Darwin, Hamilton Smith and the Lost Domestications of South America

Theories of canine domestication are beginning to take into account partial domestications of wolves, such as the 33,000 year old skeleton from a Siberian cave described by Ovodov et al. (2010), and the skull from a cave in Belgium described by Germonpre et al. (2009). Ovodov et al. argue:

�Since dog domestication almost certainly occurred multiple times without direct human selection, we suggest that it must have occasionally failed. That is, the particular set of ecological conditions associated with human settlement and hunting practices that were necessary to initiate the domestication process must have, on some occasions, existed only long enough to produce a few modified wolves (i.e., incipient dogs) with short-lived lineages.�

These researchers suggest that �ecological changes caused by progressive cooling almost certainly caused social and settlement pattern changes severe enough to have disrupted the domestication process and prevented evolution of fully domesticated dogs.� They see their results as supporting a multiple-domestication-events hypothesis for Canis familiaris.

That domestication could occur within the canid family, the Canidae, but outside the genus Canis, was demonstrated by the work of of Belyaev (1979) and his students, particularly Trut (1999), who artificially domesticated silver foxes (Vulpes vulpes), through selective breeding over 30 to 35 generations. The experiment, which continues, left open the question of whether foxes could be domesticated outside of such an experimental environment.

In the 1830s, Charles Hamilton Smith observed canids in South American that were living with indigenous peoples and performing some domesticated functions for them. In one instance, Smith suggested that some apparently different species might actually be tame and wild versions of a single species, raising the possibility that some physical differences could have begun to occur because of the domesticated state. Recent research on pointing gestures with a South American canid closely related to some of the species observed by Smith may also indicate that these lost domestications involved more behavioral change than has previously been recognized.

South American Canids Described by Naturalists in the 19th Century

Charles Hamilton Smith, after discussing the maned aguara (labeled Chrysocyon jubatus by Smith but now generally called Chrysocyon brachyurus) the largest of the South American canids he had encountered, makes some general remarks regarding the animals he will discuss in the following chapters:

�South America, when first discovered by the Spaniards, was possessed of canines absolutely indigenous, some universally wild and others liable to be partially reclaimed; all more nocturnal than the former dogs and less so than the true foxes.�

The first plate is Smith's depiction of the maned aguara. (Double click to enlarge image.) Smith suggests that the name aguara (sometimes transliterated as warrah), attached to so many of these species, may come from the sound they make, �a loud and repeated drawling cry, sounding like a-gou-a-a-a, which is heard to a considerable distance.� (Smith's drawing does not show the ratio of leg length to torso size apparent in pictures of maned wolves, suggesting they may not have been seen often at close range.)

Genetics research, as will be discussed later, has recently indicated that some of the South American canids had a North American origin, but the biological significance of the formation of the isthmus of Panama, the Great American Interchange, was not yet understood. Smith says the maned aguara, the first plate included here from Dogs: Canidae or Genus Canis of Authors, was wild, but records that he had seen various levels of domestication in the other canids:

�Several can be sufficiently tamed to accompany their masters to hunt in the forest, without however being able to undergo much fatigue; for, when they find the sport not to their liking, they return home to await the return of the sportsmen. In domesticity they are excessive thieves, and go to prowl in the forest. There is a particular and characteristic instinct about them to steal and secrete objects that attract their attention, without being excited by any well ascertained motive. All subsist upon the usual food of the wild canines, but with the addition that they eat also fish, crabs, limpets, lizards, toads, serpents, and insects. They are in general silent and often dumb animals; the cry of some is seldom and but faintly heard in the night, and in domestication often learn a kind of barking. None appear to be gregarious, but several are occasionally encountered in families. Although in company with man, the domesticated will eagerly join in the chase of the jaguar, we have never heard that they are in the same state of hostility towards feline as are their congeners in Asia and Africa.�

Smith then states that in the 35 years before completion of his great study of dogs, the natives of South American had shown a distinct preference for European dogs, so that it was by 1839 difficult to find any indigenous dogs associated with them �even in the more remote parts of the interior.�

Hoary Aguara Dog/Fox

The hoary aguara dog was labeled Dusicyon canescens by Smith, but is now generally put in a more fox-like genus and classified as Lycalopex vetulus or Pseudalopex vetulus. He had seen the animal domesticated among the Indians of Brazil, who told him that it remained wild in some areas. Smith described the animal in detail, explaining that he had looked in the mouths of some of the domesticated animals, seeing a black palate as was the edge of the lower lip.

Dalponte (2009) reports the hoary fox is widespread in Brazil and may perhaps be found in Bolivia. It is not threatened with extinction. Courtenay et al. (2006) studied the behavior of the hoary fox, finding that a breeding pair reared a single litter of five offspring per year. The pair associated with no other adults and the offspring dispersed at about ten months of age, the young dogs establishing home ranges adjacent to natal territory. The male adult �acted as chaperone to the foraging offspring, mostly in the absences of the breeding female.� Their diet is mostly made up of termites and dung beetles.

Aguara Dog of the Woods/Surinam Aguara Dog

Smith refers to some Indians having domesticated the aguara dog of the woods, to the left, which he classified as Dusicyon sylvestris, the same classification he places under the heading of the Crabodage, or Surinam Aguara Dog, making a curious reference to �the same species in the wild and domestic states.� He mentions one semi-domesticated specimen obtained from the Indians by the French naturalist, George-Louis Leclerc, Comte de Buffon (1707-1788). Smith gives no details as to the nature of the domestication in his two entries regarding these animals.

Guaraxa Aguara Fox

Smith refers to an �almost insensible degree� of distinction between the Aguara Dogs and the Aguara Foxes, the latter of which are �more completely vulpine, having tails with brushes even larger and longer than those of true foxes.� He says that they �can be domesticated, and it is believed will form cross breeds with Dusicyon and Canis familiaris.� The Guaraxa Aguara Fox, which he labels Cerdocyon guaraxa, �barks, lives on terms without restrain when brought up in the house,� and can be used in hunting �though with but indifferent docility.�

�It is to this species in particular we allude when remarking upon the singular propensity manifested by them to steal and secrete particular objects: bridles and pocket-handkerchiefs have been carried off in this manner, and subsequently found in bushes at some distance.� Smith saw them often chained because of the tendency to eat chickens.

Crabodago Aguara Fox

Smith labels the crab-eating fox Cerdocyon Azarae, but acknowledges the classification, Canis Azarae, given by Prince Maximilian of Wied. The consensus classification at present appears to be Cerdocyon thous azarae. He notes that this type was �represented in the Zoology of the Beagle.� Part of the description of the Guaraxa Aguara Fox is placed under the description of this animal, and may apply to its partially domesticated state as well.

Slater et al. (2009) find that the crab-eating fox and the hoary fox (Lycalopex vetulus) are separated by perhaps a million years, making interbreeding quite possible. They also estimate, however, that the South American canids are separated from the Canis genus by over 10 million years. All of the South American species except the maned wolf, the bush dog, and the Falkland Islands wolf are depicted by this team as having common ancestry about 5 million years ago.

Schwartz (1997) notes a report on interbreeding of a South American canid with domestic dogs, though the species of the canid is not certain:

�The Makusi of Guiana had domestic dogs but would also adopt a wild canid called a maikong. This �wild dog� probably was a crab-eating or pampas fox. A tame maikong was the most treasured possession of the Makusi. It was fed cooked flesh, fish, and ripe plantains. The Makusi said they liked to breed the maikong with their domestic dogs because the offspring make the best hunting dog. In 1925 a female pampas fox reportedly gave birth to two litters fathered by a fox terrier hybrid, but such an event must have been extremely rare. South American canids that have been tested have a different karyotype (the diploid number of chromosomes) from dogs and other members of the genus Canis, and the offspring of such crosses would usually be sterile.�

Wang and Tedford (2008) state that the raccoon dog (Nyctereutes) of eastern Asia appears to be closely related to the crab-eating fox.

The Mystery of the Falkland Islands Wolf

Darwin encountered the Falkland Islands wolf on the Beagle in 1833 and 1834:

�The only quadruped native to the island is a large wolf-like fox (Canis antarcticus), which is common to both East and West Falkland. I have no doubt it is a peculiar species, and confined to this archipelago; because many sealers, Gauchos, and Indians, who have visited these islands, all maintain that no such animal is found in any part of South America�. As far as I am aware, there is no other instance in any part of the world, of so small a mass of broken land, distant from a continent, possessing so large an aboriginal quadruped peculiar to itself.�

He noted the boldness of the animal, which would enter a tent and pull meat from beneath the head of a sleeping seaman. The Gauchos had taken to killing them �by holding out a piece of meat in one hand, and in the other a knife ready to stick them.� He noted the rapid decline in their numbers, and suspected, correctly, that they would soon be extinct. They were extinct by about 1876. Decades after visiting the island Darwin still remembered the innocence of the animal (1875, vol. 1):

�It deserves notice, as bearing on other animals as on the dog, that at an extremely ancient period, when man first entered any country, the animals living there would have felt no instinctive or inherited fear of him, and would consequently have been tamed far more easily than at present. For instance, when the Falkland Islands were first visited by man, the large wolf-like dog (Canis antarcticus) fearlessly came to meet Byron�s sailors, who, mistaking this ignorant curiosity for ferocity, ran into the water to avoid them.�

Smith called the animal the Falkland Islands Aguara Dog, and thought it similar to a coyote (Lysiscus cagottis). He related earlier accounts that �the Falkland Island wolf had originally been set on shore there by the Spaniards, with a view to prevent foreign nations finding fresh provisions at the anchorages: the information stated further, that the wolves had nearly destroyed an indigenous fox, and taken possession of its burrows.� Although Smith questioned that the Spaniards were responsible for the presence of the animals, he believed that an indigenous fox could be found on the western island. He had mistaken young for a separate species. Smith cited Louis-Antoine, Comte de Bourganville, who settled Port St. Louis in 1763, as having seen the animal burrowing along the sea and feeding on birds, emitting a feeble bark.

Juliet Clutton-Brock (1977) believed that human agency would have to explain the presence of the unique wolf on the Falklands:

�The shortest distance between the mainland and the Falkland Islands is too great (approximately 480 kilometers) to imagine the successful dispersal of this animal to the islands unless taken there by man. Nor is it likely that a single mammalian species, and a large carnivore at that, survived on the islands throughout the Pleistocene when periods of periglacial conditions prevailed.�

Clutton-Brock thus found the Falkland Islands Wolf something of an Atlantic analogue to the Australian dingo and she noted some of the same physical characteristics that suggested prior domestication: �white markings on the pelage, a wide muzzle with large somewhat compacted teeth in the premolar region, and expanded frontal sinuses.�

Slater et al., analyzing nuclear and mitochondrial DNA derived from one of the few remaining pelts, concluded that the closest remaining relative was the South American maned wolf (Chrysocyon brachurus, discussed above).

�Molecular dating analyses suggest that the Falklands wolf and several extant South American canid lineages likely evolved in North America, prior to the Great American Interchange. The Falklands wolf was the sole representative of a distinct South American canid lineage that survived the end-Pleistocene extinctions on an island refuge.�

This team of researchers noted the explanations given for the presence of the unique species on the Falklands, including that of Clutton-Brock and the idea that the animals may have been carried from the mainland by ice or logs. The team concluded that the animal �may have reached the islands by rafting or dispersing over glacial ice during the late Pleistocene and was probably able to survive into the recent past by subsisting on a rich diet of penguins, geese and pinnipeds.�

Human agency in the transportation of wolves to the Falklands has become unlikely, but the paleontological record is sparse and it is not impossible that further finds will make this theory viable again.

Additional Note:  The discussion of the Falkland Islands wolf needs to be revised in light of the findings of Austin et al. (2012), arguing that the separation of the species from the mainland most likely occurred about 16,000 years ago.  See the reference list below for additional information on this team's findings. 

Flux in the Taxonomy of South American Canids

In 1931, Cabrera described the confusion in classifying the often similar canids of South America, noting that some naturalists would put most of them in the Canis genus, while others provided for much more separation. Slater et al. suggest that at least �six exclusively South American canid lineages, including the Falklands wolf, originated prior to the formation of a Panamanian land bridge,� and that the �South American canids probably evolved from the fossil taxon Eucyon, which was widespread in North America during the late Miocene [c. 7- 5 million years before present]. The ultimate extinction of South American canid lineages in North America may have resulted from resource competition with Canis, which immigrated to the New World during the late Pliocene [5.3-2.6 million years before present].� (Additional note: see Austin et al. (2013) in sources below, for substantial revision of these findings based on DNA fossil evidence.)

The significance of this research can perhaps be visualized by superimposing some of the findings of Slater et al. on the relevant section of the taxonomic chart of Wang and Tedford. Wang and Tedford's connections are depicted as continuous lines, with Roman type on the side, while Slater et al.'s findings use dotted lines and italics on the side. The separation of the maned wolf and the Falklands wolf in terms of time is not significantly altered by Slater et al., though the connection between the crab-eating fox and the hoary fox is much more recent. Wang and Tedford indicate in one chart (Figure 7.1, p. 140) a connection between the Eucyon line and South American canids, but it appears that they do not see the connection as direct as described by Slater et al. In any case, changes to the current taxonomic arrangement are likely to result from the work of Professor Robert Wayne's group at UCLA and Professor Eduardo Eizirik's group in Brazil. Additional complexity is also added with Prevosti's 2010 phylogenetic study of extinct South American canids.

Pampas Foxes Follow Human Pointing Gestures

Three researchers from the Instituto de Investigaciones M�dicas in Buenos Aires looked at communication between humans and the Pampas fox (Lycalopex gymnocercus), similar to earlier pointing gesture research on dogs and wolves. The Pampas fox is, at least for the moment, in the same genus as the hoary fox. Seven foxes, which had been living at an experiment station for at least five years, learned that if they gazed at a human face, they received a food reward. The foxes looked longer at a human's face when a reward was offered, but less so when it was not. Once food was provided again, the animals looked longer again.

Four of the animals that had participated in the first phase were then tested for their ability to find hidden food after receiving human cues. Two opaque containers were used for the choice test. Each had a piece of liver under a false bottom to control for odor cues. Proximal and distal pointing gestures were employed--i.e., gestures in which the arm closest to the correct container pointed, and gestures where the arm crossed the body to point to the correct container. In the former situation, all four foxes performed above chance, while in the latter, three out of four did so.

The researchers concluded that "foxes, an undomesticated species with low socialization to humans, were able to follow human gestures to find hidden food." Since some of the South American canids, at the time that Smith observed them, were living in semi-domesticated states, it seems likely that they would have learned to follow human gestures for a number of purposes.

Dogs Near Coastal Argentina by 1000 AD

Smith noted the preference of Indians for European dogs as a reason for the decline in domestication of the indigenous species, yet recent archeology has found dog burials near the coast of Argentina by about 1000 AD. Prates et al. (2010) note that �the presence of prehispanic dogs in South America has always been associated with complex societies (mainly in Peru and Ecuador) and not with egalitarian hunter-gatherers.� Finds in Peru and Ecuador mostly date from 3,500 years before the present onwards. Yet, this team notes that �some reporters� and tavelers� stories from the sixteenth and seventeenth centuries indicate that Patagonian hunter-gatherers already had dogs at the time of the first Hispanic-Indian contacts.�

The team described findings in the north of Patagonia about 100 miles from the Atlantic. The culture used stone tools and dogs were found in an area of continuous occupation. A second burial, about 170 miles away in the Pampas, was in a cemetery, where a �dog was buried intentionally with a 2- or 3-year-old boy, his right arm extended near the dog�s head, in a posture remarkably similar to the famous Natufian burial from Israel. Analysis of the skeletons at these sites indicates they were �somewhat larger than the common Mesoamerican dog,� closer to a Cocker spaniel in size �and similar to a spitz-like breed or a chow chow.�

Conclusions

Except for one curious reference in Hamilton Smith, it appears unlikely that any of the fox-like canids that achieved semi-domestication had separated from wild populations sufficiently to be morphologically or genetically distinguished. The fact that the team working with Pampas foxes found that some animals were too frightened of humans to be tested may indicate that the semi-domesticated canids had to be raised with humans to be put on leashes and handled by humans, and may indicate that this had been going on for generations by the time Smith arrived in South America. Unfortunately, the cessation of any domestication process with the arrival of dogs from either prehispanic or European cultures makes investigation difficult, and absent some exceptional archeology, probably impossible.

Nevertheless, even from Smith one gets the sense of something of a continuum of the fox-like wolves, or wolf-like foxes, that involved a number of species similar in appearance and which might, depending on the environment, be confused with each other. There were undoubtedly subspecies occupying different but sometimes overlapping habitats. Smith's speculations about interbreeding of similar species are likely to have had some basis in fact.

One also gets the sense that before the increase in European dogs, a wide range of indigenous cultures formed relationships involving various degrees of domestication with these animals. They accompanied the hunters, lived in their huts and may have eaten vermin. Their generally limited vocalization would have likely made them of little use as guards and their timid and curious natures may have made them useless in combat with other tribes. (As I noted in a prior blog, if any of the South American canids, including the dogs of the Incas, had had any defensive value, the conquistadors would have had a much harder time.) Still, some connections were formed, but then easily broken when the more devoted, more domesticated dogs arrived.

Do the human-canid relationships described by Smith say anything relevant to the domestication of dogs from wolves? Some of the South American canids do not seem to have had a strong fright response to humans that had to be overcome. Nor is there any evidence that the South American indigenous cultures saw the canids as competitors for the same food. Although mentioned in connection with hunts, the animals do not seem to have been nearly as helpful as hunting dogs, probably because large game was of no interest to them. Humans may have helped them expand their own ranges, however, and may have taken them to places where birds and crabs could be found, along with the large game that interested the humans. When living with humans, they could learn to bark, and may then have had some defensive value, though this seems to have been relatively minimal. The fact that the canids would be tethered to keep them from eating chickens suggests that despite certain disadvantages, they were still considered of value.

The canids had a place in the cultures closer to that of the terrier than the large hunting dogs that are often seen as fitting the needs of Eurasian hunter-gatherer societies. If these South American domestications are to be regarded as failed, as dead ends, it might be suggested that absent the introduction of Spanish and perhaps prehispanic dogs into the cultures involved, something like the artificial domestication of Belyaev�s foxes might have been possible. These domestications of the wolf-fox-like canids may also suggest that different cultures could have different domestication events, that no single cultural prototype explains how domestication can occur.

Sources:

1. Andrews, E.B. (1894). History of the United States, vol. 1. Scribner's, New York (at 111: "The Indians of the lower Mississippi Valley, when DeSoto came, had dogs, and also what the Spaniards called hogs, perhaps peccaries, but neither brute was of any breed now bred in the country. A certain kind of dogs were native also to the Juan Fernandez and the Falkland Islands.")
2. Austin, J.J., Soubrier, J., Prevosti, F.J., Prates, L., Trejo, V., Mena, F., and Cooper, A. (2013).  The Origins ofthe Enigmatic Falkland Islands Wolf.  Nature Communications, 4(1552). DOI: 10.1038/ncomms2570 ("Previous  studies used ancient DNA from museum specimens to suggest that the [Falkland Islands Wolf] diverged from its closes living relative, the South American maned wolf (Chrysocyon brachyurus) around [7 million years ago], and colonized the islands [c. 330,000 years ago] by unknown means.  Here we retrieve ancient DNA from subfossils of an extinct mainland relative, Dusicyon avus, and reveal the [Falkland Islands Wolf] lineage became isolated only [16,000 years ago, or between 8,000 and 31,000 years ago], during the last glacial phase. Submarine terraces, formed on the Argentine coastal shelf by low sea-stands during this period, suggest that the [Wolf] colonized via a narrow, shallow marine strait, potentially while it was frozen over."). 
3. Barrera, G., Jakovcevic, A., Mustaca, A., and Bentosela, M. (2012). Learning Interspecific Communicative Responses in Pampas Foxes. Behavioral Proceses, 89, 44-51.
4. Belyaev, D.K. (1979). Destabilizing Selection as a Factor in Domestication. Journal of Heredity, 70, 301-8.
5. Cabrera, A. (1931). On Some South American Canine Genera. Journal of Mammalogy, 12(1), 54-67.
6. Clutton-Brock, J. (1977). Man-Made Dogs. Science, 197(4311), 1340-1342.
7. Courtenay, O., Macdonald, D.W., Giolingham, S., Almeida, G., and Dias, R. First observations on South America�s Largely Insectovorous Canid: The Hoary Fox (Pseudalopex vetulus). Journal of Zoology, 268(1), 45-54 (January 2006).
8. Dalponte, J.C. (2009) Lycalopex vetulus (Carnivora: Canidae) Mammalian Species, 847, 1-7.
9. Darwin, C. (1839, 1909). The Voyage of the Beagle. Collier & Son. New York; (1868, 1875). The Variation of Animals and Plants under Domestication, Vols. I and II. John Murray, London.
10. Germonpre, M, Sablin, M.V., Stevens, R.E., Hedges, R.E.M., Hofreiter, M., et al. (2009). Fossil Dogs and Wolves from Paleolithic sites in Belgium, the Ukraine and Russia: Osteometry, Ancient DNA and Stable Isotopes. Journal of Archaeological Science, 36, 475-490. 
11. Osgood, Wilfred H. (1934).  The Genera and Subgenera of South American Canids.  Journal of Mammalogy, 15(1), 45-50.  
12. Ovodov, N.D., Crockford, S.J., Kuzmin, Y.V., Higham, T.F.G., Hodgins, G.W.L., and Plicht, J. (2011). A 33,000-Year-Old Incipient Dog from the Altai Mountains of Siberia: Evidence of the Earliest Domestication Disrupted by the Last Glacial Maximum. Plos One 6(7), e22821.
13. Prates, L., Prevosti, F.J., and Beron, M. (2010). First Records of Prehispanic Dogs in Southern South America (Pampa-Patagonia, Argentina). Current Anthropology, 51(2), 273-280.
14. Prevosti, F.J. (2010). Phylogeny of the Large Extinct South American Canids (Mammalia, Carnivora, Canidae) Using a "Total Evidence" Approach. Cladistics, 26,456-481.
15. Scheider, L., Grassmann, S., Kaminski, J., and Tomasello, M. (2011). Domestic Dogs Use Contextual Information and Tone of Voice when following a Human Pointing Gesture. Plos One, 6(7), e21676.1997)
16. Schwartz, M. (1997). A History of Dogs in the Early Americas. Yale University Press. New Haven (picture of Pampas fox on p. 5).
17. Slater, G.J., Thalmann, O., Leonard, J.A., Schwizer, R.M., Koepfli, K.-P., Pollinger, J.P., Rawlence, N.J., Austin, J.J., Cooper, A., and Wayne, R.K. (2009). Evolutionary History of the Falklands Wolf. Current Biology, 19(20), R937-8.
18. Smith, C.H. Dogs, vol. I (1839), in the Naturalists Library, Mammalia. Hengry G. Bohn, London.
19. Trut, L. (1999). Early Canid Domestication: The Farm Fox Experiment. American Scientist, 87, 160-169.
20. Udell, M.A.R., Dorey, N.R., and Wynne, D.L. (2010). What Did Domestication Do to Dogs? A New Account of Dogs� Sensitivity to Human Actions. Biological Reviews, 85(2), 327-346.
21. Wang, X. and Tedford, R.H. (2008). Dogs: Their Fossil Relatives and Evolutionary History. Columbia University Press, New York.

Thanks to Linda Scheider for directing me to the research of Barrera et al. and to Yva Momatiuk for the link to Tui de Roy's excellent webpage regarding the maned wolf. Thanks to Graham Slater for sending me additional sources. Thanks to Richard Hawkins, Eric Krieger, and Brian Duggan for helpful comments. I would appreciate an email of any suggestions for further investigation of this issue, at jensminger@msn.com.