C 2006) Journal of Insect Behavior, Vol. 19, No. 2, March 2006 ( DOI: 10.1007/s10905-006-9016-5
Complex Non-Aerial Contests in the Lekking Butterfly Charis cadytis (Riodinidae) Gabriela W. Chaves,1 Claudio E. G. Patto,1,2 and Woodruff W. Benson3,4 Revised July 13, 2005; accepted November 15, 2005 Published online: June 20, 2006
Lekking butterflies typically defend territories using acrobatic aerial pursuits. Focal-method observations on marked Charis cadytis in SE Brazil revealed an unusual lek organization in which contest males disputed small core areas, whereas non-combative satellite males perched just outside their borders. Territorial interactions commonly began with two adversaries facing one another in a slow, non-contact ascending flight seemingly related to assessment. In disputes that continued, rival males perched on leaves where they engaged in one or more pushing bouts separated by short pursuits. In these sumo-like contests, obliquely facing males pushed their partially opened wings against one another until one was tilted sideways and flew off. Contest structure may be controlled by intruders that, by perching, provoke low-intensity contests that help prolong their stay in high-quality mating areas. KEY WORDS: alternative reproductive behavior; fighting; landmark encounter site; lek; lepidoptera; satellite behavior; territoriality.
INTRODUCTION Insects frequently use aggressiveness to repel rivals and thereby increase mating success (and Alcock, 1983; Andersson, 1994). In butterflies males 1Graduate
Program in Ecology, IB, Universidade Estadual de Campinas, Campinas, SP, Brazil. 2Present address: Science Department, St. Francis College, R. Belgica, ´ ˜ 399, 01448-030, Sao Paulo, SP, Brazil. 3 Depto de Zoologia, IB, Universidade Estadual de Campinas, Campinas, SP, Brazil. 4 To whom correspondence should be addressed at Departamento de Zoologia, IB, UNICAMP, C.P. 6109, 13083-970, Campinas, SP, Brazil; e-mail:
[email protected]. 179 C 2006 Springer Science+Business Media, Inc. 0892-7553/06/0300-0179/1
180
Chaves, Patto, and Benson
commonly defend mating territories (Kemp and Wiklund, 2001). These may be established near host plants (Rutowski and Gilchrist, 1988; Lederhouse et al., 1992) or adult resources (Fischer and Fiedler, 2001) or positioned along dispersal routes (Baker, 1972), places where receptive females are likely to occur. Many, however, seem to be simple landmark encounter sites (Davies, 1978; Lederhouse, 1982; Knapton, 1985) that lack resources and to which females come exclusively to mate. The latter resemble in many ways ¨ the leks of vertebrates (Hoglund and Alatalo, 1995), although, unlike some vertebrates, females do not seem to select individual mates (Andersson, 1994). The aerial contests of butterflies vary considerably across species (Baker, 1972; Wickman and Wiklund, 1983; Benson et al., 1989; Rutowski, 1992; Kemp, 2003) and can involve physical contact (Freitas et al., 1997) that at times may seriously damage opponents (Eff, 1962). Although perched butterflies may directly defend food (Owen, 1971; Murawski and Gilbert, 1986) or mates (Deinert et al., 1994), and diptera (Dodson, 1997; Bonduriansky and Brooks, 1999) and bees (Alcock and Houston, 1996) sometimes defend mating stations without flying, so far as known territorial butterflies interact using acrobatic aerial pursuits (Kemp and Wiklund, 2001) that end when all but one contestant is expelled (Davies, 1978; Alcock and O’Neill, 1986). A number of territorial butterflies show alternative mating tactics in which individuals, rather than defend territories, search for mates by patrolling in other areas (Davies, 1978; Alcock, 1994). However, unlike some other insect groups (Cade, 1979; Convey, 1989) and many vertebrates (Perrill et al., 1978; Taborsky, 1994), butterflies are apparently not known to use satellite or sneaking tactics to obtain matings. On a number of occasions we observed an unusual behavior in the riodinid butterfly Charis cadytis Hewit in which individuals, later identified as males, perched together on leaf-tops where they rubbed against one another as if courting (see Tinbergen et al., 1942; Rutowski, 1983). Charis has been previously noted by A. R. Wallace (1853) to differ from most riodinids by characteristically resting on the tops of leaves, rather than on the undersides. Like a number of congeners, the basic wing color of C. cadytis is dark iridescent blue, but it seems unique among Charis species in possessing a narrow band of yellow spots—much better developed in the female and thus allowing reliable visual separation of the sexes—traversing the forewing apex (K. S. Brown Jr., pers. com.). Closer observation of C. cadytis suggested that the behavior of rubbing together might be related to disputes over mating sites. To investigate this possibility, between October 2001 and June 2002 we marked butterflies using a non-intrusive techniques and recorded male behavior with the aid of a video camera.
Non-Aerial Butterfly Contests
181
MATERIAL AND METHODS The present study was undertaken at approximately 1000 m elevation in the Environmental Protection Zone of the Serra do Japi, a network of ˜ Paulo State, wooded ridges located just west of the town of Jundia´ı, Sao Brazil (23◦ 14 S/46◦ 56 W). At this locality male Charis cadytis form scattered aggregations associated with sunny, moist clearings along trails in the otherwise mid to late successional subtropical moist forest that dominates the area. After conducting preliminary observations, starting in November (mid-spring), 2001, two of us (GWC and CGP) working together surveyed 2-4 times a day a 3.5-km-long transect (end points at 23◦ 13 59 /46◦ 57 02 W and 23◦ 14 55 /46◦ 56 40 W) along a winding primitive dirt road that contained several persistent Charis groups (see Fig. 1). Although the butterflies are small (male forewing length c. 12 mm), they are easily distinguished from other butterflies found at the study site (Brown, 1972), and their bold color pattern, habit of resting with wings spread, and preference for low perches along sunny paths made them easy to locate and observe. Upon encountering interacting Charis males, we stopped and filmed the interactions of a focal individual using a Canon 8 ES2500 (8 mm) video camera with image stabilization. Filming sessions usually lasted between 15 and 60 min, depending upon the intensity of the interaction. On cloudy and on cool days (<20◦ C.) butterfly activity declined or was absent, and the observations were concentrated during warmer hs of sunny days when butterflies were more active. C. cadytis males, in addition to perching near the ground, are tolerant of approach to within 2 m, and often closer, so it was usually possible to film essentially all of the activities of a focal individual during a session. During the filming, the behavior of other individuals was narrated on the video-camera recorder or registered in field notes. A total of 38 h of male–male interactions were taped in this way. Later the observations were transcribed and different behaviors (see below) timed with a precision of 1 s. Behavioral sequences were analyzed using a flow chart. Data analysis was conducted using SYSTAT (Wilkinson, 1990) following Zar (1996). We present measurements in the form of means ± 1 standard error. Charis males almost always rested with their wings open and extended horizontally from the body, and many could be individually recognized because of distinctive wing damage. Because C. cadytis are small and delicate, those lacking naturally occurring marks were individually marked with a haphazard spattering of spots by blowing a fine spray of bright yellow enamel onto the wings of perched butterflies from a capillary tube. Thus,
182
Chaves, Patto, and Benson
Fig. 1. Map of the study area at Serra do Jap´ı, Brazil, showing the locations of Charis cadytis leks.
butterflies were never touched or physically constrained in a way that could adversely affect normal behavior. The minute paint marks did not noticeably alter wing appearance and marking had no visible effect of butterfly behavior or survival (C.G. Patto and G.W. Chaves, in preparation). This makes us confident that we have registered the full repertoire of commonly occurring natural behaviors in frequencies representative of the study conditions.
Non-Aerial Butterfly Contests
183
RESULTS Butterfly Distribution and Mating Behavior Charis cadytis were found aggregated at a few fixed points along the transect route (Fig. 1). Many of these points also harbored aggregations prior to 1999, suggesting that persistent habitat features are in part responsible for their location. Both here and at other sites aggregations tended to occur in moist, sunny places, often with a carpet of short, herbaceous vegetation, near forest edges. Butterflies were present in greater numbers between 10 and 15 h on sunny, warm days, with early arrivals and late departures extending this interval by about 1.5 hours. Flower visiting was frequent, especially in the early and late hs of the day when territorial activity was less intense. Flowers, however, were not especially concentrated where the butterflies clumped, and most of the time butterflies either remained perched on low vegetation or participated in more or less complex interactions with conspecifics. The aggregations, or “colonies,” of C. cadytis were composed exclusively of males, and the same individual males were present at given sites on successive days. Of the 66 individually recognizable males, 38 were reobserved on one or more days after marking, 32 exclusively at the original marking site, five at distances up to 85 m and one other 350 m from where it was first marked. Groups could contain five or more butterflies, but usually fewer were present, and it was not uncommon for a sole male to continue at a given locale for a week or more. Females, on the other hand, were infrequent (35 among 503 sightings—field sex ratio 13.4:1) and only fleetingly present at points along the transect, and, except for mating females, did not occur associated with male aggregations. Before and during the present study, six C. cadytis females were observed mating, all in places occupied by perching males. In two matings observed from the beginning, a male flew after a passing female which immediately landed and mated without further ado. In both the female was in “new” condition, apparently having just eclosed from the pupa. We therefore conclude that male C. cadytis in this population are highly sedentary and form lek-like aggregations at sites suitable for encountering newly emerged females.
Lek Structure C. cadytis aggregations were highly structured with regard to the spatial distribution of individual butterflies and the ways in which they interacted.
184
Chaves, Patto, and Benson
Agonistic interactions between rival males took the form of frontal flights, persistent pursuits and pushing contests (see below) and seemed restricted to one or two small fixed areas within a lekking site (arena). These sites were consistently occupied and disputed by particular males, whereas outside of these patches butterflies showed a high tolerance for one another and at times perched on neighboring leaves without conflict. Males at leks did not behave in ways that suggested that they were releasing pheromones. In bird leks a display site is sometimes called a “stage” (Johnsgard, 1994). We call the disputed patches at C. cadytis leks “core areas” to avoid implying that males display to attract the attention of females. Core areas of C. cadytis leks typically occurred at sunny wide places on a trail or roadside and were about 3–5 m long, a few meters deep, and dominated by 0.3–1 m tall herbaceous growth. A lek could have a single core area (e.g., site A, see Fig. 1) or two contiguous core areas (e.g., site C) if a sufficiently large clearing were available. Core areas comprised only a small portion of the total area of a lek. Each core generally contained either a single resident male Charis that spent most of its time perched on the tops of leaves or two (or more) males that repeatedly interacted like described in the paragraphs that follow. At leks with two core areas, males on each characteristically intruded and disputed with their neighbor. Bordering a core area and extending a variable distance to each side, a more or less extensive peripheral region (in one case >20 m in length) was used by other Charis males for perching and occasional flower visiting. Butterflies found outside of core areas were rarely aggressive towards one another and moved freely between perches in this region. They included both contest males—males that repeatedly interacted in core areas that were temporarily outside—and submissive individuals that seemed to avoid entering these disputed patches. Males in the peripheral area generally restricted their interactions to flying up briefly to meet conspecific males flying overhead before returning to perch. Those males that avoided core areas and interactions with contest male are tentatively interpreted to be satellite individuals awaiting an opportunity to intercept receptive females that pass through the lek periphery.
General Male Behavior on Core Areas The interactions of C. cadytis males on core areas comprised three major behavior types: face-to-face ascending flights (frontal flights), pursuits and sumo-like pushing bouts. Contest males rarely or never engage in these behaviors (except pursuit) when outside the boundaries of a core area, and putative satellite males were never observed to take part in behavior
Non-Aerial Butterfly Contests
185
Fig. 2. Two male Charis cadytis engaged in a frontal flight. The butterflies are approximately 3 cm apart and ascending vertically with their bodies almost perpendicular to the ground.
sequences with more than two behaviors (frontal flight followed by a short pursuit by the contest male) in or outside of a core area. When two males encountered one another in flight, or when a perched male first flew up to meet another that was passing by, provided one of these was a contest male, the two would often turn to face one another about 3–5 cm apart and ascend face-to-face more or less slowly with their bodies almost perpendicular to the soil (Fig. 2). Frontal flights normally began when the butterflies were 30–50 cm above the ground, the height at which males perch and fly at clearing edges, and ended when they attained a height of 1–5 m, when the butterflies separated or the interaction turned into a pursuit. During these events the two butterflies did not touch or spin around one another. In some the butterflies almost hovered, and in others the ascending flight was brisk, although still much slower than the rapid chases that sometimes develop between males. The aerial pursuits of C. cadytis were characterized by one male flying rapidly behind the other, generally through straight or curved paths inside or just adjacent to core areas, until they separate or began some other type of behavior. These chases normally occurred just above the low trail-side vegetation. In the aerial pursuits that separated bouts of “sumo” (see below), the victor of a bout seemed to follow the loser so as to engage it again at the succeeding perch. The first butterfly to leave a perch flew rapidly to a nearby leaf with the other shortly following it. At times rapid chases
186
Chaves, Patto, and Benson
carried the butterflies 5 m or more away with one following closely on the other, without, however, signs of physical aggression. These flights terminated with one butterfly abandoning the core area or with them perching together on a leaf (or, rarely, engaging in a frontal flight). Circling and spiraling maneuvers, commonly reported in other territorial butterflies (Kemp and Wiklund, 2001), were not seen. A related behavior was observed to close interactions on three occasions at site A, prior to the present film-documented study. In these an apparently victorious male followed about 20–30 cm behind a “leisurely” departing contest loser as it flew through the peripheral area away from the lek’s core. The trailing male accompanied its apparently ousted rival about 30–50 cm above the ground along the roadside to approximately 15– 20 m away from the core area, at which point the lead butterfly perched and the follower, losing itself in the trailside vegetation, returned more or less indirectly to the core area. In one of these “pursuits” the departing loser perched to rest on a low plant in mid-journey (c. 5 m beyond the limits of the core area), and the trailing male landed on a leaf 30–40 cm away, where it remained until the “leader” took flight again and resumed following. This unusual behavior, in which the winning male seems to be overseeing the departure of a rival, may signal the ousted rival its motivation or capacity to defend a core area. Immediately subsequent to short pursuits in the core area, butterflies commonly perched on the tops of low leaves and engaged in what appeared to be contests. In these, the male that first perched was shortly accompanied by the second, which landed a centimeter or two away on the same leaf top. Sometimes these events occurred so quickly that the first individual to depart a leaf had already perched again by the time the “following” butterfly took flight. Upon landing, the two butterflies immediately turned to face one another and walked forward until their partially opened wings touched (Fig. 3A). At this point the two males, each slightly offset in relation to the other, positioned their wings either (1) with the underside of a wing of one individual under and in physical contact with the corresponding wing of its opponent (Fig. 3B, “side-to-side” position) or (2) with the front edge of each wing pushing against the opposite wing of its opponent (Fig. 3C, “wings crossed” position). In the first situation, as each male opened its wings, the two butterflies pushed each other laterally and became tilted more or less sideways. Generally, one of the butterflies became increasingly inclined to the side until seemingly “pinned” against the leaf by its rival. The more tilted individual, which apparently assumed the loser role, usually flew away at this point and was immediately followed by the other male, often initiating a short pursuit that was followed by another bout of “sumo” on another leaf
Non-Aerial Butterfly Contests
187
top. A pursuit could also end with a frontal flight (rare) or with the pursuit of the losing (tilted) individual from the core area. In pushing contests with “crossed” wings, neither individual seemed able to force the other to tilt its body very much, and the contests usually ended without a clear sign as to which butterfly won. We refer to the pushing interactions of perched butterflies as “sumo contests” because of their superficial similarity to this category of oriental contest. Sumo is characterized as a “form of wrestling . . . in which a contestant wins by forcing his opponent out of the ring or by causing him to touch the ground with any part of his body except the feet” (Webster’s Encyclopedia Unabridged Dictionary). In the case of C. cadytis, the analogy seems appropriate, for contests take place on small stages (leaf tops) and when one contestant pushes over its opponent, the latter may move to
Fig. 3. (A) Two Charis cadytis on a leaf top have just perched and turned to face one another to begin a sumo-like pushing match. After positioning their wings against one another, they will push until one of the contestants tilts sideways. (B) A pushing contest in which each male has positioned its wings laterally to those of its opponent such that the underside of one wing (left or right) pushes against the underside of the corresponding wing of the other butterfly. The male on the right has gained an advantage by forcing its opponent to lean to the left. Note that each butterfly is marked with scattered minute paint spots. (C) A pushing contest in which each male has positioned its wings to push against the opposite wing of its opponent. In contests in which the wings were “crossed” in this way, pushing did not lead to a butterfly leaning conspicuously to one side, and it was normally not possible to determine the outcome of a bout. The butterflies’ bodies may be displaced to either side.
188
Chaves, Patto, and Benson
Fig. 3. Continued.
another contest site (leaf) for an additional bout or accept defeat (quit the core area). Males engaged in sumo disputes seemed to lose awareness of external stimuli and could be easily observed and photographed from a meter or less away. On occasion during male combats female C. cadytis, flower visiting, were observed to pass through the core area without being detected by either of the males. This reduction in responsiveness to outside movements
Non-Aerial Butterfly Contests
189
may be costly in terms of lost mating opportunities and risks imposed by the approach of predators such as insectivorous birds.
Male–Male Contest Dynamics Male–male interactions on core areas showed a clear temporal organization. In the following description, we define an interaction as the sequence of events or behaviors that take place between two individual butterflies, starting with their initial encounter until their separation, usually when one butterfly leaves the core area. The interactions consisted of temporal sequences of behavioral acts, or behaviors. Three are considered here: 1) frontal flights, 2) pursuits (or chases) and 3) bouts of sumo-type pushing contests. The results are based on analysis of 231 interactions involving 66 individually recognizable butterflies. Of these interactions 132 terminated with only a single behavior being recorded, whereas the most extensive interaction consisted of a sequence of 24 behavioral acts. Since Charis colonies generally contain a small number of resident dominant males, the same individual opponents participated in many of the contests, and therefore, neither acts nor interactions can be considered to be statistically independent of behavioral characteristics of individual contestants that we sampled. On the other hand, the large number of observations provides a useful characterization of the unusual contest structure in this butterfly. The interactions between male C. cadytis were rapid and dynamic, with butterflies sometimes engaging in lengthy bouts of chasing and pushing and at others shifting rapidly form one behavior to another. On October 12, 2001, reciprocal intrusions between two males on neighbouring core areas produced two unusual long interactions, each comprised of 16 behavioral acts, lasting a total of 292 and 52 s, respectively. In the first of these, sustained chases of 30 and 100 s and a sumo bout of 29 s (median act duration 14.5 s) dominated the interaction. In the second, which immediately followed, males switched quickly from one behavior to another (median duration 2 s), and longest event was a sumo that lasted 13 s. Frontal flights occurred commonly (57% of observed behavioral acts) when two individuals first encountered one another, but were relatively uncommon later on (Fig. 4). They were generally brief, with elapsed times ranging from 1 to 24 s (mean 5.0 ± 0.38 s, n = 154), and most were followed by the immediate separation of the two contestants (71% of the 130 cases observed). The few frontal flights that occurred later in interactions were generally followed by a pursuit (69% of the 35 cases observed). Successive interactions between the same two males separated by intervals of only a
190
Chaves, Patto, and Benson
Fig. 4. Flow diagram of behaviors (arrows) in male-male agonistic interactions of Charis cadytis. The diagram assumes a first-order Markov process in which the probabilities of different behaviors depend exclusively upon the immediately preceding behavior. In the diagram, circles represent different behaviors (frontal flight, pursuit, sumo) at the onset (initial) and at later points (middle) of an interaction. Solo individuals are solitary males before or after an interaction. The width of each arrow is proportional to the number of times that the sequence was observed. The percentages are conditional probabilities that specify the frequency with which one behavior is followed by the other. The number of observations upon which the conditional probabilities are based is given by the numbers (n) within circles.
few seconds were also initiated with frontal flights more than 50% of the time. Pursuits began 38% (n = 220) of all interactions and ended 62% (n = 77) of those with two or more behavior events. Pursuits starting off interactions often ended with the immediate separation of the two butterflies (41%), but many passed directly to sumo contests (see below). Pursuits were highly variable in length (1–250 s; mean 14.8 ± 1.43 s; n = 268) and also occurred frequently in the middle of interactions (46% of 628 events) where they generally alternated with sumo contests, which also occurred at high frequencies (29% of 628 events). Although Sumo contests followed 51% of initial pursuits, and the association became even stronger after an interaction was underway (68% χ2 = 6.80, df = 1, p < 0.02).
Non-Aerial Butterfly Contests
191
Pushing bouts occurred with butterflies perched on top of exposed leaves. These ranged from 1 to 149 s (mean 22.1 ± 1.42 s, n = 176). Few interactions began (5% of 225) or ended (11% of 194) with bouts of sumo, but most sumo events occurred in the middle of interactions and were separated by short pursuits. The interactions between male Charis cadytis in core areas can be summarized as follows. Contests usually started with either a frontal flight or, less commonly, a pursuit, and most interactions went no farther. Frontal flights and pursuits at the onset of interactions seemed to involve evaluations that allowed the rapid resolution of most contests. If two butterflies continued interacting after the initial contact, bouts of pushing commonly followed until one of the butterflies departed from the core area, usually after a pursuit. When a chased butterfly escaped his pursuer within a core area even for a second or two, a complete new interaction seemed to arise when the two butterflies reencountered one another. After the end of an interaction, the “defeated” individual commonly returned to the core area after a few minutes—and on successive days—to engage in new contests with the resident.
DISCUSSION This study describes an unreported type of lek-based mating system found in the riodinid butterfly, Charis cadytis. In this species dominant males vie for probable high-quality mating sites using what we term “sumolike” pushing contests, whereas non-aggressive satellite males occupy scattered perches outside disputed core areas. Our conclusion that male aggregations of C. cadytis are leks is supported by observations on females mating at aggregation sites, which in turn suggest that receptive females come to these sites to mate. Although the larval biology of C. cadytis is unknown, other Charis species oviposit on leaf litter, and larvae may feed on dead leaves (DeVries, 1997, p. 172). In seasonally dry southeast Brazil, leks of C. cadytis may be strategically located in places near water where fallen leaves remain moist throughout the year. The defense of landmark encounter sites by dominant males, with peripheral areas relegated to satellite individuals, although not previously reported for butterflies, is consistent with mating strategy theory (Gross, 1996). Satellites are expected to be males physically incapable of holding territories that withdraw to undefended border zones where there is some chance of intercepting a receptive female. In the case of C. cadytis, newly recruiting males seem to either dispute territories or to join the ranks of satellites (G.W. Chaves, field notes). This suggests that satellites are not
192
Chaves, Patto, and Benson
merely older individuals. However, more observations are needed to clarify the significance of satellite behavior in this butterfly. Males that disputed core areas usually interacted repeatedly throughout a day, and on successive days, with contests broken only by temporary absences of ousted “loser” males. However, if only one dominant male was present at a lek, it gained sole use of the core area and defended it against butterflies encroaching from the periphery. We therefore consider core areas to be territories inserted within lekking arenas, with territory sites becoming intermittently shared over sometimes long intervals when two or more dominant males were present to dispute them. Two clearly stereotyped elements of the territorial interactions of Charis cadytis were frontal flights and sumo contests. DeVries (1997, p. 177) described a behavior similar to frontal flight for Charis iris Staud. in Costa Rica. Some large butterflies, such as Hypolimnas bolina Fabr. and Caligo idomenaeus Staud., interact in territorial contests while flying face on, but these circle about one another and may strike each other with their wings as they spin (Rutowski, 1992; Freitas et al., 1997). Contest dynamics suggests that frontal flights are related to mutual assessment by adversaries, whereas sumo interactions seem to comprise the contest phase of disputes in this butterfly. The frontal flight, by placing rival males facing one another for a few seconds in near hovering flight, potentially allows each to perceive relevant traits, such as size and coordination of its rival, which may provide a measure of the chances of success in a contest. The fact that 40% of all interactions began with and immediately terminated after a frontal flight, sometimes accompanied by a brief pursuit, suggests that an evaluation is made that prompts the withdrawal of one contestant. It seems likely that the narrow yellow bands on wingtips of male C. cadytis play some role in this behavior. The most exceptional behavior shown by C. cadytis is its sumo-like pushing contests. Previously, territorial disputes in butterflies were considered exclusively aerial, with rivals characteristically flying in acrobatic chases that include conspicuous spiral and circling maneuvers (Kemp and Wiklund, 2001). The complexity of the stereotyped encounters of C. cadytis described here goes beyond that known for butterflies that defend fixed resources, such as food and female pupae (Owen, 1971; Murawski and Gilbert, 1986; Deinert et al., 1994), while seated and constitutes a challenge to explain. Pushing contests, in which the force generated by wing muscles seems important, in addition to providing cues on aerial performance, may tire adversaries and interfere with mate interception. Although pursuits may be no more than simple following from one sumo event to another, assessment and contest roles cannot be discarded. The apparent absence of acrobatic chases in C. cadytis suggests that frontal flights and sumo contests
Non-Aerial Butterfly Contests
193
may generate both the information and costs necessary to settle disputes. Our observations on C. cadytis make it clear that contest structure and assessment cues vary greatly in butterflies and give support to Kemp’s and Wiklund’s (2001) suggestion that spiraling and circling are an alternative form of contest employed by many butterflies. The sumo contests of C. cadytis evoke two important evolutionary questions: What selective advantages led to its evolution, and how does its operation relate to contest resolution? We suggest that the evolution of perching during disputes stems from adaptive tactics used by intruders of inferior competitive ability. We hypothesize that trespassing males, by landing on leaves when approached by a resident, were able to benefit relative to those engaging in aerial contests, the presumed ancestral condition. We envision three possible benefits of intruder perching: (1) reduction in damaging collisions with each other and with fixed objects, (2) diminished expenditures in energetically costly acrobatic chases, and (3) increased time in a core area and hence greater chances of being present when a female arrives. The contests of C. cadytis seem to lack potentially harmful wing contact, and they are probably less demanding energetically than aerial disputes (Nation, 2002, p. 256). Even though the sudden flights at the end of sumo bouts may present risks, we did not observe any injuries. On the other hand, butterfly species that engage in escalated fighting on the ground may become severely injured during combat (Eff, 1962; Pinheiro, 1991, pers. com.; Kemp and Wiklund, 2001), and under these circumstances selection for ritualized behaviors to avoid self-injury may be especially strong (see Silberglied, 1984). Game-theory views contest success as a product of superior opponents imposing unacceptable costs on less able adversaries (Maynard Smith, 1982). However, the nature of these costs in butterfly territoriality is controversial (Kemp and Wiklund, 2001). As seen above, physical injury does not seem to constitute a contest cost in C. cadytis. Exhaustion due to the depletion of energy reserves also seems unimportant. Energetic wars-ofattrition, in which losers may become physically drained at the end of long contests (Forsyth and Montgomerie, 1987), are well documented in some damselflies (Marden and Waage, 1990). However, in C. cadytis the same pairs of rivals engaged in disputes many times over a day, suggesting that energy reserves may be adequate even for long contests and that separate interactions are terminated for reasons unrelated to absolute energy shortage. Moreover, the subdued interactions of C. cadytis do not seem designed to exhaust quickly the energy reserves of rivals. The infrequent flower visits by C. cadytis in the intervals between interactions also
194
Chaves, Patto, and Benson
seem inconsistent with the idea that males become energy starved during contests. Although the interactions of Charis cadytis seemed neither dangerous nor directly limited by stored energy, they did appear to cause distraction and fatigue, perhaps making males less attentive to passing females and external risks. If alertness and performance degrade during contests, the male that tires most may be better off if it leaves to rest than if it continues an interaction. The idea that physical fatigue may diminish competitive ability and provide a criterion for contest settlement in C. cadytis meshes well with the information available and may provide a viable alternative for mechanisms linked to physical injury and absolute energy shortage. Whatever the true costs may be, cost reduction obtained through perching seems to be an important element in the unusual contest behavior of this butterfly. ACKNOWLEDGMENTS We thank Philip DeVries and an anonymous referee for their encouragement and helpful comments on the manuscript. Keith S. Brown, Andre´ V. L. Freitas, Helena P. Romanowski, Ronaldo B. Francini and Marcio R. Pie provided useful discussions and suggestions during earlier phases of the work. GWC and CEGP were supported by graduate fellowships from the ´ Conselho Nacional de Pesquisas e Desenvolvimento Tecnologico (CNPq). WWB was supported during early phases of the work by a CNPq Research Fellowship. Lodging at the study site and authorization to conduct studies in the Environmental Protection Zone of the Serra do Jap´ı were provided by the Municipal Government of Jundia´ı REFERENCES Alcock, J. (1994). Alternative mate-locating tactics in Chlosyne californica (Lepidoptera, Nymphalidae). Ethology 97: 103–118. Alcock, J., and Houston, T. F. (1996). Mating systems and male size in Australian hylaeine bees (Hymenoptera: Colletidae). Ethology 102: 591–610. Alcock, J., and O’Neill, K. M. (1986). Density-dependent mating tactics in the Grey hairstreak, Strymon melinus (Lepidoptera: Lycaenidae). J. Zool. 209: 105–113. Andersson, M. (1994). Sexual Selection, Princeton University Press, New Jersey. Baker, R. R. (1972). Territorial behaviour of the nymphalid butterflies, Aglais urticae (L.) and Inachis io (L.). J. Anim. Ecol. 41: 453–469. Benson, W. W., Haddad, C. B., and Zikan, M. (1989). Territorial behavior and dominance in some heliconiine butterflies (Nymphalidae). J. Lepid. Soc. 43: 33–49. Bonduriansky, R., and Brooks, R. J. (1999). Why do male antler flies (Protopiophila litigata) fight? The role of male combat in the structure of mating aggregations on moose antlers. Ethol. Ecol. Evol. 11: 287–301. ´ Brown, K. S., Jr. (1972). Borboletas da Serra do Jap´ı: Diversidade, habitats, recursos alimenta˜ temporal. In Morellato, L. P. C. (ed.), Historia res e variac¸ao Natural da Serra do Jap´ı, ´ UNICAMP/FAPESP, Campinas, Brazil, pp. 142–187.
Non-Aerial Butterfly Contests
195
Cade, W. (1979). The evolution of alternative male reproductive strategies in field crickets. In Blum, M. S., and Blum, N. B. (eds.), Sexual Selection and Reproductive Competition in Insects, Academic Press, NY, pp. 343–379. Convey, P. (1989). Influences on the choice between territorial and satellite behaviour in male Libellula quadrimaculata Linn. (Odonata: Libellulidae). Behaviour 109: 125–141. Davies, N. B. (1978). Territorial defence in the speckled wood butterfly (Pararge aegeria): The resident always wins. Anim. Behav. 26: 138–147. Deinert, E. I., Longino, J. T., and Gilbert, L. E. (1994). Mate competition in butterflies. Nature 370: 23–24. DeVries, P. J. (1997). The Butterflies of Costa Rica and Their Natural History. Volume II: Riodinidae, Princeton University Press, New Jersey. Dodson, G. N. (1997). Resource defense mating system in antlered flies, Phytalmia spp. (Diptera: Tephritidae). Ann. Ent. Soc. Amer. 90: 496–604. Eff, D. (1962). A little about the little-known Papilio indra minori. J. Lepid. Soc. 16: 137– 143. Fischer, K., and Fiedler, K. (2001). Resource-based territoriality in the butterfly Lycaena hippothoe and environmentally induced behavioural shifts. Anim. Behav. 61: 723–732. Forsyth, A., and Montgomerie, R. D. (1987). Alternative reproductive tactics in the territorial damselfly Calopteryx maculata: Sneaking by older males. Behav. Ecol. Sociobiol. 21: 73– 81. Freitas, A. V. L., Benson, W. W., Marini-F◦ , O. J., and de Carvalho, R. M. (1997). Territoriality by the dawn’s early light: The Neotropical owl butterfly Caligo idomenaeus (Nymphalidae: Brassolinae). J. Res. Lepid. 34: 14–20. Gross, M. R. (1996). Alternative reproductive strategies and tactics: Diversity within sexes. Trends Ecol. Evol. 11: 92–98. ¨ Hoglund, J., and Alatalo, R. V. (1995). Leks, Princeton University Press, New Jersey. Johnsgard, P. A. (1994). Arena Birds: Sexual Selection and Behavior, Smithsonian Institution Press, Washington, DC. Kemp, D. J. (2003). Twilight fighting in the evening brown butterfly, Melanitis leda (L.) (Nymphalidae): Age and residency effects. Behav. Ecol. Sociobiol. 54: 7–13. Kemp, D. J., and Wiklund, C. (2001). Fighting without weaponry: A review of male-male contest competition in butterflies. Behav. Ecol. Sociobiol. 49: 429–442. Knapton, R. W. (1985). Lek structure and territoriality in the chryxus arctic butterfly, Oeneis chryxus (Satyridae). Behav. Ecol. Sociobiol. 17: 389–395. Lederhouse, R. C. (1982). Territorial defense and lek behavior of the black swallowtail butterfly, Papilio polyxenes. Behav. Ecol. Sociobiol. 10: 109–118. Lederhouse, R. C., Codella, S. G., Grossmueller, D. W., and Maccarone, A. D. (1992). Host plant-based territoriality in the white peacock butterfly, Anartia jatrophae (Lepidoptera: Nymphalidae). J. Insect Behav. 5: 721–728. Marden, J. H., and Waage, J. K. (1990). Escalated damselfly territorial contests are energetic wars of attrition. Anim. Behav. 39: 954–959. Maynard Smith, J. (1982). Evolution and the Theory of Games, Cambridge University Press, Cambridge, UK. Murawski, D. A., and Gilbert, L. E. (1986). Pollen flow in Psiguria warscewiczii: A comparison of Heliconius butterflies and hummingbirds. Oecologia 68: 161–167. Nation, J. L. (2002). Insect Physiology and Biochemistry, CRC Press, New York. Owen, D. F. (1971). Tropical Butterflies, Oxford University Press, Oxford, UK. Perrill, S. A., Gerhardt, H. C., and Daniel, R. (1978). Sexual parasitism in the green tree frog (Hyla cinerea). Science 200: 1179–1180. Pinheiro, C. E. G. (1991). Territorial hilltopping behavior of three swallowtail butterflies (Lepideoptera, Papilionidae) in western Brazil. J. Res. Lepid. 29: 134–142. Rutowski, R. L. (1983). Courtship behavior of the dainty sulfur butterfly, Nathalis iole with a description of a new, facultative male display (Pieridae). J. Res. Lepid. 20: 161–169. Rutowski, R. L. (1992). Male mate-locating behavior in the common eggfly, Hypolimnas bolina (Nymphalidae). J. Lepid. Soc. 46: 24–38.
196
Chaves, Patto, and Benson
Rutowski, R. L., and Gilchrist, G.W. (1988). Male mate-locating behavior in the desert hackberry butterfly, Asterocampa leilia (Nymphalidae). J. Res. Lepid. 26: 1–12. Silberglied, R. E. (1984). Visual communication and sexual selection among butterflies. In Vane-Wright, R. I., and Ackery, P. R. (eds.), The Biology of Butterflies, Academic Press, London, pp. 207–223. Taborsky, M. (1994). Sneakers, satellites, and helpers: Parasitic and cooperative behavior in fish reproduction. Advances in the Study of Behavior 23: 1–100. Thornhill, R., and Alcock, J. (1983). The Evolution of Insect Mating Systems, Harvard University Press, Massachusetts. Tinbergen, N., Meeuse, B. J. D., Boerema, L. K., and Vorossieau, W. W. (1942). Die Balz des Samtfalters, Eumenis (= Satyrus) semele. Z. Tierpsychol. 5: 182–226. Wallace, A. R. (1853). On the habits of the butterflies of the Amazon valley. Trans. Entomol. Soc. Lond. 1854: 253–264. Wickman, P., and Wiklund, C. (1983). Territorial defence and its seasonal decline in the speckled wood butterfly (Pararge aegeria). Anim. Behav. 31: 1206–1216. Wilkinson, L. (1990). SYSTAT: The System for Statistics, SYSTAT Corp., Illinois. Zar, J. H. (1996). Biostatistical Analysis, Prentice Hall, New Jersey.