Ó Springer-Verlag 1999 Behav Ecol Sociobiol (1999) 45: 87±93 ORIGINAL ARTICLE George W. Uetz á Elizabeth I. Smith Asymmetry in a visual signaling character and sexual selection in a wolf spider Received: 26 February 1998 / Accepted after revision: 12 September 1998 Abstract Fluctuating asymmetry (FA) can indicate developmental instability in bilaterally symmetric organisms, and studies have shown that the degree of asymmetry in male secondary sexual characters in¯uences female mate choice in a number of taxa. In male Schizocosa ocreata wolf spiders, conspicuous tufts of bristles on the forelegs are a critical component of visual courtship signals, which appear to play a role in female mate choice. Previous studies have shown that females exhibit reduced receptivity to males with regenerative asymmetry, a consequence of leg loss and regeneration that causes males to be grossly asymmetric with respect to this important signaling character. We provide data on the occurrence of FA in the tufts of S. ocreata, and examine further the in¯uence of asymmetry on female mate choice. The distribution of tuft area asymmetry values from a sample of ®eld-collected males was normal, with a mean value of zero, indicating true FA. For a subset of males measured directly after ®eld collection and prior to feeding, tuft asymmetry was signi®cantly negatively correlated with measures of body size (body length) and condition (abdomen volume/cephalothorax width). Receptivity responses of females to visual signals from live males of similar size varied with the degree of asymmetry in male tufts. Since FA covaries with male body size and condition, which may also in¯uence behavioral vigor, we used video image manipulation to alter the degree of asymmetry in tufts of a courting male while holding size and condition constant. Asymmetry treatments represented values within the range of natural FA variation as well as more extreme values characteristic of regenerative asymmetry. With the confounding eects of male size, condition, and behavior held constant, female spiders exhibited reduced G.W. Uetz (&) á E.I. Smith Department of Biological Sciences University of Cincinnati. Cincinnati, OH 45221-0006, USA e-mail: uetzgw@email.uc.edu Tel.: +1-513-5569752, Fax: +1-513-5565299 receptivity responses to all experimental asymmetric video images relative to a control video stimulus. There were no dierences in the frequency of female receptivity among the various asymmetry treatments, suggesting that discrimination against asymmetry in conspeci®c male signal characters is not simply a rejection of extreme phenotypes. Results suggest that asymmetry in a key male secondary character used in visual signaling, independent of any concomitant behavioral or size factor, is an important criterion in mate choice. Key words Fluctuating asymmetry á Sexual selection á Female choice á Wolf spider Introduction Female choice based on symmetry of male characters has been an important and controversial topic in sexual selection research, because ¯uctuating asymmetry (FA) in bilaterally symmetric organisms can indicate developmental instability (Ludwig 1932; Palmer and Strobeck 1986; Leary and Allendorf 1989; Mùller 1990, 1997; Thornhill 1992; Watson and Thornhill 1994; Hill 1995; Mùller and Swaddle 1997). Numerous studies in the ®eld and laboratory have demonstrated that asymmetry plays a role in female mate choice, and that asymmetry in male characters is a good predictor of male ``quality'' and reproductive success (Mùller 1992a,b, 1993, 1997; Swaddle and Cuthill 1994; Swaddle 1996, but see Dufour and Weatherhead 1998a,b; Jennions 1998). This research area remains controversial, as measurements of FA often indicate very slight dierences among males, raising questions about detectability and biological relevance (Pomory 1997; Swaddle and Cuthill 1997; Gangestad and Thornhill 1998). Moreover, FA may covary with size and behavior, confounding interpretation and posing challenges to experimental study (Palmer 1996). Males of the brush-legged wolf spider, Schizocosa ocreata (Hentz), have conspicuous tufts of dark bristles 88 on the tibiae of the ®rst pair of legs, which are used in courtship and agonistic signaling (Aspey 1976, 1977; Stratton and Uetz 1981, 1983, 1986; Scheer et al. 1996). Previous studies with live males and video images have shown that these decorations function in species recognition and female mate choice (Scheer et al. 1996; McClintock and Uetz 1996; Uetz et al. 1996). Male S. ocreata exhibit two forms of asymmetry in their tufts: (a) regenerative asymmetry (RA), i.e., gross asymmetry arising from a missing or reduced tuft as a consequence of prior foreleg autotomy and regeneration (Uetz et al. 1996) and (b) FA, i.e., small random deviations from perfect left-right symmetry arising from developmental instability (Markow 1995; Palmer 1996). Females exhibit lower receptivity to males with RA (Uetz et al. 1996), but female receptivity in response to FA in male Schizocosa has not yet been examined. Video imaging in studies of animal behavior has shown considerable promise as an experimental tool for controlling covariates (Clark and Uetz 1992, 1993; Evans and Marler 1992; McClintock and Uetz 1996; Rosenthal et al. 1996; Uetz et al. 1996; Kodric-Brown and Nicoletto 1997). While not all species respond to video images, it is apparent that for spider species with higher visual acuity (including the Salticidae and Lycosidae), video stimuli are reasonable substitutes for live animals, and allow critical experimental manipulations in studies of species recognition and mate choice (Clark and Uetz 1991, 1992, 1993; McClintock and Uetz 1996). Because FA may covary with body size, tuft size and/or behavioral vigor (which might also in¯uence female mate choice), in this study we use video image manipulation and playback to test the hypothesis that female wolf spiders recognize asymmetry and discriminate against asymmetric males. Methods Spiders used in behavioral studies were collected as immatures from their deciduous forest litter habitat at the Cincinnati Nature Center, near Milford, Clermont County, Ohio, in April and May 1995, then brought to the laboratory and reared to maturity under controlled conditions. A second set of adult males was collected for measurement of tuft asymmetry later in the breeding season (early June) in 1995. Laboratory conditions included a 13:11 light:dark cycle, temperature range of 25±27 °C, and 60±75% relative humidity. Spiders were housed individually in opaque ``deli dish'' plastic food containers (12 cm diameter, 4 cm height) with clear lids. Spiders were fed three to ®ve newly hatched ``pinhead'' crickets (2±3 mm body length) twice a week and provided with water ad libitum. Measurement of asymmetry in male tufts Adult male spiders collected from the ®eld site (n = 47) were brought into the laboratory and maintained under conditions described previously. Males were anesthetized with CO2, and placed under a stereomicroscope for measurement of tuft asymmetry and other body dimensions. One of us (G.U.) measured the length and height of each tuft three times with an ocular micrometer, calculated tuft area as their product, then used the average area for each tuft in estimation of asymmetry (L±R). This was done to compensate for potential measurement error arising from dierential repeatability of measures (tibia length is highly repeatable and low in measurement error, while tuft height varies more along the length of the tibia and is subject to some error depending on where it is measured). The distribution of signed values of male tuft area asymmetry was tested for normality about a mean of zero, as predicted for FA (Palmer 1996). In addition to tuft area, measures of cephalothorax width (CW), body length (BL), abdomen width (AW) and abdomen length (AL) were made for a subset of ®eld-collected males (n = 16) prior to laboratory feeding (to control for dierences in body condition as a consequence of laboratory maintenance). A commonly used body condition index was calculated for these males as the ratio of abdomen volume to CW (see Jakob et al. 1996). Data on asymmetry in this subset of males was analyzed by regression and correlation with the above variables (Gangestad and Thornhill 1998). Response of females to asymmetry of live males We tested the hypothesis that female receptivity responses to male visual cues vary with the degree of tuft symmetry, using live males as stimuli. Selected male S. ocreata similar in body size (body length 7.0±7.6 mm) were paired randomly with conspeci®c females (n = 33 pairs) in an apparatus used in previous studies (Scheer et al. 1996; Uetz et al. 1996). The distribution of FA values (see above) was used to designate categories of degrees of asymmetry corresponding (approximately) to thirds of the range: low (< 0.4 mm2), intermediate (0.4±0.7 mm2) or high (> 0.7 mm2) for live males used in behavioral studies. Females were virgin individuals selected at random from the available pool, of approximately the same age (to control for dierences in behavior on account of reproductive state). Observations of spider behavior were made in adjacent clear plastic boxes (7 cm wide ´ 13 cm long ´ 7 cm high) about 5 mm apart, but on separate foam blocks to control for vibratory communication. Male courtship behavior was stimulated by placing a piece of paper with female silk in the box. An opaque paper barrier between boxes was removed when the male began courting. Trials lasted 10 min, and were videotaped for later analysis and scored blind. Data on female receptivity were analyzed using a log-linear contingency test of independence based on categories of the degree of male tuft asymmetry. Video image manipulation and playback studies We examined further the response of females to asymmetry in male tufts with video playback studies. We used video imaging because it allowed manipulation of the tufts while controlling for behavior and size dierences among males. Some spiders with acute vision (e.g., Salticidae, Lycosidae) respond to video images as if they are real (i.e., when presented with televised images of prey insects, conspeci®c and heterospeci®c spiders they behave appropriately; Clark and Uetz, 1991, 1992, 1993; McClintock and Uetz 1996; Uetz et al. 1996; Persons and Uetz 1997). Since the visual component of male courtship is by itself sucient to elicit receptivity in female S. ocreata (Scheer et al. 1996), moving video images of male courtship behavior are a reasonable substitute for live male stimuli (McClintock and Uetz 1996; Uetz et al. 1996). This test has the advantage that only the asymmetry of the tufts as a visual signal is manipulated ± all other aspects of male body size, morphology, and behavior are identical between treatments. A courting male S. ocreata was videotaped with standardized background and illumination, and images from this videotape were digitized, modi®ed, and animated as described in previous studies (Clark and Uetz 1991, 1992, 1993; McClintock and Uetz 1996; Uetz et al. 1996). The control video was an unmodi®ed courting S. ocreata male with tuft size close to the population median and fairly symmetric tufts (``low'' FA as above; L±R = 0.24). The four experimentally modi®ed video stimuli derived from this video had dierent degrees of asymmetry (Fig. 1), and were created to ad- 89 behaviors indicate that it has seen the other spider; direct approach toward the stimulus, receptive body postures and leg displays are obvious and unequivocal responses). Female display behaviors noted in previous studies (Montgomery 1903; Uetz and Denterlein 1979; Stratton and Uetz 1983) were used to determine potential receptivity as in previous studies. Just prior to copulation, females orient to males and assume a body position that allows males to mount (slow turn, settle, tandem leg extension, or lowering the cephalothorax while raising the abdomen). Copulation will usually not occur unless the female indicates receptivity by displaying one or more of these behaviors (Scheer et al. 1996; K. Delaney and G.W. Uetz, unpublished data). Results Asymmetry in male tufts Fig. 1 Video stimuli used in this study: control control male Schizocosa ocreata; removed one tuft was removed completely; reduced one tuft was reduced in size by approximately 25%; balanced both tufts were modi®ed to retain the total on-screen area of the spider's image, i.e., one tuft was reduced by approximately 25% and the other increased by the same amount; enlarged one tuft was increased in size by approximately 25%. The individual frames selected for display here are from a lateral view; the courting male on video may be viewed from various angles by the female dress dierent aspects of asymmetry manipulation: (1) removed: one tuft was removed, representing the most extreme level of FA or RA (L±R = 2.4); (2) reduced: one tuft was reduced in height such that the overall area was decreased by 25%, representing a midpoint within the range of natural FA variation (L±R = 0.9); (3) enlarged: one tuft was increased in size by approximately 25% to counteract any in¯uence of reduced tuft size (L±R = 0.9), and could also be interpreted as a male with larger, but asymmetric tufts; (4) balanced: both tufts were modi®ed to retain the total onscreen pixel area of the spider's image ± one tuft was reduced by approximately 25% while the other was increased by the same amount (L±R = 1.8). While the ``removed'' and ``balanced'' treatments may represent asymmetry levels closer to extremes of natural variation, the ``reduced'' and ``enlarged'' manipulations were well within the range of natural variation of the trait (based on the distribution of values presented below). Videos were used to test the null hypothesis of independence of female receptivity and asymmetry of the male stimulus. Videotaped images were presented to females on a Sony Watchman liquid crystal display (LCD) color microtelevision unit (2 inch diag. screen) in an arena used in previous studies (McClintock and Uetz 1996; Uetz et al. 1996). Female spiders used as subjects in video experiments were virgins selected at random from the pool of spiders collected in the ®eld (n = 143). Female spiders were used only once [a sample of 15±20 female spiders was used for each stimulus tested; unresponsive spiders (those not indicating any response to the stimulus) were eliminated from the analysis]. Female S. ocreata were presented with one of several digitized and re-animated video stimuli from a courting male S. ocreata ± either the control (symmetric) video, or one with tufts modi®ed to be asymmetric (as above). Trials lasted 10 min, during which female behaviors were videotaped for later analysis. After each trial, arenas were cleaned with an alcohol-dipped cotton swab to remove all silk and traces of pheromones from the previous occupant. Behavioral responses of female spiders were scored using qualitative measures of response (presence/absence of orientation, approach, and receptivity display behaviors). As the behavior of these wolf spiders is highly stereotyped, responses are easily scored (e.g., if a female swivels and orients to the male or the screen, those The distribution of tuft area asymmetry values from a sample of ®eld-collected males (Fig. 2) was distributed normally (Wilk-Shapiro Rankit statistic = 0.9477, P > 0.50) about a mean value of zero (one sample ttest, t = )1.18, P > 0.2), as predicted for ``true'' ¯uctuating asymmetry (Palmer and Strobeck 1986; Palmer 1996). As it was impossible to determine repeatability of FA measures for these samples, we used the post hoc method suggested by BjoÈrklund and MerilaÈ (1997), based on coecients of variation (1.18 for the sample of 47 males above, and 0.558 for the sample of 16 males described below). The results suggest a potential for FA measurement error of +55% to ±27%, respectively, less than many published values. A subset of males was measured directly after ®eld collection and prior to feeding, to assess body condition under ®eld conditions during the breeding season. These measures (untransformed) were used in correlative analyses as suggested by Gangestad and Thornhill (1998). In ®eld-collected males, tuft asymmetry was signi®cantly negatively correlated with both body size (BL, r = )0.673, P < 0.05; Fig. 3a) and the ratio index of Fig. 2 Frequency distribution of tuft area asymmetry values from a sample of ®eld-collected male S. ocreata (n = 33) 90 Fig. 4 Receptivity of female S. ocreata to video images of symmetric and asymmetric males (for treatments, see legend to Fig. 1 and text) the ``intermediate'' and ``low'' asymmetry categories (Table 1). Video image manipulation and playback studies Fig. 3 Regressions of male tuft asymmetry on body size (length in mm) (a) and condition (index of abdomen volume/cephalothorax width) (b) from a subsample of male S. ocreata measured directly after collection from the ®eld body condition (abdomen volume/CW, r = )0.782, P < 0.05; Fig. 3b). Response of females to asymmetry of live males Female receptivity responses to live males from three asymmetry categories were not independent of degree of asymmetry (log-likelihood ratio test, GWilliams correction = 10.43, P < 0.005). Females exhibited no receptivity to males in the ``high'' asymmetry category, while approximately 60% of females were receptive to males in Table 1 Number of female Schizocosa ocreata exhibiting receptivity to males with varied degrees of asymmetry. The categories correspond to approximate thirds of the range of asymmetry values in the frequency distribution in Fig. 2 Asymmetry category Receptive Not receptive Lowest third (< 0.4 mm2) Middle third (< 0.4±0.7 mm2) Highest third (> 0.7 mm2) 7 (63.6%) 9 (60.0%) 0 4 (36.3%) 6 (40.0%) 7 (100.0%) Female receptivity was not independent of the symmetry of the video stimulus (log-likelihood ratio test, GWilliams correction = 33.89, P < 0.001), and was clearly reduced in response to all asymmetric male video stimuli compared to the symmetric male video control (Fig. 4). The video treatments were treated above as qualitatively dierent categorical treatments in a contingency analysis, allowing rejecton of the null hypothesis of independence. However, while the degree of asymmetry in the video stimuli ranges from a low of approximately 0.24 (control) to a high of approximately 2.40 (``removed''), two treatments ± ``reduced'' and ``enlarged'' ± are qualitatively dierent but equal in asymmetry. An alternate analysis may therefore be appropriate, and we have taken two additional approaches: (a) we tested each of the asymmetry manipulations against the control separately, and all were signi®cantly dierent (removed: v2 = 3.88, P < 0.01; reduced: v2 = 8.42, P < 0.05; balanced: v2 = 4.40, P < 0.05; enlarged: v2 = 3.70, P £ 0.05), and (b) we tested all four asymmetry treatments in a contingency test of the null hypothesis of independence (to address the concern that previous rejection of the null hypothesis was based on females responding to extreme values); the result was not signi®cant (GWilliams correction = 2.40, P > 0.60). Discussion Numerous studies have demonstrated that asymmetry of secondary sexual characters in¯uences male mating success in a variety of animals, including insects, ®sh, birds, and mammals (Mùller 1990; Thornhill 1992; 91 Swaddle and Cuthill 1994; Watson and Thornhill 1994; Mùller and Swaddle 1997; Sheridan and Pomiankowski 1997; but see Dufour and Weatherhead 1998a,b; Ligon et al. 1998). Data presented here and elsewhere (Uetz et al. 1996) suggest that this may be true for S. ocreata wolf spiders as well. Earlier work has shown that female S. ocreata discriminate against males whose foreleg tufts are grossly asymmetric owing to limb loss and incomplete regeneration (Uetz et al. 1996). In this study, we present evidence for FA in these decorative tufts, and that it may in¯uence female mate choice. Female wolf spiders exhibited reduced receptivity responses to live males with higher asymmetry values, as well as to video images of courting males with varying degrees of asymmetry. While concerns over reduced response to extreme phenotypes might well be raised here (e.g., male wolf spiders with high asymmetry in tufts might be behaviorally impaired and exhibit less vigorous courtship, or be perceived as heterospeci®cs), the results of the video playback experiments suggest that the asymmetry of the visual signal presented by the tufts, rather than any concomitant behavioral or size factor, is the cue to which female wolf spiders are responding. Criticisms are often raised concerning FA and mate choice because few studies are able to experimentally separate the asymmetry of the male character from other covarying qualities that might also in¯uence mate choice (Palmer 1996). Video playback has been increasingly used as an experimental tool in studies of animal behavior in numerous taxa, including ®sh (Rowland 1995; Rowland et al. 1995a,b; Rosenthal et al. 1996; KodricBrown and Nicoletto 1997), birds (Evans and Marler 1992; Adret 1997), reptiles (Macedonia and Stamps 1994; Macedonia et al. 1994) and amphibians (Roster et al. 1995). Of particular value, as demonstrated in this case, is the ability to manipulate a single aspect of the visual signal to test its potential importance. The responses of female Schizocosa spiders to video stimuli from male conspeci®cs are identical to those seen with live males, and receptivity displays are clear indications of a female's willingness to mate (McClintock and Uetz 1996; Uetz et al. 1996). The video studies conducted here clearly demonstrate that with male size and behavior held constant, female wolf spiders prefer symmetrical over asymmetrical males. In several recent experiments, Swaddle and Cuthill (1994), Bennett et al. (1996) and Fiske and Amundsen (1997) successfully altered female mate preference in birds using manipulated asymmetry of colored leg bands (but see Jennions 1998). Follow-up studies demonstrated that not only mating success but also reproductive success (number of surviving ospring) are related to experimental manipulation of symmetry in leg bands (Swaddle 1996). Recent criticism of this approach by Rohde et al. (1997) has centered on the use of a highly arti®cial ornamental trait, such that ``the female's lack of preference may potentially re¯ect no more than the nonacceptance of extremely deviant or physically defective males'' (but see responses by Swaddle 1997; Swaddle and Cuthill 1997). In our study, the trait in question was not an arti®cial addition (e.g., colored leg bands), and manipulation of asymmetry in the trait was (in part) within the range of natural variation. Even so, the absence of receptivity must be interpreted with some caution. While we can be certain that a female perceives the video stimulus by the presence of orientation toward the screen, it is possible that asymmetry in the tufts may be perceived by the female as an extreme phenotype outside the range of species-typical display, and lack of response might indicate non-acceptance on that basis (Rohde et al. 1997). However, while the frequency of females responding receptively was signi®cantly reduced relative to the control across all asymmetric stimuli, some receptivity responses were seen (14±35% of females). This would suggest that asymmetric video stimuli are recognizable as conspeci®c males. Moreover, the absence of statistically signi®cant dierences between those treatments within the range of natural variation in asymmetry (``reduced'' and ``enlarged'') and the more extreme manipulations (``balanced'' and ``removed'') argues for discrimination against asymmetry, and not simply a lack of species recognition or rejection of extreme phenotypes or deviant forms. There are several plausible explanations why female wolf spiders should recognize and discriminate against asymmetric males. Visual signaling with leg-waving displays is a critical element of the signal-receiver process in S. ocreata, as their complex leaf litter habitat may constrain female perception of male vibratory courtship (Uetz 1990; Scheer et al. 1996). Asymmetry in the tufts might therefore reduce the ecacy of the male's visual signal. Studies with neural net models suggest that sensory systems may have the innate capacity to discriminate against asymmetric signals or select for symmetry as a by-product of mate recognition (Enquist and Arak 1994; Johnstone 1994; Enquist and Johnstone 1997). Alternatively, asymmetry in this highly visual secondary sexual characteristic may re¯ect some aspect of male ``quality'' and/or developmental instability, and serve as an indicator trait for female mate choice. Data presented here for live ®eld-collected males show that tuft asymmetry is negatively correlated with body condition, and thus may re¯ect a male's feeding history. If asymmetry re¯ects prior leg autotomy and regeneration owing to an aggressive encounter, it may also serve as an indicator of reduced ®ghting success (Uetz et al. 1996). Conversely, asymmetry of regenerated legs may indicate developmental stability in a direct way, as those with superior ability to regenerate autotomized limbs would be more symmetric. If asymmetry was related to any of these possibilities, female choice based on symmetry in S. ocreata might enhance female ®tness indirectly by aecting ospring success, and be subject to selection as well. Acknowledgements This research was supported by grant IBN 9414239 from the National Science Foundation, and a supplement from the Research Experiences for Undergraduates Program. We thank the Cincinnati Nature Center, Milford, Ohio, for 92 allowing us to collect spiders. We gratefully acknowledge the assistance of A. DeLay, K. Delaney, E. Hebets, M. Persons, R. Mendoza, B. Kilinc, H. Kuyper and E. Sanborn in collecting and maintaining spiders for this research project. We also appreciate statistical advice from R.Z. German, and comments on the manuscript from J. Calos, J.A. Roberts, M. Orr, D. Kroeger, M. Porter, M. Elgar, T. Amundsen and two anonymous reviewers. References Adret P (1997) Discrimination of video images by zebra ®nches (Taeniopygia guttata): direct evidence from song performance. J Comp Psychol 111:115±125 Aspey WP (1976) Response strategies of adult-male Schizocosa crassipes (Araneae: Lycosidae). Psyche 83:94±105 Aspey WP (1977) Wolf spider sociobiology. I. Agonistic display and dominance-subordinance relations in adult male Schizocosa crassipes. Behavior 62:103±141 Bennett ATD, Cuthill IC, Partridge JC, Maier EJ (1996) Ultraviolet vision and mate choice in zebra ®nches. Nature 380:433±435 BjoÈrklund M, MerilaÈ J (1997) Why some measures of ¯uctuating asymmetry are so sensitive to measurement error. Ann Zool Fenn 34:133±137 Clark DL, Uetz GW (1991) Video image recognition by jumping spiders. Anim Behav 40:884±890 Clark DL, Uetz GW (1992) Morph-independent mate selection in a dimorphic jumping spider: demonstration of movement-bias in female choice using video controlled courtship behaviour. Anim Behav 43:247±254 Clark DL, Uetz GW (1993) Signal ecacy and the evolution of male dimorphism in the jumping spider Maevia inclemens (Araneae: Salticidae). Proc Natl Acad Sci 90:11954±11957 Dufour KW, Weatherhead PJ (1998a) Bilateral symmetry as an indicator of male quality in red-winged blackbirds: associations with measures of health, viability, and parental eort. Behav Ecol 9:220±231 Dufour KW, Weatherhead PJ (1998b) Reproductive consequences of bilateral asymmetry for individual male red-winged blackbirds. Behav Ecol 9:232±242 Enquist M, Arak A (1994) Symmetry, beauty and evolution. Nature 372:169±172 Enquist M, Johnstone RA (1997) Generalization and the evolution of symmetry preferences. Proc R Soc Lond B 264:1345± 1348 Evans CS, Marler P (1992). Female appearance as a factor in the responsiveness of male chickens during anti-predator behavior and courtship. Anim Behav 43:137±145 Fiske P, Amundsen T (1997) Female bluethroats prefer males with symmetric color bands. Anim Behav 54:81±87 Gangestad SW, Thornhill R (1998) The analysis of ¯uctuating asymmetry redux: the robustness of parametric statistics. Anim Behav 55:497±501 Hill GE (1995) Ornamental traits as indicators of environmental health. Bioscience 45:25±31 Jakob EM, Marshall SD, Uetz GW (1996). Estimating ®tness: a comparison of body condition indices. Oikos 77:61±67 Jennions MD (1998) The eect of leg-band asymmetry on femalemale association in zebra ®nches. Anim Behav 55:61±67 Johnstone R (1994) Female preference for symmetrical males as a by-product of selection for mate recognition. Nature 372:172± 175 Kodric-Brown A, Nicoletto PF (1997) Repeatability of female choice in the guppy: response to live and videotaped males. Anim Behav 54:369±376 Leary RF, Allendorf FW (1989) Fluctuating asymmetry as an indicator of stress-implication for conservation biology. Trends Ecol Evol 4:214±217 Ligon JD, Kimball R, Merola-Zwartjes M (1998) Mate choice by female red junglefowl: the issues fof multiple ornaments and ¯uctuating asymmetry. Anim Behav 55:41±50 Ludwig W (1932) Das Rechts-Links Problem in Tierreich und beim Menschen. Springer, Berlin Heidelberg New York Macedonia JM, Stamps JA (1994) Species recognition in Anolis grahami (Sauria, Iguanidae): evidence from responses to video playbacks of conspeci®c and heterospeci®c displays. Ethology 98:246±264 Macedonia JM, Evans CS, Losos JB (1994) Male Anolis lizards discriminate video-recorded conspeci®c and heterospeci®c displays. Anim Behav 47:1220±1223 Markow TA (1995) Evolutionary ecology and developmental instability. Annu Rev Entomol 40:105±120 McClintock WJ, Uetz GW (1996) Female choice and preexisting bias: visual cues during courtship in two Schizocosa wolf spiders (Araneae: Lycosidae) Anim Behav. 52:167±181 Mùller AP (1990) Fluctuating asymmetry in male sexual ornaments may reliably reveal male quality. Anim Behav 40:1185± 1187 Mùller AP (1992a) Patterns of ¯uctuating asymmetry in weapons: evidence for reliable signalling of quality in beetle horns and bird spurs. Proc R Soc Lond B 248:199±206 Mùller AP (1992b) Parasites dierentially increase the degree of ¯uctuating asymmetry in secondary sexual characters. J. Evol Biol 5:691±699 Mùller AP (1993) Patterns of ¯uctuating asymmetry in sexual ornaments predict female choice. J Evol Biol 6:481±491 Mùller AP (1997) Developmental stability and ®tness: a review. Am Nat 149:916±932 Mùller AP, Swaddle JP (1997) Asymmetry, developmental stability and evolution. Oxford University Press, Oxford Montgomery TH (1903) Studies on the habits of spiders, particularly those of the mating period. Proc Acad Nat Sci Pa 55:59± 149 Palmer RA (1996) Waltzing with asymmetry. Bioscience 46:518± 532 Palmer AR, Strobeck C (1986) Fluctuating asymmetry: measurement, analysis, patterns. Annu Rev Ecol Syst 17:391±421 Persons MH, Uetz GW (1997) The eect of prey movement on attack behavior and patch residence decision rules of wolf spiders (Araneae: Lycosidae). J. Insect Behav 10:737±752 Pomory CM (1997) Fluctuating asymmetry: biological relevance or statistical noise? Anim Behav 53:225±227 Rohde PA, Amundsen T, Fiske P (1997) Fluctuating asymmetry, mate choice and experimental designs. Anim Behav 54:1030± 1033 Rosenthal GG, Evans CS and Miller WL (1996) Female preference for dynamic traits in the green swordtail, Xiphophorus helleri. Anim Behav 51:811±820 Roster NO, Clark DL, Gillingham JC (1995) Prey-catching behavior in frogs and toads using video-simulated prey. Copeia 1995:496±498 Rowland WJ (1995) Do female stickleback care about male courtship vigour ± manipulation of display tempo using video playback. Behaviour 132:951±961 Rowland WJ, Bolyard KJ, Halpern AD (1995a). The dual eect of stickleback nuptial coloration on rivals: manipulation of a graded signal using video playback. Anim Behav 50:267± 272 Rowland WJ, Bolyard KJ, Jenkins JJ, Fowler J (1995b) Video playback experiments on stickleback mate choice: female motivation and attentiveness to male colour cues. Anim Behav 49:1559±1567 Scheer SJ, Uetz GW, Stratton GE (1996) Sexual selection, male morphology, and the ecacy of courtship signalling in two wolf spiders (Araneae: Lycosidae). Behav Ecol Sociobiol 38:17±23 Sheridan L, Pomiankowski A (1997) Female choice for spot asymmetry in the Trinidanian guppy. Anim Behav 54:1523±1529 Stratton GE, Uetz GW (1981) Acoustic communication and reproductive isolation in two species of wolf spider. Science 214:575±577 Stratton GE, Uetz GW (1983) Communication via substratumcoupled stridulation and reproductive isolation in wolf spiders (Araneae: Lycosidae). Anim Behav 31:164±172 93 Stratton GE, Uetz GW (1986) The inheritance of courtship behavior and its role as a reproductive isolating mechanisms in two species of Schizocosa wolf spiders (Araneae: Lycosidae). Evolution 40:129±141 Swaddle JP (1996) Reproductive success and symmetry in zebra ®nches. Anim Behav 51:203±210 Swaddle JP (1997) Experimental design and the signalling properties of ¯uctuating asymmetry. Anim Behav 54:1034±1037 Swaddle JP, Cuthill IC (1994) Preference for symmetric males by female zebra ®nches. Nature 367:165±166 Swaddle JP, Cuthill IC (1997) The biological relevance of testing for perfect symmetry. Anim Behav 54:475±476 Thornhill R (1992) Fluctuating asymmetry and the mating system of the Japanese scorpion¯y, Panorpa japonica. Anim Behav 44:867±879 Uetz GW (1990) Habitat structure and spider foraging. In: Bell SS, McCoy ED, Mushinsky HR (ed) Habitat structure: the physical arrangement of objects in space. Chapman & Hall, London, pp 325±348 Uetz GW, Denterlein GJ (1979) Courtship behavior, habitat, and reproductive isolation in Schizocosa rovneri (Uetz & Dondale) (Araneae: Lycosidae). J.Arachnol 7:86±88 Uetz GW, McClintock WJ, Miller D, Smith EI, Cook KK (1996) Limb regeneration and subsequent asymmetry in a male secondary character in¯uences sexual selection in wolf spiders. Behav Ecol Sociobiol 38:253±257 Watson PJ, Thornhill R (1994) Fluctuating asymmetry and sexual selection. Trends Ecol Evol 9:21±25 Communicated by M.A. Elgar