Access this dataset on Dryad

The data and code found here were used for the analyses in:

DeNicola, V., Mezzini, S., Cagnacci, F. Monitoring the Effects of Ovariectomy on Seasonal Movement Behavior in Suburban Female White-Tailed Deer Using Internet of Things Enabled Devices. Wildlife Biology. https://doi.org/10.1002/wlb3.01512

The file DeNicola-2025-south-euclid-deer.zip contains:

• A data folder with the original (uncleaned) deer telemetry data (Odocoileus virginianus DeNicola South Euclid.csv), the telemetry metadata (Odocoileus virginianus DeNicola South Euclid-reference-data.csv), the cleaned telemetry data (cleaned-telemetry-data.rds), and the number of daily fixes for each deer (daily-fixes.rds).
• All R code used for the analyses in the analysis folder, except for custom functions, which are in the functions folder.

Description of the data and file structure

All missing values in the two csv files are indicated as NA.

Odocoileus virginianus DeNicola South Euclid.csv

This file contains the telemetry data before data cleaning. Also see the metadata dataset and the cleaned telemetry data below. It contains the following columns:

• event-id: numeric; the fix ID (each value is unique).
• visible: logical; all values are TRUE.
• timestamp: character; date and time of the fixes with format yyyy-mm-ddTHH:MM:SS.000000Z.
• location-long: numeric; longitude of the fix, rounded to two decimals as per Dryad's requirements.
• location-lat: numeric; latitude of the fix, rounded to two decimals as per Dryad's requirements.
• behavioural-classification: character; behavioral classification as one of STOP, START, ON_THE_MOVE, PERIODIC, or NA.
• geolocator-fix-type: character; indicates whether the geolocator fix was obtained with GPS (GPS) or WiFi positioning (wifi); is NA if the fix is sensolus (see sensolus:fix-type below).
• gps:dop: numeric; dilution of precision.
• gps:hdop: numeric; horizontal dilution of precision.
• gps-time-to-fix: numeric; number of seconds to GPS fix.
• sensolus:fix-type: character; indicates whether the Sensolus fix was obtained with GPS (GPS), WiFi positioning (wifi), or the Sigfox network (network); is NA if the fix is with a geolocator (see geolocator-fix-type above).
• sensor-type: character; indicates sensor type (always gps).
• individual-taxon-canonical-name: character; species name (always Odocoileus virginianus).
• tag-local-identifier: character; animal tag identifier (see individual-local-identifier below for animal ID).
• individual-local-identifier: character; animal ID identifier with C indicating control deer and T indicating treated (ovariectomized) deer (see tag-local-identifier above for tag ID).
• study-name: character; all values are Odocoileus virginianus DeNicola South Euclid.

Odocoileus virginianus DeNicola South Euclid-reference-data.csv

This file contains the metadata related to the telemetry data detailed above. It contains the following columns:

• tag-id: character; tag IDs containing all values from the telemetry data, plus tag YZX346 for deer C_103, which had no telemetry data.
• animal-id: character; animal ID identifier with C indicating control deer and T indicating treated deer (see tag-id above for tag ID).
• animal-taxon: character; species (all values are Odocoileus virginianus).
• deploy-on-date: date; dates on which collars were deployed.
• deploy-off-date: date; dates on which collars were removed, if available (NA otherwise).
• animal-comments: character; only non-NA value is Lost collar - out of area for deer C_103.
• animal-death-comments: character; only non-NA value 3/20/23\r\nMortality 5/30/23 for deer T_169.
• animal-mortality-date: date; mortality date if available.
• animal-sex: character; all values f to indicate female deer.
• deployment-comments: character; only non-NA value is Photo of doe without collar from 2/2023 for deer C_103.
• tag-manufacturer-name: character; tag manufacturer (all values are Sigfox).
• tag-serial-no: character; all values are NA except for LGFZWH (same as tag-id) for deer C_101.

cleaned-telemetry-data.rds

This RDS file contains the cleaned telemetry data as created by the analysis/01-telemetry-data-cleaning.R script. Deer C_101 was removed from the telemetry data because she only had 16 fixes. Deer C_103 did not have any fixes. The dataset includes the following columns:

• tag_local_identifier: character; tag ID corresponding to tag-local-identifier in the original telemetry dataset.
• animal: character; animal ID corresponding to individual-local-identifier in the original telemetry dataset.
• group: character; either Control for the control group or Ovariectomy for the treatment (ovariectomized) group corresponding to C and T in the animal ID.
• tel: list; tibbles of the telemetry data for each animal after removing problematic fixes through data cleaning.

Each tibble in the tel column contains the following columns:

• event_id: numeric; the fix ID (each value is unique).
• visible: logical; all values are TRUE.
• timestamp: datetime; date and time of the fixes.
• long: numeric; longitude of the fix, rounded to two decimals as per Dryad's requirements.
• lat: numeric; latitude of the fix, rounded to two decimals as per Dryad's requirements.
• behavioural_classification: character; behavioral classification as one of STOP, START, ON_THE_MOVE, PERIODIC, or NA.
• geolocator_fix_type: character; indicates whether the geolocator fix was obtained with GPS (GPS) or WiFi positioning (wifi); is NA if the fix is sensolus (see sensolus:fix-type below).
• hdop: numeric; horizontal dilution of precision.
• gps_time_to_fix: numeric; number of seconds to GPS fix.
• fix_type: character; indicates whether the Sensolus fix was obtained with GPS (GPS), WiFi positioning (wifi), or the Sigfox network (network); is NA if the fix is with a geolocator (see geolocator-fix-type above).
• sensor_type: character; indicates sensor type (always gps).
• species: character; species name (always Odocoileus virginianus).
• individual_local_identifier: character; animal ID identifier with C indicating control deer and T indicating treated (ovariectomized) deer (see tag-local-identifier above for tag ID).
• study_name: character; all values are Odocoileus virginianus DeNicola South Euclid.
• outlier: numeric; either 0 (to keep) or 1 (to remove before model fitting).

daily-fixes.rds

This RDS file contains the number of fixes for each deer on each day between the start and end of the tracking periods. It contains the following columns:

• group: factor; either Control for the control group or Ovariectomy for the treatment (ovariectomized) group corresponding to C and T in the animal ID.
• animal: character; character; animal ID corresponding to individual-local-identifier in the original telemetry dataset.
• deploy-on-date: date; dates on which the collar was deployed.
• date: date; date of reference.
• daily_fixes: numeric; number of fixes on the date of reference (date).
• doy: numeric; day of year of the date of reference (between 1 and 366).
• doy_cr: numeric; same as doy above, used for non-cyclical cubic regression splines in the hierarchical generalized additive model (see analysis/03-fixes-per-day-hgam.R).
• animal_year: factor; the animal ID (animal) along with the year (starting in September) to account for differences across years (2023 and 2024).

Sharing/Access information

The data and code, along with all related figures and models, are also available on GitHub at https://github.com/StefanoMezzini/south-euclid-deer-ovariectomy.

Code/Software

All analyses were run in R 4.4.0 along with the creation of all figures. The code used is available in the analysis folder, with the exception of custom functions, which are in the functions folder.

The analysis folder includes the following files:

• 00-data-overview.R: gives an overview of the two csv files and performs some data exploration; creates figure S2.
• 01-telemetry-data-cleaning.R: cleans the telemetry data and saves the output as data/cleaned-telemetry-data.rds.
• 02-fit-movement-models.R: fits the ctmm movement models and AKDEs; creates figure 3.
• 03-fixes-per-day-hgam.R: fits a hierarchical generalized additive model that estimates the difference in the number of daily fixes across groups.
• 04-hgams-without-T_169.R: fits hierarchical generalized additive models for the movement behavior of the deer, excluding deer T_169.
• 04-hgams.R: fits hierarchical generalized additive models for the movement behavior of the deer, including deer T_169; creates figures S4 and S5.
• figures/akde-contours-excursivity.R: creates figure 2.
• figures/check-data-gaps.R: checks gaps in data; creates figure S1 and a figure of T_169's excursions.
• figures/default-ggplot-theme.R: creates the default ggplot theme used for the figures.
• figures/hgam-figures.R: creates figures 4 and S3.
• figures/hgams-with-and-without-T_169.R: compares the models with and without deer T_169 and creates figures S6 and S7.
• figures/study-site-map.R: creates figure 1.
• README.md: a simple README for the folder.

The files in the analysis folder are to be run sequentially following the file name order, while the files in the analysis/figures folder are to be run after scripts 00 to 04. The two 04 scripts can be run in either order.

Movement models were fit using ctmm v. 1.2.0, while generalized additive models were fit using mgcv v. 1.9-1.

The functions contain all custom functions used in this project, namely:

• check_animal.R: creates a telemetry for animal id and calls outlier_plots() (see below)
• find_angle.R: given two vectors of coordinates x and y, finds the turning angle between each group of three temporally consecutive points
• flag_outlier.R: flags one or more points as outliers based on maximum allowed speed (max_speed), maximum allowed median displacement (max_distance), maximum allowed angle (max_angle) = 0, and maximum allowed sampling interval (max_dt). Outliers are indicated by a value > 0.
• get_cis.R: given model m, extracts estimated relative differences between the treatment and control groups, along with 95% confidence intervals and P-values.
• get_legend.R: a modified version of cowplot::get_legend() to extract legends from ggplot plot objects.
• get_preds.R: predicts from the model for a given parameter (parameter) among home range size (hr), diffusion (diff), excursivity (exc), and daily fixes (fixes).
• outlier_plots.R: creates the diagnostic plots used to clean the telemetry data.
• plot_adj.R: plots n_adj before and after fixes flagged based on minimum or maximum: speed, distance from median location (distance), angle, and sampling interval (dt).
• remove_outlier_flags.R: removes all outlier flags by changing all values of the outlier back to 0.
• window_ctmm.R: fits movement ctmm movement models to telemetry .tel using a moving window approach with a window window days and a slide of dt days.