IN species where tail autotomy can occur during

Copeia, 2004(1), pp. 165 172 Tail Autotomy in Territorial Salamanders Influences Scent Marking by Residents and Behavioral Responses of Intruders to Resident Chemical Cues SHARON E. WISE, FRANK D. VERRET, AND ROBERT G. JAEGER In territorial salamanders and lizards, tail loss may immediately reduce the resource holding power (RHP) of residents, making defense of territories more difficult. Residents of the Red-Backed Salamander, Plethodon cinereus, defend territories using scent marks and agonistic displays. We examined (1) the marking behavior of tailed and tailless residents when establishing new territories, and (2) the behavioral responses of intruders to chemical cues produced by tailed and tailless residents. In the first experiment, we counted the number of postcloacal presses (PCP: a scent-marking behavior) by tailed and tailless salamanders (future residents) placed into unmarked test chambers. Tailless salamanders performed significantly more PCPs than tailed salamanders. Increased marking may have provided a benefit for tailless residents: tailed intruders, when placed in chambers marked by tailless residents, showed significantly less aggression when the residents performed more PCPs. In the second experiment, we observed the agonistic behavior of intruders placed into test chambers previously occupied by tailed or tailless residents. Tailed, but not tailless, intruders were significantly less aggressive in chambers marked by tailed residents than in chambers marked by tailless residents. Our results suggest that scent marks, in the absence of visual cues, provide intruders with information associated with tail condition (e.g., RHP) of residents. IN species where tail autotomy can occur during predatory encounters or competitive interactions, loss of the tail can impose social costs on affected individuals (Fox and Rostker, 1982; Martín and Salvador, 1993; Wise and Jaeger, 1998) and may play a role in regulating population density (Harris, 1989). Among vertebrates, some species of lizards and salamanders are capable of tail autotomy. In the Side- Blotched Lizard (Uta stansburiana), individuals with intact tails were often more dominant during agonistic interactions than conspecifics with autotomized tails (Fox and Rostker, 1982; Fox et al., 1990). In subadult females, the tail appears to act as a status-signaling badge (Fox et al., 1990). In territorial Red-Backed Salamanders (Plethodon cinereus), tail loss seems to influence resource-holding power (RHP; Maynard Smith and Parker, 1976) with tail-autotomized (tailless) individuals being at a disadvantage in agonistic interactions relative to tail-intact (tailed) individuals; tailless, territorial residents experienced more aggression from tail-intact intruders than did tailed residents (Wise and Jaeger, 1998). For territorial species that communicate largely through visual signals, such as U. stansburiana (Fox, 1983), a resident that has lost its tail may not be able to hide this information from conspecifics; thus, an individual may not be able to deceive a visually oriented intruder about its ability for territorial defense. However, many territorial salamanders, especially the Plethodontidae, communicate using chemical signals ( Jaeger, 1986). For example, nocturnal, terrestrial Red-Backed Salamanders communicate territorial ownership largely through chemical signals, both intraspecifically (pheromones: Jaeger et al., 1986) and interspecifically (allomones: Jaeger and Gergits, 1979; Griffis and Jaeger, 1998), although visual signals are important during close encounters between contestants ( Jaeger, 1984; Townsend and Jaeger, 1998) when the environmental illumination is adequate for vision. These chemical signals may or may not provide specific information about the tail condition (and thus the RHP) of the resident. Our experiments examined (1) territorial marking by tailed and tailless salamanders placed on clean, unmarked substrata when they begin to mark new territories (Nunes and Jaeger, 1989), and (2) the agonistic behavior of tailed and tailless intruders on substrata previously marked by tailed and tailless residents (in the absence of residents). Males of P. cinereus use chemical cues as territorial markers ( Jaeger et al., 1986) for defense of cover objects (rocks and logs) that provide access to moisture ( Jaeger, 1980a,b), prey (Mathis, 1989), and probably mates (Mathis, 2004 by the American Society of Ichthyologists and Herpetologists

166 COPEIA, 2004, NO. 1 1991). The chemical cues are secreted, at least in part, from the postcloacal gland (Simons et al., 1994, 1999). This gland is located just posterior to the cloaca but is anterior to the most proximal plane of autotomy in the tail. Thus, tail autotomy does not result in the removal of the scent-marking gland that is thought to produce chemical signals used in marking a territory. However, glands on the midventral surface of the tail may also contribute to the chemical signal (Simons and Felgenhauer, 1992). Thus, the quantity and/or quality of the chemical signal may change after tail autotomy if glands on the tail contribute part of the chemical signal, or if the tail (or its absence) otherwise contributes to signals produced by glands located elsewhere on the body (e.g., if fat reserves in the tail contribute to glandular production in general). We tested the hypothesis that residents alter the rate at which they mark the substrate with chemical cues from the postcloacal gland based on their tail condition (tailed or tailless). We also pose two alternative hypotheses concerning information conferred by chemical signals during territorial defense. In the first hypothesis, the chemical signal merely informs an intruder that a territorial resident is present but provides little or no information about the status of the resident. Gosling (1982, 1986, 1990) suggested that, for mammals, territorial scent marks merely identify the authenticity of the territorial resident (Gosling s scent-matching hypothesis). That is, if the intruder can match the odor of the scent mark in a territory with the scent of an individual within that territory, then that individual must be the territorial resident. For territorial Red-Backed Salamanders, chemical signals used only for scent matching would convey no information concerning the tailed or tailless condition of the resident (assuming that tail autotomy does not entirely remove the ability of the salamander to scent mark). In the second hypothesis, scent marks provide information concerning the tailed or tailless condition (or other characteristics associated with tail condition, e.g., energetic condition) of the salamander producing the scent. We tested these two alternative hypotheses using adult male Red-Backed Salamanders with intact (tailed) or shortened tails (tailless) as territorial residents or intruders. Wise and Jaeger (1998) found that male territorial intruders with intact tails were more aggressive toward tailless residents than toward tailed residents when residents and intruders actually encountered each other on marked territories. Their results suggest that tail loss reduces the RHP of residents, perhaps because of the function of the tail in signaling during agonistic interactions ( Jaeger and Schwarz, 1991) or in energy storage (Fraser, 1980), or because of a diversion of energy to regeneration of the tail as in the California Slender Salamander, Batrachoseps attenuatus (Maiorana, 1977). However, Wise and Jaeger (1998) could not determine whether intruders assessed the tail condition (or characteristics associated with tail condition) of the residents using visual information, chemical signals, or both. We tested whether intruders can detect tail condition (or characteristics associated with tail condition) of residents using only chemical cues present on the substratum of residents territories. MATERIALS AND METHODS During August 1990, we collected adult males of P. cinereus from an area near Mountain Lake Biological Station, Giles County, Virginia. Salamanders were placed in individual containers and transported to the University of Louisiana at Lafayette. In the laboratory, they were fed Drosophila virilis ad libitum and kept in individual Petri dishes (14.5 1.5 cm) lined with damp (aged tap water) filter paper at 16 20 C on a 12:12 h light:dark photoperiod. We measured the snout vent length (SVL: from the tip of the snout to the posterior end of the cloaca) and tail length (from the posterior end of the cloaca to the tip of the tail) of each salamander. Only males with tail lengths measuring at least 75% of the SVL were used in the experiment, as in Wise and Jaeger (1998). Groups of four males (a tetrad) were matched for SVL to within 1 mm so that the effect of asymmetries in body size would be minimized by our experimental design (Mathis, 1990; Townsend and Jaeger, 1998). We randomly designated each member of a tetrad as a tailed resident, a tailless resident, a tailed intruder, or a tailless intruder. We used a total of 18 tetrads (72 males) in this study (n 18). We produced tailless residents and intruders by inducing tail loss of approximately 85% of the original tail length. For tailed residents and intruders, 2 mm of the tail was removed as a control for handling. Red-Backed Salamanders have a specialization for tail autotomy such that the vertebra detaches and blood vessels and muscles constrict just anterior to the site of pressure, whereas the skin detaches at the site of pressure on the tail, closing over the exposed tissue and resulting in little to no loss of blood (Wake and Dresner, 1967). In our experiments, we allowed salamanders to autotomize their tails by giving the tail a slight pinch with forceps at

WISE ET AL. PHEROMONES SIGNAL RESIDENT CONDITION 167 the appropriate site along the length of the tail (as in Wise and Jaeger, 1998). Tail loss is common in natural populations; up to 61% of individuals in the area where these animals were collected exhibited some degree of tail loss, depending on the time of year that sampling occurred (Mathis, 1991; Wise, 1995). Thus, this procedure seemingly puts no additional stress on the laboratory salamanders than would commonly occur in nature, including potential social costs and energy demands necessary to regenerate the tail. We induced tail autotomy of all the salamanders on the same day (25 September 1990) to be consistent between individual comparisons because each individual was used twice over the duration of the experiment (e.g., a given tailless intruder was tested once on the substratum of a tailed resident and once on the substratum of a tailless resident). Because salamanders partially regenerated tails over the duration of the study, tail lengths of tailless salamanders did not remain consistently at 85% of the original length for all tests. As a result, we examined the potential effects of regeneration on the behavior of intruders in our tests (see Appendix 1). Tests were conducted from 11 October 1990 through 19 March 1991 from 1400 2100 h. During all tests, we used the minimum amount of incandescent light necessary for the observer to see clearly the salamanders (approximately 0.2 lux, measured as reflected light 10 cm off the substratum of the test chamber). All salamanders were taken into the testing area at least 1 h before testing to acclimate to light levels. We randomized the order of testing for each tetrad and, within a tetrad, for each intruder-resident pair. As a result, each salamander had at least 23 days between tests. Test 1: The effect of tail condition (tailed and tailless) on scent marking. In preparation for the second test (see below), future residents and intruders were placed randomly into separate, glass-covered test chambers (31 17 1.5 cm plastic trays lined with two clean, brown paper towels moistened with aged tap water). Each salamander was fed D. virilis ad libitum and kept at 16 20 C. Seventy-two males were set up in this manner. Fifty-five of these males (28 tailed, 27 tailless) were randomly selected for the first test and individually observed by RGJ when each was placed into a new test chamber. For the observation period, each salamander was placed under a transparent, inverted Petri dish lid (6.0 0.75 cm) for 15 min. After this habituation period, the lid of the Petri dish was removed and the glass cover was placed over the test chamber. The number of postcloacal presses (PCP; a scent marking behavior where salamanders touch a gland located posterior to the cloaca against the substratum: Simons and Felgenhauer, 1992; Simons et al., 1994) was recorded for 30 min. We predicted that, if tail loss confers a potential disadvantage to residents, then salamanders might differ in behavior when first marking their territories. To test this prediction, we compared the number of PCPs by tailed (n 28) and tailless (n 27) salamanders using a twotailed Mann-Whitney U-test (the sample distributions were nonnormal) at 0.05. Test 2: Response of intruders to chemical signals of tailed and tailless residents. Five days (an adequate period of time for the establishment of a territory in the laboratory: Jaeger, 1981; Nunes and Jaeger, 1989) after a resident and intruder were initially placed in their test chambers, the resident was removed from its test chamber by SEW. Then, FDV placed the intruder into the center of the empty test chamber of the resident, under an inverted Petri dish bottom (6.0 0.75 cm) for a 15-min habituation period. The Petri dish was then removed, the glass lid replaced, and the behavior of the salamander was observed by FDV for 30 min. The removal of the resident by SEW ensured that the observer, FDV, was unaware of the tail condition of the resident while testing an intruder on a resident s substratum. FDV observed the amount of time that salamanders spent exhibiting (1) All Trunk Raised (ATR), an aggressive behavior in which the salamander raises its entire body off of the substratum ( Jaeger, 1984); and (2) FLAT, a submissive behavior in which the salamander lays its entire body, including its head, against the substratum ( Jaeger, 1984). After the first 15 min of observation, we introduced five D. virilis into the test chamber from a hole midway along the long side of the chamber. The flies provided foraging opportunities for intruders during the remaining 15 min of the observation period. Flies were blown into the test chamber through a tube that was inserted into a hole in the side of the test chamber (small enough so that salamanders could not escape through the hole). Salamanders did not appear to be disturbed by the introduction of flies into the chamber. We tested the null hypothesis that intruders cannot distinguish the tail condition of residents by comparing the behavior of (1) tailed intruders on the substrata of tailed and tailless residents and (2) tailless intruders on the substrata of tailed and tailless residents. For each of these analyses, we used a repeated measures

168 COPEIA, 2004, NO. 1 REPEATED MEASURES MANOVA AND SUBSEQUENT ANOVA FOR THE COMPARISON OF THE AGONISTIC BEHAVIOR OF TAILED INTRUDERS ON THE SUBSTRATA OF TAILED AND TAILLESS RESIDENTS AND FOR TAILLESS IN- TRUDERS ON THE SUBSTRATA OF TAILED AND TAILLESS RESIDENTS. Agonistic behavior includes ATR (all-trunk raised, an aggressive posture) and FLAT (a submissive posture). TABLE 1. Intruder responses to substrata of tailed/ tailless residents Tailed intruders ATR FLAT Tailless intruders * Statistically significant at 0.05. Wilks 0.583 0.995 MANOVA df F P 2, 16 5.72 0.013* 2, 16 0.037 0.964 ANOVA df F P 1, 17 1, 17 12.1 4.43 0.003* 0.051 MANOVA (SPSS 10.0) to examine the behavior of intruders (response variables were time spent in ATR and FLAT). We used tests of related data because the behavior of each intruder was compared when on the substrata of two different types of residents (one tailed and one tailless). We attempted to transform response variables, when necessary, to meet the assumptions of the repeated-measures MANOVA (Tabachnick and Fidell, 1989). We transformed FLAT (squareroot) for the comparison of tailed intruders on substrata of tailed versus tailless residents. FLAT was nonnormally distributed for the comparison of tailless intruders on substrata of tailed and tailless residents but could not be normalized using a transformation. Therefore, we compared the results of the repeated measures MANOVA (using nontransformed FLAT) with separate nonparametric, two-tailed Wilcoxon matched-pairs signed-ranks analyses for ATR and FLAT (SPSS 10.0). The results were similar, so only the MANOVA will be presented. We subsequently used repeated-measures ANOVA to determine the significance of the treatments on each response variable (ATR and FLAT). Posthoc analysis: The relationship between marking by residents and the behavior of intruders. Twentyone of the salamanders used in the first test (observed when first marking a test chamber) were residents (nine were tailed and 12 were tailless). As described in the second test, these residents were removed after five days, and a tailed intruder was placed in each of the chambers formerly occupied by these residents. The behavior of these intruders was subsequently observed. Using the data (number of PCPs) from the marking observations of the first test and the data from the subsequent behavior of the intruders (ATR and FLAT) in the second test, we performed posthoc analyses to determine whether the behavior of tailed intruders (because only tailed intruders showed significant differences in behavior in the second test) placed on the substrata of tailed and tailless residents could be predicted by the number of PCPs exhibited by territorial residents during the first 30 min in their test chambers (as a measure of the residents propensity for scent marking). We regressed the time that intruders spent in ATR and FLAT on the number of PCPs exhibited by tailed and tailless residents when first placed in their new territories. For each analysis, we initially reduced alpha to 0.025 using a sequential Bonferroni technique (Rice, 1989), because we used two behavioral patterns (ATR and FLAT) to test each hypothesis (Miller, 1981). The number of PCPs in each analysis and data for FLAT were log-transformed to meet the assumptions of the regression analysis. RESULTS Test 1: The effect of tail condition (tailed and tailless) on scent marking. Tail condition significantly influenced the number of PCPs delivered by salamanders immediately after being placed into new test chambers. Tailless salamanders exhibited significantly more PCPs than did tailed salamanders when placed on new substrata (tailed salamanders: median 4.00, interquartile range [IQR] 3.00 9.75; tailless salamanders: median 9.50, IQR 4.50 18.00; Mann-Whitney U-test: U 225, P 0.016). Test 2: Response of intruders to chemical signals of tailed and tailless residents. There was a significant effect of resident condition on the behavior of tailed intruders (repeated-measures MANOVA: Table 1). Tailed intruders spent significantly less time in the aggressive posture ATR when on substrata marked by tailed residents than when on substrata marked by tailless residents (repeated measures ANOVA: Table 1, Fig. 1). There was no significant difference in the submissive behavior FLAT (repeated measures ANOVA: Table 1, Fig. 1). Tailless intruders showed no significant difference in aggression

WISE ET AL. PHEROMONES SIGNAL RESIDENT CONDITION 169 Fig. 1. The mean ( 1 SD) amount of time (sec) that tailed and tailless intruders spent exhibiting agonistic behavior when on the substrata of tailed and tailless residents. The behavioral patterns are ATR (all trunk raised: aggression) and FLAT (submission). Data in this figure are not transformed. or submission on substrata of tailed and tailless residents (repeated-measures MANOVA: Table 1, Fig. 1). Posthoc analysis: The relationship between marking by residents and the behavior of intruders. We found no significant relationship between the time that tailed intruders spent in ATR or FLAT when on the substrata of tailed residents and the number of PCPs delivered by those tailed residents when first introduced to their test chambers (ATR: R 2 0.356, t 1.97, n 9, P 0.090; FLAT: R 2 0.323, t 1.83, n 9, P 0.111). We found a significant negative relationship between the time that tailed intruders spent in ATR (R 2 0.430, t 2.75, n 12, P 0.021; Fig. 2) but not FLAT (R 2 0.183, t 1.50, n 12, P 0.165; Fig. 2), when on the substrata of tailless residents and the number of PCPs delivered by those tailless residents when first introduced to their test chambers. DISCUSSION Tailed intruders exhibited significantly less aggression when on the substrata of tailed residents than when on the substrata of tailless residents. This result does not conform to Gosling s (1982, 1986, 1990) scent-matching hypothesis that scent marks provide information that allows intruders merely to associate a particular resident with a particular territory. Instead, our data support our second hypothesis that scent marks by residents of the Red-Backed Fig. 2. The amount of time (sec) that tailed intruders spent exhibiting agonistic behavior on the substrata of tailless residents regressed on the number (log-transformed) of postcloacal presses (PCPs) exhibited by those tailless residents for 30 min immediately after being introduced into the test chamber. The behavioral patterns are ATR (all trunk raised: aggression) and FLAT (submission). Salamander provide information associated with tail loss to intruders. The reduced aggression exhibited by tailed intruders when on the substrata of tailed residents that we observed in our study is consistent with the behavior of tailed intruders observed in Wise and Jaeger (1998). They found that in contests in which residents and intruders had direct contact (visual and olfactory information was available to residents and intruders), tailed intruders were less aggressive toward tailed residents than toward tailless residents. Game theoretic models predict that individuals are more likely to escalate in contests when asymmetries in pay-offs and fighting ability are decreased (Maynard Smith and Parker, 1976). In P. cinereus, an asymmetry in contests seems to occur because residency confers an advantage during territorial contests such that residents are likely to win territorial contests and expel intruders ( Jaeger et al., 1982). Tail loss may lead to a decrease in defensive ability (Wise and Jaeger, 1998), such that, if tail loss occurs in the resident, the probability of an intruder usurping a resident s territory may be increased (that is, asymmetries are reduced) compared to a contest in which the resident has an intact tail. A decrease in defensive ability may result from the loss of the tail as a component of aggressive displays ( Jaeger and Schwarz, 1991), the loss of energy caused by loss of fat stores in the tail or

170 COPEIA, 2004, NO. 1 diversion of energy to caudal regeneration (B. attenuatus: Maiorana, 1977; P. cinereus: Fraser, 1980), reduced balance or locomotor ability (lizards: Arnold, 1984), or some other unknown mechanism. Thus, tailed intruders may respond differently to chemical signals of tailed and tailless residents, because the probability of winning a contest is reduced when residents are tailed and increased when residents are tailless. According to our results, this response seems to be a reduction of aggressive behavior by tailed intruders when detecting the scent of tailed residents compared to tailless residents. An increased probability of winning a contest may apply only to tailed intruders, because tailless intruders are at a disadvantage in contests with both tailed and tailless residents (residents always have residency advantage). Thus, their probability of winning a contest is presumably always close to 0% (Wise and Jaeger, 1998). Based on our results, tailless salamanders did not differ significantly in agonistic behavior in any of our comparisons, perhaps because the probability of winning a contest did not change. This result differed from Wise and Jaeger (1998), where tailless intruders altered their behavior based on the tail condition of the resident in the same manner as tailed intruders. However, in the experiment by Wise and Jaeger (1998), residents and intruders were able to interact, unlike in our current study where there was no resident present. Although further research is necessary to resolve differences in behavior in these two studies (and between tailed and tailless intruders within this study), it seems reasonable to suggest that in our present study, tailless intruders could detect differences in residents associated with tail loss (just as tailed intruders apparently did), but that they did not alter their agonistic behavior based on this information. Chemical signals may be produced by a number of glands found on the tail, such that loss of all or part of the tail could affect chemical signal production or deposition and is detectable by intruders. Tailless salamanders (future residents) placed on new (unmarked) substrata exhibited more PCPs (postcloacal presses, a scent-marking behavior: Simons and Felgenhauer, 1992) than did tailed salamanders. Such behavior seems to be a response to qualitative or quantitative difference in residents own scent marks when tailed versus tailless. The effect of increased marking behavior may be compensatory (as seen in Fig. 2); we found that tailed intruders reduced their aggressive behavior based on the increased propensity of residents to mark their territories. Scent marking by territorial Red-Backed Salamanders is an important component of territorial defense ( Jaeger, 1986; Simons et al., 1994, 1997). Because tailless residents are potentially at a disadvantage in territorial contests with tailed intruders (Wise and Jaeger, 1998), broadcasting information associated with tail loss may confer a disadvantage to residents that have lost their tails, making territorial defense more difficult. Thus, tailless residents may respond to tail loss (or some other factor associated with tail loss) by increasing the rate at which they mark new territories. An increase in marking behavior by tailless residents was followed by a reduction in aggression by tailed intruders that were subsequently exposed to the residents substrata. Thus, increased marking may benefit tailless residents by reducing the aggression exhibited by intruders exposed to those scent markings. Tail loss by territorial residents of P. cinereus is common in the area in which these salamanders were captured. At Mountain Lake Biological Station, Virginia, 61% of salamanders found under cover objects in May 1994 were found to have at least some portion of the tail missing (Wise, 1995). Mathis (1991) found that 45% of presumed residents exhibited some amount of tail loss between mid-june and mid-august. Usually, these salamanders were missing only a small portion of the tail (i.e., the tail length for salamanders 30 mm SVL was generally greater than 60% of the SVL). Because we found a large proportion of presumed residents (salamanders under cover objects) with some tail loss, residents with at least some portion of the tail missing may be successful at defending cover objects. Based on evidence from laboratory interactions, resident Red-Backed Salamanders with intact tails do not seem to be at high risk of losing territories; resident P. cinereus (collected from Hawksbill Mountain, Shenandoah National Park, Virginia) expelled intruders in 74% of contests ( Jaeger et al., 1982). In 18% of these contests, intruders usurped the territories of residents, whereas in 8% of contests neither resident nor intruder was expelled. However, our results and those of Jaeger and Wise (1998) suggest that those residents experiencing 85% tail loss (tail lengths approximately 11 16% of SVL) may be at increased risk of losing their territories compared to tail-intact residents, if increased aggression by tailed intruders results in greater risk of tailless residents losing territories. Tail loss in salamanders has been shown to influence survival (Ducey and Brodie, 1983) and reproduction (Maiorana, 1977). Such a loss has been hypothesized to regulate population

WISE ET AL. PHEROMONES SIGNAL RESIDENT CONDITION 171 size (Harris, 1989). We suggest that tail loss may influence the distribution of individuals in the environment, providing potential intruders with some opportunities for usurping territories of residents. In summary, we draw the following inferences from our data. (1) Chemical cues present on the substrata of residents convey information to intruders about the tail condition, or some factor associated with tail loss (such as reduced RHP), of residents. (2) Tailed intruders detect and respond to changes in residents as a result of tail loss. (3) Residents alter territorial marking behavior (PCPs) based on their tail condition. ACKNOWLEDGMENTS We thank B. W. Buchanan, S. C. Walls, and three anonymous reviewers for reading earlier drafts of this manuscript. We thank H. M. Wilbur, director of the Mountain Lake Biological Station, for permission to collect salamanders on station property and the Department of Game and Inland Fisheries for permission to collect in the Commonwealth of Virginia (collecting permit 017047). This work was funded in part by the Department of Biology at the University of Louisiana at Lafayette and a Louisiana Board of Regents Doctoral Fellowship LEQSF (1990 95)-GF-19 to SEW through RGJ. Our research program has been approved by the Animal Care and Use Committee of The University of Louisiana at Lafayette (IACUC 2002-8717- 002). LITERATURE CITED ARNOLD, E. N. 1984. Evolutionary aspects of tail shedding in lizards and their relatives. J. Nat. Hist. 18: 127 169. DUCEY, P. K., AND E. D. BRODIE JR. 1983. Salamanders respond selectively to contacts with snakes: survival advantage of alternative antipredator strategies. Copeia 1983:1036 1041. FOX, S. F. 1983. Fitness, home-range quality, and aggression in Uta stansburiana, p. 149 168. In: Lizard ecology: studies of a model organism. R. B. Huey, E. R. Pianka, and T. W. Schoener (eds.). Harvard Univ. Press, Cambridge, MA., AND M. A. ROSTKER. 1982. Social cost of tail loss in Uta stansburiana. Science 218:692 693., N. A. HEGER, AND L. S. DELAY. 1990. Social cost of tail loss in Uta stansburiana: lizard tails as status-signalling badges. Anim. Behav. 39:549 554. FRASER, D. F. 1980. On the environmental control of oocyte maturation in a plethodontid salamander. Oecologia 46:302 307. GOSLING, L. M. 1982. A reassessment of the function of scent marking in territories. Z. Tierpsychol. 60: 89 118.. 1986. Economic consequences of scent marking in mammalian territoriality, p. 385 395. In: Chemical signals in vertebrates 4. D. Duvall, D. Müller-Schwarze, and R. M. Silverstein (eds.). Plenum Press, New York.. 1990. Scent marking by resource holders: alternative mechanisms for advertising the costs of competition, p. 315 328. In: Chemical signals in vertebrates 5. D. W. MacDonald, D. Müller-Schwarze, and S. E. Natynczuk (eds.). Oxford Univ. Press, Oxford. GRIFFIS, M. R., AND R. G. JAEGER. 1998. Competition leads to an extinction-prone species of salamander: interspecific territoriality in a metapopulation. Ecology 79:2494 2502. HARRIS, R. N. 1989. Nonlethal injury to organisms as a mechanism of population regulation. Am. Nat. 134:835 847. JAEGER, R. G. 1980a. Fluctuations in prey availability and food limitation for a terrestrial salamander. Oecologia 44:335 341.. 1980b. Microhabitats of a terrestrial forest salamander. Copeia 1980:265 268.. 1981. Dear enemy recognition and the costs of aggression between salamanders. Am. Nat. 117: 962 974.. 1984. Agonistic behavior of the Red-Backed Salamander. Copeia 1984:309 314.. 1986. Pheromonal markers as territorial advertisement by terrestrial salamanders, p. 191 203. In: Chemical signals in vertebrates 4. D. Duvall, D. Müller-Schwarze, and R. M. Silverstein (eds.). Plenum Press, New York., AND W. F. GERGITS. 1979. Intra- and interspecific communication in salamanders through chemical signals on the substrate. Anim. Behav. 27:150 156., AND J. K. SCHWARZ. 1991. Gradational threat postures by the Red-Backed Salamander. J. Herpetol. 25:112 114., D. KALVARSKY, AND N. SHIMIZU. 1982. Territorial behaviour of the Red-Backed Salamander: expulsion of intruders. Anim. Behav. 30:490 496., J. M. GOY, M.TARVER, AND C. E. MARQUEZ. 1986. Salamander territoriality: pheromonal markers as advertisement by males. Ibid. 34:860 864. MAIORANA, V. C. 1977. Tail autotomy, functional conflicts and their resolution by salamanders. Nature 265:533 535. MARTíN, J., AND A. SALVADOR. 1993. Tail loss reduces mating success in the Iberian rock-lizard, Lacerta monticola. Behav. Ecol. Sociobiol. 32:185 189. MATHIS, A. 1989. Do seasonal spatial distributions in a terrestrial salamander reflect reproductive behavior or territoriality? Copeia 1989:788 791.. 1990. Territoriality in a terrestrial salamander: the influence of resource quality and body size. Behaviour 112:162 174.. 1991. Territories of male and female terrestrial salamanders: costs, benefits, and intersexual spatial associations. Oecologia 86:433 440.

172 COPEIA, 2004, NO. 1 MAYNARD SMITH, J., AND G. A. PARKER. 1976. The logic of asymmetric contests. Anim. Behav. 24:159 175. MILLER JR., R. G. 1981. Simultaneous statistical inference. Springer-Verlag, New York. NUNES, V.DA S., AND R. G. JAEGER. 1989. Salamander aggressiveness increases with length of territorial ownership. Copeia 1989:712 718. RICE, W. R. 1989. Analyzing tables of statistical tests. Evolution 43:223 225. SIMONS, R.R.,AND B. E. FELGENHAUER. 1992. Identifying areas of chemical signal production in the Red-Backed Salamander, Plethodon cinereus. Copeia 1992:776 781.,, AND R. G. JAEGER. 1994. Salamander scent marks: site of production and their role in territorial defence. Anim. Behav. 48:97 103., R. G. JAEGER, AND B. E. FELGENHAUER. 1997. Competitor assessment and area defense by territorial salamanders. Copeia 1997:70 76., B. E. FELGENHAUER, AND T. THOMPSON. 1999. Description of the postcloacal glands of Plethodon cinereus, the Red-Backed Salamander, during bouts of scent marking. J. Morphol. 242:257 269. TABACHNICK, B. G., AND L. S. FIDELL. 1989. Using multivariate statistics. Harper Collins, New York. TOWNSEND JR., V. R., AND R. G. JAEGER. 1998. Territorial conflicts over prey: domination by large male salamanders. Copeia 1998:725 729. WAKE, D.B.,AND I. G. DRESNER. 1967. Functional morphology and evolution of tail autotomy in salamanders. J. Morphol. 122:265 306. WISE, S. E. 1995. Variation in the life history and sociobiology of a territorial salamander. Unpubl. Ph.D. diss., Univ. of Louisiana, Lafayette., AND R. G. JAEGER. 1998. The influence of tail autotomy on agonistic behaviour in a territorial salamander. Anim. Behav. 55:1707 1716. (SEW) DEPARTMENT OF BIOLOGY, UTICA COL- LEGE, UTICA, NEW YORK 13502 4892; (FDV) COLLEGE OF ENVIRONMENTAL SCIENCE AND FORESTRY, STATE UNIVERSITY OF NEW YORK, SYRACUSE, NEW YORK 13210; AND (RGJ) DE- PARTMENT OF BIOLOGY AND INSTITUTE OF COG- NITIVE SCIENCE, UNIVERSITY OF LOUISIANA AT LAFAYETTE, LAFAYETTE, LOUISIANA 70504-2451. (SEW) E-mail: swise@utica.edu. Send reprint requests to SEW. Submitted: 19 June 2002. Accepted: 19 Aug. 2003. Section editor: S. F. Fox. APPENDIX 1 Effect of tail regeneration on behavior. Tail autotomy was induced in all salamanders before testing began. Because tests were conducted over a five-month period, salamanders had differential amounts of regeneration when tested. To measure the effect of regeneration on behavior, we measured the tail lengths of tailless residents and intruders (n 36) on 10 and 22 March 1991 (at the end of the experiment) to determine the proportion of the tail that had regenerated by the end of this experiment. Tailless salamanders had regenerated 61.63 11.38% (mean 1 SD) of their tails. Thus, regeneration of tails may reduce our ability to detect differences in behavior of intruders based on tail condition. Based on the assumption that those salamanders used earlier in the experiment should have less overall regeneration than those used later in the experiment, we examined the impact of regeneration by regressing the amount of time that intruders spent in ATR (aggression) and FLAT (submission) on the number of days from the date that all salamanders were autotomized to the date that they were tested (SPSS 10.0). We used separate regressions for each behavioral pattern and for the three types of contests in which tailless salamanders were used: (1) tailed intruders on the substrata of tailless residents; (2) tailless intruders on the substrata of tailed residents; and (3) tailless intruders on the substrata of tailless residents. We did not reduce alpha for these analyses because we wanted to retain a high probability of detecting possible effects of regeneration on behavior. For all three types of contests, FLAT was log-transformed. For tests where tailless intruders were on the substrata of tailless residents, data for ATR were reflected (inverted) and square-root transformed (Tabachnick and Fidell, 1989). There was no significant relationship between the behavior of intruders (time spent in ATR or FLAT) regressed on testing date for (1) tailed intruders on the substrata of tailless residents (ATR: R 2 0.038, F 0.626, n 18, P 0.440; FLAT: R 2 0.111, F 1.997, n 18, P 0.177); (2) tailless intruders on the substrata of tailed residents (ATR: R 2 0.001, F 0.005, n 18, P 0.943; FLAT: R 2 0.007, F 0.111, n 18, P 0.743); and (3) tailless intruders on the substrata of tailless residents (ATR: R 2 0.020, F 0.323, n 18, P 0.578; FLAT: R 2 0.021, F 0.341, n 18, P 0.567). These results are similar to those found in Wise and Jaeger (1998), when residents and intruders were in direct contact with each other. These results suggest that partial regeneration of the tail did not significantly influence the results of this experiment.