An Assessment of Prairie Mound Origin Theories |
|||
 |
|||
|
Soil-forming processes primarily affect the finer components of sediment, weathering and moving clay and fine silt grains down through the profile over time. Coarser components are thus better indicators of the nature of parent material. Comparison of the relative percentages of the clay-free fraction throughout the profiles (Fig. 3, Total % sand and silt) reveals a similar parent material for both mounds and the inter-mound area. The sand percentage, especially, is nearly the same in all cores at all depths analyzed, averaging 18%. Thus the mounds appear to be composed of the same parent material as the underlying sediment without significant modification. The gravel recovered was primarily composed of rounded pebbles and concretions, with a maximum diameter of about 5 mm. Comparing sand, silt, and clay percentages (Fig. 3, Total % <2 mm) reveals an accumulation of translocated clay, representing a textural B horizon, at about the same elevation across the study site. The horizontal nature of this soil horizon, in contrast to the mounded topography of the surface, implies that it is a relict feature. If the surface had been mounded when this horizon formed, this horizon would have contoured the surface elevations, as the A horizon does. The soil is therefore polygenetic, and the textural B horizon may be considered a paleosol, in the sense that it was formed under soil-forming conditions different from those existing today.
Figure 3. Particle size by depth. Elevations in arbitrary cm beneath datum. The A horizon is nearly twice as deep within the mounds as in the inter-mound area: 60 cm as opposed to 33 cm (Fig. 4). It is not apparent from this study whether the A horizon is deeper within the mounds simply because of the topography, or because the underlying clay-rich B horizon is impeding deeper development of overlying horizons between the mounds.
Figure 4. Soil profiles. Vertical exaggeration 8x. Arbitrary datum same as in Figure 3. The mounds roughly approximate sphere segments in shape, but both are skewed very slightly to one side, and in the same direction (Fog. 2). The skew is approximately 85 cm to the south or southwest (the north or northeast sides are steeper). This may be a result of mound genesis, or it may be an artifact of subsequent weathering. It is interesting to note also that the direction of skew is roughly equivalent to the lengthwise direction of floodplain in which the mounds lie, northeast to southwest. It is unclear at this time, however, whether the floodplain formed as a result of one large channel or several small streams similar to the ones currently situated in the valley (Margaret Guccione, personal communication). The direction of water flow may or may not have paralleled the valley. Further topographic study of mounds in the region would reveal whether they are uniform in orientation or not. If they are not uniform across the mound field, this would imply that the skew is not an artifact of uniform weathering along the same aspect. Guccione et al. (1991) trenched and mapped several prairie mounds in the same mound field as the preset study. Of the seven mounds topographically mapped, six are slightly elongated and generally oriented in a north-to-south direction. Because of modern disturbances, inferring the distribution and landscape position of the original mound field from aerial photographs is problematic. All of the mounds observed on the ground and from the aerial photographs, however, occur in the floodplain.
Discussion Soil horizons within and between the mounds demonstrate that the mounded topography is younger than the surface that existed at the time the textural B horizon formed. Whether the A horizon has reached equilibrium and will form no deeper is not apparent at this time. If it has, this implies that the textural B horizon is impeding its development in the inter-mound areas, or that the topography of mounds is the primary soil forming factor governing depth of soil development in some way. Studies of prairie mounds invoking the fossorial rodent hypothesis (Cox, 1990a; Cox and Allen, 1986; Dalquest and Scheffer, 1942) generally concentrate upon such mound features as distribution, presence or absence of pebbles and cobbles, presence or absence of active rodent burrows or krotovina, depth to a hard surface, and the distribution of past or present rodents. This study sheds little light on any of these features, except to confirm the presence of a hard surface (channel deposits or bedrock) at around 330 cm beneath the tops of the mounds, and 217 cm beneath the surface of the inter-mound area. No active or relict burrows were observed. The fossorial rodent hypothesis stipulates that burrowing rodents (often thought to be gophers) differentially move soil nearer the center of their burrowing territory, although this is disputed (Berg, 1990). If an impenetrable surface exists at a depth less than the maximum normal burrowing depth of the animals, successive generations are alleged to preferentially center their territories in areas with an accumulation of soil from previous inhabitants, due to the increased depth of soil available. In this view, prairie mounds are the result of many generations of rodents centering their burrowing territories around the same spot. A useful test of the rodent hypothesis might be accomplished through size analysis of larger clasts. Bocek (1986) found that intensive rodent activity affected the distribution of artifacts at archaeological sites in predictable ways. Burrowing activity of fossorial rodents displaces clasts that are larger than the burrows down through the soil, and mixes clasts which are smaller than the burrows throughout the active burrowing zone. Soil cores offer too small of a sample to adequately test the displacement of >2 mm clasts throughout a profile. A hand-excavated bulk density column through a prairie mound containing a significant number of larger clasts might reveal whether or not intense rodent burrowing activity had occurred. The clasts need not be archaeological; naturally occurring pebbles would be affected the same way by burrowing. An accumulation of large clasts at a level consistent with the deepest zone of burrowing activity may or may not signify intense occupation by rodents, but lack of such an accumulation would indicate that an area had not been intensely modified by rodents. Studies of prairie mounds testing the seismic shaking hypothesis (Berg, 1990; Cox, 1990b) generally concentrate upon such features as uniformity of mounds within a mound field, depth to a hard surface, and correlations of mound occurrences with areas of seismic activity. The seismic hypothesis requires a hard surface shallowly buried in order to transmit seismic energy of high enough force to shake the looser overlying sediment on top into mounds (Cox, 1990). No minimum depth for this surface has been determined. This study confirms the presence of a hard surface shallowly buried beneath the mounds, as bedrock or gravelly channel deposits at about 330 cm beneath the tops of the mounds, and at 217 cm beneath the surface of the inter-mound area. Mounds at Mima prairie, Washington, are underlain by basalt bedrock at about 180 cm beneath the inter-mound areas (Washburn, 1988). Presumably the magnitude of the seismic event and nature of the loose upper material would affect the minimum depth to a hard surface necessary to form mounds in this manner. Correlations of mounds to areas of seismic activity have been disputed (Berg 1990, Cox, 1990b). Prairie mounds are found in areas with little or no current seismic activity, such as the southeast coast of Texas. A few isolated pockets of prairie mounds also occur in mid-continent regions with little seismic activity, and they do not always occur in areas where seismic activity is high. Northwest Arkansas is currently a low risk seismic area. Paleo-seismicity is often invoked by proponents of this origin theory (Berg, 1990). Studies of prairie mounds testing fluvial hypotheses generally concentrate upon such features as mound distribution, elongation, location within floodplains, and stratigraphy. Soils within and beneath the mounds in this study indicate that the mounded topography is younger than a previous, roughly horizontal surface under which the textural B horizon formed. The mounds are composed of the same parent material as the sediment beneath them. Archuleta (1974) argued that prairie mounds in Arkansas were deposited in floodwaters beneath eddies. In the current study area, the clay content and thickness of the textural B horizon imply that it was formed under a surface with an elevation at least as high as the tops of the mounds. Since the mound bases are lower than the original, horizontal surface, a fluvial deposition origin would demand uniform removal of the upper portion of the soil horizon (above the Bt), and subsequent deposition of virtually indistinguishable sediment in the form of mounds. All fluvial deposition hypotheses oblige the same unlikely scenario in this mound field. Quinn (1961) and Jenks (1960) proposed an aeolian depositional origin for mounds in northwest Arkansas, concluding that they formed as coppice dunes anchored by vegetation in a desert environment during the Hypsithermal. Aeolian and fluvial forces sort sediments differently, however, and the uniform nature of the sediments within and beneath the mounds makes this scenario unlikely. Evidence from this study weakly supports a fluvial or aeolian erosion hypothesis. Studies by Cain (1974), Cox (1994), and Guccione et al. (1991) argue that prairie mounds are erosional remnants of fluvial action. The uniformity of parent material within and beneath the mounds in this study is in accordance with this theory. The contouring of the A horizon to the ground surface, as opposed to the horizontal nature of the relict textural B horizon, also fits well within this scheme. The B horizon would have formed under the formerly flat surface, with the A horizon subsequently imposed upon and contouring the mounded topography. Cain (1974) elaborates a scenario of trees anchoring soil against erosion in a period of dry climate punctuated by flooding events. Although erosional and depositional features are generally strongly oriented in respect to flow direction, anchoring by trees or other vegetation would explain the nearly circular form of prairie mounds. The presence of a shallowly-buried surface impenetrable to water (bedrock), also lends support to a fluvial erosion origin. A surface impeding the downward movement of water raises the water table. Surface sediment in such an area is likely to be saturated more often than sediment in areas where no such surface exists, and saturated sediment is generally more susceptible to erosion. Further study of mound micro-stratigraphy, possibly through the use of thin-sections, might further support an erosional origin hypothesis. If mounds were formed through erosion, there may be some horizontal stratigraphy left that would be common to most or all mounds within a mound field – relict from the previous, horizontal surface. This study suggests, however, that such stratigraphy is not resolvable with the methods used here.
|
|||
