APPENDIX C

Investigation of Soil pH at Cox's Woods Site, 12 Or 1

Timothy M. Wright


Return to Table of Contents

[acknowledgement] [background] [method]
[result] [conclusion] [references]

Acknowledgements

I would like to begin by saying thank you to the Honors Division at Indiana University for funding this research. I would also like to express my gratitude to the staff and to all the graduate students and fellows at the Glenn A. Black Laboratory of Archaeology for graciously sharing their expertise and for giving me valuable advice and constant encouragement. I would also like to give recognition to the students of the field school in archaeology, Emily Brouwer, Chad Harvey, Jake Miller and Teresa Naomia, all of whom assisted me in the actual data collection. Lastly I would especially like to thank Dr. Brian G. Redmond and Mr. Bret Ruby.

Background
For the past four years, the IU Field School in Archaeology has focused on the Upper Mississippian Oliver Phase occupation on the East Fork of the White River, primarily in Lawrence, Martin and Orange Counties. From the excavations during this period all the bone that has been recovered has been in a uniformly poor state of preservation. When anyone asked why this was so, the standard answer was--"The soil is too acidic."
For the past year, this author has been working on a faunal collection from another Oliver Phase site located in Marion County. Although the Marion County site is about 100 years older than the East Fork sites in the study area, the bone from the Marion County site is better preserved.
Assuming that the cooking and disposal behaviors remained constant within the culture, the only explanation for this noticeable difference in bone preservation would seem to lie with the soil itself.
Soil acidity is commonly measured and expressed as pH. A pH reading of less than 7.0 is considered acidic while a pH reading greater than 7.0 is alkaline. If soil acidity, as measured on the pH scale, is the main determinant in bone deterioration, bone preservation should be at its best where pH is highest and acidity is lowest. Thus, my initial research plan was to search for a relationship between soil pH and bone preservation and to see if pH readings have any value in detecting archaeological deposits.

Return to Top of Page

Method

With the help of field school students, two separate transects were laid out across the site (See Fig.7). The first transect ran due east-west on the East 500 line of the site grid. There has been a great deal of excavation along this axis so pH readings along this transect could be associated with known deposits. This means the east-west transect serves as a control. The second transect ran diagonally across the site from the southeast to the northwest. Both transects began and ended outside the perimeter of the circular village. This perimeter had been established by earlier excavation. The village being circular, it was expected that these two transects would yield similar pH profiles.
Along these two transects, at 5-meter intervals, holes were drilled with a bucket auger. These test holes were divided into four 10-cm levels, and a pH reading was taken from each level using a Soil pH and Humidity Tester.{1} This instrument is not particularly sophisticated, but it was considered adequate for giving relative readings across the site.
On the east-west transect, 27 holes were drilled for a total of 108 pH readings. Along the southeast-northwest transect, 29 test holes were drilled and a total of 116 readings taken. During analysis, the data (8 readings) from the last two test holes on the extreme northwestern end of this transect was omitted. This seemed justified because this end of the transect was in the Haymond silt/loam soil group which is a recent fluvial deposit (Wingate 1984). Thus, this soil is an anomaly when compared to the main soil types at 12 Or 1. In addition if there were cultural deposits in this area, they would likely be buried far below the 40cm. level which is the maximum depth tested. This also gives both transects an equal number of readings which makes statistical and graphic comparisons easier.
The total sample from each 10 cm interval was screened through 1/4-inch mesh to insure good contact with the device. A reading was taken and recorded. All samples except level 1 (0-10 cm) were bagged and are now curated at the Glenn Black Laboratory.

Return to Top of Page

Results

The results of the pH readings are shown in graphic form{2} (Figures 1, 2, 3, 4, 5 & 6) and in Table 5.

On the east-west transect (Figures 1-3), levels 1 and 2 form a distinct grouping with a mean pH of 6.3, while levels 3 and 4 have a mean of 6.6 (Table 1).
Table 1.
Descriptive Statistics for East-West Transect, 12 Or 1

Level 1 Level 2 Level 3 Level 4 Transect
Maximum pH 6.5 6.5 6.9 7.0 7.0
Minimum pH 6.0 6.0 5.8 5.8 5.8
Range 0.5 0.5 1.1 1.2 1.2
Mean 6.3 6.3 6.6 6.6 6.5
Std. Dev. 0.144 0.169 0.241 0.252 0.253
Variance 0.021 0.029 0.058 0.063 0.064

This pattern of separation between the upper two levels and the lower two is persistent on the east-west transect. There are two possible explanations for this:
1)The 500 line was used as the main trail across the site by the field school over the last two seasons. This traffic may have compacted the surface which effected the soil pH, or
2)Levels 3 and 4 were tested several days after levels 1 and 2. There were several rain days in this interval which may have changed the moisture content and pH of the lower levels.
Regardless of the cause, the readings from levels 1 and 2 were viewed by this author as anomalous and excluded from any statistical tests.
The important thing to notice in the graphs is the increased variability or "static" that appears at both ends of this transect.
Readings at E525 and E520 coincide with visible earthworks at the perimeter of the village. Readings at E425 and E420 are at the west perimeter of the stockade, the exact position of which was only established through excavation. This same pattern of "static" shows up again on the Southeast-Northwest transect (Figures 4-6). In these areas (tests 3 and 4 and 23 and 25) there is no surface indication of the earthwork, and there have been no excavations to verify its exact location. If this "static" does coincide with the earthwork, one would be inclined to show the stockade line as a closed circle rather than the horse-shoe shape depicted in Figure 7.
I have no explanation for the erratic readings in the middle of the graphs, but it appears on both transects. Perhaps wild fluctuations in soil pH, whether to the high or low end of the scale, would indicate some kind of soil mixing which may or may not be associated with prehistoric human activities. It may be suggested that this central part of the village should be assigned high priority for future excavation to determine the exact agent responsible for this soil mixing.
Nowhere is the difference between the two transects more obvious than in the graphs of the linear fit (Figures 3 and 6). The east-west fit shows the split that separates levels 1 and 2 from levels 3 and 4. It should also be noted that levels 3 and 4 intersect at E470. This is interesting in that it mimics the actual stratigraphy of the site. On the east end of the site, cultural deposits are first encountered at about 20 cm below surface and exhausted by 30 cm, so the bulk of known cultural material is in level 3. The west end of the site is more deeply buried with intact features first encountered at about 30 cm below the surface. So, on this transect the best fit line constructed from pH readings gives an accurate indication of the actual slope of the subsurface deposits. In 1994, there were no excavations on the southeast-northwest transect to verify the value of this observation.
Descriptive statistics also serve to illustrate the dissimilarities in the two transects.
Table 2.
Descriptive Statistics for Southeast-Northwest Transect, 12 Or 1

Level 1 Level 2 Level 3 Level 4 Transect
Maximum pH 6.9 6.9 7.0 7.0 7.0
Minimum pH 6.2 6.4 6.4 6.5 6.2
Range 0.7 0.5 0.6 0.5 0.8
Mean 6.7 6.7 6.7 6.7 6.7
Std. Dev. 0.156 0.155 0.137 0.131 0.146
Variance 0.024 0.024 0.019 0.017 0.021

Figure 6 shows all four levels clustered around the mean of 6.7. The separation into upper and lower cluster is not nearly as extreme in the southeast-northwest transect, but it does still exist. In addition levels 3 and 4 intersect again.
In order to determine if the east-west readings have any predictive value, a correlation was run on level 4. Readings along the east-west (control) transect were the independent variable. Readings along the southeast-northwest transect were the dependent variable.{3} Results showed an extremely weak (-0.069) negative relationship with a very high probability (p< .05) that any relationship is a matter of chance.
Also, an analysis of variance was run on all four levels to determine if there was any pattern in the dissimilarity of the two transects. Since the correlation analysis showed the poor predictive value of any individual pH reading, test number was used as an independent "dummy" variable for an analysis of variance. This procedure also showed that no statistically significant relationship exists between the two transects. So, despite the apparent similarity in the graphic displays, pH readings have almost no predictive value as concerns location of subsurface archaeological deposits. Whether we look at individual readings or patterns in the entire transect, there are too many confounding factors that effect soil pH to make it an effective tool for archaeology.
So far none of this has addressed the other question of, "How does soil pH relate to bone preservation?" Throughout the entire survey process, no bone was recovered, and only three readings of pH 7.0 were recorded. Fortunately, in the course of excavation bone was recovered by Stephen Ball and his crew in Unit BBB in a matrix that read 7.0. Also in Unit QQ bone was recovered by Rex Garniewicz and his team in a matrix that tested 6.5 pH. A fluctuation in pH reading reflects an exponential change in ion exchange capacity of the soil that could have a marked effect on chemical leaching, i.e., bone deterioration. A small fluctuation in pH reading can make a big difference in the preservation potential, but it is not a direct relationship. A pH 7.0 (neutral) does not guarantee good bone preservation. There was no discernable difference in the state of preservation of bone from the two above-mentioned contexts. In both cases, the bone was in the same poor condition we have come to expect in the East Fork study area. It is apparent that pH alone is not the sole determining factor in bone preservation. In fact, it seems likely that as soil pH increases the potential for leeching also would increase, leading to a more rapid deterioration.
Bone is made up primarily of potassium and calcium. If the surrounding soil is lacking in these minerals, they will be quickly removed from the bone when H2O is available in sufficient quantities to act as the catalyst. The water retention properties of southern Indiana soils are well known. In regard to bone preservation, or the lack thereof, the mineral content of the soil matrix and water retention properties of the soil are much more important than soil acidity as measured on the pH scale (Brady,1984, Foth 1984).

Return to Top of Page
Conclusion

Based on pH readings, the soils at the Cox's Woods site cannot be characterized as acidic. Therefore, soil acidity as measured on a pH scale does not directly determine quality of bone preservation in archaeological contexts. In addition pH is a very localized and fickle property of the soil and can be effected by numerous agents other than subsurface archaeological deposits.
The parent material establishes the baseline pH and chemical characteristics of a soil group. These are then locally altered by such diverse factors as climate, vegetation, compaction, mixing, and/or the inclusion of organic matter in the form of archaeological deposits. As a result a particular pH reading can be the product of many factors. This makes the value of a pH survey questionable as a means of pinpointing subsurface archaeological deposits. This is why no statistically significant relationship could be shown between the two transects, even though we were dealing with a circular village in which archaeological deposits are in a predictable concentric ring. Excavation is still required to establish cause for any recorded variation in soil pH and to locate archaeological deposits.
Though a pH survey of this kind has obvious limitation as a predictive or interpretive tool for archaeologists, this author does not feel it was a waste of time. Soil pH is a property of all archaeological sites. As a universal feature, it should be routinely recorded in all excavations. It is not that time- consuming or expensive to do so. Because one does not see the value of a data set now does not mean someone else won't come along at a later date and reinterpret it to the benefit of the entire discipline.

Return to Top of Page
References cited

Brady, Nile C.
1984The Nature and Properties of Soil. 9th ed. Macmillan Publishing Company: New York.
Courty, Marie Agness, Paul Goldberg, and Richard Macphail
1989Soils and Micromorphology in Archaeology. Cambridge Manuals in Archaeology: Cambridge University Press.
Foth, Henry D.
1984Fundamentals of Soil Science. 7th ed. John Wiley and Sons, Inc.: New York.
Redmond, Brian G
ndbThe summer 1993 excavations of the Cox's Woods site (12 Or 1), a late prehistoric Oliver phase village in the Pioneer Mothers Memorial Forest Recreation Area, Hoosier National Forest. Glenn A. Black Laboratory of Archaeology, Indiana University. Reports of Investigations 94-17.
Wingard, Robert C., Jr.
1984Soil Survey of Orange County, Indiana. United States Department of Agriculture, Soil Conservation Service.
Table 3.
Raw Data for East-West Transect, 12 Or 1

Test#     North     East     pH1     pH2     pH3     pH4
1            500       535      6.3       6.1     6.5      6.5
2            500       530      6.2       6.5     6.8      6.8
3            500       525      6.4       6.4     6.7      6.9
4            500       520      6.4       6.0     6.7      6.3
5            500       515      6.2       6.4     6.7      6.3
6            500       510      6.5       6.0     6.7      6.7
7            500       505      6.4       6.2     6.9      6.7
8            500       500      6.4       6.4     6.8      6.8
9            500       495      6.4       6.4     6.9      6.8
10           500       490      6.4       6.4     6.6      6.6 
11           500       485      6.4       6.4     6.7      6.6
12           500       480      6.1       6.1     6.3      6.3
13           500       475      6.0       6.0     6.8      6.7
14           500       470      6.2       6.2     6.3      6.4
15           500       465      6.4       6.4     6.8      6.8
16           500       460      6.4       6.4     6.8      6.7
17           500       455      6.5       6.5     6.8      6.8
18           500       450      6.5       6.5     6.5      6.8
19           500       445      6.3       6.3     6.4      6.4
20           500       440      6.4       6.4     6.6      6.6
21           500       435      6.2       6.2     6.4      6.6
22           500       430      6.1       6.1     6.3      6.3
23           500       425      6.4       6.4     6.6      6.6
24           500       420      6.0       6.0     5.8      5.8
25           500       415      6.4       6.4     6.4      6.6
26           500       410      6.4       6.4     6.6      6.8




Table 4.
Raw Data for Southeast-Northwest Transect, 12 Or 1

Test#      North      East      pH1      pH2      pH3      pH4
1 440.00 520.00 6.8 6.5 6.5 6.6 2 443.54 516.46 6.4 6.6 6.7 6.7 3 447.07 512.93 6.4 6.5 6.4 6.5 4 450.61 509.39 6.7 6.7 6.9 6.7 5 454.14 505.86 6.7 6.7 6.7 6.7 6 457.68 502.32 6.8 6.5 6.7 6.8 7 461.22 498.87 6.7 6.8 6.7 6.6 8 464.75 495.25 6.7 6.8 6.7 6.7 9 468.29 491.71 6.7 6.8 6.7 6.7 10 471.82 488.18 6.8 6.8 6.6 6.7 11 475.36 484.64 6.5 6.7 6.7 6.8 12 478.90 481.10 6.7 6.8 6.9 6.8 13 482.43 477.57 6.8 6.9 6.9 6.9 14 485.97 474.03 6.8 6.8 6.8 6.8 15 489.50 470.50 6.5 6.4 6.7 6.8 16 493.04 466.96 6.8 6.5 6.5 6.5 17 496.58 463.42 6.5 6.5 6.7 6.7 18 500.11 459.89 6.2 6.4 6.8 6.7 19 503.65 456.35 6.8 6.9 6.8 6.9 20 507.18 452.82 6.9 6.9 7.0 7.0 21 510.72 449.28 6.8 6.9 6.9 6.8 22 514.26 445.74 6.8 6.8 6.8 6.8 23 517.79 442.21 6.8 6.8 6.9 6.8 * 25 524.86 435.14 6.7 6.7 6.6 6.5 26 528.40 431.60 6.7 6.5 6.7 6.7 27 531.94 428.06 6.7 6.7 6.6 6.5 **28 535.47 424.53 6.7 6.7 6.7 6.5 *29 539.01 420.99 6.7 6.6 6.6 6.6

* Test #24 omitted because it fell within an excavated unit.
** Test #s 28,29 dropped from graphs and regression.

Return to Table of Contents
Return to Top of Page