Quebrada Jaguay: Early South American Maritime Adaptations

See allHide authors and affiliations

Science  18 Sep 1998:
Vol. 281, Issue 5384, pp. 1830-1832
DOI: 10.1126/science.281.5384.1830


Excavations at Quebrada Jaguay 280 (QJ-280) (16°30′S) in south coastal Peru demonstrated that Paleoindian-age people of the Terminal Pleistocene (about 11,100 to 10,000 carbon-14 years before the present or about 13,000 to 11,000 calibrated years before the present) in South America relied on marine resources while resident on the coast, which extends the South American record of maritime exploitation by a millennium. This site supports recent evidence that Paleoindian-age people had diverse subsistence systems. The presence of obsidian at QJ-280 shows that the inhabitants had contact with the adjacent Andean highlands during the Terminal Pleistocene.

Because rising sea level between 18,000 and 5000 years ago submerged extensive coastal plains, archaeologists have found little evidence of how the earliest people in South America (or elsewhere in the Americas) adapted to living along the shore. The available records have tended to show “the use of maritime resources (sea mammals, fish, and shellfish) beginning only after about 10,000 years ago . . . after the Paleoindian adaptation [was] already in decline” (1). Terminal Pleistocene dates from two sites showing a predominant reliance on marine foods, the Ring Site in southern Peru (2) and the Amotape Campsites in northern Peru (3), were obtained from marine shell and may be unreliable as indicators of a settlement age. These sites are preserved where the continental shelf is narrow and rising sea level did not substantially displace the shoreline (4). A third site, Quebrada Jaguay 280 (QJ-280) (Figs. 1 and 2), had yielded a single Terminal Pleistocene date on charcoal of 10,200 ± 140 14C years before the present (yr B.P.) (5). We have now obtained an additional 12 Terminal Pleistocene dates on charcoal, ranging from 11,105 ± 260 to 9850 ± 170 14C yr B.P. [circa 13,000 to 11,000 calibrated years B.P. (cal. yr B.P.)], along with eight dates defining two Early Holocene components (Table 1). Here we describe these ages and the QJ-280 site. Similar dates have also recently been recovered from another coastal site, Quebrada Tacahuay (6).

Table 1

Radiocarbon dates from QJ-280, by sector. All dates are on charcoal. EH II, Early Holocene component II; EH I, Early Holocene component I; TP, Terminal Pleistocene component. All dates were run by the Brock University Earth Sciences Radiocarbon Lab (BGS). Cal. yr B.P. were calculated with method A of (15), with a 1σ range and a 40-year Southern Hemisphere correction.

View this table:

Site QJ-280 lies on an alluvial terrace on the north bank of the Quebrada Jaguay (Jaguay Canyon), about 2 km from the modern shoreline at 40 m above sea level (masl) (Figs. 1 and 2). At circa 11,000 yr B.P., the site would have been about 7 to 8 km from the coast (7). The local environment today is typical of the western South American coastal desert; there is little rainfall except during some El Niño years. A seasonal stream flows for a few days to a few weeks during the austral summer in the quebrada adjacent to the site. Local vegetation includes a variety of species in the quebrada bottom and xerophytic (fog-based) lomas plants on the western slopes of the foothills between 200 and 1000 masl. The marine species found at the site are typical of the modern, warm-temperate, Humboldt Current–dominated regime, which suggests that Terminal Pleistocene and Early Holocene ocean conditions in this region were similar to those of today.

Figure 1

(left). Map of southern Peru showing site QJ-280 and other geographic features and archaeological sites mentioned in the text. In the inset: 1, Amotape Campsites; 2, Quebrada Jaguay; 3, Quebrada Tacahuay; 4, Ring Site.

Figure 2

(right). Plan of QJ-280 site.

Remains of the Terminal Pleistocene and Early Holocene occupations include abundant bones and shells of fish, crustaceans, and marine mollusks. The only terrestrial animal remains are a few bones of small rodents that probably represent animals that died accidentally on site (8). The Early Holocene part of the site contains a few specimens of knotted cordage that may be parts of fishnets. Most of the Terminal Pleistocene and Early Holocene fish bones are from a single genus (Sciaenae, or drums) and are from fish with a small modal standard length of 172 mm, calculated from 534 otoliths. Thus, the site's inhabitants seem to have had a netfishing strategy focused on drums (Table 2). The molluscan assemblage is also monospecific: Over 99% of remains are from the wedge clam Mesodesma donacium. This focus on two marine taxa suggests that the Terminal Pleistocene and Early Holocene inhabitants of the site employed a specialized maritime subsistence strategy while resident at QJ-280.

Table 2

Vertebrate faunal remains from QJ-280. NISP, number of identified specimens; asterisk indicates less than 0.1%.

View this table:

Evidence for plant use at QJ-280 is minimal, probably because of poor preservation of uncarbonized soft tissue rather than nonuse of plants by the inhabitants. Plant remains from the Terminal Pleistocene deposit consist only of burnt pieces of dicotyledonous stems and branches. In the Early Holocene deposits, plant remains also include three gourd rind fragments.

Lithic remains include chipping debris and a few broken or unfinished tools exhibiting bifacial workmanship, but no projectile points. Lithic raw materials in both the Terminal Pleistocene and Early Holocene deposits include some petrified wood and small pieces of obsidian. To identify the likely sources of obsidian, we analyzed 30 obsidian flakes using Instrumental Neutron Activation Analysis. Twenty-six of the flakes came from the Terminal Pleistocene component and one from the Early Holocene II component (three could not be assigned to a component). The trace element data indicate that all 30 specimens from QJ-280 probably came from the Alca source (Fig. 3) (9). At 2850 masl, Alca is the closest source to QJ-280 and can be reached by traveling up the Quebrada Jaguay to its head, crossing a series of passes to the Cotahuasi River, and then moving upstream 16 km—a total distance of 130 km (Fig. 1). Another source, near Chivay at 4900 masl, is less than 20 km further (10); the absence of Chivay obsidian at QJ-280 may indicate that this source was covered by a glacial readvance during the Younger Dryas (circa 11,000 to 10,00014C yr B.P.).

Figure 3

Plot of Mn versus Na content for QJ-280 obsidian artifacts, compared with 95% confidence ellipses indicating known southern Peruvian obsidian sources and chemical types whose geological sources are not yet identified; ppm, parts per million.

At the base of sector II and covered by levels dating between about 10,500 and 10,800 years B.P., a series of circular holes may be evidence of posts for a building. The Early Holocene II deposit in sector I contained part of a circular semisubterranean house, about 5 m in diameter, with a central hearth.

We have found 55 smaller sites in the immediate vicinity of QJ-280, and we dated samples from 17 of them. No date fell in the Terminal Pleistocene, but many sites have basal dates in the Early Holocene (Table 3); these sites contain Mesodesma clams, have bifacially worked stone, and tend to occur below 250 to 300 masl. Other sites contain a wider variety of mollusk species, have basalt grinding stones, and tend to occur above 250 masl; a date on one of these sites was 3895 ± 80 14C yr B.P.

Table 3

Radiocarbon dates from sites located near QJ-280. Cal. yr B.P. were calculated with method A of (15), with a 1σ range, using dataset 1 and a 40-year Southern Hemisphere correction for atmospheric samples and dataset 3 for marine samples (reservoir correction ΔR = 190 ± 40) (16). All dates were run by BGS except as noted.

View this table:

The evidence from QJ-280 demonstrates that its Paleoindian-age inhabitants relied on ocean resources while resident at the site. This conclusion supports other recent studies showing early diversification of subsistence beyond an economy based on the hunting of large game (11). In North America, comparable dates for maritime resource use are from Daisy Cave, on an island off the California coast (12). It has been suggested that some early inhabitants of the Americas migrated along the Pacific coasts of North and South America, separately from those who moved through the interior (13). As the earliest well-dated maritime occupation in South America, QJ-280 offers insight into that issue. We suggest, with caution, that the initial Terminal Pleistocene inhabitants of QJ-280 were transhumant foragers who spent part of their time in the highlands and part of the time on the coast exploiting marine resources—an idea that T. F. Lynch proposed almost 30 years ago, though he did not suspect such great antiquity (14). The absence of any other Terminal Pleistocene sites in the region, the location of QJ-280 on the banks of a quebrada offering an easy travel route to the highlands, the seasonal nature of local waterflow, and the presence of highland obsidian in the Terminal Pleistocene component all support this hypothesis. By the early Holocene, coastal occupation may have become year-round, as suggested by the near absence of obsidian, the more substantial structures, and the explosion of local sites dating to this time. Thus, building on the achievements of their Terminal Pleistocene predecessors, the Early Holocene inhabitants of southern Peru seem to have developed a fully maritime adaptation.


View Abstract

Navigate This Article