obsidian artifacts from El Chayal, Ixtepeque, Otumba,. Ucareo, San Martin Jilotepeque, Zacualtipan, Paredón,. Zaragoza, and Pico de Orizaba (Figure 2). These.
Using Portable X-ray Fluorescence to Determine Source Provenience of Obsidian from Lubaantun and Nim li Punit, Toledo District, Belize v
James T. Daniels, Jr. and Geoffrey E. Braswell Introduction
Research Objectives
This study presents the results of the use of This study attempts to test the efficacy of the Portable X-ray fluorescence (pXRF) spectrometry to nondestructive technique of pXRF spectrometry as a analyze the chemical composition of 612 obsidian viable alternative to more costly and destructive artifacts from Nim li Punit and 203 obsidian artifacts from means of geochemical sourcing. Lubaantun, Toledo District, Belize. The raw elemental This study also seeks to determine whether the data for each artifact were calibrated and quantified into distribution of obsidian sources at Lubaantun and Nim concentrations in parts per million (ppm) for statistical li Punit suggest two separate procurement and analyses. Various statistical routines including group consumption strategies, and how their strategies may classifications using Mahalanobis distances were be different from those in other parts of the Maya area. conducted with sample sets of obsidian artifacts with known proveniences to obtain geological provenience information for the unknown samples from Lubaantun and Nim li Punit. Lubaantun and Nim li Punit are located in the “South Toledo District” (Hammond 1975:1) circumscribed to the north and west by the foothills of the Maya Mountains, to the east by a thin coniferous pine ridge forest, and to the south by the marshy terrain of the Temash and Sarstoon rivers (Figure 1). Both sites are located on hilltops with Lubaantun positioned on the Rio Grande, which connects it to the Caribbean Sea. Nim li Punit is situated just above the edge of the coastal plain in a location that strategically controls north to south travel. Nim li Punit was founded arond A.D. 400-450 as indicated by pottery recovered from the site and was Figure 2: Map of the South Toledo District with the location of the sites of the occupied until some time after A.D. 830 as evinced by current study, Lubaantun and Nim li Punit. carved dates on stelae and ceramics (Fauvelle et al. 2013). Chronological data from Lubaantun, including Materials and Method Obsidian from Lubaantun, Nim li Punit, and nine stylistic evidence from three ballcourt markers (Morley 1938:IV:5-10) and ceramic cross-dating (Hammond known geological sources were analyzed using the 1975), suggest occupation there began after A.D. 700 Bruker Tracer III-V Portable XRF Analyzer that uses energy dispersive X-ray fluorescence (EDXRF) and ended around A.D. 900. technology with an X-ray tube as the excitation source and a high-resolution, Peltier cooled, Silicon PIN (Si-PIN) diode detector (Kaiser and Wright 2008:8). The data were calibrated, and concentration values in parts per million (ppm) for elements Mn, Fe, Zn, Ga, Th, Rb, Sr, Y, Zr, and Nb for each obsidian sample were recorded. The known source reference samples included obsidian artifacts from El Chayal, Ixtepeque, Otumba, Ucareo, San Martin Jilotepeque, Zacualtipan, Paredón, Zaragoza, and Pico de Orizaba (Figure 2). These artifacts with known source proveniences were used as reference groups for comparisons with the samples with unknown source provenience from Lubaantun and Nim li Punit. The data were imported into GAUSS language based software and a series of statistical routines designed at MURR were performed, including principal components analysis (PCA), canonical discriminant analysis, and calculation of group membership probabilities using Mahalanobis distances. Figure 1: Map of the South Toledo District with the location of the sites of the current study, Lubaantun and Nim li Punit.
Results
PCA was performed on the data set from Lubaantun and Nim li Punit to explore its structure. The elements Mn, Ga, and Th were eliminated from the statistical analyses because Mn is below the optimized level of excitation, Ga often has very low and dubious values, and the L-line of Th may not be accurately detected. One extreme outlier was identified, sample N381, which was visually sourced to the Mexican source of Pachuca because of its green color. The sample was removed from the data set in order to clarify its structure and to determine if there werre other clusters in the distribution of the data (Figure 3). The results of the PCA demonstrate that there are two main source groups represented in the unknown samples from Lubaantun and Nim li Punit, which, based on the visual sourcing performed by Braswell, should be El Chayal and Ixtepeque. A third smaller cluster of artifacts is also distinguishable from the two main groups predominately by a negative correlation to strontium concentrations. The loadings of elements on principal component one suggest the two main groups are separated by Zr and Rb. A bivariate plot of using these two elements with the 95% confidence intervals of the source reference groups indicate the two main geochemical groupings do indeed represent the Chayal and Ixtepeque sources (Figure 4). The third smaller group is Ucareo (Figure 5). The results also indicate one sample from Lubaantun came from the Zacaultipan source, one sample from Nim li Punit came from San Martin Jilotepeque,and three others are from the Otumba source in Mexico. The canonical discriminant analysis provides even better separation of the source groups (Figure 6). Group membership probabilities using Malhalanobis distances were calculated for each artifact from the sites. The results indicate that of the 612 pieces of obsidian from Nim li Punit, 374 are from El Chayal, 226 from Ixtepeque, seven from Ucareo, three from Otumba, one from San Martin Jilotepeque and one piece was visually sourced to Pachuca. Of the 203 pieces from Lubaantun 165 artifacts are from El Chayal, 28 from Ixtepeque, nine from Ucareo, and one from Zacaultipan (Table 1).
Conclusions
Figure 3: PCA of obsidian samples from Lubaantun and Nim li Punit, Belize.
Figure 4: PCA including source reference groups.
Figure 5: PCA including source reference groups.
The current study demonstrates the effectiveness and efficiency of pXRF as a viable alternative for other more costly and destructive sourcing techniques. Although the pXRF study was able to assign the obsidian samples from Lubaantun and Nim li Punit to specific geological sources, there are some issues with accuracy in measurement of chemical concentrations as identified by Nazaroff et al. (2010). However, the precision of the technology is quite good, even for small samples. The efficiency is illustrated by the fact that over 800 artifacts were analyzed in just a few days’ time. The overall density of El Chayal obsidian at both sites is relatively similar, but the density of Ixtepeque obsidian at Nim li Punit is about three and one-half times greater than that at Lubaantun. The differences between the two assemblages as identified with the pXRF data are further evidence that the sites were indeed independent polities with different economic connections as suggested by Braswell et al. (2008:60). Early Classic contexts at Nim li Punit exhibit lower overall abundance of obsidian but Ixtepeque obsidian was the predominan. During the Late Classic and Terminal Classic I phases, El Chayal was the principal source of obsidian. During the Terminal Classic II period, the relative proportion of Ixtepeque obsidian to increasedsomewhat. Nim li Punit is likely to have been an independent polity, completely separate from Lubaantun. During the Early Classic it likely received obsidian from inland trade routes or via coastal distribution where obsidian moved down the Motagua River from El Chayal and Ixtepeque through Quirigua. During the Late Classic, the influx of Ixtepeque obsidian may have declined due to political strife in the southeastern Maya periphery, including Quiriguá’s revolt against Copán in A.D. 738. The increase of Ixtepeque in Terminal Classic contexts at Lubaantun and Nim li Punit may have been the result of stabilization in the southeastern periphery and expansion of coastal trading.
References
Figure 6: Canonical discriminant analysis showing the distribution of source materials represented at the sites of Lubanntun and Nim li Punit, Belize.
Braswell, Geoffrey, Cassandra R. Bill and Christian Prager 2008 Exchange, Political Relations, and Regional Interaction: The Ancient City of Pusilha in the Late Classic Maya World. Research Reports in Belizean Archaeology 5:51-62. Fauvelle, Mikael DH, Megan R Pitcavage and Geoffrey E Braswell 2012 Dynatic Capital, Minor Center, or Both? Recent Investigations at Nim li Punit, Toledo District Belize. Paper presented at the 9th Annual Belize Archaeology Symposium, San Ignacio. Hammond, Norman 1975 Lubaantun: A classic Maya realm. Peabody Museum of Archaeology and Ethnology, Harvard University Cambridge, Mass. Kaiser, Bruce and Alex Wright 2008 Draft Bruker XRF Spectroscopy User Guide: Spectral Interpretation and Sources of Interference. Morley, Sylvanus G 1937-1938 1937-1938. The inscriptions of Peten. 5 vols. Washington, DC: Carnegie Institution of Washington Publication 437. Nazaroff, Adam J, Keith M Prufer and Brandon L Drake 2010 Assessing the applicability of portable X-ray fluorescence spectrometry for obsidian provenance research in the Maya lowlands. Journal of Archaeological Science 37(4):885-895.
