The ephemeral record: The role of opportunistic animal resources in the archaeology of Pampa and Patagonia

Significance

The consumption of opportunistic foods is part of the behavioral repertoire used by hominins for subsistence. The acquisition of opportunistic animal resources is sometimes associated with stressful conditions. But, there are other conditions under which they were selected as part of established foraging circuits in well-known land. Archaeological sites from Pampa and Patagonia show that opportunistic foods—particularly bogged guanacos (Lama guanicoe)—had a role in Holocene human adaptation that goes beyond stressful conditions. Even when many examples of the use of opportunistic animal resources are related with events of stress or initial colonization, the use of these resources should be considered a facultative strategy that can be used at any time under a variety of circumstances.

Abstract

The acquisition of opportunistic animal resources by hunter-gatherers—such as scavenged carcasses—is a well-known subsistence strategy. It is frequently mentioned in the context of the history of early human evolution, but not regularly considered among the strategies utilized by more recent foragers of the Southern Cone of South America. Historical and ethnographic information presented here suggests that the use of opportunistic animal resources was a strategy used under a number of conditions but only partially documented in the archaeological literature. We also present archaeological evidence from four sites—Guardia del Río, Paso Otero 1, Ponsonby, and Myren—in different Pampean and Patagonian settings where relevant guanaco (Lama guanicoe) bone assemblages were recovered. These sites record minimal anthropic involvement, basically cut marks on some guanaco bones accompanied by few stone tools, that we interpret as evidence for access and use of water logged immobilized or recently dead animals. Archaeological evidence for the use of these scavenging strategies is difficult to obtain at large archaeological sites that usually result from multiple occupations, since the distinction between the acquisition of actively pursued versus opportunistic animal resources is not straightforward. One conclusion from our review is that the best places to find and recognize this evidence are archaeological sites resulting from ephemeral occupations. The inclusion of these sites gives us access to crucial and rarely documented evidence for the long-term survival of hunter-gatherers.

During the 1834 expedition lead by Captain Robert Fitz Roy up the Santa Cruz River, Patagonia, the first well-documented entry to south mainland Patagonia, Charles Darwin was one of the hunters providing meat. Guanaco (Lama guanicoe) was the main prey, but it was not always easy to find or hunt. When the hunting failed, the expedition relied on half-rations of salt-meat. After a few days of hard work hauling against the rapid current of the river, a dead guanaco was found in the water, and “some of our party, hungrier or less squeamish than the rest, immediately proposed dividing and eating it; and hunger carried the day …” (1, p. 344). Darwin’s observations of the event were not included in his famous Journal of Researches describing the voyage of the Beagle (2), but were recorded in his Diary, “A Guanaco was found dead under water, but in a shallow place; the meat was quite fresh … after a few doubtful looks it was voted by the greater number better than salt meat, & was soon cut up & in the evening eat (sic)” (3, p. 235).

This is a historically interesting document about the importance of an opportunistic resource, but it is not an isolated example. Similar histories can be found among narratives of early explorers of Southern South America. One particularly poignant example is provided from the final days of Captain Alan Gardiner and his companions in 1851, during their efforts to find a suitable place for an Anglican Mission for the South American Missionary Society in the Beagle Channel, Tierra del Fuego. The story, by Gardiner himself in his diary found at his death site, is that after inspecting several places they camp at Hispaniard Harbor in the peninsula Mitre, where they finally starved to death. Gardiner informs that when hunger started, they were forced to eat unusual foods like mice and a fox, until they were restricted to consume half-eaten shags or fish found dead on the beach (4, pp. 58, 79).

These examples can be seen in the light of the “learning the landscape” mode of behaving (5) and also of what groups do under subsistence stress conditions. Effectively, both the Beagle and the South American Missionary Society expeditions were exploring lands unknown to Europeans, for which minimal information about its resources existed, and their strategies to recover resources failed. Regardless of the fact that in both cases the possibility of observing the behavior of others was a potential help, the Gardiner example shows that this was not enough to survive in unfamiliar land. The anthropological literature has many examples of people who had a hard time trying to adapt to unknown resources in spite of the possibility of imitating other people (i.e., ref. 6). Cultural factors become important in these cases, since some resources require special search and processing strategies. The situation would have been even more difficult for the first humans arriving to uncharted territory, as it likely happened with the first settlers of Ice Age America (5). However, exploration and subsistence stress are not the only circumstance under which opportunistic animal resources were utilized.

Opportunistic Animal Resources

The acquisition of most foods involve relatively complex steps derived from the availability and distribution of resources, some of which can be accessed at very low search and processing costs. One important characteristic of opportunistic resources is that they do not constitute a separate subset of resources, but are defined by their mode of acquisition. The exploitation of immobilized or dead animals is an opportunistic strategy that permits the relatively expedient access to resources normally acquired by more costly means. In this paper, we will concentrate our discussion on the acquisition of relatively large mammals found dead or naturally trapped and will only mention smaller resources, but it must be stressed that small-package resources like insects, fish, and others can also provide substantial amounts of food (7, 8).

The classic example of prey that can be acquired at low cost are animals that are also preyed by carnivores. They result in already opened and partially processed carcasses that can be actively contested by human foragers, though perhaps at some cost. In contrast, passive access to carcasses occurs when foragers have access to already abandoned carcasses, or wait for the carnivores to be satiated, obtaining free access to the remaining tissues (9). Animal carcasses can also result from other processes, like drowning while crossing rivers, or falling into natural traps, many times giving access to complete carcasses for exploitation (10). Some of these scenarios are available in a variety of environments, present foods that can be easily accessed by humans, and constitute attractive strategies.

Scavenging in particular is sometimes seen as a situational strategy used during the initial colonization of an area, when human explorers were learning about the resources and geography of new lands (5, 11). Basic discussions of Plio-Pleistocene adaptations were derived from Binford’s (12) assessment of the capabilities of early hominins and were followed by a wave of studies about the merits of hunting versus scavenging strategies (i.e., refs. 10 and 13). Among others Bordes and Thibault (14) and Gamble (15) offer examples of the potential role of freeze-dried carcasses in the colonization of glacial Europe. However, the distinction between different modes of meat and fat acquisition remains elusive because we lack unambiguous markers for scavenging.

Use of opportunistic animal resources is only partially documented for recent times in the Southern Cone. We are exploring this issue because its recognition will help us to complete our image of past human adaptations in the Pampas and Patagonia. Some of the few published examples deal with large extinct Pleistocene mammals like Megatherium or Mylodon, that weighs between one and three tons (16, 17). Certainly, given the risks implicated by hunting these megamammals, their exploitation can be better explained if they were found dead or immobilized (11, pp. 373–374). As Grayson and Meltzer (18, p. 178) explained in relation with some Late Pleistocene sites from North America, “it is fully possible that those animals were scavenged, not hunted, by the people who utilized their remains,” but it is a difficult distinction in the archaeological record.

My previous examples refer to people getting acquainted with new lands and resources. However, there is also an ample ethnographic record about the access of carcasses of immobilized or recently dead animals by people with generations of experience in the area; in other words, people with good knowledge of their land, that complemented their subsistence with opportunistic resources. We believe that this is also what the archaeological record shows and is the main subject of this contribution.

Ethnography

Numerous examples in the ethnographic record show the ubiquity of scavenging among modern foragers disputing ungulate carcasses with lions or other carnivores (19). The most detailed study resulted from 188 d of observations among the Hadza from Tanzania, a period during which 54 carcasses were acquired, 11 of which “were obtained by scavenging” (19, p. 358). Bunn and collaborators studied Hadza acquisition of carcasses for 66 d, during which 24 animals were hunted and five were scavenged (17%) (20). The ethnohistorical account of an Inuit group that relied in part on “restes de phoques abandonnés sur la glace par les ours’’ (seal remains abandoned by bears on the ice) during their long march between Baffin Land and NW Greenland is notable (21, p. 44). These examples show the relative importance of acquiring food from animals found dead as a complement to resources acquired by hunting (22). As explained by O’Connell et al. (19, p. 361), the “amount of scavenged animal tissue available to the Hadza would have been small indeed,” but it was dependable. What is clear from these examples is that it is a significant subset of the consumed resources.

However, there are also examples of acquisition of large amounts of opportunistic resources. Effectively, the exploitation of stranded whales is among the best documented cases in the world (23–27). Individual strandings of whales weighing from one to twenty tons are regularly recorded in Tierra del Fuego, where shallow bays with large intertidal differences like San Sebastián or Lomas facilitated the process (28). The ethnographic record shows that Selk’nam shamans in Tierra del Fuego, whose status was dependent on their success in predicting good news, used their knowledge about potential stranding places, sightings of jets from whale blowholes in the distance and the sounds of whales to predict strandings (29). In that way, shamans cemented their role within the society as providers of food (24).

I will present in some detail the case for Tierra del Fuego and the Southwestern archipelagos of South America (24). One important example is presented by Fitz Roy and Darwin, who transcribed an observation made by Captain William Low, who described a group of native maritime hunter-gatherers who were enduring a bad weather famine. One day some members of the group left camp, and returned 4 d later loaded with large pieces of “putrid whales-blubber with a hole in the middle through which they put their heads” (2, p. 203, also ref. 1, p. 195). It must be stated that decomposed meat is not necessarily inadequate for human consumption and was many times favored by people (27). Contrary to many interpretations of access to dead animals as a very marginal strategy producing only bits of food (12), individual stranded whales provided large quantities of meat and blubber (28). In fact, whale strandings produced a surplus of food that might also require storage (26). When mass strandings occurred, which are not unusual in Tierra del Fuego, the availability of food largely exceeded human needs. Ethnography shows that storage was usually located in the intertidal zone, with meat and blubber covered with stones and memorized for future use (24, 26, 30).

Both individual and mass strandings can provide sufficient food for foragers, enabling them to concentrate in other social activities, including rituals that involved large numbers of people, otherwise difficult to organize. This was the case of the Hain among the Selk’nam of Northern Tierra del Fuego, or the Chiexaus among the Yamana of the Beagle Channel, Southern Tierra del Fuego. In both cases long initiation rites were scheduled depending on the availability of adequate supplies of food (24). Both ceremonies use stranded whales when available, with the Chiexaus more dependent on them. Accordingly, the duration of the latter was prolonged until a beached whale had been consumed (31, p. 85).

There is good information about the use of whales during the 18th Century. The wreckage of the Spanish ship Purísima Concepción in 1765 at the Peninsula Mitre, Tierra del Fuego produced the oldest ethnographic description of foragers interacting with beached whales. On this occasion the castaways, who spend months on the coast, observed the stranding of a large whale and saw how the local people made use of its meat, blubber, and bones. At least two forager bands converged on the place and spent several weeks exploiting the carcass (32). This case shows how fast and well organized the circulation of information about whale strandings was, and how people from far away places responded to the appeal of large amounts of meat and blubber. Detailed testimonies preserved at the Real Academia de Historia in Madrid, Spain, describe how about 60 individuals cut pieces of decomposed whale meat and blubber, transported those parts to nearby places, and buried some pieces. They also saw how they broke the whale mandibles using wedges and stone hammers (32). It is of the highest interest to know that remains of the Purísima Concepción and probably of the castaways camp were recently identified by archaeologists working at Caleta Falsa in Peninsula Mitre (Fig. 1), opening the way to detailed archaeological examination of this case (33). In general, it is very difficult to demonstrate past consumption of whale meat and blubber, as only the use of their bones for tools was archaeologically well supported (26). Even so, given its potential to shed light on this issue, the search and study of the camp of the Purísima Concepción castaways is eagerly waited.

Fig. 1.

Main sites and geographic features mentioned in the text. References: 1. Paso Otero 2. García del Río; 3. Ponsonby; 4. Marazzi; 5. Myren; 6. Lanashuaia; 7. Mischihuen.

The importance of this particular opportunistic resource can be gauged by the fact that when dying whales—usually victims of attacks from Orcinus orca—were drifting into the inner channels of the Southern archipelagos, the Kawésqar helped facilitate the stranding process by moving the whale into a convenient beach at high tide (34).

The ethnographic evidence for the opportunistic use of beached whales accompanied by storage of meat and blubber is clearly related with planning, that serves not only as a security food, but also to organize the long-term use of the land. A strategy of acquisition of animals found dead can be organized, perhaps by individuals patrolling potential areas where dead or immobilized animals can be found, or attending to signs of their presence. In consequence, good knowledge of the environment and its resources provides more and better opportunities for acquiring fast and cheap food. All these cases inform that chances of access to opportunistic meat and fat were diverse in Pampa and Patagonia. On the basis of the number of guanacos killed by winter stress and the frequency of whale mass strandings on the coasts of Tierra del Fuego, these resources may have been relatively frequent (26, 35, 36) (SI Appendix).

High yields of food are not only restricted to large animals like whales. Strandings of schools of small fishes like sardines (Sprattus fuegensis) and hakes (Macruronus magellanicus) also produce massive amounts of food. Fish strandings were recorded both on the coasts of the Beagle Channel (8, pp. 46–47) and the Western archipelagos (37, p. 125). In each case, thousands of fish of diverse sizes were available to be collected—and disputed with birds and foxes—on the beaches. Similar high yields of food were obtained from collecting naturally occurring windfalls of hoppers or other insects in different parts of the world (7, 38).

The use of most of these sources of abundant opportunistic food was by peoples well acquainted with their lands and resources. Simply speaking, the availability of large packages of meat and blubber (whales) or large amounts of small resources (sardines, hakes, and insects) attracted foragers, even when they were not suffering dietary stress. But, these resources do not need to be abundant to be attractive. We will now turn to the Holocene archaeological record, by reviewing four cases from different regions of Southern South America. This will illustrate the potential importance of strategies of guanaco acquisition either during the initial exploration of an area or when people were already established.

Archaeological Results

Paso Otero 1 (PO1).

PO1 is a site located on the banks of the Quequén Grande River, in the Eastern grassland plains, Argentina (Fig. 1). Five guanaco bone piles were recovered in wet meadows of the alluvial plain, the result of at least two depositional events. An excavation of 22 m² produced over 3,500 guanaco bones, associated with few rodent bones, eight small flakes (<4 cm), and one bipolar tool (Table 1). Gutiérrez and Kaufmann (39) interpreted the bone assemblage (SI Appendix, Table S1) as transported by the river and quoted Darwin’s comments about thousands of cattle rushing into a Pampean river, and “being exhausted by hunger they were unable to crawl up the muddy banks, and thus were drowned … their bodies when putrid were seen floating down the stream” (2, p. 126).

Table 1.

Artifacts associated with bone assemblages (40–43)

	Paso Otero 1	García del Río	Ponsonby	Myren 2
Side-scrapers				5
Cores				5
Bipolar	1
Cutting tool				1
Bolas			1†	3
Hammers			5	1
Flakes	6	1*		14
Debitage, microflakes	2			4
Manuports				3
Bone tools			4
Total	9	1	6	36

^*Inserted in a guanaco sacral vertebra.

^†Fragment.

One subsample of PO1 was analyzed, producing a minimum number of 30 guanacos. The mortality profile corresponds to a live population with a slight overrepresentation of neonates. The deposit, in an area with shallow water, and good trapping conditions created by the presence of hydrophic vegetation, appears to have been the result of a large scale flooding event. Rapid burial and minimal carnivore activity were inferred, and a large number of complete at burial bones were recovered at a “channel structure … [with] a turbulent flow” (39, p. 170). The bones were recovered from a context of wet meadows with grasses and reeds, an interpretation confirmed by the invertebrate fauna, which include freshwater gastropods such as Biomphalaria peregrina and Littoridina perchapi (44). Abundant root marks from marsh plants reflect this environment. Water sorting is indicated by the list of bones recovered at the piles, basically those resistant to water transport (45). There is some variation though, with one of the piles—located at a place where bones and other particles moving with the current are trapped—probably representing a transported group of bones either by flotation or saltation (40, p. 209). Significantly, minimal evidence of access by carnivores and rodents is recorded in the bone piles, which is consistent with the glut of available food and rapid burial. Several dates on organic sediments from a series of three buried A-horizons in the Holocene Río Salado Member of the Luján Formation were produced. The horizon associated with the lower bone bed was dated ∼5882 to 5320 cal B.P., and the horizon associated with the uppermost bone bed was dated ∼3320 to 2935 BP cal B.P. (46) (SI Appendix, Table S2). Bad preservation of collagen made it difficult to date the bone piles; however, a 3356 to 3070 BP cal B.P. result was obtained on a guanaco tooth (SI Appendix, Table S3).

In addition to the few lithic artifacts found associated with bone piles, there was only minimal human modification identified on the bones. Cut marks were recorded on just 1% of the sample, and scarce helical fractures interpreted as bone elements from guanaco carcasses processed by hunter-gatherers “in the surrounding area … added to those of the animals that died naturally” (39, p. 172). I propose an alternative interpretation to explain some of these anthropic markers: That is, these were the waterlogged remains of recently died or dying guanacos found and utilized by foragers in sectors of the river. This interpretation does not necessarily replace that presented by Gutiérrez and Kaufmann (39), whose model of naturally deposited guanaco bones is well grounded. Indeed, natural guanaco bone accumulations exist at this basin, but we think that a very small subsample of cut-marked bones associated with few lithic tools can be the result of human scavenging. Given the intensive Late Holocene human occupation of the basin, the small number of cut-marked bones recovered at PO1, and its association with few tools, we hypothesize that the guanaco carcasses attracted human foragers moving along the river who use them for food. The presence of carcasses of drowned animals in the Pampas rivers should not have been unusual (2, p. 126). Similar uses of drowned animals are known from other regions, for example the consumption of drowned buffaloes recorded among Missouri tribes (47, pp. 496–497).

García del Río (GdR).

The bone assemblage found at GdR in the Southern Pampas, Argentina (Fig. 1) was originally considered paleontological. An assemblage of 87 Lama guanicoe bones was recovered within the floodplain of the middle valley of the Napostá Grande Creek (41) (Table 2 and SI Appendix, Table S4). The deposits in which the bones were found correspond to the Agua Blanca Sequence, Upper Section (Cristina Bayón, pers. comm.). These deposits resulted from waterlogging in the valley that created vegetated water bodies, as indicated by the presence of Holochilus sp., large vertebrates and pollen (48–51). The age of the Agua Blanca Sequence, Upper Section is 5994 to 5467 cal B.P. at Bajo San José (52) and 2343–2056 to 2095–1611 cal B.P. at other localities (51). One date of 2463 to 2140 cal B.P. made on a right unciform of Lama guanicoe recovered at GdR is consistent with the age of the Formation (SI Appendix, Table S3).

Table 2.

Number of bones recovered at bone assemblages (40–43)

	Number of identified specimens (NISP)	Minimal number of individual animals (MNI)
Ponsonby	988	12
Myren	536	9
Paso Otero 1	3,781	30
García del Río	87	3

The recovered bones correspond to a minimum of five individuals, whose population structure (three adult individuals, one juvenile/subadult, and one newborn) suggests the presence of a family group or part of a mixed group. About 62.1% of the bones were complete, and the 38.0% that were incomplete consist mainly of vertebrae of newborns. About 13.8% of the remains were found in articulated position. The recovered bones were interpreted as the result of drowned guanacos. Following the results of artificial flume experiments with guanaco bones (45), Tomassini and collaborators found a predominance of elements with low hydraulic-transport potential (41).

Evidence of human manipulation is scarce but compelling, and “includes two deep incisions with a linear outline and a V-shaped section … identified on a sacral vertebra, affecting the articular facets in the anterior and posterior portions of the element” (41, p. 573). Also, a flake was found inside an incision on the sacral vertebra, making it difficult to dispute human actions on the carcass (Table 1). Tomassini and collaborators (41, pp. 574–575) inferred that the dead animals attracted people who used them as food.

Ponsonby.

Ponsonby is an open air site located near the Fitz Roy channel in SW Patagonian, Chile (Fig. 1). It is a key site to understand the interaction between terrestrial and maritime foragers in the southwestern channels of South America. Forager visits to Ponsonby are recorded since the beginning of the Middle Holocene. Human occupations sampled at two loci showed that the lower layer D, dated between 8372–8035 BP and 7568–6855 cal B.P. (53, p. 367) (SI Appendix, Table S3) includes a few lithics—a fragment of a bola and five hammers (42, p. 216) –, and bone tools—an awl, one bipoint, one pendant on a pinniped canine and a possible whale bone wedge tinted with ocher (54, p. 226) (Table 1). Two peat deposits were discovered at the bottom of the archaeological sequence where an extensive accumulation of guanaco bones, “un véritable cimetière de guanacos” (a real cemetery of guanacos) (55, p. 21), was uncovered in one of the peat layers. A total of 988 guanaco bones were recovered (Table 2), most of them complete and in anatomical position (56, p. 417), representing at least 12 individuals, including four adults, three young and one perinatal. They are the probable result of natural deaths near a lagoon, and are dated 7969 to 7623 cal B.P. (SI Appendix, Table S3). The presence of cut marks on five bones—left coxal, left talus, lumbar vertebrae, right femur, right metacarpal (one of the few broken long bones)—suggests that those carcasses were scavenged (56, p. 417).

Ponsonby is dated several thousand years after the initial human arrival to South Patagonia, but probably represents the first human exploration of that remote western area. In this case, the evidence is consistent with indications of low acquaintance with the local environment. The basis for this inference is the minimal use of the most important local raw material—green obsidian—and the low exploitation of sea mammals that characterize human occupations in the Ponsonby area. Also, thumbnail endscrapers made on chalcedony were interpreted as curated tools transported from the eastern pampas where they are frequent (57, p. 199). Even when it is clear that foragers in Southern Patagonia were already well familiarized with guanacos, Ponsonby represents one of the first cases of human exploration of forested areas near the archipelagos. Not only is this an area where guanacos are less abundant, but also where they developed short migratory adaptations that contrasted with the more territorial mode from the steppes (58), suggesting that foragers would have needed some adjustments for the full utilization of the area. The Ponsonby assemblage probably represents opportunistic use of drowned guanacos during the Middle Holocene by people not fully acquainted with the region, whose archaeological signature markedly contrast with the local adaptation.

Myren 2.

A large guanaco bone assemblage was recovered in a muddy area at Myren 2, North Tierra del Fuego, Chile (43) (Fig. 1). This area was near an annual spring where cows and sheep were regularly trapped, leading the ranchers to fence the area. Published evidence indicates that the spring was active at least since the beginning of the Late Holocene. A sedge (Carex) peat bog was formed on top of a clay substratum, within which lens of complete and well-preserved guanaco bones were found. This can be interpreted as the result of animals bogged or drowned at the spring. A few bones of Ctenomys, Pseudalopex culpaeus, and cattle were also recovered. A total of 536 guanaco bones from a minimum of nine individuals were recovered, 92.4% of which were identifiable (Table 2), at least three of which are males (two adults, one juvenile) and two females (43). One subsample showed that 10.4% present carnivore marks and 31% present cut marks (SI Appendix, Table S5). Five cores, 10 tools, and debitage were laying on the clay level at the bottom of the peat layer (Table 1). The bone bed is dated ∼4572–4295 to 4350–3987 cal B.P. by three radiocarbon dates (SI Appendix, Table S3). Many of the bones with cut marks are on portions of the axial skeleton, which is not the usual pattern in Patagonia for kill or habitation sites, usually dominated by cut-marked bones from the appendicular skeleton (59, 60). Myren 2 is a place where parts of drowned guanacos were opportunistically exploited at the beginning of the Late Holocene.

Discussion

Most discussions about the acquisition of opportunistic resources during prehistoric or modern times are focused on animals killed by predators, but other mechanisms—like animal deaths by drowning, freezing or trapping—need to be considered. One difference is that these mechanisms offer access to complete carcasses (10). These and other alternatives of immobilized animals (36, 61, 62) are among the expected sources of opportunistic meat and fat in Southern South America. In fact, the ethnographic and archaeological examples presented thus far involved animals deposited either in shallow water or bogs, or beached on the ocean coast. Given the importance of guanacos in most of the presented cases, it must be mentioned that they are occasional drinkers (58), not dependent of surface water. What attracts guanacos to boggy areas is the availability of forage and, in Tierra del Fuego, shelter from the wind.

The attraction exerted by long-term natural preservation—weeks to months—of guanaco meat and fat in a bog or the bottom of a pond can be entertained on the basis of the archaeological cases presented here. Given the low level of exploitation of the carcasses, it is difficult to explain those bone assemblages as resulting from concerted efforts to drive guanacos. Instead, they can be best explained as the result of opportunistic exploitation of bogged or drowned guanacos.

Beyond the modern observations of cattle at Myren 2, or those of Darwin in the Pampas, there is information about the importance of animals becoming mired in sites of Late Pleistocene age (i.e., ref. 63) and recent times (64, 65). During the harsh 1877 winter, the famous Patagonian guide William H. Greenwood found that two of his three horses “had struck a boggy place and were lying down in helpless state, and half bogged” (64, p. 97), a recurrent situation in alluvial lowlands (Fabiana M. Martin, pers. comm.). In 1898 J. B. Hatcher observed “a full grown guanaco that had become mired in a spring and was unable to extricate itself” (65, p. 184). The reiterated deaths of hundreds of guanacos by winter stress, or the recurrent mass strandings of whales attest to the abundant availability of these opportunistic resources (SI Appendix). These and other observations help to understand the formation of bone assemblages at places like Ponsonby and Myren and speak to the availability of immobilized animals for foragers.

Opportunistic Resources, Health, and Security.

The consumption of animals found dead is usually considered dangerous for at least two reasons. The first is the danger of contesting carcasses if the predators that killed the animal are still present. It is true that Hadza women are able to displace leopards from their kills (19), but there are also cases where the attempt to do so injured people (9). Large human group size and waiting for carnivores to be satiated are among the most effective tactics to minimize risk (13). Indeed, the perception of risk when interacting with carnivores is culturally determined, and no doubt varies according with the degree of acquaintance with the local carnivore guild and the number of carnivores involved.

Disputing the meat of animals killed by hyper-carnivores—Panthera onca mesembrina, Smilodon sp., Arctotherium sp—during the Late Pleistocene human colonization of Patagonia, was probably dangerous (66, 67). In some cases, the behavior of these animals is unknown to us, making our analyses of the interaction unclear. Regardless, the strategy was less risky during the Holocene, when the carnivore guild was restricted to pumas (Puma concolor), small canids, and felids. Even though fatal encounters between humans and pumas are known, they are rare. They usually resulted from humans ignoring existing knowledge or common sense about how to behave in their presence (68). The low levels of prey consumption by pumas, who leave behind abundant meat for secondary consumers, are well known (61, 69, 70). Its importance for foragers was demonstrated by Greenwood when he was unsuccessfully trying to hunt during a harsh winter in 1877. He found a puma feasting on a recently killed deer (Hippocamelus bisulcus), scared away the puma, and “took the best of the meat he had abandoned” (64, p. 95). This is a classic pattern of interaction in which pumas are easily scared away.

The second potential risk involved possible health consequences when consuming decomposed food (13), in part resulting from contact with “bacteria from the saliva of the predator, from the digestive tracts of insects, and from the carcasses’ own gut” (22, p. 477). Whale meat and blubber are considered among the safest nonhunting foods to consume, because whales rarely result from carnivore kills. Nonetheless, there is a reference by Lothrop of people concentrated to feast on a dead whale that floated ashore at Lanashuaia in the Beagle Channel (Fig. 1), that “were poisoned by the meat” (34, p. 87; also ref. (24), p. 281). Initial excavations at Lanashuaia disputed this interpretation (71), but subsequent work indicated that the site in which they were working was older than Lothrop’s reference. Moreover, biomolecular analyses at Lanashuaia by Evans et al. (30) showed the presence of bones of several whale species, and found no evidence of human aggregation, concluding that “it is highly probable that all the bones were obtained from stranded whales or beached whale carcasses” (30, p. 9).

Anyway, most doubts about health concerns derived from ingesting animals found dead have been dispelled by Speth’s recent work showing how putrid meat and fat may be not only safe but desirable to eat, particularly—but not exclusively—in cold environments (27, 72). Particularly relevant in evaluating Lothrop’s reference is that Speth and Morin (72) show that references to botulism (Clostridium botulinum) are present only from the 1970s onward for coastal arctic and subarctic foragers, probably the result of changes introduced by Westerners in the way the foods were fermented.

Opportunistic Animal Resources and Climate.

Cold climates prolong the circumstances during which acquisition of certain opportunistic resources is possible. In winter, when the landscape is covered by ice and snow, a menu of raw and decomposed meat and fat also have the additional advantage of not requiring fuel, which is difficult to obtain in that season. Even partially frozen meat can be consumed sometimes (27, p. 63). Greenwood “supped on raw meat” in winter in the 19th Century, and informs how he was obliged for a couple of days to eat raw hawk meat, which is a clear advantage in areas or seasons where fuel is not abundant (64, pp. 95–96). But animals found dead in winter are not always advantageous, especially if the carcass is fully frozen and its consumption required access to fuel (not only to defrost the carcass, but to access it), which increased the costs of the food. This situation suggests that timing of access to a carcass is crucial for the exploitation of immobilized or recently dead animals in Fuego-Patagonia, and also explains why few individuals from mass winter deaths are exploited.

Strategies of acquisition of immobilized or recently dead animals were probably helpful during cold winters in general, but particularly important during colder intervals like the Little Ice Age (c. 1500 to 1850 AD) or the Antarctic Cold Reversal (ACR) [c. 13,900 to 13,580 cal B.P.] (36, 69). It must be noted that it was probably at the end of the ACR that humans arrived in the region.

Opportunistic Resources and Stress.

An association between use of opportunistic animal resources and stressful circumstances is well recognized (73), and this is apparently supported by some examples presented here (1, 4, 21) But this association hardly describes all the conditions under which foragers could have access to opportunistic animal resources. Independently of the stress factor, good knowledge of the environment and its resources simply provides better subsistence opportunities.

There is no a priori reason why guanacos recovered in bogs or mud, killed by winter stress, or fallen into natural traps were only used at times of subsistence stress. It is relevant to remember that seasonal fluctuations of prey resources, like those characteristic of northern and southern latitudes, implicate that the prey themselves were under stress in late winter and early spring, times “when animals may become fat depleted” (74, p. 151). Given foragers’ requirements of fats, expectations of intensive processing—i.e., bone marrow—must be associated with such stressful cases. No evidence of intensive processing can be observed in the archaeological cases presented here, where limited exploitation of the available resources is manifest. People acquainted with the local resources would have known where to look during the times of necessity—stressful or not –, and make useful decisions. To patrol the landscape looking for indications of recent carnivore kills, like the flight of raptors or the calls of carnivores, or to visit selected beaches of the Strait of Magellan, constitute basic tactics to access opportunistic foods. In all the reviewed cases, the main distinction is between the stop-and-go use of carcasses (Ponsonby, García del Río and Paso Otero 1), and the probably more systematic use of places like Myren 2. These are two very different tactics to achieve the same goal of acquiring meat, fat, and hide, and the main distinction is between opportunistic and planned use of resources. Planning not only involved situations of return to the same place. The ethnographic information from the southwestern archipelagos shows how beached whales led people to organize their subsistence, social and ritual schedules. Given that scavenging of whales provided the conditions for long-term initiation ceremonies, this utilization clearly falls outside the realm of stressful circumstances. Indeed, the distribution of food caches resulting from large fluxes of whale meat and blubber supports the effort to secure subsistence stability providing at the same time means to negotiate potential starvation crises and ways to organize human mobility in the maze of channels and islands.

Significance of the Archaeological Evidence.

The potential association with opportunistic resources is possibly not restricted to the four sites discussed here. Other potential archaeological cases exist, many involving guanacos or whales, but they will require more complete information. For example at Marchant, south of Lago Argentino, Santa Cruz, Argentina a cluster of side-scrapers was recovered within layers of peat, dated 2351 to 2141 cal B.P. (75, p. 72) (SI Appendix, Table S6), in a situation similar to Myren 2. Animals are frequently bogged in this extensive peat at the foot of the South Patagonia Ice Field (pers. observ.).

Relevant information from sites on the coasts of the Strait of Magellan like Marazzi 32 or Punta Catalina 3 (29) (SI Appendix, Table S6) shows that whale bones were transported as raw material for tools (26, 76). This could be the result of combing the beaches looking for whale mandibles, the most appropriate bone for tools. When animals were intercepted a few days or weeks after their stranding, meat and blubber were probably consumed, but there is no archaeological evidence of this use. The Beagle Channel presents some fascinating sites probably associated with whale strandings. A notable case is the recent component of Mischihuen I, dated between 954–573 and 1176–733 cal B.P. (SI Appendix, Table S6), where 31 barnacles (Coronula diadema), which are specific to humpback whales (Megaptera novaeangliae), were found. The presence of these cirripeds that probably reached the site attached to the skin of a whale, strongly suggests “the possibility that the settlement was related to a whale stranded” (77, pp. 131–132, also ref. 25) (Fig. 1).

We already emphasized that the large amount of meat and blubber provided by whales was associated with the behavior of caching and storing food. Shallow pits in the intertidal zone appear to have been extraordinarily important deposits of whale blubber during historical times, but there is no archaeological evidence in support of this inference. All that can be said for the moment is that some of the sites excavated near the Strait of Magellan or the Beagle Channel hold the potential of being related with stores of whale meat and blubber. Only by asking the questions involved in the construction and use of storing pits (27, pp. 50, 78), we will be able to recognize them in the future.

Conclusions

The archaeological evidence presented here suggests that opportunistic access to waterlogged resources was a plausible strategy for Holocene foragers. In spite of being a safer and cheaper strategy, acquisition of opportunistic meat, fat, and hide is not usually considered as a regular strategy for the Holocene in the Southern Cone, where it is regularly mentioned for the Late Pleistocene. It is not clear if the reason is that researchers, with the few exceptions presented here, had a negative perception of scavenging (79, p. 317), had not considered this possibility, or result from the difficulties for recognizing these strategies.

This is more noticeable given that, contrary to the common perception that opportunistic foods provide only marginal resources, whales, and sometimes even beached fish or drowned guanacos offer abundant sources of food. The importance of these riches is reflected, for example, in how the exploitation of stranded whales impacted the social life of foragers facilitating the completion of the ritual calendar. Also, storage related with a surplus of meat and fat could have been important in planning the annual subsistence. Opposite to what was sometimes assumed, there is no basis to consider the utility of these strategies as marginal. This is one of the reasons why these are not second-rate, but crucial strategies for the long-term survival of people.

It is only by considering opportunistic resources as valid subsistence alternatives that it is possible to be alert to the subtle indications of their use in the past. Some of the archaeological cases presented here were originally considered as examples of exploitation of opportunistic resources (43, 56), but were treated more as exceptions than as a regular mode of acquisition of meat, fat, and hide. However, it is possible that when their existence is normally accepted, new sites will be recognized. Natural traps are diverse, and so are opportunities for easy access to varied resources. For example, Pomi analyzed natural vertical cavities (“marmits”) in river terraces of the Pampa that operated as bone traps (63). His study of bone assemblages recovered at those traps suggested criteria to recognize accumulations of complete drowned animals or anatomical parts in the paleontological record that provide—together with sedimentological analyses—basic elements for discussion of their use since the end of the Pleistocene. Similarly, work at the Pali Aike Volcanic Field (62) identified a number of natural traps for animals not restricted to fluvial and peat areas, but also including lava tubes. Different environments present a diversity of potential natural traps for the immobilization of animals, creating easy and difficult to ignore access to resources. Both the ethnographic evidence involving different kinds of animals and the archaeological guanaco examples discussed here testify to their use, but they rarely were the target of archaeological research. Scavenging sites can be spatially extensive, as it happens with sites involving megafauna. However, the size of the occupied area is probably a result of the intensity of use of the resource. As shown here, with few exceptions Holocene sites are basically small sites used for a short time, where no specialized tool-kits were required. These sites are not only small, but present few artifacts and few cut-marked bones. In other words, these are places difficult to recognize as archaeological sites. Only by considering sites where only few bones display cut marks, it was possible for us to compile the short list of Pampean, Patagonian, and Fuegian cases presented here, including some in which the archaeological context was not obvious. The difficulty of discovering these sites is augmented because not many marks from meat stripping are expected from processing carcasses of animals found dead, since peeling off bones is relatively easy when the meat is decomposed and macerated (27, p. 63). This is especially true when the bones were collected from waterlogged deposits. Only in situations of redundant use like Myren 2, the archaeological signal will be relatively strong. One possible implication from our examples is that archaeologists will need to fully incorporate small assemblages with few obvious human traces on bones or few artifacts within search models. This is important not only because they represent human activities, but also because in some cases they constitute the only evidence for some behaviors.

Finally, foraging is a way of life that cannot ignore potential sources of food. Both the ethnographical and the archaeological records shows that opportunistic resources—particularly those derived from already dead or naturally trapped or immobilized animals—were used. We suggest that they were part of the latent strategies maintained by human groups, strategies to be used opportunistically or to be regularly used when other tactics were not feasible. The use of opportunistic resources should not be considered an obligate strategy restricted to events of stress, but a facultative strategy that can be used under a variety of circumstances.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.