Kirsanow

FTIR, DNA, AND THE FUTURE OF ARCHAEOLOGICAL SOIL ANALYSIS : 

FTIR, DNA, AND THE FUTURE OF ARCHAEOLOGICAL SOIL ANALYSIS References: Chapman, S.J. , S.D. Campbell, A.R. Fraser, and G. Puri. 2001. FTIR Spectroscopy of peat in and bordering Scots pine woodland: relationship with chemical and biological properties. Soil Biology and Biochemistry 33:1193-1200. Cox, R.J., H.L. Petersen, J. Young, C. Cusik, and E.O Espinoza. 2000. The forensic analysis of organic soil by FTIR. Forensic Science International 108: 107-116. Ding, G. J.M. Novak, D. Amarasiriwardena, P.G. Hunt, and B. Xing. 2002. Soil organic matter characteristics as affected by tillage management. Soil Science Society of America Journal 66: 421-429. Ellerbrock, R.H., A. Hohn, and J. Rogasik. 1999. Functional analysis of soil organic matter as affected by long-term manurial treatment. European Journal of Soil Science 50: 65-71. Ellerbrock, R. H., A. Hohn, and H.H. Gerke. 1999. Characterization of soil organic matter from a sandy soil in relation to management practice using FT-IR spectroscopy. Plant and Soil 213:55-61. FTIR as an Analytical Tool in the Field The primary goal of this study was to determine the effectiveness of FTIR spectroscopy as a field technique for the chemical analysis of archaeologically interesting soils. Modern FTIR equipment can produce rapid-non-destructive analysis of fresh or preserved material, an advantage over time-consuming laboratory methods (Chapman et al. 2001,). It is important to determine, however, whether this method can reliably differentiate soils that have experienced different archaeological land-use patterns. This analysis was performed in conjunction with a study of the phosphate content of soils having different land-use histories, and the resultant phosphate data were checked for covariance with features of the FTIR spectra. It should be emphasized that the purpose of this study was to compare, rather than characterize, the soils from the different historical land-use areas. The question investigated can most neatly be summarized as “Do easily extractable soil organics produce FTIR spectra that differ according to land-use type?” In this study, we analyzed dried soils from three different historical land-use areas, as determined through historical data and phosphate analysis: previously cultivated soils, improved (manured and plowed) pasture, and permanent woodlot from the Pierce Farm area of the 380 ha Prospect Hill tract of the Harvard Forest in Petersham, Massachusetts. Principles: Evidence from long-term field experiments indicates that the composition and quantity of SOM is affected by agricultural management practices, especially the addition of field amendments such as fertilizers and manures. These amendments can produce differences in the content and spatial arrangement of functional groups within the SOM (Ellerbrock et al. 1999). FTIR (Fourier Transform Infrared Spectroscopy) can be used to analyze the organic fraction of soils by identifying infrared bands (absorption peaks) characteristic of various functional groups Methods Dry soil samples for FTIR analysis were taken from the 0-10 cm and 15-25 cm depth intervals of soil cores of three different land-use areas of the Pierce Farm. It should be noted that these core intervals do not represent natural soil depths, as the soils were compacted during the coring process. 200 µl of EDTA and 100 µl of 2-propanol were added to centrifuge capsules containing 0.25 g soil samples. The samples were centrifuged in an Eppendorf centrifuge for 15 minutes at 10000 rpm, and then syringe-filtered through a 45 µm PTFE filter. This process was repeated 3 times. 100 µl of the supernatant liquid was pipetted onto Spectra-Tech PTFE (Polytetrafluoroethylene) and PE (Polyethylene) ST-IR cards. The cards were set out to dry for 24 hours under a fume hood, and then analyzed in a Thermo Nicolet IR 100 FTIR spectral analyzer. Absorbance spectra from samples on PE and PTFE were compared to minimize the deleterious effects of the different intrinsic interference regions of the two ST-IR card types. PE cards are not useful between 2918-2849 cm-1, 1480-1430cm-1, and 740-700 cm-1. PTFE cards can show interference in the 1270-100 cm-1 and 660-460 cm-1 regions. These interference regions are marked by grey bands in Figures 1-3, which display spectra obtained on PTFE cards. FTIR Spectra and Analysis Although the purpose of this study was to compare rather than characterize the spectra from the different land-use areas, a bit of FTIR geography is useful in getting our bearings in the spectra. The broad peak seen in all three figures at 3500-3200 cm-1 is due to O-H stretching vibrations in alcohols, phenols, and water (Ellerbrock et al., 1999). Because an alcohol was used in the preparation of soil samples for analysis, and because water content can vary between samples even after careful preparation, this peak is not as informative as it seems. The tall peak seen around 1600 cm-1 represents unsaturated ketones or amides, and the second tall peak around 1400 cm-1 is due to carboxylic and carbonylic groups (Ellerbrock et al., 1999). In many of the spectra, a smaller peak can be seen at 1081 cm-1 which represents certain groups found in cellulose (Chapman et al., 2001). Interestingly, bands at 1690 cm-1 and 1710 cm-1 commonly associated with manure amendments were not observed in these spectra, suggesting that these peaks are not long-lived enough to be archaeologically informative. Comparisons of peak presence/absence and peak height between spectra from different land-use types suggests that the region between 1600 and 1400 cm-1,and the region between 100 and 800 cm-1 may be the most informative for differentiating soils having different land-use histories. The inter-area variance in these regions can be seen in Figure 1 below. DNA Extraction and Analysis To Infinity and Beyond Work is continuing on identifying the organisms represented by the DNA extracted from the Harvard Forest soils. Another several waves of subtractive FTIR analyses are planned using different extractants in order to attempts to reduce the intra-landuse area variance seen in the soil spectra (see Figure 2 above). Absorbance spectra were also collected for samples from each land-use area which had been heated at 650 C for 8 hours In a muffle furnace. The muffling process destroys organic material, leaving the inorganic soil fraction behind. These samples were used to represent the mineral component of the soil spectrum. Spectra produced by the muffled samples were subtracted from those of their non-muffled counterparts from the same land-use area (after Cox et al., 2000). The subtracted spectra obtained by this method represent the organic component of the soils. An example of a subtracted spectrum is shown in Figure 3. FTIR and DNA Analysis at the Harvard Forest The purpose of our fieldwork at the Harvard Forest was to develop and implement methods to differentiate archaeological soils according to land-use history. Our primary technique was phosphate analysis, but we have also begun preliminary work in developing alternative methods for characterizing soil organic matter (SOM). A protocol for using FTIR (Fourier Transform Infrared Spectroscopy) to distinguish soil samples having different archaeological land-use histories is in preparation. FTIR can be used to produce visual representations of soil composition, a potentially useful analytical tool. We have also extracted DNA from the microbial biomass of the soils from the study site in the Harvard Forest. This DNA will be further analyzed in order to characterize the microbial complements of soils having different land-use histories. Additional investigation of the variable regions mentioned above Is also being performed in the hope of identifying consistent differences in absorbance peak position or intensity between land-use areas. These differences should be consistently greater than intra-landuse area variation in order to be useful.