The human-canine bond, as we see it today in the United States, is the modem manifestation of a most important and unusual event which occurred over 11,000 years ago in the Shanidar Cave region of Iraq. This event was the domestication of the dog. As the first case of animal/plant domestication and the only one to occur during the hunter-gatherer stage of man's cultural development, this early establishment of the human-canine bond paved the way for a process that has since provided man with a wide variety of food-producing, transport and companion animals which have helped him to cope with environmental and climatic change and even the social stress associated with living in today's complex society. The purpose of this presentation is to suggest that an understanding of the changes which occurred in both the dogs themselves and in their relationship with their human consorts during the first several thousand years following the initial man-dog contact, are still relevant to a number of problematic aspects of the human-canine bond as we see it in our society today. These aspects include but are not limited to: problems with dog bites, domestic dog health and psychological well-being, pet overpopulation and more. Most efforts to study and understand the domestication of the dog and the associated early development of the human-canine bond have focused on an archaeological approach to these issues.1,2,3 An alternative but complimentary approach has involved anthropological studies of man-dog relationships within present day aboriginal or culturally simple human societies.4,5 We propose here a third approach to studying the early human-canine bond — namely the documentation of the basic biology, behavior and ecology of present-day populations of primitive free-ranging or semi-confined dogs. The subjects of such studies are dogs whose characteristics suggest a close descent of type, if not direct genetic relationship to those canines which participated in the initiation and early development of the human-canine bond many thousands of year ago. In particular, we will describe two forms of primitive dogs, emphasizing studies of their basic ecology and behavior under semi-confined or free-ranging conditions. These forms include the New Guinea Singing Dog (Canis lupus dingo)6,7,8 and the so-called Carolina Dog, a registered domestic breed developed from the captive breeding of wild-caught individuals showing a primitive long-term feral/pariah phenotype, acquired from several locations in the southeastern United States.^9,10,11

The Domestication and Dispersion of Primitive Dogs
   A number of accounts summarize the evidence for the domestication of the dog in the Middle East and present details of the developing relationship between neolithic humans and sympatric wild Canis of that region.^1,2,3,12 Most of these domestication scenarios propose one or more southwest Asian subspecies of the wolf (eg. Canis lupus pallipes, C. l. arabs) as the canids most likely to have been involved in this process and this position is generally accepted by the available archaeological material. However, this initial contact was followed by the rapid spread of these primitive man/wolf-dog symbionts out of the Middle East, moving southward into Africa and eastward across the Indian subcontinent into southeast Asia, and there is relatively little useful archaeological material available from these regions. There is thus a notable lack of information about the characteristics of these early domesticating wolf-dogs and their relationship with their hunter-gatherer consorts during the "long walk" of this critical phase of the early development of the human-canine bond. What information is available however, points to a high degree of uniformity in both the cranial/skeletal features and the external body phenotypes shown by the canids associated with that dispersion.¹³ This degree of morphological uniformity converging on an external phenotype typified by that of the Australian Dingo, is even more remarkable considering the broad biogeographic range involved.^5,13 The animals most closely associated with this early dispersal phase of the domestication process show the generalized "poutiatini" cranial morphotype of the so-called "southern pariahs"² Interestingly enough however, the uniform external body morphotype shared by all of these canids throughout the far-flung reaches of their dispersion, bears little resemblance to those of the wolf subspecies which are generally considered to have begun the domestication process in the Middle East. The general appearance and body-type of the southern pariahs are rather those of the prototypic "yaller" dog: a sharp pointed muzzle with erect pointed ears, giving a distinctively foxlike appearance, a characteristically fish-hook-shaped tail usually showing a whitish or pale coloration on the underside, and a uniform reddish-yellow to ginger body color with a short, dense pelage.^a Rather than typifying southwest Asian wolves, these features rather characterize the appearance of two other wild canid species that also occur within the same biogeographic region as the location of the dog's early domestication and initial early dispersion. These species are the Simien Wolf (Canis simensis) of North Africa and the Dhole [Cuon (now Canis) alpinus] of the Indian subcontinent of southeast Asia.^b Although neither of these species have ever been formally considered as possible ancestors of the domestic dog, the above issues of similarity of appearance, biogeography and close behavioral/social overlap with free-ranging domestic dogs^14,15 suggest the importance of considering the possible roles of one or both of these two species, perhaps through introgressive hybridization of symbiotic pariah Canis lupus from southwest Asia, in giving rise to the generalized southern pariah type of today's domestic dog. Whatever their ancestry and wherever they may occur however, populations of free-ranging dogs which approximate the southern pariah/poutiatini long-term feral morphotype offer the potential for studies which may help in understanding how this morphotype and any of its uniquely associated behavioral and/or ecological characteristics became established and subsequently may have contributed to the development of the human-canine bond. We now describe here ongoing studies of this kind for semi-confined populations of two of these primitive southern pariah types.

The New Guinea Singing Dog
   Although primitive dogs have been known to occur throughout the island of New Guinea (currently comprised of the two nations of Papua New Guinea and Irian Jaya) for many thousands of years, the native range of the most primitive and feral form, the New Guinea Singing Dog (NGSD), has been restricted since the time of its discovery to the higher cloud forests and associated alpine/subalpine habitats of the island's Central Highlands and other mountainous habitat.⁸ The remote and isolated nature of this area delayed this dog's discovery and recognition by the scientific community until the mid-late 1950's.^7,8 When first discovered, the NGSD was initially described as the new species Canis hallstromi, based on the external appearance of the first two captured dogs which were subsequently held and bred at the Taronga Park Zoo in Sydney, Australia.⁸ Offspring from this initial pair were subsequently bred and their progeny were distributed widely throughout many of the world's major zoos. Later, however, taxonomic analyses based on skeletal and cranial characteristics suggested that this animal was not a new species but rather only a primitive feral form of the domestic dog (then designated as Canis familiaris).⁶
   The "rediscovery" of the NGSD as a unique taxonomic form, distinct perhaps at the subspecific level from all other feral or domestic dogs of the Canis lupus-familiaris complex, occurred in the early 1990's and was based in part on studies of a newly outcrossed captive population of dogs in North America which included bloodlines of new founders captured by a German expedition to a remote region of Irian Jaya in the late 1970's.^7,8 These studies included descriptions of unique patterns of reproduction, behavior, social behavior and particularly vocalizations under semi-confined captive conditions. This work, together with a review of several previously-published molecular genetic studies, resulted in the proposal to describe the NGSD as belonging to the same subspecies as the Australian Dingo, Canis lupus dingo.⁸ Subsequent unpublished multilocus studies of genomic DNA have further supported the validity of this designation.^c
   Having evolved for many thousands of years in an environment free of other members of the Canis (eg. wolves and coyotes), the Australian Dingo and particularly the NGSD, offer a unique opportunity to describe and study some of the most primitive characteristics of the first members of the dog-wolf complex to participate in the formation and early development of the human-canine bond.^d Of particular importance in this regard is the degree of apparent social monogamy shown by the NGSD under semi-confined captive conditions.⁸ Since all of the wild canids which were likely involved in the domestication process are generally highly social in nature, and most all other domestic and feral dogs, including the Australian Dingo are also highly pack-oriented, the non-pack monogamous status of the NGSD may represent the low point in a U-shaped continuum of social complexity which developed throughout the process of formation of the human-canine bond.⁸ The fact that dogs were likely to have reached New Guinea and other islands of the Australasian region through travels with early seafaring peoples, may have contributed to a more monogamous nature in these animals. Certainly single pairs of dogs would have been easier to maintain than larger packs on such voyages, and many island colonization events likely involved only a pair or single monogamous family group. The consequences of such a possibility for the later development of the human-canine bond in island vs. larger continental habitats (eg. Australia) should be explored further - particularly through studies of the more primitive forms of dogs which may still exist in a free-ranging state, free of genetic/social contact with more modem forms of domestic dogs and/or other wild species of Canis.

The Carolina Dog
   Unlike the situation described above for the New Guinea Singing Dog, the first primitive dogs to accompany humans across the Bering Land Bridge and into North America about 8,000 years ago entered a continent already inhabited by at least two and possibly three wild species of Canis (the Gray Wolf, C. lupus; Coyote, C. latrans; and Red Wolf, C. rufus). It has now been well-established that hybridization eventually occurred between most all of these canids.^16,17 These hybridizations together with crosses to modem European dogs which were subsequently established on the continent over the past 500 years, have created a situation in which it is unlikely that more than a few, if any, remnants of the original primitive dog type still exist in a genetically pure state. Recently however, in the southeastern United States, certain free-ranging dogs have been discovered whose external body phenotype closely resembles that of the Australian Dingo and other primitive Australasian feral/pariah dogs.^11,18 This discovery has provided an opportunity to test the hypothesis that such animals may represent close descendants of type if not direct genetic ancestry, from dogs which first crossed the Bering Land Bridge with primitive humans and subsequently helped to shape the early development of the human-canine bond in North America. To date, most attempts to test the predictions of this hypothesis have involved behavioral/ecological studies of wild-caught dogs and/or their first- or second generation captive-bred progeny.⁹ While the majority of these studies have not yet produced definitive/publishable results, they have already begun to describe a number of traits that have never before been recorded for any other member of the genus Canis. In most cases moreover, these unique traits seem to be those that would suggest some form of adaptation to the ecological niche occupied by these dogs in rural/uninhabited areas of the southeastern United States. An example of such a trait is an unusual pattern of changes in the spacing of estrous cycles throughout the lifetime of individual females (Figure 1). An extraordinarily high frequency of estrus (up to three cycles per year) at earlier ages, followed by longer interestrous periods at older ages, would be a reasonable (ie. adaptive) pattern for females which are almost certainly under a strong selection pressure to produce one or more litters quickly before succumbing to any of what must be a large number of mortality factors (eg. heartworm disease) that would take their toll at ages of two to three years.

Conclusion: An Agenda for Future Research and Conservation Priorities
   Taken together, even the preliminary results of the studies that have so far been conducted on these dogs are beginning to suggest that during the early development of the human-canine bond in North America and elsewhere throughout the world, many of the behavioral and ecological traits shown by the dogs involved were the result of environmental selection pressures or selective breeding choices made by the dogs themselves, rather than being the result of artificial selections imposed upon the dogs by their human consorts. The ability to identify and value such traits, as displayed within remnant populations of primitive free-ranging dogs, could provide a useful approach to current-day problems such as unwarranted canine aggression, public health issues, genetic diseases of modem domestic dogs and pet overpopulation. The extent to which this can be done however, is dependent on the existence of genetically isolated populations of primitive dogs whose ecological characteristics and relationships with any local human populations still continue to approximate those of the earliest dogs that participated in the formation of the human-canine bond. Such populations of dogs are only found today. in those parts of the world where few if any modem domestic breeds of dogs are found and where local human populations still maintain basic lifestyles and attitudes towards dogs and other livestock. Today, most such areas are found in the Central Highlands of Papua New Guinea and Irian Jaya⁸ and in remote regions of the Australian Outback.⁵ In North America, the existence of several relatively large isolated tracts of land from which public access and free-ranging modern domestic dogs are largely excluded, may also still provide opportunities for small pockets of genetically isolated primitive dogs to exist. Such land holdings in the southeastern United States include acreage associated with military bases and government weapons production facilities such as the United States Department of Energy's Savannah River Site (SRS), near Aiken, South Carolina.¹⁹ Studies at this site for example, have confirmed that the proportional numbers of free-ranging dogs whose external body phenotype closely resemble those of the Australian Dingo, are significantly higher in the lands on or directly adjacent to this 780 km² site than in surrounding habitats (Figure 2).^f Studies of dogs captured from lands surrounding the SRS, and their first- and second-generation captive-bred offspring, have begun to reveal unusual behavioral traits, some of which seem to be seasonally adapted to the dogs local environmental conditions.^e Until recently, primitive dog populations of the SRS and other tracts of isolated land in the southeastern United States have remained free from contact with any native wild species of Canis. However, the recent invasion of this region by the Coyote now poses a significant threat to the continued existence of primitive dog populations in these areas.¹⁹ Pedigreed captive populations of primitive dogs can provide a temporary protection for the unique genomes of these animals.^g However, even when based on documented wild-caught founders, such continued management under conditions of captive breeding cannot be expected to maintain those traits which set these animals apart from all other domestic dogs.
   Concerns for the conservation of these and other forms of primitive feral/pariah canines are now becoming shared worldwide. However threats from the loss of isolated habitat, the introduction of domestic dog diseases, and the potential for genetic contamination continue to increase in those regions of the world that still offer a refuge to these unique animals.^5,8 Only concerted public awareness and concern, coupled when necessary with appropriate legislative protective measures will assure the continued existence of these unique opportunities to understand the early development of the human-canine bond.

  
+ Support for manuscript preparation was provided in part by Financial Assistance Award Number DE-FC09-96SR 1 8546 from the U.S. Department of Energy and the University of Georgia. back
  
A From the Savannah River Ecology Laboratory back
  
B From the Department of Zoology and Wildlife Science, Auburn University, Auburn, AL 36849-5414 USA. back

Footnotes
  
a A detailed description of the generalized southern pariah external body phenotype is represented by the breed standard of the Carolina Dog, as published by the American Rare Breed Association (The Rarity 1993; 3:18-19.). back
  
b The proposal to place Cuon within the genus Canis has been made by: Anderbjorn A, Kleist T. A phyogenetic classification of the family Canidae. Abstract, 6th International Theriological Congress, University of New South Wales, Sydney, Australia, 1993. back
  
c Gergits WF, Brisbin IL, Jr. A taxonomic reassessments of the New Guinea Singing Dog. Abstract 181, 75th Annual Meeting, American Society of Mammalogists, University of Vermont, Burlington, VT, 1995. back
  
d Clutton-Brock J, Department of Zoology, The Natural History Museum, London, England: Personal communication, 1995. We are grateful to Dr. Clutton-Brock for this insight. back
  
e Brisbin IL, Jr., Risch TS. Studies of an American Dingo: digging behavior of free-ranging Canis lupus familiaris from the southeastern United States. Abstract 167, 75th Annual Meeting, American Society of Mammalogists, University of Vermont, Burlington, VT, 1995. back
  
f Data for Figure 2 are taken from: Brisbin IL, Jr., Morphological and behavioral characterizations of free-ranging Canis familiaris inhabiting an area of restricted public access. Abstract 84, 68th Annual Meeting, American Society of Mammalogists, Clemson University, Clemson, SC, 1988. back
  
g Both the New Guinea Singing Dog and the Carolina Dog are now recognized breeds which are registered with both the American Rare Breed Association and the United Kennel Club. Stud-book registries for wild-caught founder stock and their captive-reared progeny are maintained for both breeds by the latter organization and the International Species Inventory System. back

Jane Gunnell
Banbury Cross Farm
Phone: (803) 215-6166 or
            (803) 649-0045
jane@gforcecable.com