Workflow for constructing social networks from automated telemetry systems

Abstract

1.Advances in datalogging technologies have provided a way to monitor the movement of individual animals at unprecedented spatial and temporal scales, both large and small. When used in conjunction with social network analyses, these data can provide insight into fine scale associative behaviors. The variety of technologies demand continuous progress in workflows to translate data streams from automated systems to social networks, based on biologically relevant metrics. 2.Here we present a workflow for generating flexible association matrices from automated radio-telemetry data that can be parsed into both spatial and temporal dimensions. These can then be used to generate and compare social networks across space and time.3.We illustrate this workflow using data collected from an automated telemetry study of acorn woodpeckers (Melanerpes formicivorus), a cooperatively breeding bird. The data were collected continuously over two years at base stations placed within social group territories. We use this system to demonstrate how this flexible data structure can be used to answer a number of biological questions, specifically 1) how assortative are social associations at the population scale, 2) how do association patterns among territory visitors vary across territories, 3) and how does seasonality affect assortative affiliation within groups?4.This flexible method allows one to generate social networks that can be used to ask a variety of biological questions pertinent to a wide range of animal systems, exploiting the investigative power that can be gained by using automated radio-telemetry in conjunction with social network analyses.

A comparison of low‐cost behavioral observation software applications for handheld computers and recommendations for use

“…The more recent technological advancement of automated methods, such as GPS trackers and proximity loggers, provides an exciting avenue to collect behavioral and ecological data (Shizuka et al, 2020;Smith & Pinter-Wollman, 2020). Nevertheless, traditional observation methods are necessary to validate these automated methods (Gelardi et al, 2020) and to describe and collect data on the contexts of behaviors (something most automated methods still cannot collect).…”