Font Size:  Small  Medium  Large

Exploring the utility of path analytics to make inferences about dynamic interactions between pairs of moving objects

Jennifer Miller

Abstract


In animal movement ecology, ‘dynamic interactions’ are interactions defined by proximity in both space and time and are important for understanding spatial behaviours such as mating, predation, and territoriality as well as phenomena resulting from these behaviours such as disease spread. The two main approaches of quantifying dynamic interactions consider the locations as discrete points or as paths that are inferred as trajectories between subsequent points.
Point-based metrics consider proximity or distance between individuals explicitly, but their calculation is dependent upon subjective values such as a distance threshold or home range estimation. Path-based metrics compare derived movement parameters such as step length and azimuth. They require fewer subjective inputs but are inherently aspatial, as they are based on movement similarity irrespective of relative location.
This research explores the utility of quantifying interactions by combining spatiotemporal point proximity with analysis of path parameters to infer interaction. The movement parameters measured one or more of speed, direction, or change of speed/direction. This hybrid method was tested using data for fifteen black-backed jackal (Canis mesomelus) dyads comprised of twelve different individuals in Etosha National Park, Namibia.
Results indicated that combining path analytics with spatio-temporal proximity may be more appropriate for studying dynamic interactions that are likely to be in response to another individual but are asymmetric. Potentially interaction-related movement parameters between individuals in a dyad were not always correlated, which suggests that path-based dynamic interaction metrics should allow for asymmetry.

Full Text: PDF

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.