Texas tortoise microhabitat associations in a southern Texas coastal rangeland
Data files
Jul 10, 2026 version files 141.86 KB
-
G_Berlandieri_Microhabitat_R_Code_20260125.R
34.33 KB
-
LMM_data_ms_2026.csv
16.59 KB
-
microhabitat_data_ms_2026.csv
70.49 KB
-
README.md
20.44 KB
Abstract
This data was used assessed Texas tortoise (Gopherus berlandieri) microhabitat use and its relationship with carapace temperature during the Texas tortoise active (April–October) and inactive seasons (November–March) between January 2023–February 2024 in a thornscrub savanna in southern Texas, USA. The data collected include vegetation ground cover and overhead canopy cover data at locations of telemetered tortoises and paired random locations. Additionally, carapace temperature data (degrees Celsius) was collected using iButton temperature loggers adhered to a rear marginal scute of each tortoise. We compared microhabitat variables by location type to determine if tortoises associated with certain cover types compared to what is available. Then, we assessed the relationship between microhabitat variables at tortoise locations and carapace temperature using linear mixed models.
Dataset DOI: 10.5061/dryad.hmgqnk9zm
Description of the data and file structure
This dataset was collected to assess microhabitat use by Texas tortoises (Gopherus berlandieri), tracked via radio telemetry, during different seasons of activity on a coastal rangeland in southern Texas. The information was collected to specifically examine the vegetation cover types the tortoises are using (locations where tortoises were observed) and how it compares to the cover types available to them (random locations), as well as the relationship between carapace temperature and microhabitat cover variables at tortoise locations.
Files and variables
There are two csv data files for analysis.
The first file is named microhabitat_data_ms_2026.csv and contains microhabitat data collected at tortoise locations and at random locations during unique tracking occasions for different individual tortoises across seasons of activity. These data were used in Wilcoxon rank sum, non-metric multidimensional scaling, and permutational analysis of variance tests. The variables include the following:
| Variable | Description |
|---|---|
| Date | Date data was collected in YYYY-MM-DD format |
| Time | Time of day of data collection in 24-hour format |
| CarapaceTemp | Carapace temperature recorded at time of data collection in degrees Celsius. Collected with an iButton temperature logger adhered to a rear marginal scute on the carapace of the tortoise. For random locations, where carapace temperature does not exist, or for tortoise locations where carapace temperature data was missing, cells are left blank. |
| Plot | Plots refer to predefined areas within the study area. This column provides a plot identifier for the location in which the data was collected for that instance. Data collected within predefined areas are indicated by a number assigned to each predefined area, while data collected outside of predefined areas are indicated by "NA" in the column. This variable should be treated as a factor/categorical variable. Here, "NA" counts as a data value (i.e. category) and should not be removed from analysis. |
| Season | Tortoise activity season in which the data was collected. Tortoises have two behavioral seasons of activity: the active season (when tortoises move around, breed, etc.; approximately April-October) and the inactive season (when tortoises enter a state of rest called brumation during the coldest months of the year, approximately November-March). This variable specifies in which season of activity the data was collected. |
| ActivityStatus | Applies to active season data only. This indicates whether a tortoise was observed to be active or inactive at the time of data collection. Tortoises walking, feeding, interacting with other tortoises, or otherwise out from under cover were considered active, and tortoises resting in pallets or under cover were considered inactive (Rose and Judd 2014). For inactive season data, cells are left blank, as tortoises were always inactive during those periods. |
| PointType | Indicates whether data was collected at a tortoise location, indicated by "Tortoise", or a paired random random location, indicated by "Random". Data was collected at two paired random locations for each tortoise location. |
| Canopy | Percentage overhead canopy cover, measured photographically to the nearest whole number using the mobile application Percentage Cover (Mignanelli 2018) with the camera pointed skyward 1 meter above the ground |
| Grass | Percentage of ground covered by grass within a 1 meter square quadrat, estimated to the nearest 5%. |
| Forb | Percentage of ground covered by forbs within a 1-meter square quadrat, estimated to the nearest 5%. |
| Woody | Percentage of ground covered by woody-stemmed vegetation within a 1-meter square quadrat, estimated to the nearest 5%. |
| Litter | Percentage of ground covered by litter within a 1 meter square quadrat, estimated to the nearest 5%. |
| Bare | Percentage of ground covered by bare ground within a 1-meter square quadrat, estimated to the nearest 5%. |
| Tortoise_ID | Unique identifier indicating which individual tortoise in the study the line of data is associated with |
The second CSV file is used for linear mixed model analysis examining the relationship between tortoise carapace temperature and microhabitat variables. This file contains data collected at tortoise locations only that contained carapace temperature measurements. This file is named LMM_data_ms_2026.csv and contains the following variables:
| Variable | Description |
|---|---|
| Date | Date data was collected in YYYY-MM-DD format |
| DailyAvgTempC | Average temperature for that date in degrees Celsius, retrieved from the Port Mansfield weather station |
| Time | Time of day of data collection in 24-hour format |
| Hour | Numeric hour of day based on 24-hour format |
| CarapaceTemp | Carapace temperature recorded at time of data collection in degrees Celsius. Collected with an iButton temperature logger adhered to a rear marginal scute on the carapace of the tortoise. |
| Plot | Plots refer to predefined areas within the study area. This column provides a plot identifier for the location in which the data was collected for that instance. Data collected within predefined areas are indicated by a number assigned to each predefined area, while data collected outside of predefined areas are indicated by "NA" in the column. This variable should be treated as a factor/categorical variable. Here, "NA" counts as a data value (i.e., category) and should not be removed from analysis. |
| Season | Tortoise activity season in which the data was collected. Tortoises have two behavioral seasons of activity: the active season (when tortoises move around, breed, etc.; approximately April-October) and the inactive season (when tortoises enter a state of rest called brumation during the coldest months of the year, approximately November-March). This variable specifies in which season of activity the data was collected. |
| ActivityStatus | Applies to active season data only. This indicates whether a tortoise was observed to be active or inactive at the time of data collection. Tortoises walking, feeding, interacting with other tortoises, or otherwise out from under cover were considered active, and tortoises in pallets or under cover were considered inactive (Rose and Judd 2014). For inactive season data, cells are left blank, as tortoises were always inactive during those periods. |
| Canopy | Percentage overhead canopy cover, measured photographically to the nearest whole number using the mobile application Percentage Cover (Mignanelli 2018) with the camera pointed skyward 1 meter above the ground |
| Grass | Percentage of ground covered by grass within a 1-meter square quadrat, estimated to the nearest 5%. |
| Forb | Percentage of ground covered by forbs within a 1 meter square quadrat, estimated to the nearest 5%. |
| Woody | Percentage of ground covered by woody-stemmed vegetation within a 1-meter square quadrat, estimated to the nearest 5%. |
| Litter | Percentage of ground covered by litter within a 1-meter square quadrat, estimated to the nearest 5%. |
| Bare | Percentage of ground covered by bare ground within a 1-meter square quadrat, estimated to the nearest 5%. |
| Tortoise_ID | Unique identifier indicating which individual tortoise in the study the line of data is associated with |
Daily average temperature data was acquired from the Port Mansfield, Texas weather station and then converted from degrees Fahrenheit to degrees Celsius. The data was retrieved at the following link: https://www.weather.gov/wrh/Climate?wfo=bro
Code/software
Data were analyzed using RStudio and Program R, version 4.1.2. An R Script file is included that runs through the analyses the data were used for. The R script file is named:
G_Berlandieri_Microhabitat_R_Code_20260125.R
Notes and steps for analysis are included in the code file. The R packages read in for analysis are as follows:
library(ggplot2)
library(gridExtra)
library(tidyr)
library(tidyverse)
library(lme4)
library(readxl)
library(dplyr)
library(readr)
library(vegan)
library(ggvegan)
library(corrplot)
library(writexl)
library(AICcmodavg)
library(MuMIn)
library(lmerTest)
library(arm)
Access information
Other publicly accessible locations of the data:
- not applicable
Data was derived from the following sources:
- Daily average air temperature data was retrieved from the NOWData repository (provided by National Weather Service, National Oceanic and Atmospheric Administration, and U.S. Department of Commerce) for the Port Mansfield, Texas weather station, and then was converted from Fahrenheit to Celsius. This data is accessible at https://www.weather.gov/wrh/Climate?wfo=bro