Homing through ecological barriers in Balkan pond turtles
Data files
Mar 11, 2025 version files 425.21 KB
-
clusters.ref.df.csv
245.72 KB
-
cnl_agamon_buffer_Uniondis.zip
108.14 KB
-
data_sup_for_paper.xlsx
60.88 KB
-
metrics.all.csv
1.19 KB
-
origin.df.csv
4.33 KB
-
README.md
4.96 KB
Abstract
Homing behaviour, the tendency to return home from a place outside an individual’s home range, requires an ability to navigate by integrating inputs from natural cues. While navigation has been extensively studied, it remains taxonomically biased, mainly focusing on birds. We used the ATLAS tracking system to test homing behaviour in the Balkan pond turtle (Mauremys rivulata), a semi-aquatic reptile. Thirty-six turtles were displaced up to 2.5 km from their capture location and tracked to test for homing behaviour. Another five individuals were released in their capture location as a baseline for local movement. A correlated random-walk model was developed, simulating movement patterns fitted to match turtle movements. Most displaced turtles successfully returned or headed towards their home range within 1–3 days, traveling significantly greater daily distances Homing turtles showed a more direct, and shorter delay than simulated or undisplaced turtles. Our high-resolution tracking revealed that water conduits served as eco-hydrological barriers, causing substantial delays or preventing turtles from passing through the water. Taken together, we demonstrate the navigation capability of the Balkan pond turtle placing this species as a potential non-avian model system for navigation research and for conservation biology, specifically as an indicator of wetland connectivity.
https://doi.org/10.5061/dryad.p5hqbzkzq
Description of the data and file structure
Localizations from ATLAS transmitters:
Turtles were tagged with ATLAS transmitters, a radio tracking system which provides high resolution localizations in real time using lightweight, low-cost tags. This study is the first to ever deploy ATLAS tags on reptiles, in this case, taking the technology to the extreme (close to the ground or water surface). The sample frequency of the transmitter was configured to 1 Hz to maximize the chance of detection when the individual is above the water surface.
Data filtering and analysis:
Movement analysis procedures were executed in R (R Development Core Team). Tracking data was filtered by velocity (10 m/sec) and Std (20 m). Following this filter, we applied a geographic filter, including the interception of data points, with a water body layer created in ArcGIS 10.8 (Esri). For canals, which are presented in the GIS layer as a line, we added a 20-m buffer to include the entire waterbody. Due to the turtles’ behaviour, wherein they surface for thermoregulation, the data is constructed as clusters of points in these locations. Between these clusters, when turtles are underwater, there is no signal transmission, leading to movement data gaps. Therefore, data was interpreted as a collection of stopping locations and metrics were calculated between them, using a designated R package (ToolsForAtlas by Sivan Margalit). Restricted simulations through waterbodies were made using the SimRiv R package . We examined navigation and movement using the following metrics: distance from home, cumulative daily distance, homing duration (total time to arrival at the home range (days)), straightness (calculated for the displaced turtles as the cumulative distance / displacement distance, as a proxy for directionality of movement). Statistical analysis, including ANOVA and mixed models (turtle ID included as a random effect), was done using JMP statistical software (SAS Institute Inc. 2018).
Files and variables
File: data_sup_for_paper.xlsx
Description: Daily movement of turtles
Variables
- TAG: Transmitter number
- date: Date in current location
- capture.X: longitude (ITM)
- capture.Y: Latitude (ITM)
- daily.dist.from.home: Minimum daily distance from capture location (m)
- daily.dist: daily distance travelled (m)
- is.home: True/False
- days.since.release (nomber of days)
- relocated: Yes/No
- did.return: True/False
File: metrics.all.csv
Description: Data used in script - summary of displaced turtle metrics
Variables:
- id: Chronological ordered id number
- TAG: Transmitter number
- capture.X: East (ITM)
- capture.Y: North (ITM)
- init.dist.from.home: Initial distance from capture location (=release distance) (m)
- min.dist.from.home: Minimum distance from capture location (m)
- did.return: True/False
- straightness.index: Straightness score for movement route (=displacement distance/cumulative distance)
- time.to.return: Duration until arrival (d)
- total.time.tracked: (d)
- dist.traveled: Distance traveled (m)
File: clusters.ref.df.csv
Description: Data constructed using R packages and used in script- information on movement. This file contains information on clusters of coordinates
- id: Chronological ordered id number
- TAG: Transmitter number
- date: Date in current location
- visit_ID: Chronological count of visits in this cluster
- cluster_ID: id of cluster
- cnt: cumulative count of visited cluster
- avgX: average of all x coordinated within a cluster
- avgY: average of all Y coordinated within a cluster
- sum.segments: edges within a cluster
- firstTime: entrance to a cluster
- lastTime: exit from a cluster
- dist: distance traveled between clusters (m)
- capture.X: East (ITM)
- capture.Y: North (ITM)
- capture.dist: Distanse from location in cluster to the capture location (m)
File: origin.df.csv
Description: data constructed and used for script - capture site information
Variables:
- id: Chronological ordered id number
- date: Capture date
- capture.lat: latitude (WGS)
- capture.long: longitude (WGS)
- sex: M\F
- transmitter.number: Tag number from Atlas system software use (has prefix code)
- TAG: Tag number
- coords.x1: East (ITM)
- coords.x2: North (ITM)
- capture.Y: North (ITM; same as coords.x2 )
- capture.X: East (ITM; same as coords.x1 )
File: cnl_agamon_buffer_Uniondis.zip
Description: files for canal GIS layer - shp file used for script
Code/software
Movement analysis procedures including simulations were executed in R (R Development Core Team). R scripts are provided here with a water body layer created in ArcGIS 10.8 (Esri). - used for restricted simulation.