Pyragra nigrescens Brindle 1977

Pyragra nigrescens Brindle

This species was collected under loose bark. The male cerci are very similar to those of the females (Figure 68).

We taped 13 high-level and 26 low-level aggressive interactions. Because cercus morphology was virtually identical in males and females, some battles may have involved females that we could not distinguish. Intense interactions began when antennal contact was followed by a very rapid 90° or 180° turn and a lateral slam or series of slams (Figure 69). On several occasions the tips or lateral surfaces of the cerci struck the opponent as this turn was executed, and, on two occasions, the opponent was displaced by such an impact (Figure 70).

Some slams involved twisting the abdomen, and, in some cases, a strike resulted in a pinch of the opponent's abdomen (Figure 71-4,5). On five occasions, one individual pinched the opponent and then lifted it in the air and shook it briefly (Figure 72). One pinched individual apparently wrenched itself free by twisting the other's abdomen (Figure 71) as in Anisolabis and Carcinophora. Sustained simultaneous pinches did not occur.


Mutually aggressive individuals both turned, each backed rapidly toward the other, their cerci interlaced, and each gave brief pinches to the other on the distal and middle portions of the other's abdomen (Figure 73-1). Often they also pushed briefly and slammed laterally; the points and lateral margins of the cerci contacted the opponent as they struggled (Figure 73-5,6,7). In one case in which confounding stimuli from other individuals were not present, it appeared that an individual “wound up” to deliver a forceful lateral slam by bending its abdomen slightly in one direction before slamming in the other (Figure 74).



When one individual had won a battle, it often chased the loser, striking at it with its cerci (Figure 75). Occasionally strong rapid vibrations of the body were seen in aggressive contexts while the two individuals were out of contact.

C. COURTSHIP AND COPULATION

Pyragra nigrescens

We taped five courtships, none of which led to copulation. In at least two interactions an apparent male rubbed with the basal portion of his cerci or the sides of his cerci on the abdomen of the other individual (Figure 96). Prior to and during rubbing, the apparent male sometimes also vibrated his body and antennae rapidly in apparent courtship movements.

Discussion

A. FUNCTIONAL MORPHOLOGY OF MALE CERCI AND PYGIDIA

Animals generally failed to respond to each other until they made physical contact, suggesting that visual stimuli are of little importance in either aggression or courtship. The generally claustral and/or nocturnal habits of most species reinforce this conclusion. We will therefore attempt to deduce possible functions of cerci and associated structures on the basis of the ways in which they contacted other earwigs during our observations.

Before beginning a detailed discussion of possible functions, it should be noted that in only two species (Doru taeniatum and Vostox quadripunctatus) was male-male aggression observed in the field, and only in Doru was courtship seen in the field. Our observations thus give neither the biological context of fights (e.g., near females, food, or a refuge) nor the physical sites where fights and courtship occur (e.g., in closed tunnels, small chambers, open sites, or on vertical trunks). The kinds of behavior used in fights clearly were influenced by the physical setting (tunnels vs. open petri dishes) in those species in which we made observations in both settings (Doru, Vostox, Paralabella, Anisolabis). In order to permit interpretations of probable functions, we made the assumption that the behavior patterns of males in our observation chambers are similar enough to those that occur in nature. Although this assumption is justified in the one species we observed in detail in nature (Doru), further studies are needed on other species.

This limitation is less severe in the species found in relatively flat, enclosed spaces (Sparatta, Paralabella, to a lesser extent Vostox). In Sparatta, the flattened body design, the lack of abdomen twisting or dorsal flexion associated with slams and strikes, and the attempts of males to turn themselves upside down both when fighting and when courting lead us to believe that individuals of this species normally interact in spaces that are dorsoventrally compressed. The same is probably true of Paralabella, in which the body also is relatively compressed dorsoventrally. The acceptance posture of females of this species, however, is only feasible where there is some space dorsally. On the other hand, both the morphology and behavior of Metrasura, Skalistes, and Ancistrogaster suggest that they must fight in relatively open spaces. Our field observations are in accord with this, as all three were found in open sites at night.

Some general preliminary statements can be made regarding the functional significance of the forms of some structures. The males of many species have cerci whose internal margins have teeth and other irregularities, especially near the base of the cercus. Males of all species pinch opponents with their cerci at least occasionally. It seems probable that the teeth serve to increase the friction against the surface of the opponent and/or to produce discomfort. The small teeth of Doru are highly corrugated (Figure 97), probably facilitating the proposed friction function. There is a seta on the posterior side of each small tooth of Doru (Figure 97). These setae probably are deflected when the cercal tooth touches another object, and they may play an important role in sensing contact with opponents during cercal tapping and squeezing. Only in Paralabella dorsalis did courtship behavior clearly suggest that a prominence on the inner margin of the cerci was used in interactions with females.

The largest cercal teeth usually occur near the middle of the cercus, as in Metrasura, Sparatta, and Ancistrogaster, or near the tip, as in Doru. The reasons for these differences are not clear. Larger teeth probably allow the male to give more forceful pinches to their opponents than they can with the tips of the cerci, because the teeth reduce the effective length of the cercus.

Table 1 summarizes the possible functional significance of different male-specific cercus and pygidium morphologies. Species-by-species discussions follow that explain how we reached the conclusions in the table.

1. Doru taeniatum

In high-level interactions, opponents usually were defeated by pushing and lifting movements. Thus, details of male-specific cercus and pygidium morphology seem not to be especially useful. Occasionally movements associated with a pinch were used; in at least some pinches the tooth on the inner edge of the cercus probably served to grip the opponent (Figure 12). In others, the length of the cercus may have provided increased mechanical effectiveness of the male's lifting or slamming attack (e.g., Figures 10, 12). On the other hand, many (most) slams did not involve pinches, and the force on the opponent apparently was exerted by the pygidium (e.g., Figure 9), so there were many fights in which male-specific cercus morphology did not seem to confer mechanical advantages.

The pointed pygidium probably does not function as an effective penetrating weapon. It undoubtedly struck the opponent in many backward strikes, but the posterior portion of the male abdomen is heavily sclerotized, and there are no pits or grooves into which an opponent's pygidium could fit so as to produce a tighter mesh with the opponent. Perhaps the point functions as a mechanism for increasing contact in dorsal slams or to intensify the stimuli associated with backward strikes, and thus it functions as a threat device.

Tapping behavior always occurred before and usually during the time cerci were meshed. Although tapping was probably partly exploratory in function, it also may serve as threat behavior. Some interactions ended with one male fleeing after being tapped. During tapping, cercal length, cercal teeth, and the pygidium presumably could all have been sensed by the opponent as they touched him. Thus, these male-specific characters may function as threat devices.

Possible courtship functions for some male-specific cercal characters also are possible. Although males usually tapped females with only the ventral, proximal portions of their cerci, females sometimes mouthed male cerci and sometimes pinched them with their own cerci. Female pinching of male cerci probably, however, was a signal to the male that the female was ready to copulate, as it usually was followed immediately by tilting of the abdomen by the female to facilitate intromission. The male pygidium was not contacted by females and played no apparent role in male-female interactions.

In sum, we cannot confidently assign functions to several male-specific characters; at least some may function as threat devices.

2. Skalistes inopinata

Two male-specific characters that seem to be involved in male-male aggression rather than courtship are the dorsal triangular projections at the bases of the cerci and the strongly curved, smooth, tong-like distal portions of the cerci. The dorsal projections appear to be used in threat displays, and the tongs probably serve as weapons in fights. The male abdomen usually is turned and moved so that the dorsal projections tap on the opponent during low- and moderate-level male-male aggression. The general lack of hard, dorsally directed slamming movements, the rarity of forceful movements in which the projections were brought to bear on the opponent's body, and the lack of any other mechanical contact or meshing of the projections during pinching movements all indicate that the projections are not weapons used to gain physical advantage.

The tong-like form of the distal portion of the cercus probably allows the cerci to fit snugly around the opponent's body so that he can be turned upside down when the attacker's abdomen is twisted.

Most courtship is performed by tapping with the ventral, basal portions of the cerci. The female is not in position to sense either of these male-specific characters during such taps.

3. Metrasura ruficeps

The cerci of male Metrasura appear to be used as weapons in fights in a number of different ways. Opponents were thrown after being pinched (Figure 19), and they also were twisted or pried away from the substrate (e.g., Figures 21, 22). Both twisting and shaking often were combined with prying movements (in various combinations, e.g., Figures 22, 23) in apparent attempts to weaken or break the opponent's foothold. The extreme length of the cerci in this species probably increases the effectiveness of these fighting tactics.

In contrast to species like Carcinophora spp., pinches per se played little role in the fights of Metrasura males. Pinches were rare, and those that occurred usually were short and combined with prying and/or twisting. Presumably the tooth on the inner margin of the male cerci serves to increase the male's ability to hold opponents when they are seized and pried (Figure 18) and when they are thrown (Figure 19).

Thus, the extreme length of the male cerci in this species seems to serve to make the cerci more effective as levers. The cerci are used in combination with blows from the abdomen to dislodge the opponent or even fling him. The male cerci have thus largely lost the primitive function of delivering forceful pinches, a function for which their great length makes them unsuited.

It also is possible that the extensive cercus-cercus and cercus-abdomen contact that always occurred while males were in intense pygidium-to-pygidium fights functions as threat behavior. If this is true (we see no compelling reason to either accept or reject the possibility), then the extreme length of male cerci, which repeatedly touch the opponent during such interactions, also could function to increase the effectiveness of the threats.

In contrast to these uses in male-male aggression, the basal and ventral rather than the distal portions of the male cerci generally were used to tap on females during courtship, so the long length of the cerci, the teeth on their inner surfaces, and pygidium characteristics are unlikely to be important in courtship.

4. Ancistrogaster scabilosa

The lateral extensions of the male's posterior abdominal tergites did not come into contact with the female during courtship, but they apparently did make contact with the extensions of other males during middle-level fights in which both males raised their abdomens (Figures 30, 31). Because the extensions gave no obvious mechanical advantage, we suspect that they are threat or assessment devices, analogous to the lateral prolongations of the heads of some flies (e.g., McAlpine, 1979; Dodson, 1989).

The greater length of male cerci and their wider separation at the base may aid in fights by permitting the male to pinch at a greater distance and more forcefully. The presence of a large tooth near the tip of the cercus is in accord with the idea that the cerci are used to pinch in this way (see discussion of teeth above). Both of these characters also could function as threat devices. Neither appeared to be involved in courtship, as only more basal portions of the male's cerci touched the female.

5. Sparatta bolivari

Understanding the significance of male-specific characters on the cerci and pygidium of this species is difficult, perhaps due to the small number of high-intensity fights we observed. Possibly the lateral prongs of the longer male pygidium function to make it easier for males to pinch the cerci of opponents (an otherwise unusual behavior that was common in this species; Figure 35), but such pinches did not seem to influence the outcomes of battles. The pygidia of fighting males sometimes meshed (Figure 36), but again the significance for the outcomes was unclear. It also was unclear whether the form of the male pygidium could be sensed by the female during courtship (Figure 84). Perhaps the pygidium and the somewhat longer cerci of males both function as threat devices.

6. Vostox quadripunctatus

The only male-female interaction that led to copulation resembled mid-level male-male aggression, so it is not possible to distinguish between the possible effects of courtship and aggression on male cercus morphology. The marked difference with the courtship behavior of V. apicedentatus, in which the male taps, rubs, and presses the female with his cerci (Moore and Wilson, 1993), suggests that our observations may have been of aberrant behavior.

Cerci and pygidia were used in three different ways in male-male fights: to deliver hard lateral and sometimes partially dorsal slams that could knock the opponent off his feet; to push with the pygidium against the opponent's pygidium or the base of one of his cerci so as to push him away (pushing of this sort seemed to determine winners in many pairings); and, more rarely, to pinch him with the entire length of the cercus.

The wide space between the bases of male cerci, the outward curve of the cercus at the base, and the prominent pygidium may all function to bring a male's pygidial area to bear on his opponent in pushing contests. Assuming that fights in tunnels were typical with respect to body parts that are brought together in pushing attempts in other contexts, the exact site on the pygidium that contacted the opponent during battles varied, as did the site on the opponent where contact was made. Thus, there appears to be no precise mesh during pushing battles; the function of the rectangular form of the pygidium is not clear.

We can offer only even more tentative speculations on the functional significance of secondary sexual differences in other details of male cercus morphology. The inward curve at the tip and the point on the tip may serve to make pinches more painful and/or less likely to slip on the opponent's body. The straight middle portion and its relatively thick diameter may increase the effectiveness of the cercus as a rod with which to deliver slamming blows and/or to pry or hold the opponent's abdomen so that backward pushes will be more effective. Any or all of these characteristics could be sensed by opposing males during pushing fights or the preceding interactions, and thus they also could serve as threat devices.

The aggressive behavior of a second species of Vostox, V. apicedentatus, is quite different (Moore and Wilson, 1993) (their terms are denoted parenthetically). Pushing contests apparently do not occur, whereas pinching (pinch and grab) is more common. Slams (bats) involve lifting the abdomen dorsally and then striking ventrally with the cerci, rather than swinging them laterally. We did not observe several types of behavior seen in V. apicedentatus: pressing down on the opponent with the cerci (press); prying the opponent up with the cerci (lift); and biting the opponent with the mouthparts (bite). Several of the differences involve greater use of dorsoventral movements by V. apicedentatus. This pattern may be related to the habitat of this species (in and around large rotting cacti), which is probably less dorsoventrally compressed than that of V. quadripunctatus (under loose bark).

7. Paralabella dorsalis

The blunt teeth near the bases of male cerci were not consistently brought into play during male-male fights. Their design did not seem appropriate to aid males mechanically in any of the contexts observed in male battles. On the other hand, the proximal edges of the teeth appeared to mesh with the dorsal surfaces of the female's cerci, forming a clamp when the male seized the female just prior to intromission. Thus, these teeth seem likely to function in interactions with females rather than with males. Because the male seizes the female with his cerci only after an extended courtship, and after the female has raised her abdomen to an otherwise unusual acceptance position, it is unlikely that the teeth function as devices with which the male physically forces the female to copulate. Instead, they probably are analogous to the nongenitalic holding and clasping devices of the males of many other animals (Eberhard, 1985), which probably function as contact courtship devices.

Male pygidia made extensive contact with opponents in battles. The breadth of the pygidium could increase its effectiveness as a defensive shield against the opponent's pinching attacks (Figure 48), but this interpretation is speculative. The greater inward curvature of male cerci probably is related to the wide separation at their bases associated with the wide pygidium. Greater curvature allows the cerci to grasp and squeeze forcefully on the body of an opponent during fights (Figure 49) despite their wide basal separation. We are unable to explain why male pygidia have their particular rounded form.

8. Pseudomarava prominensis

The wide, flattened pygidium and the wide curve of the smooth male cerci make it possible for the male to grasp and hold another individual's abdomen. Other individuals’ abdomens are squeezed in this way just prior to copulation (Figure 90) and during male-male fights (Figure 53). In contrast to Paralabella dorsalis, the female often was seized while apparently resisting the male. Thus, the male design may function to force the female to copulate, rather than to court her as in P. dorsalis. We cannot determine if the design to grasp another individual arose in the context of male-male fights and was incidentally useful in male-female interactions, or vice versa.

The clearest use of cerci in fights between males was to pinch, and the sharp points of male cerci may make pinches more painful and/or more tenacious.

9–12. Anisolabis maritima and Carcinophora spp.

The short length and inward curvature of one cercus make the cerci mechanically strong and able to produce strong pressure when they pinch. The behavior of these species supports the idea that these male features evolved as weapons. Males often began fights by immediately striking and attempting to pinch. These were the only genera in which fights were seen that resulted in physical damage to males (one puncture wound in each genus). Male courtship behavior involved tapping with the ventral surfaces of the cerci, not pinching or clamping.

13. Pyragra nigrescens

The lack of clear sexual dimorphism in the cerci of this species is associated with relatively simple fighting behavior, which included quick pinches and slams.

B. EVOLUTION OF AGGRESSIVE BEHAVIOR

Male-male aggressive behavior generally varied more between species than did courtship. In all species, however, intense interactions included pinches. Females and nymphs generally use their cerci to pinch defensively (see “Introduction”). Their relatively simple cercus designs resemble more closely the cerci of fossil Dermaptera (Carpenter, 1992) and other orthopteroid insects. Thus, pinching in male-male battles probably represents the retention of a primitive trait.

We documented a number of additional, probably derived aggressive behavior patterns. Some of these probably have their evolutionary origins in behavior designed for other purposes. For instance, both turning to direct the cerci toward a disturbing stimulus and swinging the abdomen from side to side in an exploratory fashion while backing up are widespread and probably ancient behaviors. The strong lateral slams, which were especially frequent in Vostox, Sparatta, and Pyragra, and the strong pushes, such as those of Doru and Paralabella, probably derive from these behavior patterns. Because they are executed so rapidly and energetically, they acquire the added function of inflicting unpleasant stimuli on an opposing male. Prying the opponent (Metrasura), flipping him over (Skalistes), and raising the abdomen to tap him (Ancistrogaster) may represent further modifications of pushing behavior. Only in Pseudomarava and perhaps Sparatta were the male's other obvious potential weapons, his mouthparts, used in aggression.

Males of Pseudomarava employed aggressive behavior in an additional, presumably derived context. By biting the legs and antennae of females, they manipulated them into situations in which they could bring their genitalia into position for copulation. When the female bent her cerci toward a male to defend herself against the bite, the male seized her abdomen with his tong-shape cerci (Figure 90) and thus achieved intromission.

Homosexual courtship was observed several times in each of three forficulid species and in one labiid. The males’ lack of contact with other earwigs for at least one week before our observations may have caused them to make an increased number of behavioral mistakes. However, in at least two species (Skalistes and Metrasura) the male's behavior immediately before or after he “courted” another male was sometimes clearly aggressive, suggesting the possibility that homosexual courtship may actually be an aggressive display. Apparently aggressive use of male courtship behavior (or the opposite, i.e., use of aggresive behavior as courtship) occurs frequently in natural situations in a long-horned orthopteran (Field and Sandlant, 1983). Homosexual courtship also is known in captive orthopterans (Alexander, 1964; Steinberg and Willey, 1983) and (in the presence of female odor) in bees (O'Neill and Bjostad, 1987). The possibility that males of Skalistes use their genitalia as weapons in intraspecific battles, a very unusual trait (Eberhard, 1985), merits further study.

C. GENERAL COMMENTS

Of the seven male-specific morphological characters to which we have assigned functions with relative confidence, three are weapons in male-male battles, two are threat devices in these battles, one is a courtship device, and one is a device to force copulations on females. Of the 14 assignments with somewhat less confidence (“yes?” in Table 1), all are weapons. The predominant pattern seems to be for these features to function as weapons, as do horns in beetles (Eberhard, 1979). This conclusion must be tentative, however, even for the species described here, because more subtle uses, such as threat devices, are very difficult to document. There are many other designs of earwig cerci (Figure 1) that were not included in our sample, and future studies may find different trends.

Similarly, our level of comparison was not fine enough to attempt explanations of more subtle differences in cercus and pygidium design in closely related species. For instance, we can hardly even speculate on the reasons for the differences in relative pygidium length and bluntness or the minor differences in patterns of teeth on the inner edges of cerci or cercus shape commonly seen in different species of a genus (e.g., Doru; Figure 98).

Although it is possible that hardened cerci first evolved in Dermaptera when they were used by males to function in sexual contexts, and were only later acquired by females and immatures for defense, we suspect that the opposite sequence occurred. Probably hardened cerci and the ability to close them forcefully originally arose in earwigs as a defense against predators, because males, females, and nymphs of all the species we observed, as well as many others (see “Introduction”), pinch with their cerci when disturbed. Presumably early male cerci were similar in form and size to those of females and nymphs (e.g., females in Figures 7, 16, 26, 40, 51, 57, 63). Probably males began to use their cerci in different contexts associated with competition for sexual access to females. Given that, as we have observed, cerci can inflict wounds and combat can result in dislodgement, there would have been selection on males to judge the fighting capacities of opponents before entering into serious combat. This situation would create selection favoring signalling devices best able to intimidate rivals, and it could produce rapid evolution in the signalling characteristics (West-Eberhard, 1983), involving both the cerci themselves (for instance, the horns of Skalistes and possibly the extreme elongation of Metrasura) and other parts of the body (such as, lateral abdominal projections in Ancistrogaster and the vibrations of the entire body in Pseudomarava). In addition, battles using cerci would create selection favoring cercal structures improving the male's ability to fight (for instance, the teeth on the inner margins of several species). Some or all weapons also may function as signalling devices (e.g., the teeth of Doru could function during cercal tapping as do the knobs on the claws of male fiddler crabs, Crane, 1975), but our level of analysis was not adequate to demonstrate this.