Introduction

The order Artiodactyla comprises three suborders: Nonruminantia, Tylopoda, and Ruminantia (Haltenorth, 1963). The Nonruminantia comprise three families: Suidae (five genera and eight species), Tayassuidae (one genus and two species), and Hippopotamidae (two genera and two species). The Tylopoda consist of only one family: Camelidae (two genera and four species). The Ruminantia comprise five families: Tragulidae (two genera and four species), Cervidae (eleven genera and thirty-two species), the Giraffidae (two genera and two species), the Antilocapridae (one genus and one species), and the Bovidae (forty-two genera and ninety-nine species). Numbers of subfamilies, genera, and species vary somewhat with different classification systems; the figures given above represent the minima.

At present, the discussion of communication in Artiodactyla suffers from certain difficulties. The Nonruminantia and the Ruminantia have rather different physical structures and means of communication. Also, there are considerable differences within Ruminantia with respect to size, physical structure, habitat, life habits, and social organization.

Furthermore, our present knowledge of behavior of Artiodactyla is limited. Some information is available on behavior of Hippopotamidae, Tayassuidae, and a few Suidae species. Virtually nothing is known about behavior of the Tragulidae. In only about seven cervid species has behavior been studied intensively enough to allow description and discussion of the phenomena and problems of communication. Good information is available on the behavior of Tylopoda, Giraffidae, and Antilocapridae; however, they comprise relatively few species. More investigations of communicative behavior have been carried out on bovid species than on other groups of Artiodactyla, but the approximately thirty species investigated make up less than one-third of all bovid species.

Information from studies on behavior of artiodactyl species is usually rather good on visual displays (postures and gestures), considerably less so on acoustical and olfactory behavior, and poor on tactile communication. (This is provided that one takes the term "communication" seriously and does not consider any form of physical contact to be a communication.)

Further difficulties arise from the general problems of expressive behavior and intraspecific communication. Expression can be a phenomenon or an epiphenomenon (Eibl-Eibesfeldt, 1957). As a phenomenon it is a special and well-defined display (movement, posture, or vocalization), like the threatening presentation of horns toward an opponent in many bovid species. In the case of an epiphenomenon, a basically nonexpressive behavior is performed in a special manner. For example, in a stiff-legged walk, the special manner (stiff-legged), not the behavior pattern per se (walking), adds an expressive character to the performance. Such expressive epiphenomena can, at least occasionally, be attributed to almost any behavior. A discussion of them could easily lead to a discussion of behavior in general. For this reason, it appears advisable to focus this presentation on expressive phenomena. On the other hand, we cannot completely exclude expressive epiphenomena since some of them are important in communication and/or may contribute to a better understanding of comparative and evolutionary aspects of certain (special and well-defined) displays.

Furthermore, the realm of expressive behavior is not confined to social communication (Leyhausen, 1967). In other words, there is also expressive behavior without function in intraspecific communication. For example, "Rehmen" (Fig. 1) is a very common expressive behavior (especially in males) of many artiodactyl species (Schneider, 1930, 1931, 1934), but, except for occasional contagion, it has no effect on conspecifics. An expressive behavior with a clear function in social communication is addressed (but not necessarily directed) to a definite conspecific recipient (occasionally also to an animal of another species when the latter is treated more or less as a conspecific partner by the sender) and it releases a definite response by the recipient (provided the latter had become aware of the sender's action, or does not deliberately ignore it). It would follow from these statements that expressive behavior without function in social communication is not addressed to and does not release a definite response from a partner. Generally, this is correct; however, there are behavior patterns which are important in communication but which are not addressed to definite partners but to potential recipients, "to whom it may concern." Other displays are clearly addressed to definite partners, but do not release marked responses by the latter. In certain situations, of course, no response is a response. This seems to be especially true in mating rituals where many artiodactyl females show no special reactions toward certain courtship displays of the males. Sometimes it may be difficult to distinguish such displays from expressive behavior having no function in social communication.

Fig. 1. The male Uganda kob (Adenota kob) shows the lip curl (flehmen) after having smelled the female's urine. (After photo by H. Buechner.)

Another difficulty arises from contagion. This means that an animal that is (presumably) in the right mood for a given behavior performs it when this behavior is exhibited by another animal close by. This contagion has to be distinguished from a response by the same behavior (for example, a threatened recipient returning the threat using the same display as the sender). Mere contagion, however, can hardly be said to be a response. One animal simply does the same thing as the other (yawning, eating, lying down, grooming, urinating), and neither addresses this behavior to the other; nor does the behavior itself call for a response. Occasionally, almost any behavior can have contagious effects. Thus, as with expression as an epiphenomenon, a presentation of contagious effects would inflate this paper to a general discussion of behavior. On the other hand, contagion can contribute quite remarkably to intraspecific communication in certain cases, as in the coordination of group activities. Therefore, it cannot be completely excluded from this discussion.

The statement that there is expression without function in communication can also be reversed, as communication does not necessarily depend on expressive behavior. For example, tracks, excrement, and other spoor that indicate that an animal of a given species, sex, and age is or has been present in a given area may have communicative functions, but they are not expressive behaviors. Moreover, at least in the broadest sense, the term "transmission of communication" could even be extended to cases where the behavior of the partner is influenced by merely mechanical means, as in a fight. When one thinks of the transitions in fighting behavior from all-out fights to ritualized fights, to playful sparring, and to gentle, but slightly aggressive pushing of the partner (tactile communication!), one may easily understand that it is hard to make a clear-cut distinction, even when one thinks of "communication" as primarily implying the imparting of information by signs and signals. In effect, this would mean the inclusion of fighting behavior, copulatory behavior, nursing, cleaning the young, social grooming, etc., in our discussion. This again would lead to an unwieldy inflation of this presentation. However, one has to be aware that there are such transitions and that the boundary between communicative and noncommunicative behavior is sometimes vague.

In short, the following presentation will focus on the elaborate and well-pronounced displays (expressions as phenomena) that release definite and clear responses in (conspecific) recipients. Expressions without a clear function in social communication, expressive epiphenomena, and behavior patterns that have only contagious effects on conspecifics will be discussed only as far as it is necessary for a better understanding of certain general problems of communication. Also, the discussion of communication without expressive behavior will be restricted to a minimum.

A last difficulty arises with respect to classification of the behavior patterns under discussion. An approach frequently used in the literature is the presentation and classification according to sense impressions (visual, acoustical, olfactory, and tactile behavior). This approach is clear and simple, but is also superficial and unsatisfactory, somewhat resembling a "classification" of plants by the colors of their flowers. Moreover, this approach cuts natural and well-integrated units of communication into pieces (although an occasional separation of behavior patterns that belong together can hardly be avoided by any kind of classification). Another approach is the classification of displays according to functional circles (Funktionskreise: von Uexküll, 1921). Thus, one may speak about aggressive displays, sexual displays, alarm signals, etc. This approach appears to agree better with biological situations than one using the sense impressions; however, it also has its problems. In the artiodactyles expressive displays often have one basic meaning and message, two (or a few more) effects on the recipient, and multiple social functions. For example, threat displays of certain species express the readiness of the sender to become aggressive. Depending on how equal the addressee and the sender are, such threats may have either challenging or intimidating effects on the recipient. In differing situations the same threat display may be used to establish or maintain a territory, establish or maintain a position in a social hierarchy, coordinate group activities, reject sexual approaches (as when used by a female toward immature males), or prevent strange young from suckling and soliciting milk. Thus, a classification of displays according to functional circles can also result in considerable and poorly founded separations between similar or even identical behavior patterns. Much the same is true of a classification based on involved partners (♂:♂, ♂:♀, ♀:♀), as was suggested by Carpenter (1942).

Another approach would be a classification based on phenomenological characteristics of behavior patterns such as head-up postures, head-down postures, broadside positions, etc. This approach is certainly a good one, but difficulties arise from the position of a given behavior pattern within the entire behavioral inventory of a species. Since behavioral inventories vary with the species, the meaning and message of phenotypically very similar behavior patterns can be different and even opposite in different species. For example, in certain bovid species, the opponents routinely drop to their "knees" (carpal joints) during a fight. If an animal drops to its knees in an agonistic encounter without establishing horn contact with its opponent, this action can be an intention movement for fighting and can thus be a threat. Other bovid species do not drop to their knees in fighting; when one of them exhibits this behavior during an agonistic encounter it is usually an intention movement for lying down, i.e., a submissive behavior. Thus, the meaning and message of the same behavior pattern (dropping to the knees) in the same situation (agonistic encounter) can be very different depending on its "bedding" in the species-specific behavior inventories. (By the way, this is also the root for misunderstandings in encounters between animals of different species.) Occasionally, similar problems may arise with respect to certain behavior patterns even within the same species. In short, a classification under merely phenotypical aspects can easily unite behavior patterns that are different in meaning, origin, and function, especially when the discussion includes many different species.

Possibly a classification according to phylogenetic origin of expressive displays would be helpful. However, our present knowledge of this subject is limited and often still in the stage of speculation.

The best approach may be to try a classification of displays with communicative functions according to their meanings and messages by using some aspects of functional circles for the outlines and, as much as possible, incorporating phenomenological and phylogenetical aspects (Walther, 1974) and limiting the use of categories derived from sense impressions. I propose that communication of artiodactyles be broken into the following categories: (1) advertising presence, position, state, and status, (2) excitement activities, (3) alarm and flight signals, (4) advertising readiness for social contact and group cohesiveness, (5) mother-offspring signals, (6) orientation relative to the partner and signals of direction, (7) threat displays, (8) space-claim displays, (9) dominance displays, (10) courtship displays, (11) submissive and appeasement behavior.

A few remarks on organs and parts of the body that play a role in the expressive behavior of artiodactyles may complete this introductory discussion. Artiodactyles are primarily pantomimers. In particular, the position of the neck relative to the body (stretched forward, erected, lowered) and the position of the head relative to the neck (head held in one line with the neck, chin tucked in toward the throat) often have very definite meanings and can signal a multitude of information to conspecific partners (Schloeth, 1961). In some species the torso can also show special postures (e.g., lordosis or kyphosis of the back).

Two aspects of the physical structure of animals of the artiodactyl type are noteworthy with respect to orientation of the entire animal, or at least its head, toward the addressee. First, the eyes of these animals are located much more laterally on the head than, say, in humans, monkeys, and certain carnivores. For this reason, a broadside position or a sideward turn of the head in Artiodactyla does not necessarily result in the loss or avoidance of eye contact with the partner. It often is one-eyed fixation of the other. Even when such an animal turns its body or head so far that it almost faces away from its partner, this movement is functionally comparable to a slight sideward inclination of the human head.

Second, these animals present their full breadth when standing in lateral position to an addressee, whereas they offer a relatively small silhouette when standing in frontal orientation. In humans the situation is just the opposite. Linked with the broadside position is the presentation by some species of striking color patterns (such as black or white stripes and bands) and/or additional structures such as beards and manes (which sometimes are extended over the entire back and can be erected).

Movements of the legs come after the postures of the head, the neck, and the torso with respect to importance in communication. Bending or stretching the legs can contribute to the appearance of body postures; and a slow-motion walk—sometimes combined with an exaggerated lifting of the forelegs—and various forms of symbolical kicking with the forelegs and stamping and scratching the ground are used as means of expression.

Facial expression is not lacking in artiodactyles, but apparently it does not play as great a role as it does, for example, in primates. Movements of the mouth and the mouth organs are quite common. For example, wide-open mouths in hippos and symbolic biting in Suidae and Tayassuidae ("squabbling": Schweinsburg and Sowls, 1972) are threats. Tongue flicking occurs in courting males of quite a number of species (especially in the cervids, but also in certain bovids). Folding the skin of the nose or inflating the nose region is found in certain species, such as gazelles (Walther, 1958, 1966a, 1968a). Eye movements—including a straight look, a look from the corner of the eye, and a pop-eyed look —do occur in artiodactyles. However, they have not been thoroughly investigated, and thus virtually nothing is known about their effects on recipients.

Ear movements occur rather frequently in connection with expressive displays in artiodactyles (Freye and Geissler, 1966). Examples of common ear movements are: the laying back of the ears in threat and courtship behavior; the "ear drop" (similar to the permanent ear attitude of Indian cattle) in courting males of Indian blackbuck (Antilope cervicapra), as well as of the gnus and hartebeests; the "pointing" with one ear toward the opponent in certain threat and dominance displays of Antilopinae and Hippotraginae species; and holding the ears sideways during fights in many bovid species. However, it is uncertain that ear movements are signals to conspecifics and release responses in them since they usually are combined with other, more striking postures or gestures of the head, neck, torso, or legs. Thus, the probability is great that the recipients may react primarily to these other behavior patterns.

The same may be said for most tail movements. Striking movements of the tail are combined with flight behavior in many artiodactyl species. However, they occur either when the animal is already fleeing or immediately preceding flight. Thus, it is hard to say whether conspecifics become alarmed by these tail movements or by the first animal's running away, its alarm posture, or its alarm calls. The communicative role of tail movements is even more dubious in courtship and threat displays. Here, the displaying animal is frequently frontally oriented to the recipient, and the latter cannot see the sender's tail.

According to Tembrock (1959, 1963, 1964, 1965), who has done pioneer work in studies of vocalization in mammals, the acoustical transmission of information in Artiodactyla has developed to different degrees. Species living in dense vegetation possess more differentiated sound patterns than those living in open landscapes, where optical information prevails. In the Suidae (and possibly other groups of Artiodactyla) three trends in differentiating vocalization seem to be significant in the development of information transmission: transmission of short sounds to long sounds, adding rhythm to the short sounds, and transformation of the frequency range. Besides a relatively few loud, striking sounds such as whistling, roaring, and barking, vocalizations are often very soft in the artiodactyles and can only be heard at very close range.

Kiley (1972) has emphasized that vocalizations in horses, pigs, and cattle are generally not discrete displays conveying specific messages but rather convey information on the general motivational state of the animal (Table 1). Certainly, there is much truth in these statements (which in part refer to certain general problems of expressive behavior discussed above). However, it would be too sweeping a generalization to deny the display character of all vocalizations and their conveyance of specific messages.

Vocalization can be brought about by the combined activities of the larynx and the mouth organs. Some sounds appear to be closely related to belching. Sometimes, special postures of the neck and/or head appear to be necessary to produce certain sounds (for example, the rutting call of red deer). In a number of artiodactyles (such as gazelles and some other bovids), sounds uttered through the nose are quite common. In this case, certain cartilaginous structures of the nose (vibrating organs), skin folds and skin bags in the nose region that can be enlarged, and the opening or closing of the nostrils produce these vocalizations. Some noises can also be made with the teeth.

Artiodactyla are considered to be macrosmatic. Besides urine and feces (and possibly saliva), secretions (pheromones) of skin glands are supposed to be important in intraspecific communication. Skin glands are frequently found in artiodactyl species. However, it has to be emphasized that in the majority of cases, the importance of these glands and their secretion in intraspecific communication is at present postulated on the fact that artiodactyles have such glands and a keen sense of smell. Except for the occasional sniffing at glands or secretions, and adding of urine, feces, or secretion to an already existing marking site, convincing observations or responses to these scents are very rare. Assumptions on repellent effects of secretion marks to potential territorial competitors should be treated with particular caution. The studies of Müller-Schwarze (1967, 1969, 1971) have recently brought a clearer picture of the social functions and effects of certain pheromones in mule deer (i.e., in one species of Artiodactyla). As far as I can see no statements or only very limited ones can presently be made on the communicative functions of the following glands: mental, crural, circumcaudal, infracaudal, circumanal, proctodeal, prevulval, preputial, inguinal, tibial, parungular, interdigital, and occipital. However, all of them may possibly play a role in communication of certain artiodactyl species.

Table 1

A. Syllables of cattle calls and the way they combine to form the calls (Kiley, 1972). 1 = low, m = medium, h = high. B. Interrelationships of the calls in cattle (Kiley, 1972). Note that amplitude, frequency, length, and/or repetition increase with excitement.

In tactile encounters such as nosing, licking, or rubbing, the epiphenomenal mode (gentle or violent, brushing or knocking, with increasing or decreasing pressure, etc.) appears to be of greater importance with respect to transmission of communication than are the behavior patterns themselves, but no studies on this subject are available at present. For technical reasons discussed above, it appears inopportune to include in the transmission of communication other tactile stimuli, such as horn or antler contact, biting, shoving, or throwing the body on the partner.

I will now discuss the displays according to the categories suggested.

Advertising Presence, Position, State, and Status

It is a basic presupposition for any communication that both partners are aware of each other's presence and position, and often also their social status (such as territorial or nonterritorial, high ranking or low ranking in a social hierarchy) or of their general motivational state (such as alarmed, relaxed, in migratory mood). Apparently, the mere presence of an animal is not necessarily sufficient for these purposes, but additional advertising devices are required. Thus, the basic message of such advertisement behavior is, "Here is an animal of this given species," often combined with the modification "And it is up and doing" and/or "It is in this particular mood, social status, sex, or age."

Emphasizing presence always means selfexposure of the animal, either relative to other conspecifics (social self-exposure) or to the environment or a combination of the two. Visual selfexposure relative to conspecifics is only found in gregarious or semigregarious artiodactyles, mainly in adult males. The self-exposing animal separates itself from the group most of the time. In species that form temporary and/or seasonal harem groups, such as Grant's gazelle (Gazella granti), pronghorn (Antilocapra americana), topi (Damaliscus lunatus), and impala (Aepyceros melampus), it is commonly the male who, linked with his "shepherd" role, keeps himself separated from the females. This means that he usually stands, moves, and rests at the periphery of the group, often ten to thirty meters or more from the females. Thus, his presence can easily be recognized by other males. Visual self-exposure is sometimes combined with courtship or dominance displays. It appears likely that the selfexposure forms the base for certain dominance displays (p.690).

In the case of nonterritorial species, such as red deer (Cervus elaphus), or in species where territorial behavior is combined with harem behavior and males have very large territories, such as Grant's gazelle, the male remains separated from the females but moves with them. In species such as wildebeest (Connochaetes taurinus: Estes, 1969), Thomson's gazelle (Gazella thomsoni: Walther, 1964a; Estes, 1967), and Uganda kob (Adenota kob: Buechner, 1961; Leuthold, 1966), territories are comparatively small and female herds visit the territorial males for only a few hours per day (pseudo-harems). In this case, territorial males do not participate in daily movements of female herds (or, of course, in those of bachelor groups), but they remain behind as solitary individuals. This separation is characteristic of the territorial status of these males (but it is neither the only nor an unmistakable indication of territoriality).

Visual self-exposure relative to the environment means that an animal does not make use of cover (otherwise a very common strategy in animals), but instead stands on the rim of a slope or on top of a rock, hill, or termite heap. This standing freely on elevated ground can serve several functions. It may sometimes be used for better observation of the surrounding area. In hot climates, this position may allow the wind to cool the legs. On the other hand, this position also emphasizes the animal's presence, allowing some communication—be it repellent to rivals and competitors or attractive to potential sexual partners.

Social and environmental exposure can also be combined. For example, in chamois (Rupicapra rupicapra) and in pronghorn the male often separates himself somewhat from a female group during rutting season, but he stands or rests above them on the slope of a mountain or hill.

Acoustical advertising of presence and position is sometimes combined with visual displays or advertising, as when territorial bulls of brindled wildebeest (Connochaetes taurinus) combine their groaning-croaking calls ("ugh") with a head-up posture, and black wildebeest (Connochaetes gnou) males combine their shrill, blaring call ("he-it") with a throwing upward of the head. In some species, acoustical advertising may possibly be used because visual displays are not effective enough in the habitat (water, high grass, thicket, forest) in which these animals live. Examples of this kind might be the trumpeting grunts of hippo (Hippopotamus amphibius), the roaring rutting call of red deer (Cervus elaphus hippelaphus), the bugling of elk (Cervus elaphus nelsoni)_f and the long roaring of bulls in certain bovines (Schloeth, 1961).

In a few species, such sounds are apparently uttered primarily by territorial males and may advertise their territorial status. Possible examples are the whistling of Uganda kob (Buechner and Schloeth, 1965), the burring and rhythmically repeated "pferrr" (usually five to a stanza) in Grant's gazelle, or the loud, strophic panting in pronghorn.

Perhaps most acoustical signals function as vocal contacts that work on the principle of feedback. In this case, the advertisement of presence and position ("I am here!") calls for the partner's answer ("Where are you?"). In other words, the advertisement of presence and position extends into the realm of advertising the readiness for social contact.

Olfactory advertisement of presence has two principal aspects. First, many artiodactyles can emit scent from skin glands, some of which, the inguinal glands, for example, are open and produce secretion more or less constantly. Thus, this scent is always with the animal. Other glands secrete only during certain periods of the animal's life, as when the belly gland of musk deer (Moschus moschiferus), ejects a strong-smelling secretion during rut. Again, other glands are opened automatically when the animal performs certain movements; for example, the interdigital glands are more or less closed while the two hooves of one foot are close together, but they open when the hooves are spread, i.e., when the feet are abruptly forced against the underlying surface during galloping, jumping, stamping, or pawing the ground (commonly with one foreleg in artiodactyles). They are also spread when the animal emphatically stretches a foreleg, as in the foreleg kick (p.705). Certain glands are sometimes opened in combination with visual and acoustical displays, as when a roaring red deer automatically opens its preorbital glands when opening its mouth and stretching its head and neck forward and upward in the typical ruttingcall posture. On the other hand, males of certain species such as Thomson's gazelle and Indian blackbuck can open their preorbital glands at will. In this case, the (presumable) emission of scent can be timed and restricted to special occasions, primarily agonistic and sexual encounters.

Another possibility of olfactorily emphasizing presence is self-impregnation with urine (sometimes also with sperm). This is used by the males of some artiodactyl species mainly during rut. Males of Gray's Waterbuck (Onotragus megaceros) and Capra spp. splash urine into the hair (beard) of the throat region. Odocoileus spp. (Fig. 2a) as well as moose (Alces alces; Geist, 1966b) and reindeer (Rangifer tarandus; Espmark, 1964), and possibly all cervids within the Telemetacarpalia group urinate on their hind feet (besides "normal" urination). The urine runs over the tarsal glands, and the animal rubs its hind feet together, apparently mixing the urine scent with that of the secretion from these glands (Müller-Schwarze, 1971). This seems to be especially common in dominant individuals. Thus, another animal can learn something about the social status of the bearer when sniffing at its tarsal glands. The European bison (Bison bonasus) will wallow in its own urine (Hediger, 1949). During rutting season, male moose may lie on the spot where they have previously urinated (Kakies, 1936), and territorial males of blesbok (Damaliseus dorcas phillipsi) often rest in the middle of their dung piles (Walther, 1969a).

Fig. 2. a. Mule deer (Odocoileus hemionus) in rub urination, b-d. Typical sequence of (b) pawing the ground, (c) urination, and (d) defecation in a male pronghorn (Antilocapra americana). The hoof scratches on the ground, the urine, and the feces are all deposited at the same spot.

The second method for olfactory advertisement of presence and position is by object marking. In this case, the animal deposits odoriforous substances on definite spots of the environment. Since these marks remain for some time after the animal has left, their message is not only "I am here!" but also "I was here!" Again, the secretion of skin glands as well as urine and feces are used by certain artiodactyl species for this purpose. The remains of other activities such as pawing the ground and rubbing or goring of ground and vegetation with horns or antlers may also serve a similar purpose. Pawing the ground is often combined with urination and/or defecation in males of many bovid species. In cervids, gland marking is often combined with aggression against inanimate objects.

Besides some of the glands already mentioned in another context (e.g., interdigital glands), of the skin glands the subauricular glands (in pronghorn), dorsal glands (e.g., in peccary), frontal glands (e.g., in roe deer), postcornual glands (e.g., in chamois and mountain goat), and, above all, the preorbital glands (in many bovid and cervid species) are of special importance in marking, and deposition of the secretion often requires special movements and postures (Fig. 3a). It has to be emphasized, however, that not all artiodactyles that possess such glands mark with them. This varies even within a subfamily or genus. For example, all Antilopinae species have preorbital glands, but only some of them, such as Thomson's gazelle, goitered gazelle (Gazella subgutturosa), red-fronted gazelle (Gazella rufifrons), blackbuck, and gerenuk (Litocranius walleri), mark objects with them, whereas other Antilopinae species, such as Grant's gazelle, Soemmering's gazelle (Gazella soemmeringi), and dorcas gazelle (Gazella dorcas), do not. In mountain gazelle (Gazella gazella) there are even variations within the same species (provided that the classification system presently used is correct): Gazella gazella benetti marks, Gazella gazella gazella does not. In all Antilopinae that do mark, only the males do so, and the same is true for a great number of other bovids and cervids. However, in klipspringer (Oreotragus oreotragus) and blesbok the females also mark occasionally.

Territorial animals mark their territories, but gland marking does not necessarily imply that the individual is territorial. For example, many cervid species mark rather intently with their preorbital glands. Axis deer (Axis axis) do it in a very striking fashion (often combined with object aggression) by rising on their hind feet and depositing the mark on a branch, as high above the ground as possible. However, the majority of cervid species are not territorial. A number of them have—obviously prematurely—been assumed to be territorial, but territoriality is a proven fact at present only in roe deer (Hennig, 1962). Even within the same species, object marking is not restricted to territorial individuals. For instance, the territorial males of Thomson's gazelle mark their territories with preorbital gland secretion, but nonterritorial males in bachelor groups and mixed migratory herds show the same—if less frequent—behavior (Walther, 1964a). Thus, the mark perse only indicates the presence or past occurrence of a male but does not necessarily indicate his territorial status. It is the concentration of marks by the same individual within a limited area and the specific marking system, such as a belt of marks along the territorial boundary, that make marking indicative of territoriality (Walther, 1964a).

Fig. 3. Marking with preorbital glands, a. Indian blackbuck (Antilope cervicapra) marking the top of a branch, b. Male dibatag (Ammodorcas clarkei) marking the female.

It is tempting to assume that these secretion marks would have an intimidating or repellent effect on potential competitors. Occasionally one can observe such reactions; however, on the whole they are rare. In Thomson's gazelle, nonterritorial males as a matter of course enter wellmarked territories without paying any attention to the marks. In territorial marking, the marks are possibly more important for the orientation of the owner of the territory than for other animals. The most commonly observed reaction by an animal to another's secretion mark is to sniff the marked object; sometimes the newcomer will mark it, too, or mark another one close by. Thus, the first animal leaves an indication of his presence, and the newcomer adds his own "visiting card."

Almost all that has been said about effects and functions of secretion marks is also true of marking with urine and/or feces. Urination and defecating are normal physiological processes in all mammals. Therefore, one can speak about a special function of excrements only when other conspecifics clearly react to them, and/or when an animal deposits urine and feces on definite places (resulting in dung piles in the case of defecation), and/or when urinating and/or defecating behavior shows features exceeding the mere need to excrete digestive waste. For example, the tylopods of the New World have a certain "defecation ritual" (Pikers, 1954): an animal will sniff at a fixed dung pile, stamp and paw it, then add its own droppings to it. Sometimes several animals do so simultaneously. Male hippo and pigmy hippo (Choeropsis liberiensis) splash their feces around with swirling movements of their tails. In some artiodactyles (such as Gazella and Antilocapra), urination and defecation postures of males are exaggerated (Fig. 2c); scratching the ground with a foreleg often precedes the process, and defecation follows immediately after urination. Thus, urine, feces, and tracks of scratching (and possibly, some secretion of interdigital glands) are deposited at the same spot. In other genera, such as Oryx, Addax, and Damaliscus, urination and defecation are separated from one another. Only defecation is preceded by pawing the ground and, in adult male oryx and addax, it is performed in a deeply crouched posture. Again, in other species and/or in other situations only the urine conveys information on state of the animal. This is especially true for urinating by females sexually driven by males. By sniffing or licking the female's urine, followed by flehmen in most of the Tylopoda and Ruminantia, the males can obviously find out whether the female is in or close to estrus.

To establish relatively large dung piles, an animal must be in the same area for some time. This is easily achieved in territorial species. In many, as in vicuna (Lama vicugna), Kirk's dikdik (Rhynchotragus kirki), topi (Damaliscus lunatus topi), and the Gazella species, dung piles are found either in the approximate center of the territory (apparently close to the owner's preferred resting place) or along the boundary (linked with agonistic encounters). Sometimes several individuals may use the same dung pile, making it larger than one animal alone could do. This may happen with the owners of neighboring territories, but may also occur with apparently nonterritorial species such as nilgai (Boselaphus tragocamelus). Generally, there are several tendencies that may contribute to the establishment of dung piles in artiodactyles: Animals often defecate after a long rest. Thus, when a species has fixed resting places in a territory or a home range, there is often an accumulation of dung near each. Furthermore, animals often defecate in excitement. Thus, animals often defecate at the same spot in an area where agonistic encounters frequently take place (such as along the boundaries of a territory) or where a frequently used trail leaves a relatively safe area to enter more dangerous terrain. In quite a number of artiodactyl species there is also a tendency to place droppings where another conspecific has previously defecated. Finally, these animals often show a preference to establish dung piles on bare ground, such as on their trails and even on human trails, where they are very visible and exposed. All these points indicate that dung piles play a certain role in the social communication of these animals. However, it seems to be a rather "anonymous" kind of information, which is not commonly addressed to definite recipients but to potential partners, "to whom it may concern" (except for marking in connection with agonistic encounters).

Excitement Activities

The term "excitement activities" refers to movements and vocalizations that indicate that an animal is in a state of agitation. They all occur in several heterogenous situations, and some occur in many such situations—when an animal watches an enemy at a distance, when it is forced to cross unfavorable terrain, when it is separated from a familiar group or mate, when it is involved in an agonistic encounter, or when it is expecting food or is prevented from getting food in captivity—but do not contribute to the solution of these situations.

Excitement activities in the described sense, are often termed "displacement activities" (Tinbergen, 1940). I hesitate to use this term for a number of reasons: It does not cover the whole range under discussion; the theoretical conception underlying this term implies that the animal is in an inner conflict that in some cases, at least, cannot be observed and thus remains an open question; and finally, when using this term one has to take the conception of functional circles (or major instincts) so literally and stick with it so tightly that it does not appear to be justified in the light of certain facts, such as the multiple functions of expressive behavior.

Excitement activities frequently seen in Artiodactyla are self-grooming; scratching and shaking (not as a reaction to itching or insects); the volte (stepping around in a narrow circle); stationary vertical jumps; and, in some species, also stamping (with a foreleg). Kiley (1972) lists a considerable number of vocalizations (grunts, squeals, snorts, "mm" calls) in various artiodactyl species that generally fall in the category of excitement activities..She also uses the term "excitement" and states that these vocalizations reflect its level (Table 1). Thus, the excitement activities that mainly advertise the general motivational state of the animal have a certain importance for communication insofar as they can be contagious and can bring a conspecific animal into the same mood as the sender. In general, however, they are not addressed to a definite partner, and even in situations where this could be possible, they do not elicit special and definite responses (above the level of mere contagion).

Alarm and Flight Signals

Alarm signals can be considered as special kinds of excitement activities, as they correspond to very high levels of excitement. This excitement, however, is often not just of a general and unspecific nature. I think it is necessary to look at these behavior patterns in an evolutionary context. Apparently, they are "on their way" from mere expressions of high excitement to becoming special displays; however, none have yet completely reached this final state in the artiodactyles. It follows, then, that all these behavior patterns occur at least occasionally in situations having nothing to do with alarm, and merely express a high level of excitement. For example, this is probably true in the cases of so-called displacement alarm (Estes, 1969; David, 1973) when behavior patterns commonly seen in alarm situations also show up in agonistic encounters of certain species. The degree of perfection in this change from unspecific excitement activities to specific alarm signals varies within different species as well as with respect to single behavior patterns. Thus, these behavior patterns are probably not alarm signals by nature and origin, nor do they exclusively convey and release alarm; however, the latter is one of their striking and common functions. In this aspect, it is not out of place to use the term "alarm signals" (provided one is aware of the relativity of this terminology).

At least in certain situations, every fast-running or leaping animal—not necessarily even a conspecific—may attract the attention of other animals and release alarm or flight reactions in them. This "running away" can be made more conspicuous by striking locomotor patterns (such as stotting), special movements or postures of the tail, ruffling hairs of the (white) rump patch, emitting scents from certain skin glands, and distress cries. The last should be distinguished from alarm calls (p.668). Distress cries are uttered only when an animal is captured or is very close to being captured. In most artiodactyl species, distress cries sound like a roaring or bleating "aaaaa" (a as in höre) or "uuuuu" (u as in mwrder). They are long, loud sounds made with an open mouth, are absolutely situationspecific, and usually have a strong alarming effect on other animals. However, release of alarm in conspecifics is clearly a secondary effect of a distress cry since it is also uttered when a pursued or captured animal is completely alone. Although distress cries are sometimes apparently caused by pain, they generally appear more closely related to fear. For instance, some of my captive dorcas gazelle regularly gave loud and persistent cries when being captured for veterinary treatment but very seldom cried out during the treatment itself, although it was often likely to be painful. Although distress cries may be uttered by adults in mortal fear, they are more commonly heard from young animals. Distress cries of the young release a mother's defense against a predator, provided it is not too big and dangerous. In the latter case, the mother may perform distracting maneuvers, such as crossing several times at full gallop between her fleeing young and the pursuing predator. In certain species mothers often cease their defense when the young is killed and is no longer crying (Walther, 1969b).

In mule deer (Odocoileus hemionus), metatarsal scent is discharged in fear-inducing situations (Müller-Schwarze, 1971); and in springbuck (Antidorcas marsupialis), an emission of scent from the dorsal gland is likely during stotting (or "pronking," as the modified form of stotting of the springbuck is often termed: Bigalke, 1972). Other olfactory alarm signals appear to be possible in artiodactyles; however, at present, this is more or less subject to speculation.

Among alarm-releasing locomotor patterns, the so-called stotting (Fig. 4c) deserves mention. This special, striking kind of jumping (usually not used for clearing obstacles), often results in a chain of leaps, during whichthe animal bounces up and down with all four legs rather stiffly stretched. Stotting is common in Antilopinae species but is also found in pronghorn and in certain cervid species such as fallow deer (Dama dama) and mule deer. It apparently corresponds to a high—but not the highest—level of running and flight excitation. Thus, it occurs when excitation is rising or falling (Walther, 1964a, 1969b). It is predominantly seen at the beginning of flight (provided the pursuer is not too close) and at the end (when the enemy has ceased pursuit). It is more frequent in young animals and in females than in adult males.

Fig. 4. Alarm behavior in Thomson's gazelle (Gazella thomsoni). a. Looking around, relaxed (adult female), b. Long-neck posture of alertness (adult male), c.-e. Stotting (adolescent female): c. Normal stotting gait. d. Paddling with the hind legs in extremely high stotting. e. Landing from high stotting.

Ruffling of rump patch hairs (springbuck also ruffle the white hair in the pouch of the croup, which has a dorsal gland inside) is combined with flight (and also stotting) or may precede flight in most species having a rump patch. The same is true for certain tail movements of some species. A relatively common movement is the vertical erection of the tail, found in various artiodactyl species such as warthog (Phacochoerus aethiopicus), mountain gazelle, dibatag (Ammodorcas clarkei), and white-tailed deer (Odocoileus virginianus). During flight Tragelaphus species curl the tail up so that the tip almost touches the root. An important aspect of the signal character of such tail movements in some species (e.g., in Tragelaphus and Odocoileus) is the exposure of the white underside of the tail when it is erected or curled.

Of course, there are also other tail and body movements linked with flight. For example, giraffe (Giraffa camelopardalis) presses its curved tail laterally to its hindquarters at the beginning of flight, and certain other species, such as Thomson's gazelle, Grant's gazelle, and Kirk's dikdik, often shake their flanks before flight.

However, as stated above—with the exception of distress cries—all these behavior patterns can also occur in situations other than flight, where they do not release alarm in conspecifics and sometimes do not even attract attention.

When watching a potential danger, all artiodactyles tense their muscles and stand as motionless as a statue (posture of alertness). They are oriented frontally toward the dangerous object, with ears turned forward. In many species, the animal stiffly erects its neck to its maximum height ("long-neck" posture, Fig. 4b). It sometimes also stamps with a foreleg and frequently utters special sounds (alarm calls). In some artiodactyl species, alarm calls are produced in the mouth, as in the loud, doglike barking of many cervids and of the genus Tragelaphus. Others are produced in the nose (which, by the way is also true for quite a number of other sounds, especially in bovids). In the latter case, alarm calls can also be very loud, like for example, the whistling of ibex (Capra ibex), chamois, and reedbuck (Redunca redunca); or they can be of medium volume, like the snorting of wildebeest and topi; or they can even be rather soft, like the "quaking" alarm calls of some gazelle species. Especially in Thomson's gazelle, the alarm call, "quiff" (i as in hull), is so soft that a human cannot hear it beyond a distance of 30 m. Nose calls are apparently vibration sounds the timbre of which may vary considerably with distance.

The significance for communication of an animal standing in the posture of alertness, uttering alarm calls, and conspecifics reacting by becoming alert, is naturally much clearer than for many other behavior patterns that may also occur in situations of alarm and flight. On the other hand, the posture of alertness and, at least in certain species, the vocalizations under discussion may also occur in certain other (exciting) situations, and even when the animal is alone. Thus, in most cases, they do not appear to be addressed to a definite partner.

Advertising Readiness or Need for Social Contact and Group Cohesiveness

Signals advertising readiness or need for social contact or group cohesiveness can be considered as modifications of signals advertising presence, position, and general mood. In common with the latter group of signals as well as with those indicating alarm and flight, they are usually not addressed to definite partners.

Perhaps the closest relation between signals of presence and those facilitating social contact is found in the visual field. This is especially true in open plains areas, where the figure of an animal (of the size of an ungulate) is often the most striking sight in its vast and uniform surroundings. Here, body markings (such as black stripes or bands, white rump patches) and sometimes tail movements (like the almost perpetual tail wagging of certain gazelles) may play a role in making an animal more recognizable to conspecifics. Like landmarks, the striking figure of an animal on the open plain attracts the attention of other animals, which often turn toward it and approach it (Walther, 1972). In this way, conspecifics join to form groups and, eventually, the large herds so typical of many artiodactyl species (bison, wildebeest, springbuck, gazelles, reindeer) on open plains.

Because of their annual and circadian rhythms, all the animals in such herds are in approximately the same mood, and many behavior patterns can become contagious, contributing to the coordination of activity within the group. This synchronization of group activities is of special importance during moves and migrations. Obviously, a moving conspecific easily causes others to follow. One may think here of a modification of the infantile following reaction. In this regard, it is certainly important that in most artiodactyles, herds march in file, with one animal behind the other, at least during moves of some length. This means that the animal behind always has the preceding animal's rump in front of him. Presumably, rump patches and tail movements play an additional role in releasing this following reaction. This appears valid even in some solitary-living species, where social attraction is generally restricted to sex partners or offspring. An example is the blue duiker (Cephalophus monticola), which flips its little tail up and down, almost like a reflector when moving. Since the underside of the tail is white, it may act like an intermittent light in the dim forests in which these little creatures live.

Sounds used in vocal contact are also related to signals advertising an animal's presence and position. In some cases, they are more or less identical with the latter. An acoustical signal for social contact easily releases and, in a sense, calls for a partner's vocal response, which is often given in the same sound as that uttered by the sender. These sounds are frequently repeated, sometimes in a rhythmical manner. Relatively soft grunting sounds in pigs (Hainard, 1949; Snethlage, 1957), cattle (Schloeth, 1961; Kiley, 1972), red deer (Darling, 1937; Burckhardt, 1958a; Kiley, 1972), fallow deer (Gilbert, 1968), and axis deer (Schaller, 1967); roaring and growling in camels; and bleating in llamas (Pikers 1954, 1956) may be mentioned as examples. In large herds of certain gregarious artiodactyl species, these vocal contact sounds are not addressed to definite partners, but are expressions of particular moods (general motivational states) of an animal. Herd members may generate and sustain a particular mood throughout the group by reciprocal uttering and repeating of these sounds. They may also be important in species recognition and cohesiveness of conspecifics. Perhaps the most impressive example of this is provided by migratory herds of wildebeest, which are almost constantly vocalizing relatively loud croaking calls, so that the herds are enveloped in "clouds" of (familiar) noises.

In some species, the members of a group may olfactorily impregnate each other. For example, in peccary (Tayassu tajacu), members of the same group pair up in reverse-parallel position and rub their heads on each other's dorsal glands. Since it is likely that all members of a group eventually exchange and mix their individual scents in this way, such behavior may result in creating a mutual "group scent." Male gerenuk and dibatag (Fig. 3b) mark females with their preorbital glands, and male Gray's waterbuck urinate on the long hair of their throats and rub this wet region on the backs of females. In these last cases, the communication tends to become more specifically addressed than the other, rather "anonymous" actions discussed above.

Signals in Mother-Offspring Relations

The communication between mother and offspring adds few new aspects to the discussion of advertising presence and position, readiness or need for social contact, and alarm signals. The means used in mother-offspring relations are only special cases and are often identical with the signals mentioned above. Thus, a difference is not so much in the signals per se but in their being clearly addressed to a definite partner and releasing rather pronounced and special responses of that partner. This is closely linked with the bond between and the individual recognition of mother and young. Although these topics exceed the realm of a discussion on communication, they are so important to an understanding of communication between mother and offspring that one must at least describe them briefly.

In all artiodactyl species, the young follow the mother (Fig. 5b), but there are species-specific differences in form and intensity of this following behavior. Neonate artiodactyles follow moving objects close to them and larger than themselves. Under natural conditions, it is highly probable (although not absolutely certain) that the object will be the mother. Later the young show a strong preference for following their mothers. The mother usually has only to move away to get her young to follow. If, for some reason, this does not work, she may call her young, or walk back to it and touch it with her nose, or circle around and pass it in a fast gait from behind. The last method appears to be most effective in releasing the following reaction (Walther, 1969a).

Apparently, the individual bond of a young artiodactyle to its mother is due to imprintinglike processes. It is uncertain whether there is visual imprinting in artiodactyles, but acoustic and olfactory imprinting to an individual maternal partner is more certain. The young apparently learn to recognize their own mother's voice, and, at least in certain situations, they react to it by approaching (Walther, 1959). Sounds used by mothers in calling their young— like that for nursing—do not seem to differ from vocalizations commonly used in group contact by such species as blackbuck and dorcas gazelle. In other species, special maternal calls have been described: growling and bleating in camel (Pikers, 1954, 1956), bleating in fallow deer (Tembrock, 1968), guttural grunting in cattle (Schloeth, 1958, 1961).

Fig. 5. a. Head bobbing of a female caribou (Rangifer tarandus) to a neonate calf. (After W. D. Berry in Pruitt, 1960.) b. Chamois (Rupicapra rupicapra) fawn following the mother.

Visual displays may serve the same or similar purposes as these maternal calls. For example, in reindeer, Pruitt (1960) describes how a mother can lure her neonate calf by head bobbing (Fig. 5a). Also, I could make my tame blesbok calf approach me by (silent) bowing movements, similar to those that adult blesboks frequently perform with the head and neck (Walther, 1969a). Apparently, in these cases, the effects of vocal and visual "calling" of the young act cumulatively when displayed simultaneously or alternately.

The young may also contact its mother vocally, and there is often a "question-answer" vocalization between mother and offspring. Sounds made by the young are usually higher in pitch than the calls of the adult animals, but they may often be the infantile forms of adult vocal contact sounds. The mother reacts to the call of her young by calling back and/or by approaching it. When moving, she may stop and wait for it. Reactions of mothers to distress calls of their offspring have already been mentioned.

A blesbok calf I raised (Walther, 1966a, 1969a) was strongly imprinted to me (individually) in the olfactory realm. One may assume that in natural conditions, this is also true with the mother, and it probably holds true in a number of other artiodactyl species.

In certain duikers, mothers have been observed marking their young with their preorbital glands (Frädrich, 1964). This may facilitate recognizing their own offspring and distinguishing them from strangers. The licking of the young by the mother (i.e., wetting the young with the mother's saliva and, thus, possibly impregnating them with her scent), which is widespread in cervids and bovids, may also have a similar result (as a secondary effect—the primary functions of licking the young are different). Generally, licking may contribute to the establishment and maintenance of the bond between mother and young in many Ruminantia. It may also convey certain messages. However, little is known as to the identity of these messages.

Orientation Relative to the Partner and Direction Signaling

All expressive behaviors indicate momentary psychosomatic states of the sender. Whether and how the recipient responds depends, at least to some extent, on its central-nervous evaluating mechanisms and psychosomatic state. With respect to communication, an additional problem arises in addressing the partner. Apparently, there are three possibilities in artiodactyl communication:

(1) The behavior is not addressed to definite partners, but only to potential ones. It is left almost completely to the potential partners whether they relate the performer's behavior to themselves and whether to respond to it. This "to-whom-it-may-concern" type of communication is found in most of the signals advertising presence, position, state, and status; excitement activities; alarm and flight signals; and in many signals advertising readiness or need for social contact and group cohesiveness.

(2) An individual bond is established between the partners (for example, by imprinting). Based on this special attachment, each partner automatically relates the signals to itself and responds to them. This is true for some signals advertising readiness or need for social contact and, above all, for signals used in motheroffspring relationships. (In this latter case, however, orientation components are sometimes involved; see below.)

(3) The sender clearly addresses a definite partner. The "I-mean-you" component is largely brought about by a change in the sender's orientation relative to the addressee (usually) at close range. Frequently (but by no means necessarily), the sender approaches the recipient before or after the new orientation is achieved. In certain cases, the addressee's position relative to the sender also plays a role. Having moved to a new position, the sender remains in it for a while. Thus, the previous movement contributes to addressing the partner, but apparently the sender's orientation per se also has a definite meaning and message.

The three basic orientations of the sender to the recipient are: frontal, reverse, and lateral. In each case, the sender may orient its whole body toward the recipient. Sometimes, however, the animal only brings its head into the corresponding position, in which case, the head obviously substitutes for and represents the body (pars pro toto).

Especially when the sender is moving, orientation of the hindquarters toward the recipient may easily release the following reaction of the latter, as discussed above. It has to be added that the same orientation when shown by an inferior animal in agonistic situations is suitable either to releasing the superior combatant's pursuit or to diminishing its aggressiveness. There can also be a sexual component in it, since it is the usual orientation of the female toward the male in mounting and copulation, and females of many artiodactyl species walk in front of the male in the course of the mating ritual. Generally, in artiodactyles, the orientation of the hindquarters toward the partner frequently expresses peaceful intentions or even inferiority.

Frontal orientation is often indicative of hostile intentions. (The most remarkable exception to this rather rough and general rule is, in many species, the frontal orientation of the mother toward her young when calling to nurse.) This is very understandable since most organs used for fighting (teeth, tusks, horns, antlers, neck, forelegs) are located on the anterior part of the body in the Artiodactyla. Thus, when turning, approaching, or standing in frontal orientation toward a conspecific, an artiodactyl has directed practically all its potential weapons toward the partner.

Expressis verbis or implicitly, the broadside position has sometimes been considered as resulting from a conflict between aggression (frontal orientation) and escape tendencies (hindquarters position) (e.g., Fraser, 1957; Ewer, 1968). However, at least in Artiodactyla, this is rather unlikely. There are a few artiodactyl species, such as mountain goat (Oreamnos americanus: Geist, 1965), giraffe (Backhaus, 1961), and Barbary sheep (Ammotragus lervia: Haas, 1959), that fight either regularly or occasionally in parallel or reverse-parallel position. In these species, the broadside position is clearly related to aggressive behavior only. Geist (1966a) has expressed the opinion that fighting in parallel or reverseparallel position is a phylogenetically old fighting technique in artiodactyles. Under this assumption, it is also possible to consider broadside orientation as a phylogenetic relic of aggressive intentions in those recent species that no longer fight in this position.

Above all, however, an animal of the physical constitution of Artiodactyla can block a partner's or opponent's path by assuming the broadside position in front of him. This blocking of the path is not a theoretical assumption or postulation, for it can actually be observed in quite a number of the species. Apparently in a relatively few artiodactyles, such as cattle (Schloeth, 1958), greater kudu (Tragelaphus strepsiceros), and sitatunga (Tragelaphus spekei: Walther, 1964b), the young may block its mother's path so that it may suckle. More numerous, by far, are the cases where a superior animal blocks an inferior's path by broadside position. This has been observed in lesser (Tragelaphus imberbis) and greater kudu (Walther, 1958, 1960a), oryx (Oryx gazella beisa: Walther, 1958), Grant's gazelle (Walther, 1972, 1974), Thomson's gazelle (Walther, 1974), mountain gazelle, hartebeest (Alcelaphus buselaphus: Gosling, 1974), tsessebe (Damaliscus lunatus lunatus: Joubert, 1972), warthog (Frädrich, 1965), and others. By taking this lateral position in front of a recipient, the performer forces the other to stop and often also to withdraw or at least deviate from its original course.

All these orientations can occur together with special displays. Typically, frontal orientation is frequently combined with threat and courtship displays, broadside position with dominance and, to a lesser extent, with courtship displays, and hindquarter orientation with behavior of inferiority. This will be discussed in detail later. However, it must be emphasized that these positions per se (i.e., without additional displays) can, in certain situations, release responses in conspecifics. Apparently, their major effect is to inform the recipient of the direction it is expected to take or not take. Presuppositions, facilitating such direction signaling, are that the sender is superior or at least not clearly, inferior to the recipient, and that the latter is somewhat ready to "obey" the sender's intentions. To date, these problems have been studies in detail for only one species, Grant's gazelle. However, there is reason to assume that the results obtained are valid for quite a number of other artiodactyl species.

In Grant's gazelle, direction signaling is especially obvious when a relatively stable group of females (and their offspring) remains for several months in the (large) territory of a male (Walther, 1972). In such a harem group, only the male is territorial and only he is aware of territorial boundaries, which do not exist for the females. Thus, the females will transgress the boundaries without hesitating if the male does not prevent them from doing so. When females are at the point of leaving the territory, the male's efforts to block their path and to herd them back are very striking (Fig. 6).

This situation, however, is an extreme case. More commonly, the male tries to direct their course longbefore they near the boundary, simply by using his position relative to the females and placing himself between the harem and the boundary. He shows no particular display, but stands, moves, or, most commonly, continues grazing. He often directs and relates his position to only one of the females, the one that has moved farthest in a given direction. Her position relative to him can also modify the meaning of his position. The females definitely react to the male's behavior (as long as they do not try to leave him and the area deliberately). Such permanent direction signaling works so inconspicuously and effectively that it took me a shamefully long time to become aware of it. Although it may appear to be somewhat childish and anthropomorphical, the simplest and perhaps even the only way to characterize meanings and messages of these positions is to "translate" them into words of the human language, as is done in Fig. 7. This figure, of course, gives only the major positions that may occur in such situations. Transitions between these positions (e.g., between male frontally oriented toward the female and male in broadside position) are possible.

Fig. 6. Blocking the way in Grant's gazelle (Gazella granti). (After 16 mm film at 24 frames/sec.) a. Frame 1: Several females and their offspring approach the boundary of a territory (marked by a termite heap), coming from the center. The territorial male stands at the opposite side of the termite heap, at one side of the females' path. b. Frame 32: The male turns and starts walking to a position in front of the females, c. Frame 72: The male moves into broadside position. The females stop or turn 90°. d. Frame 152: The male is in full broadside position, blocking the females' path. e. Frame 164: The females walk back or move parallel to the boundary. The male is still in broadside position, f. Frame 182: Still in broadside position, the male assumes an erect posture and turns his head in the direction of the females (head flag). Most of them have turned and walked back to the center of the territory.

Fig. 7. Direction signaling by the position of a (territorial) male Grant's gazelle (black) relative to that of a female (white) of his harem—"translated" into human language, a. "Stop! Do not advance in this direction!" b. "Do not turn around!" c. "Stop or go ahead, but do not turn in my direction!" d. "Continue in your direction!" e. "Continue in your direction and speed up!" f. "Turn around and withdraw!" g. "Go ahead!" h. "Follow me!"

This example of ("silent") herding behavior in Grant's gazelle may demonstrate the importance of mere orientation between partners in social communication. Such orientations are also involved in all the displays discussed below. In some of these displays, the role of the orientation component is very significant; in others it appears to be of minor importance, but it is never lacking in any that are aimed and addressed to definite recipients. The reader is asked to keep this in mind since I will describe these displays without coming back to details of the previously discussed orientation components.

Threat Displays

Threat displays indicate readiness for fighting ("I am going to fight you!"). This distinguishes them from dominance displays (p.690). However, there are some transitional cases that may be termed "threat-dominance displays" (Droh-Imponieren in German literature), in which features of both dominance and threat displays are combined. One may distinguish between "symbolic" actions, in which an animal uses the same behavior patterns as in fighting but does not touch its opponent, and more or less ritualized intention movements, where the performance is restricted to the very initial movements of beginning a fighting action. Most ritualized are those threat displays where intention movements are "frozen" into postures. Since there are offensive and defensive fighting techniques, there are corresponding offensive and defensive threat displays, beside others that can be used both ways. In an agonistic encounter, when only one opponent shows an offensive threat and the recipient responds with a defensive threat, the latter is a sign of inferiority (Fig. 8).

Whether threat, dominance, and space-claim displays (p.687) challenge or intimidate the recipients depends on whether they are equal or inferior to the senders and also on the situation. For example, during migration even recipients equal to the sender will often simply "obey" the latter's threat without any counterdisplay. In a very considerable number of encounters the threat remains one-sided (Fig. 9). This means that only one of the animals involved shows a threat display. In a relatively few cases the other animal may immediately attack the sender or simply ignore its threat. Usually, however, a recipient will withdraw, or show submissive behavior, or even flee in a one-sided threat encounter. Since threat displays occur under the same conditions that can lead to overt aggression, they can substitute for and save fighting in such cases. Especially in encounters between peers, the sender's threat releases the addressee's counterdisplay. Also, in some of these reciprocal threat encounters, one opponent may eventually give in and withdraw. However, the probability that reciprocal threat encounters will end in a fight is high in the artiodactyles. The aggressiveness of both opponents obviously is heightened by the reciprocal displays, finally culminating in overt fighting. Thus, fights are not prevented in such cases. However, each opponent has become aware of the other's hostile intentions by the previous displays, and both are prepared to fight. Surprise attack, the most dangerous form of aggression, is effectively avoided by these reciprocal threats.

Fig. 8. Dominant bull (right) of oryx antelope (Oryxgazella beisa) in broadside position, displays with (erected) head-up posture, angling the horns and "pointing" the ear toward a subordinate bull (left), who responds by lowering the head and withdrawing. Lower left: Frontal view of an oryx in (pronounced) head-low posture. Lower right: Oryx antelope in relaxed "normal" posture (for comparison).

Fig. 9. Coordination of group activity by horn threat and pursuit march in adult male Thomson's gazelles. (After 16 mm film at 24 frames/sec.) a. Frame 1 : Within a bachelor group (which was changing from grazing to moving), a male (right) approaches one of the males still grazing (left), b-c. Frames 16 and 24: The challenger (right) threatens with a high presentation of horns. The recipient (left) ceases grazing and starts moving, passing in front of the threatening male at an angle of 90°. d-g. Frames 32, 48, 56, and 72: Without interrupting the high presentation of horns, the challenger stands and lets the addressee pass in front of him. h-k. Frames 80, 88, 96, and 120: Still threatening continuously, the challenger places himself behind the withdrawing recipient and follows him in a pursuit march.

The statements above, especially the thesis on theintimidating and challenging effects of threat displays, may be substantiated by the example of a quantitative analysis on the outcomes of one-sided and reciprocal threat encounters in Thomson's gazelle (Table 2). The data were collected during a two-year study (1965-66) in Serengeti National Park. In the table, the term "horn threats" refers predominantly to medial and high presentation of the horns and to symbolic butting, but it also includes (rarer) cases of symbolic downward and sideward blows and of head-low postures. "Fight" means any form of horn contact. "Other forms of aggression" mainly include air-cushion fights and grazing rituals (see below), but also the rarer cases of (one-sided or reciprocal) aggressions toward inanimate objects and the very exceptional cases of body attacks. "Withdrawal, flight, or submission of one of the opponents" are listed according to the relative frequency in which they occur after a threat. In one-sided threats, it is, of course, always the (nonthreatening) addressee who withdraws or flees or shows submissive behavior. "Ending in other ways" refers to those cases in which the recipients did not show any reaction to the threats and/or the sender(s) eventually ceased threatening, and in which both animals involved continued with clearly nonagonistic activities such as herding females, running plays, or relaxed standing.

Artiodactyla

Table 2

The one-sided threat encounters ended with withdrawal, flight, or submissive behavior of the addressee in 84.0 percent of the 1,680 observed cases, clearly demonstrating the intimidating effect of the threats. On the other hand, the reciprocal threat encounters (both opponents displaying, usually with the same form of threat) led to fights in 70.7 percent of the 738 observed cases. (This proportion is even greater in encounters between completely equal opponents.) These statistics clearly demonstrate the challenging effect of these displays. While the intimidating effect of threat displays has been acknowledged frequently and readily in ethological literature, the challenging effect has rarely been pointed out expressis verbis. However, it definitely exists, it is by no means rare, and it should be distinguished from the intimidating effect since it does not make sense to speak about intimidating when the threats lead to fighting and obviously none of the opponents have been intimidated.

Threats, as well as fighting, may serve a multitude of social functions in artiodactyles. The most important, many of which occur in a single species, are: territorial establishment and ratification, defense against territorial invasion, maintaining or enlarging individual distance (especially in grazing), coordination of group activities (especially when a group changes from one activity to another, e.g., from resting to moving), "voting" to determine marching direction and order, pushing during movement (i.e., keeping the migration going), establishment and maintenance of social hierarchies, herding (male:female), defense against sexual approaches (female:male), and soliciting milk (young:mother) and defense against it (female: young). Also in the mating rituals, threat and dominance displays of the courting male play an important role in many species; while in others, male courtship displays are apparently related to agonistic behavior. In short, there is hardly any realm of the social life of artiodactyles in which aggressive displays are not involved. With little exaggeration, one may say that the entire social organization and communication of these animals is based on aggression.

Among "symbolic" actions (= full performance of the aggressive action without touching the opponent) we may first mention two forms of so-called redirected aggression (Moynihan, 1955): aggression against an inanimate object and aggression against an animal other than the one that released this aggression (= "Radfahrer" Reaktion: Grzimek, 1949). Neither type necessarily consists of addressed threat displays. For example, the object aggression may sometimes be a play with inanimate objects. Moreover, it appears that there is a connection between object aggression and marking behavior and, thus, as in all marking activities, object aggression frequently occurs when no potential addressee is present. Even when an addressee is present, the second animal often does not react to the performer's action but either ignores it or simply watches it. Thus, the effect of object aggression on a recipient is sometimes dubious. On the other hand, there are cases in which object aggression is addressed to a definite recipient and clearly releases reactions in that animal.

Common objects to be attacked are trees, branches, bushes, rocks, grass, the ground, and, in captivity, fences and feeders. Object aggression is often due to rather complex situations. For example, when a territorial male or a "master" of a harem group sees a potential rival at a distance, it would be very much against his "interests" to leave his territory or his females in order to threaten or attack his rival. But since his aggressiveness is stimulated by his adversary's presence, he remains where he is and begins fighting an inanimate object close by. In the same situation, a male may also become aggressive toward a group member, for example, an immature male, whose presence was tolerated or ignored before. These actions are usually rather striking, so that the potential rival releasing them can notice them and, subsequently, will also notice the presence of a potential opponent even at a considerable distance. The rival may then stay away.

Redirected aggression against conspecifics is present in all artiodactyl species. Object aggression, however, has yet to be reported from Hippopotamidae, Suidae, and Tayassuidae. There are only a few reports of object aggression in captive Camelidae and Giraffidae (Pikers, 1954, 1956; Backhaus, 1961). In these groups it apparently has no function in social communication. Cervidae and Bovidae frequently perform object aggression with their antlers and horns. Object aggression may possibly be of greater importance to social communication in cervids than in bovids. In some cervid species such as mule deer, the sound caused by beating trees or bushes with the antlers may release flight in immature or subordinate males during rut (Geist, pers. comm.). I do not know of any corresponding events in bovids. In some bovid species, however, object aggression obviously has become ritualized. Beating and goring the grass and the ground, alternately to the right and left, has led to a rhythmical and persistent "weaving" (Fig. 10a) in certain species such as Grant's gazelle (Walther, 1965). On the whole, it seems that redirected aggression represents an intermediate stage between addressed and unaddressed behavior as well as between threat displays and dynamic visual marking (Hediger, 1954), on the one hand, and advertising presence, position, state, and status, on the other.

Fig. 10. Some "symbolic" threat behaviors, a. Object aggression ("weaving") in Grant's gazelle, b. Dropping down on the carpal joints ("knees") in wildebeest (Connochaetes taurinus). (After photo by R. D. Estes.) c. Rising on the hind legs in ibex (Capra ibex). d. Undirected neck winding and snapping of a female situtunga ( Tragelaphus spekei) in response to the male's sexual approach.

The following behavior patterns are clearly addressed to definite partners in the overwhelming majority of cases. Chasing occurs in its severest form after one of two combatants has been completely defeated in a fight, but it can hardly be considered a means of communication. However, there is also a kind of "symbolic" chasing in certain artiodactyl species. With no previous fighting, one animal will run after another as if the latter had previously been defeated. In a sense, this symbolic chasing anticipates victory. Typically, it is most frequently used by animals of a high social status, such as territorial males, toward hopelessly inferior partners, especially females or nonterritorial, immature males. In some species the males show no intention of actually attacking the addressee during these symbolic chases. They may utter certain sounds, however, such as roaring in impala (Schenkel, 1966) or, in Thomson's gazelle, a very typical "chasing call," a strophic "pshorr-pshorrpshorr," uttered through the nose, which almost always causes the inferior addressee to flee as fast as it can.

Related to the symbolic chase is the feint attack, in which one animal approaches the other in a rush. This action is usually combined with other species-specific intention movements for attacking (e.g., open mouth, lowered horns or antlers), but the animal stops just before touching its opponent (if it has not already fled). Feint attacks occur at least occasionally in practically all artiodactyl species.

When both opponents perform feint attacks and continue with such offensive and corresponding defensive maneuvers as would occur in a true fight but without touching each other, one may speak about an "air-cushion fight," in which it appears as if there is an invisible cushion between the opponents. These air-cushion fights are very frequent in certain artiodactyles, such as gazelles and topi. They can occur as intermezzos between true fighting, but may also substitute completely for an overt fight. Air-cushion fights as well as feint attacks and other symbolic actions can be combined with all the sounds commonly heard during the fighting of certain species: loud roaring in hippo; loud growling in warthog (Frädrich, 1967) and peccary (Schweinsburg and Sowls, 1972); growling, gargling, and roaring in tylopods (Pikers, 1954, 1956); and rather soft growling sounds in certain bovid and cervid species.

Biting as a threat behavior in the Artiodactyla commonly takes the form of symbolic snapping (i.e., snapping in the direction of an opponent without touching it). It is very pronounced in tylopods, especially in camels (Pikers, 1956). In hippo (Verheyen, 1954) and peccary (Frädrich, 1967), the opening of the mouth as a threat display may lead to a yawning-like performance. Squabbling, tooth clicking, and tooth chattering, which in severe threat encounters are intensified to a staccato snapping of the jaws, have been described in peccary (Schweinsburg and Sowls, 1972), and similar phenomena have been observed in wild boar (Sus scrofa: Frädrich, 1967). Symbolic snapping is also quite common in certain cervid species, but apparently not in all of them. I have seen symbolic snapping quite frequently in red deer, axis, and fallow deer, but never in white-tailed deer. Symbolic (as well as actual) snapping is found in only some of the Bovidae species. Here, it is apparently negatively correlated with the presence, development, and use of horns (Walther, 1960a). Symbolic snapping occurs mainly in bovids with small horns, such as the Cephalophinae, and in hornless females of certain species (Fig. lOd), such as Tragelaphus and Kobus. In sitatunga it was also observed in young males as long as they had very small horns or no horns at all. The only bovid species I know of in which females with large horns occasionally show symbolic snapping is the eland antelope (Taurotragus oryx)—interestingly enough, this species is very closely related to Tragelaphus (a genus in which the females have no horns).

The grinding of teeth brought about by exaggerated sideward movements of the lower jaw in several cervid species (Schneider, 1930) is probably also a ritualized form of biting behavior.

Symbolic (and actual) pushing with the mouth shut is apparently closely related to snapping. In the Suidae (Fig. 11a), symbolic pushing is a more or less pronounced, relatively slow upward movement of the head in the direction of an opponent, usually causing him to withdraw (Frädrich, 1967). In the Tylopoda, Cervidae, Giraffidae, and a number of Bovidae species, the movement is a short, but relatively violent, horizontal push forward with the snout and head in the direction of an opponent. As with symbolic snapping, this pushing with mouth shut is especially frequent in female bovids without horns. In the genus Tragelaphus it is the most common defense of the female against an approaching or driving bull at the beginning of the mating ritual (Walther, 1958, 1964b). Occasionally, pushing with the mouth shut (as well as snapping) may be combined with stretching the whole neck forward in the direction of the opponent. This is possibly the origin of the head-and-neckstretched-forward posture, which is a common dominance or courtship display in many artiodactyles (see below).

Fig. 11. a. Symbolic pushing with mouth shut (right) in wild boar (Sus scrofa). (After Frädrich, 1967.) b. Threatening with wide-open mouth in hippo (Hippopotamus amphibius).

In connection with mouth and head movements, we may mention spitting of stomach contents, which is exclusively found in the New World tylopods (Pikers, 1954, 1956). They spit only when an adversary is close enough, and then they aim for the opponent's head. Having received the "full load" in its face, the recipient may show a grimace of "loathing" (Pikers, 1954, 1956).

Species that practice neck fighting (pressing down or lifting up of the opponent with the neck) may show corresponding symbolic neck movements—forward stretching alternating with lowering and steeply erecting and winding of the neck (Fig. lOd). They are especially evident in females of sitatunga (Walther, 1964b) and, to a lesser extent, in greater kudu (Walther, 1958, 1964b). Postures derived from neck fighting (p. 683), however, are obviously commoner than symbolic movements.

In fighting, certain artiodactyl species rise on their hind feet (Fig. 12a). Then the animal either throws its body on its opponent (Suidae and Tylopoda; also female nilgai: Frädrich, 1967; Pikers, 1954, 1956; Walther, 1966a), or beats the opponent with its forelegs (especially the Cervidae: Müller-Using and Schloeth, 1967), or "dives down" into a horn clash (many Caprinae species: Fig. 10c) (Walther, 1960b, 1966a; Geist, 1966a). The symbolic form, occurring at a distance from which the animal cannot reach its adversary, is a rising on the hind legs or a more or less pronounced jump with the anterior part of the body (Drohsprung: Walther, 1960a) in the direction of an adversary (the hind feet remain on the ground). This behavior is shown by the species mentioned above and by others, including Soemmering's gazelle, Grant's gazelle, chamois, and the Tragelaphus species, which have none of the aforementioned fighting techniques in their recent behavior inventories.

Fig. 12. a. An aggressive jump with biting intentions (right) is countered by an extreme nose-up posture (left) in guanaco (Lama guanicoe). (After Pikers, 1954.) b. Rising on the hind legs and (symbolic) beating with the forelegs in red deer (Cerbus elaphus).

Especially in cervids, where striking out with the forelegs (with or without rising on the hind feet) is common in agonistic encounters, symbolic beating or kicking with the forelegs (Fig. 12b) can be used as a threat. Kicking with the hind feet is, on the whole, rare in the fights of the Artiodactyla (it plays a certain role in the fights of the tylopods), and the corresponding symbolic performance is even rarer.

Some species (especially wildebeest, hartebeest, topi, nilgai, oryx, addax, and roan and sable antelope) tend to drop down on the "knees" (carpal joints) of their forelegs during fighting. This can also occur symbolically when the animal is still some distance from its rival (Fig. 10b). In this case, it is often, although not necessarily, combined with goring the ground (object fighting), or it may pass into grazing.

In Cervidae and Bovidae (i.e., species in which at least the males have antlers or horns), symbolic butting (Fig. 14a), i.e., a pronounced nodding movement of the head in the direction of the addressee, is widespread. It is more frequent in females and juveniles (animals with no horns or smaller horns) than in adult males. Head throwing, i.e., nodding head movements (like the exaggerated affirmation of humans), apparently is a rhythmically repeated form of the butt. Head shaking (like that of humans in negation) can be considered the symbolic form of twisting the head and horns (Fig. 14b) left and right (horns interlocked in fighting animals). All these threat movements occur in rather similar forms in almost all bovid and cervid species; however, there are considerable differences as to the frequency of the movement and how pronounced it is. For example, in gazelles head shaking is rare and head throwing even more so, but these movements are frequent in wildebeest, hartebeest, and topi.

Other behavior patterns that are used as fighting techniques but may also occasionally occur as symbolic threat movements are the downward and sideward blow (Fig. 14c, d). In the symbolic downward blow, the animal brings its head and horns or antlers down from an upright position in a violent movement. The forehead may touch the ground, and the horns (or antlers) then point forward (toward the opponent). Occasionally, the head and horns may be kept at the lowest point of this movement for several seconds (low presentation of horns). In the sideward blow, the head and horns are rapidly moved sideward from either an upright or a lowered position. These two movements occur with some frequency only in relatively few species— for example, downward blows in blackbuck (Walther, 1959; Schaller, 1967) and sideward blows in oryx (Oryx gazella). The horn sweep is a combination of the symbolic downward and sideward blow. The whole performance resembles weaving, but is shorter and more violent. I have seen horn sweeps frequently only in adult male Grant's gazelle, where they occurred predominantly in very severe threat encounters between peers.

The threat displays discussed above are movements. Other threat displays are postures in which an animal—in contrast to symbolic actions—does not show the entire aggressive action but just the first intention of it, although often in an exaggerated form. Moreover, these intention movements are "frozen" for several seconds, in extreme cases for one or even several minutes. Vocalizations are not commonly heard in connection with these postures, but when there are any, they are usually rather soft growling sounds. When animals have skin glands that can be opened at will—the preorbital glands of (males of) certain bovid and cervid species are especially important in this regard—they are widely opened. Apparently, there are few threat postures in Nonruminantia—one could think of the opening of the mouth in hippo (Fig. lib) or of a nose-up posture with turning of the cheek toward the opponent in peccary, which Schweinsburg and Sowls (1972) interpret as an intention movement for biting with the side of the mouth. In Tylopoda and Ruminantia, however, threat postures are well pronounced, frequently used, and obviously the most important means of threat.

In the head-low posture (= Kopf-tief-Halten or Kopf-tief-Drohen: Walther, 1958, 1964b, 1966a), the neck and head are stretched downward and forward (Fig. 14h). The nose is close to the ground in an attitude similar to that of grazing, and sometimes the animal may switch to grazing. Apparently, meaning and origin vary with the species, and the head-low posture is of a somewhat ambivalent nature even within the same species. For example, male guanaco (Lama guanicoe) may approach a rival with a head-low posture (Pilters, 1956), which appears to be a rather offensive threat in this species. It has possibly evolved from biting an adversary's forelegs (a common fighting technique in tylopods) and/or from a special form of neck fighting (getting under the opponent's body). In moose (Geist, 1963) the head-low posture combined with raising the hair on the neck, withers, and rump and holding the ears down (inside toward the opponent), and, sometimes, with a very loud roar, appears to be a defensive threat, frequently used in intra- and interspecific encounters. Also, in many bovid species, the head-low posture is apparently a defensive threat. The horns, which are directed backward and upward and more or less parallel to the neck, are in an ideal position to parry an opponent's butt or downward blow. At least in some bovids, such as Gazella and Oryx, the head-low posture may turn into a submissive posture (Fig. 14i) that is similar or almost identical with it. On the whole, here, the head-low posture often expresses some kind of inferiority, especially when used as a response to an offensive threat display of a challenger.

The head-and-neck-stretched-forward posture (Fig. 15a) is common in courting artiodactyl males (p.702), but is much rarer in agonistic encounters. It occurs sporadically over a range of some rather different species. It also has different origins and meanings. For example, in white-tailed deer, this posture has been termed the "hard look" (Thomas, Robinson, and Marburger, 1965). However, the more we learn about this behavior in white-tailed deer, the more it appears to be a display of inferior animals toward superior opponents. (Geist, pers. comm.), possibly related to the head-low posture of other artiodactyl species. Thus, it may not be such a "hard look" as it was initially assumed to be. In species that practice neck fighting, such as giraffe (Backhaus, 1961) and nilgai (Walther, 1958), the head-and-neck-stretched-forward posture can be considered an intention movement for a stroke with the neck (in giraffes) or for neck fighting since these animals can and do place the neck over the opponent's from this posture and press the rival down (Fig. 20a). Here, this posture is a display between peers, but it is also frequently used by superior animals toward inferiors.

Closely related to neck fighting are the head- and-neck-forward/upward posture and the noseup posture. The latter posture (Fig. 13b) is a mirror image (upward) of the head-low posture. In a pronounced nose-up posture, the neck and head are stretched upward stiffly, the nose pointing skyward (Fig. 12a). In llama, the ears lie back, and both postures show a relation to pushing with mouth shut (Pilters, 1954, 1956). Pilters even considers pushing with mouth shut to be the origin of at least the head-and-neck-forward/upward posture. The major point, however, seems to be that the head-and-neck-upward/forward posture is the perfect initial posture and the swing-out movement (Ausholbewegung) for placing the neck over an opponent's neck or body. This becomes clear from descriptions and pictures of behavior of tylopods (Pilters, 1954, 1956) and giraffe (Backhaus, 1961), and, evidently, is true of okapi (Okapia johnstoni: Walther, 1962). Thus, in these animals the head-and-neck-upward/forward posture is an offensive threat, usually used by a superior opponent or in encounters between peers.

In tylopods and in (hornless) nilgai females (Walther, 1960a, 1966a), the nose-up posture was observed as a fighting technique. Here it is used to parry an aggressor's jumping attack (throwing the anterior part of the body on the opponent) and immediately push it back with the chest and long side of the neck, in a special kind of neck fighting. Thus, it is a basically defensive maneuver, but one that allows the defender to counterattack immediately. These features also determine the character of the corresponding symbolic performance (threat display), which Pikers (1954, 1956) very adequately interprets as the expression of strong resistance in Lama. Much the same is true in okapi (Walther, 1960c, 1962); whereas in giraffe (Backhaus, 1961), the nose-up posture (also shown in lateral position to a rival) may possibly be more offensive in nature relative to the particular fighting technique of this species (sideward strokes of the neck and head against an adversary's neck, shoulder, body, or hindquarters: Backhaus, 1961).

Fig. 13. a. Erect posture as intention movement for rising on the hind legs in guanaco. (After Pikers, 1956.) b. Head-and-neck-forward/upward posture (right) and turning away (left) in erect posture in llama (Lama guanicoëglama). (After Pikers, 1956.)

Phenotypically, the erect posture (Fig. 13a) is intermediate between head-and-neck-forward/ upward and the nose-up posture. In the erect posture, the animal stretches its neck straight upward as in the nose-up posture, but the head and nose point forward or forward/upward, in similar or even identical fashion to the head-and-neck-forward/upward posture. Pikers (1954, 1956) interprets it in llama as the "utmost readiness for defence." However, I have frequently seen it preceding the jump attack in llama, and, thus, I consider it more offensive in nature. Of course, "utmost readiness for defence" can be said to be the point where defense verges on offensive action. Being an intermediate stage between an (offensive) head-and-neck-forward/upward posture and a (defensive) nose-up posture, it is very possible that the erect posture is used both offensively and defensively. In addition, very slight changes in this posture (neck somewhat more forward or backward) may shift its meaning more to the offensive or the defensive side, respectively. In Marco Polo sheep (Ovis ammon poli: Walther, 1960b), as well as in a number of cervid species (Müller-Using and Schloeth, 1967; Geist, 1966b), the erect posture precedes an (aggressive) rising on the hind feet, and I am inclined to interpret the erect posture generally and primarily as an intention movement for rising on the hind legs (Walther, 1960a). This derivation, however, excludes neither a neck-fight component nor a connection to pushing with mouth shut nor beating with the forelegs in certain species, since all these behavior patterns may occur in combination with rising on the hind legs. The similarity to the head-and-neck-forward/upward and the nose-up posture has occasionally led to mistaking one for the other or, in the earlier literature, to not clearly distinguishing between them. (I readily admit being among the "sinners.") Despite the doubtlessly close relationship and transitional stages between these postures, one should, as far as possible, try to distinguish them, in the interest of a better analysis.

In species with horns or antlers, i.e., all the Bovidae and Cervidae, presentation of horns (Hörnerpräsentieren: Walther, 1958) and antlers toward an opponent is a frequent and very important form of threat. It can occur in four forms: low, medial, and high presentation and sideward-angling of horns or antlers. In a pronounced low presentation, the horns are held parallel to and on the ground, with the tips pointing toward the opponent. As a posture (i.e., not as a momentary phase in a movement such as a downward blow or horn sweep), this pronounced form is rare. However, when it occurs, it is a very severe threat. In a very common but less-pronounced form of the low presentation, the head is not held so low, but is still clearly below body level, and the chin is tucked in toward the throat so that the horns point forward/ upward. This posture is sometimes similar to the head-low posture, and the two may change into one another. Interestingly enough, in pronghorn, a posture frequently seen in herding males corresponds to this less-pronounced, low presentation of bovids and cervids with respect to the height at which the head is held. Otherwise, however, it is a head-low posture, i.e., the chin is not tucked in toward the throat, apparently because of the position of the horns relative to the skull axis. In pronghorn, the horns "lean" forward more than those of bovids. Thus, the horn tips point forward when a male pronghorn holds his head downward/forward, whereas a bovid or cervid (in which the horns or antlers stick straight upward or, in some species, even "tilt" backward) has to tuck in his chin for the same effect.

In the medial presentation (Fig. 14e), the neck is held forward at body level, the chin is tucked in toward the throat so that the nose points approximately vertically to the ground, and the horns or antlers subsequently point upward or somewhat forward. This kind of presentation is the commonest and is used by almost all bovid and cervid species.

Fig. 14. Some common threat displays in Bovidae and Cervidae, demonstrated by the examples of Thomson's gazelle (a-f.) and Grant's gazelle (g-i.). a. (Symbolic) butting. When repeated and exaggerated, this results in head throwing, b. (Symbolic) head twisting. When repeated (to the right and the left) this results in head shaking, c. (Symbolic) sideward blow, d. (Symbolic) downward blow. The end phase corresponds to a low presentation of horns, e. Medial presentation of horns, f. High presentation of horns, g. Sideward angling of horns, h. Head-low posture (subadult male), i. Submissive posture (adolescent male).

Imagine that the addressee is standing in front of the threatening animal (in b. and c. the reader is in the position of the adversary), except in g., where the recipient is to the side of the displaying animal (in the position of the animal in h.).

In the high presentation of horns (Fig. 14f) the neck is held erect so that the head and horns are carried distinctly above body level and the chin is tucked strongly in toward the throat. The horns, which tower above the head, point upward or upward/forward. The high presentation of horns or antlers occasionally occurs in many bovid and cervid species. As a frequent and pronounced display, however, it is found in relatively few, the most important of which are the gazelles (Walther, 1958, 1964a, 1968a), but it is also described in oryx (Walther, 1958) and in sable (Hippotragus niger: Huth, 1970) and roan antelope (Hippotragus equinus: Joubert, 1970). In the Gazella species it is a very common and rather serious, offensive threat, used predominantly by adult and, in particular, territorial males. It is an open question whether the high presentation of horns is to be considered a pure threat (swingout movement for a downward blow) or a combination of a threat (horn presentation) and a dominance display (erect posture, p.697).

Sideward angling of horns (Fig. 14g) in broadside position as well as the broadside position itself are threat displays in those artiodactyl species (p.673) that fight from a lateral (parallel or reverse-parallel) position. However, the combination of broadside position and sideward angling of horns also appears in a number of species that do not do so, but in which the rivals fight in frontal orientation to each other. In these cases, obviously, a dominance display (broadside position, p.693) and a threat display (angling of horns toward an opponent) are combined. Such a sideward angling of horns or antlers is occasionally found in almost all bovid and cervid species. However, I have only observed it to be a frequent and elaborate display (always combined with an erect or head-up posture) in a few species, including Grant's gazelle (Walther, 1965), oryx (Walther, 1958, 1966a), and Waterhuck (Kobus defassa). Joubert (1970) recorded it in roan antelope, and Estes (1969) in wildebeest. In medial and high presentation, as well as in sideward angling of horns, one ear is often emphatically turned ("pointing") in the addressee's direction. In presentation of horns, it is usually directed forward; in sideward angling of horns it points sideward toward the opponent.

To the human observer, all these forms of presentation of horns and antlers appear to refer strongly to the presence and potential use of these weapons. Observations by Hediger (1946) of certain cervids and by Geist (1971) of bighorn sheep (Ovis canadensis) support the view that this is also true for a conspecific recipient. In other species, however, there is evidence to the contrary. In a captive, dehorned oryx bull, the postures of high presentation and sideward angling of horns had strongly intimidating effects on conspecifics (Walther, 1958), and I have had occasion to observe precisely the same thing in a dehorned male dorcas gazelle. Thus, the role of horns in these displays should be considered with caution. At least in certain species, the recipients may react more, and perhaps even primarily, to the corresponding postures of the head and neck and pay little attention to the challenger's horns in threat encounters.

With the exception of sideward angling of horns or antlers, the sender is usually frontally oriented toward an addressee in all the threats described above. Sometimes the animal will not turn its whole body but only its head and neck in the opponent's direction. The addressee also often stands frontally to the sender. In principle, however, the addressee's position does not matter. In other words, its flank or hindquarters can also face the sender. This last orientation is regularly found in pursuit marches, where the inferior animal walks away in a "normal" or a submissive attitude while the superior one follows with a threat or a dominance display. Occasionally, all the threats described above can be combined with broadside displays (p.693). For example, the sender may block the opponent's path in lateral position and may show a high presentation of horns (without turning its head and horns in the opponent's direction). Such cases must be distinguished from the combination of threats and inferior behavior patterns, e.g., when the inferior animal stands frontally toward its challenger and responds to the latter's threat with a counter display, at the same time turning its head away from the challenger and avoiding a clearly directed threat; or when a withdrawing opponent continues to threaten forward while the pursuing dominant marches behind it. We must emphasize that these cases are mosaiclike combinations of threat displays with tendencies for withdrawal or turning away. The latter are expressed by the orientation component. However, the threat displays themselves cannot be explained as a result of an inner conflict between aggression and escape tendencies as certain authors (Tinbergen, 1952; Hinde, 1966, Ewer, 1968) have assumed. They are expressions of pure aggression (Walther, 1974).

Displays of Space-Claim

Besides threat and dominance displays (p. 690), some additional behavior patterns, which are certainly not forms of aggression, may show up in agonistic encounters of artiodactyles. On the other hand, they do fit meaningfully into an agonistic situation. They are connected to behavior patterns used in relation to space, the environment, inanimate objects, or the ground. Thus, they are not restricted to hostile encounters but they may also occur (even primarily) in other situations. When used in hostile encounters, the animal may claim occupancy of a place —literally, in the case of a territorial animal; symbolically and only very temporarily, in the case of a nonterritorial animal. The intimidating or challenging effects on recipients are incomparably less pronounced than in threat and dominance displays. One can imagine that these behavior patterns may have a somewhat self-insuring effect on the performer and may perhaps signal to the opponent: "This is my place (you'd better stay away)." This view is supported by the absence of these displays in situations where other threat and dominance displays may be used but where a space-claim would not agree with the requirements of the situation (as in courtship, mother-infant relationships, coordinating group activities). Such behavior patterns are predominantly object aggression, marking with skin glands, urination and/or defecation, pawing the ground, and possibly also grazing in agonistic encounters.

Object aggression is so closely related to aggressive behavior that it was discussed in connection with threat displays. However, its occurrence in nonagonistic situations, its relative ineffectiveness as a threat in some species, the possibility that in a few species the secretion of certain glands may be deposited in connection with this behavior, and the possibility that it may leave visually recognizable marks on vegetation or the ground make probable a connection with marking behavior.

Gland marking is quite common in connection with agonistic encounters (when such glands are present). The way in which certain species, such as chamois (Hediger, 1949) or mountain goat (Geist, 1965), deposit the secretion of their postcornual glands (rubbing the object with the forehead or the region between the horns) makes even an original connection with object aggression rather likely. However, in many species that mark with their preorbital glands, deposition of secretion shows no resemblances to aggressive behavior. On the other hand, the frequency of preorbital gland marking often increases strikingly just before, after, and sometimes during agonistic encounters. Furthermore, the males of some species, such as Thomson's gazelle, goitered gazelle, and blackbuck, open their preorbital glands wide during fights and threat and dominance displays (and during courtship). Finally, these and other species, such as pronghorn, blesbok, and axis deer, frequently rub their horns or antlers more or less intensively on the same object before or after marking it with their glands. Thus, it does not appear too farfetched to assume a connection with agonistic behavior in these animals as well. In the case of territorial males, this marking in agonistic encounters may serve to make the position of the boundary more recognizable to an opponent and to the owner himself. In a hostile encounter, this means the owner can, if necessary, retreat to the marked area, as this is his territory, into which no opponent will easily follow. If an animal is not territorial, it may create a fixed starting point for the combat by marking —comparable to the "corner" in a boxing ring.

Practically the same aspects are true for urination and/or defecation in connection with agonistic encounters. In species in which urine and/or feces have a role in marking territory, these behavior patterns occur more frequently in encounters of territorial than nonterritorial individuals. On the other hand, urination and/or defecation in agonistic encounters is not restricted to territorial individuals or species, and, since it is not dependent on the presence of certain glands, it is more widespread than gland marking in Artiodactyla. However, it is not used as frequently by the same individual in the same encounter, probably because the "material" is not as easily at hand as gland secretions.

In some species, pawing the ground with a foreleg in agonistic encounters shows a relation to pawing before urination and/or defecation. In species that do this, pawing is neither obligatory (an animal may also urinate and/or defecate without pawing) nor restricted to hostile encounters. However, it is apparently somewhat more frequent in agonistic situations than in connection with "normal" urination/defecation. There is an interesting trend in this behavior, which may be demonstrated by its occurrence in Thomson's gazelle, Grant's gazelle, and oryx antelope. In agonistic encounters of territorial male Thomson's gazelle, pawing the ground precedes urination/defecation. Occasionally, a male may paw the ground without urination/defecation following. In all the observed cases, however, there was strong circumstantial evidence that the buck had urination/defecation "in mind" but did not actually perform it because of the pressure of the situation. The same is true for Grant's gazelle. In some relatively rare cases, however, this explanation is unlikely; for example, in a very severe fight between adult Grant's males, one or both combatants may paw the ground while their horns are interlocked. In these and a few other cases, pawing the ground does not seem to be linked to urination/defecation, but appears to be a display of its own. In oryx antelope, pawing the ground during threat encounters without defecation following (urination is not combined with defecation in the oryx) is even rather frequent. Thus, there is obviously an increasing separation (or, depending on the direction in which this trend is considered to go, an increasing combination) of pawing the ground and urination and/or defecation. With this separation, pawing the ground is more likely to become a special means of threat.

A last behavior possibly belonging to the category of space-claims is grazing in agonistic encounters, in which a mechanism similar to or identical with that of a transitional action (Lind, 1959) seems to be involved. The head of an ungulate is close to the ground in a number of aggressive behavior patterns such as the headlow posture, low presentation of horns, downward blow, and actual fighting. Thus, the head is generally in a position proper for grazing, and the animal may then sometimes switch to grazing, which in certain situations and/or species appears to be a behavior of an inferior animal; in others there are no indications of inferiority. Unfortunately, its relation and significance to agonistic interactions has not been realized until recently, apparently because observers always assumed that grazing would not belong to an agonistic interaction and that the encounter was over when one or both combatants started grazing. So, at present, although this behavior is certainly widespread in artiodactyles, only the very thorough studies of Estes (1969) on agonistic grazing in wildebeest, David's (1973) very similar results in bontebok (Damaliscus dorcas dorcas), and my own observations in Thomson's and Grant's gazelle are available for a somewhat detailed discussion.

In Grant's gazelle, grazing in connection with agonistic encounters occurs relatively often on the part of inferior opponents. They frequently respond by grazing to a superior challenger's approach and/or threat and dominance displays before they withdraw. Occasionally, however, a superior combatant or both opponents may also graze. When both rivals graze at the end of an encounter, they may move away from each other while doing so. Thus, the meaning of agonistic grazing appears to be somewhat ambivalent in Grant's gazelle. On the whole, however, one may say that it frequently, although not necessarily, indicates inferiority in this species.

The situation is clearly different in Thomson's gazelle. Here, agonistic grazing is primarily a behavior of adult males, especially of territorial males, and both opponents generally perform it simultaneously. What makes me use the term "grazing ritual" in the case of this species (Walther, 1968a) is a very pronounced and predictable change of position between rivals (territorial neighbors). Grazing almost uninterruptedly, they go through a frontal position, a parallel or reverse-parallel position, and a reverse position (hindquarters to hindquarters). These three changes in position are the minimum. Sometimes the opponents may graze side by side along the entire boundary of a territory, or one or both may turn again to a frontal position after having grazed for a while in parallel or reverse-parallel position. In short, there can be many variations and repetitions in such a grazing ritual, and it may last from a few minutes up to half an hour or longer. In Thomson's gazelle, a grazing ritual rarely precedes a fight. Incomplete attempts are frequent during pauses in fights. Especially when grazing in frontal position, the rivals can easily change from grazing back to fighting. The complete, pronounced ritual, however, is seen only following or in place of a fight.

When grazing in frontal position after a clash, male Thomson's gazelle immediately enlarge the distance between one another by stepping backward. In this species, as in many artiodactyl species, individual distance (Hediger, 1954) depends on age, sex, and activity. It reaches its largest extent in adult males during grazing. Thus, by simultaneous frontal grazing immediately after a horn clash, the opponents signal to one another for a return to the (large) grazing distance. If necessary, the grazing posture allows them to begin fighting again immediately. The parallel or reverse-parallel position probably has the same origin and basic meaning as other broadside displays—blocking the path—as it is rather obvious in grazing along the territorial boundary. Finally, in the hindquarters-to-hindquarters position, agonistic grazing may allow an exit without "losing face," the battle, or territorial status for both combatants. Thus, agonistic grazing is closely related to fighting in Thomson's gazelle, is most commonly shown by territorial peers (i.e., nothing speaks for a behavior of inferiority), and appears to belong to that category of expressive behavior that is connected with space relations (taking place right on the boundary or, at least, in the boundary zone of neighboring territories).

In wildebeest (Estes, 1969), one territorial male may approach another and may intrude into his territory while permanently grazing. (This was never observed in Grant's or Thomson's gazelle.) When they approach each other, the owner of the territory often threatens the intruder by sideward angling of his horns. The intruder keeps his present grazing attitude, i.e., behaves like a subordinate male Grant's gazelle. On the other hand, the owner of the territory may also start grazing, and both animals may go through changes in position similar to those of territorial "Tommy" males (except that this occurs inside the territory belonging to one of them and not at the territorial boundary). They may even circle around each other in grazing, something gazelle males do not do. Furthermore, in wildebeest, grazing apparently does not turn as easily and as readily into fighting (or vice versa) as in Thomson's gazelle. Thus, agonistic grazing of wildebeest offers an even more complicated picture than that of the two gazelle species; however, it also seems to be restricted to territorial males in wildebeest (Estes, 1969) and in bontebok (David, 1973). Thus, it could have something to do with space relations in these two species as well.

Possibly, the key to a general interpretation may be seen in the close relationship of agonistic grazing to horn fighting, as it was found in Thomson's gazelle. Biting and snapping as threat and fighting behavior are found in a number of artiodactyl species as discussed above. One may assume that snapping generally was a fighting technique of the ancestors of recent artiodactyles. It even (still?) occurs in hornless or antlerless females of certain bovid and cervid species. Interestingly enough, symbolic snapping (directed toward the ground) was observed in combination with boxing the opponent with the forehead in greater kudu and sitatunga females (Walther, 1964b). Thus, one could imagine that in some of the recent bovid species also (which primarily fight with their horns), the biting could become activated in agonistic encounters and (since the head is lowered according to the head-low posture or the low or medial presentation of horns in agonistic encounters) that it may turn into grazing (= biting the grass). At first sight, this may appear to be a rather bold speculation, but it would not only explain the occurrence of grazing in agonistic encounters but also make understandable the numerous species-specific variations of agonistic grazing: It is likely that such a common ancestral behavior has become differentiated in a variety of ways in the single species during phylogentic evolution. It is also understandable that such a relatively mild form of aggressiveness can sometimes occur in a dominant animal, sometimes in an inferior one, depending on the situation. Of course, more data are required to substantiate these assumptions.

Dominance Displays

In contrast to threat displays, dominance displays (= Imponieren in German literature; display threat: Lent, 1965; bravado display: Geist, 1966b; present threat: Geist, 1971) do not indicate immediate readiness to fight. Commonly an animal demonstrates its height and/or breadth or shows other striking postures or movements, none of which are related to recent fighting techniques of the species. Sometimes, weapons may also be presented, but not in a position suitable for fighting. Thus, these displays indicate the claim of superiority over the addressee ("I am the boss!") without showing fighting intentions. On the other hand, dominance displays have the same effects—intimidation or challenge—on the recipient as threat displays, and they occur in the same situations. Typically, however, they are infrequent in females and juveniles (animals that often play a subordinate role in artiodactyl societies) and in agonistic situations predominantly or exclusively found in these age or sex classes (soliciting milk and defense against it, defense against sexual approaches, playful encounters). In encounters between unequal opponents, these displays are typical of superior, stronger, older, or higher-ranking (also territorial) animals, but are rarely used by inferiors, whereas the latter may quite frequently show (defensive) threat displays. When the encounter cannot be settled by dominance displays (a situation especially frequent in reciprocal encounters, as when both rivals show the same displays), the opponents may change from dominance displays to threat displays and in some cases from these to fighting. The rivals rarely move immediately from reciprocal dominance displays to fighting, whereas that direct change is very common in reciprocal threat displays. Thus, in such cases, the reciprocal dominance displays contribute to the prolongation of the opening phase before a fight. When the display remains one-sided, the (inferior) recipient usually withdraws upon receipt of a dominance display by the superior sender.

These statements may be substantiated by some quantitative records on certain reciprocal threat and dominance displays in Grant's gazelle, a species that has threat displays (the following presentation refers predominantly to medial and high presentation of horns); at least two dominance displays (head-sideward inclination in frontal approach or in parallel walk, and the head flag in erect posture and in broadside position—for detailed discussion of these displays see below); and an intermediate form, a "threat-dominance display," in which features of dominance displays (broadside display and circling) are combined with a threat component (sideward angling of horns toward the laterally standing opponent).

Table 3 shows which behavior followed immediately after these displays. Cessation of aggression by both opponents (including cases where both are grazing) is relatively rare after reciprocal displays. It is rarest after threats (1.1%). Withdrawal or (the even rarer) submissive behavior of one of the opponents is relatively infrequent (in reciprocal encounters) after the head-sideward inclination (10.8%). Here, however, one has to take into account the fact that the approach with head-sideward inclination is especially frequent in the opening phase of an agonistic encounter in this species, and, thus, the encounter is likely to be continued after this display. With this exception, encounters are most frequently settled and decided (i.e., one of the opponents gives in) after the head flags, i.e., dominance displays (28.8%), and rarest after horn threats (10.1%). Sideward angling of horns as an intermediate behavior is right in the middle (19.5%). As mentioned above, encounters are frequently continued with other displays—such as head flag, sideward angling or presentation of horns, horn sweep, and object aggression—after an approach or parallel walk with head-sideward inclination (75.4%). The same is true, although to a lesser extent, for (reciprocal) head flags (60.8%) and sideward angling of the horns (53.7%). It is considerably rarer after horn threats (11.2%). The most conclusive figures are provided by the frequency of fights immediately following the displays under discussion: In the two dominance displays none occurred after head-sideward inclination and fights followed only 2.6% of the reciprocal head flags. However, fights followed "threat-dominance displays" (sideward angling of horns) in 24.2% of the cases, but they followed reciprocal pure threat displays (horn threats) in frontal position in 77.5% of the cases. These figures clearly demonstrate in Grant's gazelle that threat displays are significantly closer than dominance displays to overt fighting, and the same appears to be true in other Artiodactyla.

Table 3

It is certainly noteworthy that an animal exposes itself to a possible attack ("daring") by the rival much more during dominance displays than during threat displays. In this context, there is a direct connection between the dominance displays and advertising and emphasizing presence. On the other hand, the commonest and most important dominance displays of artiodactyles are not without connection to threats and fighting behavior. Sometimes features of threat and dominance displays occur in combination (Droh-Imponieren = threat-dominance displays), as mentioned above. Moreover, the alternative "threat or dominance display" does not depend on the posture or movement per se, but on its relation to the recent fighting behavior of a given species. It follows from this that the same posture can be a threat display in one species and a dominance display in another. For instance, in species like mountain goat, which fight from a broadside position, assuming a lateral position is an intention movement for fighting and, thus, a threat display. However, when a broadside display appears in the hostile encounters of a species like Grant's gazelle, which fights only from a frontal position, the broadside position is a dominance display.

In Artiodactyla, the relationship between threat and dominance displays and the evolution of the latter are rather obvious. Paleontologists (e.g., Thenius and Hofer, 1960) generally agree that the recent artiodactyles have evolved from forms that had not developed special organs for fighting. There are some recent species, such as the Tylopoda, that still lack special armament and use their legs, teeth, and neck in fighting. Others, like the Suidae and hippos, have specialized already existing organs (the teeth) for use as weapons. Still others, mainly Cervidae and Bovidae, have developed special organs (antlers and horns) for intraspecific aggression. However, these species have also evolved from hornless/ antlerless ancestors (Archeomeryx optatus or related prehistorical ungulates).

It is certainly correct to presume that these "unarmed" ancestors had their fights and threat displays, but they could not use horns or antlers since they had none. It is not out of place to assume that their fighting techniques and threat displays were similar, if not identical, to those shown by recent "unarmed" artiodactyles (e.g., tylopods). With the development of horns and antlers as special means of intraspecific aggression and with the corresponding development of phylogenetically "new" fighting methods and threat displays, ancestral fighting techniques have been replaced in the bovids and cervids. Of course, this has varied among the species according to differences in advancement (size, mass, shape, permanent or temporary usefulness as weapons) of horns and antlers (Walther, 1960a, 1966a). In some species, apparently, although the ancestral fighting techniques have more or less completely disappeared, the corresponding threat displays have remained. Hence, even though these ritualized intention movements originated in fighting techniques, they now have no connection with the recent fighting behavior of these species and are now dominance displays in the defined sense. Because of their separation from recent fighting techniques, they are milder forms of challenge or intimidation compared to the "new" forms of threat, which refer directly to actual fighting behavior. When, in a species, both "modern" threats and displays derived from ancestral aggressive behavior coexist, a more subtle gradation in the forms of challenge and intimidation becomes possible, which is certainly advantageous (allows more shades) to social communication and makes the "survival" of these phylogentically old displays understandable.

We will come back to this hypothesis in the discussion on courtship displays. At the moment, it is sufficient that the reader is aware of the possible relationships between threat and dominance displays, and can understand why certain postures or movements are considered threat displays in some species and dominance displays in others, and why dominance displays generally are more frequently found in the more highly advanced artiodactyles than in the more primitive forms.

The broadside position was mentioned previously as a means of blocking another's path and as an initial position for fighting in lateral (parallel or reverse-parallel) position—two aspects that are not mutually exclusive. Apparently both gave rise to the broadside attitude as a dominance display. One can speak of a special broadside display only when additional behavioral features are added to the basic broadside orientation, making it more striking to conspecifics (Fig. 15c). This can be achieved in many different ways: In warthog, bushbuck (Tragelaphus scriptus), and nyala (Tragelaphus angasi), the animal ruffles up the manes on its neck and back. In gaur (Bibos gaurus), cattle (Bos primigenius), and greater kudu, the displaying animal may hump its back by extending its hind feet further under its belly than usual. And/or it may stretch its head and neck forward (e.g., nilgai) or forward/downward (e.g., greater kudu and gaur), or in some species (e.g., blackbuck, Grant's gazelle, and lesser kudu) the broadside position may be combined with an erect posture or even with a nose-up posture.

An exaggeration of the broadside position by a sideward inclination of the head (see below) away from the opponent is very frequently seen, e.g., in nilgai, mountain gazelle, Grant's gazelle, red deer, and elk. This movement is often related to sideward angling the horns, e.g., in wildebeest, oryx, and roan antelope.

When both rivals show a broadside display, they are automatically brought into a parallel or (more often) a reverse-parallel position (Fig. 17). In many species, they then begin to circle, always keeping their flanks toward each other. As the encounter becomes more severe, they may change from circling in reverse-parallel position to frontal orientation and may eventually attack one apother frontally (provided they do not belong to a species that fights from a lateral position anyhow). In some species, such as Grant's gazelle, circling can also be seen in one-sided encounters. This means the displaying animal moves around the (nondisplaying) recipient. Apparently, this "encircling" has a strongly intimidating effect.

Fig. 15. Some of the dominance (or threat-dominance) displays that are used as fighting techniques or threat displays in other, behaviorally more primitive, artiodactyl species, a. Neck stretch of a female caribou directed frontally toward the addressee, corresponds to a bite or a (very intensive) push with mouth shut and possibly to neck fighting. (After Berry, in Pruitt, 1960.) b. Manchurian sika (Cervus nippon dybowskii), with nose-up posture and grinding of teeth, directed frontally toward the addressee, corresponds to hooking with (elongated) upper canines and/or to neck fighting. (After Schneider, 1930.) c. Broadside display (i.e., the reader is in the position of the addressee) of a greater kudu bull (Tragelaphus strepsiceros), combined with head-low posture, humped back, and bent hind legs positioned under the belly. This corresponds to fighting in lateral position (jostling with the rump or shoulders, sideward blow with the head and neck) and possibly also to a swing-out movement to an aggressive jump (throwing the body on the opponent while simultaneously turning toward the adversary).

Some Suidae species, such as wild boar, fight in a lateral position (Frädrich, 1965). With respect to a derivation of broadside displays from ancestral fighting behavior, it is noteworthy that pressing and jostling with the shoulder or the whole side of the body against the opponent's flank plays a large role in these fights. In this species, however, assuming the broadside position is a threat rather than a dominance display. On the other hand, one could definitely think of the broadside position as a dominance display in other Suidae, such as bush pig (Potamochoerus porcus) and warthog, which commonly fight in frontal position but which show the broadside attitude as an expressive behavior in agonistic encounters (Frädrich, 1965, 1967). The broadside position is also present in at least some tylopods, such as vicuna (Pikers, 1954, 1956), and in the Giraffidae. However, in these species it is more a threat than a dominance display because of its close relationship to the fighting behavior of these animals, as previously discussed. A few Bovidae fight in parallel or reverse-parallel position (Fig. 17a), but since the majority of bovids and cervids do not when broadside displays are found in a considerable number of bovids and cervids (as listed above), they are dominance displays in the outlined sense (Fig. 17c).

In the head-sideward inclination (Fig. 16b), the head is turned sideward at an angle of about 45°, and in the head-sideward turn (Fig. 16a) at an angle of about 90° or even slightly more. Apparently, these two behavior patterns have not been distinguished from each other in previous literature. However, it appears that one is considerably more frequent in a given species than the other; for example, the head-sideward inclination is frequent and the head-sideward turn rare in Grant's gazelle and mountain gazelle, while the opposite is true in Uganda kob and impala. This fact and the possibility that there may be some differences in messages and meanings make a distinction desirable, although the two behavior patterns are closely related and transitional cases are possible. Both can be combined with other species-specific dominance displays, such as the erect posture in mountain gazelle, Grant's gazelle, and topi, the head-and-neck-stretched-forward posture in nilgai (Fig. 17b), and, occasionally, with threat displays such as the high presentation of horns. Obviously, the head-sideward inclination has a tendency to turn into the sideward angling of horns. This tendency is less pronounced in the head-sideward turn.

Fig. 16. Several forms of dominance displays, a. (Erected) head-up posture (right) and sideward turn of the head (left) in an encounter between two territorial topi bulls (Damaliscus lunatus topi), b. Sideward inclination of the head (left) in a dominant male Grant's gazelle during a pursuit march behind the inferior opponent (right) in submissive posture, c.-e.: Sequence of behavior patterns in the head-flag display of Grant's gazelle. (The position of the addressee is marked by X at the right of the picture.) c. The buck assumes an erect posture in broadside position, d. Head-sideward inclination as a swing-out movement away from the recipient, e. Head flag (hohes KopfZuwenden) toward the recipient, after which the animal turns its head back to the initial position (c).

Generally, when an animal makes a turn, the head precedes the body and the movement or the posture of the head alone is often sufficient for and of special importance to social communication. In this respect, the sideward inclination and sideward turn of the head can be considered as intention movements for turning to a broadside position (when the animal stands with its body frontally oriented toward its rival before turning) or as an emphasis or exaggeration of the latter position (when the animal is already oriented with its flank toward the addressee). In combination with the broadside position, the head is always turned away from the opponent. As it became clear from many observations of different species, however, the animal always keeps its opponent in sight, watching the rival carefully from the corner of its eye, and immediately reacts to the slightest movement of its opponent (one-eyed fixation and lurking watch, respectively).

Fig. 17. Some examples of reciprocal broadside positions in agonistic encounters, a. Fighting mountain goat ( Oreamnos americanus). Note the ruffling of the hair on the back, especially on the withers and the croup, and the swing-out movement of the lowered head for a sideward blow against the opponent's flank. (After Geist, 1966a.) b. Nilgai bulls (Boselaphus tragocamelus) circling each other. The neck stretch can be considered an intention movement for the neck fight, which actually occurs in this species. Thus, the neck stretch in combination with the broadside position is probably a threat-dominance display in nilgai, c. Reciprocal head flag in two (standing) bucks of Grant's gazelle. Since neither broadside position and erect posture, nor the head flag, itself, belong to the (recent) fighting techniques of Grant's gazelle, this is an example of a pure dominance display.

As with the broadside position, certain authors seem inclined to consider every form of sideward turning of the head as indicative of involvement of escape tendencies and thus consider that such displays are brought about by inner conflict between aggression and flight. Of course, these head movements can also initiate turning around for withdrawal, and sometimes this is definitely the case in artiodactyles. However, this is only one possibility. It is as possible —and this happens fairly frequently—that the animal turns into a full broadside display from a head-sideward inclination or a head-sideward turn in, at first, frontal orientation. One may even see a very effective swing-out movement for a reinforced turn toward the rival in the initial turn of the head away from him. Furthermore, as with all dominance displays, both behavior patterns are primarily displayed in encounters between high-ranking or territorial peers, and, in encounters between unequal opponents, are displayed by the clearly superior one. Last, but not least, in certain species, such as Grant's gazelle and mountain gazelle, a sideward inclination of the head is rather persistently shown by the pursuer (!) when following a withdrawing inferior. This truly does not speak in favor of the involvement of escape tendencies in this display. On the whole, I am inclined to consider the head-sideward inclination as well as the head-sideward turn abbreviations of the broadside display. The head-sideward inclination is possibly a somewhat more offensive form and more closely related to threat behavior in the strict sense than the sideward turn of the head.

One common principle in dominance displays is to make the animal appear as long and broad as possible. This is verified by broadside displays and related behavior patterns. Another widespread principle is to make an animal appear as tall as possible. This is mainly achieved by the erect posture (Fig. 16a, c), the nose-up posture, and the head-and-neck-forward/up ward posture. Because these three are closely related and their differences were pointed out in the discussion on threat displays, I will lump them here and refer to them as "erect displays." Erect displays can be combined with broadside displays or either may occur independently.

Since most artiodactyl species rise up on their hind feet in sexual mounting, one could consider the erect postures to be intention movements for the latter. However, one can also think of a development parallel to that of the broadside displays, inferring that these erect postures are phylogenetic relics of ancestral fighting behaviors. A point that speaks in favor of a connection with sexual behavior is the fact that a combatant sometimes mounts its rival from an erect posture in an agonistic encounter. On the other hand, Tragelaphus species show erect postures as dominance displays, but do not mount in erect postures. The male puts his head and neck on the female's back during sexual mounting (Walther, 1958, 1964b). Hence, their erect displays are obviously not closely related to sexual mounting. Or, in certain highly advanced cervid species, such as red deer, Manchurian sika (Sika nippon dybowskii: Schneider, 1930), and barasingha (Cervus duvauceli: Schaller, 1967), an erect posture with almost vertically raised nose (Fig. 15b) and gnashing teeth apparently corresponds to the attitude adopted by related primitive species (Moschus, Hydropotes, Muntiac) when threatening or attacking an opponent with their prolonged upper canines (Antonius, 1939).

Finally, it is possible that sexual mounting— of the artiodactyl type—itself is related to and has originated from aggressive jumping and throwing the body on the partner. In this connection, it may be mentioned that all Tylopoda copulate lying down. Thus, their copulatory posture has nothing to do with rising on the hind legs. A male tylopod, however, frequently jumps at and throws his body on the female to force her to the ground—a behavior that is similar to or even identical with that shown in fights between rivals. It is this aggressive jumping at the female, not the copulatory posture of tylopods, that corresponds to the mounting behavior of other artiodactyles. It is possible that tylopods, which appear to be behaviorally primitive artiodactyles in more than one regard, may have kept this general feature of ancestral artiodactyl behavior. In short, it is impossible at present to decide conclusively whether the erect dominance displays of certain artiodactyles have evolved from sexual mounting or from ancestral fighting techniques. However, the odds appear to be in favor of the latter.

Since a connection between erect displays and recent fighting techniques is very likely in Tylopoda (jumping at the other, neck fighting), Giraffidae (pronounced neck fighting in several forms), and Cervidae (rising on the hind legs and beating with the forelegs), erect postures as true dominance displays occur mainly in certain Bovidae. A possible exception appears in the genus Ovis, where erect postures are apparently related to rising on the hind feet as a fighting technique (Walther, 1960b), bringing them more on the side of threat displays. In discussing certain other bovids, some difficulties arise from the major role of the neck in erect postures. In a species with a relatively long, movable neck, its erection is an essential component in these displays and makes their recognition very easy. In a species with a relatively short, massive neck, however, it cannot be erected to any great extent. It is then doubtful whether one can speak about an erect posture and, if so, whether it can be directly compared to the displays of long-necked species.

In the wild oxen, for example, the anatomical structure of the neck largely inhibits the performance of pronounced erect displays. Raising the head in these animals usually results in a noseforward posture, and it is doubtful whether it can be directly compared to erect postures of other species. Erect postures as clear dominance displays, often combined with a more or less pronounced lifting of the nose, have been observed in species such as impala (Schenkel, 1966), greater kudu, lesser kudu, topi, blackbuck, and Grant's gazelle. I do not hesitate to add to this list the "head-up" displays of species with less movable necks, such as wildebeest (Talbot and Talbot, 1963; Estes, 1969), oryx antelope (Walther, 1958), and tsessebe (Damaliscus lunatics lunatics: Joubert, 1972).

In all species under discussion, the erect displays occur in frontal orientation toward an addressee as well as in combination with the broadside position. Tsessebe bulls displaying broadside in erect posture with lifted nose in front of an opponent sometimes jump up into the air with their forelegs (Joubert, 1972). This example is of special interest, since one might doubt that erect postures are intention movements for rising to the hind feet, arguing that this view makes no sense when the erect posture is displayed in a lateral orientation toward the recipient. Regardless of whether it is logical in human terms, however, jumping up with the forefeet from an erect posture while in broadside position is a fact.

Some additional behavior patterns can be considered dominance displays (for example, the protruding of a "goulla bag" from the mouth of camels), but I will discuss only one more in detail—the unique modification of the erect posture in Grant's gazelle. Standing in broadside position with vertically erected neck, the displaying animal turns its lifted head and nose sideward toward the addressee (hohes Kopf-Zuwenden: Walther, 1965) and then forward again. The tail is often horizontally stretched during this display and may swing to the right or the left. This head flag (Figs. 16c-e, 17) is commonly initiated by a head-sideward inclination or sometimes even a sideward turn of the head away from the rival as a swing-out movement. The tension of neck muscles (Estes, 1967) and/or the white throat patch may have an additional effect in this display. However, I think that the term "neckintimidation display," suggested by Estes (1967), conceals the fact that the head turn toward the rival is its most important component. The head flag is always delivered while standing and may be repeated several times. The releasing situation apparently comes when the sender is at the point of passing the receiver or, vice versa, when the receiver is passing or has just passed the sender. The sender can stand broadside in front of or parallel to the recipient or, most frequently, in reverse-parallel position (Fig. 16c). In reciprocal encounters, the reverseparallel position of the opponents is almost obligatory. It provides a good chance for one of the opponents to cease displaying after a while and walk forward in a normal, common way. Thus, the display provides a "golden bridge" for an inferior to retreat "without losing face." One may speculate that this "golden bridge" is a principle in all highly ritualized dominance displays (Walther, 1965).

Interestingly enough, a perfect analogy occurs in human behavior when we signal—in a rather rude and arrogant way—to a (usually inferior) person to go away by a sideward swing of the head. In Grant's gazelle, the head flag can be understood as a(phylogenetically) "new" orientation ("pointing") movement from an erect posture, adapted and specialized to a situation in which the recipient is standing in reverse-parallel position to the sender. Thus, the sender anticipates and demands the recipient's withdrawal in the direction opposite to its own. Because of the highly stereotyped nature of the head flag, "failures" occasionally occur, i.e., when the sender's position relative to the recipient is inadequate for the orientation of the head flag (comparable to "failures" in orientation of the inciting behavior of female mallards: Lorenz, 1963). Possibly, such a "failure" gave rise to the nonobligatory but relatively frequent "undirected" head flag of female Grant's gazelle during the mating ritual (when she is standing before the male and facing in the same direction as he), which apparently provokes the male's mounting (Walther, 1965).

In barren ground caribou (Rangifer tarandus arcticus) a head-and-neck-stretched-forward posture (Fig. 15a) in agonistic encounters (Pruitt, 1960) could be considered a dominance display (provided that reindeer do not bite, an action that I have not seen mentioned in the literature). It fits the picture that male reindeer show the same posture in courtship (since dominance displays can be used in the courtship of certain species). On the other hand, some of the situations that are described for this behavior (females' warding off strange fawns, females' warding off other adults from their fawns) are atypical for dominance displays and agree better with threat behaviors. Possibly, this is a case of a threat behavior "on its way" to becoming a dominance display. Also, in the Ovis species, the head-and-neck-stretched-forward posture and the kick with the foreleg can occur in agonistic encounters (Walther, 1960b; Geist, 1968, 1971). Geist (1971), who did a very intensive study on these behaviors, considers them primarily courtship displays in Ovis. Therefore, they will be discussed more in detail in this context. It may be mentioned, however, that certain species rotate the head about its long axis in this posture so that one cheek almost points toward the ground. In bighorn sheep, this behavior was termed the "twist" by Geist (1971), and he considers it a horn display (the horns expand laterally in sheep rams). However, the fact that practically the same behavior can also be seen in the hornless females of greater kudu (Walther, 1964b) contradicts this interpretation. At least, it cannot be generally applied to all artiodactyl species that show this behavior. It may also be that this "twist" (as a phylogenetic relic) is more closely related to biting behavior since in many mammalian species turning the head about its long axis is frequently combined with biting.

Since dominance displays and certain other agonistic behavior patterns (such as offensive threats, symbolic chases, pursuit marches, and mounting) are frequently shown by superior animals, and since similar (even identical in some species) behavior patterns occur in males of certain species in sexual encounters, the question may be raised whether these behavior patterns are basically sexual in nature and whether a superior male may treat an inferior as if it were an (estrous) female. In a study on the behavior of bighorn sheep, Geist (1971) has strongly argued in favor of this view. In studies on the behavior of oryx (Walther, 1958), several Tragelaphus (Walther, 1964b), and several Antilopinae species (Walther, 1968a), however, I came to an almost opposite conclusion. I fully agree with Geist that in a number of artiodactyl species an adult male treats inferior conspecifics more or less alike, regardless of sex (as far as they are of interest to him—in many species, adult males do not pay attention to the young, and in some species they are not interested in females either if they are not in estrus). Since females are smaller and lighter and usually have smaller horns (in some species, no horns at all), they are naturally inferior partners; thus, the adult males treat them as inferiors in sexual encounters, just as they treat younger and weaker males as inferiors in agonistic encounters. Although the addressee frequently accepts the inferior role when challenged by a superior partner, there are cases in which he (sometimes also she) does not do so, but reacts with (defensive) counterdisplays or even fights back. This behavior is easily understood as a reaction by a recipient to a basically agonistic behavior of a challenger (i.e., treating the addressee—also the estrous female in sexual encounters—as an inferior opponent). What is very difficult to understand is that an addresseemale or female—would react by submission or defense to a basically sexual behavior, i.e., being treated like a female by a superior companion (it is unlikely that these animals share the "male bias" of certain humans).

Courtship Displays

Like "alarm signals," the term "courtship displays" refers to only one function, but one that is particularly important. These displays are used almost exclusively by males and addressed almost exclusively to females. Obviously, there are great differences within the Artiodactyla. In the tylopods, for example, the behavior of males toward females strongly resembles male behavior in encounters with other males and may even be more or less identical with it. In species such as blackbuck, males use the same (or almost the same) dominance displays toward females in courtship as they use in agonistic encounters with (male) rivals, but they do not use the fighting techniques and the threats. In still other species, for example, Thomson's gazelle, the behavior inventory of a male encountering other males is (almost) entirely different from that occurring in encounters with females. I speak about "courtship displays" here to distinguish them from threat and dominance displays, which either occur in encounters between partners of the same sex or are used toward partners of both sexes.

Interestingly enough, there are situations in which these courtship displays also work like threat or dominance displays. For example, when Thomson's gazelle are migrating (in large, mixed herds, and usually in file), if the animal in front of an adult male stops or slows down it is often "pushed" from behind by the adult male in an attempt to speed it up and keep the migration going. If the animal in front is another male, the "pusher" will threaten him with high or medial presentation of horns. Since in Thomson's gazelle these displays are not normally used by a male toward a female, if the animal in front is a female, the male will not threaten her but will show "courtship" displays (head-and-neckstretched-forward posture, nose-up; see below). A "Tommy" male uses these behavior patterns only very rarely toward another male. In this situation (and in a few others), the courtship displays affect females as threats do males.

It follows that (1) the so-called courtship displays can serve functions other than mating (they can be especially important in herding, soliciting urine, coordinating group activities, and soliciting milk); and (2) although courtship displays can be phenotypically different from (recent) threat and dominance displays in certain species, there are unmistakable connections between these courtship displays and agonistic behavior. One may say that the most elaborate courtship displays in artiodactyles are special kinds of dominance displays that probably originated in ancestral fighting behavior. As discussed in the section on dominance displays, the hypothesis states that, especially in the bovids, ancestral fighting techniques were largely replaced by the "modern" horn fight during evolution, whereas expressive displays (intention movements) related to ancestral fighting techniques have remained. Such displays owe their aggressive nature to their origin, but they are milder forms of aggression than threat displays, which refer to the recent fighting techniques of these species. Expressive behavior patterns that mildly challenge or intimidate the partner—-just enough to diminish somewhat its aggressiveness (intimidation) or avoidance tendencies (challenge) released by the other's close approach, but not enough to release serious aggression or flight— are tailor-made for mating rituals of the artiodactyl type, in which the male has to approach the female and establish some kind of dominance over her. Male dominance is apparently a prerequisite for successful mating in these animals. If, for some reason (e.g., the male is too young) the female turns out to be superior, that usually means the end of the mating activity. In the more highly-evolved artiodactyl species, male dominance is achieved by the male's courtship displays and in some species by dominance displays as well; in more primitive artiodactyles, true threats or fighting may serve the same purpose.

The reactions of the females are in full agreement with the basically aggressive nature of the male courtship displays. Most commonly, the females respond by withdrawal; in some species or in certain situations also by flight at a gallop, submissive behavior, defensive threats, sometimes even fighting, and, in rare cases, offensive threats or dominance displays. All these are clearly responses to aggressive displays on part of the males. It is noteworthy that there are no genuine female courtship displays or sexual behavior patterns in the artiodactyles, except for some directly related to copulation, such as lifting the tail, standing for the male's mount, and leaning into the male's mount. What artiodactyl females show during courtship rituals are either male behavior patterns (e.g., the female's mounting the male, which, by the way, is not very frequent in nondomesticated artiodactyles), or agonistic displays, or behavior of inferiority as discussed above.

A few examples may substantiate these statements. In the tylopods, the males use the same displays toward females in courtship as they show in agonistic encounters with male rivals (Pikers, 1954, 1956). Also neck fighting, jumping at the female and throwing the body on her to force her down to the ground, and biting her forelegs are more or less obligatory. Sometimes a tylopod female may defend herself by the same behavior patterns; then, the mating ritual, at best, differs from a true fight only in intensity. In wild boar, the male may put his snout under the female's belly and lift her up (Frädrich, 1956), a behavior that is also known as a fighting technique in this species. Driving giraffe males push the female's shoulder, flank, and hindquarters with their "horns" (Backhaus, 1961). Even in certain bovid species, for example, mountain gazelle, the driving male uses his horns in a ritualized manner. Chasing the female at a trot or gallop occasionally occurs in the majority of artiodactyl species. In some, such as roe deer (Capreolus capreolus), it is obligatory, and in others, such as pronghorn, it is very frequent. More commonly, the male drives the female ahead at a walk (Fig. 19), as he would an inferior male opponent after a hostile encounter. This mating march, i.e., the ritualized withdrawal of the female from the pursuing male (which can often turn into her true withdrawal or even flight, especially when the female has not yet reached the peak of the heat), is a basic component of the mating rituals in many artiodactyl species. Obviously, the partners synchronize their readiness for mating in this way.

In oryx, addax, sable and roan antelope, the mating march has been largely replaced by a more stationary performance, the mating whirlaround (Paarungskreisen: Walther, 1958), in which the male and female step around each other in reverse-parallel orientation (Fig. 18b) (a behavior that also occurs in agonistic encounters between male opponents in these species, see Fig. 8)—the female showing a head-low posture all the time (like an inferior male opponent in an agonistic encounter). Especially in oxen, the mating ritual has become even more stationary. Here, male and female stand in reverse-parallel position throughout most of the ritual ("guarding" =Hüten: Schloeth, 1961). Generally, broadside displays are frequently used in courtship rituals of Artiodactyla—be it in reverse-parallel position (e.g., in guarding or in the mating whirlaround), or with the male blocking the female's path by standing in broadside position in front of her (observed, for example, in bison, bush pig, and greater kudu) or walking in lateral escort with the female.

Fig. 18. Some of the agonistic elements in the courtship rituals of certain Bovidae. a. Erect posture of the male (right) and intention movement for pushing with mouth shut of the female (left) in an early phase of the mating ritual of lesser kudu ( Tragelaphns imberbis). b. Mating whirl-around in oryx antelope. Note the head-up posture of the male and the headlow posture of the female (compare Fig. 8).

Also, in those species that show special courtship displays in the defined sense, there are hardly any that do not occur as a threat or dominance display in other artiodactyl species. The difference is that in certain species, these displays are almost exclusively used by males toward females. As is true of dominance displays, this is most pronounced in the Bovidae.

Fig. 19. Frequent male courtship displays in Bovidae, demonstrated by the example of Thomson's gazelle. a. Neck stretch, b. Head-and-neck-forward/ upward posture, c. Nose-up movement with foreleg kick.

The head-and-neck-stretched-forward posture (Fig. 19a, 20c, 21a, b) or—as an abbreviation of this precise, but very long term—the neck stretch (=Kopf-Hals-Vorstrecken: Walther, 1964a, 1968a; Überstrecken: Walther, 1958; low stretch; Geist, 1971 ) is one of the commonest attitudes of courting artiodactyl males. In wild boar and warthog, the male tries to keep his nose in contact with the female's genitals. This is accompanied by rhythmical sounds resembling the starting of an outboard motor (Frädrich, 1967) and by rhythmical tongue flipping from the slightly opened mouth. Finally, the male warthog may put his chin on the female's hindquarters (precopulatory posture: Simpson, 1964). In Cervidae, the males approach the females from behind with the neck-stretch posture and tongue flicking in order to touch the females' genitals (naso-genital testing: Müller-Using and Schloeth, 1967; Geist, 1963, 1966b).

In the Bovidae, the neck-stretch is also very common in courtship, but with certain speciesspecific differences. Probably in all Bovidae species, the male may occasionally show such a posture when touching, licking, or sniffing a female's vulva. In oryx, roan, sable antelope, and Grant's gazelle, this movement is only used to reach the female's genitals. It is neither an obligatory component of the courtship ritual nor a special display. In sheep and goats, nilgai, hartebeest, impala, and mountain goat, the neckstretch is definitely a display, frequently combined with tongue flicking or with various species-specific additions: in nilgai with vertical erection of the tail; in ibex with flapping the tail over the back (Fig. 21a); in sheep ("low stretch": Geist, 1971) with uttering roaring sounds and rotating the head around its long axis (Fig. 21b); in Capra and Ovis species and in dorcas gazelle with foreleg kicks (see below).

In Tragelaphus species, the neck stretch ("Überstrecken"'. Walther, 1958), combined with very soft sounds (like "imm—imm—imm") and occasional tongue flicking, is an intention movement for neck fighting (Fig. 20). This is very clear in greater kudu, where a driving male accompanies the female in lateral escort, or may frontally approach and place his neck over her neck from the neck-stretch posture (Walther, 1964b). This ritualized neck fight during courtship is less frequent and pronounced in other Tragelaphus species; however, the males frequently rub the sides of their stretched necks with winding movements on the females' hindquarters, and also show a pronounced neck stretch during the lateral escort. This connects this posture to neck fighting, and it makes a connection to naso-genital testing unlikely. The relationship between neck stretch and neck fighting is not impossible in other bovids, of course, although it is not as clear as in Tragelaphus and Taurotragus (Walther, 1960a, 1966a). In Gazella species, tongue flicking during the neck stretch is rare. In Thomson's gazelle (Fig. 19a), the neck stretch in combination with opening the preorbital glands and uttering soft "bl—bl—bl" sounds frequently transforms into or alternates with a head-and-neck-forward/upward posture (Walther, 1964a). Of course, this sequence (Fig. 19) also does not agree with a derivation of the neck stretch from naso-genital testing. A somewhat similar change between lowering the head and bobbing it up is described in mule deer (Geist, 1966b).

Fig. 20. Ritualization of the neck fight (Halskampf) as a courtship display, a. Neck fight in an agonistic encounter between a subadult (left) and an adult (right) bull of nilgai antelope, b. Neck fight (strongly diminished in intensity) between male and female (here, in frontal position) in an early phase of the mating ritual of greater kudu. c. Neck stretch as an intention movement for neck fighting in a greater kudu bull (here, in lateral escort to the female).

Two other male courtship postures found in a number of artiodactyl species are the erect posture (Fig. 18a) and the head-and-neck-forward/ upward posture (Fig. 19b). Both can be temporarily exaggerated by a nose-up movement in which the nose points almost vertically upward. In principle, all three are displays in their own right, and in certain species, such as okapi (Walther, 1960c), they are easily distinguishable. On the other hand, there are transitions that are sometimes difficult to distinguish clearly. For example, Antilopinae males frequently display during walking. The neck then leans somewhat forward, making the performance rather similar to a head-and-neck-forward/upward posture, and it is hard to say whether it is now the latter posture or a modification of the erect posture. Perhaps an even more difficult distinction to make is the one between the head-and-neck-forward/upward posture and the nose up; often the latter appears to be nothing but an exaggerated form of the former (Fig. 19c).

The erect posture may occasionally be seen in mating rituals of many artiodactyl species since, being an intention movement for rising onto the hindlegs, it is related to mounting. As a truly elaborate display (one that not only precedes mounting but also is retained during large parts of the mating ritual) it is not so widespread.

Erect posture, head-and-neck-forward/upward postures, and/or nose-up movements are typical in courting males of guanaco (combined with laying back the ears and grunting: Pilters, 1954, 1956), giraffe (Backhaus, 1961), okapi (Walther, 1960c), and Antilopinae species (Walther, 1968a). In erect posture, the males of chamois utter strange grunting sounds (the "Blädern" in the terminology of German hunters). The erect posture also appears in certain cervids, such as red deer; however, here, strictly speaking, it probably belongs more to herding than to courtship.

Antilopinae are especially interesting in this respect. For example, in blackbuck, an erect display—apparently closer to the head-and-neck-forward/upward posture than to the erect posture in the strict sense—(with ear dropped and tail flapped over the back) is more or less the same in courtship and in agonistic encounters between male rivals. In Grant's gazelle, the erect posture occurs in courtship and in agonistic encounters; however, there are some differences (Walther, 1965, 1968a). For example, in courtship, the erect posture frequently transforms into a head-and-neck-forward/upward posture, and both are displayed almost continuously (i.e., easily for quarter of an hour to one hour and longer); whereas in agonistic encounters, the erect posture does not transform into a head-and-neck-forward/upward posture, and it usually precedes or follows head flagging (i.e., it is held only for seconds). In Thomson's gazelle, it is difficult to speak about an erect posture in the strict sense; however, the head-and-neck-forward/upward display is very common (it is very briefly held in this species and is more a movement than a posture) and is restricted to encounters with females. Thus, there is no similar display in the agonistic encounters between males.

Frequently the nose-up movement is combined with foreleg kicks (see below) in the Antilopinae (Fig. 19c). In some other bovid species, the erect displays are combined with exagger ated walking. The most pronounced case of this kind is perhaps the prancing with high lifting of angled forelegs in topi (Walther, 1968b).

Head-sideward inclinations and/or head-sideward turns (frequently in combination with erect postures) of the driving male when standing behind a female may occasionally occur in many artiodactyl species. In some such as eland, they are frequently and even regularly seen during courtship. I am aware of the sideward turn of the head as a truly striking and pronounced courtship display only in pronghorn (Fig. 22b), where males show this behavior when approaching a female and when following her. Possibly this behavior is linked to the black patch on the cheek and/or the presence of a cheek patch gland in this species.

Another typical male courtship display is kicking(Fig. 21) with the forelegs (Laufschlag or Laufeinschlag: Walther, 1958), which is found in okapi and a large number of Bovidae species (Walther, 1960c). With respect to a possible origin of courtship displays in ancestral fighting techniques in combination with and in correlation to the latter's replacement by "modern" fighting techniques, it is certainly of interest that the foreleg kick as a courtship display is not found in those groups of recent artiodactyles that often fight with their forelegs, i.e., the Tylopoda and, above all, the Cervidae. In some cervids, such as fallow deer, white-tailed deer, and axis deer, kicking with the forelegs occurs (apparently without function) in the young during suckling—not for soliciting milk when the young is following behind its mother (which happens in certain bovids) but while the fawn is actually nursing. It is presentlyunknown whe this behavior of young cervids might be related to the foreleg kick in the courtship of certain bovids, and, if so, how to interpret this relationship.

Fig. 21. Several forms of the foreleg kick (Laufschlag) and its combination with someother displays. The postures in combination with the foreleg kick are typical for these species (except for the mounting intention in oryx, which is not necessarily combined with the foreleg kick); however, the kick with the foreleg can also be performed in "normal" posture in all of the species. The positions of the females relative to the males and the distances between the sex partners shown in these pictures do not imply species-specific differences.

a. (Knock-kneed) foreleg kick with neck stretch, tongue flipping, and tail flapped over the back in ibex. The (resting) female responds with symbolic butting in this case. b. Foreleg kick with neck stretch and "twist" of the head in Punjab urial (Ovis orientalis cycloceros). The male is in lateral escort to the female in this case. c. Foreleg kick (between the female's hind-legs) with head-up (erect) posture and mounting intentions (bending the hind legs, characteristic tail posture) in oryx antelope.

Bovidae groups in which the kick with the foreleg is not found are the bovines, Tragelaphus species, Tauro tragus, Boselaphus, Connochaetes, Aepyceros, and Ammotragus. It is also lacking (or only infrequent and weakly pronounced) in Alcelaphus, Rupicapra, and some of the Damaliscus species. However, it frequently occurs in the majority of Caprinae, Antilopinae, Neotraginae, Hippotraginae, Reduncinae, Cephalophinae, and in Oreamnos and Damaliscus lunatus topi. It is most frequently seen when the male is standing or walking behind the female, but it also occurs occasionally when he is face to face with her or is oriented toward her flank. Especially in oryx, addax, and roan and sable antelope, the males often perform the kick with the foreleg during the mating whirl-around in reverse-parallel position. Sometimes the male touches the female's hind legs and, occasionally, even her belly, with his foreleg. More often the female is not touched —the male moves his hind leg between her hind legs (Laufeinschlag — kick-in-between) in certain species, or he does not raise his foreleg high enough, or he performs the foreleg kick at a distance from which he cannot possibly touch her.

Sloppy "knock-kneed" performances (Fig. 21a) occasionally happen in any species. In scimitar-horned oryx (Oryx gazella tao), they are even the rule; otherwise the foreleg is rather stiffly stretched in a "good" kick. The most pronounced performances (raising one foreleg to approximately 90°) are seen when the male is standing. However, he also can deliver the kick with the foreleg while walking, but the leg is then raised only about 45° (Fig. 19c). In Thomson's gazelle, the male can kick alternately with the right and the left foreleg during walking ("drumroll": Walther, 1964a). In Grant's gazelle and blackbuck, the foreleg kick is reduced to a big, stiff-legged step (but the full foreleg kick was recently observed in juvenile males of blackbuck: Benz, 1973). In Soemmering's gazelle only an occasional tripping of the courting male resembles the foreleg kick of other Antilopinae species.

In all bovid species under discussion, the foreleg kick can be delivered when the male is standing or walking in a "normal" or an erected attitude. In the Hippotragus and Kobus species, this is the rule. In some species, such as topi and oryx antelope (Fig. 21c), it can be combined with mounting intentions (bending the hind legs). In Caprinae (Fig. 21a, b) and in dorcas gazelle, it is frequently combined with the neck stretch. In many Antilopinae species, there is a strong tendency to combine it with a nose-up movement. In particular, the foreleg kick is used to make the female continue after she has stopped walking during the mating march or the mating whirlaround. Linked to this, it can also be used as a "last inquiry" before mounting: when the female does not react (by walking ahead) to the foreleg kick, she is ready to accept and tolerate the male's mounting.

As mentioned above, the kick with the foreleg is often combined with the neck stretch in sheep. The two behavior patterns are also frequently used in this combination as well as independently between males in sheep (Geist, 1971). This is the only presently known case in which behavior patterns primarily or exclusively serving as courtship displays in most bovid species play an important role in agonistic encounters. Here, they obviously serve the same function as dominance displays since they typically are shown by dominant males in encounters with inferior ones.

Visual courtship displays appear to be important in artiodactyles, but other types of stimuli play significant roles too. In Suidae, acoustical displays are perhaps as important or even more important than the visual ones; playing back the tape-recorded mating grunts of a male can release "immobilized" standing of a female in heat (Signoret and du Mesnil, 1960). The tylopods are also rather noisy during courtship. In most Ruminantia, however, either courtship is silent or the calls of the sex partners are soft and apparently occur only in combination with and in addition to pronounced visual displays. Scents are probably emitted by glands opened during courtship (preorbital glands, possibly also subauricular glands, and interdigital glands in connection with the foreleg kick), again in addition to the visual displays. Tactile stimulation of the female by licking and touching her, especially her genitals and hindquarters, with the tongue, mouth, chin, or neck is widespread. For example, in Suidae it apparently plays a major role; and in Cervidae, at least, the licking of the female is very pronounced. In Bovidae, there is a whole range —from species where courting males frequently and intensively touch females (for example, Tragelaphus with its neck fight and related behavior) to species that hardly touch the female at all during the mating ritual (for example, Grant's and Thomson's gazelle), except, of course, in mounting, but even this is restricted more or less to contact of the genitals in Antilopinae and Neotraginae. In short, the importance of tactile stimulation during courtship varies widely in artiodactyl species, whereas visual displays are found in almost all of them.

In view of the relevance of visual courtship displays, it is surprising that the list of the most important ones is so short: neck stretch, head-and-neck-forward/upward posture, erect posture, nose-up movement, head-sideward turn, and foreleg kick. Obviously, the species-specific character of courtship rituals in artiodactyles is not demonstrated by a multitude of different displays but by differences in frequency of single displays (ranging down to the absence of certain displays in certain species), by differences in the elaboration, and by combinations within these relatively few displays. The degree of specialization achieved in this simple way is astonishing. Of course, the species-specific differences become more pronounced when one takes into account additional features, such as ear and tail movements and tongue flicking. However, it is possible to characterize the courtship behavior of many species simply by using the postures and movements listed above. For example, in eight Antilopinae species investigated (Table 4), there are no two with courtship behavior that is completely alike, although the basic components are only neck stretch, erect posture, head-and-neck-forward/upward posture (and/or nose up), and foreleg kick. In Table 4, horn threat and sideward turn of the head are added for completion, although the courtship behavior of a single species can be distinguished without them.

Furthermore, as can be seen from Table 4, two or three (sometimes even four or five) of these displays usually occur in the courtship of a single species. This is valid for many but not all artiodactyles, for example, courting bontebok males show only one of them, the neck stretch (David, 1973).

Besides these male displays that apparently (mildly) intimidate (or sometimes also challenge) the female, appeasing behavior plays a certain role in courtship of some artiodactyl species. Appeasement is mainly achieved by licking the partner's head, neck, or shoulders (Fig. 22a), and the males are at least as active, if not more active, in this regard as the females. Licking the female's genitals or her croup and touching the female's croup with the chin by the male may contribute to the female's sexual arousal. The majority of displays in courtship of artiodactyles, however, are either clearly aggressive behavior or have possibly/probably originated from such behavior. In any case, the small proportion of genuine sexual displays in the mating rituals of these animals is surprising. On the whole, here, the relationship of sexual drive to courtship displays is analogous to that between the French nation and the Foreign Legion: The legionaries served and fought for France, but most of them were not French. Correspondingly, the courtship displays occur in the service of sexual drive, but most of them are not sexual behavior.

a = posture or movement similar to medial presentation of horns or butting, b = neck stretch, c = neck stretch + foreleg kick, d = foreleg kick in "normal" or moderately erect posture, e = nose up + foreleg kick (including drumroll and big step), f = nose-up movement and/or head-and-neck-forward/upward posture, g = erect posture (in the strict sense), h = sideward turn of the head, ++ = pronounced and frequent, + = pronounced and moderately frequent, (+) = aberrant performance or situation, ? = not well pronounced and/or questionable whether a special display or accidental, blank = not observed, but possibly exceptionally and/or accidentally occurring, - = not observed, probably lacking, 1) = gentle touching of the female with the horns, 2) = head flag (belongs more to herding than to courtship in the strict sense, occasionally also during courtship), 3) = inconspicuous tripping steps, 4) = occasionally when marking the female with preorbital glands, 5) = threatening and beating the female with the horns in herding (not during courtship in the strict sense).

Submissive and Appeasement Behavior

Most submissive displays in artiodactyles are in every way the antithesis of dominance and offensive threat displays. Sometimes, there are connections with defensive threats. Submissive displays indicate the acceptance of an inferior role. In a sense, they anticipate defeat and lack features that could possibly challenge an opponent and release its aggression. The effect on a (superior) recipient may range from a diminution to complete cessation of aggression. Submissive displays and appeasement behavior enable an inferior animal to remain with a group and/or in a familiar terrain despite the presence of superior and aggressive conspecifics. This is very important for females in the courtship rituals; and for juvenile animals, it may often be essential to their survival.

According to Frädrich (1967), submissive gestures are unknown or dubious in Suidae and Hippopotamidae. They appear in the other groups of artiodactyles; however, their intensity, frequency, and importance vary considerably with the species. The commonest forms that can occur singly or in combination are lowering the head (Fig. 23a), turning 180°, and lying down (Fig. 23b). Following one after the other, they may express increasing degrees of submission.

Fig. 22. a. Licking to appease the female (left) by the courting male (right) in okapi (Okapia johnstoni). b. Head-sideward turn (alternately to the right and the left) in the courting male pronghorn.

In agonistic encounters, turning 180° (with the hindquarters toward an opponent) is usually an intention movement for withdrawal or flight. One can consider it a special display only when it is combined with other features of submission and/or when the animal does not withdraw or flee but remains in the superior opponent's vicinity. Since females are oriented with their hindquarters toward the males in sexual encounters, one might also think that an inferior male might mimic a female through this behavior in an agonistic encounter. However, the same submissive behavior can also be seen in encounters between females. It is unlikely for a (subordinate) female to "mimic" a female—which she is, after all. With respect to the 180° turn and other submissive behavior patterns, the opposite is more probable, i.e., that they are behavior patterns of inferiority by origin and nature and that they may also appear on the part of a female in sexual encounters because she is inferior to (adult) males in most of these species, as discussed above.

Fig. 23. Submissive behavior, a. Head-foreward-downward posture in a male dorcas gazelle (Gazella doreas) in response to the threat of a superior opponent (right), b. Lying down with the head and neck on the ground and with the hindquarters toward the superior in black wildebeest (Connochaetes gnou).

Lying down with the head and neck stretched forward on the ground, often with the hindquarters toward the opponent, apparently occurs considerably more frequently in captivity than in the wild. In free-ranging artiodactyles one may see it most frequently when females presumably not yet in heat are sexually approached by males. In some species, such as black wildebeest, lying down in submission is sometimes accompanied by uttering sounds frequently heard from calves (Walther, 1966b). This and the resemblance to the infantile lying-out (lying in seclusion) behavior of certain artiodactyl species make it tempting to assume that an animal lying down in submission may be mimicking a baby and that this infantile behavior may stop the superior's aggression (Burckhardt, 1958b).

However, there are several objections to this hypothesis. It is obvious that most artiodactyles can distinguish an adult animal from an infant (by visual appearance and scent) regardless of its posture and position. Thus, it is unlikely an adult animal can assume a baby's identity simply by lying down. It is also doubtful that certain vocalizations in artiodactyles are so typical of juveniles that they definitely identify an animal as an infant. The sounds uttered in connection with submissive behavior are often distress cries. These, of course, are relatively frequently heard from juveniles, but they are neither genuine infantile vocalizations nor restricted to juveniles. Moreover, submissive lying down as well as other forms of submission are by no means necessarily combined with sounds. On the contrary, silent submission is very common in Artiodactyla. Finally, submissive lying down is a behavioral response to a threat by very superior conspecifics, who, very frequently, are adult males. In most artiodactyles, however, adult males have hardly any contact with the young and pay little attention to them. Thus, it remains obscure why such a male should react to infantile behavior. Finally submissive lying down is also shown by species, such as black wildebeest, whose young do not lie out. In short, it is more probable that lying down with the head and neck stretched forward functions as a submissive behavior simply because it is the perfect antithesis to the self-exposure and emphasizing of an animal's presence in dominance and offensive threat displays, and because the inferior animal blends into the ground and, in a sense, leaves little for its opponent to attack.

Lowering the head ranges from performances identical to the head-low posture described as a defensive threat to postures in which the animal stretches its head and neck more forward-downward (e.g., dorcas and mountain gazelle), or turns its horns somewhat away from the challenger (e.g., oryx antelope), and/or holds its neck in a rather strikingly curved fashion (e.g., guanaco: Pikers, 1956).

The back often appears slightly humped. As in the head-low posture, submissive lowering of the head may sometimes change into grazing. It is often shown during withdrawal and sometimes during flight. The most pronounced and even exaggerated performances, however, can be seen when an animal does not withdraw but remains close to its superior partner. In this case, the lowering of the head can be combined with a 180° turn, but it may also be displayed in any other orientation relative to the superior, including frontal orientation.

Besides these submissive displays, there is also appeasing by licking in a number of artiodactyl species. Typically, the inferior animal grooms the superior one, especially its head, sometimes also the neck and shoulders, as Schloeth (1961) has pointed out in an excellent study on the behavior of Camargue cattle. Of course, this behavior only occurs when the hostilities are not too severe.

The possibility, or even probability, of a relationship between submissive displays and threat behavior, as well as the occasional switch by an inferior animal from a submissive behavior to a threat display and vice versa, have been mentioned several times in the literature (Lorenz, 1935; Tinbergen, 1959). The comparative study of artiodactyl behavior may possibly provide further insight into this mechanism. As mentioned above, the head-low posture occurs as a defensive threat display in certain bovids and cervids. When an animal challenged by another's offensive threat or dominance display (high presentation of horns, erect posture, etc.) shows a defensive head-low threat, this means that it does not ("dare") respond to the challenge in an equivalent way and is not ready to attack its opponent, but is only ready to defend itself if attacked. Hence, a defensive threat when used in response to a challenger's offensive threat or dominance display, comes very close to a behavior denoting inferiority (Fig. 8). In the particular case of the head-low posture, furthermore, the animal has only to stretch its lowered head and neck somewhat more forward and/or turn the horns away from the opponent to assume a more pronounced submissive attitude. Thus, this difference is only marginal. It may be mentioned here that in the behavioral inventory of a species there are also other relationships between threat and submissive displays, which apparently depend on the position of certain displays. For example, in hartebeest and topi, instead of a head-low posture, an attitude very similar to medial presentation of horns serves as an expression of submission; and in fallow deer, submission is expressed by fast, repeated snapping movements of the mouth in combination with lowering the head or lying down. Finally, there are cases of mosaiclike combinations of threat and submissive displays—as in humans, when an inferior individual may withdraw and/ or bow when charged by a feared superior, but may curse and clench the fists at the same time. For example, in okapi (Walther, 1962), submissive lying down may occasionally be combined with vertical erecting of the head and neck, which is a (defensive) threat behavior (related to neck fighting) in this species.

In conclusion, I wish once more to emphasize the prevailing role of agonistic displays and of aggressively tinted behavior patterns in the communication and social life of Artiodactyla. If it were possible to eliminate the intraspecific aggression from the behavioral repertoires of these animals, their communication as well as their social organization would break down.

References

Antonius, O., 1939. Über Symbolhandlungen und Verwandtes bei Säugetieren. Z. Tierpsychol., 3:263-78.

Backhaus, D., 1961. Beobachtungen an Giraffen in Zoologischen Gärten und in freier Wildbahn. Brussels: Inst. Pare. Nat. Congo et Ruanda Urundi.

Benz,M., 1973. Zum Sozialverhalten der Sasin (Hirschziegenantilope, Antilope cervicapra L. 1758). Zoolog. Beitr., 19:403-66.

Bigalke, R. C., 1972. Observations on the behaviour and feeding habits of the springbok (Antidorcas marsupialis). Zoolog. Afric., 7:333-59.

Buechner, H. K., 1961. Territorial behavior in Uganda kob. Science, 133:698-99.

Buechner, H. K., and Schloeth, R., 1965. Ceremonial mating behavior in Uganda kob (Adenota kob thomasi Neumann). Z. Tierpsychol., 22:209-25.

Burckhardt, D., 1958a. Observations sur la vie sociale du cerf (Cervus elaphus) au Parc National Suisse. Mammalia, 22:226-44.

Burckhardt, D., 1958b. Kindliches Verhalten als Ausdrucksbewegung im Fortpflanzungszeremoniell einiger Wiederkäuer. Rev. Suisse Zoolog., 65:311-16.

Carpenter, C. R., 1942. Societies of monkeys and apes. Biol. Symposia, 8:177-204.

Darling, F. F., 1937. A Herd of Red Deer. London: Oxford University Press.

David, J. H. M., 1973. The behavior of the bontebok, Damaliscus dorcas dorcas (Pallas 1766), with special reference to territorial behaviour. Z Tierpsychol., 33:38-107.

Eibl-Eibesfeldt, I., 1957. Die Ausdrucksformen der Säugetiere. Handb. Zoolog., 8, 10(13):l-88.

Eibl-Eibesfeldt, I., 1967. Grundriss der vergleichenden Verhaltensforschung. Munich: R. Piper & Co.

Espmark, Y., 1964. Rutting behaviour in reindeer (Rangifer tarandus). Anim. Behav., 12:420-26.

Estes, R. D., 1967. The comparative behavior of Grant's and Thomson's gazelles, f. Mammal., 48:189-209.

Estes, R. D., 1969. Territorial behavior of the wildebeest (Connochaetes taurinus Burchell, 1823). Z. Tierpsychol., 26:284-370.

Ewer, R. F., 1968. Ethology of Mammals. London: Logos Press.

Frädrich, H., 1964. Beobachtungen zur Kreuzung zwischen Schwarzrückenducker, Cephalophus dorsalis Gray, 1846, und Zebraducker, Cephalophus zebra Gray, 1838. Z. Säugetierk., 29:46-51.

Frädrich, H., 1965. Zur Biologie und Ethologie des Warzenschweines (Phacochoerus aethiopicus Pallas), unter Berücksichtigung des Verhaltens anderer Suidae. Z. Tierpsychol., 22:328-93.

Frädrich, H., 1967. Das Verhalten der Schweine (Suidae, Tayassuidae) und Flusspferde (Hippopotamidae). Handb. Zoolog., 8, 10(26): 1-44.

Fraser, A. F., 1957. The state of fight or flight in the bull. Brit. J. Anim. Behav., 5:48-49.

Freye, H. A., and Geissler, H., 1966. Das Ohrenspiel der Ungulaten als Ausdrucksform. Wiss. Z. Univ. Halle, 5:893-915.

Geist, V., 1963. On the behavior of the North American moose (Alces alces andersoni Peterson 1950) in British Columbia. Behav., 20:377-416.

Geist, V., 1965. On the rutting behavior of the mountain goat./. Mammal., 45:551-68.

Geist, V., 1966a. The evolution of hornlike organs. Behav., 27:175-214.

Geist, V., 1966b. Ethological observations on some North American cervids. Zoolog. Beitr., 12:219-50.

Geist, V., 1968. On the interrelation of external appearance, social behavior and social structure of mountain sheep. Z. Tierpsychol., 25:199-215.

Geist, V., 1971. Mountain Sheep. Chicago: University of Chicago Press.

Gilbert, B. K., 1968. Development of social behavior in the fallow deer (Dama dama). Z. Tierpsychol., 25:867-76.

Gosling, L. M., 1974. The social organization of Coke's hartebeest (Alcelaphus buselaphus cokei). In: The Behaviour of Ungulates and its Relation to Management. V. Geist and F. R. Walther, eds. Morges, Switzerland: IUCN publications, no. 24, vol. 1:488-511.

Grzimek, B., 1949. Die "Radfahrer-Reaktion." Z. Tierpsychol., 6:41-44.

Haas, G., 1959. Untersuchungen über angeborene Verhaltensweisen beim Mähnenspringer (Ammodorcas lervia Pallas). Z. Tierpsychol., 16:219-42.

Hainard, R., 1949. Les mammifères sauvages d'Europe Paris: Délachaux 8c Niestlé.

Haltenorth, T., 1963. Klassifikation der Säugetiere: Artiodactyla. Handb. Zoolog., 8, 10(26): 1-167.

Hediger, H., 1946. Zur psychologischen Bedeutung des Hirschgeweihs. Verh. Schweiz. Naturf. Ges. Zürich, 1946:162-63.

Hediger, H., 1949. Säugetierterritorien und ihre Markierung. Bijdr. Dierk., 28:172-84.

Hediger, H., 1954. Skizzen zu einer Tierpsychologie im Zoo und im Zirkus. Stuttgart: Europa.

Hennig, R., 1962. Über das Revierverhalten der Rehböcke . Z. Jagdwiss., 8:61-81.

Hinde, R. A., 1966. Animal Behaviour. London: McGraw-Hill.

Huth, H. H., 1970. Zum Verhalten der Rappenantilope (Hippotragus niger Harris 1838). D. Zoolog. Gart., 38:147-70.

Joubert, S. C. J., 1970. A study of the social behaviour of the roan antelope (Hippotragus equinus equinus Desmarest, 1804) in the Kruger National Park. Masters thesis, University of Pretoria.

Joubert, S. C. J., 1972. Territorial behaviour of the tsessebe (Damaliscus lunatus lunatus Burchell) in the Kruger National Park. Zoolog. Afric., 7:141-56.

Kakies, M., 1936. Elche zwischen Meer und Memel. Giessen: Brühl.

Kiley, M., 1972. The vocalizations of ungulates, their causation and function. Z. Tierpsychol., 31:171-222.

Lent, P. C., 1965. Rutting behavior in a barren-ground caribou population. Anim. Behav., 13:259-64.

Leuthold, W., 1966. Variations in the territorial behaviour of Uganda kob (Adenota kob thomasi, Neumann 1896). Behav., 27:214-57.

Leyhausen, P., 1967. Biologie von Ausdruck und Eindruck. Psychol. Forsch., 31:113-227.

Lind, H., 1959. The activation of an instinct caused by a "transitional action." Behav., 14:123-35.

Lorenz, K., 1935. Der Kumpan in der Umwelt des Vogels. J. Ornith., 83:137-413.

Lorenz, K., 1963. Das sogenannte Böse. Vienna: Borotha-Schoeler.

Moynihan, M., 1955. Some aspects of reproductive behaviour in the black-headed gull (Larus ridibundus L.) and related species. Behav., Suppl. 4:1-201.

Müller-Schwarze, D., 1967. Social odours in young mule deer. Am. Zoolog., 7:807.

Müller-Schwarze, D., 1969. Complexity and relative specificity in a mammalian pheromone. Nature (London), 223:525-26.

Müller-Schwarze, D., 1971. Pheromones in blacktailed deer (Odocoileus hemionus columbianus). Anim. Behav., 19:141-52.

Müller-Using, D., and Schloeth, R., 1967. Das Verhalten der Hirsche. Handb. Zoolog., 8, 10(28): 1-60.

Pikers, H., 1954. Untersuchungen über angeborene Verhaltensweisen bei Tylopoden, unter besonderer Berücksichtigung der neuweltlichen Formen. Z. Tierpsychol, 11:213-303.

Pikers, H., 1956. Das Verhaken der Tylopoden. Handb. Zoolog., 8, 10(27): 1-24.

Pruitt, W. O., 1960. Behavior of the barren-ground caribou. Biol. Pap. Univ. Alaska, 3:1-43.

Schaller, G. B., 1967. The Deer and the Tiger. Chicago: University of Chicago Press.

Schenkel, R., 1966. On sociology and behaviour in impala (Aepyceros melampus suara Matschie). Z. Säugetierk., 31:177-205.

Schloeth, R., 1958. Über die Mutter-Kind-Beziehungen des halbwilden Camargue-Rindes. Säugetierk. Mitt., 6:145-50.

Schloeth, R., 1961. Das Sozialleben des Camargue-Rindes. Z. Tierpsychol., 18:574-627.

Schneider, K. M., 1930. Das Flehmen (I. Teil). D. Zoolog. Gart., 3:183-98.

Schneider, K. M., 1931. Das Flehmen (II. Teil). D. Zoolog. Gart., 4:349-64.

Schneider, K. M., 1934. Das Flehmen (V. Teil). D. Zoolog. Gart., 7:182-201.

Schweinsburg, R. E., and Sowls, L. K., 1972. Aggressive behavior and related phenomena in the collard peccary. Z. Tierpsychol., 30:132-45.

Signoret, J. P., and du Mesnil, F., 1960. Rôle d'un signal acoustique de verrant dans le comportement reactionnel de la truie en oestrus. C. rendues séances Acad. sei., 250:1355-57.

Simpson, C. D., 1964. Observations on courtship behaviour in warthog (Phacochoerus aethiopicus Pallas). Arnoldia, 1:1-4.

Snethlage, K., 1957. Das Schwarzwild. Hamburg: Paul Parey, p.3.

Talbot, L. M., and Talbot, M. H., 1963. The wildebeest in western Massailand, Tanganyika. Wildl. Monogr. 12.

Tembrock, G., 1959. Tierstimmen. Wittenberg: A. Ziemsen.

Tembrock, G., 1963. Grundlagen der Tierpsychologie. Berlin: Akademie.

Tembrock, G., 1964. Verhaltensforschung. Jena: Gustav Fischer, p.2.

Tembrock, G., 1965. Untersuchungen zur intraspezifischen Variabilität von Lautäusserungen bei Säugetieren. Z. Säugetierk., 30:257-73.

Tembrock, G., 1968. Artiodactyla. In: Animal Communication, T. A. Sebeok, ed. Bloomington: Indiana University Press, pp.383-404.

Thenius, E., and Hofer, H., 1960. Stammesgeschichte der Säugetiere. Berlin: Springer.

Thomas, J. W.; Robinson, R. M.; and Marburger, R. G.; 1965. Social behavior in a white-tailed deer herd containing hypogonadal males./ Mammal., 43:462-69.

Tinbergen, N., 1940. Die Übersprungbewegung. Z. Tierpsychol., 4:1-40.

Tinbergen, N., 1951. The Study of Instinct. London: Oxford University Press.

Tinbergen, N., 1952. "Derived" activities, their causation, biological significance and emancipation during evolution. Rev. Biol., 27:1-32.

Tinbergen, N., 1959. Einige Gedanken über "Beschwichtigungsgebärden. " Z. Tierpsychol., 16:651-65.

Uexküll, J. von, 1921. Umwelt und Innenwelt der Tiere. Berlin.

Verheyen, R., 1954. Monographie éthologique de V Hippopotame (Hippopotamus amphibius L.). Brussels: Inst. Pare. Nat. Congo et Ruanda Urundi.

Walther, F. R., 1958. Zum Kampf- und Paarungsverhalten einiger Antilopen. Z. Tierpsychol., 15:340-80.

Walther, F. R., 1959. Beobachtungen zum Sozialverhalten der Sasin (Hirschziegenantilope, Antilope cervicapra L.). fahrb. G. v. Opel-Freig., 2:64-78.

Walther, F. R., 1960a. Entwicklungszüge im Kampfund Paarungsverhaken der Horntiere, fahrb. G. v. Opel-Freig., 3:90-115.

Walther, F. R., 1960b. Einige Verhaltensbeobachtungen am Bergwild des Georg von Opel-Freigeheges. fahrb. G. v. Opel-Freig., 3:53-89.

Walther, F. R., 1960c. "Antilopenhafte" Verhaltensweisen im Paarungszeremoniell des Okapi (Okapia johnstoni Sclater, 1901). Z. Tierpsychol., 17:188-210.

Walther, F. R., 1962. Über ein Spiel bei Okapia johnstoni. Z. Säugetierk., 27:245-51.

Walther, F. R., 1964a. Einige Verhaltensbeobachtungen an Thomsongazellen (Gazella thomsoni Günther, 1884) im Ngorongoro-Krater. Z. Tierpsychol., 21:871-90.

Walther, F. R., 1964b. Verhaltensstudien an der Gattung Tragelaphus De Blainville, 1816, in Gefangenschaft, unter besonderer Berücksichtigung des Sozialverhakens. Z. Tierpsychol., 21:393-467.

Walther, F. R., 1965. Verhaltensstudien an der Grantgazelle (Gazella granti Brooke, 1872) im Ngorongoro Krater. Z. Tierpsychol., 22:167-208.

Walther, F. R., 1966a. Mit Horn und Huf. Berlin: Paul Parey.

Walther, F. R., 1966b. Zum Liegeverhaken des Weissschwanzgnus (Connochaetes gnou Zimmermann, 1780). Z. Säugetierk. 31:1-16.

Walther, F. R., 1968a. Verhalten der Gazellen. Wittenberg: A. Ziemsen.

Walther, F. R., 1968b. Kuhantilopen, Pferdeböcke und Wasserböcke. In: Grzimeks Tierleben, B. Grzimek, ed. Zurich: Kindler, 13:437-71.

Waither, F. R., 1969a. Ethologische Beobachtungen bei der künstlichen Aufzucht eines Blessbockkalbes (Damaliscus dorcas philippsi Harper, 1939). D. Zoolog. Gart., 36:191-215.

Waither, F. R., 1969b. Flight behaviour and avoidance of predators in Thomson's gazelle (Gazella thomsoni Günther 1884). Behav., 34:184-221.

Walther, F. R., 1972. Social grouping in Grant's gazelle( Gazella granti Brooke 1872) in the Serengeti National Park. Z. Tierpsychol., 31:348-403.

Walther, F. R., 1974. Some reflections on expressive behaviour in combats and courtship of certain horned ungulates. In: The Behaviour of Ungulates and its Relation to Management: V. Geist, and F. R. Walther, eds. Morges, Switzerland: IUCN publications, no. 24, vol. 1:56-106.