Studies on the development of communication and its role in social organization in canids have been conducted primarily on captive animals. Schenkel (1947, 1967) and Zimen (1974) provide the most detailed observations on communication and social organization in the wolf (Canis lupus), and Fox (1971a) reviews many aspects of the ontogeny and phylogeny of communication in several canid species. These studies focus principally on visual displays, and with the exception of Tembrock's research on the vocalizations in Vulpes vulpes and overview of other species (Tembrock, 1968), there is a dearth of literature dealing with vocal, tactile, and olfactory communication in this family. Kleiman (1966) gives some observations of scent-marking behavior in a few captive species, but no systematic studies have been reported on this important aspect of communication. Recent interest in mammalian pheromones and in the possible application of antipredator chemicals to protect livestock from coyotes may rectify this gap in our knowledge. Canids do have well-developed scent glands and vomeronasal, or Jacobson's, organ, and it is surprising that to date no studies have been conducted on this well-defined and accessible organ.
In summary, most work has been done on the visual modality of communication (i.e., displays) in the Canidae. An exhaustive survey by the authors of field studies of various species (many of which are reviewed by various authors in Fox, 1975a) unearthed little new material to add to our knowledge of the comparative ethology and communication in canids. It is unfortunate that few field biologists are trained to perceive and record significant communication phenomena and that most controlled studies are on captive specimens. One exception is L. D. Mech (pers. comm.), who is now investigating in detail both vocal and chemical communication in the wolf in northern Minnesota.
Before considering various modalities of communication, it should be emphasized that often more than one channel of communication may be utilized at the same time, e.g., localization plus visual display, such as a growl and threat gape. Also the same signal(s) may be given in different contexts (growl as a threat, or as a warning of danger) and signals of one modality may be simultaneously combined (superimposition), such as in the growl/scream and defensive gape, or they may be successively combined as in growl-whine and approach-withdrawal movements in play soliciting. Important variables to consider are age, sex, season, prior experience, and expectations and relations between interactees. Also, the terrain may influence the modality of communication utilized and the intensity or amplitude and frequency of the signal, as well as the proximity of conspecifics.
The Canidae include a wide range of species that differ in degrees of sociability, which is related to their socioecology and communication patterns. Species such as the wolf (Canis lupus), jungle whistling dog, or dhole (Cuon alpinus), and Cape hunting dog (Lycaon pictus) are highly gregarious and hunt in packs. These have been designated Type III canids and have a complex repertoire of subtly graded signals, which complement a more complex social organization than in other canids (Fox, 1975b). Type II canids have permanent pair bonds but do not usually form packs, the young dispersing at around ten months of age, as in the coyote (C. latrans) and golden jackal (C. aureus). The least gregarious, the Type I canids, as exemplified by the red fox (Vulpes vulpes), in which there is usually only a temporary pair bond and the young disperse around five to six months of age, have a more stereotyped communication repertoire than the more gregarious types (for further details see Fox, 1975b), especially of close-proximity visual displays. With this perspective in mind, the various modalities of communication in canids will now be reviewed and compared.
Included in this category are body postures often combined with different tail positions and movements, and facial expressions, including various positions of the ear and orientations of the eyes. Body markings are also important in enhancing certain displays.
Various gross body postures associated with changes in emotional/motivational state are schematized in Fig. 1. Darwin's principle of antithesis is exemplified by the stiff, upright threat posture and the crouched posture of submission. Changes in the direction of the lean of the body, distribution of weight in fore and hind legs, extension or flexion and turning of head and neck and movements of the tail can be identified as component units of these displays. Golani (1973) has developed a method for analysis of such movements and for interaction sequences between subjects.
Forepaw raising may be a submissive signal, expressing the intention to roll over (see Fig. 2), or a defensive warding-off reaction. Pawing directed at a conspecific is also associated with play soliciting.
Back arching as a threat display is well developed in red and gray foxes (i.e., in vulpine or foxlike canids), where the body orientation may be side-on to the rival. Possibly correlated with more frontal face-to-face threat orientation in wolves and dogs (and other more doglike canids), back arching is less frequently seen and is usually of low amplitude. It is more common in the coyote.
It is important to note not only the kind of signal that is being given but also the orientation of interactees where a "T" posture may be assumed in threat or courtship and an "L" posture in active submission. A subordinate wolf usually approaches the head of a higher-ranking wolf side-on, while a coyote threatening another will approach its side and may place its chin or forefeet on the other's shoulders.1
A common action in greeting is to twist the trunk so that the groin region is presented to a conspecific, the "C" posture, and in some species such as the gray fox ( Urocyon cinereoargenteus), coyote, and culpeo (Dusicyon culpaolus) the hind leg nearest the partner may be raised (see Fig. 3). This is the inguinal response, a derived socio-infantile display (Fox, 1971b). In the wolf, dog, and coyote it may be followed by rolling over into lateral recumbency (passive submission) and submissive urination. The interactee may respond by genital investigation, and in the wolf, by genital grooming (Fox, 1971b).
Neonatal-and infantile-derived activities in adult canids.
Other possibly derived activities are listed in Table 1, and are to be added to the repertoire of visual displays. It is to be noted that all derivedsocio-infantile activities are associated with reduction of social distance and maintenance of proximity in adults.
Changes in body position and movements of extremities are enhanced by differences in fur length and color, e.g., longer hairs on back ("hackles" in agonistic piloerection); pale hairs in external auditory meatus and black or pale tail tip enhance movements; pale-colored cheeks may orient ritualized attacks in some species; and the pale ventral body surface along neck, chest, and abdomen may enhance signals of submission, such as looking away and rolling over (Fox, 1969).
The more foxlike canids have more felinelike displays than other canids. Although all canids wag or flag the tail in greeting, the red fox has vertical and "J"-shaped positions associated with intense arousal, threat, and play-chasing very similar to the cat (Fox, 1974). Also foxes have not only the play-soliciting bow but also a roll-ing-over display in play analogous to that of the domestic cat. Coyotes have a variation of this— a combined play-bow-diving roll as a play-solic-iting signal (see Fig. 4A).
Head movements are associated with forward bite intention threat signals (see below) and turning away and breaking eye contact, as in submission, or in a higher-ranking wolf s "daring" a subordinate to attack (Schenkel, 1967), or simply ignoring it (Fox, 1971a).
Some body postures may be interpreted as compromise postures, as in the play-bow where approach and avoidance may be simultaneously combined and "frozen" into a distinct attitude. Another interpretation of the play-bow is that it is a highly ritualized form of stretching, which communicates a "relaxed" state. Such actions as yawning, grooming, preening, blinking, and eye closure in other species may be associated with relaxation and may serve secondary social functions in ritualized forms.
Stretching and yawning may be relaxed actions preceding greeting and may contribute independently to the ontogeny of the play-bow. This display may be coupled with incomplete forward and backward leaps or intention movements, i.e., approach/withdrawal ambivalence. The metacommunicative "mood" that "this is play"2 is enhanced by the open-mouth play face (see below) and vocalizations (play panting or barking in C. familiaris).
Stretching and looking away may be signals of ambivalence, the former being a displacement like the ground sniffing or brief grooming of a cat facing an adversary (Fox, 1974). Licking (i.e., repeated extrusion of the tongue) may be directed toward a conspecific in greeting in the "canine" species but is a sign of ambivalence if the tongue is curled back and touches the nose or lips.
Certain expressive movements of canids may have an habitual quality and give the animal in question a certain style or character, i.e., the "hang dog" expression of the obsequious pet dog or the "perpetual puppy" behavior of an indulged lap dog. This capacity to assume a particular demeanor in certain contexts or social situations is a quality of the more social canids and is best exemplified in the wolf pack, where the omega or lowest-ranking wolf can be easily identified by its carriage and behavior. Interestingly, when there is a change in rank, there is a change in the individual's behavior and character or body demeanor. Thus in a relatively stable social situation, certain individuals may take on a characterological set of expressive movements, a point which has considerable diagnostic significance in the somatic analysis and treatment of psychiatric disorders in man (Lowen, 1970; Fox, 1976).
Some facial expressions may vary in intensity, as in threat, where the aggressive pucker or mouth-closed, lips-forward expression gives way to a more open-mouth expression with the lips retracted vertically to display the teeth. This latter action is absent in "vulpine" canids. Besides successive shifts in intensity, facial expressions associated with different contexts or emotional states may also occur simultaneously, as in the superimposition of the submissive grin (horizontal retraction of lips) with an agonistic gape and snarl (see Fig. 5, 9).
In the vulpines, the facial expressions lack the subtle gradation of intensity of the more doglike canids, being more stereotyped and with a lower incidence of simultaneously combined expressions. Tables 2 and 3 list the components of various expressions observed in canids together with their occurrence in different species.
Components of expressions associated with increase of social distance and aggression.
Components of expressions associated with decrease of social distance and submission.
The variety of facial expressions that canids can display is considerable; at least ten distinct categories have been identified—almost as many as described by van Hooff (1967) in primates (see Table 4).
As in most displays, these expressions serve to increase, decrease, or maintain a certain distance from one or more conspecifics (see Figs. 6 and 7). Some may also serve as alarm signals in certain contexts. In general they serve to express the intentions of the animal. The complete lack of any overt signal may also have significance. Fox et al. (1974) use the term "passive" indifference for the lack of any overt reaction by a dominant (alpha) wolf when threatened by the alpha wolf of another pack in the latter's territory. Other wolves displayed submission or defensive aggression.
On the basis of spectrographic evidence (Cohen and Fox, 1976), eight basic sound types of canids have been identified: whines (including shorter "yips" and yelps, and longer, softer whimpers), screams, barks, growls, coos, howls, mews, and grunts. Variation within these sound categories may occur at both the individual and the interspecies levels, although the degree of such variation must be largely dependent on the structural limitations of the vocal apparatus. The sounds may vary in duration, frequency, intensity, cyclicity, and context.
Not all these basic sound types are included in the vocal repertoire of every canid species, and the same sound may be used by different species in very different contexts (see Table 5). Foxes, for example, are the only canids known to emit a pure scream while greeting conspecifics, and domestic dogs will bark in many situations (e.g., threat, play soliciting, contact seeking), while foxes bark only in threat.
The following brief descriptions of the eight major canid sounds will point out some of the similarities and differences between certain physical properties of the sounds:
Whines. Whines are commonly heard in wolves, coyotes, foxes, and domestic dogs. They are wide-banded, cyclic sounds of short duration and moderate frequency variations. The clear horizontal stratification typical of the whine is evident in Figure 8A.
Screams. The scream, like the whine, is common among canids. Although it too is a wide-banded sound, it differs in that its frequency variations are much greater, it may be of longer duration, it is noncyclic, and it is delivered with much greater intensity. It also exhibits strong horizontal stratification (see Fig. 8B).
Comparison of facial expression in primates (after van Hooff 1967) and canids.
Barks. Barking may also be heard in most canids, although the context may vary greatly. The bark is a short wide-banded sound with few frequency variations. It is often cyclic in wolves, coyotes, and especially dogs, but is nöncyclic in foxes (see Fig. 8C).
Growls. The growls of most canids are wide-banded, with the lower frequencies often carrying more energy than the higher ones. The duration and intensity of the growl may vary greatly with motivation. It is noncyclic and exhibits strong vertical stratification (see Fig. 8D).
Coos. Foxes are the only canids whose vocal repertoire is known to include the "coo."4 This sound is typically characterized as a short call with a wide frequency range and short (vertical) frequency variations of moderate degree. Although the sound may be repeated, it is not of a cyclic nature and often occurs when the animal is socially isolated (see Fig. 8E).
Howls. The howl is commonly heard in wolves, coyotes, and some breeds of domestic dog. It is not heard in foxes. Lower frequencies are more prominent than higher frequencies. It is characterized as a long-to-extended sound (i.e., 1.0+ sec) with relatively few frequency variations. It may be repeated but is of a noncyclic nature (see Fig. 8F).
Mews. The mew is heard in neonates of wolves, dogs, coyotes, and foxes, generally when the animal is in some way distressed (i.e., cold, hungry, etc.). It is a short, wide-banded, cyclic sound with moderate-to-great frequency variations. Foxes are the only common canids known to produce this sound in the adult stage (see Fig. 8G).
Grunts. Grunts are often heard in wolves, coyotes, and dogs. The frequency range may vary from narrow to wide. The sound is very short (usually less than 0.5 sec), cyclic, and with few frequency variations (see Fig. 8H).
Further description and discussion of these eight sound types may be found in Tembrock (1960, 1968), Theberge and Falls (1967), and Cohen and Fox (1976).
In addition to these vocal sound types, all species communicate to some extent by means of more "mechanical" sounds such as the guttural "clicking" heard in foxes, or the tooth-snapping heard in wolves, coyotes, and dogs. Because these "mechanical" sounds are of relatively low volume and do not carry very far, they are reserved for close-contact situations and are often heard during agonistic encounters.
The above sound types may be mixed in one of two ways: either by superimposition of two vocal sounds or by successive emission of two or more types (vocal and/or mechanical) (see Fig. 9A-F). A combination of these two processes is also possible; for example, a mixed sound might follow a pure sound. The phenomenon of sound mixing correlates well with the superimposition of "body language" postures and facial expressions described earlier. A wolf faced with a situation simultaneously eliciting fear and aggression, for example, may concurrently display facial and body characteristics expressive of both motivations (Schenkel, 1947). The accompanying vocalizations are likely to include components that may be heard separately in pure form in fearful and aggressive contexts, respectively.
The ontogeny of tactile communication involving contact-seeking behavior for warmth and care (grooming, etc.) has not been systematically studied in canids. Contactual circling (Fox, 1971b) and body leaning have been identified as a derived socio-infantile action in wolves and dogs; adult wolves may run together, for example, frequently in contact or bumping each other. The inguinal response, a socio-infantile derived activity, may evoke or solicit genital grooming by a conspecific in the wolf. Wolves of a particular allegiance may engage in this activity (Fox, 1972), or it may occur in the absence of any identifiable sexual or filial bond. The groomer may assume the role of "parent" (or care giver) in this ritual, irrespective of rank since the alpha wolf may solicit genital grooming from a subordinate.
Although we have not yet quantified the frequencies of bouts of face and body grooming in captive canids, we have the impression that the vulpines engage in more social grooming, especially during the breeding season, than do the canine types.5
Forelimb pawing and upward pushing with the muzzle or nose are friendly gestures that are visual displays at a distance but at closer proximity lead to contact. The body region of the conspecific that is touched may have some significance. Touching the shoulder region may be intimidating or a threat gesture and is used by males to "test" the readiness of the estrus female to stand for copulation. The sides of the mouth and groin region are associated with submissive and friendly contact in canids (associated with derived infantile food soliciting [greeting] and genital grooming).
Touching during social investigation is usually confined to certain discrete body regions (see Fig. 10). Some of these regions may be rich in scent or apocrine glands, notably the perioral, anal, genital, and aural regions, and dorsum of the tail (supracaudal gland). In some dogs (male mute, husky) and some wolves, a submandibular gland has been detected, and this gland, in the region of the cheek, may evoke cheek-oriented investigation by conspecifics. Little is known about the social significance of these glands and the area is open for further studies.
Probably one of the least-understood phenomena in the study of mammalian behavior is the role of the olfactory sense. Tembrock (1968), Ewer (1968), Gleason and Reynierse (1969), and Eisenberg and Kleiman (1972) have reviewed various aspects of scent marking and olfactory communication in vertebrates. Intraspecific odors may give information about individual identity, sex, tribe or pack, age, and physiological status; and in marking behavior one sees self-marking, partner-marking, and marking of territory. To this list of potential information transmissions we might add that social status or rank as well as the emotional state of the animal might be identifiable, for many mammalian species deposit various odors when they are alarmed. They are not usually deposited at a specific marking spot, although the actual composition of the secretions deposited on a traditional or specific marking spot may be altered by such emotional reactions.
Ewer (1968) proposes that the animal's own smell restricts it to a familiar home range and scent marking tends to space animals out: a stranger's odor may be threatening and the animal's own odor reassuring. One of the few works dealing with the subject in Canidae is by Kleiman ( 1966), who presents the following basic theories regarding the action of scent marking in canids:
(1) Scent marking originated as a device for familiarizing and reassuring the animal when it entered a strange environment.
(2) Secondary, associated functions have arisen, such as the bringing together of the sexes and the maintenance of territory, which are important to the survival of the species.
(3) There exist species-selective patterns of response to visual and odor-bearing objects; i.e., some species may use one marking method to the exclusion of others. Species also differ in their marking postures.
(4) Scent marking may serve to inform animals of the population density of conspecifics and the number of rivals, estrus females, etc.
(5) Scent-marking postures in conjunction with other behavioral and anatomical characteristics can be used to establish taxonomic relationships. Sprague and Anisko (1973) have studied elimination behavior in laboratory beagles and present useful observations on various patterns of elimination, but add little to our understanding of the role of marking behavior per se.
The reactions of captive wolves and red, gray, and Arctic foxes to propionic acid (a major component of anal gland secretion in red fox; Albone and Fox, 1971), skatole, and urine and feces of familiar and strange conspecifics, other canid species, rat, and impala were studied in our laboratory. Findings are summarized in Table 6.
The response of defecation during the tests with conspecific feces and urine exclusively (Table 6) can be interpreted in several ways. Defecation may be a means of intraspecific communication, a scent deposited on the strange conspecific odor to warn an intruder away from the den site. Another interpretation has it that such behavior is instrumental in the making of new acquaintances (i.e., leaving a "calling card"). The action could also simply mask a "threatening" strange conspecific odor by covering it with the animal's own odor; a reassurance function might then be proposed, as was done by Ewer (1968). All these interpretations are warranted when the wolf is compared with the other canids in which defecation occurred in response to a wider range of stimuli; a social function may therefore be subserved by fecal marking in the wolf and to a lesser extent in the less-social canids such as the red fox (Table 6). In the latter species this form of marking is more generalized and includes the marking of novel objects with feces.
THE GRAY FOX
The gray foxes were noted for their use of rubbing almost exclusively as a scent-marking device (see Table 6). No urination on the test plate was ever observed, and the sole occurrence of defecation was in response to their own urine, perhaps just a reflex response to a novel stimulus, the plate. Rubbing was used extensively in the tests with strange conspecific odors. In this species the scent-marking behavior of rubbing may haveselectively evolved as a specialized pattern. As Kleiman (1966) points out, some canid species tend to use one marking method to the exclusion of others. Our contention that the gray fox rubbing behavior is socially based rather than being a simple reduction of strangeness is supported by two observations. One is that in most instances, rubbing was followed by mutual sniffing and licking of the neck, face, and anal regions. Secondly, there is the absence of behavior such as urination and defecation and burying, which can be more easily interpreted as actions to hide or mask the odor or as simple responses to the novelty of the stimulus. Further support of a specialized social function comes from the observations that gray foxes, unlike other fox species studied, engage in much social grooming and appear to maintain the pair bond outside the breeding season. Behavior such as mutual licking and grooming (as occurs after a strange scent is added by rubbing) is well known among many species of mammals to have the function of renewing and strengthening social bonds.
THE RED FOX
In the test of the red foxes with strange conspecific feces, the smaller female urinated on the plate immediately after sniffing the feces of the larger female, who had defecated on the plate. It is our interpretation that this urination is specifically a scent-marking behavior elicited in response to the conspecific's action of defecating rather than being just a response to the odor on the plate. This is the sole occurrence of urination in all our tests with the red foxes. The larger female also defecated on the smear sample of impala feces. It is perhaps relevant here that Burrows (1968) in his studies of the red fox in England found that novel objects such as a trap or a dead dog would be covered with fox feces; this can only be interpreted as a marking response evoked by a novel stimulus in an otherwise familiar territory. Arousal that triggers marking may be a generalized phenomenon in some mammalian species. It might be argued that as arousal increases toward the edge of the territory or in unfamiliar territory or in the presence of novel objects, the probability of marking with urine, feces, or some specific glandular secretion is increased.
THE ARCTIC FOX
The Arctic foxes (Alopex lagopus) differed from the wolves and the gray and red foxes in their overall use of urination and, to a lesser extent, defecation as a marking method. Why the Arctics should prefer this one pattern and the reds and grays should not is a matter for further study.
Defecation was only used as a response in two tests, one with impala feces and one with the artificial scents. Rubbing was not observed in this species in any of the tests, but in a later study, rubbing was elicited in Arctic foxes by gray fox feces. The low frequency of rubbing in this species might be construed as evidence that there is little direct social function in the Arctic's rubbing behavior. We have never observed these foxes engaged in social grooming, and social investigation is of much lower frequency than in the wolf and gray fox.
The most striking observation that can be drawn from our tests with the wolves is the exclusive occurrence of defecation in response to strange conspecific odors. This sharply contrasts the use of urination and rubbing for interspecific, prey, and artificial odor tests. We think that this distinction may be related to the highly developed sociability of the wolf. Rubbing may enhance social integration and communication in the pack. One could hypothesize that there could be less tendency to add to the body the scent of a conspecific, foreign to the pack.
In contrast, the distinction in the red and gray foxes was less clear-cut. Defecation occurred in response to the foxes' own odors, the control plate, conspecific tests, and prey scents, but there was no pattern or regularity as seen in the wolf. We tend to ascribe less of a social function to defecation by the foxes, and interpret the actions as a response to a novel stimulus but which may also serve to mask a strange odor.
Kleiman (1966) has defined scent marking in canids as follows: "Scent marking has been defined as urination, defecation, or rubbing of certain areas of the body which is (1) oriented to specific objects, (2) elicited by familiar conspicuous landmarks and novel objects or odors, and (3) repeated frequently on the same object." Using the data from our study and the framework of the above definition, we would define scent marking in terms of our experiment as (1) urination, defecation, and/or rubbing of the body oriented toward the test plate, (2) response to a novel, unfamiliar object or scent (note that while not denying that scent marking is elicited by familiar, conspicuous objects, this part of Klei-man's definition has little bearing on our study). The frequently repeated response to familiar objects was observed during the testing situations in our subjects but had no relation to the test stimulus. The male gray fox, for example, was observed to urinate continually on the middle, back steel part of his cage, resulting in an accumulation of crusted matter, probably a combination of urine and sawdust.
Kleiman (1966) also observes that
Although dogs and foxes could use feces as material for marking, defecation is not used by the members of most species as a means of depositing odor on a specific object or landmark. The fennec fox and Arctic fox were occasionally observed defecating in response to novel stimuli: however, the act was never repeated more than once.... In the case of the Arctic foxes and fennec fox, feces were considered to be a marking substance because they were directed onto novel objects.
We would like to point out that in contrast with Kleiman's statement, our wolves and red and gray foxes did use defecation as a scent-marking response; in addition, in most cases we ascribe more significance to this action than to a simple response to a novel stimulus. The response of defecation may well have some value in intraspecific communication, especially in the wolf. This hypothesis requires further investigation.
To explain the rubbing and rolling of canids (see Fig. 10), we offer four possible hypotheses: (1) The animal, wolf or otherwise, rolls in the novel odor to familiarize itself with it; by rubbing, the animal adds the strange odor to its own body odor in order to make it less strange. This hypothesis does not account for the fact that the animal is attracted to the odor; also, an underlying motivation to become familiarized with the novel odor by "wearing it" is less tenable than the second hypothesis. (2) Rolling and rubbing may have important social consequences. In the wolves, adding a new scent to the body might influence within-pack social interactions. D. Guthrie (pers. comm.) proposes that "wearing" an odor (either a strange odor or the animal's own urine, as in ungulates) may be related to status-identity and self-assertive ("this is me") display. The odor might increase social investigative behavior and attention, and reduce aggression or assertion of dominance by socially superior individuals. (3) The animal imparts some of its own body odor onto the material in which it rolls, and, as happens when a conspicuous object (a rock or a tree) is marked, other individuals will be attracted to the material and learn of the first animal's presence. This hypothesis is dubious in that a more efficient way of marking in canids is to use urine or feces. (4) This behavior is simply a response to a pleasurable stimulus: wolves will often roll and rub against a deer carcass before they eat.
One interesting ontogenetic phenomenon is the control of urination. During the first three weeks in canids it is under maternal control, for the cub does not urinate until it is licked by the mother, who ingests urine and feces. This must certainly help in keeping the den clean and dry while the mother is away. When the neonate's bladder becomes extremely distended it will be spontaneously evacuated, however. With the subsequent development of voluntary control of urination and defecation, the young canid begins to evacuate independently at some distance from the den, often at a particular spot. Possibly as a consequence of emancipation and ritualization (Fox, 1971b), urination can be evoked by a conspecific; usually the latter is older and socially dominant. Submissive urination by a subordinate is seen during greeting or active submission (Schenkel, 1967) and is often associated with passive submission. This has been observed by the authors while being greeted by domestic dogs and hand-raised coyotes, golden jackals, and wolves. An intriguing, possibly partially emancipated submissive urination is seen in hand-raised red foxes. While greeting each other or their handler, they will briefly squat and urinate. This action is very brief, lacking the clear lateral recumbency or sideways twist with genital presentation seen in the aforementioned more social canids, and appears to be less of a social display than a nonritualized arousal or excitement reaction. The latter interpretation is supported by the fact that one female red fox not only urinates but also invariably defecates in a greeting-play soliciting display to one particular handler.
After urinating, many canid species will scrape6 the ground with the feet. The probability of occurrence of this behavior is increased by the presence of a rival conspecific. Scraping may add a visual signal to the mark (analogous to tree clawing in the leopard), or an additional olfactory cue may be deposited on the ground by interdigital scent glands, which are present in red foxes (Tembrock, 1968) and possibly in other canids.7
Summary and Conclusions
Details of visual, auditory, tactile, and chemo-olfactory communication in various canid species are reviewed from an ethological standpoint. Comparisons between species add to our understanding of socio-ecology, where the behavior patterns of communication are related to the type of social organization that has evolved as an adaptation to a particular niche or life style (Fox, 1975b).
Future research could be advantageously directed toward more detailed studies of chemo-olfactory communication, an area where the least amount of work has been done in canids and in mammals in general. Rather than describe differences in behavior in various contexts (sexual, maternal, agonistic, etc.) this review focuses on comparisons between species of various modalities of communication since the same signals may occur in different contexts. In studying canid communication, it is considered important for the observer to know the nature of the relationships between interactees as well as their prior experiences, and imperative that the context be clearly defined.
Since many canid species will hybridize, cross-breeding studies may throw further light on the inheritance of certain actions and communication signals; the role of experience and of genetic factors in the encoding and decoding of visual, auditory, and chemical signals await future research. At this stage it may be concluded that while the expression of most signals (displays and vocalizations) are experience independent (i.e., relatively environment resistant), their decoding may be more experience-dependent.
In other words, how a canid communicates may indeed be innate, but how it responds and to whom and when may be significantly modified by early experience. This variable of early experience must be considered in studying captive and/or hand-raised animals, and, whenever possible, field studies should complement those of the laboratory. A wild canid also knows many of the signals of other species sharing its habitat and is exposed to a much more complex and variable environment in infancy than one raised in captivity. In the more social species, protocultural influences must be considered, where individual differences reflect differences in rank and social role, as differences between species reflect social adaptations to different sets of ecological factors.
Albone, E., and Fox, M. W., 1971. Anal gland secretion in the red fox. Nature, 233:569-70.
Bekoff, M., 1972. The development of social interaction, play, and metacommunication in mammals: an ethological perspective. The Quarterly Review of Biol., 47:412-34.
Burrows, R., 1968. Wild Fox. Newton Abbot, England: David and Charles.
Cohen, J. A., and Fox, M. W., 1976. Vocalizations of wild canids and possible effects of domestication. Behavioural Processes 1:77-92.
Eisenberg, J. F., and Kleiman, D. G., 1972. Olfactory communication in mammals. Annual Rev. Ecology and Systematics, 3:1-32.
Ewer, R. F., 1968. Ethology of Mammals. New York: Plenum Press.
Fox, M. W., 1969. The anatomy of aggression and its ritualization in canids. Behaviour, 35:242-58.
Fox, M. W., 1970. A comparative study of the development of facial expressions in canids, wolf, coyote and foxes. Behaviour, 36:49-73.
Fox, M. W., 1971a. Behavior of Wolves, Dogs, and Related Canids. New York: Harper Sc Row.
Fox, M. W., 1971b. Socio-infantile and socio-sexual signals in canids: a comparative and ontogenetic study. Z. Tierpsychol., 28:185-210.
Fox, M. W., 1972. The social significance of genital licking in the wolf, Canis lupus. J. Mammal., 53:637-39.
Fox, M. W., 1974a. Understanding Your Cat. New York: Coward McCann.
Fox, M. W., 1974b. The behaviour of cats. In: The Behaviour of Domestic Animals, 3d ed., E. S. E. Hafez, ed. London: Bailliere.
Fox, M. W., ed., 1975a. The Wild Canids. New York: Van Nostrand Reinhold.
Fox, M. W., 1975b. Evolution of social behavior in canids. In: The Wild Canids, M. W. Fox, ed. New York: Van Nostrand Reinhold.
Fox, M. W., 1976. Between Animal and Man. New York: Coward McCann.
Fox, M. W.; Beck, A. M.; and Blackman, E.; 1975. Behavior and ecology of a small group of urban dogs. Applied Animal Et hol., 1:119-37.
Fox, M. W., and Clark, A. L., 1971. The development and temporal sequencing of agonistic behavior in the coayote (Canis la trans). Z. Tierpsychol., 28:262-78.
Fox, M. W.; Lockwood, R.; and Shideler, R.; 1974. Introduction studies in captive wolf packs. Z. Tierpsychol., 35:39-48.
Gleason, K. K., and Reynierse, J. H., 1969. The behavioral significance of phereomones in vertebrates. Psychological Bull., 71:58-73.
Golani, I., 1973. Non-metric analysis of behavioral interaction sequences in captive jackals (Canis aureus L.). Behaviour, 44:89-112.
Kleiman, D., 1966. Scent marking in the Canidae. Symp. Zool. Soc. London, 18:167-77.
Lowen, A., 1970. Depression and the Body. New York: Coward, McCann.
Schenkel, R., 1947. Expression studies of wolves. Behaviour, 1:81-129.
Schenkel, R., 1967. Submission: its features and functions in the wolf and dog. Amer. Zool., 7:319-30.
Sprague, R. H., and Anisko, J. J., 1973. Elimination patterns in the laboratory beagle. Behaviour, 47:257-67.
Tembrock, G., 1960. Spezifische Lautformen beim rotfuchs (Vulpes vulpes) und ihre beziehungen zum verhalten. Säugertierkunde. Mitt., 8:150-54.
Tembrock, G. 1968. Land mammals. In: Animal Communication, T. A. Sebeok, ed. Bloomington: Indiana University Press, pp.338-404.
Theberge, J. B., and Falls, J. B., 1967. Howling as a means of communication in wolves. American Zoologist, 7:331-38.
van Hooff, J. A. R. A. M. 1967. The facial displays of catarrhine monkeys and apes. In: Primate Ethology,D. Morris, ed. Chicago: Aldine, pp.7-68.
Zimen,E., 1974. Social dynamics of the wolf pack. In: The WildCanids, M. W. Fox, ed. New York: Van NostrandReinhold.
1. Standing-over is an agonistic posture in the coyote (Fox and Clark, 1971) and is also seen in the golden jackal, wolf, and domestic dog. It is to be differentiated from clasping, where one seizes the other around the waist or chest between its forelimbs. This frequently occurs in aggressive contexts but is most often identified in sexual contexts. Reciprocal clasping or "hugging" is seen during contact-play or play-fighting and courtship in "canine" types, while rearing up together and pushing with both forelimbs is common to both vulpine (foxlike) and canine (doglike) canids in agonistic interactions.
2. See Bekoff (1972) for discussion of metacommunication and play in canids and other animals.
3. For details of the ontogeny of facial expressions in canids, see Fox (1970, 1971a).
4. The Asiatic wild dog (Cuon atpmus) emus a whistle call for contact or assembly, but without spectrographic analysis we cannot determine if this sound is unique or a higher-frequencv form of the "coo."
5. A red fox caged temporarily with a golden jackal groomed the latter's eyelids until they were raw and hairless. Whereas in the red fox the groomee stops the groomer when it has had enough, the opposite is true in the jackal, who remains motionless until the groomer has finished.
6. Scraping may also occur in the absence of urination in the dog, wolf, and coyote during agonistic (threat) interactions.
7. Canids may also mark over, that is, they urinate where a mate or companion has just marked (Fox et al., 1975).