Geochemical Characterization of Lapita Pottery Via Inductively Coupled Plasma–mass Spectrometry (ICP–MS)

Abstract

We have obtained high-resolution elemental data on Lapita ceramics (3200 -2700 cal year BP ) from Fiji, Tonga and New Ireland using chemistry-based inductively coupled plasma-mass spectrometry (ICP-MS). These data show clear elemental distinctions between Lapita pottery manufactured in Fiji, Tonga and New Ireland, and demonstrate significant elemental variation in Fijian ceramics collected from settlements in close proximity to one another. Therefore, we anticipate that ICP-MS will become an effective technique for tracking the transfer of Lapita pottery within and between different island groups in Oceania.

Pottery production, regional exchange, and state collapse during the Middle Horizon (A.D.500–1000): LA-ICP-MS analyses of Tiwanaku pottery in the Moquegua Valley, Peru

“…Chemical characterization of archaeological materials is now routinely used to investigate the production and movement of goods in the past, and has been particularly effectively applied to studies of obsidian and pottery as well as to metals and glass (Arnold et al 1991;Burger et al 1994;Burger et al 1998aBurger et al , 1998bBurger et al , 1998cBurger 2000;Burger and Glascock 2000;Cecil 2004;Glascock et al 2007;Gliozzo and Memmi Turbanti 2004;Gratuze 1999;Kennett et al 2004;Kennett et al 2001;Mallory-Greenough et al 1998;Mirti et al 2004;Niedershlag et al 2003;Perez-Arantegui et al 1996;Shortland 2002;Tykot 1997;Vaughn and Neff 2004). Studies identifying the presence of non-local goods contribute to interpretations of long-distance trade and interaction in a wide geographical and temporal range of archaeological contexts.…”

C Horizon

“…The coupling of ICP with mass spectrometry (ICP-MS) enables isotopic analyses to be made (e.g., Tyler and Olsson, 2001). An accessible description of the technique is given in Soltanpour et al (1996) and Kennett et al (2001).…”

Investigating Compositional Variation of CeramicMaterials during the Late Neolithic on the GreatHungarian Plain – Preliminary LA-ICP-MS Results

“…Secondly, Mahalanobis distance probability calculations [63,70] were carried out using the values of the first eight principal components to identify a core reference group and refine ceramic chemical groups [63,64,66,71]. This process evaluates hypothetical group assignments [72] and shows the statistical likelihood of chemical group membership for each sample.…”

“…Having completed this step, GAUSS was used to identify potential sample groupings in the compositional data, which can be accomplished using several methods. The first method employed generated a dendrogram [62,63] based on the hierarchical cluster analysis using the first eight principal components (see also, [64][65][66][67]). The dendrogram used the Euclidean distance for the distance measure on the x-axis and the average linkage for the clustering algorithm for the y-axis, and it was used only as an exploratory method of analysis [68].…”

“…Using GAUSS, ppm concentrations were transformed to log base-10 values in order to minimize gross variations between major, minor, and trace elements [4, [61][62][63]. Principal components (PC) analysis was conducted on the logged data using an R-Q factor analysis based on the correlation matrix, and it was determined that there were 10 principal components with eigen values greater than one, often considered a cut-off point for determining which principal component values to use [64] (also see [49,[51][52][53]). These 10 principal components ac-counted for 84.12% of the compositional variance in the ceramics; however, due to the limited overall sample size and because some statistical procedures require there to be at least two more samples than variables (including Mahalanobis distance probability calculations), it was necessary to decrease the number of principal components used from 10 to eight.…”