am ± ;;v ... •' •• •' ■ ZOOLOGIC A SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 58 • ISSUE I • SPRING 1973 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President Charles W. Nichols, Jr. Vice-President John Pierrepont T reasurer Augustus G. Paine Secretary William G. Conway General Director ZOOLOGICAL PARK William G. Conway , Director & Chairman, Dept, of Ornithology ; Joseph Bell, Curator, Ornithology ; Donald F. Bruning , Associate Curator, Ornithology ; Hugh B. House, Curator, Mammalogy; James G. Doherty, Associate Curator, Mammalogy; F. Wayne King, Curator, Herpetology; John L. Behler, Assistant Curator, Herpetology; Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus; Grace Davall, Curator Emeritus AQUARIUM James A. Oliver, Director; George D. Ruggieri, S.J., Associate Director; William S. Flynn, Curator; H. Douglas Kemper, Associate Curator; Christopher W. Coates, Director Emeritus; Louis Mowbray, Research Associate, Field Biology OSBORN LABORATORIES OF MARINE SCIENCES George D. Ruggieri, S.J., Director & Experimental Embryologist; Ross F. Nigrelli, Senior Scientist; Martin F. Stempien, Jr., Assistant to the Director & Bio-Organic Chemist; Jack T. Cecil, Virologist; Paul J. Cheung , Microbiologist; Joginder S. Chib, Chemist; Kenneth Gold, Marine Ecologist; Myron Jacobs, Neuroanatomist; Klaus D. Kallman, Fish Geneticist; Kathryn S. Pokorny, Electron Microscopist; Eli D. Goldsmith, Scientific Consultant; Erwin J. Ernst, Research Associate, Estuarine & Coastal Ecology; Martin F. Schreibman, Research Associate, Fish Endocrinology CENTER FOR FIELD BIOLOGY AND CONSERVATION George Schaller, Research Zoologist & Co-ordinator; Thomas Struhsaker, Roger Payne, Research Zoologists; Donald F. Bruning, Research Associate ANIMAL HEALTH Emil P. Dolensek, Veterinarian; Consultants: John Budinger, Pathology; Ben Sheffy, Nutrition; Gary L. Rumsey, Avian Nutrition; Kendall L. Dodge, Ruminant Nutrition; Robert Byck, Pharmacology; Jacques B. Wallach, Clinical Pathology; Edward Garner, Dennis F. Craston, Ralph Stebel, Joseph Conetta, Comparative Pathology & Toxicology; Harold S. Goldman, Radiology ; Roy Bellhorn, Paul Henkind, Alan Friedman, Comparative Ophthalmology; Lucy Clausen, Parasitology; Jay Hyman, Aquatic Mammal Medicine; Theodore KazimirofF, Dentistry; Alan Belson, Resident in Pathology Z00L0GICA STAFF Simon Dresner, Editor & Curator, Publications and Public Relations Joan Van Haasteren, Associate Editor & Assistant Curator, Publications & Public Relations Louise A. Purrett, Assistant Editor & Feature Writer F. Wayne King Scientific Editor AAZPA SCIENTIFIC ADVISORY COMMITTEE Ingeborg Poglayen, Louisville (Kentucky) Zoological Gardens, Chairman; William G. Conway, New York Zoological Society; Gordon Hubbell, Crandon Park Zoo, Miami; F. Wayne King, New York Zoological Society; John Mehrtens, Columbia (South Carolina) Zoological Park; Gunter Voss, Metro Toronto Zoo, Canada © 1973 New York Zoological Society. All rights reserved. Observations on the Acoustic Behavior of Crocodilians (Figures 1-6) Howard W. Campbell Department of Zoology, University of Florida, Gainesville, Florida 3260 1 1 Laboratory and field observations were obtained for some acoustic behaviors of Alligator mississippiensis , Caiman croeodilus , Crocodylus acutus , and Melanosuchus niger. Tape re- cordings and sonographs of several call types were obtained. The young of each species vocalize with a characteristic "bark,” which appears to function to announce a threat or food or some other environmental perturbation of significance. No species-specific differences in this call were observed and the spectral characteristics of the call did not vary, except in intensity, with the situation. Introduction There is ample documentation of the extent of vocal behavior among the croco- dilians; few accounts of any species fail to note its occurrence. Crocodiles are reported to roar and bellow, to hiss, to bark, and to chirp and grunt under a wide variety of ecological and social circumstances. To date, however, no one has attempted to collate these various reports and construct anything approaching an “acoustic ethogram” for any of the species. The vague, con- tradictory, and often wildly imaginative accounts that comprise much of the literature on crocodil- ians present a serious hindrance to any such attempt. This report evaluates some aspects of the eco- logical and social significance of the acoustic behavior in several species of crocodilians from which laboratory and/or field data were person- ally obtained. Recordings of the spontaneous vocalizations and of the vocal responses to sev- eral experimental situations were obtained from each species in the laboratory and these responses have been verified in natural populations of all species except one (Melanosuchus niger). Re- cordings were usually made with a Sony 400 tape recorder at speeds of three-and-one-quarter or seven-and-one-half inches per second (Ips). Ex- 1 Present address: Office of Endangered Species, Department of the Interior, Bureau of Sport Fisheries and Wildlife, Washington, D.C. 20240. ceptions will be noted. Recordings were analysed using a Kay Electric Co., Missilyzer sound spec- trograph. I am most appreciative of the kindness of sev- eral individuals in allowing me access to crocodil- ians in their care, particularly Peter C. H. Pritchard and Jill Goodman. A. Stanley Rand generously assisted me during my field work in Panama and allowed me to report on recordings he obtained. The staff of the Middle America Research Unit, Canal Zone, was of invaluable assistance in every phase of my activities in Panama. Field work in Panama was supported by NIH Grant Ex-00139 to the Center for the Biol- ogy of Natural Systems, Washington University. Support for field work in Mexico was from the American Philosophical Society, Penrose Fund, and the New York Zoological Society. The Theo- dore Roosevelt Memorial Fund of the American Museum of Natural History supported, in part, my studies in Florida. William E. Evans, Naval Undersea Research and Development Center, San Diego, has been of frequent encouragement and assistance in my acoustic studies. Alligator mississippiensis The American alligator, both as an adult and when young, is highly vocal. A variety of sounds are produced; the adults characteristically “roar” or “bellow” while the young produce a variety of sounds usually referred to as “barks” or “grunts.” The functions of the various calls are open to some question and a diversity of opinions abound. 2 New York Zoological Society: Zoologica, Spring 1973 The adult “bellow,” for example, has variously been considered a mating call, usually ascribed to the male in the breeding season. Neill (1971), however, has pointed out that the call is given by both sexes and at various times of the year. My own observations would support this position, although the call does appear to be involved in courtship and is given by both sexes at this time (Carr, pers. comm.). All references to these calls are anecdotal and no detailed acoustic analysis has been undertaken to determine whether or not differences exist between the bellows of males and females or between the males of different sizes or social status. Such an analysis may reveal differences in the calls which encode such infor- mation. It is, of course, not unusual to find an animal call which transmits information in a social con- text which varies as a function of the circum- stances attending its production and reception. The breeding song of male frogs, for example, encodes vastly different information for another male of the species, an egg-laden female, and a spent female. In squirrel monkeys the relative social status of the recipient individual has been shown to determine the message content of acoustic signals (Winter, Ploog, and Latta, 1966). A similar relationship may well be involved in the social use of the bellow by alligators, both for territoriality and courtship, depending on the hor- monal state and social experience of the individ- uals involved. The sounds produced by the young also appar- ently serve diverse functions. They are reported to attract the adult (parent only?) to the defense of the young when they are endangered or alarmed (Neill, 1971). Vocalizations produced by the young before they emerge from the egg are believed to attract the female to the nest and stimulate her to open it to assist in the emergence process (Mcllhenny, 1935). These have alterna- tively been interpreted as a synchronizing stimu- lus to assure a well coordinated hatch (Lee, 1968). Inasmuch as the female is known to open the nest, on many occasions at least, at the proper time, the first interpretation seems most probable. There is no data to support the contention that hatching in alligator eggs is well coordinated under natural conditions. It is not inconceivable, however, that both functions are involved: the audible vocalizations which begin shortly before hatching serving to attract the female to the nest, and the earlier vibratory stimuli or subaudible vocalization serving to regulate developmental synchrony at earlier periods, much as is the case in many birds (Vince, 1969). The pre-hatching vocalizations are clearly audible for a distance of 50 feet or more from the nest, even before it has opened, and might well attract predators at a highly vulnerable stage in the species’ life history. The adult’s presence at the nest and protective response to these vocal- izations would certainly be an advantage during this process. It would appear, in fact, that the possibility of attracting predators would consti- tute a selective disadvantage for this behavioral trait were it not for the presence of the adult at the nest at this time. In the laboratory, young Alligator vocalized readily under several circumstances: during the excitement of group feeding; when startled, frightened, or grasped; and in response to the vocalizations of other crocodilians, either in the flesh or played back from a tape recorder. In the laboratory the initiation of vocalization via tape- recorded playback would result in a series of vocal responses from the captive individuals and a movement toward the loudspeaker. This vocal response and orientation behavior of young alli- gators could be elicited by a variety of stimuli in addition to the playback of their own calls. The recorded calls of Caiman crocodilus, Melano- suchus niger, and Crocodylus acutus all were equally effective, as were recordings of several lizard vocalizations, the aggressive bark of Gecko gekko , and the call of Ariestelliger praesignis. In the field, vocalizations were usually accom- panied by short lunges away from the source of disturbance. This would be repeated throughout the group of young with a resulting net group movement away from the disturbance. This vocal- ization, the characteristic “gnuc,” or “bark,” of the young, is illustrated in figure 1. No consistent differences detectable on sonograms can be ob- served between calls given in response to other calls, at feeding stations, or under alarm situa- tions. The one call appears to serve to alert the group members to some environmental charac- teristic of interest, food, a potential predator, etc., and to maintain some group cohesiveness in the response. The frequent references in the general literature to “distress” and “alarm” calls would appear to reflect the writer’s interpretation of the call’s significance rather than the alligator’s. The groups of similar aged young (= pods) remain together for at least one season, and per- haps two or three, but nothing is known regard- ing their composition, whether formed from the young of only one nest or from several; their internal social structure and internal stability, whether pods exchange individuals or not; or what survival function, if any, they subserve. The above discussion, however, would strongly sug- gest that acoustic signals are significant in main- taining intragroup stability and contact. At night there appears to be an increase in low amplitude spontaneous vocalization by members of the pods. Inadequate data are available to substantiate this point, but one pod monitored for seven hours over three nights and six hours over two days averaged 20 such spontaneous vocalizations (i.e., no environmental stimulus apparent to the inves- Campbell: Acoustic Behavior of Crocodilians 3 ZHM ui Aousnbajj Figure 1. Two separate juvenile calls of a hatchling Alligator mississippiensis. Time in Seconds 4 New York Zoological Society: Zoologica, Spring 1973 tigator) per hour during the night observation periods and only 1 1 per hour during the daylight periods. This suggests that these low amplitude signals are utilized to a greater degree when visi- bility is reduced. Caiman crocodilus Young Caiman are highly vocal in captivity, though perhaps not as much so as young alliga- tors. Caiman vocalized readily when startled or restrained and in the midst of group feeding activity, but were less inclined to respond to the vocalizations of other crocodilians. Whereas alli- gators would respond to a variety of recordings, Caiman would only rarely respond to any record- ings and then only to recordings of their own species. These observations, while preliminary, may indicate a selective advantage for a species- specific response to acoustic signals in an area where several crocodilian species are in ecologi- cal contact (South America) as opposed to an area where only one species exist (central Flor- ida). This, while an appealing hypothesis, is not supported by obvious differences in the calls of the various associated species which might serve for species identification (figure 6). The behavior associated with the vocalizations was identical to that described for Alligator, a vocal response associated with a series of short lunges; this would be picked up by the group with a resulting net group movement. Under natural conditions in the field in Panama, the response to an intruder was fre- quently quite different. On being approached there, usually in ponds or river swamp, most individuals would quietly, or after one vocaliza- tion, submerge several inches, back up a few inches, and turn to the side, much as described for Crocodylus novaeguineae (Neill, 1971). When surprised on land, however, or in very shallow water, escape lunges associated with vocalizations were normally observed. The method of escape utilized appeared to correlate with the depth of water in which the individual was located and, perhaps, with the suddenness of the appearance of the investigator-predator. Inasmuch as auditory determinations are avail- able for this species, it may be profitable to compare the known auditory sensitivity as deter- mined by electrophysiological methods with the frequency parameters of their call. Manley (1970), studying single units in the cochlear nucleus; Konishi and Campbell (unpubl. data), using the N, response recorded at the round window; and Wever (1971), using the micro- phonics recorded at the round window, deter- mined the frequency zone of maximum sensitivity to lie approximately between 0.2 and 2.0 KHz. This is quite in line with the frequency param- eters of the call as illustrated in figure 2. Here the fundamental begins at about 0.7 KHz and sweeps downward to about 0.2 KHz in one-ninth of a second. Several harmonics are present, the first beginning at 1.3 K Hz and sweeping downward to 0.3 KHz, and the second beginning at 1.7 KHz and sweeping down to 0.6 KHz. Crocodylus acutus Judging from the literature and the behavior of captive young, the American crocodile is much less vocal than either of the preceding species. The young do vocalize before and during the hatching process, however, and recordings of this behavior were obtained by Dr. A.S. Rand of the Smithsonian Tropical Research Institute in the Canal Zone. Vocalization began prior to the actual emergence from the eggs and continued throughout the process. Figure 3 illustrates one of the calls from Rand’s recording. The fundamental begins at about 0.6 KHz and downshifts to about 0.3 KHz in one-fifth of a second. Several harmon- ics are strongly indicated. To the observer the call sounds highly similar to the distress call of very young Alligator mississippiensis and Caiman crocodilus and to the calls produced by hatching Alligator. Young American crocodiles also utilize a “contact” and/or “distress” call under much the same laboratory circumstances as reported above for Alligator mississippiensis and Caiman croco- dilus. Crocodylus acutus frequently utters a “snarl” when attacking non-prey objects. This has been observed on several occasions with different individuals, all under three feet in total length. This call is usually given with mouth agape, facing the object, and immediately precedes, or is concurrent with, an attacking lunge. Figure 4 illustrates an example of this call given by a 30- inch specimen from Jamaica when attacking a large turtle which had been placed in its home tank. The call begins at a frequency of 0.3 KHz in one-fourth second, then changes to 1.5 KHz for one-fourth second before abruptly dropping off. Similar “snarls” have been reported in aggres- sive situations for other crocodilian species, C. niloticus (Cott, 1960) and Melanosuchus niger (Neill, 1971), for example. Interestingly, if the initial aggressive response does not result in an alteration of the “threat” situation, the crocodile will alter its behavior to attempted escape or avoidance from the threat- ening object and begin to vocalize with the juve- nile call. Figure 5 illustrates a juvenile call, or “grunt,” of the above mentioned Jamaican indi- vidual after its initial attacks on the large turtle were unsuccessful. These calls were uttered as the crocodile attempted to climb out of its pen or sat submerged and tail-to the turtle. This call is very similar to the so-called distress calls re- corded from the other species, beginning at 0.5 KHz and downshifting to approximately 0.2 KHz Campbell: Acoustic Behavior of Crocodilians 5 *HM ui Aouanbajj Figure 2. Two separate juvenile calls of a 12-inch Caiman crocodilus. Time in Seconds 6 New York Zoological Society: Zoologica, Spring 1973 ZH>| w Aouanbajj Figure 3. Hatching call of Crocodylus acutus recorded in Panama by A. S. Rand. Individual was partially emerged from the egg. Time in Seconds Campbell: Acoustic Behavior of Crocodilians 7 zH>j ui ^ouanbaij Figure 4. Aggressive growl or snarl of 30-inch Crocodylus acutus. Time in Seconds New York Zoological Society : Zoologica, Spring 1973 ZHM ui ^ouanbaj-j Figure 5. Juvenile call of 30-inch Crocodylus acutus (same recording session as figure 4). Time in Seconds Campbell: Acoustic Behavior of Crocodilians 9 Figure 6. Response of an 18-inch Melanosuchus niger (second call) to a playback of the juvenile call of a 12-inch Caiman crocodilus (first call). 10 New York Zoological Society: Zoologica, Spring 1973 over one-third of a second. This full sequence of behaviors was observed on two occasions with this Jamaican specimen and has since been ob- served in several other individuals. MELANOSUCHUS NIGER One specimen of Melanosuchus was available for study. It was an 18-inch juvenile and had been in captivity for 1 1 months in the company of other crocodilians prior to its use. Overall, this individual was much less inclined to vocalize than the specimens of Alligator mis- sissippiensis and Caiman crocodilus studied. Numerous attempts to obtain recordings were unsuccessful. The specimen would grunt or “bark” on occasion when disturbed, but record- ings were not obtained. The specimen showed a positive orientation to playbacks of the calls of Alligator and Caiman crocodilus , orienting to- ward the speaker and moving to the close end of the tank, but on only one occasion did it vocalize in response to any recordings. The vocal response to a call of Caiman crocodilus is illustrated in fig- ure 6. The first signal is that of the Caiman , the second that of the Melanosuchus. The call is very similar to that of the Caiman but extends to lower frequencies, to 0.2 KHz as opposed to 0.3 KHz. This reflects the larger size of the Melanosuchus rather than any species-specific difference in the calls. As crocodilians grow, their calls, at least the juvenile calls, deepen and include an increased range in the lower frequencies. The calls of 20- inch Caiman crocodilus also extend down to 0.2 KHz as does the juvenile call of Crocodylus acu- tus illustrated in figure 5, from a 30-inch speci- men. This relationship between body size and pitch has previously been described in other animal groups (Collias, 1960). Discussion Crocodilians vocalize under a variety of cir- cumstances, both in the field and in the labora- tory. While species-specific variations in vocal tendencies are suggested, the data currently are too anecdotal and incomplete to allow any con- clusions on this point. Several types of calls are well established: the bellowing of adults, of un- verified function; and a variety of calls produced by juveniles. There appears to be no obvious dis- tinction in structure between the juvenile calls of the various species examined, and the behavioral and environmental correlates of the vocalizations are similar in all species for which data are available. The most often reported, and most easily evoked, vocalization of the young crocodilian is the “distress” call, a segmented call with an initial down sweep in frequency followed by a short plateau. This resembles the segmented “ground predator” or down sweeping “distress” calls as described by Collias (1960), and appears to function to alert nearby individuals and per- haps to coordinate the group escape response, as well as attracting adults to the area. The vocalizations produced by hatching indi- viduals of Crocodylus acutus resemble the down sweeping distress call of Collias (1960) and func- tion to attract the female parent to the hatching site. Species-specific differences in this behavior may be present, for example Alvarez del Toro (1969) reports both parents participating in the emergence process in Caiman crocodilus while only the female is implicated in Crocodylus nilot- icus (Cott, 1960) and the situation in Alligator mississippiensis is open to debate (Neill, 1971), or variable. More careful observations on this point are needed. The other vocalizations reported here, the aggressive “snarl” of Crocodylus acutus , and the call given in the presence of food by all species examined, which does not appear to differ from the “contact” or “distress” call, are mentioned throughout the literature for a number of other species. In addition to those mentioned here, a number of other vocalizations have been reported for crocodilians ranging from “coughs” to hisses. At the present time it is difficult to assess the social or ecological function of most reports of crocodilian vocalizations despite the frequency with which these reports occur. The evidence does indicate that the crocodilians are a highly vocal group which utilize acoustic signals in a wide spectrum of behavioral /ecological contexts, and future studies in this area should prove highly rewarding. Campbell: Acoustic Behavior of Crocodilians Literature Cited Alvarez del Toro, M. 1969. Breeding the spectacled caiman in the Tuxtla Gutierrez Zoo. lnt. Zoo Yrbk., 9:35-36. COLLIAS, N. 1960. An ecological and functional classifica- tion of animal sounds. In Animal sounds and communication. W. E. Lanyon and W. N. Tavolga, eds. AIBS., pp. 368-391. Cott, H. B. 1960. Scientific results of an inquiry into the ecology and economic status of the Nile Crocodile (Crocodylus niloticus) in Uganda and Northern Rhodesia. Trans. Zool. Soc. London, 29: 211-356. Lee, D.S. 1968. Possible communication between eggs of the American alligator. Herpetolog- ica, 24( 1): pp. 88. Manley, G. A. 1970. Frequency sensitivity of auditory neur- ons in the caiman cochlear nucleus. Z. vergl. Physiol., 66: 251-256. McIlhenny, E. A. 1935. The alligator’s life history. Christopher Publishing House, Boston. 117 pp. Neill, W.T. 1971. The last of the ruling reptiles. Columbia University Press, New York. pp. ix-xvii, 1-486. Vince, M. A. 1969. Embryonic communication, respiration and the synchronization of hatching. In Bird vocalizations, R. A. Hinde, ed. Cambridge University Press, Cam- bridge. pp. vii-xiii, 1-394. Wever, E. G. 1971. Hearing in the crocodilia. Proc. Nat. Acad. Sci., U.S.A., 68(7): 1498-1500. Winter, P., D. Ploog, and J. Latta 1966. Vocal repertoire of the squirrel monkey ( Saimiri sciureus), its analysis and sig- nificance. Exp. Brain Res., 1: 359-384. 2 The Behavior of the Red-winged Tinamou, Rhynchotus rufescens (Figures 1-5, Table 1) Sam E. Weeks Laboratory of Ornithology, Cornell University, Ithaca, New York 14850 An ethological study was done on captive red-winged tinamous, Rhynchotus rufescens. Descriptions of their behavioral repertoire are given, including maintenance, locomotory, feeding, agonistic, vocal, and reproductive behavior. Some qualification and comparison of behavior with other species of tinamous are made. Introduction The red-winged tinamou is one of approximately 45 species of tinamous which comprise the monofamilial avian order, Tinamiformes (Meyer de Schauensee, 1966). The family, Tinamidae, is neotropical. All the species, however, are found in South Amer- ica, and only five have discontinuous distributions into Central America with four of these extend- ing as far north as Mexico. Within this wide area, they are the dominant ground birds, filling most of the terrestrial habitats from Mexico, south to Tierra del Fuego. The tinamids are divided into nine distinct genera (Meyer de Schauensee, 1966). These are listed on Table 1 with the num- bers of species, comparative size of the birds, habitats, and distributions. They range in size from as little as a small quail to as large as a domestic chicken. Their coloring and behavior render them extremely cryptic and may explain why they have been so little studied. Usually males and females are similarly marked. In general appearance and behavior they may be imagined to resemble something between a rail and a gallinaceous bird. Although equipped with a keeled sternum and capable of flight, their closest relatives are most probably among the ratites. The Rheiformes seem the most likely (Simpson, 1958), for they share such things as certain behavioral traits (sole male incubation and some displays), some structures (copulatory organs, aftershafted feath- ers, fine parts of the palate and rhampotheca; Parkes and Clark, Jr., 1966), and a not too dis- similar restricted distribution. Both tend to have a very reduced tail. They also share some egg shell characteristics (Tyler and Simkiss, 1959). Clay (1957), however, did not find Mallophaga helpful in elucidating the relationship between these two groups. Jehl (1971) could find nothing in the comparative plumage patterns of chicks that would indicate an affinity with any of the ratites, while Sibley and Frelin (1972), on the basis of egg-white proteins, would place tina- mous near the kiwis. As with much of the world’s fauna, these birds are facing a double assault on their existence by man’s ever increasing population. Their eggs and flesh are eaten by man. The meat in particular is considered a delicacy. But most important is the destruction of habitats which denies these birds a place in which to live. Many of these forms are headed for extinction in the not too distant future. Shy and secretive, tinamous are reluctant fliers, preferring to run and hide when disturbed. A few species have been reported to perch or roost in trees (Skutch, 1959; and Wetmore, 1965) or to fly spontaneously without being hard pressed (Pearson and Pearson, 1955; and Lancas- ter, 1964). Tinamous tend to be omnivorous, eat- ing seeds, fruits, insects, other invertebrates, and small vertebrates. 13 14 New York Zoological Society: Zoologica, Spring, 1973 Table 1. Summary of the Family, Tinamidae, by Genera Genera Number of Species Size of Birds Habitat Distribution 1. Tinamus 5 Targe Tropical and subtropical forests Southern Mexico, Central America, northern and central South America 2. Nothocercus 3 Medium Tropical and subtropical forests Central America, northwestern South America 3. Crypturellus 20 Medium to small Tropical and subtropical forests Southern Mexico, Central America, central and northern South America 4. Rhynchotus rufescens 1 Targe Tall grasslands and pampas Brazil, Bolivia, Para- guay, Uruguay, and Argentina 5. Nothoprocta 6-7 Medium From semiarid thorn forests up to sub- alpine Puna The Andes from Ecuador south to Chile and Argentina 6. Nothura 5 Small Short grasslands Southern South America 7. Taoniscus nanus 1 Very small Secondary forest- savannah Sao Paulo State of Brazil 8. Eudromia 2 Targe Arid regions Southern South America 9. Tinamotis 2 Targe From temper- ate Puna up to alpine regions Southern Andes from southern Peru south to Chile and Argentina A unique feature of tinamous is the nature of their eggs which have a very glossy, mirrored sur- face. Although tinamous are ground nesters, the eggs are one solid color without the protective camouflage that one might expect. Even the colors seem maladaptive as, depending on the species, they are pastel pink, blue, pea green, or various shades of violet or brownish-purples. The function and evolutionary significance of the gloss and colors is unknown but its maintenance has required some behavioral adaptations. For those species which have been studied, it is usual for more than one female to contribute to a single clutch of eggs which is solely incubated by the male. The young are precocial and remain for a time with the attentive male. Methods, Animals, and Facility The purpose of this investigation was to de- scribe the behavioral repertory of the red-winged tinamou and also to obtain some comparative data from the other tinamid species present. I have observed the following species of tina- mous in captivity: Crypturellus obsoletus Crypturellus noctivagus Rhynchotus rufescens Nothoprocta ornata Nothoprocta perdicaria Nothoprocta pentlandii Nothura darwinii (one specimen) Nothura maculosa Eudromia elegans Brown tinamou Yellow-legged tinamou Red-winged tinamou Ornate tinamou Chilean tinamou Andean tinamou Darwin’s nothura Spotted nothura Elegant crested tinamou Of the over 200 observational hours for this study, most were spent on six individuals of Rhynchotus rufescens , the red-winged tinamou. These birds were wild-caught, presumably as young in the state of Sao Paulo, Brazil. Individ- uals were marked with leg bands and were wing- tagged with colored nylon Saflag fabric fastened around the humerus. Weeks: Red-winged Tinamou, Rhynchotus rufescens 15 The aviary was a 6.4 x 30.5 m greenhouse (Sheldrake, 1969) consisting of a wooden frame covered with plastic. It was heated with two thermostatically controlled, forced-air gas heat- ers, one at each end. Ventilation was accom- plished by a thermostat and a time controlled two-speed fan. This operated in such a way that for five minutes of every hour the fan was oper- ated by the timer, regardless of the temperature. Anytime other than this, when the temperature inside the aviary rose above the thermostatic set- ting, the fan was again activated. Fresh, incoming air was equally distributed throughout the aviary by means of a perforated plastic tube, five feet in circumference, that ran the length of the ceiling. This flexible tube re- mained closed until the negative pressure caused by the exhaust fan’s operation inflated it with air from the outside. This arrangement functioned to prevent birds from being exposed directly to a draft, especially important during the cold winter temperatures. Initially the red-winged tinamous had free run of the entire eviary with other species of tina- mous. During the first winter the birds contracted a respiratory viral infection. While this was fatal to some species, most of the red-winged tinamous were only mildly affected, but a few required force feeding for a time. With one indicated ex- ception, no data from sick individuals were used in this paper. Rhynchotus rufescens spent the first summer in an. outdoor pen. They were then given their own enclosure within the aviary. The floor plan and dimensions are shown in figure 1. Rye grass seed was sown over a portion of the aviary. Passion flower vines (Passiflora species) grew along and up the walls. Some locally wild grasses such as timothy were transplanted for additional cover, as well as clovers, mustard, etc. The probing and grazing behavior of these ani- mals required a regular program of planting and transplanting to keep green areas. To simulate tall grass clumps, long dried grass stems and dried forbs were gathered and tied into bundles with a stake in the middle. These “artificial grass clumps” were placed along the walls and in line- of-sight rows which converged on my observation station. They gave the birds a sense of cover with- out obscuring them from view. In addition, to further break up the visual field of the birds which could see each other most of the time, low cardboard and burlap baffles that were also laid out in line-of-sight orientation were erected. When the birds could no longer see each other, there was marked increase in their vocalizations. Most of the data on the red-winged tinamou were collected after the birds had been in captiv- ity for about 18 months, five of which they had spent in their own enclosure. Tinamous are very difficult to observe in the wild. Observations in the field are short term, chance sightings, or perhaps the result of luring birds briefly out from cover with an imitation of their call (Lancaster, 1964a). Lancaster (1964a) was able to determine the direction of movement of a calling male and put himself in its path in order to observe it as it passed. Continuous obser- vations would be very rare with most species. Pearson and Pearson (1955) had some success in open, short grass grazing areas in Peru with Nothoprocta ornata , and Schafer (1954) suc- ceeded with Nothocercus bonapartei when the birds accepted him as a part of the habitat. Other investigators in the field were not so fortunate (Lancaster, 1964a, 1964b, and Beebe, 1925). In 1892, Hudson described the habitat of Rhynchotus rufescens as being largely dominated by only a few coarse grass species, 1 to 2.5 m in height and growing in large tussocks. The most notable of these are Cortaderia (- Gynerius ) species. More recently. Bump and Bohl ( 1965) state that tall (40 to 90 cm) clump grass is a favor- ite cover for these birds. It is these grass tussocks that structure the environment and the behavior of the red-winged tinamou. Alterations of the habitat to facilitate observations would probably render it unsuitable for the birds. Hudson (1920) reported that wherever the shorter European varieties of grass were planted, the red-winged tinamou disappeared. A field study of this species, then, would probably require many hours spent for very little behavioral data. Complete se- quences of behavior might never be seen, and manipulations of the birds or the environment would be difficult or impossible. Wetmore (1926) mentions Rhynchotus rufescens as being difficult to see and collect in areas where they were common. Aviary studies do have shortcomings, such as the possibility of disease and the chance that the unnatural environment will cause unnatural be- haviors. One can obtain an accurate picture of most if not all of the behavioral repertoire of many birds from captive studies, if he has an understanding of their habitat and general be- havior. There may be some behaviors which will not take place in an aviary due to the lack of proper stimuli, such as the reaction to certain predators, and there may be some behaviors caused in captivity due to stimuli not found in the wild (Kaufmann and Kaufmann, 1963). The lat- ter sometimes are of interest in understanding the evolution of behavior. Work both in the field and in the aviary are important and the discrep- ancies between the two are of heuristic value. The captive data which predictably vary the most from field observations are those of frequencies of behavior, while descriptions of species-specific behavioral patterns should be very similar. This paper is mainly concerned with these behavioral 16 New York Zoological Society: Zoologica, Spring, 1973 £ CN Figure 1. Floor plan of the Rhynchotus rufescens aviary. Weeks: Red-winged Tinamou, Rhynchotus rufescens 17 patterns. In many cases of this kind, it is more efficient to have the aviary study precede the field work because familiarity with a species as to how it hides, its vocalizations, signs of feeding, droppings, etc., aid in locating it in the field and the recognition of incomplete behavioral se- quences maximizes the data collected. The shortcomings of the aviary can be made to work in one’s favor. Assuming that behavior has changed only quantitatively and not qualitatively in captivity, behaviors rare in the wild may be repeated over and over by caged birds. This allows very careful recording of the behavior and sometimes the determination of what is causing it to happen so often. By manipulation one may reveal something of the dnve(s) controlling this particular event. Also the birds may be bred out of season by light stimulation. Photography and other means of recording behavior are more eas- ily accomplished in captivity since one knows where, within limits, it will occur and can in some cases increase the probability of its occurrence. Caution should be used in predicting what goes on, and how, in the wild based solely on captive studies (Kaufmann and Kaufmann, 1963). Description Rhynchotus rufescens is tan, brown, tawny, and black in a cryptic, barred pattern (see figure 2). This, combined with the animal’s preference for sitting beside and partly under the overhang of a clump of grass, makes it extremely hard to see. The rufous of the primaries and initial sec- ondaries is seen only when the wings are ex- tended. Adults weigh about one kilogram, with the females tending to outweigh the males. Other- wise the sexes are identical. Positive sex identifi- cation by cloacal examination was attempted without much success in the red-winged tinamou, but was useful in other species (Bump and Bohl, 1965; and Bump and Bump, 1969). The sex of the individual Rhynchotus rufescens was determined behaviorally. Considerable variation has been reported in the color of these birds from different geographic locations (Wetmore, 1926; Laubmann, 1930; and Naumburg, 1930). The birds from the state of Sao Paulo, Brazil, are darker than specimens from the north or farther south. Laubmann (1930) states that the birds farther south take on a more grayish cast. This probably reflects areas in which the climate is drier and the grayer tone of the plumage is adaptive. The birds used in this study came from southeastern Brazil and showed considerable variation in coloring. These speci- mens are now at the Laboratory of Ornithology, Cornell University, Ithaca, New York. Locomotion and Stationary Postures Rhynchotus rufescens has strong legs with what appear to be small feet in comparison with other terrestrial birds of similar size. Its feet likely have become reduced as an adaptation for cursorial locomotion through dense vegetation. Walking and running As the bird walks the head is moved back and forth in a horizontal, anterior-posterior direction in a mid-sagittal plane, similar to many walking birds. Bangert ( I960) studied the coordination of the head movements with foot movements in the domestic chicken. These head movements prob- ably function in giving a bird a series of stable visual images (Daanje, 1950). In Rhynchotus rufescens , these head movements are always associated with walking and never occur without foot movements as they do in some other tina- mous, Nothura maculosa and Nothoprocta ornata. This was noted in the latter species by Pearson and Pearson (1955) in the field and was observed in both species in the aviary. In these two, the head movements of locomotion have taken on some signal function, become typically exagger- ated, and may occur even when the birds are standing. It was probably from this source that the bowing head movements of Eudromia elegans evolved. The birds which use these behaviors live in fairly open environments, and these move- ments may serve in species recognition or take the place of contact calls; and, as with most visual displays, it is used only when the birds are out in the open. The erect walking gait of the red-winged tinamou varies from slow steps to a rapid walk. When the bird runs, the head is held low with the neck outstretched almost horizontally. The back and forth head component disappears as the bird typically plunges through the grass. A run usually culminates in a fast turn with the bird immediately sitting behind a clump of grass and under its overhang. Jumping These birds jump with little effort and do so to get food, to escape from an aggressive encounter, or to attack another bird, but never to clear an obstacle. A jump may be taken backward, for- ward, or to the side. In order to jump, the bird must have its feet close together; it apparently cannot perform a running jump. Several times I have seen a running bird trip over a dead branch or some other obstacle. 1 did not see the red-winged tinamou jump over the aviary’s 45 to 60 cm high barriers. On the other hand, Crypturellus noctivagus usually walked around the 45 cm high barriers in its enclosure but when pressed easily hopped over. Crypturel- lus species are smaller than Rhynchotus and in- habit forests where solid barriers are commonly encountered. Rhynchotus plunges into its environmental matrix rather than jumping or flying over it. It prefers to move on the level, and will walk 18 New York Zoological Society: Zoologica, Spring, 1973 Figure 2. Rhynchotus rufescens adult. Weeks: Red-winged Tinamou, Rhynchotus rufescens 19 around stones and other such obstructions. Flying When in a bare aviary, these birds take to wing at the slightest disturbance, but when provided with some ground cover, run for it. Their wings are short and rounded much like those of the ruffed grouse, Bonasa umbellus. Their flight is noisy and accompanied by a high, shrill, pulsed whistle. When forced to flight they typically burst up unexpectedly at your feet with fast, continuous wingbeats. The wings make a rattling sound and the whole effect is well de- scribed by Hudson (1872) as similar “to the rat- tling of a light vehicle driven at great speed over a hard road.” Even when 1 have seen a bird sit- ting on the ground and have been fairly certain that it would fly as I approached, it was still startling when the bird burst up from the ground. 1 am sure it would have a similar effect on a predator. Hudson (1920) reports that the birds can make only two or three flights in succession before they are too exhausted to fly or to run very far. Alert posture The birds assume the alert posture when mildly disturbed, as by a strange sound. In this posture the feet are brought fairly close together, prob- ably to gain added height. The neck is extended upward and the back is almost vertical. They can “freeze” in this position for several minutes or may turn their heads and look around. They normally stand and walk with their heads about 30 cm off the ground but, in the alert pos- ture, they can stretch to a height of at least 45 cm. Nothura maculosa also has an alert posture when it is not extremely disturbed. Crypturellus noctivagus does not have a pronounced upright posture; for this forest bird it is probably more important to look around environmental clutter than to attempt to see over it. This species does perform an interesting display when slightly dis- turbed which I call wing-flicking. It was easy to elicit simply by whistling and a bout was usually started by a bird in the open. These birds carry their wings slightly drooped, and a wing-flick is accomplished by a quick folding of the primaries. This produces a soft rustling sound and birds in cover come out and look around. This behavior is both mimetic and synchronized. They are soon moving about nervously, wing-flicking in almost perfect unison. Crouching and sitting While feeding, resting, or preening, the red- winged tinamou will occasionally crouch. The weight of the body is on the heels and tarsi; the body orientation is nearly horizontal. In addition, there is a post-copulatory crouch and a full pos- ture, mentioned later, in which the male’s body is held in a more upright position. While sitting, the bird’s breast is on the ground with its feet and legs folded under; the neck is held in a tight “S” curve above the breast. High stepping and creeping walk A peculiar type of walking behavior is seen when a bird is leaving the nest. Individual birds differ with respect to the distance traveled by this method, varying from a few steps to 4 m. The differences did not correlate with the sex of the individual. While walking away from the nest, the body is about the same distance from the ground as in normal walking, perhaps a bit lower. The feet are raised unusually high as though the animal were stepping over some obstacles. 1 was reminded of the walking behavior of a cat in wet grass. The area around the nest was not cluttered; sometimes well packed hay and straw formed the substrate, other times it was only bare, dry earth. I call this behavior “high stepping.” This gait is faster than average walking and the bird is always moving directly away from the nest in a fairly straight line. High stepping may function to reduce noise around the nest and may signal other birds as to the nest’s location. Other birds noticed a high stepping individual and would cease momentarily what they were doing. Another type of locomotion termed the “creep- ing” walk was seen only a few times performed only by males. The body, head, and neck are held low to the ground as the bird moves along. 1 be- lieve this is a submissive behavior of males to particularly dominant individuals and will be discussed under agonistic behavior. Standing Rhynchotus , Crypturellus noctivagus, and other tinamous (possibly all other tinamous) rarely stand with their feet side-by-side. This con- dition has been noted by Raikow (1968) in the rhea. I have seen C. noctivagus standing with one foot on top of the other. Maintenance Behavior and Comfort Movements Head and bill care Rhynchotus spends a great amount of time digging in the ground with its bill, and dirt, mud, and other debris frequently stick to it and some- times to the head feathers. Head shaking, a sharp flicking of the head and bill, is usually the first method used by the bird to remove it. If unsuc- cessful, the bird will bring its foot up and scratch the head or bill. The foot is never placed over the wing. For very sticky items, the bird resorts to bill wiping on the substrate. The mucous secreted by earthworms is frequently removed in this way. Another way is to peck into soft soil several times. Stretching There are three types of stretching behavior. In 20 New York Zoological Society: Zoologica, Spring, 1973 one, the wings while remaining folded are raised over the back. In another, one wing is extended downward until the primaries touch the ground while the ipsilateral leg is extended posteriorly. And finally, both legs may be extended together until they are straight. This raises the rump of the bird, but the head and neck remain low and may be slightly extended anteriorly. Ruffle-shake In this behavior plumage is ruffled and the body shaken from side to side in a rolling motion. It is used to straighten the feathers after preening, dustbathing, being mounted by another bird, and other plumage-disturbing activity. A wet bird will shake off some of the water by this method. At times this behavior would seem to have aggres- sive implications when a bird which had not been preening or disheveled in any way would sud- denly ruffle and shake. Dustbathing This behavior occurs only during sunny periods of spring and summer. It may start abruptly when the sun breaks out from behind clouds, and it is most apt to occur between the hours of 10 and 1 1 a.m. and between 2 and 3 p.m. The bird selects a bare area where the soil is dry and dusty. The site is out in the open, never shaded or near vegetation or aviary walls. While walking, or occasionally standing, a bird may make intention movements for dustbathing by lowering its body and pecking obliquely to the side at the ground. These pecks, averaging about four, may function to test the suitability of the soil for dustbathing. Even when the bird is moving, all pecks are directed very near the same spot. To accomplish this, the first peck is made with the neck out- stretched, and each additional peck is oriented more to the side compensating for the forward motion. The last peck usually requires an appar- ent awkward bend in the neck. The bird may then straighten up and continue walking, or it may turn sharply and sit very near the area of investi- gative pecks. From this position the bird continues to peck at the soil in front of its breast and to either side. Then, without rising, the bird scrapes with alternate feet and makes a quarter to a third turn, and then begins another bout of pecking at the soil in front of its breast. This sequence may be repeated several times until the soil is worked up and the bird is settled in a dustbathing scrape, a depression 4 to 8 cm deep. The length of time the bird spends in the pecking, scraping, and turning bouts seems to depend on the degree to which the soil is worked up, how fine the soil is, and whether a depression is present from previous dustbathing. A bird en- tering an old dust scrape, or one starting in an area of fine, loose soil, may only peck, scrape, and turn a few times. While in a new area where the soil may be more hard packed, it takes much longer. Finally, the bill is jabbed into the loosened earth and flicked caudally and laterally so that dust, pebbles, and dirt clumps are sent flying over the bird’s back. The feathers of the back and rump are raised slowly after a time. Dust is thrown alternately over the shoulders. As dust- bathing progresses, one folded wing is raised in such a way that the ends of the primaries slide over high on the back as the bird turns slightly on the opposite side. Now dust is directed only to the side newly exposed. Eater the other side of the body is usually afforded the same treatment after the bird has turned in the other direction. Upon righting itself in the dust scrape, material may be directed at the back again, and the dorsal surface may become completely covered with dust. Frequently, the bird rests in this position for several minutes. At any point in the dusting se- quence, further bouts of pecking, scraping, and turning may occur. Dustbathing is terminated by the bird’s rising and taking a few steps from the dusting scrape and ruffle-shaking. The bird emerges from the re- sulting cloud of dust usually to ruffle-shake again a few steps later. On a few occasions, a bird leapt from its dusting scrape aggressively at another bird that had come too close. Dustbathing is mimetic and other birds are drawn to the area to dustbathe. Not all birds that come to an active site dustbathe themselves. Some approach an actively dusting bird, sit be- hind and off to one side, apparently to benefit from the dust it throws. If a dust scrape is vacated, one of these satellite birds may enter it and begin active dustbathing. The length of dustbathing bouts can vary from a couple of minutes to well over half-an-hour. Eliminating the many abortive starts (those last- ing less than 30 seconds) gives an average of about 13 minutes. Mounting and copulatory behavior were asso- ciated with dusting especially during the bird’s reproductive period. This usually occurred when a male would approach a dustbathing female. An attempted mount would usually cause her to leave her scrape. Frequently when the female left her scrape, she would solicit for copulation within the vicinity of the dusting area. A fresh dustbathing scrape is easily identified. It is a depression about 20 cm in diameter and 4 to 8 cm deep. Around the outside of this circular cup, there is usually a trough about 3 cm wide and 1 cm deep resulting from the action of the bill. The factors controlling dustbathing seem to be temperature of the soil and the air, sunshine or artificial light and radiant heat, humidity, and the presence of Mallophaga. Dustbathing does not take place until temperatures on the ground are Weeks: Red-winged Tinamou, Rhynchotus rufescens 21 above 74° F and most often around 81° F. I could regulate the air temperatures to some degree by means of the heaters. The temperature of the ground could not be controlled and re- mained quite cold throughout the winter. I found I could increase the chances of dustbathing oc- curring in the very early spring, when dusting was rare, by increasing the air temperature to near 80° F. I did not attempt to raise the humid- ity, which during periods of frequent dustbathing is quite high. Further, when an infra-red lamp was installed in an enclosure containing Eudro- mia elegans and Nothoprocta perdicaria and Nothoprocta pentlandii, the area under the lamp was warm to the touch, and this same area be- came an active dustbathing site. It is probable that the factors which cause or are most apt to bring on dustbathing are the same factors which cause Mallophaga to become active. Rothschild and Clay (1952) mention the sensi- tivity of Mallophaga to temperature. I have no- ticed while doing avian surgery that the first indication of a failure in an operation was when the Mallophaga changed their preference from the bird to me. Without exception, this was an indication that the bird had expired. I assume that the cue for this emigration was a slight dif- ferential in temperature (Bair, 1950). The presence of Mallophaga on the tinamous was noted repeatedly. I agree with Rothschild and Clay ( 1952) that the function of dusting is primar- ily to remove ectoparasites. They report that Mallophaga have been found in dusting scrapes. Silicon dioxide powder can be used for the collec- tion of ectoparasites by lightly dusting a bird’s feathers with it. Observations 1 have made on a cinder path indicate that the path was far more attractive to wild birds and even small mammals for dustbathing than were two adjacent areas where the soil was dry, fine, and dusty. Attempts to collect Mallophaga with the dust from the cinder path were not fruitful, but this may have been done at a time when the birds had cleaned themselves of most of their ectoparasites. Dustbathing in other species of tinamous seems to be mimetic also. It is common to see one dustbathing bird soon joined by another. Eu- dromia elegans and Nothura maculosa were seen dustbathing in groups of twos and threes. No- thura maculosa dusting behavior was markedly different from Rhynchotus. The dust was stirred up and into the feathers mostly by rapid foot scraping movements and not at all by the bill which was ineffective in this way. Unlike many other birds, such as the galliforms, for example, the tinamous do not use their wings to get dust into their plumage. Nor do they use their bills to pick up dusting material as do rheas. Lancaster (1964) describes dustbathing in Nothoprocta cinerascens which is quite similar to the red-wing. The two forest forms, Crypturellus obsoletus and Crypturellus noctivagus, were never seen to dustbathe. No sign of a dusting scrape was ever seen in their pen. Eudromia elegans does not always ruffle-shake after a dusting bout. Several times 1 have seen them walking with a layer of dust on their backs, and one incubating male was liberally covered with dust. This could render a bird and the nest more cryptic in their dry habitat during incuba- tion. Upon examining a red-wing one morning before any dustbathing had occurred, 1 lifted the feathers of the rump and found the skin covered by a layer of dusty grit. Waterbathing Eudromia elegans and Rhynchotus bathe very rarely. Also, Lancaster (1964) observed bathing only once in his observations on Nothoprocta cinerascens. Perhaps waterbathing may be ac- complished by walking through wet vegetation. These birds, it was noted, are active dustbathers. Crypturellus noctivagus was the most active waterbather of the tinamous that I observed. It bathes in pools as well as in the spray from a hose nozzle simulating rain. It slowly walks back and forth through the spray, shaking the head. The bird lowers its body and creeps, shakes its head, and eventually sits. It may get up, creep around, and sit again. There is little ruffling of the plumage. At this time the bird will shake its head, peck at a leaf, or the sub- strate, wipe its bill, but most often perform a peculiar head movement. The head and neck are rapidly placed flat on the ground in front of the individual and are simultaneously rotated a quarter-turn in one direction with the bill remain- ing in the axis of the straightened neck. The bird then quickly rotates its head a half-turn in the opposite direction and withdraws the head and neck. This may be repeated, or head shaking, pecking, or bill-wiping may occur. Soon the bird will turn slightly on one side and lift the upper, folded wing high over the back in such a way that it looks as though the wing were broken. As the wing approaches the midline, the tips of the pri- maries actually cross over the back. Enhancing the “broken wing” look is the bird’s head which now rests between the body and the raised wing, specifically between the wrist of the raised wing and the breast. Again, the bird will extend and rotate the head, then withdraw it. The bird will eventually turn back to a more normal sitting position and may creep about some more and sit again, or it may remain sitting. It is not uncom- mon for the same side to be bathed again or the bird may alternate, but in a full bout of spraybath- ing both sides eventually are exposed. When a bout is over, the bird walks off slightly ruffling and then relaxing the plumage. The odd motion of the head in spraybathing 22 New York Zoological Society: Zoologica, Spring, 1973 may be better understood when looking at pool- bathing. The bird enters a shallow pool and may peck at the surface or drink. The actual bathing may begin by a lowering of the body into the water in a rhythmical bobbing motion. But this is not always seen. The bobbing wets the ventral surface of the bird. Next the body is lowered into the water tilting slightly forward on the breast. Again, the head movements occur, but the head and neck are stretched out parallel with just the tip of the bill into the water’s surface. The rapid rotation throws a few drops of water some of which land on the bird’s plumage. The wing is raised as in spraybathing but there is no bill wip- ing or head shaking, although there is sometimes a rare peck at the water’s surface. These two types of bathing are not mutually exclusive and may be combined. The plumage of both the Crypturellus species I observed is quite different from the other tina- mous. Besides being much darker, the feathers have a dusty coating that repels water and indi- cates abundant powder downs (Welty, 1973). This covering washes off and remains on the sur- face of the water like minute bits of wax. From this 1 was able to tell whether a Crypturellus had bathed since the water pans had last been filled. I could tell which birds had bathed recently be- cause their plumage appeared considerably darker. The color change was due to the removal of the waxy dust. The Crypturellus like water and would even rest in it for several minutes at a time. The most active bather of the openland- grassland forms was Nothura maculosa. Under a spray of water, it goes down on its breast, throws back its wing on one side, and turns on the other. Following a period of deprivation of water spray, these birds would go down and turn with such momentum that they would roll completely over. Bump and Bump (1969) noted this species’ eager- ness to spraybathe. Nothura maculosa also bathed in puddles or pools; this was not seen at close quarters, but the birds did run purposefully into a standing puddle, stop, ruffle, crouch, and then run out again. Rhynchotus is drawn to the spray from a hose because of the movement caused in the grass by the water, rather than by the spray itself. The birds avoid getting wet most of the time. Occa- sionally a bird would run through the spray or through the water pan. But on only three occa- sions did I see a bird hesitate in the water, ruffle its plumage, and bob up and down in the water or make a few tossing motions with the bill. Only once did a red-winged tinamou crouch in the spray and perform a few desultory dustbathing- like motions before moving on. These bouts were all very short. Nothoprocta ornata spraybathe by going into a sitting position, rolling over on one side. Huffing the plumage, and raising the upper wing. This same procedure is then repeated on the other side. Sunbathing This behavior was not observed in the red- winged tinamou. Crypturellus noctivagus is an active sunbather. Even when light intensity was low, it went into sunbathing postures in the forced air stream from the heater, despite the fact that it was sometimes difficult to hold that posi- tion in the strong draft. Nothoprocta ornata would also sunbathe in response to the same stimulus. Lanyon (1958) found a similar behavior of birds in complete darkness. A spotlight would also elicit sunbathing in Crypturellus noctivagus. The light was strong and so close to the ground that the area under it was warm. A bird would wander under the light and seem to become caught by it, stopping directly under it and standing for a time. Then it would turn and sit and slowly roll over to one side. The wing on the upper side would then take one of two positions. Either it is raised over the back while kept folded and held in that posi- tion, or it is held fully extended out across the side of the bird and parallel to the ground. Due to the cant of the wings, only the tips of the pri- maries rest on the substrate. Gradually the eyes are closed and the head allowed to sink slowly until the bill rests on the ground. A sunbathing bout may be interrupted by some other mainte- nance behavior such as scratching the bill with the foot. To summarize bathing behavior: 1) In captivity most openland-grassland birds do not sunbathe. Nothoprocta ornata is an excep- tion. Bohl (1970) reports that Eudromia elegans sunbathes, which is a second exception. The non- sunbathers are exposed to the sun much of the time, while the forest birds which are active sun- bathers, are not. Welty (1963) discusses the pos- sible function of sunbathing. 2) Tinamous of the openland-grassland regions do dustbathe. They live in a drier environment than the forest forms which do not dustbathe. 3) Birds of the forest are more active water- bathers. Their environment has a higher rainfall and thus water would be more abundant. Nothura maculosa is an exception. Eudromia elegans was not seen by me to waterbathe but Bohl (1970) says that they rainbathe. It must be kept in mind that the condition of captivity could greatly influence whether a type of bathing occurs and the degree to which it is used. Tameness, for example, can affect bathing, for only a bird that is sufficiently habituated to its environs and the presence of an observer will be relaxed enough to perform it. Powder downs and oil glands Welty (1963) states that birds without oil glands frequently have abundant powder downs. Rhynchotus has few powder downs and a large preen gland while Crypturellus noctivagus has Weeks: Red-winged Tinamou, Rhynchotus rufescens 23 many powder downs and a small gland. Welty states further that powder downs grow in dense yellowish patches, especially on the breast, belly, or flanks of a number of birds including tinamous. In the tinamous that 1 have examined, these feathers seem more dense on the rump and are diffusely spread over the body. In addition, they are the same grayish color as the downy base of the feathers and aftershafts. Thus they are not easily separable. Yawning It is not uncommon to see the red-winged tinamou yawn, clearly displaying its wide gape. The function of yawning has been studied in other birds (Sauer, 1967). I can assign no display function to it and it does not seem to be mimetic. In Crypturellus noctivagus the incidence of yawn- ing increased markedly with the starting of the heater which blew hot air into the flight. Dilger (pers. comm.) has noticed that rheas in captivity seem to yawn when jet aircraft take off in fairly close proximity to the birds, perhaps in irritation to the sound or intense vibrations. Wing flap Usually without warning, a Rhynchotus rufes- cens would simultaneously go up on its toes and quickly raise its half-folded wings over the back and bring them down sharply making a loud flap. This is performed by both sexes. Variations include a double wing-flap or a jump as the wings are flapped down and finally a sharp, short cry after the flap. 1 was not able to determine a function for this behavior although I feel it has some aggressive display function. In Eudromia elegans, a similar behavior is used in aggressive encounters where one of the flapping wings hits another bird. In Crypturellus nocti- vagus, again the function is not obvious. At least twice, when I was imitating the call of Rhynchotus rufescens , the dominant male stopped what he was doing and walked to a point in front of me and seemed to direct a wing flap toward me and then returned to where he had been before. Sleeping posture and roosting behavior Well before dark, the red-wings begin prepar- ing for the night. Each bird has a favorite roosting place but usually a fresh roosting scrape is pre- pared each evening. The bird may be already resting near the site of the roosting scrape or approach it directly. If this site is in an area of abundant straw, the bird will make several turns while lifting the feet higher than usual, but some- times these same turns are performed on bare ground. The bird may use this behavior to push down the grass and to select the actual site of the scrape. Eventually the animal settles down and performs scraping and turning motions as those of dustbathing. Some birds repeatedly select sites next to others, and might perform pushing-under bouts. These are most often females pushing underneath a bird already in its roosting scrape or in the preparation stage. These birds would commonly choose a roosting site near those birds under whom they had pushed. This behavior is reminiscent of a chick pushing under and into the plumage of a parent and would seem to be the evolutionary source for the behavior. Also at this time some precopulatory behavior occurs. Red-wings sleep sitting with the neck in a tight “S” curve and the bill resting on the upper breast. There were only two exceptions: one bird was seen sleeping while standing on one leg: another had its head tucked under the scapulars like most other birds. It could be that these tinamous never felt relaxed enough to tuck their heads in my presence and did so after I left, or at least during a deeper part of the sleeping period. Roosting scrapes are placed near objects such as aviary walls, vegetation, or barriers. They can further be distinguished from dusting scrapes by the abundant droppings found around their edges. Schafer (1954) reports that Crypturellus nocti- vagus roost in trees at night. I placed a number of perches including a small tree in their flight, but only saw perching when a bird was put to flight by an aggressive cage-mate and it would inadvertently nutter down into thick vines. I repeatedly saw Crypturellus noctivagus individ- uals take off for no apparent reason and fly to the ceiling of their enclosure and Rutter down to the ground. Also the yellow-legged tinamous were the most relaxed tinamou during the day, but at night they became the most Eighty. Almost any movement would send them into a panic. Eying to the ceiling and Euttering down in the dark. This may indicate improper roosting facilities for this species. As mentioned earlier, other species of tinamous have been reported perching in trees at night (Skutch, 1969; and Wetmore, 1965). Defecation This behavior seemed elaborate and exagger- ated enough to have some signal function, but my investigations did not elucidate any. It occurs most often 15 to 30 minutes after a feeding bout when an individual who is usually walking will stop, rufRe the plumage, may take a step back- ward, half-crouch and extend the head and neck forward, slightly raise the wings, and defecate. The bird rises, may ruffle-shake, and walk on. Crypturellus noctivagus commonly defecates while stretching both wings over the back. This was rare in the red-wing. Preening As noted earlier, Rhynchotus has a well- developed oil gland with a tuft of specialized feathers projecting out about I cm. In some individuals this appeared to be four separate tufts, in others two, and in one bird it looked like one large tuft. 24 New York Zoological Society: Zoologica, Spring, 1973 Preening in the red-wing is very much like that of Anseriformes. The preen gland is pinched and the bill and head are rubbed over the surface of the body and the preening of the flight feathers, tail, and coverts was also similar to that of a duck. Preening follows bouts of waterbathing, dust- bathing, and copulation as well as most feeding bouts. Preening before preparation of the eve- ning roost scrape was also noted. Food, Feeding, and Drinking The feeding apparatus One of the peculiarities of the family Tinamidae is their broad gape. The red-wing’s gape is 2.5 cm between the commissural points. In addition, this gape seems to cut deeply into the head, so when viewed from the side, the commissural point is situated well posterior to the eye. This wide gape permits the swallowing of large food items. The bill of Rhynchotus is the longest in the family, tapering 4.5 cm from the forehead to the tip and almost 6 cm from the commissural point to the tip. It is dorso-ventrally flattened and slightly decurved. The tip is depressed also so that it is sharply pointed in its lateral profile and broadly rounded from above. This bill and the bird’s long, powerful neck constitute an effective digging mechanism. The toes have short rounded claws adapted for scrap- ing and scratching the substrate. However, the scratching common in the feeding behavior of many terrestrial birds has not been observed by me in any of the tinamous. Rhynchotus , at times, uses its feet to scrape while feeding, but this is in an unusual way and does not function to uncover food but serves only as an accessory to the dig- ging of the bill. One would expect that a member of a family with an unusual bill would feed in a different way, exploiting another food source. This is indeed the case. Diet The red-winged tinamou can almost be classi- fied as omnivorous, for it is capable of consuming a wide variety of food materials, the only excep- tion perhaps being carrion. Yet the greatest quantity of its diet is vegetable matter. Through the examination of stomach contents of 53 specimens of Rhynchotus taken at different times of the year from diverse areas of the state of Sao Paulo, Brazil, Hempel (1949) concluded that the principal part of their diet was vegetable material consisting mainly of tubers, tender roots, and seeds. Some fruit was present, especially Byrsonima intermedia , which is known locally as the “fruit of the red-wing.” Another favorite fruit was guava or aracas do campo (Psidium species). Fruits and seeds of Smilax species, seed of Con- volvulus and of Desmodium species (Papiliona- ceae) were also found. The most common seeds, according to Hempel, were those of legumes (Leguminosae) and sec- ondly those of grasses (Gramineae or Poaceae). The animal part of the diet was made up mostly of insects. Grasshoppers were the most common. Termites ( Syntermes silvestrii and Syntermes par- allels), crickets, hemipterans, coleopterans, and lepidopteran chrysalises and caterpillars made up most of the remainder. Spiders and earthworms were also present. Finally, Hempel noted that during the winter when insects are scarce, the red-wing procures more vegetable food. Schubart, et al (1965), report the stomach contents of four specimens from the state of Minas Gerais, Brazil, that varies little from Hempel’s findings. Comi (1927) and Renard (1924) found that Rhynchotus occasionally take mice. Comi felt it occurred more often in areas where grain was not abundant and the birds were forced to take this unusual food item. Aviary diet The following diet was supplied ad libidum : High protein pellets — Game and Turkey Grower Pellets, Agway, Inc. (21 percent crude protein). Mixed seeds — Wild Bird Seed Mix, Agway, Inc.; contains: milo, red and white millet, wheat, sun- flower seeds, shelled peanuts, canary seed, and buckwheat. Growing grass of varying ages, sprouting grass seeds, clover, alfalfa, and other plants. Calcium sources — (a) cuttle bone, whole and crushed; (b) oyster shell; (c) Calcite crystals; and (d) chicken egg shell. Regularly supplementing the above diet were the following: Fresh greens such as various types of lettuce, cabbage, mustard leaves, and spinach. Vegetables such as potato and carrot. Fruit such as apples and grapes. Live animal material: (a) earthworms, (b) mealworm beetles and larvae ( Tenebrio molitor), (c) crickets, and (d) pillbugs (isopods). All of the live animal food was cultured in the aviary and was available the year around. Earth- worms were also collected and brought into the aviary when they were in season. New food items were continually being tried both as a supple- ment to their diet and to obtain information on feeding behavior. Small sections of the aviary floor were spaded up and mixed with peat moss and kept moist. At these locations grass seed was routinely sown and it was there that the earthworms were presented. Those which were not immediately eaten would remain in these moist “oases” and were available to the birds who dug them later. During the summer months, quantities of in- sects and spiders were collected with a sweep net for use as a further dietary supplement. Weeks: Red-winged Tinamou, Rhynchotus rufescens 25 Feeding behavior Rhynchotus almost always consumed live food at once. A bout of feeding could be stimulated at any time of the day and feeding times were more determined by my schedule than by any natural preference that the birds might have had. 1 have the impression, however, that there was generally an early morning feeding bout before 10:00 a.m., and an irregular afternoon feeding time. In addi- tion, just before making their evening scrapes there would be increased feeding at the food bowls. Appetitive feeding behavior is obvious as the birds actively forage while walking with their heads low to the ground. At first seeds and other food items which were broadcast over most of the flight floor were taken in preference to that of- fered in bowls. Random pecks, mostly into the moist areas, occurred at this time. This is explor- atory or incipient digging behavior. After a period of this active foraging and feeding behavior, the bird will find a spot for digging. Digging behavior In a soft substrate such as moist earth, the bill, in one continuous motion, is jabbed in and brought sharply backwards by the flexion of the neck. A lateral component is added by a slight turning of the head at the same time. This causes most of the dislodged material to be thrown back and to the side at about a 45 degree angle from the body line. The side component alternates from one side to the other. If the substrate is very light and soft, as is peat moss, the bill appears to be open during the throwing part of digging. When undisturbed in a good feeding location, the bird will crouch and continue digging. As this goes on, some debris will build up in front of the breast. The feet are then alternately brought up carefully forward to scrape this material back from the breast; at the same time, it may advance the bird’s position, as the foot may not return exactly to the original resting position but to a somewhat more anterior point. The stimulus for this slow scraping foot move- ment is dirt touching the breast. As a bird’s breast comes in contact with the rim of a feeding bowl, the same behavior is observed. The feet come forward and scrape as though to pull down the side of the bowl. While digging, the birds eat earthworms, grubs, ants and ant pupae, roots, and sprouted seeds. Blaauw (1896) also observed Rhynchotus obtain- ing earthworms in this manner. Digging bouts usually followed the presentation and eating of fresh earthworms. Newly-spaded earth stimu- lated digging also. Jumping In addition to digging, Rhynchotus , like some of the other tinamous, increases its potential three dimensional feeding space by jumping up and picking insects off vegetation. Tests to deter- mine the limits of this ability showed that they could catch insects 1 m off the ground, although accuracy falls off markedly between 85 and 100 cm. Crypturellus noctivagus was similarly tested and was found to be accurate up to 75 cm. In jumping for insects, the leap is made almost straight up, as the animal positions itself almost directly in front of the item to be taken. This dif- fers from agonistic jumping which has a much greater horizontal component. Handling of food items Very small items such as rye grass seed, clover seed, small ants, and flour beetles ( Tribolium confusum ) are not eaten, while things slightly larger such as canary seed and millet are taken readily. Most insects, seeds, pellets, and other food types are handled in much the same way. The object is pecked and seized between the tips of the mandibles. The head is withdrawn, and then thrust forward as the bill is opened and the item is thrown into the gullet. More precisely, the gaping mouth is thrown around the food object. When several small food bits are collected in the mouth, they are swallowed. Targer food items are downed whole and swallowed separately. One of the largest items taken in this way was a cerambycid beetle, Prionus species, which meas- ured about 3.8 cm long and 1 cm wide. Sunflower seeds may be shelled, especially if the seed has lain on damp soil and softened. These seeds are tested by a loud clicking mandibulation. If the mandibulation continues, the seed may be hulled, eaten whole or it may be dropped. Shelled peanuts, even though smaller than items swallowed whole, are commonly broken up, by one or two pecks and the pieces are then con- sumed one at a time. Seeds are usually eaten off the ground, but I did see birds taking seeds from maturing grass heads. Also the red-wings do eat some blades of grass and leaves, especially new growth, but they concentrate more on the roots and shoots which they expose while digging. Small earthworms 5 cm or less are swallowed whole. Sometimes when tossed, not all the worm goes into the gullet and another swallowing act is required. The larger earthworms will not fit in the mouth and therefore require several swallowing attempts. Most commonly the large earthworms ( Lumbricus species) are pecked and given one or more shakes and snipped in two. There is some variation in whether these shakes included rub- bing the worm on the ground. The largest are commonly given these sharp shakes while being held against the ground. These worms are usually snipped into three pieces before being ingested. 26 New York Zoological Society: Zoologica, Spring, 1973 The birds show a preference for the smaller worms. An interesting example of special handling before eating is the treatment dealt to the woolly bear caterpillars (Arctiidae, Isia isabella). The larva is quickly grasped and at the same instant the bill makes rapid lateral motions in the sub- strate in an arc about 5 cm long. The larva is dropped, picked up, and rubbed again. If the sub- strate is gravel, the bird will attempt to rub the item, but then move to a finer substrate. When most of the setae are rubbed off, the larva is thrown into the maw and swallowed. A remark- ably similar behavior was reported by Morton (1968) in robins which were feeding on the same larvae. A food source which is rejected by most birds but is eaten readily by Rhynchotus is slugs (Limax, Arion , and Deroceras species). Centi- pedes, millipedes, as well as terrestrial snails are swallowed whole. The birds would not eat carrion, canned dog food, or chopped or stripped fresh calves liver. The stripped liver was rejected even when thrown out to the birds alternately with earth- worms of the same size and dimensions. I have seen Rhynchotus and Crypturellus nocti- vagus chase and catch low flying insects when they land, but these again have to be above a certain minimum size to be of interest. Their catching moths has been most conspicuous. The red-winged tinamou is an extremely oppor- tunistic feeder and can capitalize on an abundant food source by shifting its intake temporarily and almost exclusively to whatever is plentiful. This was seen many times in the aviary and is illus- trated by the study of stomach contents by Hempel (1949). He counted 45 grasshoppers in the stomach of one bird, 275 termites (Syntermes species) in another, and 707 in a third. Lieber- mann (1935) claims that Rhynchotus , like most other tinamous, eats locusts (grasshoppers) and that during times of locust invasions (grasshopper outbreaks), the red-wings eat little else. In the aviary it seemed virtually impossible to satiate these birds with live material. In one morning the six individuals had eaten over 1 kg of earthworms and they were by no means equally distributed among them. Perhaps not all the protein is avail- able to the bird’s metabolism from an eating orgy such as this, for after such excesses I would find very dark blackish-brown droppings of the con- sistency of heavy grease. Mealworm larvae are always eaten in any quantity provided, as are crickets. Because of this voracious appetite, at- tempts to establish ant colonies and cockroaches in the aviary were unsuccessful. Predation on vertebrates From the reports of mouse eating in the literature (Comi, 1927; and Renard, 1924), 1 tried presenting small mice to the birds. They proved to be determined, if not well-equipped, predators. Although able to deliver hard-driving pecks, they seem unable to break the skin of even a mouse. They are likewise unable to subdivide the flesh. Thus, it takes them a long time to kill an animal and quite a long period of repeatedly pecking the carcass until it is swallowed. At that point the carcass is very limp and soft. It is taken head first and the first thrust of swallowing lodges the mouse’s head within the mouth of the bird. After a long bout of swallowing, the tail of the mouse finally disappears. A 30 cm garter snake ( Thamnophis species) was brought into the aviary to check the reaction of the birds. At first the birds were quite put off by the coiling and striking behavior, but soon they were pecking at it, mostly on the tail, and running with it. When dropped, the snake cov- ered its head with its body so many of the pecks were not too detrimental. The snake was taken back and forth through the aviary by various birds, one snatching it from another. Eventually the snake, though not quite dead, was eaten head first, again requiring a long bout of swallowing. Later, snakes up to 35 cm long and of various species (mostly Thamnophis species and one milk snake, Lampropeltis doliata) were eaten with less expenditure of energy and the snakes were much more alive when swallowed. One individual swal- lowed two snakes tail first. Both of these tried to crawl out or at least stop the downward motion toward the bird’s stomach without success. When the snakes approach 30 cm in length, a peculiar problem arises. When running with the snake, unless it has been grasped near the middle, it will be stepped on by the bird’s foot which pulls it out of the bird’s mouth. The snake is quickly grabbed by another bird, only to have the same thing happen. After a bird has swallowed a large animal, such as a snake or mouse, it assumes a “full posture.” This is a crouch in which the body is held more upright, and the head and neck are maneuvered out, down, and up, apparently to aid in the pas- sage of the consumed animal down the throat and into the crop. There was never any sign of regurgitated bones, nor was there any bone or hair in the droppings, so I assume that the birds are capable of digest- ing these items. Other vertebrates taken as food include small frogs (Rana pipiens and Ran a clamitans) and toads (Bufo americanus). Although somewhat similar in appearance, these were dealt with in different ways. The frogs were treated in the same way as the mouse, with repeated hard pecks; but the toads were pecked and rubbed in the dirt, pecked again and rubbed again, and finally swal- lowed head first. The toads less than 2.5 cm are pecked and swallowed whole. Tadpoles were also taken from the water pan. Weeks: Red-winged Tinamou, Rhynchotus rufescens 27 These birds had a tendency to eat their own eggs, especially if they were laid outside the nest area. A chicken egg placed in the middle of the aviary was eaten by the end of the second day. Five dove eggs were placed in a small depression on the ground and were eaten in one afternoon. Five Coturnix quail eggs, which have a mottled brown color pattern, remained for several days without being touched and were then removed. In general then, Rhynchotus differs from the other openland-grassland tinamous in the degree to which it feeds on material found under ground and its almost complete lack of grazing on mature grass leaves. Eudromia elegans, Nothura macu- losa, Nothoprocta ornata, Nothoprocta pentlandii, and Nothoprocta perdicaria are all grazers and probe only shallowly into the ground. They are most apt to turn over small rocks. Crypturellus noctivagus and Crypturellus obsoletus are leaf turners and tossers. Rhynchotus was never seen to eat feathers, while Crypturellus noctivagus does eat them. Hempel (1949) found feathers in the stomach of Crypturellus parvirostris. Drinking Although water was provided ad libidum , the tinamous drank very little, except when they were sick and long bouts of drinking occurred. Drinking behavior is scattered quite evenly throughout the day and it is very uncommon to see one individual drink more than once a day. Drinking in the red-wing is always from stand- ing water and never from the foliage as with Nothura maculosa , Nothoprocta perdicaria , and Nothoprocta pentlandii. These two Nothoprocta species seem to prefer drinking from the foliage rather than from the water pan. Koepcke (1963) noted this preference in Nothoprocta pentlandii. Rhynchotus dips its bill into the water at a very low angle. There is some variation in how deep the bill is immersed in the water, but it is usually about 1 cm. The gullar area is moved in and out and the head is raised high. This may be repeated up to three or four times. Agonistic Behavior The red-winged tinamou, it was noted, is a bird of the tall, open grasslands. Hudson (1920) some- what ambiguously describes them as being soli- tary and yet with “many birds usually found living near each other.” I assume this latter state- ment was made on the basis of his hearing “many individuals answering each other” during periods of calling. Rhynchotus, then, lives in a rather dense envi- ronment of tall, tufted, or clumped grasses and is in at least auditory contact with several neighbors during parts of the day. The area available for an individual’s home range would probably be measured in hectares and not in square meters as in the aviary. Marler and Hamilton (1966) state that crowding can affect dominance systems and that in many cap- tive vertebrates a shift from territorial dominance to individual dominance is seen as densities in- crease. My findings support this. The crest All of the tinamous I observed have erectile crests, but only in Eudromia , the crested tina- mous, have these markedly specialized, forming a graduated crest about 5 cm long. It is undoubt- edly not coincidental that in the tinamids this well developed, highly mobile, visual signal is found in the only species which is known to occur in flocks (Wetmore, 1926). These birds display a flock cohesiveness which I have not seen in other tinamous. The movements of one individual at times are determined completely by the move- ments of the others, especially when they are disturbed. This is unlike the other tinamous which, if they happen to be together, are most apt to disperse when alarmed. The red-wing’s crest feathers are black, streaked with tawny brown. When viewed from Rhynchotus eye-level, the erected crest appears mostly black. These feathers are about 1 cm long, shorter on the forehead and longer on the back of the head. Thus, when fully erect, there results a marked increase in the apparent size of the head. As has been demonstrated in a wide variety of birds, the position of these feathers can change quickly, showing the fluctuations in the motiva- tions of the bird. Crest-raising or crest-up indi- cates aggressive motivation while crest-lowering or crest-down is a sign of fear. Within a few sec- onds the crest can change from one extreme to the other. This is predictably seen, for instance, when the birds are drawn into a small area by a concentration of a favorite food item such as mealworms. The crest plays an important role in the routine display of dominance and submission. A meeting between a dominant and a submissive bird results in the dominant one raising its head and crest, while the submissive bird lowers its head and may sleek the crest. Also, when one individual chases another, the pursuer has its crest raised most of the time and the fleeing bird’s head is usually sleeked. These birds are capable of individual recogni- tion which is obvious from the dominance hier- archies which are formed. These hierarchies are not often linear and can change rapidly during the reproductive period. An indication of the importance of the crest in the role of dominance may be illustrated by the fact that the least dom- inant bird was a male which had had the top of his head skinned and lacked a crest. Eventually when he tried to exert himself aggressively, he used one of the more extreme displays not entail- 28 New York Zoological Society: Zoologica, Spring, 1973 ing the use of the crest, but even then it appeared that he was ignored by the other birds. This could imply that in Rhynchotus learning plays an important part in aggressive encounters, as has been shown in other vertebrates (Marler and Hamilton, 1966). An obvious explanation might be that an individual unable to display day- to-day dominant aggression toward any of the other red-wings eventually loses his aggressive identity. Thus on occasions when this bird with- out a crest was very aggressively motivated and used extreme threat displays, other red-wings would apparently feel no fear and make no effort to avoid him. Threat displays The threat displays of the red-wing may be scaled in order of the degree of their aggressive- ness. Starting with the least aggressive, they are as follows: 1) The mildest threat is when one individual orients its head toward another. The threatening bird may be sitting, crouching, standing, or walk- ing. If walking, the bird usually stops as it orients the head in threat. The head may be turned in any direction, even posteriorly over the back. 2) If the neck is slightly withdrawn and the bill opened as the bird orients toward another, a higher degree of threat is indicated. This can be given from a standing or sitting position and is most often directed anteriorly or laterally. Peck feigning is probably an extension of this and is discussed below. 3) A much more extreme threat is displayed when, added to the above attitude of neck pulled in and gaping bill, the bird orients the body and assumes a crouched or nearly crouched position. This is a forward leaning posture and it varies in the degree to which the bird goes forward and down on its breast. In the most threatening form, the bird’s breast touches the ground. From this position, a bird can easily jump at an opponent. It is similar to the threat of the common rhea (Rhea americana ) as described by Raikow (1969). The rhea, however, is more nearly in a standing position. Crypturellus noctivagus is more rhea- like in its standing posture but it does not gape. It alternates this tilted forward, neck withdrawn posture with a few stalking steps toward the bird on which it is oriented. 4) If the above threat display of Rhynchotus is not readily effective and the offending bird con- tinues to approach or remains nearby, another component is added. The aggressive bird goes into a full forward crouch with the breast on the ground and the feathers of the rump begin to rise in broad plaques of twos and threes. 5) Finally, the highest level of threat is for a crouched bird, leaning forward on its breast, neck withdrawn, bill gaped and rump feathers elevat- ing, to give a soft shrill cry as though to call fur- ther attention to itself and accent its aggressive state. Only males go beyond level 3 and it is rare for a female to perform at a low level 3. Peck feigning is related to level 2 in the aggres- sive scale above, but it is most used in fast mov- ing, transient encounters when an aggressive bird with little time to react makes an irresolute at- tempt to peck an intruder. Many times the bird is sitting and the peck attempt is obviously futile since the bird would have had to move in order to land the blow, as for example when the feigned peck is directed posteriorly over the back at an intruder. Overt aggression in the red-wing is restricted to pecking and/or chasing another bird. Chasing is usually a short burst of running that rarely goes beyond 3 to 5 m. Longer distances are covered by interspersing chasing bouts with walking pursuits. Pecks are mainly directed at the rump and back of another bird. These are hard enough to be audible but in the red-wing, they do not often result in the loss of feathers. In Crypturellus noc- tivagus an aggressive peck is commonly com- bined with feather plucking. After a chase and peck, the aggressive bird is left standing with a contour feather in its bill while the other bird runs off. After a pause, the plucked feather is carefully mandibulated and then ingested, calamus first. The assailant seems surprised and distracted by the presence of this object in its bill, and pauses as though to “decide” what to do with it. Perhaps other Crypturellus species do this as well for, as noted earlier, Hempel (1949) reports feathers in the stomach of Crypturellus parvirostris. Figure 3 is a graph showing the frequencies of overt aggression and threat displays according to the sex of the aggressor and the individual elicit- ing that aggression. The most common encounters are males displaying aggression toward females. I had assumed that the most common encoun- ters would be male-male interactions as would be true for many vertebrates so confined. I expected that the males might do little else than fight among themselves, and watched carefully for any indication that they might injure one another. Lancaster’s observation (1964) of Nothoprocta cinerascens that the only means of territorial defense by the male is calling bouts, may give an indication as to the paucity of chases, pecks and visual threat displays between red-wing males. This will be discussed further in the section deal- ing with vocalizations. The lack of much female- female interaction was about what I expected. The similarity of frequencies of threats, male- to-male, female-to-female, and female-to-male is remarkable and in sharp contrast to that of males- to-females by seven-fold. This does not mean that the males can readily distinguish between the 200 Weeks: Red-winged Tinamou, Rhynchotus rufescens 29 o o o if) © «o o u CO 'O E CP 1— O •y O CO E *” CN 0) v*- \ a. CO J) o u E *■*- \ _cu o E 00 CN 00 K o CN O h- CN c □ 3 _c o u “D jO c 0) d5 k+— 0 at -C >/) i— t/i