ANALYSIS AND DISCUSSION


ANALYSIS


The sites were plotted using a GIS and their distribution was correlated to several environmental parameters which might have affected their differential placement: (1) proximity to major rivers; (2) geology; (3) vegetation cover; (4) watersheds; and (5) soil type. Of these factors, proximity to rivers, watersheds, and soil type proved to be the most interesting in helping to understand prehistoric settlement patterns in the area.

For the purposes of analysis, 'unknown' and 'historic' sites were disregarded since they do not provide us with any real insights into the prehistory of the region (table 1).

Phase Number of Sites
Non-ceramic 14
Early Ceramic 156
Late Ceramic 233
Generic Ceramic 15
Check Dams 14

Table 1. Number of sites in the CARP dataset
assigned to each phase (check dams falling
within the ceramic period).


While little can securely be inferred from the quantity of Non-ceramic sites, the patterns for the Early and Late Ceramic sites show a clear concentration away from major rivers (table 2), thus suggesting that, following the adoption of agriculture, people would have moved away from the unstable area located near major rivers, favoring instead less ecologically diverse, but more stable, highlands (see also map 1).

Phase 0-0.5 km 0.5-1 km 1-3 km >3.0 km
Non Ceramic 29% 29% 29% 14%
Early Ceramic 15% 12% 54% 19%
Late Ceramic 15% 14% 54% 16%
Generic Ceramic 0% 7% 47% 47%
Check Dams 14% 0% 36% 50%

Table 2. Percent of sites in each chronological phase located within
buffers of increasing distance from major water courses.


The fact that there are very few Early or Late Ceramic sites between 0.5 and 1.0 km further suggests that the shift in land-use patterns that accompanied the adoption of agriculture and ceramic technology was a rapid rather than gradual process. This is also suggested by the distribution of sites identified as simply "Ceramic."

Map 1: Sites , major rivers, and 0.5, 1.0, and 3.0 km buffers.

Most of the sites in this survey (68.7%) are located in the Chevelon Canyon watershed (table 3 and map 2). This implies that the sites used in this study come from more or less directly comparable settings. Interestingly, all of the non-diagnostic Ceramic sites are located outside the Chevelon Canyon drainage, suggesting perhaps that Plog and colleagues might have been using a different sampling and/or recording strategy in that part of their survey.

Phase Chevelon Canyon Middle Little Colorado River Silver River
Non Ceramic 86% 0% 14%
Early Ceramic 99% 1% 1%
Late Ceramic 100% 0% 0%
Ceramic 7% 20% 73%
Checkdams 100% 0% 0%

Table 3. Percent of sites in each chronological phase located within Colorado Plateau watersheds

Map 2: Sites, major rivers, and river drainages.


In terms of their geological context, almost all of the sites under consideration here were associated with Permian geological formations (table 4). The 'outlying' Non-Ceramic and Generic Ceramic sites that occur on Cretaceous and Tertiary formations most probably represent those sites found in the Silver or Middle Little Colorado River watersheds. Likewise, those sites associated with a vegetation cover other than Pinyon-Juniper (which is by far the norm [table 5]), can also be tied to those sites that fall outside Chevelon Canyon (map 3).

Phase Triassic Permian Cretaceous Tertiary Quaternary
Non Ceramic 0% 86% 7% 7% 0%
Early Ceramic 1% 97% 0% 1% 1%
Late Ceramic 4% 96% 0% 0% 0%
Generic Ceramic 0% 13% 73% 13% 0%
Check Dams 0% 100% 0% 0% 0%

Table 4. Percent of sites in each chronological phase located on geological formations of varying age


Phase Cottonwood-Willow Communities Mixed Grass Communities Mixed Meadow Communities Pine Communities Pinyon-Juniper Communities Shrub-Grass Disclimax Communities
Non Ceramic 0% 7% 0% 21% 71% 0%
Early Ceramic 0% 8% 0% 5% 86% 1%
Late Ceramic 1% 6% 0% 10% 80% 4%
Ceramic 0% 0% 7% 73% 20% 0%
Checkdams 0% 0% 0% 0% 100% 0%
Table 5. Percent of sites in each chronological phase associated with various vegetation communities

Map 3: Sites, major rivers, and vegetation types

Lastly, there appears to be a gradual shift in the soil types on which sites were identified over time (table 6). In all periods, most sites are found on either Overgaard Elledge Telephone- (OET) or Winona Boysag-Rock Outcrop-type (WBRO) soils. However, from being located predominantly on WBRO soils during the Non-Ceramic period, the majority of them shift to OET settings by the Late Ceramic period. Interestingly, this pattern appears to be more gradual than that suggested by the relationship between Period and distance from major rivers. This is because Early Ceramic sites are to be found less frequently on WBR soils than Non-Ceramic sites while they are still more common in those contexts than Late Ceramic sites (see map 4).

Phase Moenkopie-Shalet-Tours Overgaard-Elledge-Telephone Showlow-Distrheff-Cibeque Tours-Navajo Winona-Boysag-Rock Outcrop
Non Ceramic 0% 29% 0% 0% 71%
Early Ceramic 2% 39% 0% 1% 57%
Late Ceramic 2% 54% 0% 2% 42%
Ceramic 0% 80% 7% 0% 13%
Checkdams 0% 71% 0% 0% 29%
Table 6. Percent of sites in each chronological phase located on soils of different types

Map 4: Sites, major rivers, and soil types.



INTERPRETATION


As mentioned above, the main factors that seem to have differentially influenced site placement before and following the adoption of agriculture and ceramic technology are:

  1. proximity to a major river;
  2. vegetation; and
  3. soil type.

Based on these features, two basic patterns emerge from this study:

Non-ceramic: within 1.0 km of a major river; predominantly pinyon-juniper vegetation; with some pines; and WBRO soils.

Ceramic: more than 1.0 km of a major river; predominantly pinyon-juniper vegetation; decrease in WBRO soil types in favor of OET ones.

The patterns highlighted in this study suggest that pre-agricultural (i.e., non-ceramic using) groups in northern Arizona settled preferentially in unstable landscapes near the bottom of canyons. From a geomorphological perspective, these areas represent the most active and ecologically diverse zones. On the Colorado Plateau, where most of the sites in this study are located, the runoff from spring snowmelt and summer monsoons is captured and concentrated within these narrow canyons and before heading northwards towards the Little Colorado River. Canyon bottoms are thus highly unstable and unpredictable environments that can change little for several years, only to be completely reconfigured over the course of a storm that only lasts a few hours.

Clearly, this kind of ecological instability is unfavorable to the investment of large amounts of time and resources in the construction of permanent structures and the clearing of fields, all of which risk unexpected destruction. For sedentary groups willing and able to invest considerable effort in the construction of houses and the clearing of fields, the geomorphologically stable highlands would provide an ideal setting. There, the abundant pinyon nuts would have constituted a more dependable subsistence base that could have supported a population in the process of adopting or developing an agricultural lifeway.

On the other hand, for groups of mobile hunter-gatherers, the wide and shifting range of animal and plant resources that can be found in the canyons makes them ideal locales for short-term occupation. In other words, “canyon landscapes encourage human inhabitants to be flexible opportunists, but discourage them from investing much in a particular place or depending too much on a particular resource” (Schmich et al. 2001).

IntroductionMethodsProblemsAbout me!