Using Gaussian Mixture Model clustering for multi‐isotope analysis of archaeological fish bones for palaeobiodiversity studies

Abstract

The GMM clustering method is applicable to complex multi-dimensional stable isotope data sets established by isotope ratio mass spectrometry (IRMS). This exemplary application resulted in an identification of habitat preferences and non-local individuals. Depending on the scientific question to be solved, restriction of the cluster size could lead to a better reproducibility; however, with loss of dissolution. Copyright © 2016 John Wiley & Sons, Ltd.

Ancient DNA reveals the Arctic origin of Viking Age cod from Haithabu, Germany

“…For instance, in medieval and postmedieval Europe, the onset of urbanized market economies has been linked to the growth of longrange trade by historical and archaeological evidence (3,4). The exploitation of increasingly distant fish populations has proven to be one of the clearest demonstrations of this ecological globalization (5)(6)(7)(8)(9). Fisheries around the coastal regions of the Lofoten Archipelago in Norway have a particularly long history; in this region, the regular arrival of seasonal spawning aggregations of Atlantic cod-migrating southwards from the Arctic Barents Sea (10, 11)-coincides with those climatic conditions ideal for the freeze drying and long-term preservation of cod without the use of expensive salt.…”

“…Fourth, the cod finds from Haithabu occurred alongside species such as saithe (Pollachius virens), ling (Molva molva), and halibut (Hippoglossus hippoglossus), which is more consistent with fishing in the North Sea or North Atlantic than in the Kattegat or Baltic Sea (38)(39)(40). Finally, isotope analyses using bone collagen (8,41) or bone carbonate (9) suggest that the Haithabu cod were not locally caught, although their origin remains ambiguous. This previous research pinpoints the fish bones from Haithabu as ideal material with which to test the hypothesis of Viking Age transport of cod from northern Norway by using genomic methods.…”

Which isotopes should we choose? Entropy‐based feature ranking enables evaluation of the information content of stable isotopes in archaeofaunal material

“…For example, cluster analysis of isotopic data of fish from Haithabu and Schleswig using Gaussian Mixture Model (GMM) clustering (see 1.3 and 2.3) not only separated fish according to their habitat (freshwater, brackish, marine) but also revealed a fourth cluster of probably non‐local fish from a colder environment. These groups could not be detected in the bivariate plots . Therefore, it is advisable to use multi‐dimensional isotopic data whenever possible.…”

“…For an illustration of the feature ranking results, optimal combinations of features were visualized using Gaussian Mixture Model (GMM) clustering (see 2.3). Cluster analysis based on GMM clustering has already been tested for multi‐isotope data . GMM clustering is a clustering method representing data as a mixture of multivariate normal (Gaussian) distributions.…”

“…The GMM clustering procedure used in the present study (R package “mclust”) uses an expectation maximization (EM) algorithm to detect the maximum likelihood of the model. For additional information on GMM clustering see Göhring et al and references therein.…”