Beyond the enclosure: A decade of monitoring reveals altered traits in a European ground squirrel colony with implications for a recovery program
Data files
Mar 18, 2026 version files 49.68 KB
-
Data_10.1002wlb3.01647.xlsx
45.18 KB
-
README.md
4.50 KB
Abstract
Effective integrated conservation programs critically depend on high-quality animals from ex situ breeding, specifically those that are genetically diverse and physically healthy, yet insufficient scientific rigor in husbandry protocols can lead to unintended, detrimental consequences that compromise reintroduction success. Drawing on over a decade of long-term monitoring data for the European ground squirrel (Spermophilus citellus), we compared a semi-captive, predator-free and supplementary-fed colony with an adjacent free-ranging population to assess the effects of enclosure management on key population traits such as body mass and female reproduction. We found that the semi-captive system, despite one of its goals of producing animals for release, unintentionally altered traits in ways that could negatively impact its long-term viability and condition of released animals, particularly in terms of body mass and reproductive success While males tended toward higher body mass (though not significantly), only non-lactating females exhibited a statistically significant difference, with a mean increase of 28.4% compared to free-ranging counterparts. Conversely, juveniles born inside the enclosure were significantly lighter, suggesting they might be outcompeted by adults for limited resources such as food and space. Furthermore, a high proportion of non-lactating females inside the enclosure suggests that either high density or surplus food resources may negatively affect breeding success, though the distinct ecological impacts of these factors require further investigation. Practitioners need to implement science-based husbandry protocols and develop feeding strategies that minimize adult-juvenile competition, emphasizing that these recommendations are derived from empirical evidence presented in this study. A rigorous, integrated approach including systematic long-term monitoring must be adopted to mitigate unintended negative consequences, thereby maximizing conservation outcomes and preventing resource waste, with particular attention to adaptive management informed by ongoing scientific evaluation.
Dataset DOI: 10.5061/dryad.jdfn2z3rm
Description of the data and file structure
Dataset overview
This dataset contains demographic, morphometric, and spatial data collected during a 12-year monitoring period (2011–2022) of the European ground squirrel (Spermophilus citellus; EGS) at the Prague Zoo. The data support an analysis of a conservation project aimed at establishing a free-ranging colony of this critically endangered species through a managed enclosure and subsequent "soft-release" into the surrounding steppe habitat. The study focuses on a colony established between 2006 and 2011 with 73 founder animals from four natural populations in the Czech Republic. The dataset includes a supplementary-fed population within a robust enclosure and a self-sustaining, free-ranging population. Data points cover both the managed "inside" population and the "outside" population that colonized the zoo's perimeter and adjacent animal pens. The dataset includes records for 242 transponder-marked individuals and additional unmarked animals. Key variables include: (1) capture data (date, location and age), (2) biological metrics (sex, age class and body mass), and (3) reproductive status of females. Statistical analyses (GLMMs and LMMs in R) were employed to compare the two sub-populations. The data reveal how different environmental conditions, specifically the contrast between managed care and natural exposure, affect body mass trends, recapture probabilities, and female reproductive success. While the populations remained largely distinct with limited interchange, the dataset documents the successful expansion of the species into the surrounding landscape, providing a model for future soft-release reintroduction strategies.
Files and variables
File: Data_10.1002wlb3.01647.xlsx
List Capture age summary
- Year
- Age
- Number
List Capture condition summary
- Year
- Location
- Number
List Recaptures
- Year
- Sex (m = male, f = female)
- Age (ad = adult, juv = juveniles
- Location
- ID = animal transponder number
- Recapture (1 = captured, 0 = not captured)
List Female mass and reproduction
- Year
- Location
- Label = animal shortcut
- ID = animal transponder number
- Sex (f = female)
- Reproduction (Lactating, Non-lactating)
- Reproduction1 (1 = Lactating, 0 = Non-lactating)
- Body mass (g) = body mass in grams
List Male mass
- Year
- Location
- Label = animal shortcut
- ID = animal transponder number
- Sex (m = male)
- Body mass (g) = body mass in grams
List Juvenile mass
- Year
- Sex (m = male, f = female)
- Location
- Body mass (g) = body mass in grams
List Adult sex ratio
- Year
- Sex (1 = male, 0 = female)
- Location
- ID = animal transponder number
List Juvenile sex ratio
- Year
- Sex (1 = male, 0 = female)
- Location
List Adult mass trends
- Year
- Sex
- Location
- Body mass (g) = body mass in grams
List Juvenile mass trends
- Year
- Sex
- Location
- Body mass (g) = body mass in grams
List Female reproduction trends
- Year
- Captured enclosure
- Captured outside
- Reproduced enclosure
- Reproduced outside
- Reproduced enclosure % - proportion of reproducing females in percentages
- Reproduced outside % - proportion of reproducing females in percentages
Code/software
Code/Software
Required Software To view, process, and replicate the analyses in this dataset, the following open-source software is required:
- R Statistical Software: Version 4.2.2 or higher (R Core Team).
- Integrated Development Environment (IDE): RStudio is recommended for managing the provided scripts and workspace.
R Packages The analysis relies on several specific R libraries. Users should ensure the following packages (and their dependencies) are installed:
- Data Manipulation & Visualization:
tidyverse(specificallydplyrfor data cleaning andggplot2for graphical visualization). - Statistical Modeling:
lme4(Linear Mixed-Effects Models),lmerTest(p-value calculation for LMMs), andemmeans(post-hoc pairwise comparisons and Tukey adjustments). - Model Diagnostics:
DHARMa(residual diagnostics for hierarchical models) andnipnTK(specifically for sex ratio and data utility tests).
