VOLUME 73, No.4 DECEMBER 1998 THE QUARTERLY REVIEW O!BIOLOGY

EXPLOITATION OF SEXUAL SIGNALS BY PREDATORS AND PARASITOIDS

MARLENE ZUK AND GITA R. KOLLURU Department ofBiology, University ofCalifornia Riverside, California 92521 USA E-MAIL: [email protected] and [email protected]

ABSTRACT Signals used to attract mates are often conspicuous to predators and parasites, and their evolution via sexual selection is expected to be opposed by viability selection. Many secondary sexual traits may represent a compromise between attractiveness and avoidance of detection. Although such signal exploitation appears to be widespread, most examples comefrom species that use acoustic or olfactory matingsignals, and relatively few cases ofvisual signal exploitation can be substantiated. Because males are usually the signalingsex, they are more at riskfrom predators orparasitoids that locate prey or hosts by sexual signals; this differential selection on the two sexes can affect the intensity of sexual selection on male ornamental traits. The notable exception to male signaling and female attraction occurs in pheromone-producing insects, particularly lepidopterans, which show an opposite pattern offemale odor production. Exploitation ofsuch sex pheromones is relatively rare. We discuss reasons for the reversal in sex roles in these species and its implicationsfor signal exploitation. Changes in signals that appear to be adaptations to avoidpredation include the use ofdifferent signal modalities, changes in signaling behavior, loss ofsignals, and alteration ofsignal characteristics such as pitch. Selection pressure from signal exploiters could lead to the production ofa novel signal and thusfacilitate speciation. Relatively little work has been done on adaptations on thepart ofthe exploitingspecies, butsuch adaptations could indirectly influence the mating system ofthe predator or parasitoid. Signal exploitation is also expeded to be a fruitful source ofexamples of coevolution. Finally, plants emit attradants analogous to secondary sex characters in animals, and may also be vulnerable to signal exploitation.

INTRODUCTION Burk 1982; Sakaluk 1990; Verrell 1991; Endler 1992). Otte (1974) called such unintended re­ ANY SCIENTISTS have recognized that cipients "illegitimate receivers," and Dicke Mthe signals used by animals to attract and Sabelis (1992) discussed further subdivi­ mates are also conspicuous to potential preda­ sions ofsignal interception, including "spies," tors and other natural enemies (Darwin 1871; "stowaways" and "boasters." Most researchers

The Qy,arterly Review ofBiology, December 1998, Vol. 73, No.4 Copyright © 1998 by The University of Chicago. All rights reserved. 0033-5770/98/7304-0001$2.00

415 416 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73 agree that animals producing mate attraction nature of sexual signals and the possible con­ signals are faced with a conflict between mat­ straints on their evolution (Zuk 1991). ing success and survival, and many secondary In this article we attempt a comprehensive sexual traits are thought to represent a com­ review of the phenomenon of exploitation of promise between attractiveness to mates and mating signals by other species, and address avoidance of detection by enemies. This risk the following questions: has been' examined in a wide range of taxa 1. To what extent does the sensory modality using several signaling modalities, including of a signal determine its likelihood of be­ acoustic (e.g., calling crickets attracting para­ ing exploited? sitoid flies; Cade 1975),visual (e.g., coloration 2. How has selection by the exploiter shaped in guppies associated with presence of visual the evolution of the victim's sexually se­ predators; Endler 1980), and olfactory/ pher­ lected signal? omonal (e.g., use ofpheromones by egg para­ 3. Does selection act differently on the signal­ sitoids; Noldus et al. 1991a, 1991b). ing sex and the responding sex? Recent work, in particular on acoustically­ 4. Whatare the adaptationsfor exploitingand orienting parasitoids ofcalling insect hosts, has for avoiding exploitation? highlighted several issues of evolutionary sig­ The topic ofsignal evolution in the context nificance. These parasitoid flies use the song ofexploitation has many implications in addi­ of male crickets or other orthopterans to lo­ tion to those mentioned above. We will not cate a host; the female fly then deposits larvae include interesting but tangential topics, such on the cricket. The larvae burrow into the as the exploitation of host plant chemicals by cricket's bodycavity and developfor 7 to 10 days, insects; the general risks ofcopulation and mate afterwhich they emerge and pupate in the soil searching, including the attraction of rivals; (Cade 1975; Walker and Wineriter 1991; Zuk the energetic or aerodynamic costs ofsignals; etal. 1995). The auditorysystem ofone species the causes ofdifferential mortality of the sexes; of ormiine, Ormia ochracea, is closely tuned to the exploitation ofnonmating signals, such as the peak of the energy emission spectrum of aggregation pheromones, by natural enemies; the calling song ofthe host species (Robert et and the evolution of reduced conspicuous­ al. 1992, 1994), suggesting evolutionary con­ ness ofpredators to their prey. Some of these vergence between the parasitoid and its host. issues are covered in more specialized reviews, This specificity provides an opportunity for including those of Verrell (1991), Magnhagen studies not only of the convergence itself, but (1991), Sakaluk (1990), Burk (1982), and also of the potential for speciation based on Stowe et al. (1995). We consider only those variation in host signaling and on differential signals that appear to be the results of sexual attractiveness of signals to females. selection via either intrasexual competition or Despite questions raised about signal ex­ intersexual mate choice, and not simply pri­ ploitation, much of the literature on the sub- mary sexual traits used in mating. Note that ject has appeared in works that have either a our use of the word "exploitation" is distinct taxonomic or a sensory modality focus; work­ from the idea of sensory exploitation or sen­ ers on sex pheromones ofmoths, for example, sory bias (Ryan and Rand 1990), which we will and those who study visual or acoustic signals, discuss in a later section. particularly in vertebrates, rarely if ever cite one another's research. Itis therefore difficult SURVEY OF SIGNAL EXPLOITATION to determine how general the findings from Since at least the 17th century, naturalists research on a particular taxon are likely to be. have recognized that predators may be at­ Such a restriction has also hindered the devel­ tracted to the mating signals of their prey opment ofgeneral theory about the evolution (Lloyd 1966and references therein). Erasmus ofsexual signals in the contextofexploitation. Darwin, for example, described frogs that at­ Similarly, those who studysignals and their use tacked live coals they presumably mistook for by prospective mates and potential enemies flashing fireflies (cited in Lloyd 1966). More sometimes neglect the literature on sexual se­ recently, exploitation of mating signals by lection, much ofwhich is concerned with the predators and parasitoids has been reported DECEMBER 1998 EXPLOITATION OF SEXUAL SIGNALS 417

TABLE 1 Exploitation ofvictim-produced mating signals bypredators andparasitoids Signal type Exploiter Victim References Visual firefly (Photuris sp.) firefly (Photznus sp.)

firefly (Photinus collustrans) various lycosid spiders 2

trout (Salmo clarki) stickleback (Gasterosteus aculeatus) 3

prawn (Macrobrachium crenulatum), guppy (Poeczlza reticulata) 4 several predatory fishes

Sparrowhawk (Accipiter nzsus) Pied Flycatcher (Fzcedula hypoleuca) 5

Acoustic Arthropods

tachinid fly (Euphasiopteryx ochracea field crickets (Gryllus rubens. G. lineahceps. 6 =Ormza ochracea) G. integer)

tachinid fly (Euphaszopteryx depleta mole cricket (Scapterzscus spp.) 7 = Ormia depleta)

tachinid fly (Ormia ochracea) field cricket (Teleogryllus oceanicus) 8

tachinid fly (Ormza lineifrons) tettigoniid orthopteran 9 (Neoconocephalus robustus)

tachinid fly (Homotrixa sp.) tettigoniid orthopteran (Sczarasaga quadrata) 10

tachinid fly (Therobza leonzdei) tettigoniid orthopteran (Poeczlzmon spp.) 11

sarcophagid fly cicada (Okanagana rzmosa) 12 (Colcondamyza audztrzx)

tachinid fly (Euphaszopteryx ochracea mole cricket (Scapterzscus acletus) 13 = Ormia ochracea), chaoborid fly (Corethrella wzrthz)

chaoborid fly (Corethrella spp.) tree frog (Hyla avzvoca) 14

Vertebrates

gecko (Hemidactylus tursicus) cricket (Gryllodes supplzcans) 15

snapping turtle (Chelydra serpentina) bullfrog (Rana catesbeiana) 16

Little Blue Heron (Florida coerulea) short-tailed cricket (Anurogryllus celerznzctus) 17

in a numberoftaxa thatuse various signal mo­ rid flies on toads (Bufo typhonius) , G RBourne, dalities (Table 1). Many of the examples are pers. comm.; sarcophagidflies (Emblemasoma) anecdotal, based on counts of invertebrate attracted to cicada song, TJ Walker, pers. predators and parasitoids attracted to host comm.; FloridaScrubJays (Aphelocoma coerules­ pheromone-baited traps (e.g., Hardie et al. cens) foraging on singing orthopterans, JP 1991; Mendel et al. 1995). Other researchers Hailman, pers. comm.; grass snakes (Natrix na­ have noted the attraction of natural enemies trix) feeding on calling European tree frogs to various acoustically-signaling animals [pho- (Hyla arborea) , P Edenhamn, pers. comm.]. 418 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

TABLE 1 connnuatlon Exploltanon ofVlcnm-produced manng sIgnals by predators andparasltolds Signal type Exploiter Victim References (Acoustic) bats (Mlcronycterzs negalotis, tettigoniid orthopterans 18 M hlrsuta, Tonatla sylvicola, Trachops clrrhosus)

bat (Trachops clrrhosus) frogs (Hyla boulengerz, 19 Physalaemus pustulosus. Smllzsca slla)

bat (Tonana sylvlcola) tettigoniid orthopteran 20

bat (Eptesicus fuscus) tettigoniid orthopteran 21 (Pterophylla camellifolza), frog (Acrzs crepitans)

opossum (PhIlander opossum) frog (Physalaemus pustulosus) 22

cat (Felis domesncus) various orthopterans 23

Sharp-shinned Hawk Song Sparrow (Melosplza melodla) 24 (AccipIter strzatus)

Olfactory tachinid fly (Trzchopoda pennipes) pentatomid bug (Nezara Vlrzdula) 25

phaslid fly (Gymnosoma rotundatum) pentatomid bug (Plautla stalz) 26

tachinid flies (Euclyna flava, pentatomid bugs (Podisus maculzventrzs, 27 Hemyda aurata), vespid wasp P. fretus) (Vespula maculifrons)

aphelinid wasp (Encarsla pernlclosl) diaspidid scale (Quadraspldlotus pernlclosuS) 28

braconid wasps (Praon volucre, aphid (AphIS spp.) (synthetic pheromone) 29 P. a~lectum, P. dorsale)

parasitoid wasps (Aphytis afrlcanus, diaspidid scale (Aonldlella aurannl) 30 A melinus, A cohenl)

trichogrammatid wasps on eggs noctuid moth (Helzothls zea) 31 (Trzchogramma evanescens. T pretlosum)

trichogrammatid wasps on eggs noctuid moth (Mamestra brasslcae) 32 (Trzchogramma evanescens, T pretlosum)

Because human interest in controlling ag­ ary perspective to determine whether hosts can ricultural pests has fueled an enormous body escape detection (Tumlinson et al. 1993) by of research aimed at new biological control "spies" (Dicke and Sabelis 1992). For example, strategies, chemicals are continually being some hosts have evolved pheromone blends tested for their effectiveness in attracting pests. that either reduce the risk of being attacked Often these substances are kairomones, host­ (Raffa and Klepzig 1989), or represent the re­ produced chemicals thatattractenemies (Dicke sult ofpast selection by parasitoids (Aldrich et and Sabelis 1992). Some of these discoveries al. 1989). Although pheromone studies may have been further explored from an evolution- not be familiar to evolutionary biologists be- DECEMBER 1998 EXPLOITATION OF SEXUAL SIGNALS 419

TABLE 1 connnuanon Exploltanon ofvlcnm-produced manng signals bypredators andparasltolds Signal type Exploiter Victim References (Olfactory) scehonid wasp on eggs noctuld moth (Spodoptera fruglperda) 33 (Telenomus remus)

scehonld wasp on eggs lymantriid moth (Euprocns talwana) 34 (Telenomus euprochdls)

clerid beetle (Enoclerus lecontei), scolytid beetle (Ips confusus) 35 ostomId beetle (Temnochlla Vlrescens chlorodla)

clend beetle scolytid beetle (Ips typographus) 36 (Thanaslmus formlcarrus)

clerid beetle (Thanaslmus dublUS) scolytld beetle (Dendroctonus frontalrs) 37

anthocorid bug matsucoccid scale 38 (Elatophllus hebralcus) (Matsucoccus jOSephl)

vanous entomophagous and scolytid beetle (Dendroctonus frontahs) 39 parasitic msects

Olfactory/ vespid wasp (Vespula germanlca) tephritld fly (Ceranns capltata) 40 Visual several aphelinid wasps diaspldid scale (Quadraspldlotus pernlclosuS) 41 References: 1) Lloyd and Wing 1983; Lloyd 1997 2) Lloyd 1973; Daly 1978; Wing 1988. 3) Moodie 1972. 4) Endler 1978, 1980, 1982, 1983, 1991. 5) Slagsvold et al. 1995 6) Cade 1975; Walker 1983, 1986, 1993; Cade et al 1996; Wagner 1996. 7) Mangold 1978; Fowler 1987; Fowler and Garcia 1987; Parkman et al. 1996. 8) Zuk et al. 1993. 9) Burk 1982 10) Allen 1995a, 1995b 11) Heller and von Helversen 1993; Lehmann 1996; Lehmann and Heller 1997. 12) Soper et al 1976. 13) Mangold 1978. 14) McKeever 1977 15) Sakaluk and Belwood 1984. 16) Halliday 1980. 17) Bell 1979. 18) Belwood and Morris 1987 19) Tuttle and Ryan 1981, 1982; Ryan et al 1982; Tuttle et al. 1982 20) Tuttle et a1. 1985. 21) Buchler and Childs 1981. 22) Tuttle et al. 1981 23) Walker 1964. 24) PK Stoddard, pers. corom. 25) Mitchell and Mau 1971. 26) Moriya and Masakazu 1984. 27) Aldnch et al. 1984, 1986; Aldrich 1985. 28) Kypanssoudas 1987 29) Hardie et al. 1991; Powell et al 1993 30) Stemlicht 1973; Samways 1988 31) Noldus et al. 1991a, 1991b. 32) Lewis et al. 1982. 33) Nordlund et al. 1983. 34) Arakaki et al. 1996. 35) Wood et a1. 1968; Rice 1969. 36) Hansen 1983. 37) Vite and Williamson 1970 38) Mendel et a1. 1995. 39) Dixon and PaYne 1980. 40) Hendrichs et al 1994. 41) McClain et a1. 1990

cause they often appear in taxon-specific or that attract predators, and work on fireflies applied entomology sources (e.g., Aldrich (Lloyd 1966, 1973, 1997; Lloyd and Wing 1985; Kyparissoudas 1987), theyare extremely 1983), which has shown how the females of valuable in establishing the occurrence and one species respond to the courtship flashes intensity ofexploitation ofmating signals. of the male of a prey species. Although it is Visual signals are thought to be particularly unlikely that visual signals are not subject to susceptible to detection by predators (Alcock exploitation, it is probably more difficult to 1984), and many are classic examples ofsexu­ demonstrate their role in attractingpredators, ally selected ornaments. Interestingly, how­ perhaps because most visual signals are pro­ ever, exploitation ofvisual mating signals has duced continually. In addition, visual traits rarely been demonstrated (Olsson 1993). No­ such as bright colors often have functions, table exceptions include the long-term studies such as thermoregulation or territorial display of guppies (Endler 1978, 1980, 1983; Endler (Endler 1978), that are under their own selec­ and Houde 1995), which have revealed the tion pressures; these pressures may mask the specific components ofmale guppycoloration effects ofselection to avoid predation. 420 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

Acoustic mating signals can be detected at backgroundinwhich an animal signals, whereas night, are easily localized, and travel quickly conspicuous patterns must deviate from the over long distances (Alcock 1984; Sakaluk background. He then quantified the conspic­ 1990). These characteristics make transmis­ uousness ofguppycolor patterns under differ­ sion of acoustic signals easy, but also make entbackgrounds, and showed thatsignalers in mating songs, such as those produced by or­ areas of high predation intensity had better thopteran insects and frogs, detectable by a background color matching (i.e., less conspicu­ variety of invertebrate and vertebrate natural ousness) thansignalers in areas oflow predation enemies (Table 1) .Acoustic signals have been intensity. Similarvariation in conspicuousness examined both from a mechanistic stand­ ofmating signals with predation intensity was point, as in the parasitoid ear morphology found by Heller (1995) for bushcrickets. Fe­ studies by Robert et al. (1992, 1994) and male guppies also show reduced preference for Lakes-Harlan and Heller (1992), which have bright males under high predation (Endler shown that tachinidflies have evolved the nec­ and Houde 1995; Houde 1997). essary specialized morphology to detect or­ Environmental conditions can affect court­ thopteran songs, and from an evolutionary ship behavior as well. Potential victims may re­ perspective (Gwynne and Morris 1983). A duce predation risk by signaling in areas where well-known example ofthe latter approach in­ (or at times when) detection by predators is cludes work on the tungara frog (Physalaemus minimized. For example, some lekking birds pustulosus) and its acoustically-orienting bat display themselves in light environments that predator (Tuttle and Ryan 1981; Ryan 1985). maximize conspicuousness, butremain incon­ Like the guppy studies, work on the tungara spicuous at other times (Endler and Thery frog has revealed the compromise between 1996). Guppies also switch courtship tactics sexual selection and natural selection that can from displaying themselves to sneak copula­ result from mating signal exploitation, thus tions performed without courtship, depending demonstrating that the same signal compo­ on perceived predation risk (Endler 1987; nents are attractive both to potential mates Godin 1995). Furthermore, male guppies from and to unintended signal receivers. high predation localities are more likely to re­ duce courtship in the presence of a predator ENVIRONMENTAL VARIATION IN than males from low predation localities (Ma­ SIGNAL CONSPICUOUSNESS gurran and Seghers 1990), and within a local­ Understanding the evolution of signal ex­ ity, large males are more likely to reduce court­ ploitation relies on determining how conspic­ ship displays at high light intensities than small uous a mating signal is to a predator or para­ males, possibly because theyface a greater risk sitoid. Both Darwin (1871) and Endler (1978, of predation (Reynolds et al. 1993). Females 1991) emphasized that signal conspicuous­ attracted to conspicuous males can be targets ness is relative: the same mating display can ofpredation (Sakaluk and Belwood 1984; Pock­ be noticeable under certain environmental lington and Dill 1995), and mayalso alter their conditions to certain receivers, but cryptic un­ behavior under different environmental con­ der other conditions to other receivers. Light ditions. Both female crickets and tungara and turbidity levels, for example, can affect pre­ frogs respond differently to male songs, de­ dation on fish (Moodie 1972). Signal detection pending on the perceived risk of predation by predator and prey may differ enough so (Hedrick and Dill 1993; Csada and Neudorf that a signal conspicuous to a potential mate 1995; Rand et al. 1997), and female guppies is not as easily detected by a predator (Endler reverse their preference for conspicuous 1978, 1983, 1991). Ifthis is the case, then cer­ males in the presence of a predator (Godin tain aspects of a mating signal such as color and Briggs 1996; Gong and Gibson 1996). may be less susceptible to exploitation, and sex­ Signaling systems using nonvisual cues also ual selection may thus favor aspects less easily provide useful examples of conspicuousness detected by the predator (Endler 1978, 1992). that varies with environmental conditions. A Endler (1978) noted that cryptic color pat­ number ofstudies have addressed the optimal terns must resemble a random sample of the conditions for transmission ofacoustic signals DECEMBER 1998 EXPLOITATION OF SEXUAL SIGNALS 421

with respect to background noise and the acoustic, visual or olfactory-is also associated songs ofother individuals (Wiley 1991; Endler with a sex difference in signaling practice. 1992; Romer 1993; Badyaev and Leaf 1997). Acoustic and visual signals are most often pro­ Male Smilisca frogs, for example, tend to call duced by males at a given location, with fe­ from areas that have higher ambient noise lev­ males traveling to stationary groups or territo­ els generated by waterfalls, which are avoided rial individuals, as evidenced in taxa as diverse by predatory bats, possibly because the noise as crickets and katydids (Gwynne and Morris interfereswith the bats' ability to detectcalling 1983; Thornhill and Alcock 1983), lekking frogs (Tuttle and Ryan 1982). It would be in­ birds and mammals (Hoglund and Alatalo teresting to know whether other acoustically­ 1995), fruit flies (Spieth 1974), and many an­ signaling animals co-occurring with phono­ uran amphibians (Howard 1988; Sullivan 1989). tactic predators or parasitoids signal under Even when males do not signal from a fixed conditions that reduce the risk ofexploitation position, they are still the sex assumed or found (Endler 1993). to pay the price ofhaving conspicuous mating signals. Forexample, singingmale crickets use THE SIGNALING SPECIES: several times more energy than those at rest IMPLICATIONS FOR SEXUAL SELECTION (Prestwich and Walker 1981), and the meta­ Mortality sources associated with sexual sig­ bolic power output of several species of hylid naling obviously influence the evolutionary frogs is many times greater than their resting ecology of the target (Burk 1982; Sakaluk metabolic rate (Prestwich et al. 1989). Ener­ 1990). Such effects will differ for males and getic costs ofvisual signals are more difficult to females because ofthe ways in which selection determine, but studies of barn swallows have acts on the two sexes. According to classical suggested that males that are more fit are bet­ sexual selection theory, males maximize re­ ter able to produce the long tail feathers that productive success by obtaining as many mat­ attract females, which implies that the trait is ings as possible, while females are limited by costly (M011er 1994). Numerous papers in the the numberofoffspring they can produce and sexual selection literature are devoted to the rear; male variance in reproductive success is origin, measurement and consequences of likely to be much higher than that offemales costly male sexual ornaments and displays. because of the larger parental investment by females in most animal species (Trivers 1972; FEMALE PHEROMONES: Thornhill andAlcock 1983; Andersson 1994). FALLACY, FACT, OR TEST OF VIGOR? This dichotomy is often said to accountfor the In contrast to male visual and acoustic dis­ usual male role ofrisky signal production and plays, long-range olfactory sex signals-sex the usual female role of signal reception, pheromones-are usually produced by the fe­ travel to the signalingmale, andeventual mate male, although odors may be produced by choice. In sexually dimorphic species, males both sexes during courtship (Thomhill1979). are usually more brightly colored, larger, This striking reversal ofthe usual signalingsex more likely to possess specialized ornamenta­ has been met with reactions ranging from as­ tion or weaponry (such as horns and antlers), tonishment to indifference. Williams (1992: and more commonlyproduce courtship songs Ill) states, "The world is full ofmales display­ and calls (Andersson 1994). Traits that are ab­ ing tofemales with brightcolors andloudsong solutely necessary for reproduction, such as and conspicuous actions, and of females dis­ the gonads, are not usually considered to be playing to males with odors. This is strange. sexually selected characters. Or, more likely, wrong." He goes so far as to suggest that the sex-attractant pheromones SENSORY MODE AND THE SIGNALING SEX used by moths and other insects are not sexu­ The cost of producing a signal is generally ally selected signals per se, because he does not assumed to determine which sex produces it, see that females usually exhibit specialized be­ andfor the reasons summarizedin the preced­ haviors orstructures for signal emission, and be­ ing paragraph, males usually bear that cost. In­ cause males appear to have been strongly se­ terestingly, however, the type of signal used- lected to distinguish even tiny concentrations 422 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

of the odor molecules in the air. Hence, he ally thought to involve more than species recog­ argues, males are merely capitalizing on a trait nition (cf. Paterson 1985) because, in addition that happens to reveal sexual receptivity in fe­ to being a member ofthe appropriate species, males, and he terms the phenomenon of fe­ individuals preferred in sexual selection must male pheromones a "fallacy" because a true also win in sexual competition. This competi­ sex pheromone should be a signal, like the tail tion will lead to exaggerated ornaments, such of a peacock, that has itself been subject to as long tails in many birds (Andersson 1994). sexual selection. The supposed sex reversal of Contrary to the directional selection produc­ signaling therefore does notexist. Otte (1974) ing these exaggerated ornaments, stabilizing presents a somewhatmilderversion ofthis idea. selection is usually the form invoked for the Perhaps because workers on signaling in evolution of olfactory mating signals (Carde one modality or taxon tend to communicate and Baker 1984; Phelan 1997a). In any case, mainlywith those in the same area, this radical exploitation offemale pheromones is seen to position has received surprisingly little com­ be unlikely by most researchers because odor mentary in the pheromone literature (but see detection is highly specific to particular com­ Phelan 1997a, 1997b). At the same time, sev­ pounds or combinations of compounds; there­ eral authors have addressed the question of fore predators and parasitoids are less likely role reversal in olfactory signaling (Landolt to be able to "eavesdrop" on prey signals. 1997). Most ofthem have concluded that pro­ Greenfield (1981) and Boake etal. (1996) sug­ ducing pheromones is not particularly costly, gest that pheromones may even have evolved whereas responding to the odor and traveling as long-range attractants precisely because to its source involves relatively more risks they are rarely exploited. (Carde and Baker 1984; Dicke and Sabelis We do not agree with Williams (1992), but 1992; Svensson 1996; Phelan 1997b). The are still not convinced that the exploitation risky behavior is therefore taken on by males, of pheromones has been easily or completely as usual, while females are not in danger of de­ explained. Two issues arise from Williams's tection or exploitation by predators and para­ declaration that female pheromones are not sites because they emit only minute amounts of truly signals. First, he suggests that female in­ highly specific chemicals (Carde and Baker sects often lack specialized apparatus for the 1984). Indeed, the intensity oflong-range pher­ production or transmission of odors, unlike omones produced by females is dramatically male crickets, for example, which have modi­ less than that of male-produced pheromones fied wing structures used to produce and am­ (Greenfield and Coffelt 1983). Greenfield plify sound. Such a deficiency would imply (1981) concurredwith this viewpoint, andsug­ that, although males can use sex-specific odors gested that female moths are not competing to distinguish females in reproductive condi­ among themselves in the way that male pea­ tion, these odors may not have evolved as an cocks or crickets may be by signaling; byemit­ adaptation on the part of females to attract ting such low intensity signals, females might males. Closer examination ofthe literature on even be presenting a passive filter to test male pheromones suggests, however, that although response, such that only those males able to concentrations ofsex attractants are indubita­ detect minute concentrations of odors can bly small, females frequently assume particu­ find a mate. This testwould onlywork, ofcourse, lar "calling" postures when emitting phero­ if detection ability is linked to the viability of mones, and manyspecies have glands near the the male, perhaps because more sensitive ovipositor that are specialized for pheromone males can also locate host plants more easily. production (Carde and Baker 1984; Phelan Other authors have not viewed female sig­ 1997b). In addition, females may adjust the naling as a departure from a more conven­ amount of pheromone emitted, depending tional pattern, and simply assume that because on how much sperm they have received (Mc­ females invest more in individual offspring, Neil et al. 1997). It therefore seems plausible their signals must be canalized mechanisms that selection has acted on females to produce for ensuring species recognition (Carde and appropriately alluring signals, although the Baker 1984). Sexual selection, however, is usu- lack of exaggeration of those signals, unlike DECEMBER 1998 EXPLOITATION OF SEXUAL SIGNALS 423

those in males, remains intriguing. What would attack adult lepidopterans; most such parasit­ constitute an elaborated scent? Our own rela­ oids attack eggs or larvae. It may be economi­ tive insensitivity to olfactory cues may hinder cally unwise for a parasitoid to locate a phero­ our ability to imagine the odor equivalent of mone-emitting female, only to wait until she a bird ofparadise's plumage. Certainly little is mates and lays eggs. Thus, except for a few known about the ancestral state ofpheromones special cases such as the attraction ofegg para­ amongthose insects thatproduce them, making sitoids to the noctuid moths, Heliothis zea and comparative studies even more problematic. Mamestra brassicae, at the time of oviposition The second issue is whether a role reversal (Table 2; Noldus et al. 1991a, 1991b), most in signaling occurs in animals using long­ sex pheromones that attract natural enemies range pheromones to attract mates. If the are male-produced pheromones in aggregat­ pheromone is not costly to produce, and if ing species such as bark beetles (Table 1; males compete with one another by searching Wood et al. 1968; Hansen 1983). This argu­ and incur costs as they travel to the female, ment does not explain, however, why preda­ then Greenfield's (1981) idea about using tors ofadult insects are thought to be unlikely pheromones as a filter to test males is appeal­ to use odor cues. ing because males are still performing the Why do so few examples of odor detection costly part ofmating. The problem is that vir­ by predators exist, compared to the detection tually no dc:)-la on the energetic costs ofphero­ ofother sensory cues? We suggest that the dif­ mone production are available (Dicke and ference may lie in a distinction drawn by May­ Sabelis 1992). Interestingly, female moths nard Smith (1958, 1991) between "notices" sometimes produce greater concentrations of and "advertisements." If the interests of sig­ pheromone as they age (Greenfield 1981); naler and receiver do not coincide, the evolu­ this is consistent with the idea that odor pro­ tion of a costly advertisement is likely. Most duction is expensive since females should be interactions between males and females fall more willing to pay costs as their reproductive into this category because of the disparity of value decreases and less of their reproductive parental investment between the sexes and lifespan remains (Williams 1975). Ifincreased subsequent male competition for females signal intensity is not costly, but attracts males (Trivers 1972). However, some signals, such more effectively, why has selection not in­ as the railway timetable or bee waggle dance creased signal intensity at all ages? Lundberg (Maynard Smith 1991), are not selected to be and Lofstedt (1987) discussed variation in costly because both sender and receiver bene­ pheromone production in the context of in­ fit from accurate transfer of information. If traspecific competition, and suggested that female moths do not compete among them­ ecological constraints control emission rate. selves for males, and ifodors are not energeti­ Phelan (1997a) emphasized the importance cally costly to produce, sex pheromones may of stabilizing selection in the evolution of qualify as notices, and hence not be conspicu­ pheromone signaling, but it seems to us that ous to natural enemies in the way, for exam­ as long as male responses are linked to female ple, that cricketsongis. It is interesting to note signals, directional selection and subsequent that in the few cases oflong-range sex attract­ exaggeration of the odor ought to be at least ants in noninsects, including mammals such as likely. Information about the costs ofmanu­ as dogs, females are again the sex that pro­ facturing and releasing pheromones is sorely duces the odor (Thornhill 1979). Whether needed. predators are attracted to such mammalian­ The other cost, besides an energetic one, is produced scents and whether these odors are the subjectofthis article: exploitation by pred­ similarly less costly to produce remains to be ators or parasites. If odors are not likely to at­ seen. tract natural enemies, then females take no Finally, it has been suggested that predators risks by producing them. Greenfield (1981) might simply find it more difficult to exploit suggested that the apparent rarity with which odors because of the precise composition of parasitoids locate female-emitted pheromones most pheromones. We are skeptical ofthis ex­ may reflect the rarity with which parasitoids planation, given the remarkable adaptations 424 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

TABLE 2 Exploitation offemale-producedpheromone signals bypredators andparasitoids Victim Exploiter Pheromone type References aphid brachonid wasps (Praon volucre, synthetic (Aphis spp.) P. abjectum, P. dorsale)

noctuid moths trichogrammatid wasps on eggs natural 2 (Heliothis zea, (Trichogramma evanescens, Mamestra brassicae) T. pretiosum)

lymantriid moth scelionid wasp on eggs natural and synthetic 3 (Euproctis taiwana) (Telenomus euproctidis)

noctuid moth scelionid wasp on eggs natural 4 (Spodopterafrugiperda) (Telenomus remus)

diaspidid scale aphelinid wasps (Aphytis africanus, natural and synthetic 5 (Aonidiella aurantii) A. melinus, A. coheni)

matsucoccid scale anthocorid bug synthetic 6 (Matsucoccusjosephi) (Elatophilus hebraicus)

diaspidid scale aphelinid wasp (Encarsia perniciosi) synthetic 7 (Quadraspidiotus perniciosus) References: 1) Hardie et al. 1991; Powell et al. 1993. 2) Noldus et al. 1991 a, 1991b. 3) Arakaki et al. 1996. 4) Nordlund et al. 1983. 5) Sternlicht 1973; Samways 1988. 6) Mendel et al. 1995. 7) Kyparissoudas 1987; McClain et al. 1990.

seen in othernatural enemieswith special sen­ VICTIM ADAPTATIONS sitivities to acoustic frequencies or otherprop­ The mating signals of numerous species erties ofsignals (Robertetal. 1994). Sanderford have been exploited by natural enemies, as de­ and Conner (1995) suggested that the acous­ tailed in Table 1, and here we consider the tic courtship signals given by both males and avenues of escape taken by the victim. Later females in the moth Syntomeida epilais are pos­ we also examine the interaction from the ex­ sible because ofa release from predation pres­ ploiter's pointofview. Table 3 contains a sum­ sure by bats; whether other species are under maryofsignal characteristics thathave beensug­ similarconstraints is as yetunknown. The liter­ gested to be adaptations for avoiding detection ature yielded numerous examples of female­ by parasites or predators. Forest-dwelling katy­ produced pheromones that are exploited by dids (Tettigoniidae) subject to predation by parasitoids and predators (Table 2), most of foliage-gleaning bats show reduced calling ac­ which were discovered in the last few years. tivity and unusually high ultrasonic carrierfre­ More research may lead to the abandonment quencies, and utilize substrate vibration instead ofthe idea that olfactory cues are inconspicu­ ofairborne calling (Morris 1980; Belwoodand ous to unintended receivers. We agree with Morris 1987; Belwood 1990; Morris etal. 1994; the conventional view that pheromones are Heller 1995). Similarly, members of 11 insect signals, butwe also agree with Williams (1992) orders produce substrate vibration in lieu of that the phenomenon of female-produced airborne calling songs; such "silent singing" is odors is strange. We hope thatfuture workwill particularly noteworthy in lacewings (Henry address the questions ofcosts ofpheromones 1994). Other forms of antipredator behavior and the sensory capabilities of predators that in acoustically-signaling insects are discussed utilize them. by Bailey (1991). Surprisingly, incidences of DECEMBER 1998 EXPLOITATION OF SEXUAL SIGNALS 425

visual cues modified as a result of predation to the evolution of alternative reproductive be­ pressure are harder to document, although havior, which may involve less risky, "sneaky" the cryptic plumage offemales in many sexu­ means of gaining fertilizations, as shown by ally dimorphic birds, for example, is often at­ some males within a population (Gadgil1972; tributed to selection by predators against the Austad 1984; Andersson 1994). The alterna­ conspicuous coloration of males (Promislow tives may yield the same reproductive success, et al. 1994). The ancestral state of plumage in which case they may be genetically prede­ coloration in sexually dichromatic species is termined; or theymaynot, in which case males generally supposed to have been dull or cryp­ with inferior developmental histories may be tic rather than bright, since the usual pattern "making the best of a badjob." For example, is for males to be more colorful (Butcher and female-mimicking males ofbluegill sunfish are a Rohwer 1989). Although little evidence is avail­ smaller, morphologically distinct class that does able on this point, a striking exception is the notdefend territories (Dominey 1980). Instead, work byEndler (1983, 1991, 1992),who demon­ such males wait until a female is about to de­ strated that male guppies in predator-rich en­ posit eggs onto the nest of a territorial male, vironments have duller orange patches than then swim quickly into the territory, release males in streams relatively free from predators. sperm, and leave. Adoption ofa female-mimic An interesting and little considered aspect or territorial male strategy appears to be rela­ ofescape from predation or parasitism by sig­ tivelyfixed (Dominey 1980). Cade (1975,1980) naling animals is the possibility that such an found some male field crickets (Gryllus integer) escape may include the production of novel thatdid notcall butwere seen nearcallers; this signals, which could instigate or facilitate spe­ species is subject to an acoustically-orienting ciation via sexual selection on the new signal parasitoid fly (Ormia ochracea) , which primar­ (Lande 1981; West-Eberhard 1983; Verrell ily attacks calling males. Cade (1975) called 1991). Iffemale preference for a signal is cor­ such silent males "satellites" and suggested related with genes for signal production, as that they were intercepting females as they manymodelsofsexual selection suggest (Kirk­ moved toward callers, thus avoiding parasitiza­ patrick and Ryan 1991), females may "follow" tion. Similar satellite males were observed near males as they evade detection by natural ene­ male moths (Syntonarcha iriastis) producing mies. Alternatively, if a risky signal contains ultrasound by genital stridulation (Gwynne elements favored by females, either because and Edwards 1986), and females sometimes they exploit her sensory systems (Christy mated with silent male wax moths (Achroia gri­ 1995) or because the risk constitutes a test of sella) found near ultrasound-producing males male fitness, such rapid isolation of popula­ (Greenfield and Coffelt 1983). tions is less likely. In the tungara frog Physalae­ Evasion ofpredation on the more conspicu­ mus pustulosus, females prefer the portion of ous signaling males is often thought to be a the song most easily detected by predatory benefit of adopting such an alternative strat­ bats (Ryan etal. 1982; Ryan 1985), but the gen­ egy, but unless such an advantage is demon­ erality of this finding remains unexplored. strated, it may be unwarranted to assume that More work on the role of female preference it is. For example, the parrotfish Sparisoma ra­ in shaping the opposing selection forces on dians has both conspicuous territorial and sexual signals is needed. cryptic schooling males within a population In addition to showing altered signals, sev­ (Clifton and Robertson 1993). Although one eral taxa modify their behavior in response to might assume that the cryptic males enjoy a predation on the signaling sex (Table 3; Burk more risk-free existence, examination of the 1982). Displaysites, spacing patterns, and tem­ stomach contents ofthe major predatorofthis poral shifts in signaling of both birds and in­ species, the yellow jack Caranx bartholomaei, sects all may reflect selection by predators or showed that both male morphs were eaten in parasites (Burk 1982; Lloyd and Wing 1983; proportion to their availability, with a shift Trail 1987; Sakaluk 1990; Endler and Thery over the course of the day to selective preda­ 1996). In its most extreme form, such behav­ tion on spawning males, whether gaudy or ioral differences among individuals have led cryptic (Clifton and Robertson 1993). Simi- TABLE 3 TYICtl111 adaptations that counter exploitation ofl1latlng signals Signal type Victim Victim adaptation References Visual spider (Dolol1ledes triton) surface-,vave mating signal frequency characteristics like that of nonprey rather than prey

firefly (Photl nus spp.) evolution offlashing signal instead ofconstant glo,v: paucity of sedentary aggregations in the U.S. 2 (,vhere predator occurs): delayed signaling activity until sunlight is reduced during sunuuer (,vhen predator is active)

firefly (Pyractol11ena sp.) nlale drops to the ground after female flash response instead offlashing again 3 ~ poeciliid fishes evolution of decrease in number and size of sexually selected color patches in populations ,vith high 4 tt-j (guppy, Poecllza retlculata: predation intensity: more frequent displays at lo,v light intensities and use ofalternative nlating tactics at ;a Phalloceros caudlnlaculatus) high light intensities: large nlales nlore likely to reduce courtship displays at high light intensities than §2 small males, possibly because they face a greater risk ofpredation: evolution of reduced courtship display ~ in populations ,vith high predation ~ >:3 ~ Guianan Cock-of-the-Rock nlating displays perfornled in groups (leks) to reduce raptor predation 5 ~ (Ruplcola niplcola) ~ ;S lekking birds luinimized conspicuousness ,vhen not displaying: luales are either more chromatic or brighter than 6 tt-j ~ (Ruplcola rupicola, background, but never both Coraplpo gutturalzs, ~ Lepidothrix serena) ~ a~ ~ Acoustic Arthropods a CJ ~ tettigoniid orthopteran reduced proportion oftime spent calling in presence of predator: reduced airborne calling in favor of 7 (Copiphora rlllnoceros) substrate vibration in presence ofpredator

tettigoniid orthopteran presunset calling: frequent movenlent. lack ofassociation ,vith anyone plant species 8 (Sclarasaga quadrata) tettigoniid orthopteran cessation ofultrasonic luating calls ,vhen predator is detected 9 B (lnsara covrlleae) t-t C ~ trl field cricket (Gr.vllus Integer) sotue luales renlain silent and opportunistically tuate "'ith feluales attracted to other tuales' songs 10 '-l VJ U field cricket (Gryllus rubens) phase shifting ofpulses in calling song to reduce detection by parasitoid 11 trl Cj trl ~ field cricket beginning and ending singing more abruptly in parasitized populations: singing ITIOre slo\vly In parasitized 12 to (Teleogryllus oceanlcus) populations trl ~ J--l ~ field cricket (Gryllus Integer) delays calling until sunrise and preceding hours when parasitoid is not active 13 ~ 00 mole cricket changes in time ofyear that signal is produced 14 (Scapteriscus spp.)

snowy tree cricket singing in choruses so each individual is less likely to be detected by predator or parasitoid 15 (Oecanthus fultonl) ~ ~ lesser wax moth ceases producing ultrasonic mating calls and switches to pheromone calling when predator is detected 16 a (ArchrOla grisella) ~ S2 eleven insect orders production ofsubstrate vibrations instead ofairborne songs 17 ~ a ~ r~ertebrates ~ frogs ceases calling when predator approaches: dives down into water ifpredator is very close: 18 ~ (Physalaemus pustulosus, synchronized calling ~ Smilisca sila) §2 ~ Olfactory Pentatomid bug selective release ofpheromone during daylight hours: "silenC strategy adopted by some males: males 19 ~ (Podisus nlaculzventrzs) do not signal, but instead mate with females attracted to other males' pheronlone: seasonal decline in CJ attraction ofbugs to pheromone traps when parasitoid becomes active ~ ~ VJ Pentatomid bug evolution of shorter preoviposition period and longer larval period in parasitized populations 20 (Nezara vlrzdula) (may dampen effects ofparasitoid)

scolytid beetle (Ips plni) ability to produce and respond to a variety ofpheromone blends: may enable "escape" from exploitation 21

scolytid beetle (Ips plni) parasitoid more attracted to local than distant prey populations, possibly due to evolution ofchemical 22 differences in prey pheromone References: 1) Bleckmann and Bender 1987. 2) Lloyd 1983. 3) Lloyd 1966. 4) Farr 1975:,Endler 1978, 1980, 1982. 1983, 1987, 1991: Reynolds et a1. 1993. 5) Trail 1987. 6) Endler and Thery 1996. 7) Morris 1980: Belwood and Morris 1987. 8) Allen 1995b. 9) Spangler 1984. 10) Cade 1975. II) Walker 1993. 12) Zuk et a1. 1993: Rotenberry et a1. 1996. 13) Cade et a1. 1996 14) Mangold 1978 15) Walker 1969 16) Spangler 1984. 17) Henry 1994. 18) Tuttle et a1. 1982: Tuttle and Ryan 1982 19) Aldrich et a1. 1984: Aldrich 1985: Aldrich 1995. 20) Aldrich et a1. 1989. 21) Raffa and Klepzig 1989. 22) Raffa and Dahlsten 1995 428 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

larly, the field cricket Teleogryllus oceanicus is dinner), obviously applies here. We expect also subject to parasitization by the same pho­ stronger adaptations for avoiding predators notactic parasitoid fly that attacks Gryllus inte­ than for overcoming these avoidance mecha­ ger, and silent males are common in parasit­ nisms. Nevertheless, examining the interac­ ized populations; these silent males, however, tion from the exploiter's viewpoint is also are actually more likely to harbor parasitoid worthwhile, especially when parasitoids spe­ larvae than are calling males (Zuk etal. 1995). cialize on a single host species. Table 4 lists Perhapsbecause ofthe relatively recentassoci­ exploiter adaptations in a variety ofsystems. ation between Ormia ochracea and T. oceanicus, The life/dinner principle becomes less ap­ parasitized males may not have evolved de­ plicable as the predator becomes more spe­ fenses thatwould allow them to continue call­ cialized, and especially when parasitoids are ing despite the presence ofthe parasitoids (Zuk involved (Thompson 1994). The disparity be­ et al. 1993; Rotenberry et al. 1996). Instead, tween the costs to each side oflosing an evolu­ males in populations where the flies are pres­ tionary arms race is lessened in these situa­ ent show shifts in the time ofday when calling tions, because highly specialized predators starts and stops and in the structure of the and mostparasitoids mustfind an appropriate song, comparedwith unparasitized populations preyor host; ifone chance is lost, another may of the same species (Zuk et al. 1993; Roten­ not arise. As an example, contrast the feeding berry et al. 1996). ofbats on tungara frogs with the use ofcalling Finally, although most studies ofbright col­ crickets as hosts by ormiine flies. Ifa bat does oration and other sexual signals assume that not detect a frog, it can eat other prey, orfind these evolved via sexual selection, the unprof­ one by using another means. A gravid female itable prey hypothesis (Baker and Parker 1979) ormiine, however, must locate a calling male maintains that conspicuous colors actually serve cricket from one or a few appropriate species to indicate unpalatabilityorawareness ofa pred­ in order to reproduce at all, and her window ator, and hence are not a risk at all (Lloyd of time for doing so is probably quite narrow. 1966; Baker and Parker 1979; Gotmark 1994; Itcomes as no surprise, therefore, that the flies Gotmarkand Unger 1994). Andersson (1994) possess a highly unusual tympanal hearing ap­ provided a discussion of the recent literature paratus. As mentioned earlier, the auditory on this topic, and concluded that while a few system of female Ormia ochracea is closely dichromatic species mayshowaposematic col­ tuned to the calling song of the host genus oration, and a few others may experience more (Robert et al. 1992,1994), suggesting conver­ predation on the less conspicuous sex in ac­ gent coevolution between the parasitoid and cordance with the unprofitable prey hypothe­ its hosts. Indeed, Robert et al. (1992) pointed sis, this notion is not likely to be a general ex­ out that the female O. ochracea must be able to planation for the evolution ofshowy male traits. do exactlywhat a female cricket does, namely, find calling male crickets. To our knowledge, THE DETECTING SPECIES: no such coevolved structure exists in the hear­ EXPLOITER ADAPTATIONS ingapparatus ofbats thatfeed on tungara frogs. The degree ofspecialization on a particular Having an ear that is similar to a cricket's is host or prey type will constrain the sensorysys­ obviously helpful for finding a host. But what tem ofthe parasite orpredator, as well as influ­ about the need to find a mate, which is of ence the signaling of its prey. Earlier reviews course not a cricket but another fly? How has on the exploitation of sexual signals have selection for prey detection and exploitation mainly focused on victim adaptations, but as of mating signals constrained the signaling with other predator-prey interactions, both abilities, notofthe exploitedspecies butofthe sides of the relationship are expected to be exploiter? We know of no studies along these affected. The "life/dinner principle" (Daw­ lines, but it seems at least plausible that the kins and Krebs 1979), which states that the mating system of the predator or parasitoid consequences of being eaten (losing one's could be affected by the need to be sensitive life) are more important than the conse­ to the visual, auditory or olfactory range emit­ quences of missing a prey item (losing one's ted by the prey species, as well as to signals TABLE~ Predator andparasitold adaptatlonsfor exploiting victinl fllating signals Signal type Exploiter Exploiter adaptation References Acoustic sarcophagid fly (Co/condanlyla auditnx) prey cicadas muted after parasitization to prevent superparasitism

tachinid fly (Ornlia ochracea) tympanal ear allo,v5 hearing high (4-5 kHz) frequencies 2

tachinid fly (Euphasiapteryx dep/eta = activity period corresponds to victim calling periods 3 Ormia dep/eta)

bat (Eptesicusfuscus) ability to hear lo,v frequency, long-range sounds offrog and insect choruses 4

bat (Trachops cirrhosus) ability to discern suitable prey frogs by their songs 5

bats (Trachops cirrhosus~ Tonatia sy/vicola) resource partitioning by exploiting calls ofeither orthopteran insects or frogs 6

Olfactory trichogrammatid ,vasps on eggs arrestment offlight in presence ofprey pheromone (more advantageous than flying 7 (TrichogranlJJlQ evanescens, T. pretiosu111) toward pheromone because pheromone is not exactly ,,,here eggs are): preferential searching for prey eggs on underside of leaves~ ,vhere they are deposited

brachonid wasps (Praol1 volucre, ability to recognize sexual female aphids, ,vhich Inay be the last chance for 8 P. abjectUJ11, P. dorsa/e) parasitoid to find suitable host for overwintering

clerid beetle (Thanasimu$' dubius) ability to recognize a variety ofprey pheromone blends (expressed as high local 9 variation in response to blends)

clerid beetle (Thanasimus fOr1nicanus) antennal olfactory receptors as sensitive as the prey receptors to prey pheromone 10

aphelinid wasp (Encarsla perniciosi) start ofseasonal flight coincident ,vith that ofprey Inales 11

Olfactory/Visual vespid ,vasp (f "espula gennanica) switch fronl olfactory detection in luorning ,vhen victitn lekking peaks~ to visual 12 detection later in day ,vhen victilu felnales are ovipositing References: 1) Soper et a1. 1976. 2) Lakes-Harlan and Heller 1992: Robert et a1. 1992,1994: Edgecolub et a1. 1995. 3) Fo,vler 1987. ~) Buchler and Childs 1981. 5) Tuttle and Ryan 1981. 6) Tuttle et a1. 1985. 7) Noldus et a1. 1991a, 1991b. 8) Hardie et al. ]991. 9) Hernls et a1. ]991. 10) Hansen 1983. 11) Kyparissoudas 1987. 12) Hendrichs et al. 199~. 430 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

produced by the opposite sex. On the other shown that costly male ornaments will evolve hand, predators may use exploitation to locate via the handicap principle even if they pose potential mates that are also attracted to the a cost to the choosing female (e.g., if she is victim's mating signal (Vite and Williamson attacked by the exploiter responding to the 1970; Dixon and Payne 1980). The situation is male's signal), provided the handicap is "re­ complicated by our lack ofinformation about vealing" or "condition-dependent" (Grafen the natural lives of many parasitoids at times 1990; Maynard Smith 1991;Johnstone 1995). other than host location and larval deposition In otherwords, ifmales can produce the hand­ (Godfray 1994) . Presumably a predator or para­ icap only ifthey are fit, then the handicap trait sitoid cannot be finely tuned to two different will evolve. frequency curves, and so one might expect that it either uses the same sensory window COEVOLUTION as its prey, or switches to a different sensory Ever since itwas first emphasized by Ehrlich modality entirely for locating mates (e.g., ol­ and Raven (1964), coevolution has been con­ faction if it locates prey acoustically, or vice troversial, largely owing to the lack ofa consis­ versa). This idea is speculative, but worth ex­ tent definition (Janzen 1980). Every mutu­ ploring. alism or predator-prey association that involves adaptations is not an example of coevolution SIGNAL HONESTY (Janzen 1980; Schemske 1983). An adequate We have discussed the inadvertent attraction demonstration of coevolution requires evi­ ofpredators or parasitoids as a cost ofproduc­ dence that the traits in question have evolved ing conspicuous mating signals (Magnhagen specifically to aid in the interaction described, 1991). Why would signals that are potentially and are not the products of past evolution fatal to the signaler evolve? This question may (Janzen 1980). Even this restrictive definition be asked ofany costly mating signal, such as a has yielded several convincing exampIes ofre­ long tail that reduces male flying ability (Evans ciprocal evolution (Thompson 1994 and ref­ and Thomas 1992). One convincing answer is erences therein). thatsuch traits serve as indicators ofthe signal­ Signaling systems that evolve under selec­ er's quality to the receiver (Zahavi 1975). If tion pressure imposed by exploiters may yield the expense of the signal ensures that it can other examples of coevolution. Signaling ani­ only be produced by high-quality males, then mals are under selection to avoid detection by it is an "honest" indicator ofthe signaler's qual­ illegitimate receivers, and the predators and ity; signal honesty is therefore advanced as a parasitoids are in turn evolving better ways to necessary condition for "good genes" models eavesdrop on the signals of their victims; the of sexual selection (Andersson 1994; J ohn­ result is a coevolutionary arms race between stone 1995). victim and exploiter (Burk 1988). Both preda­ As noted already, sexual ornaments are "ad­ tors and parasitoids may evolve specializations vertisements," not "notices" (Maynard Smith that detect signals, but because parasitoids are 1991); the production of advertisements in­ forced to live a large part, or all, of the life volves a conflict of interest between the sig­ cycle on a single host individual, they are ex­ naler and the receiver that is not present in pected to evolve even more highly specialized the production of notices. Advertisements thus abilities to detect suitable, high-quality victims involve a cost to the signaler because the cost (Thompson 1994). Indeed, manyofthe exam­ maintains signal reliability and prevents low­ ples ofexploiter adaptations in Table 4 involve quality males from cheating by displaying the parasitoid insects. ornament without having the accompanying Pheromones are often produced in ex­ high fitness. The handicap principle (Zahavi tremely small quantities, possibly as a mecha­ 1975) states thatexaggerated male ornaments nism for preventing exploitation (Greenfield will evolve via sexual selection because they 1981; Boake et al. 1996). As a consequence, indicate a male's ability to breed despite being organisms that use host pheromones as kairo­ burdened with a trait (the handicap) that mones must evolve specialized mechanisms to threatens survival. Theoretical models have detect such minute quantities of chemicals. DECEMBER 1998 EXPLOITATION OF SEXUAL SIGNALS 431

Forexample, Hansen (1983) provided anatomi­ tion may have constrained the evolution ofvi­ cal and electrophysiological evidence suggest­ sual signals in particular, and more than is ing that the antennae ofpredatory clerid bee­ commonly assumed, but because predation is tles have developed sensitivity to their prey's rarely observed in nature, this has been diffi­ pheromones that is equivalent to the prey's cult to document. Taxa that have been ne­ own antennal sensitivity. glected in this regard include acoustically-sig­ The most striking example of host-detection naling fish, which are conspicuous in their morphology currently known, however, is the own environment but have been little studied ears oftachinid flies which phonotacticallyori­ (Bass 1992). Seeking examples of signal ex­ ent to their singing orthopteran hosts (Lakes­ ploitation in new situations may help resolve Harlan and Heller 1992; Robert et al. 1992, some of the controversies and test hypotheses 1994). These flies, particularly Ormia ochracea, about its evolution. For example, if phero­ have evolved tympanal ears that are conver­ mone-producing insects have evolved signals gent with the ears of their hosts and that are in a very narrow "frequency band" because of unlike the auditory organs ofother closely re­ selection pressure from predators, then pher­ lated flies (Edgecomb et al. 1995). Orthopter­ omone-producinganimals thatare notsubject ans in turn may have evolved mechanisms to reduce parasitization, such as restricting their to such predation should have more general­ singing period to times ofdarkness and reduc­ ized signals. Carnivores at the top of the food ing various temporal song components (Zuk chain, such as tigers, might be interesting sub- et al. 1993; Allen 1995b). Before the associa­ jects for studies in this regard (Brahmachary tion between orrniine flies and their hosts can et al. 1992), and researchers should look at a be called coevolution in the strict sense, how­ diversity of taxa within particular signaling ever, the heritabilities of the traits involved modes. and the histories of the associations must be Studies on coevolution should also look elucidated (Schemske 1983). toward signal exploitation for new sources of Many species have evolved adaptations to examples. Much of the current literature on counter exploitation by natural enemies (Table coevolution relies on plant-pollinator relation­ 3); however, most of these traits, such as re­ ships, but signal exploitation should yield many maining silent in the presence of a predator, other potential cases of reciprocal changes in minimize conspicuousness in general but are signal production and detection. As discussed, not specializations against specific predators the high degree ofspecialization found in many or parasitoids. Probably other examples of natural enemies of signaling species opens the specific traits that reduce the risk of exploita­ way for coadaptations. Studies ofexploitation tion will be found in acoustic and olfactory of sexual signals may provide tests for some signaling systems, because these signals are of the currently intractable hypotheses about produced discreetly and often require special­ patterns of coevolution (Thompson 1994). ized structuresfor detection. Visual signals, on The role ofsignal exploitation in speciation the other hand, are conspicuous to a variety was discussed by Verrell (1991), who pointed oforganisms, and may not require specialized out that arms races between signalers and natu­ detection organs beyond what most species have evolved in order to see conspecifics. This ral enemies can lead to rapid divergence ofpop­ generalization should be treated with caution, ulations in both taxa. Ifpopulations ofsignalers however, because it may reflect human bias are subject to different exploiters, evasion of towards visual orientation; animals vary, for the predator or parasitoid might generate iso­ example, in their abilities to perceive certain lation from other populations of signalers as wavelengths of light (Endler 1983), and so the signal changes (Verrell 1991). Although specialization may be equally possible in visu­ several authors have suggested that sexual se­ ally signaling systems. lection can drive rapid speciation in certain groups, such as the Hawaiian drosophilids CONCLUSIONS (Kaneshiro and Boake 1987) or in theoretical Signal exploitation is widespread among models (Lande 1981; West-Eberhard 1983), animals. It occurs in many taxa and uses vari­ less attention has been paid to signal exploita­ ous signaling modes. We suspect that preda- tion as a part of the sexual selection process. 432 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

Classical biological control, involving the acter is subject to detection and exploitation use of native parasitoids to control pest spe­ by a natural enemy, variation in its characteris­ cies, has long been appealing because it does tics may arise independently of geography. not involve pesticides and because the para­ Conversely, pressure from the exploiter may sitoids are often host-specific (Pimentel 1963; exaggerate existing variation if the exploiters Nechols and Kauffman 1992). However, more are present in some areas and not others, as recently researchers have argued that natural is the case for the phonotactic parasitoid fly enemies may not be as effective in biological Ormia ochracea that uses the cricket Teleogryllus control as novel ones, because pests may have oceanicus as a host (Zuk et al. 1993). Exploita­ evolved adaptations to avoid enemies with tion of signals will also influence their costli­ which they have co-occurred (Pimentel 1963; ness, and hence their reliability and usefulness Hokkanen and Pimentel 1984). Because para­ as honest indicators. sitoids often exploit mating signals to locate Although our review focused on animals as hosts, researchers interested in determining both signaling and exploiting species, there is which parasitoid to use in biological control no a priori reason why plants should not emit must understand the degree to which native signals that might be used by exploiters, such hosts have evolved adaptations to avoid exploi­ as nectar robbers orherbivores, that capitalize tation. For example, some predators are more on the need to attract pollinators and seed dis­ highly attracted to the pheromones of novel persers. Sexual selection in plants is now hosts than native ones (Aldrich 1995; Raffa widely acknowledged (Willson and Burley 1983; Andersson 1994; Grant 1995), and thus and Dahlsten 1995), andvariation amonghost perhaps the time has come to recognize the populations has been suggested to be the re­ potential for further study ofsexually-selected sult of selection pressure imposed by eaves­ signals in these organisms. Regardless of dropping enemies (Aldrich et al. 1989; Raffa whether authors agree on the definitions of and Klepzig 1989); however, in other cases the sexual competition and secondary sexual char­ natural enemies are more attracted to native acters in plants (Grant 1995), conspicuous hosts (Raffa and Dahlsten 1995). Careful stud­ visual and odor attractants are widespread ies of which parasitoids and predators occur among them, and should be examined for un­ with which hosts (Hokkanen 1986), how re­ wanted visitors. Exploitation of sexual signals cent the associations are, and how signal ex­ is a unifying force in sexual selection that we ploitation has evolved are necessary to estab­ hope will receive even more attention and syn­ lish effective control programs. thesis from biologists in many disciplines. Finally, signal exploitation has implications for the study ofsexual selection itself. For ex­ ACKNOWLEDGMENTS ample, geographic variation in secondary sex­ We are grateful to T Burk, RT Carde, and MD ual characters has received considerable at­ Greenfield for helpful discussion, to C Sassaman for tention in the literature (Endler 1983; Zuk et reading an earlierversion ofthe manuscript, to KA McKean for help with references, and to all who al. 1993; Endler and Houde 1995; Heller responded to our ABSnet posting asking for exam­ 1995). This variation is of interest partly be­ ples of signal exploitation. MZ is supported by cause it may contribute to speciation, as de­ grants from the National Science Foundation and scribed already. If the secondary sexual char- the University ofCalifornia, Riverside.

REFERENCES

Alcock]. 1984. AnimalBehavior: AnEvolutionary Per­ Aldrich] R. 1995. Testing the "new associations" spective. Third Edition. Sunderland (MA): Si­ biological control concept with a tachinid nauer. parasitoid. Journal of Chemical Ecology 21:1031­ Aldrich] R. 1985. Pheromone of a true bug (He­ 1042. miptera-Heteroptera): attractant for the preda­ Aldrich] R, Kochansky] P, Abrams C B. 1984. At­ tor, Podisus maculiventris, and kairomonal ef­ tractant for a beneficial insect and its parasit­ fects. In Semiochemistry: Flavors and Pheromones, oids: pheromone ofthe predatoryspined soldier edited by TE Acree and DM Soderlund, pp 95­ bug, Podisus maculiventris (Hemiptera: Pentatomi­ 119. New York: Walter de Gruyter. dae). EnvironmentalEntomology 13:1031-1036. DECEMBER 1998 EXPLOITATION OF SEXUAL SIGNALS 433

AldrichJ R, LusbyW R, KochanskyJ P. 1986. Identi­ selection in relation to pest-management strate­ fication of a new predaceous stink bug phero­ gies. AnnualReview ofEntomology 41:211-229. mone and its attractiveness to the eastern yel­ BrahmacharyRL, Sarkar M P, DuttaJ. 1992. Chemi­ lowjacket. Experientia 42:583-585. cal signals in the tiger. In Chemical Signals in Verte­ AldrichJ R, Lusby W R, Marron B E, Nicolaou K C, brates, Volume 6, edited by RL Doty and D Hoffman M P, Wilson L T. 1989. Pheromone Muller-Schwarze, pp 471-475. New York: Ple­ blends ofgreen stink bugs and possible parasit­ num Press. oid selection. Natunvissenschaften 76: 173-175. Buchler E R, Childs S B. 1981. Orientation to dis­ Allen G R. 1995a. The biology of the phonotactic tant sounds by foraging bigbrown bats (Eptesicus parasitoid, Homotrixa sp. (Diptera: Tachinidae), fuscus). Animal Behaviour29:428-432. and its impact on the survival ofmale Sciarasaga Burk T. 1982. Evolutionary significance of preda­ quadrata (Orthoptera: Tettigoniidae) in the field. tion on sexuallysignallingmales. FloridaEntomol­ Ecological Entomology 20:103-110. ogist 65:90-104. Allen GR. 1995b. The calling behaviour and spatial Burk T. 1988. Acoustic signals, arms races and the distribution of male bushcrickets (Sciarasaga costs of honest signalling. Florida Entomologist quadrata) and their relationship to parasitism by 71 :400-409. acoustically orienting tachinid flies. Ecological Butcher G S, Rohwer S. 1989. The evolution ofcon­ Entomology 20:303-310. spicuous and distinctive coloration for commu­ Andersson M. 1994. Sexual Selection. Princeton: nication in birds. In Current Ornithology, Volume Princeton University Press. 6, edited by DM Power, pp 51-108. New York: Arakaki N, Wakamura S, Yasuda T. 1996. Phoretic Plenum Press. egg parasitoid, Telenomus euproctidis (Hymenop­ CadeW. 1975. Acoustically orienting parasitoids: fly tera: Scelionidae), uses sex pheromone of tus­ phonotaxis to cricket song. Science 190:1312-1313. sock moth Euproctis taiwana (Lepidoptera: Ly­ Cade W. 1980. Alternative male reproductive be­ mantriidae) as a kairomone.Journal ofChemical haviors. Florida Entomologist 63:30-45. CadeWH, Ciceran M, MurrayA-M. 1996. Temporal Ecology 22:1079-1085. Austad S N. 1984. A classification of alternative re­ patterns of parasitoid fly (Ormia ochracea) at­ traction to field cricket song (Gryllus integer). Ca­ productive behaviors and methods for field-test­ nadianJournal ofZoology 74:393-395. ing ESS models. American Zoologist 24:309-320. Carde R T, BakerT C. 1984. Sexual communication BadyaevA V, LeafE S. 1997. Habitat associations of with pheromones. In Chemical Ecology ofInsects, song characteristics in Phylloscopus and Hippolais edited by WJ Bell and RT Carde, pp 355-383. warblers. The Auk 114:40-46. New York: Chapman and Hall. Bailey W J. 1991. Acoustic Behaviour ofInsects. Lon­ Christy J H. 1995. Mimicry, mate choice, and the don: Chapman and Hall. sensory trap hypothesis. The American Naturalist Baker R R, Parker GA. 1979. The evolution ofbird 146:171-181. coloration. Philosophical Transactions ofthe Royal Clifton K E, Robertson D R. 1993. Risks ofalterna­ Society ofLondon Series B 287:63-130. tive mating tactics. Nature 366:520. Bass A. 1992. Dimorphic male brains and alterna­ Csada R D, NeudorfD L. 1995. Effects ofpredation tive reproductive tactics in a vocalizing fish. risk on mate choice in female Acheta domesticus Trends in Neurosciences 15:139-145. crickets. Ecological Entomology 20:393-395. Bell P D. 1979. Acoustic attraction of herons by Daly M. 1978. The cost ofmating. The American Nat­ crickets.Journal ofthe New York Entomological Soci­ uralist 112:771-774. ety 87:126-127. Darwin C. 1871. The Origin ofSpecies and the Descent BelwoodJJ. 1990. Anti-predator defences and ecol­ ofMan. New York: Random House. ogy ofneotropical forest katydids, especially the Dawkins R, KrebsJ R. 1979. Arms races between and Pseudophyllinae. In The Tettigoniidae: Biology, within species. Proceedings of the Royal Society of Systematics andEvolution, edited byWJ Baileyand London Series B 205:489-511. DC FRentz, pp 8-26. Bathurst (Australia): Dicke M, Sabelis M W. 1992. Costs and benefits of Crawford House Press. chemical information conveyance: proximate BelwoodJ J, Morris G K 1987. Bat predation and and ultimate factors. In Insect Chemical Ecology, its influence on calling behavior in neotropical edited by BD Roitberg and MB Isman, pp 122­ katydids. Science 238:64-67. 156. New York: Chapman and Hall. Bleckmann H, Bender M. 1987. Water surface Dixon W N, Payne T L. 1980. Attraction of ento­ waves generated by the male pisaurid spider Do­ mophagous and associate insects of the southern lomedes triton (Walckenaer) during courtship be­ pine beetle to beetle- and host tree-producedvola­ havior.Journal ofArachnology 15:363-369. tiles. Journal of the Georgia Entomological Society Boake CRB, ShellyTE, Kaneshiro KY. 1996. Sexual 15:378-389. 434 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

Dominey W J. 1980. Female mimicry in male blue­ Gryllotalpidae) by Euphasiopteryx depleta (Dip­ gill sunfish-a genetic polymorphism? Nature tera: Tachinidae). Revista Brasileira de Biologia 284:546-548. 47:371-374. Edgecomb RS, RobertD, Read M P, HoyRR. 1995. Gadgil M. 1972. Male dimorphism as a conse­ The tympanal hearing organ of a fly: phyloge­ quence ofsexual selection. TheAmericanNatural­ netic analysis of its morphological origins. Cell ist 106:574-580. and Tissue Research 282:251-268. Godfray HCJ. 1994. Parasitoids. Princeton: Princeton Ehrlich P R, Raven P H. 1964. Butterfliesand plants: University Press. a study in coevolution. Evolution 18:586-608. Godin J-G J. 1995. Predation risk and alternative EndlerJ A. 1978. A predator's view ofanimal color mating tactics in male Trinidadian guppies (Poe­ patterns. Evolutionary Biology 11: 319-364. cilia reticulata). Oecologia 103:224-229. Endler J A. 1980. Natural selection on color pat­ GodinJ-GJ, Briggs S E. 1996. Female mate choice terns in Poecilia reticulata. Evolution 34:76-91. underpredation risk in the guppy. AnimalBehav­ EndlerJ A. 1982. Convergent and divergent effects iour51:117-130. ofnatural selection on color patterns in two fish Gong A, Gibson R M. 1996. Reversal of a female faunas. Evolution 36:178-188. preference after visual exposure to a predator EndlerJ A. 1983. Natural and sexual selection on in the guppy, Poecilia reticulata. AnimalBehaviour color patterns in poeciliid fishes. Environmental 52:1007-1015. Biology ofFishes 9:173-190. Gotmark F. 1994. Does a novel bright colour patch Endler J A. 1987. Predation, light intensity and increase or decrease predation? Red wings re­ courtship behaviour in Poecilia reticulata (Pisces: duce predation risk in European blackbirds. Pro­ Poeciliidae). Animal Behaviour 35:1376-1385. ceedings of the Royal Society of London Series B 256: Endler J A. 1988. Sexual selection and predation 83-87. risk in guppies. Nature 332:593-594. Gotmark F, Unger U. 1994. Are conspicuous birds Endler J A. 1991. Variation in the appearance of unprofitable prey? Field experiments with hawks guppy colorpatterns to guppies and their preda­ and stuffed prey species. The Auk 111:251-262. tors under different visual conditions. Vision Re­ Grafen A. 1990. Biological signals as handicaps. search 31:587-608. Journal ofTheoretical Biology 144:517-546. EndlerJ A. 1992. Signals, signal conditions, and the Grant V. 1995. Sexual selection in plants: pros and direction of evolution. The American Naturalist cons. Proceedings ofthe NationalAcademy ofSciences 139(supplement):S125-S153. USA 92:1247-1250. EndlerJ A. 1993. Some general comments on the Greenfield M D. 1981. Moth sex pheromones: an evolution and design ofanimal communication evolutionary perspective. Florida Entomologist 64: systerns. Philosophical Transactions ofthe RoyalSoci­ 4-17. ety ofLondon Series B 340:215-225. Greenfield M D, CoffeltJA. 1983. Reproductive be­ EndlerJ A, Houde A E. 1995. Geographic variation haviour of the lesser waxmoth, Achroia grisella in female preferences for male traits in Poecilia (Pyralidae: Galleriinae): signalling, pair forma­ reticulata. Evolution 49:456-468. EndlerJ A, Thery M. 1996. Interacting effects oflek tion, male interactions and mate guarding. Be­ placement, display behavior, ambient light, and haviour84:287-315. color patterns in three neotropical forest-dwell­ Gwynne D T, Edwards E D. 1986. Ultrasound pro­ ing birds. The American Naturalist 148:421-452. duction by genital stridulation in Syntonarcha iri­ Evans M R, Thomas AL R. 1992. The aerodynamic astis (Lepidoptera: Pyralidae): long-distance sig­ and mechanical effects ofelongated tails in the nalling by male moths? Zoological Journal of the scarlet-tufted malachite sunbird: measuring the Linnean Society 88:363-376. cost ofa handicap. AnimalBehaviour43:337-347. Gwynne D T, Morris GK 1983. Orthopteran Matzng FarrJ A. 1975. The role of predation in the evolu­ Systems: Sexual Competition in a Diverse Group of tion ofsocial behavior ofnatural populations of Insects. Boulder (CO): Westview Press. the guppy, Poecilia reticulata (Pisces: Poecilii­ Halliday T. 1980. Sexual Strategy. Chicago: Univer­ dae). Evolution 29:151-158. sity ofChicago Press. Fowler H G. 1987. Field behavior of Euphasiopteryx Hansen K. 1983. Reception of bark beetle phero­ depleta (Diptera: Tachinidae): phonotactically mone in the predaceous clerid beetle, Thanasi­ orienting parasitoids of mole crickets (Orthop­ musformicarius (Coleoptera: Cleridae).Journal of tera: Gryllotalpidae: Scapteriscus). Journal of the Comparative Physiology 150:371-378. New York Entomological Society 95:474-480. HardieJ, Nottingham S F, Powell W, Wadhams LJ. Fowler H G, Garcia C R. 1987. Attraction to synthe­ 1991. Synthetic aphid sex pheromone lures fe­ sized songs and experimental and natural para­ male parasitoids. EntomologiaExperimentalis etAp­ sitism of Scapteriscus mole crickets (Orthoptera: plicata 61:97-99. DECEMBER 1998 EXPLOITATION OF SEXUAL SIGNALS 435

HedrickA V, Dill L M. 1993. Mate choice by female Lakes-Harlan R, Heller K-G. 1992. Ultrasound-sen­ crickets is influenced by predation risk. Animal sitive ears in a parasitoid fly. Naturwissenschaften Behaviour46: 193-196. 79:224-226. Heller K-G. 1995. Acoustic signallingin palaeotrop­ Lande R. 1981. Models ofspeciation by sexual selec­ ical bushcrickets (Orthoptera: Tettigonioidea: tion on polygenic traits. Proceedings ofthe National Pseudophyllidae): does predation pressure by Academy ofSciences USA 78:3721-3725. eavesdroppingenemies differ in the Palaeo- and Landolt P ]. 1997. Sex attractant and aggregation Neotropics?Journal ofZoology (London) 237:469­ pheromones of male phytophagous insects. 485. American Entomologist 43: 12-22. Heller K-G, von Helversen D. 1993. Calling behav­ Lehmann G. 1996. The influence ofa parasitoid fly ior in bushcrickets of the genus Poecilimon with on southern European bushcrickets [abstract]. differing communication systems (Orthoptera: XX International Congress ofEntomology, Firenze, Tettigonioidea, Phaneropteridae).Journal ofIn­ Italy. sect Behavior 6:361-377. Lehmann G, Heller K-G. 1997. Song structure in Hendrichs], Katsoyannos B I, Wornoayporn V, bushcrickets is related to predation by a parasit­ Hendrichs M A. 1994. Odour-mediatedforaging oid fly [abstract]. Proceedings ofthe German Zoolog­ by yello\\jacket wasps (Hymenoptera: Vespidae): ical Society 90:244. predation on leks of pheromone-calling Medi­ Lewis W], Nordlund D A, Gueldner R C, Teal PEA, terranean fruit fly males (Diptera: Tephritidae). Tumlinson] H. 1982. Kairomones and their use Oecologia 99:88-94. for management of entomophagous insects. Henry C S. 1994. Singing and cryptic speciation in XIII. Kairomonal activity for Trichogramma spp. insects. Trends in Ecology & Evolution 9:388-392. ofabdominal tips, excretion, and a synthetic sex Herms D A, Haack RA, Ayres B D. 1991. Variation pheromone. Journal of Chemical Ecology 8: 1323­ in semiochemical-mediatedprey-predatorinter­ 1331. action: Ips pini (Scolytidae) and Thanasimus du­ Lloyd] E. 1966. Signals and mating behavior in sev­ bius (Cleridae). Journal of Chemical Ecology 17: eral fireflies (Coleoptera: Lampyridae). Coleop­ 1705-1714. terists Bulletin 20:84-90. Hoglund ], Alatalo R V. 1995. Leks. Princeton: Lloyd] E. 1973. Firefly parasites and predators. Princeton University Press. Coleopterists Bulletin 27:91-106. Hokkanen H. 1986. Polymorphism, parasites, and Lloyd] E. 1983. Bioluminescence and communica­ the native area of Nezara viridula (Hemiptera, tion in insects. Annual Review ofEntomology 28: Pentatomidae). Annales Entomologici Fennici 52: 131-160. 28-31. Lloyd] E. 1997. Firefly mating ecology, selection Hokkanen H, Pimentel D. 1984. New approach for and evolution. In The Evolution ofMating Systems selectingbiological control agents. CanadianEn­ in Insects and Arachnids, edited by] C Choe and tomologist 116:1109-1121. B] Crespi, pp 184-192. Cambridge (UK): Cam­ Houde A E. 1997. Sex, Color, and Mate Choice in Gup­ bridge University Press. pies. Princeton: Princeton University Press. Howard RD. 1988. Reproductive success in two spe­ Lloyd] E, Wing SR. 1983. Nocturnal aerial preda­ cies ofanurans. In Reproductive Success, edited by tion of fireflies by light-seeking fireflies. Science TH Clutton-Brock, pp 99-117. Chicago: Univer­ 222:634-635. sity ofChicago Press. LundbergS, Lofstedt C. 1987. Intra-specific compe­ Janzen D H. 1980. When is it coevolution? Evolu­ tition in the sex communication channel: a se­ tion 34:611-612. lective force in the evolution of moth phero­ Johnstone R A. 1995. Sexual selection, honest ad­ mones? Journal ofTheoretical Biology 125:15-24. vertisement and the handicap principle: re­ Magnhagen C. 1991. Predation risk as a cost of re­ viewing the evidence. Biological Reviews of the production. Trends in Ecology & Evolution 6:183­ Cambridge Philosophical Society 70: 1-65. 186. Kaneshiro KY, Boake CR B. 1987. Sexual selection Magurran A E, Seghers B H. 1990. Risk sensitive and speciation: issues raised by Hawaiian Dro­ courtship in the guppy (Poecilia reticulata). Be­ sophila. Trends in Ecology & Evolution 2:207-212. haviour 112:194-201. Kirkpatrick M, Ryan M]. 1991. The evolution of Mangold] R. 1978. Attraction ofEuphasiopteryx och­ mating preferences and the paradox of the lek. racea, corethrella sp. and gryllids to broadcast Nature 350:33-38. songs ofthe southern mole cricket. Florida Ento­ Kyparissoudas D S. 1987. The occurrence ofEncar­ mologist 61 :57-61. sia perniciosi in areas of northern Greece as as­ Maynard Smith]. 1958. Sexual selection. In A Cen­ sessed by sex pheromone traps ofits host Quad­ tury ofDarwin, edited by SABarnett, pp 231-244. raspidiotus perniciosis. Entomologia Hellenica 5:7-12. London: Heinemann. 436 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

Maynard Smith]. 1991. Theories of sexual selec­ Nordlund D A, Lewis W], Gueldner R C. 1983. Kai­ tion. Trends in Ecology & Evolution 6:146-151. romones and their use for managementofento­ McClain D C, Rock G C, Woolley] B. 1990. Influ­ mophagous insects XIV. Response of Telenomus ence of trap color and San]ose scale (Homop­ remus to abdominal tips of Spodopterafrugiperda, tera: Diaspididae) pheromone on sticky trap (Z)-9-tetradecene-l-01 acetate and (Z)-9-dode­ catches of 10 aphelinid parasitoids (Hymenop­ cene-l-01 acetate. Journal of Chemical Ecology tera). EnvironmentalEntomology 19:926-931. 9:695-701. McKeever S. 1977. Observations of Corethrella feed­ Olsson M. 1993. Nuptial coloration and predation ing on tree frogs (Hyla). Mosquito News 37:522­ risk in model sand lizards, Lacerta agilis. Animal 523. Behaviour46:410-412. McNeil] N, Delisle], Cusson M. 1997. Regulation Otte D. 1974. Effects and functions in the evolution of pheromone production in lepidoptera: the ofsignalingsystems. AnnualReview ofEcology and need for an ecological perspective. In Insect Pher­ Systematics 5:385-417. omone Research: New Directions, edited by RT Parkman] P, Frank] H, Walker T], Schuster D]. Carde and AK Minks, pp 31-41. New York: 1996. Classical biological control of Scapteriscus Chapman and Hall. spp. (Orthoptera: Gryllotalpidae) in Florida. En­ Mendel Z, Zegelman L, Hassner A, Assael F, Harel vironmentalEntomology 25:1415-1420. M, Tam S, Dunkelblum E. 1995. Outdoor at­ Paterson HE H. 1985. The recognition concept of tractancy of males of Matsucoccus josephi (Ho­ species. In Species and Speciation, edited by ES moptera: Matsucoccidae) and Elatophilus hebra­ Vrba, pp 21-29. Pretoria (South Mrica): Trans­ icus (Hemiptera: Anthocoridae) to synthetic vaal Museum. female sex pheromone of Matsucoccus josephi. Phelan P L. 1997a. Genetics and phylogenetics in Journal ofChemical Ecology 21:331-341. the evolution ofsex pheromones. In Insect Phero­ Mitchell W C, Mau RF L. 1971. Response of the moneResearch: New Directions, edited by RT Carde female southern green stinkbugandits parasite, and AK Minks, pp 563-579. New York: Chap­ Trichopoda pennipes, to male stink bug phero­ man and Hall. mone.Journal ofEconomic Entomology 64:856-859. Phelan P L. 1997b. Evolution of mate-signaling in M011er A P. 1994. Sexual Selection and the Barn Swal­ moths: phylogenetic considerations and predic­ low. Oxford: Oxford University Press. tions from the asymmetric tracking hypothesis. Moodie GEE. 1972. Predation, natural selection In The Evolution ofMating Systems in Insects and and adaptation in an unusual threespine stickle­ Arachnids, edited by] C Choe and B] Crespi, pp back. Heredity 28:155-167. 240-256. Cambridge (UK): Cambridge Univer­ Moriya S, Masakazu S. 1984. Attraction ofthe male sity Press. brown-winged green bug, Plautia stali Scott Pimentel D. 1963. Introducing parasites and preda­ (Heteroptera: Pentatomidae) for males and fe­ tors to control native pests. Canadian Entomolo­ males ofthe same species. AppliedEntomology and gist 95:785-792. Zoology 19:317-322. Morris G K. 1980. Calling display and mating behav­ Pocklington R, Dill L M. 1995. Predation on females iour of Copiphora rhinoceros Pictet (Orthoptera: or males: who pays for bright male traits? Animal Tettigoniidae). AnimalBehaviour 28:42-51. Behaviour49:1122-1124. Morris G K, Mason A C, Wall P, Belwood]]. 1994. Powell W, Hardie], Hick A], Holler C, Mann], High ultrasonic and tremulation signals in neo­ MerrittL, Nottingham SF, Wadhams L], Witthin­ tropical katydids (Orthoptera: Tettigoniidae). rich], WrightAF. 1993. Responses ofthe parasit­ Journal ofZoology (London) 233:129-163. oid Praon volucre (Hymenoptera: Braconidae) to Nechols] R, Kauffman W C. 1992. Introduction and aphid sex pheromone lures in cereal fields in overview. In Selection Criteria andEcological Conse­ autumn: implications for parasitoid manipula­ quences of Importing Natural Enemies, edited by tion. EuropeanJournal ofEntomology 90:435-438. WC Kauffman and] RNechols, pp 1-5. Lanham Prestwich K N, Brugger K E, Topping M. 1989. En­ (MD): Entomological Society ofAmerica. ergy and communication in three species of Noldus LP]], Potting RP], BarendregtH E.1991a. Hylid frogs: power input, power output and effi­ Moth sex pheromone adsorption to leafsurface: ciency. Journal ofExperimental Biology 144:53-80. bridge in time for chemical spies. Physiological Prestwich KN, WalkerT]. 1981. Energetics ofsing­ Entomology 16:329-344. ing crickets: effects oftemperature in three trill­ Noldus LP]], van Lenteren] C, Lewis W].1991b. ing species (Orthoptera: Gryllidae). Journal of How Trichogramma parasitoids use moth sex Comparative Physiology 143:199-212. pheromones as kairomones: orientation behav­ Promislow D, Montgomerie R, Martin T E. 1994. iour in a wind tunnel. Physiological Entomology Sexual selection and survival in NorthAmerican 16:313-327. waterfowl. Evolution 48:2045-2050. DECEMBER 1998 EXPLOITATION OF SEXUAL SIGNALS 437

Raffa K F, Dahlsten D L. 1995. Differential re­ scale Aonidiella aurantii (Mask.) (Hom., Diaspid­ sponses among natural enemies and prey to idae) and its Aphytis spp. (Hym., Aphelinidae) bark beetle pheromones. Oecologia 102:17-23. parasitoids. Journal of Applied Entomology 105: Raffa KF, Klepzig KD. 1989. Chiral escape ofbark 483-489. beetles from predators responding to a bark Sanderford M V, ConnerW E. 1995. Acoustic court­ beetle pheromone. Oecologia 80:566-569. ship communication in Syntomeida epilais Wlk. Rand AS, Bridarolli M E, Dries L, Ryan Mj. 1997. (Lepidoptera: Arctiidae, Ctenuchinae). Journal Light levels influence female choice in tungara ofInsect Behavior 8: 19-31. frogs: predation risk assessment? Copeia 1997: Schemske D W. 1983. Limits to specialization and 447-450. coevolution in plant-animal mutualisms. In Co­ Reynoldsj D, Gross M R, Coombs Mj. 1993. Envi­ evolution, edited by MH Nitecki, pp 67-109. Chi­ ronmental conditions and male morphologyde­ cago: University ofChicago Press. termine alternative mating behaviour in Trini­ Slagsvold T, Dale S, Kruszewicz A. 1995. Predation dadian guppies. AnimalBehaviour45:145-152. favours cryptic coloration in breeding male pied Rice R E. 1969. Response of some predators and flycatchers. AnimalBehaviour50:1109-1121. parasites ofIps confusus (Lee.) (Coleoptera: Sco­ Soper R S, Shewell G E, Tyrrell D. 1976. Colconda­ lytidae) to olfactory attractants. Contributions myia auditrix nov. sp. (Diptera: Sarcophagidae), from Boyce Thompson Institute 24:189-194. a parasite which is attracted by the mating song RobertD,Amorosoj, HoyRR.1992. The evolution­ of its host, Okanagana rimosa (Homoptera: Ci­ ary convergence of hearing in a parasitoid fly cadidae). Canadian Entomologist 108:61-68. and its cricket host. Science 258:1135-1137. Spangler H G. 1984. Silence as a defense against Robert D, Read M P, Hoy R R. 1994. The tympanal predatory bats in two species of calling insects. hearing organ of the parasitoid fly Ormia ochra­ Southwestern Naturalist 29:481-488. cea (Diptera, Tachinidae, Ormiini). Cell and Tis­ Spieth H T. 1974. Mating behavior and evolution sue Research 275:63-78. ofthe Hawaiian Drosophila. In GeneticMechanisms Romer H. 1993. Environmental and biological con­ ofSpeciation in Insects, edited by Mj D White, pp straints for the evolution of long-range signal­ 94-101. Dordrecht (Holland): Reidel. ling and hearingin acoustic insects. Philosophical Sternlicht M. 1973. Parasitic wasps attracted by the Transactions ofthe Royal Society ofLondon Series B sex pheromone of their coccid host. Entomo­ 340:179-185. phaga 18:339-342. Rotenberry j T, Zuk M, Simmons L W, Hayes C. Stowe M K, Turlings T Cj, Loughrinj H, Lewis Wj, 1996. Phonotactic parasitoids and cricket song Tumlinson j H. 1995. The chemistry of eaves­ structure: an evaluation of alternative hypothe­ dropping, alarm, and deceit. Proceedings ofthe Na­ ses. Evolutionary Ecology 10:233-243. tional Academy ofSciences USA 92:23-28. Ryan Mj. 1985. The TungaraFrog: A Study in Sexual Sullivan B K. 1989. Passive and active female choice: Selection and Communication. Chicago: University a comment. AnimalBehaviour37:692-694. ofChicago Press. Svensson M. 1996. Sexual selection in moths: the Ryan M j. 1990. Sexual selection, sensory systems role of chemical communication. Biological Re­ and sensory exploitation. Oxford Surveys inEvolu­ views ofthe Cambridge Philosophical Society 71:113­ tionary Biology 7:157-195. 135. Ryan Mj, Rand AS. 1990. The sensory basis ofsex­ Thompsonj N. 1994. The Coevolutionary Process. Chi­ ual selection for complex calls in the tungara cago: University ofChicago Press. frog, Physalaemus pustulosus (sexual selection for Thornhill R. 1979. Male and female sexual selec­ sensory exploitation). Evolution 44:305-314. tion and the evolution of mating systems in in­ Ryan Mj, Tuttle M D, RandAS.1982. Batpredation sects. In Sexual Selection and Reproductive Competi­ and sexual advertisement in a neotropical an­ tion in Insects, edited by MS Blum and NA Blum, uran. The American Naturalist 119:136-139. pp 81-121. NewYork: Academic Press. Sakaluk S K. 1990. Sexual selection and predation: Thornhill R, Alcockj. 1983. The Evolution ofInsect balancing reproductive and survival needs. In Mating Systems. Cambridge (MA): Harvard Uni­ Insect Defenses: Adaptive Mechanisms and Strategies versity Press. ofPrey and Predators, edited by DL Evans andj 0 Trail P W. 1987. Predation and antipredator behav­ Schmidt, pp 63-90. Albany: State University of ior at Guianan Cock-of-the-Rock leks. The Auk NewYork Press. 104:496-507. SakalukS K, Belwoodjj. 1984. Gecko phonotaxis to Trivers R L. 1972. Parental investment and sexual cricketcallingsong: a case ofsatellite predation. selection. In Sexual Selection and the Descent of AnimalBehaviour 32:659-662. Man, 1871-1971, edited byB Campbell, pp 136­ Samways Mj. 1988. Comparative monitoring ofred 179. London: Heinemann. 438 THE QUARTERLY REVIEW OF BIOLOGY VOLUME 73

Tumlinson J H, Turlings TC J, Lewis W J. 1993. siopteryx ochracea) in north Florida. Florida Ento­ Semiochemicallymediated foraging behaviorin mologist 69:678-685. beneficial parasitic insects. Archives ofInsect Bio­ Walker TJ. 1993. Phonotaxis in female Ormia ochra­ chemistry and Physiology 22:385-391. cea (Diptera: Tachinidae), a parasitoid of field Tuttle M D, Ryan MJ. 1981. Bat predation and the crickets.Journal ofInsect Behavior6:389-410. evolution offrog vocalizations in the neotropics. Walker T J, Wineriter SA. 1991. Hosts ofa phono­ Science 214:677-678. tactic parasitoid and levels of parasitism (Dip­ Tuttle M D, Ryan MJ. 1982. The role ofsynchro­ tera: Tachinidae: Ormia ochracea). Florida Ento­ nized calling, ambient light, and ambient noise, mologist 74:554-559. in anti-bat-predator behavior of a treefrog. Be­ West-Eberhard M J. 1983. Sexual selection, social havioralEcology and Sociobiology 11:125-131. competition, and speciation. The Quarterly Re­ Tuttle M D, Ryan MJ, BelwoodJJ. 1985. Acoustical view ofBiology 58:155-183. resource partitioning by two species of phyllo­ Wiley R H. 1991. Associations of song properties stomid bats (Trachops cirrhosus and Tonatia sylvi­ with habitats for territorial oscine birds of east­ cola). Animal Behaviour 33:1369-1371. ern North America. The American Naturalist Tuttle M D, Taft L K, Ryan M J. 1981. Acoustical 138:973-993. location ofcalling frogs by philander opossums. Williams G C. 1975. Sex and Evolution. Princeton: Biotropica 13:233-234. Princeton University Press. Tuttle M D, Taft L K, Ryan MJ. 1982. Evasive behav­ Williams G C. 1992. Natural Selection: Domains, Lev­ iour ofa frog in response to bat predation. Ani­ els, and Challenges. Oxford: Oxford University Press. mal Behaviour30:393-397. Willson M F, Burley N. 1983. Mate Choice in Plants: Verrell P A. 1991. Illegitimate exploitation ofsexual Tactics, Mechanisms, and Consequences. Princeton: signalling systems and the origin of species. Princeton University Press. Ethology, Ecology & Evolution 3:273-283. Wing S R. 1988. Cost of mating for female insects: ViteJ P, Williamson D L. 1970. Thanasimus dubius: risk ofpredation in Photinus collustrans (Coleop­ prey perception. Journal of Insect Physiology 16: tera: Lampyridae). The American Naturalist 131: 233-239. 139-142. WagnerWE,Jr. 1996. Convergentsongpreferences Wood D L, BedardW D, Tilden P E, Silverstein R M, between female field crickets and acoustically RodinJ 0.1968. Response ofIps confusus to syn­ orienting parasitoid flies. Behavioral Ecology 7: thetic sex pheromones in nature. Science 159: 279-285. 1373-1374. Walker T J. 1964. Experimental demonstration ofa Zahavi A. 1975. Mate selection-a selection for a cat locating orthopteran prey by the prey's call­ handicap. Journal of Theoretical Biology 53:205­ ing song. Florida Entomologist 47:163-165. 214. Walker T J. 1969. Acoustic synchrony: two mecha­ Zuk M. 1991. Sexual ornaments as animal signals. nisms in the snowy tree cricket. Science 166:891­ Trends in Ecology & Evolution 6:228-231. 894. Zuk M, Simmons L W, Cupp L. 1993. Song charac­ Walker T J. 1983. Diel patterns ofcalling in noctur­ teristics ofparasitized and unparasitized popula­ nal Orthoptera. In Orthopteran Mating Systems: tions ofthe field cricket Teleogryllus oceanicus. Be­ Sexual Competition in a Diverse Group ofInsects, ed­ havioralEcology and Sociobiology 33:339-343. ited by DT Gwynne and G K Morris, pp 45-72. Zuk M, Simmons L W, RotenberryJ T. 1995. Acous­ Boulder (CO): Westview Press. tically-orienting parasitoids in calling and silent Walker T J. 1986. Monitoring the flights of field males of the field cricket Teleogryllus oceanicus. crickets (Gryllus spp.) and a tachinid fly (Eupha- Ecological Entomology 20:380-383.