"Zooarchaeological datasets are often large, complex, and difficult to visualize and communicate. Many visual aids and summaries often limit the patterns that can be identified and mask interpretations of relationships between contexts, species, and environmental information. The most commonly used of these often include bar charts, pie charts, and other such graphs that aid in categorizing data and highlighting the differences or similarities between categories. While such simplification is often necessary for effective communication, it can also obscure the full range of complexity of zooarchaeological datasets and the human-environment dynamics they reflect. In this paper, we demonstrate the utility of formal network graphs to capturing the complexity of zooarchaeological datasets and to effectively highlighting the kinds of relationships between contexts, time, and faunal assemblages in which zooarchaeologists are primarily interested. Using a case study from southwestern Florida (USA), we argue that network graphs provide a quick solution to visualizing the structure of zooarchaeological datasets and serve as a useful aid in interpreting patterns that represent fundamental reflections of human-centered ecosystems."
From https://pmc.ncbi.nlm.nih.gov/articles/PMC10729997/#abstract1
Construction
“Networks are built from simple data tables, or matrices, such as a species list indicating MNI or presence/absence for each species within particular contexts. In some cases, ties represent the presence of a specific species in a specific context or habitat. In other cases, ties represent a certain cutoff point of percent MNI for a specific context. Another example may be a similarity matrix, where rows and columns both represent species, and the values within the matrix represent how similar those species are in regard to the contexts in which they are found.” “We use two different kinds of networks: two-mode networks contain two different kinds of nodes with ties between different classes of nodes; a one-mode network is a transformation of a two-mode network where only a single kind of node is depicted.” “For the graphs we build here, nodes might represent specific contexts, specific species, or specific habitats.” “All data is archived on Zenodo at https://doi.org/10.5281/zenodo.7948824.”
deFrance S, Walker KJ. The zooarchaeology of Pineland. In: Marquardt WH, Walker KJ, editors. The archaeology of Pineland: A coastal southwest Florida site complex, AD 50–1710. Monograph 4, Institute of Archaeology and Paleoenvironmental Studies, University of Florida: Gainesville; 2013. pp. 305–348.