Recording the human presence in the central Andes
Most civilizations in the world probably had their kick-start subsequent to the bizarre accident of cultivating cereals - wheat, rice, maize or millet. In the Andes, a major centre for the origin of numerous cultivated plants (Vavilov 1926; 1992) including potatoes, it may instead have been a diverse array of tubers and pseudo-cereals that led to an early population expansion. It was not until much later that maize arrived from Central America to supplement and probably to replace many of these indigenous crops.
The central Andes was the centre of the Inca empire which stretched from southern Colombia to central Chile, AD 1290-1537; the endpoint of a whole series of civilizations, chiefdoms and kingdoms which have left archaeological remains all through the Andes (Keatinge 1988; Bruhns 1994). There were no historical records until the arrival in 1532 of the Spanish who caused the collapse of the Inca civilization, including the loss of sophisticated farming systems and the introduction of Eurasian domesticated animals. Even in the historical period from 1532 to the present, reliable documentation is largely missing prior to 1850. When only archaeological remains are studied for human impact on the landscape, the record is invariably disjunct in time and space, and frequently compounded by the problem of dating events precisely.
Continuous records are needed to document these changes accurately in this area across this time-interval. Well-dated sequences obtainable from lakes within major areas of archaeological importance can provide continuous records of vegetation and climatic changes, and anthropogenic impact, such as at Marcacocha.
The Marcacocha sequence
The Patacancha Valley (Peruvian central Andes: 13 [degrees] 13[minutes]S, 72 [degrees] 12[minutes]W, altitude 3300 m), containing the infilled lake at Marcacocha, is located within the central area of the Pan-Andean civilizations (Keatinge 1988; Bruhns 1994). Reputedly once a sacred lake according to the local people, its potential palaeo-ecological significance was recognized by the Cusichaca Trust, directed by Ann Kendall; preliminary analysis established that it contained 8 m of sediments (Tim Holden pers. comm. 1989). The valley is a minor tributary of the Urubamba river, extending from the confluence at Ollantaytambo (2800 m), once a major Inca city, up to a pass at 4500 m, c. 12 km to the north [ILLUSTRATION FOR FIGURES 1 AND 2 OMITTED]. It is today heavily cultivated and almost completely deforested. Wooded remnants are mostly secondary, introduced types, even growing on Inca terrace systems, but fragments of native woodland occur in steep-sided ravines. The infilled lake of Marcacocha, only 40 m in diameter and located 100 m east of the Patacancha river, remains wet in the dry season and it probably formed as a nivation hollow at the end of the last Ice Age.
Previous investigations of the Holocene record of the central Andes, providing a regional scheme for vegetational change, unfortunately have mostly come from above the forest limit at altitudes greater than 4000 m, in Peru at Lagunas Junin, Tuctua, Pomacocha, Jeronimo (Hansen et al. 1984; Hansen et al. 1994), and in Bolivia (Graf 1981). The tree-line at this altitude is c. 3900 m (Ellenberg 1979), but at Marcacocha the forest has been largely replaced by scrubland communities.
Only Laguna Paca (Hansen et al. 1994) at 3600 m altitude offers a broad, but valuable comparison with Marcacocha (altitude 3300 m). Laguna Paca lies near the town of Jauja, close to the archaeological site of Pancan in the valley of the Mantaro River in the province of Junin, with cultivation up to the lake-shore. The Mantaro River valley and Junin plain have been a focal area for studies investigating the early domestication of plants and have provided abundant ethnobotanical remains (e.g. Pearsall 1980; 1988; Hastorf 1988). …