telem_assembly: harmonizing wildlife telemetry data (version 1.0.0)

This software was developed using the statistical program R (R Core Team 2024) to compile multiple wildlife tracking datasets that contain locations of animals recorded through time. The software is organized through a series of scripts that run hierarchically through script sourcing. To demonstrate how the software works, we simulated four wildlife tracking datasets, each representing different dataset scenarios, including different types of errors, different data attributes, and different structural characteristics. We then apply the software to the four datasets, combining them into a single coherent database object with a set of standardized database fields that can be used for querying and filtering. Our software runs the following pipeline: (1) formatting individual datasets to follow a common structured template; (2) merging datasets into a single database object; (3) error checking database; and (4) filtering database. It is critical to note that the software scripts, particularly for data formatting, will require modification by end users to customize code to fit real dataset scenarios. Therefore, our software should be thought of as a scripting template for demonstration purposes.

One important attribute of avian telemetry datasets that include track locations recorded during the breeding period is the presence of nest locations embedded in the animal tracks. Individuals with nests will repeatedly visit these static locations as they incubate the eggs and, in the case of altricial species, feed and brood the young. These locations may be very important to identify because they provide information on breeding areas which may be of specific conservation interest, and because they are locations that represent distinct behavioral states of the animal. Our compilation pipeline includes some additional capabilities to isolate and identify these locations, although the processing steps will depend on specific characteristics of each dataset. For example, telemetry datasets that record nest locations in a separate file require a merging process, but nest locations might also be unknown, which requires a predictive modeling approach (Picardi et al. 2020). Our software demonstrates ways to handle four distinctly different types of data scenarios when telemetry tracking datasets contain information on nest locations. We represent these four scenarios by simulating datasets at four hypothetical study areas that correspond to each scenario (i.e., sites A, B, C, and D).