After 2000 B.C., there was a major change in the lifeways of Titicaca Basin peoples. Browman notes ``the adoption of new technologies such as ceramics, the development of new techniques in architecture, and the increasing reliance upon a wide range of domesticated plants'' ([Browman 1984]:119). We refer to this time of early time of agricultural, ceramic-using and occasionally village-dwelling peoples as the Early Formative Period. This cultural and economic watershed, which occurred locally around 1500 B.C., also marks the first definable occupation of the Taraco Peninsula.
There seems to have been no detectable occupation of the Taraco Peninsula during the long preceramic or Archaic period of Titicaca Basin prehistory. In the entire survey not a single aceramic debitage scatter nor a single Archaic style point was recovered. Although this contrasts sharply with reports of intensive Archaic period occupations in other parts of the Basin, it is consistent with the fact that Albarracín-Jordan and Mathews ([Albarracín-Jordan and Mathews 1990]: 51-53) found only two isolated Archaic points in their Tiwanaku Valley survey. Although a small number of Archaic period points were recovered from the excavations at Chiripa, these were highly worn and were encountered in Formative levels; they seem to have been collected and curated by Formative Period inhabitants. In sum, there is no evidence for a measurable human presence on the Taraco Peninsula (indeed, in the Tiwanaku heartland) during the Archaic period.
There are several reasons why this might be the case. First, recall that the spine of the Taraco Peninsula is formed by the Taraco Formation, a loose and poorly sorted deposit of fluvial origin. The hills of the peninsula are covered with nodules of quartzite and chert, and the entire area constituted an extensive quarry during the whole prehispanic period. That is to say that the entire peninsula is a low-density debitage scatter. This would have the effect of obscuring ephemeral Archaic period lithic scatters which might be identifiable in a different geological context. This cannot account for the extreme rarity of Archaic period projectile points, however.
Second, geologists have demonstrated ([Abbott et al. 1997a,Abbott et al. 1997b]) that the little lake (Lago Wiñaymarka) was completely dry prior to about 1500 B.C. In the Archaic period, then , the peninsula would have been considerably colder and less hospitable to agriculture than it is today, lacking as it would have the climatic amelioration effects of the lake. Also, the area of the lake now covered with water would have been a vast pampa or grassland. Since pastoralism seems to have been the primary focus of the economy, at least in the later or terminal Archaic, it could be that the local inhabitants lived on these pampas, perhaps near the small meandering rivers, as was the case in the Ilave valley. If this was so, then the remains of their settlements and camps would now be inundated and archaeologically invisible.
Although it is impossible to evaluate this archaeologically at present, it is possible to imagine what could have happened when the lake abruptly rose to near-modern levels around 1500 B.C. We may postulate a landscape populated by small pastoral groups practicing a very extensive form of tuber agro-pastoralism on the vast pampas of what is now Lake Wiñaymarka. The rise in lake level - which seems to have been relatively abrupt - would have 1) eliminated a vast expanse of grazing land, 2) made available rich lacustrine resources, in the form of fish and lake birds, and 3) created a climate more amenable to agricultural production, as a result of the thermal effects of the lake. All of these changes would have encouraged the development of a more sedentary fishing and agricultural economy by simultaneously lowering the region's carrying capacity for a primarily pastoral economy and raising it for a mixed lacustrine and agro-pastoral economy. Assuming that adjacent grazing areas were already populated - thus limiting the possibility of outmigration - this would have constituted a strong incentive for local populations to intensify their fishing and agricultural activities. Furthermore, since the Taraco Peninsula is a peninsula, pastoral groups inhabiting the entire surrounding pampa would have been forced onto what became a narrow spit of land with very poor grazing potential. This is the sort of situation we might imagine resulting in relatively dense populations (relative to the rest of the Titicaca Basin) and a marked economic shift away from pastoralism and toward agriculture and lacustrine fishing/collecting.
Testing of this model is, of course, impossible since sites of this terminal Archaic time period - if they do exist - are submerged. It is nevertheless compelling since it explains several otherwise anomalous archaeological facts. First is the lack of any evidence of an Archaic period occupation on the Taraco Peninsula. In the scenario outlined above virtually all such evidence would be inundated at the present time. Second, it explains the character of the Early Formative occupation of the peninsula, which, as we shall see shortly, does not correspond to what we might expect of an initial colonization of a region.