ABSTRACT

This chapter explores the potential of network alignment to measure the conservation needs of complex systems. Networks describe how a system's components relate to each other, and changes in the structures of these relationships may offer more sensitive diagnostics than any of the system's individual components. Network alignment algorithms measure such changes by comparing pairs of networks, turning them into a single number akin to a ruler measuring the distance between them. Alignments between many pairs of networks can be used to track a system's behavior by creating a state-space where more similar networks are closer. Plots of these state-spaces can help us visualize the system's behavior and identify times and places of instability. As an example of how to do this, I have aligned 1.14 billion pairs of networks describing 19 measures of the Chesapeake Bay's water quality at 133 monitoring stations recorded from 1985 to 2015. Altogether these 1.14 billion distances map a state-space of the Bay's behavior. Unlike the 19 component variables, this alignment state-space indicates the Bay is less stable closer to land, and may be destabilizing up north near Washington D.C. and Baltimore. The chapter ends with simplified open source R code to help you apply this approach to any system.