arnqvist 2002). SAC and CFC predictions vary in some contexts, and the
Such clamping helps the male hold onto the female during massive data bank on genital evolution that is available in
her energetic struggles when he mounts, and is a necessary the taxonomic literature permits the use of huge sample
prelude to intromission. There is a cross-specific correla- sizes to test them. I will discuss what I consider the two
tion between the relative development of several different most powerful sets of evidence available regarding the
male structures, including elongate grasping male genitalia, likely generality of SAC and CFC explanations for genital

123
10 Genetica (2010) 138:5–18

evolution (for a more complete discussion, see Eberhard those in which males station themselves near oviposition or
2009). Both tests strongly favor CFC over SAC. feeding sites and attempt to mate with arriving females,
and those in which males station themselves at sites where
Groups in which males can and cannot coerce females females are emerging from pupae and mate with then while
to mate they are still relatively defenseless. For each genus in
which behavioral evidence suggests that females are con-
This test is based on a SAC prediction made by Alexander sistently either protected or unprotected, taxonomic studies
et al. (1997), who distinguished between coercive and non- were consulted to determine whether male genitalia are or
coercive circumstances in which males attempt to obtain are not useful in distinguishing closely related species.
copulations. Grasshopper males were cited as mating The data clearly failed to conform to the SAC prediction
coercively, because they often jump onto females that are (Fig. 1a). A total of 75.4% of 236 genera with protected
engaged in other activities, and without any preliminaries females have species-specific male genitalia, while 68.8%
attempt to grasp the female’s genitalia with their own. of 125 genera with unprotected females have species-spe-
Females often struggle forcefully to dislodge males and to cific male genitalia (data from 113 families in 10 orders).
prevent genital coupling. The example of non-coercive The difference is not significant (X2 = 1.82, df = 1,
mating was in the cricket genus Gryllus, in which males P = 0.17), and the trend was in the opposite direction from
produce a calling song; the receptive female, with no overt that predicted by SAC. Re-analyses designed to correct for
coercion by the male, approaches the male and positions possible biases in the data (over-use of genitalia by tax-
herself to allow him to couple with her. The female cannot onomists due to tradition, under-use of genitalia due to the
be physically coerced, because she only encounters the difficulty of studying them, inadvertent biases with respect
male if she seeks him out (i.e., she is protected from to which groups were included in the study, and phyloge-
unwanted male attentions). Alexander et al. reasoned that netic inertia) all failed to result in the predicted trend.
SAC in male and female genitalia would be more likely to Analyses of more taxonomically restricted groups, such as
occur in grasshoppers because male and female interests the large fly family Chironomidae, in which additional
were more clearly in conflict. The two groups they dis- behavioral details increase confidence in the lack of male–
cussed fit their prediction: male genitalia are often species- female conflict, also failed to fit the SAC prediction. These
specific in grasshoppers, but are simple and not divergent numbers, in fact, strongly underestimate the strength of the
in species of Gryllus (Alexander et al. 1997). evidence against SAC, because data from the large order
Subsequently this sample was greatly enlarged by using Lepidoptera (approximately 250,000 species) also contra-
publications on behavioral ecology and taxonomy to dis- dict SAC predictions but were omitted from the totals.
criminate between SAC and CFC (Eberhard 2004a). Female lepidopterans are nearly all protected from pre-
Discrimination is possible, because CFC predicts that no copulatory male coercion, because females throughout the
trend should occur: female use of male genitalia to bias order attract males with long distance attractant phero-
paternity could occur equally well in species with protected mones (Phelan 1997). Contrary to SAC predictions, the
or unprotected females (unless unprotected females are genitalia of most lepidopterans are nevertheless elaborate
more likely to be monandrous due to male manipulations, and species-specific in form, as shown in taxonomic
in which case the prediction is the opposite—there should compendia that review thousands of species in the North
be greater genital divergence in non-coercive mating American and Palaearctic fauna (Amsel et al. 1965–2000;
systems). Dominick et al. 1971–1998; Forster and Wohlfahrt 1952–
A list of groups (in most cases genera) in which females 1981; Huemer et al. 1996).
are more or less likely to be coerced into mating by males The SAC prediction of Alexander et al. (1997) of a
was compiled from publications on the behavioral ecology negative correlation between species-specificity in genita-
of insects and spiders. Protection of females from coercion lia and protection of females from male harassment needs
was assumed in species in which males attract females by to be tempered, however, because even though a female is
chemical signals or singing, females attract males with receptive to copulation and thus to genital coupling, she
attractant pheromones, females emit light signals at night might not be receptive to genital manipulations performed
in response to light signals from the males that allow the after copulation has begun. To estimate how frequently
male to find them, males form leks or swarms that are not different species-specific male genital structures function
associated with resources needed by females such as ovi- in different ways, I made a separate literature survey of
position or feeding sites, and in spiders in which males are studies of the functional morphology of male genitalia in
dwarfs in comparison with females (and the female can the order Diptera. The results indicated that the prediction
thus easily kill a harassing male). In contrast, species in of Alexander et al. is reasonable, though not absolute.
which females are not protected from harassment included Of 105 cases in which a function was attributed to a

123
Genetica (2010) 138:5–18 11

a b Female coevolution
possibly defensive
Female coevolution
Female protected not defensive

Female not protected No female coevol.

100 100
Genera with species-specific genitalia (%)

102
224
80 236
196 80
125 236 98
125
236125
60
60

40 40

20 20

Totals Very Small “No Original Genitalia non-

small mention” list (N=43) Genitalia
(N=63)

Fig. 1 Summary of data from two large survey studies documenting at tops of bars area sample sizes). The SAC prediction that the dark
failures of confirm predictions of the SAC hypothesis. a Percentages bars would be higher was not confirmed. b Conservative estimates of
of genera in which male genitalia are and are not species-specific, fractions of the 84 taxonomic groups with species-specific male
contrasting groups in which non-receptive females are protected from genitalia (left) and non-genital contact structures (right) that did
sexual harassment by males with groups in which non-receptive (black) and did not (grey and white) conform to predictions of
females are not protected. The totals (left pair of bars) include all species-specific defensive female coevolution from physical coercion
groups examined; the other pairs of bars represent data that were versions of the SAC hypothesis (after Eberhard 2009; data from
modified in different ways to attempt to take into account different Eberhard 2004a, b)
possible biases in the data against SAC predictions (see text; numbers

species-specific male genital structure (in 43 species in 22 of this trend constitutes strong evidence against SAC as a
families), 85.7% were functions in which male–female general explanation.
conflict should be reduced or absent in species with pro- In sum, data from literally hundreds of thousands of
tected females (39.0% of the attributed functions involved species failed to show the negative relation between spe-
clasping the female, and 46.7% facilitating penetration and cies-specificity in genital morphology and female
sperm transfer; Eberhard 2004a). These precise percent- protection from male harassment predicted by SAC, even
ages are not especially meaningful, because of probable when analyses were modified to attempt to take into
biases in the ease of documentation of these particular account possible biases against SAC. If anything, the trend
functions compared with others (e.g., stimulation), and a was in the opposite direction.
bias in the possible functions that were considered by the
authors. But the percentages show clearly that the SAC Female defensive coevolution with males
predictions should have been met in an appreciable number
of genital traits. Unless Diptera are somehow unrepresen- A second broad survey (Eberhard 2004b) examined a dif-
tative of other insects in this respect (there is no obvious ferent set of predictions in 61 families, mostly of insects
reason to suspect this; species-specific claspers are com- and spiders, in which studies are available on the functional
mon, for instance, in Lepidoptera), these percentages morphology of species-specific male structures. Species
indicate that the proposed survey was indeed a valid test of were included in which studies have determined the site on
SAC predictions. The large sample in this study, made the female that is contacted by the species-specific portions
possible of course by the huge taxonomic literature on of the male structure, and the mechanical details of the fit
genitalia, is rare in comparative evolutionary studies. A between them. The physical coercion version of the SAC
sample of this size should have been sufficient to reveal hypothesis makes several clear predictions for these
even a weak trend in the predicted direction. Thus the lack structures: the female morphology should generally

123
12 Genetica (2010) 138:5–18

coevolve with the species-specific aspects of the male; the SAC hypothesis, facultatively defensive structures, was
species-specific female structures of related species should completely absent (0% of 106). An extended search for
interact mechanically with the species-specific portion of defensive designs in an additional, large set of spider
the male; and the designs of the species-specific aspects of species (in which drawings of the female genitalia are
the female structures should often be appropriate to defend routinely included in taxonomic descriptions) failed to
her against the male, and in particular against the action of reveal a single example of such a facultatively deployable
his species-specific structures. Female structures that can defensive device among thousands of species included in
hold the male away or impede his access should be com- general faunal studies and recent reviews.
mon. Finally, because under SAC females need to mate at It would be possible to rescue the physical coercion
least once even though they resist other males, an espe- version of the SAC hypothesis from these apparently
cially likely design would be species-specific female contradictory data if it turned out that in the many species
structures that can facultatively defend against males. in which females that lack species-specific defensive
Moveable structures such as erectable spines, inflatable morphology, females use species-specific defensive
sacs, or sliding barriers that could be moved out of the way behavior instead of morphology, and that this behavior
to facilitate the copulation or copulations needed to obtain selects for diversity in male contact structures (Eberhard
sperm, but interposed to reject others are expected. 2004b). To my knowledge, however, not a single case of
CFC, in contrast, predicts that external female mor- such female behavior has ever been documented (though it
phology will often (but not always) not vary when females must be admitted that perhaps female behavior is seldom
are screening males on the basis of the stimuli they pro- studied with sufficient detail). In addition, the details of
duce. Rather females are expected to coevolve with respect male–female interactions in 21 genera of the 84 taxonomic
to their sense organs and perhaps even more likely with groups allow confident rejection of the female defensive
respect to how their CNS processes the information from behavior possibility; species-specific female resistance that
these sense organs. Females can also screen males on the could select for the variant designs seen in different species
basis of their morphological fit with the female, however, of males is either mechanically impossible, or female
so some cases of male–female morphological coevolution behavior has been observed with sufficient detail to rule it
are expected. The designs of females are expected to often out (Eberhard 2004b). In nine other genera, it is the female
be ‘‘selectively cooperative’’, rather than defensive, as that approaches the male and actively maintains contact
expected with the physical coercion version of SAC. I with him, rather than vice versa; she is thus free to break
assembled a collection that included 43 species-specific away at any time, so the female has no need for special
male genital structures in 34 taxonomic groups and 63 ‘‘resistance’’ behavior that could have favored the male’s
species-specific male non-genital contact structures in 53 species-specific morphology (Eberhard 2004b). In sum,
taxonomic groups, mostly insects and spiders. The female morphology in an entirely independent, large
assembled groups were then checked for female traits. sample of taxa also fits CFC predictions much better than
Once again, the SAC predictions clearly failed. Of 106 predictions of the physical coercion version of the SAC
structures in 84 taxonomic groups (Fig. 1b), in more than hypothesis. The stimulation version of SAC does not sup-
half (53.8%) female morphology was inter-specifically pose male–female morphological coevolution, so is not
uniform even though the male morphology was species- affected by these observations.
specific (the respective percentages for genital and non-
genital structures were 34.9% of 43, and 68.3% of 63). In
addition, the designs of over half of those female structures Conclusions regarding CFC and SAC
that did coevolve with species-specific structures of males
did not have the predicted defensive designs: among 49 The current balance of evidence is tilted against SAC as a
coevolving female structures in 39 taxonomic groups, general explanation (for a more complete discussion, see
55.1% were not even feasible as defensive devices (57.1% Eberhard 2009). This is not to argue that SAC on genitalia
of 28 genital structures and 52.4% of 21 non-genital and non-genital contact structures never occurs. Even in
structures). The female designs seemed to be selectively cases in which SAC seems especially unlikely to have
cooperative in many species (grooves and furrows used by shaped current morphology, it may nevertheless have
a male with the appropriate design as sites to support or occurred at certain moments in evolution. Take, for
strengthen their grip on the female) rather than defensive. instance, the sepsid flies, a relatively well-studied group
In total, females failed to confirm to these SAC predictions in which SAC seems especially unlikely to explain the
in 79.2% of 106 structures. present-day morphology of the modified male front legs
In addition, the female design that constituted arguably that clamp the female’s wing. The area near the base
the strongest prediction by the physical coercion of the of the female wing that is contacted by the elaborate,

123
Genetica (2010) 138:5–18 13

species-specific clamping devices on the males’ front These conclusions are not completely convincing,
legs is relatively uniform in different species and genera, however. The molecular evidence that a single male sires
and seldom has potentially defensive traits, so female all of a female’s offspring does not necessarily demonstrate
morphological coevolution with males has generally not the strict female monandry; the female might have geni-
occurred (Eberhard 2001a; Ingram et al. 2008). In addition, talic contact with other males, some copulations might
experimental data support the female choice rather than the occur without insemination, or some copulations may
SAC hypothesis (Eberhard 2002a). Nevertheless, it is quite involve complete dumping of sperm from previous males.
possible that SAC played a role at early stages in the Perhaps further observations will reveal attempted cou-
evolution of the clamping structures of male sepsids. plings in which the male stipes, volsella and squama
Energetic female shaking behavior that could dislodge contact other portions of the female that fail after contact
males is widespread in sepsids (Eberhard 2005; Ingram and thus give evidence of female rejections on this basis
et al. 2008; Parker 1972; Ward 1983), and also in other (contact seems to usually lead, however, to copulation—
related flies in which the male’s front legs are not modified P. Schmid-Hempel, pers. comm.). The displacement of
(Crean and Gilburn 1998; Eberhard 2000). Shaking may sperm from previous males (despite apparent mating
have originally evolved in sepsids due to male-inflicted plugs—Sauter et al. 2001) might obliterate traces of
losses to females when males began to ride them for long polyandry, and save the CFC hypothesis from these
periods at oviposition sites; a female with a riding male is apparently contradicting data. Alternatively, further
probably less able to avoid predators. Early modifications understanding of Bombus may lead to a new theory.
of the male’s femur that allowed him to couple his leg Another paradoxical group is the carabid beetle genus
more tightly to the female’s wing may have represented an Platynus, in which changes in female genital traits
antagonistic coevolutionary male response to female (development of a dorsal pouch of the bursa its subsequent
shaking behavior; similar modifications have also occurred sclerotization and narrowing) have not occurred in step
in the males of some other related flies (Dodson 2000). with the evolution of associated traits of the male genitalia
Subsequently, however, it is likely that further modifica- (various modifications of the tip of the median lobe) that fit
tions of the sepsid male legs that resulted in the great into the pouch; instead, changes in females have preceded
diversity of forms in modern species evolved under sexual those in males (Liebherr 1992). Perhaps the female pouch
selection by female choice (Eberhard 2004b). has other functions; evolution of female structures adapted
only to male structures that have not yet evolved is para-
doxical under any of the hypotheses. This group merits
New frontiers further study.

Speculating on where scientific research will go in the Genital dialogues between males and females
future is difficult. There are, however, some types of
missing data that would clearly help solve present prob- The emphasis in studies of copulation behavior has gen-
lems in understanding. erally been on the behavior of males. For instance, my own
study of copulatory courtship in insects and spiders (Eb-
Paradoxical species erhard 1994) concentrated on the possibility that males
perform courtship during copulation, and neglected the
Further study of species that seem anomalous under the possibility that females also perform communicatory
two presently popular hypotheses is likely to be especially behavior during copulation that can influence paternity.
rewarding. The bumblebee genus Bombus appears to fal- Nevertheless, female behavior during copulation that could
sify predictions of both CFC and SAC hypotheses. The be communicatory in function is apparently common. A
complete lack of modification of the portion of the female’s conservative count indicates that it occurred in at least
abdomen that is contacted by the diverse, species-specific 12.2% of 131 species of insects and spiders in that study
stipes, volsella and squama of the male genitalia (Richards (see Peretti et al. 2006). It seems likely that many of the
1927) clearly contradicts the prediction of the physical female signals during copulation are functionally related to
coercion version of SAC that female morphology will the behavior of male’s genitalia, and that they may repre-
coevolve with that of males. Female Bombus are also sent a window on understanding what the male genitalia
thought to be strictly monandrous but male genitalia are are doing. A recent study of stridulation by a female spider
nevertheless complex and strongly species-specific in form during copulation showed that understanding both the
(see Eberhard 2009 for details), contradicting CFC pre- male’s genital behavior and his paternity success was
dictions and failing to follow the less strict trended improved by taking into account this female communica-
predicted by SAC. tive behavior (Peretti et al. 2006).

123
14 Genetica (2010) 138:5–18

Direct observations of the behavior of genitalia Experimental manipulations of male and female
structures
Direct observations of the behavior of male genitalia are
often difficult, because they are generally hidden within the A general weakness of most supposed demonstrations of
female during copulation. Nevertheless, the dream of CFC-type effects of male genitalia on female reproductive
directly observing their behavior is now in reach. The processes is that they have only documented correlations,
removal of the male’s head or anesthesia (sometimes rather than direct cause and effect relations. Experimen-
combined with mild stimulation of the male genitalia) tation, in particular by alteration of the species-specific
releases genital behavior in some species (Roeder (1967) aspects of the male and the corresponding traits of the
on mantids; West-Eberhard (1984) on a wasp; Eberhard female, is needed to establish cause and effect (e.g.,
and Pereira (1995); Briceño et al. (in preparation) on flies). Rodriguez et al. 2004). Such experiments are feasible in
In addition, it is possible to observe male genital behavior groups with male genitalia that are large and tough
within the female in some species such, as the tsetse fly enough to resist experimental modifications (e.g., removal
G. pallidipes (Briceño et al. in preparation), using real time of at least portions of particular, non-inflatable structures)
phase contrast synchrotron X-ray videos; some of the same without disabling the male; the use of lasers to modify
coordinated and surprisingly energetic movements seen in small structures is particularly promising (M. Polak, pers.
headless males occur inside the female during copulation comm. on Drosophila). A good example of the possible
with intact males. payoffs from this type of experiment comes from a recent
study of the tsetse fly Glossina pallidipes. Males have
Female sense organs sexually modified, species specific cerci and other genital
structures that clasp the female’s abdomen during copu-
The CFC hypothesis predicts that in all groups, other than lation (Briceño et al. 2007). Alteration of the species-
those lacking rigid species-specific female structures that specific form of the male cerci and of his abdominal
might be filtering males on the basis of mechanical fit (e.g., sternites elicited changes in post-copulatory female
most spiders—see Eberhard and Huber 2009), females reproductive responses, suggesting that the male struc-
should have sense organs in the area contacted by species- tures are under sexual selection by cryptic female choice:
specific portions of male genital structures. Female sense ovulation and the amount of sperm arriving in the sper-
organs are also possible, though not necessarily predicted, mathecae were reduced, and female tendency to remate
on rigid structures that are contacted by species-specific was increased (Briceño and Eberhard 2009). In addition,
male structures (e.g., the wing bases of sepsid flies). This inactivation or modification of possible sense organs at
prediction thus constitutes a strong test of CFC for some the sites on the portions of the female’s external surface
types of female structures, but female sense organs have where these male grasping structures contact her produced
almost never been studied (see, however, Battin (1993) and similar female responses. Together with data from phys-
Robertson and Paterson (1982) on the thorax of damselflies; iological studies showing that physical stimulation of the
Córdoba-Aguilar (2005) on the oviduct of a damselfly; Eb- female during copulation rather than male seminal prod-
erhard (2001a, 2005) and Ingram et al. (2008) on the wings ucts or products released from spermathecae that contain
of sepsid flies; M. Djernaes et al. (unpublished) on genital sperm elicit these female responses (Gillot and Langley
sclerites in four species of cockroaches). It is also not clear 1981; Saunders and Dodd 1972), these experiments give
whether females utilize generalized receptors with other strong evidence that CFC explains the evolution of these
functions that were already present in the area that is con- species-specific male structures. The sites on the female
tacted by the male, or whether they tend to evolve special show no sign of coevolution with the male; they lack
sensors whose placements or other characteristics coevolve defensive species-specific adaptations that could impede
with the form of the male. Both distributions are compati- male attempts to grasp her, and species-specific resistance
ble with CFC, because even without specially-positioned behavior to these structures does not occur, arguing
receptors in each species, female preferences could result against SAC.
from differences in processing deeper in the CNS. Receptors Non-genital contact devices of males are particularly
that have other functions in addition to sensing males would attractive objects for further experimental studies of this
seem less likely to show special distributions that reflect the type to test CFC predictions: their forms are relatively
shapes of the male organs that stimulate them. The presence easily modified; their mechanical functions are frequently
or absence of sensors is not a test for the stimulation version easily understood (usually clasping the female); and the
of the SAC hypothesis, which is compatible with both sites on the female that they contact are also easily
(Arnqvist 2006). modified.

123
Genetica (2010) 138:5–18 15

Limited usefulness of experimental measurements of balance under natural conditions. Attempts to solve the
fitness ‘‘ecological realism’’ problem by using organisms that
have spent many generations in captivity (Orteiza et al.
Experimental alteration of selective regimes on genitalia 2005) are also problematic, because adaptations to the new
and female monogamy/polyandry in captive populations has captive environments are likely to be only partial (for a
illuminated aspects of genital evolution and male–female specific case in Drosophila, see Eberhard 2009). Female
conflicts (e.g., Miller and Pitnick 2003, Arnqvist and Rowe reproductive processes cannot be assumed to be finely
2005). Experimental manipulation of the intensity of male– adjusted to conditions in captivity.
female conflicts (e.g., by imposing strict monogamy on In sum, partial evolutionary responses in captivity will not
males and females) is also a powerful tool (Arnqvist and be enough to justify confident conclusions regarding the
Rowe 2005). Nevertheless, I do not share the optimism of ambitious undertaking of obtaining measurements of the
some of the leading workers on CFC and SAC (e.g., Hosken balance between absolute values of direct and indirect effects
and Stockley 2004; Holman and Snook 2006; Moore et al. on female fitness that can be used to evaluate the relative
2003; Orteiza et al. 2005; Pizarri and Snook 2003; Rice and importance of the SAC and CFC evolutionary theories. Tests
Chippendale 2001) that another use of studies of captive in captivity can be (and often are) very sophisticated tech-
populations, to study the overall reproductive costs and nically; nevertheless they are only relatively crude in their
benefits to females in the laboratory, is likely to resolve the theoretical implications for the SAC–CFC controversy.
relative importance of SAC and CFC in the evolution of
genitalia (or other traits). I say this despite the fact that the
crucial difference between CFC and SAC models hinge on Conclusion
the balance between a female’s overall gains from traits that
result in rejecting some males in terms of direct fitness. Genitalia are surely the best known traits in the animal
Direct gains (fewer offspring lost when she avoids male kingdom with respect to how species-level divergence
manipulations) are expected to be larger if SAC is operating evolves. With what other traits could one imagine the pos-
(Arnqvist and Rowe 2005; Chapman et al. 2003), while sibility of further, yet-to-be-done tests of SAC and CFC
indirect gains (improved offspring quality) are expected to ideas using data that have already been collected, with a
be larger if CFC is operating (Eberhard 1996). The most taxonomic sweep ranging from leeches, molluscs, and mites
direct means of resolving the CFC–SAC controversy would to primates, snakes and bats to nematodes? The profound
seem to be to measure these costs and benefits. But any direct coverage down to the level of species for literally hundreds
comparison requires accurate measurements of the magni- of thousands of species is unparalleled in any other trait.
tudes of both types of fitness to determine the sign of the Genitalia may be especially useful in understanding sexual
difference between them. It is not enough (as is often true in selection action on ‘‘arbitrary’’ male signals to females
many evolutionary studies) to simply determine the sign of (Eberhard 1993). Many related questions remain unan-
the effect—whether it is positive or negative. Pizarri and swered, and answers may have important implications for
Snook (2003, 2004) make a related point: it is necessary to understanding genital evolution and sexual selection in
utilize male and female fitness, rather than arbitrary phe- general. Happily this potential is beginning to be realized.
notypic traits, if experimental approaches to testing SAC are Much work is yet to be done to harvest the abundant payoffs
to be useful. promised by the pioneering discoveries of taxonomists.
Measuring fitness accurately is not child’s play, to say
the least. A large part of the difficulty stems from the trite Acknowledgments I am extremely grateful to Michael Schmitt and
Dominique Joly for honoring me by organizing a symposium on genital
but true idea that the costs and benefits to the female must
evolution. I also thank Daniel Briceño, Marie Djernaes, Rudolf Meier,
be measured under ‘‘ecologically realistic’’ conditions Alfredo Peretti, Hojun Song, and Nick Tatarnic for access to unpub-
(Cordero and Eberhard 2003). The costs and benefits must lished work, Y. Kamimura and P. Schmid-Hempel for permission to
be similar in sign and magnitude to those under which the quote personal communications, David Hosken, Santosh Jagadeeshan,
Dominique Joly, Rafael Lucas Rodriguez, and two referees for useful
animals evolved, if one wishes to make arguments con-
comments. My research was supported financially by the Smithsonian
cerning why some traits and not others occur in present-day Tropical Research Institute and the Universidad de Costa Rica.
organisms. Unfortunately, precise measurements of both
direct and indirect payoffs in the field are extremely diffi-
cult to obtain; they will be impossible in model species References
such as Drosophila melanogaster and Tribolium castae-
Alexander RD, Marshall DC, Cooley JR (1997) Evolutionary
neum, in which the natural habitat is not even known. perspectives on insect mating. In: Choe JC, Crespi BJ (eds)
There is no guarantee that the balance of gains and losses The evolution of mating systems in insects and arachnids.
under captive conditions is a reliable indicator of the Cambridge University Press, Cambridge, pp 4–31

123
16 Genetica (2010) 138:5–18

Amsel HG, Gregor F, Reisser H (eds) (1965–2000) Microlepidoptera (Tephritidae): phylogeny and evolution of behavior. CRC Press,
Palaearctica. Goerg Fromme, Vienna Del Ray, pp 175–184
Arnqvist G (1998) Comparative evidence for the evolution of Dominick RB, Ferguson DC, Franclemont JG, Hodges RW, Munroe
genitalia by sexual selection. Nature 393:784–786 EG (eds) (1971–1998) The moths of America North of Mexico.
Arnqvist G (2004) Sexual conflict and sexual selection: lost in the E. W. Classey Limited and RBD Publications, London
chase. Evol Int J Org Evol 58:1383–1388 Eberhard WG (1985) Sexual selection and animal genitalia. Harvard
Arnqvist G (2006) Sensory exploitation and sexual conflict. Philos University Press, Cambridge
Trans R Soc B 361:375–386. doi:10.1098/rstb.2005.1790 Eberhard WG (1990) Genitalic courtship in Acmaeodera impluviata
Arnqvist G, Danielsson I (1999) Copulatory behavior, genital (Coleoptera: Buprestidae). J Kans Entomol Soc 63:345–346
morphology, and male fertilization success in water striders. Eberhard WG (1993) Evaluating models of sexual selection by female
Evol Int J Org Evol 53:147–156. doi:10.2307/2640927 choice: genitalia as a test case. Am Nat 142:564–571. doi:
Arnqvist G, Rowe L (2002a) Antagonistic coevolution between the 10.1086/285556
sexes in a group of insects. Nature 415:787–789 Eberhard WG (1994) Evidence for widespread courtship during
Arnqvist G, Rowe L (2002b) Correlated evolution of male and female copulation in 131 species of insects and spiders, and implications
morphologies in water striders. Evol Int J Org Evol 56:936–947 for cryptic female choice. Evol Int J Org Evol 48:711–733.
Arnqvist G, Rowe L (2005) Sexual conflict. Princeton University doi:10.2307/2410481
Press, Princeton Eberhard WG (1996) Female control: sexual selection by cryptic
Aubertin D (1933) Revision of the genus Lucilia R.-D. (Diptera, female choice. Princeton University Press, Princeton
Calliphoridae). Linn J Zool 38:389–436 Eberhard WG (2000) Sexual behavior in the medfly, Ceratitis
Baena ML, Eberhard WG (2007) Appearances deceive: female capitata. In: Aluja M, Norrbom A (eds) Fruit flies (Tephritidae):
‘‘resistance’’ behaviour in a sepsid fly is not a test of the male’s phylogeny and evolution of behavior. CRC Press, Del Ray, pp
ability to hold on. Ethol Ecol Evol 19:27–50 457–487
Battin TJ (1993) The odonate mating system, communication, and Eberhard WG (2001a) The functional morphology of species-specific
sexual selection. Boll Zool 60:353–360 clasping structures on the front legs of male Archisepsis and
Birkhead T (1996) In it for the eggs. Nature 383:772. doi:10.1038/ Palaeosepsis flies (Diptera, Sepsidae). Zool J Linn Soc 133:335–
382772a0 368. doi:10.1111/j.1096-3642.2001.tb00630.x
Briceño RD, Eberhard WG (2009) Experimental modifications of Eberhard WG (2001b) Species-specific genitalic copulatory courtship
male genitalia support cryptic female choice hypothesis for in sepsid flies (Diptera, Sepsidae, Microsepsis). Evol Int J Org
genital evolution. Proc Natl Acad Sci USA (submitted) Evol 55:93–102
Briceño RD, Eberhard WG, Robinson AS (2007) Copulation behav- Eberhard WG (2001c) Genitalic behavior during copulation in
iour of Glossina pallidipes (Diptera: Muscidae) outside and Hybosciara gigantea (Diptera: Sciaridae) and the evolution of
inside the female, with a discussion of genitalic evolution. Bull species-specific genitalia. J Kans Entomol Soc 74:1–9
Entomol Res 97:471–488. doi:10.1017/S0007485307005214 Eberhard WG (2002) Physical restraint or stimulation? The func-
Briceño RD, Chinea-Cano E, Wegrzynek D, Eberhard WG (in tion(s) of the modified front legs of male Archisepsis
preparation) New technique opens a new field of study, genital diversiformis (Diptera, Sepsidae). J Insect Behav 15:831–850.
behavior during copulation doi:10.1023/A:1021161915227
Chapman RF (1969) Insects structure and function. English Univer- Eberhard WG (2003) Sexual behavior of male Themira minor
sity Press, London (Diptera, Sepsidae), and movements of the male’s sternal lobes
Chapman T, Arnqvist G, Bangham J, Rowe L (2003) Sexual conflict. and genitalic surstyli. Can Entomol 135:569–581
Trends Ecol Evol 18:41–47. doi:10.1016/S0169-5347(02)00004-6 Eberhard WG (2004a) Male–female conflicts and genitalia: failure to
Cordero C, Eberhard WG (2003) Female choice of sexually antagonistic confirm predictions in insects and spiders. Biol Rev Camb Philos
male adaptations: a critical review of some current research. J Evol Soc 79:121–186. doi:10.1017/S1464793103006237
Biol 16:1–6. doi:10.1046/j.1420-9101.2003.00506.x Eberhard WG (2004b) Rapid divergent evolution of sexual morphol-
Cordero C, Eberhard WG (2005) Interaction between sexually ogy: comparative tests of sexually antagonistic coevolution and
antagonistic selection and mate choice in the evolution of traditional female choice. Evol Int J Org Evol 58:1947–1970
female responses to male traits. Evol Ecol 19:111–122. Eberhard WG (2005) Sexual morphology of male Sepsis cynipsea
doi:10.1007/s10682-004-7918-2 (Diptera: Sepsidae): lack of support for sexually antagonistic
Córdoba-Aguilar A (2005) Possible coevolution of male and female coevolution and lock and key hypotheses. Can Entomol
genital form and function in a calopterygid damselfly. J Evol 137:551–565
Biol 18:132–137. doi:10.1111/j.1420-9101.2004.00796.x Eberhard WG (2009) Genitalic evolution: theory and data. In:
Crean CS, Gilburn A (1998) Sexual selection as a side-effect of Leonard J, Cordoba-Aguilar A (eds) The evolution of primary
sexual conflict in the sea-weed fly, Coelopa ursina (Diptera: sexual characters in animals. Oxford University Press, Oxford
Coelopidae). Anim Behav 56:1405–1410. doi:10.1006/ Eberhard WG, Gelhaus J (2009) Genitalic stridulation in a male
anbe.1998.0932 tupulid fly. Rev Biol Trop (in press)
Danielsson I, Askenmo C (1999) Male genital traits and mating Eberhard WG, Huber BA (2009) Spider genitalia: precise maneouvers
interval affect male fertilization success in the water strider with a numb structure in a complex lock. In: Leonard J,
Gerris lacustris. Behav Ecol Sociobiol 46:149–156. doi:10.1007/ Cordoba-Aguilar A (eds) The evolution of primary sexual
s002650050604 characters in animals. Oxford University Press, Oxford (in press)
Darwin C (1871) The descent of man and selection in relation to sex, Eberhard WG, Pereira F (1995) The process of intromission in the
6th edn. Modern Library, New York (Reprinted) medfly, Ceratitis capitata (Diptera, Tephritidae). Psyche (Stuttg)
Dixson AF (1987) Observations on the evolution of the genitalia and 102:101–122
copulatory behavior in male primates. J Zool (London) 213:423– Eberhard WG, Pereira F (1996) Functional morphology of male
443 genetic surstyli in the dungflies Archisepsis diversiformis and
Dixson AF (1998) Primate sexuality. Oxford University Press, Oxford A. ecalcarata (Diptera: Sepsidae). J Kans Entomol Soc 69:43–60
Dodson G (2000) Behavior of the Phytalmiinae and the evolution of Forbes WTM (1941) Does he stridulate? (Lepidoptera: Eupterotidae).
antlers in tephritid flies. In: Aluja M, Norrbom A (eds) Fruit flies Entomol News 52:79–82

123
Genetica (2010) 138:5–18 17

Forster W, Wohlfahrt TA (1952–1981) Die Schmetterlinge Mitteleu- Mühlhäuser C, Blanckenhorn W (2002) The costs of avoiding
ropas. Franckh’sche Velagshandlung, Stuttgart matings in the dung fly Sepsis cynipsea. Behav Ecol 13:359–365.
Ghiselin M (2009) Darwin’s view. In: Leonard J, Cordoba-Aguilar A doi:10.1093/beheco/13.3.359
(eds) The evolution of primary sexual characters in animals. Orteiza N, Linder JE, Rice WR (2005) Sexy sons from remating do
Oxford University Press, Oxford not recoup the direct costs of harmful interactions in the
Gillot C, Langley PA (1981) The control of receptivity and ovulation Drosophila melanogaster laboratory system. J Evol Biol
in the tsetse fly, Glossina morsitans. Physiol Entomol 6:269– 18:1315–1323. doi:10.1111/j.1420-9101.2005.00923.x
281. doi:10.1111/j.1365-3032.1981.tb00271.x Otronen M (1990) Mating behavior and sperm competition in the fly,
Gwynne DT, Edwards ED (1986) Ultrasound production by genital Dryomyza anilis. Behav Ecol Sociobiol 26:349–356.
stridulation in Syntonarcha iriastis (Lepidoptera: Pyralidae): doi:10.1007/BF00171101
long distance signaling by male moths? Zool J Linn Soc 88:363– Parag A, Bennett NC, Faulkes CG, Bateman PW (2006) Penile
376. doi:10.1111/j.1096-3642.1986.tb02253.x morphology of African mole rats (Bathyergidae): structureal
Hedin M (1997) Speciational history in a diverse clade of habitat- modification in relation to mode of ovulation and degree of
specialized spiders (Araneae: Nesticidae: Nesticus): inferences sociality. J Zool (London) 270:323–329. doi:10.1111/j.1469-
from geographic-based sampling. Evol Int J Org Evol 51:1929– 7998.2006.00141.x
1945. doi:10.2307/2411014 Parker GA (1972) Reproductive behavior of Sepsis cynipsea (L.)
Holland B, Rice WR (1998) Chase-away sexual selection: antago- (Diptera: Sepsidae). I. A preliminary analysis of the reproductive
nistic seduction versus resistence. Evol Int J Org Evol 52:1–7. strategy and its associated behaviour patterns. Behaviour
doi:10.2307/2410914 41:172–206. doi:10.1163/156853972X00257
Holman L, Snook RR (2006) Spermicide, cryptic female choice and Parker GA (1979) Sexual selection and sexual conflict. In: Blum MS,
the evolution of sperm form and function. J Evol Biol 19:1660– Blum N (eds) Sexual selection and reproductive competition in
1670. doi:10.1111/j.1420-9101.2006.01112.x insects. Academic Press, New York, pp 123–166
Hosken DJ, Stockley P (2004) Sexual selection and genital evolution. Parker GA (2005) Sexual conflict over mating and fertilization: an
Trends Ecol Evol 19:87–93. doi:10.1016/[Link].2003.11.012 overview. Philos Trans R Soc B 361:235–259. doi:10.1098/
Hosken DJ, Minder AM, Ward PI (2005) Male genital allometry in rstb.2005.1785
Scathophagidae (Diptera). Evol Ecol 19:501–515. doi:10.1007/ Peretti A, Eberhard WG, Briceño RD (2006) Copulatory dialogue:
s10682-005-1023-z female spiders sing during copulation to influence male genitalic
House CM, Simmons LW (2003) Genital morphology and fertiliza- movements. Anim Behav 72:413–421. doi:10.1016/
tion success in the dung beetle Onthopahgus taurus: an example [Link].2006.01.014
of sexually selected male genitalia. Proc R Soc Lond B Biol Sci Phelan PL (1997) Evolution of mate-signalling in moths: phyloge-
278:447–455. doi:10.1098/rspb.2002.2266 netic considerations and predictions from the asymmetric
Huber BA, Eberhard WG (1997) Courtship, genitalia, and genital tracking hypothesis. In: Choe J, Crespie B (eds) Mating systems
mechanics in Physocyclus globosus (Araneae, Pholcidae). Can J in insects and arachnids. Cambridge University Press, Cam-
Zool 74:905–918. doi:10.1139/z97-109 bridge, pp 240–256
Huemer P, Karsholt O, Lyneborg L (1996) Microlepidoptera of Pizarri T, Snook RR (2003) Perspective: sexual conflict and sexual
Europe. Apollo Books, Stenstrup selection: chasing away paradigm shifts. Evol Int J Org Evol
Ingram KK, Laamanen T, Puniamoorthy N, Meier R (2008) Lack of 57:1223–1236
morphological coevolution between male forelegs and female Pizarri T, Snook RR (2004) Sexual conflict and sexual selection:
wings in Themira (Sepsidae: Diptera: Insecta). Biol J Linn Soc measuring antagonistic coevolution. Evol Int J Org Evol
Lond 93:227–238 58:1389–1393
Kokko H, Brooks R, Jennions M, Morley J (2003) The evolution of Rice WR, Chippendale AK (2001) Intersexual ontogenetic conflicts. J
mate choice and mating biases. Proc R Soc Lond B Biol Sci Evol Biol 14:685–693. doi:10.1046/j.1420-9101.2001.00319.x
270:653–664. doi:10.1098/rspb.2002.2235 Richards OW (1927) The specific characters of the British humble-
Leonard J, Córdoba-Aguilar A (2009) The evolution of primary bees (Hymenoptera). Trans Roy Ent Soc Lond 75:233–265
sexual characters in animals. Oxford University Press, Oxford Richards OW (1978) The social wasps of the Americas. British
(in press) Museum (Natural History), London
Lewis CT, Pollock JN (1975) Engagement of the phallosome in Robertson HM, Paterson HEH (1982) Mate recognition and
blowflies. J Ent (A) 49:137–147 mechanical isolation in Enallagma damselflies (Odonata: Coen-
Liebherr JK (1992) Phylogeny and revision of the Platynus degallieri agrionidae). Evol Int J Org Evol 36:243–250. doi:10.2307/
species group (Coleoptera: Carabidae: Platini). Bull Am Mus 2408042
Nat Hist 214:1–115 Robinson JV, Novak KL (1997) The relationship between mating
McAlpine DK (1988) Studies in upside-down flies (Diptera: Neuro- system and penis morphology in ischnuran damselflies (Odonata:
chaetidae). Part II. Biology, adaptations, and specific mating Coenagrionidae). Biol J Linn Soc Lond 60:187–200
mechanisms. Proc Linn Soc N S W 110:59–82 Robson GC, Richards OW (1936) The variation of animals in nature.
Merrett DJ (1989) The morphology of the phallosome and accessary Longmans, Green and Co., London
gland material transfer during copulation in the blowfly, Lucilia Rodriguez V (1994) Fuentes de variación en la precedencia de
cuprina (Insecta, Diptera). Zoomorphology 109:359–366. espermatozoides de Chelymorpha alternans Boheman 1854
doi:10.1007/BF00312276 (Coleoptera: Chrysomelidae: Cassidinae). Master’s thesis, Uni-
Mikkola K (2008) The lock-and-key mechanisms of the internal versidad de Costa Rica
genitalia of the Noctuidae (Lepidoptera): how are they selected Rodriguez V, Windsor DM, Eberhard WG (2004) Tortoise beetle
for? Eur J Entomol 105:13–25 genitalia and demonstration of a sexually selected advantage for
Miller GT, Pitnick S (2003) Sperm–female coevolution in Drosophila. flagellum length in Chelymorpha alternans (Chrysomelidae,
Science 298:1230–1233. doi:10.1126/science.1076968 Cassidini, Stolaini). In: Jolivet P, Santiago-Blay JA, Schmitt M
Moore AJ, Gowaty PA, Moore PJ (2003) Females avoid manipulative (eds) New developments in the biology of Chrysomelidae. SPB
males and live longer. J Evol Biol 16:530–532 Academic Publishing, The Hague, pp 739–748

123
18 Genetica (2010) 138:5–18

Roeder KD (1967) Nerve cells and insect behavior. Harvard Stockley P, Preston BT (2004) Sperm competition and diversity in
University Press, Cambridge rodent copulatory behavior. J Evol Biol 17:1048–1057.
Roig-Alsina A (1993) The evolution of the apoid endophallus, its doi:10.1111/j.1420-9101.2004.00742.x
phylogenetic implications, and functional significance of the Tuxen L (1970) A taxonomist’s glossary of genitalia of insects. S-H
genital capsule (Hymenoptera, Apoidea). Boll Zool 60:169–183 Service Agency, Darien
Ronkainen K, Kaitala A, Huttenen R (2005) The effect of abdominal Verrell PA (1992) Primate penile morphologies and social systems:
spines on female mating frequency and fecundity in a water further evidence for an association. Fol Primat 59:114–120.
strider. J Insect Behav 18:619–631. doi:10.1007/s10905-005- doi:10.1159/000156648
7015-6 Ward PI (1983) The effects of size on the mating behaviour of the
Rowe L, Arnqvist G (2002) Sexually antagonistic coevolution in a dung fly Sepsis cynipsea. Behav Ecol Sociobiol 13:75–80.
mating system: comparative approaches to address evolutionary doi:10.1007/BF00295078
processes. Evol Int J Org Evol 56:754–767 Ware A, Opell BD (1989) A test of the mechanical isolation
Saunders DS, Dodd CHW (1972) Mating, insemination, and ovulation hypothesis in two similar spider species. J Arachnol 17:149–162
in the tsetse fly, Glossina morsitans. J Insect Physiol 18:187– Wenninger EJ, Averill AL (2006) Influence of body and genital
198. doi:10.1016/0022-1910(72)90119-9 morphology on relative male fertilization success in oriental
Sauter A, Brown MJF, Baer B, Schmid-Hempel P (2001) Males of beetle. Behav Ecol 17:656–663. doi:10.1093/beheco/ark013
social insects can prevent queens from multiple mating. Proc R Werner M, Simmons LW (2008) The evolution of male genitalia:
Soc Lond B Biol Sci 268:1449–1454. doi:10.1098/rspb. functional integration of genital sclerites in the dung beetle
2001.1680 Onthophagus taurus. Biol J Linn Soc Lond 93:257–266
Scudder G (1971) Comparative morphology of insect genitalia. Ann West-Eberhard MJ (1984) Sexual selection, social communication,
Rev Ent 16:379–406. doi:10.1146/[Link].16.010171.002115 and species-specific signals in insects. In: Lewis T (ed) Insect
Shapiro AM, Porter AH (1989) The lock-and-key hypothesis: communication. Academic Press, New York, pp 284–324
evolutionary and biosystematic interpretation of insect genitalia. Wood DM (1991) Homology and phylogenetic implications of male
Ann Rev Entomol 34:231–245. doi:10.1146/[Link].34.0101 genitalia in Diptera. The ground plan. In: Weismann, Orszagh,
89.001311 and Pont A (eds) Proceedings of the Second International
Simmons LW (2001) Sperm competition and its evolutionary Congress of Dipterology. The Hague, pp 255–284
consequences in the insects. Princeton University Press,
Princeton

123