Huamachuco

-
ANALISYS OF BOTANICAL REMAINS

We tested our hypotheses about the contents of the different qollqa types in Huamachuco by examining micro and macrobotanical remains (Chiswell 1984, 1986). In addition to the remains of food products themselves, other types of botanical remains may indicate the storage of particular crops. In some qollqa at Huánuco Pampa, Morris (1967: 92) found evidence that tubers were stored in layers of ichu grass, which he refers to as pirwa. These bundles of grass and tubers were probably similar to the modem chipa (bundles of poles or bark and leaves) used to transport fruits to market (cf. Gonzalez Holguin 1952: 111). We reasoned that if pirwa storage were used in Huamachuco, the presence of higher than expected quantities of ichu or other grass and leaves in a qollqa might indicate tuber storage. A second possible indicator of stored crops in qollqa is phytoliths.

Archaeologists have only recently begun making extensive use of the analysis of biogenic silica. Numerous problems have been addressed with phytoliths, ranging from environmental reconstruction to analysis of diet (Pearsall 1982; Rovner 1983; Piperno 1987). A significant contribution has been the identification of specific cultivated crops such as maize (e.g., Pearsall 1979: 135-50). The potential for identifying particular plants was one of the primary reasons for selecting phytolith analysis for the Huamachuco study.

Phytoliths are microscopic silica bodies that form in the cells of a plant. Upon decomposition of the plant, they are deposited in the soil, where they are resistant to alteration. Phytoliths are less susceptible than pollen to wind transportation, and studies indicate that they move relatively little within undisturbed soils (e.g., Beavers and Stephens 1958). Thus, one would expect to find concentrations of the phytoliths of stored crops in qollqa soils.

Because of specific storage requirements, most Inka crops were probably stored separately, and a limited variety of plants should be represented in a single qollqa. The amount of plant material present should also be greater than in nonstorage contexts. Polo de Ondegardo (in Murra 1980: 133) states that annual additions were made to qollqa even though some materials deteriorated before use. The recovery of macrobotanical remains in other qollqa (e.g., D'Altroy and Hastorf 1984: 345-46, and Chapter 9) also suggests that some plants remained in storage long enough to contribute their phytoliths to the soil.

The first step of the analysis was to examine a variety of plants to assess their silica content. The quantity and type of phytoliths vary from one part of a plant to another, so it was necessary to evaluate the storable portion(s) of the plants studied. We identified a total of 35 plants or parts of plants as storage possibilities, including the edible portion of several crops, as well as nonfood plants that may also have been present in qollqa (e.g., ichu grasses and unidentified plants used in modern chipas in the Huamachuco market).

The first observation was simply to note the presence or absence of silica. Four samples-peanut, peanut shell, pumpkin seed, and potato flesh-had no significant silica whatsoever. Most of the remaining samples had nondiagnostic types of silica, but few short cells, the type of phytolith that research has found to be most usefully studied. However, seven samples-four varieties of maize cob, ichu grass, caña brava, and an unidentified species of grass-had short cells in abundance. In these samples, the number of each short cell form was counted.

Next, we extracted the phytoliths from 22 soil samples collected . at Cerro Mamorco and Cerro Cacañan. These samples included two "controls" taken near each group of qollqa and 20 archaeological samples from six structures. Each sample was analyzed in the same way as the plant samples-the short cell phytoliths in each sample were counted.

The resulting data were submitted to two kinds of comparisons. First, we compared all the soil samples from a qollqa with the corresponding control sample. This comparison identified two structures at Cerro Cacañan (Qollqa B and Building C) in which none of the soil samples could be distinguished from the control sample. Either the soil samples from these two structures were contaminated, or there was simply insufficient plant decomposition and phytolith accumulation in these structures to be detected.

We next compared the remaining soil samples with the plant samples to look for evidence of specific plants. The sample from the Qollqa A on Cerro Cacañan did not resemble any of the tested plant samples, but samples from all three of the Cerro Mamorco qollqa resembled the ichu grass sample.

The phytolith analysis thus suggests there was an unexpectedly large amount of ichu grass in the qollqa of Cerro Mamorco. Based on Morris's association of tubers with large amounts of ichu grass in Huanuco Pampa, we can suggest that tubers were being stored at Cerro Mamorco. It will be recalled that two of the Cerro Mamorco qollqa have subfloor channels that we suggested above were used to create the high humidity conditions necessary for tuber storage.

Macrobotanical remains recovered by flotation augment these observations. After we collected the small samples needed for phytolith extraction, the rest of each soil sample was floated in detergent and water, agitated by hand, and the organic material separated in filters. We then examined this material, both visually and under low-power magnification.

The samples from Cerro Cacañan produced very small amounts of fragmentary wood charcoal. The Cerro Mamorco samples yielded slightly more material, including carbonized fragments of two unidentified (presumably nonfood) seeds from Qollqa A, a few burned maize fragments from Qollqa C, and small amounts of fragmentary wood charcoal from all three qollqa. The best recovery of botanical material was from the Cerro Santa Barbara samples. A sample from Qollqa 3 produced large quantities of carbonized maize and wood charcoal. Although the very small flotation sample from Qollqa 4 failed to produce maize, the field notes indicate that carbonized maize was even more abundant there than in Qollqa 3. This supports the predicted association of maize storage with the raised floor qollqa of the Cerro Santa Barbara structure designed for low humidity storage, as suggested above.

Our predicted distribution of stored items in the Huamachuco qollqa seems to hold. The facilities on Cerro Santa Barbara that were thought to be devoted to dry storage produced evidence of maize. The facilities on Cerro Mamorco (Qollqa A and B) that were predicted to be intended for tuber storage produced evidence of ichu grass that may have been used to pack the tubers; Qollqa C had macrobotanical evidence for maize storage as well as ichu grass phytoliths and may have served to store both grains and tubers. Unfortunately, there. was less conclusive botanical evidence of the contents of the Cerro Cacañan qollqa, but the subfloor channels in these qollqa implies they, too, were used for tuber storage.

Retornar Avanzar