The MoveTraits Initiative

The MoveTraits Initiative is a team of researchers dedicated to developing and harmonizing movement traits and making them readily available in the MoveTraits database.

Movement traits are well-defined, common, and standardized movement metrics from animal tracking data, such as displacement distances, range sizes and the diurnality index. Unlike morphological traits, which are relatively static within the individual, behavioral traits, including movement traits, are labile, allowing animals to respond to their environment.

MoveTraits is an open access behavioural trait database with the aim to make the behavioral information contained in movement and bio-logging data more readily accessible to facilitate a better understanding of species interactions, ecological dynamics and the role behaviour plays for species’ acclimation and adaptation capacity to environmental variability and long-term change. MoveTraits provides movement metrics summarised from GPS tracks for individual animals at three levels: species, between individual and within individual. For more information see Beumer et al. (2025).

Why does MoveTraits matter?

MoveTraits has three goals:

• Bridge a critical gap in ecological research: MoveTraits aims to facilitate a closer integration of animal behaviour into ecological research, from species interactions to global change ecology and ecosystem functioning in a rapidly changing world.

• Maximise data potential: MoveTraits extracts standardized metrics from animal movement data to make them readily usable and interpretable for researchers, conservationists, and policymakers.

• Support conservation and policy goals: Behavior-based indicators have significant potential to contribute to global biodiversity assessments, for instance as Essential Biodiversity Variables. See for example the MoveBON initiative.

How to participate

MoveTraits currently provides trait data for more than 50 land mammals and more than 100 bird species. To unlock its full potential, it needs to grow considerably in the number of species and individuals included. We invite data owners to contribute their data. We will soon approach owners of data on Movebank with an invitation to join MoveTraits via an automated workflow. If you want to contribute data that are not yet on Movebank, please email us at movetraits@gmail.com.

If your data are not public, access will remain restricted, with only the derived movement traits at the scale of your choice shared in the public MoveTraits database You will also become a coauthor of the database and will thus be attributed each time MoveTraits is used. In its current form, the MoveTraits database already features > 50 co-authors. We will regularly update the database and co-author list, while providing a stable DOI and citation tracking functionality.

The vision: a growing global database

We envision MoveTraits as a “living” database that grows alongside the millions of new animal location records deposited Movebank and other data platforms every day. While we currently focus on selected movement traits based on GPS tracking data, we plan to add workflows to calculate additional movement traits and integrate other types of bio-logging data over time. MoveTraits is currently limited to land mammals and birds, but we strive for an extension into the marine and aquatic realms in the near future.

Workflow

MoveTraits relies on a standard workflow to efficiently process movement data and calculate movement traits. Only data that are publicly accessible or are explicitly shared with MoveTraits by their owners are included.

1. Import

After data are contributed to the MoveTraits database, they will be transferred and run through the MoveTraits calculation pipeline.

2. Data cleaning and quality control

The first step in our pipeline entails steps of data cleaning, standardization and filtering for outliers and duplicates.

3. Trait metric calculation

MoveTraits resamples GPS data to 1 position per hour (1h), 1 position per day (24h), and 1 position per week (7d). These resampled data are used to calculate movement traits in a standardized way. Data recorded at coarser resolutions will only be used for the calculation of traits at lower temporal resolution—for example, 3-hourly fixes will be used to calculate daily, weekly, monthly and annual, but not hourly, metrics.

In its current form, MoveTraits contains five distinct animal movement traits, which are reported at different temporal resolutions (from hourly to annually) and annotated with the spatial location and time window covered.

Displacement distance
Displacement distance is the straight-line distance between two recorded GPS locations. Displacement distances can vary depending on the state of the animal (e.g., hungry or sleepy), the behavioral mode (e.g., foraging, travelling, fleeing), or how the environment is constraining or facilitating movement (e.g., elevation, infrastructure, vegetation density). Displacement distance is an informative ecological trait, representing routine movement distances covered by an animal in an hour or day (Hertel et al 2025), and can encode for seed dispersal distances (Graf et al. 2024), and reveal barrier effects (Xu et al. 2023). Displacement distances are therefore widely used in movement ecological research.

Maximum displacement distance
Maximum displacement distance is the maximum extent of a movement trajectory over a given time period. Unlike displacement distance, the extent is not based on sequential/consecutive fixes, but on the maximum distance between any two fixes within that period. Maximum displacement therefore tells us something about the ranging propensity of an individual.

Range size
The range size summarizes the area used by an animal over a given time period. To calculate (home) ranges, many different estimators exist. We here use simple 95% minimum convex polygons, encompassing 95% of locations in a given time period, and report the area used during a given time period.

Intensity of use
The intensity of use is the rate between a path length and the square root of the area covered during that same time (Almeida et al., 2010). It therefore is an area-standardized measure of the directionality of animal movements (Mumme et al. 2023). Low values close to zero are indicative of straight, directional movements, e.g., migration or dispersal, or a large home range area. High values characterize tortuous movement paths clustered in smaller areas.

Diurnality
We calculate the diurnality index following Hoogenboom (1984) based on hourly displacement distances. The index calculates the ratio of distances travelled during the day versus night (determined based on spatial location on the planet and the day of year). The index ranges from -1 (fully nocturnal) to 1 (fully diurnal).

4. Trait metric summary

MoveTraits uses movement data to calculate movement traits repeatedly at the individual level, but summarises and provides the traits at three different hierarchical levels:

• Within-individual level: the underlying repeated movement trait records for each individual over time. The resulting database is nested for each trait. Each trait contains multiple rows per individual, one for each repeated record of the trait.

• Individual level: repeated individual movement trait records are summarised as mean, median, coefficient of variation, and 5th / 95th percentile at the individual level. The resulting database contains one row per individual.

• Species level: traits are summarised across all individuals of a given species as mean, median, coefficient of variation, and 5th / 95th percentile. The resulting database contains one row per species.

5. Data annotation

For each individual, we annotate traits with the individual's sex, life stage, and body mass, as recorded in Movebank, as well as the study ID, contact person, and citation for the source study.

Animals adjust movement first and foremost to the environment (Nathan et al., 2008). To facilitate ecological insights, measured trait values need to be annotated with environmental conditions. We annotate traits with the spatial location and temporal context to facilitate annotation by the end user.

In the current version of the database, we annotate traits with averaged spatial coordinates. To allow for the sharing of traits at a coarser spatial resolution, we are developing an annotation using standardized spatial grids, as illustrated by the example below for caribou and reindeer (Rangifer tarandus). We plan to implement this gridded annotation in the next update of the database.

The resolution of these spatial annotations varies by the level and metric for each trait calculation:

• At the species level, trait summaries are annotated with standardized grid cells of 10,000 km2 (distance between cell centroids ~100km). The user can therefore assess how much of the species’ range is represented. No recorded GPS locations are revealed.

• At the individual level, trait summaries are annotated with the tracking duration, tracking start and end date, and the grid cells occupied by the individual at a 100km (mandatory) or 10km (optional) spatial accuracy. No recorded GPS locations are revealed.

• At the within-individual level, repeated individual trait measurements are annotated either with the observation scales coordinates where the trait measurement was observed, the 1km grid cell in which the trait measurement was observed, or the 10km grid cell in which the trait measurement was observed. The spatial resolution of these annotations depends on the trait metric: For hourly and daily displacement distance, the observed leading and end fix coordinate are shared. Recorded locations are revealed. For maximum displacement distance, home range size, intensity of use and diurnality, the mean fix coordinate across all fixes taken during the period are shared. No recorded GPS locations are revealed.

Data owners can choose to agree to trait calculations only at the species and individual levels, excluding within-individual level calculations, to avoid publicly sharing recorded species locations. Observation-level GPS locations provide the most flexibility for end users in downstream environmental annotation and analysis. We therefore encourage sharing at this level. However, sharing traits at coarser resolutions is still valuable for studies at broader spatial scales, e.g. annotation with global climate data products.

6. MoveTraits data repository

We will regularly update the database and co-author list, while providing a stable DOI and citation tracking functionality. The MoveTraits database is currently hosted on figshare, we reserve the right to move to a different provider in the future.

Governance, funding and maintenance

MoveTraits is an initiative led by the Implementation Team (see participants) and supported by the Senckenberg Biodiversity and Climate Research Centre (SBiK-F) and the Max-Planck Institute of Animal Behavior Radolfzell (MPI-AB).

References

Concept manuscript (currently undergoing peer review):

Beumer LT, Hertel AG, Royauté R, Tucker MA, Albrecht J, Beltran RS, Cagnacci F, Davidson SC, Dejid N, Kays R, Kölzsch A, Lohr A, Neuschulz EL, Safi K, Scharf AK, Schleuning M, Wikelski M, Mueller T. 2025. MoveTraits – A database for integrating animal behaviour into trait-based ecology. bioRxiv. https://doi.org/10.1101/2025.03.15.643440

Database:

Hertel AG, Beumer LT, Royaute R, Tucker MA, Albrecht J, Beltran RS, Cagnacci F, Davidson S, Dejid N, Kays R, Kölzsch A, Neuschulz EL, Safi K, Scharf AK, Schleuning M, Wikelski M, and Mueller T. 2025. MoveTraits database. Figshare. https://doi.org/10.6084/m9.figshare.28611890.v1

Contact

For more information, contact movetraits@gmail.com.

Participants

Implementation team: Larissa Beumer, Anne Hertel, Anne K Scharf, Kamran Safi, Sarah Davidson, Martin Wikelski, Roland Kays, Francesca Cagnacci, Marlee Tucker, Kevin Frac, Thomas Mueller

Co-authors of initial concept paper: Larissa Beumer, Anne Hertel, Raphaël Royauté, Marlee Tucker, Jörg Albrecht, Roxanne Beltran, Francesca Cagnacci, Sarah Davidson, Nandia Dejid, Roland Kays, Andrea Kölzsch, Ashlee Lohr, Eike Lena Neuschulz, Kamran Safi, Anne K Scharf, Matthias Schleuning, Martin Wikelski, Thomas Mueller

Co-authors of figshare database: Anne Hertel, Larissa T. Beumer, Raphael Royaute, Marlee A. Tucker, Jörg Albrecht, Roxanne S. Beltran, Francesca Cagnacci, Sarah Davidson, Nandintsetseg Dejid, Roland Kays, Andrea Kölzsch, Eike Lena Neuschulz, Kamran Safi, Anne K. Scharf, Matthias Schleuning, Martin Wikelski, Susanne Åkesson, Yuriy Andryushchenko, Mathieu Basille, Guillaume Bastille-Rousseau, Nyambayar Batbayar, Jerrold Belant, Oded Berger-Tal, Dominique Berteaux, Joanna T. Bialas, Jacqueline Bishop, Wayne Boardman, Randall Boone, Andre Botha, Mark Boyce, Evan Buechley, Patrik Byholm, Simon Chamaillé, Peter Desmet, Amanda Droghini, Łukasz Dylewski, Ron Efrat, Kyle H. Elliott, Andrea Flack, Morten Frederiksen, Brock Geary, Robert Heinsohn, Marco Heurich, Ladd M. Irvine, Zuzanna Jagiello, Jana W. E. Jeglinski, Frédéric Jiguet, Matthew Johnson, Scott LaPoint, A. David M. Latham, Leif Egil Loe, U. Lötberg, Mark Mallory, Stephani Matsushima, Erling Meisingset, Candice Michelot, Tanja Milotic, Ronaldo Morato, Atle Mysterud, Rohit Naniwadekar, Sarah Olson, Federico Ossi, Daniel M. Palacios, Simona Picardi, Ivan Pokrovsky, Laura Prugh, Nathan Ranc, Yan Ropert-Coudert, David Roshier, Shay Rotics, Nir Sapir, Daniel Schmidt-Rothmund, Laurel Serieys, Sherub Sherub, Johannes Signer, Orr Spiegel, Richard Temple, Grigori Tertitski, Kasper Thorup, Marcin Tobolka, Anders P. Tøttrup, Cody Wallace, Luke Wilde, Christopher Wilmers, Thomas Mueller

Move BON-MoveTraits working group: Briana Abrahms, Larissa Beumer, Luca Borger, Francesca Cagnacci, Emily Cohen, Eduardo Cuevas, Sarah C Davidson, Diego Ellis Soto, Weijian Gao, Anne Hertel, Lacey Hughey, Roland Kays, Dongmin Kim, Johannes Krietsch, Matthias Loretto, Thomas Mueller, Ruth Oliver, Meredith S Palmer, Laura Pollock, Nicholas Russo, Christian Rutz, Jared Stabach, Stella Uiterwaal, George Wittemyer, Weiqian Xu, Scott Yanco

*Please contact us if you have contributed to the MoveTraits effort and want to be listed as a participant.

Participating studies