Analyze revisits in animal movement trajectory data
The paper in Ecography is here: https://doi.org/10.1111/ecog.03618
Movement is a defining characteristic of animals, and it connects with many important life history events, from finding food to avoiding predators to finding mates. As part of these movements, animals can revisit areas for a number of reasons, many of which are ecologically significant. For example, animals may return to regularly used resources, such as foraging patches or watering holes, and the return interval could be relevant to the animal's internal state or the resource level. Revisits may involve caring for young, such as returns to a den or nest, which could also be involved in central place foraging.
The seeds for this paper began in informal discussions in our working group about how to quantify animal movements returning to an area. A quick lit search revealed a few papers, but no straightforward easy-to-use method to apply to a movement trajectory. It sounded like a fun project to write a bit of code to try out some ideas. This also coincided with my teaching at the AniMove course, so I decided to wrap the code up into an R package to present during the course. We were pleasantly surprised by the enthusiastic reception of the students there. Realizing how broadly applicable the package could be, we decided to turn it into a paper.
took the example of Leo, a turkey Vulture, from a data set available
on Movebank. Leo migrates
between breeding grounds in Saskatchewan, Canada and wintering
grounds in Venezuela. Given the foraging ecology of vultures as
scavengers who use roads to locate road kills, we examined the
correlation of number of revisits to a location’s distance to
nearest road. Next we focused on the highly visited sites (i.e. nests
or roosts), and examined the temporal patterns of visits to those
locations at several scales. At a seasonal to interannual scales,
different sites were important in different years and multiple sites
were used at different but overlapping times within a year. At
shorter scales, visits partitioned into short‐duration
non‐overnight visits and longer duration overnight visits with a
departure around dawn. The package also enables detailed visit-level
analysis beyond the spatial pattern of revisits, including time of
day, duration, and time between visits.