This temporal period covers the calendar years 3300 BCE through AD400, incorporating a time that includes dramatic changes in the region from the Terminal Archaic through the Late Formative following the Titicaca Basin chronology. As was described in Chapter 3, during this time period the sweeping changes that occurred in the larger region are linked at the Chivay obsidian source through evidence of intensification of obsidian production, the establishment of regular llama caravan traffic linking far-flung parts of the south-central Andean highlands, and a dynamic political environment in the Titicaca Basin consumption zone that appears to have influenced behavior at the source.

In the Terminal Archaic, distinctive economic changes that were accompanied by social and political developments were brought about in a broad suite of transformations that appear to have been co-evolutionary. Briefly, these changes began with transitional or low-level food production (Smith 2001) in the form of animal husbandry and a greater reliance on seed-plants. These changes were also associated with greater evidence of sedentism (Aldenderfer 1998;Craig 2005) and these settlements were linked through expanding interregional networks that appear to have interfaced through regular llama caravan transport (Browman 1981;Dillehay and Nuñez 1988). By the end of the phase described here as "Early Agro-Pastoralists" a number of regional centers in the Lake Titicaca region emerged during the Late Formative (Stanish 2003: 137-164) that were controlled by a socio-political elite in a socially complex, ranked society. The influence of these powerful political and religious elite were most evident at the primate centers of Late Formative Tiwanaku and Pukara, centers that grew in influence through, some have argued, the harnessing of labor to produce food, build monuments, and control interregional exchange.

In this research, features used to differentiate the Early Agropastoralists period occupations from the preceding Archaic Foragers and from the later Late Prehispanic periods are comprised by evidence from artifactual data and settlement pattern data. With the beginning of the Early Agropastoralists period one can expect the reliance on herding in the Upper Colca to have changed the settlement pattern in the direction of areas that are suitable for herding, in particular reliable water sources and bofedales for alpacas. The Series 1-4 projectile points (except 4c and 4e) are diagnostic to the Archaic Foragers period and so components associated these projectile points are therefore pre-pastoralist by the definition used here. A ceramics style termed "Chiquero" is thought to be diagnostic to the Formative Period based on surface distributions in the main Colca valley (Wernke 2003), but in the course of the 2003 Upper Colca fieldwork, ceramics in the style similar to Chiquero persisted into Middle Horizon period strata in one test unit in Block 3, suggesting that this ceramic style persisted in use in the higher altitude regions of the Colca. Thus, the presence of sand-tempered, unburnished ceramics can be used to differentiate the Early Agropastoralists period from the preceding preceramic time period, but not the Middle Horizon end of the Early Agropastoralists period in the Upper Colca. The end of the Early Agropastoralists period with the onset of the Titicaca Basin Tiwanaku period (AD 400) is more difficult to differentiate because the pastoral economic basis of the periods in the Upper Colca are not significantly transformed, and therefore the settlement pattern may be largely unchanged.

Diagnostic ceramics of the Middle Horizon, Late Intermediate Period, and Late Horizon have been described comprehensively by Wernke (2003) and therefore the presence of these ceramics is one way to differentiate certain components as "Late Prehispanic".

Based on the site classifications used for the pre-pastoralist Archaic, the Early Agropastoralist sites in the Upper Colca are considered here in terms of (1) residential bases, (2) logistical camps, and (3) isolates. These site groupings have analogs in the settlement pattern of modern pastoralists, where residential bases consist in distributed estanciasof varying sizes with adjacent grazing areas where permanent or seasonal residence is established, and logistical camps or "herding posts" that consist of regular travel stops during travel between estancias and other settlements (Nielsen 2001: 193-196;Tomka 2001). These types correspond with "main residences" and "herding posts" described with archaeological correlates by Nielsen (2000: 478-482).

Table 6-47. Classifications for Early Agropastoralist components of sites

Criteria used for identifying pastoral camps were available in the course of this study (Figure 6-48). Unlike the Upper Colca area, some of the ethnoarchaeological data used for constructing these categories come from regions, like eastern slopes of the Andes in Bolivia (Nielsen 2000: 480-483), that had low site reoccupation and spatial redundancy due to an abundance of suitable pastoral camp areas. Ethnoarchaeological evidence suggests that short-term herding posts with little site structure but many reoccupation episodes will often contain many hearths with associated overlapping middens that result from groups moving the hearth facilities to avoid debris from earlier occupants (Nielsen 2000: 480-483).

Pastoral bases and herding posts are characterized by extensive but shallow refuse discard. Bases may contain greater density contrasts in refuse discard because of longer term occupation and site maintenance activity, and areas around rich resource patches are likely to contain "large and dense archaeological sites resulting from multiple, non-contemporaneous, partially overlapping, and perhaps functionally diverse occupations" (Nielsen 2000: 482-483). Short term caravan stops do not necessarily contain corrals, as camelids do not require corralling every night.

Nielsen (2000) observes that short term herding posts (logistical camps in the typology used here) and shelters appear to prioritize the following features, beginning with the well-being of the caravan animals:

1. Grazing opportunities.
2. Water for the herd.
3. Shelter from the wind.
4. Other human concerns such as hunting opportunities and proximity to the residences of exchange partners.

Herders will take advantage of existing structures, such as wind breaks and sometimes existing hearths, but they virtually never construct new features in their short term occupation of overnight camps (Nielsen 2000: 452).