Department of Biology, 167 Castetter Hall, University of New Mexico, Albuquerque, New Mexico 87131, USA Rooster crows are loud, stereotyped vocalizations that appear to function as assertions of social dom- inance over conspecific males in wild Red Junglefowl (Gallus gallus) and their domestic counterparts (Mil- ler 1978, Leonard and Horn 1995). Roosters establish linear dominance hierarchies that reflect individual fighting ability, and dominant males have preferen- tial sexual access to females in a flock (Miller 1978, Leonard and Horn 1995). Honest signaling theory (Grafen 1990, Zahavi 1991) predicts that assertions of status should entail costs that reveal phenotypic quality, rendering such signals accurate, condition-dependent advertise- ments of resource-holding potential (Maynard Smith 1982). Costly signals are sometimes referred to as "handicap" signals (Zahavi 1991, Folstad and Karter 1992). Natural selection should favor honest adver- tisements of resource-holding potential because both participants in such systems (i.e. signal sender and signal receiver) reduce their risk of unnecessarily en- gaging in dangerous fighting behaviors. By an- nouncing high resource-holding potential, dominant individuals reduce the incidence of fights and the fit- ness effects that even victorious animals can face during combat, whereas receivers avoid challenging individuals that they are unlikely to overpower, thereby avoiding injury or death. Parameters of agonistic acoustic displays pro- duced among males in species as varied as tarbush grasshoppers (Ligurotettix planum), toads (Bufo bufo), and red deer (Cervus elephus) predict which contes- tants will prevail in combat (Davies and Halliday 1978, Clutton-Brock and Albon 1979, Greenfield and Minckley 1993; also see Arak 1983). It has long been suspected that the rooster's crow similarly advertises fighting ability (Fisher 1930). Rooster crows recently have been shown to entail no significant metabolic production costs, raising doubts about this signal's honesty (Chappell et al. 1995, Horn et al. 1995). How- ever, other mechanisms may ensure a condition-de- pendent quality of crows, such as immunological handicapping by testosterone (Folstad and Karter 1992), social costs (attacks on crowing subordinates; Leonard and Horn 1995) or neurodevelopmental in- tegrity (Furlow 1997). Hence, rooster crows may honestly advertise resource-holding potential to so- cial competitors despite their low immediate pro- duction costs (Kodric-Brown and Brown 1984). To make a preliminary assessment of whether crows signal resource-holding potential, we ana- lyzed the crows of 20 adult male Red Junglefowl. We measured three condition-dependent markers of phenotypic quality known or presumed to be cor- related with fighting success: (1) body size, (2) size and asymmetry of fighting spurs, and (3) length of fleshy head combs. We then compared the acoustic parameters of crows with these phenotypic mea- sures. Methods.--Twenty sexually mature, 1-year-old male Red Junglefowl were measured and recordings were made of their crows. Rearing conditions are de- scribed in Kimball et al. (1997). The roosters were free-ranging from about six weeks until seven months of age, at which time they were individually penned. At the time of recording, all roosters used in this study had been individually penned for approx- imately five months. Each rooster's comb length and fighting spurs were measured to the nearest 0.1 mm using dial cal- ipers. Standardized body size was obtained by sum- ming the standardized lengths of the humerus, tar- sus, tibia, and femur. The fluctuating asymmetry (FA) of spurs, a measure of developmental integrity and a possible correlate of fighting success in birds (Moller 1992), was calculated as the difference be- tween right and left measures divided by mean spur length (Parsons 1990; see also Kimball et al. 1997). Spur asymmetry met the statistical definitions of FA (Palmer 1994, Kimball et al. 1997). Because spur length and FA were not correlated, we calculated ab- solute (vs. relative) FA values (Palmer 1994). Doing this rendered a half-normal distribution for FA val- ues from our sample (see Palmer 1994). Crows were recorded at a distance of 1 to 3 m with a Radio Shack VSC-2002 recorder and an Electret 33- 3007 unidirectional microphone. Recordings were digitized on an Apple Macintosh computer using Cornell University's Canary 1.1 bioacoustical analy- sis software. Sampling rate was 22 kHz and display range 11 kHz. Sample size for Canary software is set at 8 bits (see Charif et al. 1993). The software was used to generate spectrographs from which the crow's physical features were mea- sured. Crow duration, mean fundamental frequency, and dominant frequency (Hz at peak amplitude) were measured. We did not measure amplitude be- cause of the variation in recording distances (record- ing distance is unlikely to affect relative measures such as which frequency has the highest amplitude, but comparisons of absolute amplitudes of different individuals could be affected). Only one crow was analyzed for each rooster, because junglefowl crow acoustics exhibit high individual stability (Miller 1978). Statistical analyses were performed using SAS 6.04 software. Dominant frequency was split into two cat- egories because the crows of all roosters fell into a low (782 to 869 Hz) or high (1,740 to 2,170 Hz) dom- inant-frequency category. We compared phenotypic differences between the low and high dominant-fre- quency categories. Fundamental frequency and crow duration were continuously distributed; therefore, these data were analyzed with correlation analyses. We used parametric tests for comb length (t-test), standardized body size (Pearson correlation), and spur length (Pearson correlation), and nonparamet- ric tests for spur asymmetry (Wilcoxon rank sum and Spearman correlations). Because we tested for pos- sible relationships among multiple acoustic and physical variables, all P-values were subsequently adjusted with sequential Bonferroni corrections (Rice 1989). Means, ranges, and correlations between trait sizes, as well as calculations of fluctuating asymmetry error, are detailed in Kimball et al. (1997). Results and Discussion.--Roosters that produced low dominant-frequency crows had longer combs (œ = 82.7 ñ SD of 5.1 mm) than those that produced high dominant-frequency crows (œ = 75.3 _+ 3.8 mm; t = 3.6, P < 0.05 after Bonferroni adjustment). No other variables (i.e. body size, spur length, and spur asymmetry) varied significantly with acoustic pa- rameters of rooster crows. Comb length is a condition-dependent signal of plasma testosterone levels and immunological health (Zuk et al. 1990, 1995, Folstad and Karter 1992) and has been shown to correlate with fighting ability in the population of Red Junglefowl used in this study (Ligon et al. 1990). Junglefowl comb turgidity is maintained by connective tissue fibroblasts that pro- duce viscous, capillary-dilating mucoid in response to testosterone (see Ligon et al. 1998). Males with larger combs have relatively high levels of circulat- ing testosterone and relatively low levels of circulat- ing lymphocyte (Zuk et al. 1995), as predicted by the immunological handicap hypothesis (Folstad and Karter 1992). Another galliform species, the Gray Partridge (Per- dix perdix), has a syringeal structure very similar to that of Gallus (Gaunt and Gaunt 1985). Testosterone administration in this species thickens tracheal and bronchial lumina membranes during development (Beani et al. 1995) and lowers the dominant frequen- cy of vocalizations (Fusani et al. 1994, Beani et al. 1995). A similar mechanism may be responsible for the correlation between testosterone-dependent comb length and the dominant frequency of jungle- fowl crows. If so, then crows offer a "snapshot" of androgen levels during syringeal development, whereas comb quality conveys information about a rooster's current immunological and hormonal sta- tus (see Zuk et al. 1995). Hence, crow quality may re- flect an ontogenetic, Zahavian "pay now, display lat- er" strategy. Junglefowl crows may be crudely indic- ative of phenotypic quality, whereas comb quality is a more precise indicator of current condition. Why multiple signals evolve has been the subject of considerable interest (e.g. Moller and Pomian- kowski 1993, Omland 1996, Ligon et al. 1998). In ag- onistic signaling contexts, sequential assessment theory (e.g. Enquist et al. 1990) may help explain the evolution of multiple signals. Sequential assessment theory posits that contestants should escalate ago- nistic encounters from cheap, general signals to more costly and more accurate (i.e. difficult to falsify) sig- nals of resource-holding potential. Well-matched contestants continue to escalate until the interaction becomes violent, whereas power asymmetries are detected during earlier assessment steps such that the individual less likely to win retreats from the en- counter Hence, multiple signals with overlapping information domains in some cases may be ex- plained as components of sequential agonistic-as- sessment displays. The evolved audience of crows in the wild may have included conspecific territory intruders. If dis- tant males were of roughly equivalent quality as that indicated by a territorial male's crow, closer inspec- tion of a more accurate signal--comb size--may have been the next step in assessing the territorial male's resource-holding potential. It is also possible that crowing attracts unaffiliated hens in the wild. In a previous study, Leonard and Horn (1995) found no significant relationship between comb length and fundamental frequency of crows among healthy domestic roosters, but they did identify a significant correlation between status and funda- mental frequency (dominants had higher-pitched crows than subordinates.) The authors did not ana- lyze dominant frequency. Their roosters were inter- acting and had established dominance hierarchies enforced by subordinate-directed attacks. Roosters in our study had been individually penned for five months prior to recording (roosters were not in vi- sual contact but could hear one another's crows, ren- dering their social isolation incomplete). It is possi- ble that aspects of an individual rooster's crow acoustic structure (such as fundamental frequency) vary with social conditions and motivational states, as Leonard and Horn (1995) found to be the case with crow rate, whereas other acoustic characteristics con- stitute a relatively "fixed" marker of hormone levels and other aspects of phenotypic quality during de- velopment. A longitudinal study of crow acoustic quality of roosters whose social status is experimen- tally manipulated is needed. Unfortunately, we were unable to perform social manipulations. Future stud- ies should compare the crows of socially isolated roosters (and phenotypic measures) with those of roosters in dominance hierarchies to differentiate be- tween the effects of social dominance (status) per se versus intrinsic phenotypic quality. The quality of rooster crows is related to comb length--a trait previously shown to be a testoster- one-dependent signal of immunocompetence and a correlate of fighting success in the population of Red Junglefowl that we studied (Ligon et al. 1990, Zuk et al. 1990). Both immunological health and aggres- siveness (another possible correlate of testosterone levels) are probable correlates of fighting success among roosters. The preliminary results that we re- port suggest that resource-holding potential is ac- curately indicated by the dominant frequency of rooster crow vocalizations, despite the low metabolic expense of crowing. Acknowledgments.--We thank J. David Ligon for ac- cess to his junglefowl flock and helpful comments about our hypothesis and interpretations. Matthew Marshall and Rachel Campbell are thanked for lo- gistical support, and Tara Armijo-Prewitt is thanked for her assistance with acoustical data collection. Paul Watson, Randy Thornhill, Astrid Kodric-Brown, Marty Leonard, Andy Horn, Ivan Folstad, Mike Fur- low, and anonymous reviewers are thanked for their critical reviews of this paper. This research was made possible by a previous NSF grant. LITERATURE CITED ARAK, A. 1983. Sexual selection by male-male com- petition in natterjack toad choruses. Nature 306: 261-262. BEANI, L., G. PANZICA, E BRIGANTI, P. PERSICHELLA, AND E DESSi-FULGHERI. 1995. Testosterone-in- duced changes in call structure, midbrain and syrinx anatomy in partridges. Physiology and Behavior 58:1149-1157. CHAPPELL, M. A., M. ZUK, t. H. KWAN, AND t. S. JOHNSEN. 1995. Energy cost of an avian vocal display: Crowing in Red Junglefowl. Animal Be- haviour 49:255-257. CHARIF, g., S. MITCHELL, AND C. CLARK. 1993. Ca- nary 1.1 user's manual. Cornell Laboratory of Ornithology, Ithaca, New York. CLUTTON-BROCK, T. H., AND S. D. ALB¸N. 1979. The roaring of red deer and the evolution of honest advertisement. Behaviour 69:145-169. DAVIES, N., AND T. HALLIDAY. 1978. Deep croaks and fighting assessment in toads, Bufo bufo. Nature 274:683-685. ENQUIST, M., O. LEIMAR, T. LJUNGBERG, Y. MALLNER, AND N. SEGERDAHL. 1990. A test of the sequen- tial assessment game: Fighting in the cichlid fish Nannacara anomala. Animal Behaviour 40:1-14. FISHEI, R. A. 1930. The genetical theory of natural selection. Clarendon Press, Oxford. FOESTAD, I., AND A. KARTER. 1992. Parasites, bright males, and the immunocompetence handicap. American Naturalist 139:603-622. FtRLOw, E B. 1997. Neurodevelopmental integrity and Zahavian bioacoustics. Trends in Ecology and Evolution 12:34. FUSANI, L., L. BEANI, AND E DESSi-FULGHERI. 1994. Testosterone affects the acoustic structure of male call in the Grey Partridge (Perdix perdix). Behaviour 128:301-310. GAUNT, A. S, AND S. L. L. GAUNT. 1985. Syringeal structure and avian phonation. Current Orni- thology 2:213-245 GRAFEN, A. 1990. Biological signals as handicaps. Journal of Theoretical Biology 144:517-546. GREENFIELD, M.D., AND R. L. MINCKLEY. 1993. Acoustic dueling in tarbush grasshoppers: Set- tlement of territorial contests via alternation of reliable signals. Ethology 95:309-326. HORN, A. G., M. L. LEONARD, AND D. M. WEARY. 1995. Oxygen consumption during crowing by roosters: Talk is cheap. Animal Behaviour 50: 1171-1175. KIMBALL, R. T., J. D. LIGON, AND M. MEROLA-ZWART- JES. 1997. Fluctuating asymmetry in Red Jungle- fowl. Journal of Evolutionary Biology 10:441- 457. KODRIC-BROWN, A., AND J. H. BROWN. 1984. Truth in advertising: The kinds of traits favored by sex- ual selection. American Naturalist 124:309-323. LEONARD, M. L., AND A. G. HORN. 1995. Crowing in relation to status in roosters. Animal Behaviour 49:1283-1290. LIGON, J. D., R. T. KIMBALL, AND M. MEROLA-ZWART- JES. 1998. Mate choice by female Red Junglefowl: The issues of multiple ornaments and fluctuat- ing asymmetry. Animal Behaviour 55:41-50. LIGON, J. D., R. THORNHILL, M. ZUK, AND K. JOHN- SON. 1990. Male-male competition, ornamenta- tion and the role of testosterone in sexual selec- tion of the Red Junglefowl. Animal Behaviour 40:367-373. MAYNARD SMITH, J. 1982. Do animals convey infor- mation about their intentions. Journal of Theo- retical Biology 97:1-5. MILLER, D. 1978. Species typical and individually distinctive acoustic features of crow calls of Red Junglefowl. Zeitschrift ffir Tierpsychologie 47: 182-193. MOLLER, A. P. 1992. Patterns of fluctuating asym- merry in weapons: Evidence for reliable signal- ling of quality in beetle horns and bird spurs. Proceedings of the Royal Society of London Se- ries B 248:199-206. MOLLER, A. P., AND A. POMIANKOWSKI. 1993. Why have birds got multiple sexual ornaments? Be- havioral Ecology and Sociobiology 32:167-176. OMLAND, K. E. 1996. Female Mallard mating pref- erences for multiple male ornaments. 1. Natural variation. Behavioral Ecology and Sociobiology 39:353-360. PALMER, g. 1994. Fluctuating asymmetry analyses: A primer. Pages 335-364 in Developmental insta- bility: Its origins and evolutionary implications (T. Markow, Ed.). Kluwer Publishers, Dordrecht, The Netherlands. PARSONS, P. 1990. Fluctuating asymmetry: An epige- netic measure of stress. Biological Reviews of the Cambridge Philosophical Society 65:131-145. RICE, W. 1989. Analyzing tables of statistical tests. Evolution 43:223-225. ZAHAVI, A. 1991. On the definition of sexual selec- tion, Fisher's model, and the evolution of waste and of signals in general. Animal Behaviour 42: 501-503. ZUK, g., T. S. JOHNSEN, AND T. MACEARTY. 1995. En- docrine-immune interactions, ornaments and mate choice in Red Jungle fowl. Proceedings of the Royal Society of London Series B 260:205- 210. ZUK, g., K. JOHNSON, R. THORNHILL, AND J. D. LI- GON. 1990. Parasites and male ornaments in free-ranging and captive Red Junglefowl. Be- haviour 114:232-248. Received 25 April 1997, accepted 3 November 1997. Associate Editor: E. Greene