The strongest evidence of human activity during the Archaic in the Upper Colca valley segment of the survey comes from projectile point distributions.

Table 6-34. Diagnostic Projectile points Series 1-4 from Blocks 3, 6 and the valley portion of Block 5 identifed by ArchID number.

Projectile point counts shown in Table 6-34 include the Upper Valley areas of Block 3 and Block 6, and the counts also include projectile points from Kakapunku [a03-1000] that lay in the valley bottom segment of Block 5 on the perimeter of Block 3. The Upper Colca valley segment of the survey had relatively low numbers of Early, Middle, and Late Archaic Period occupation evidence as compared with other segments of the survey. As is evident from projectile point counts, only 26 of the 56 projectile points (46%) from this Upper Valley derive from the Archaic Foragers period, a 5700 year long time period.

Obsidian projectile points are well-represented throughout prehistory, although over half are Series 5 points from the Terminal Archaic and onwards. It is notable that obsidian is so well-represented in finished projectile point material types because evidence of production or maintenance in the form of obsidian flaking debris is far less common than chert in this area. Chert is found in the riverbeds in Block 3 and therefore it is the predominant local material. However, among projectile points in Block 3 obsidian predominates in all point styles except type 3D where obsidian is a close second. By count, obsidian is most frequent, but by weight obsidian points are smaller even if Series 5 point types are excluded.

(a) (b)

Figure 6-23. Projectile Point weights and lengths (when complete) by material type for Block 3, Block 6 and adjacent valley areas of Block 5. Series 5 points are excluded and chalcedony is included with chert.

The projectile points made from the immediately-available chert material type are larger than points of either obsidian or fine-grained volcanics (andesite). The difference in mean weights was assessed using the Kruskal-Wallis non-parametric test because the distributions were not normally distributed. The Kruskal-Wallis test showed that the observed difference in means between the three groups is significant (c²= 8.235, .01 > p> .05).