Locomotor performance of the Cerrado lizard Micrablepharus atticolus: effects of body temperature and caudal autotomy

Associated publication

Silva NA, Caetano GHO, Campelo PH, Cavalcante VHGL, Godinho LB, Miles DB, Paulino HM, Silva JMA, Souza BA, Silva HBF, Colli GR. 2021. Effects of caudal autotomy on the locomotor performance of Micrablepharus atticolus (Squamata, Gymnophthalmidae). Diversity 13:562. https://doi.org/10.3390/d13110562

Description of the dataset

This dataset contains individual-level locomotor performance data for the gymnophthalmid lizard Micrablepharus atticolus. The data were collected to evaluate how sprint performance varies with body temperature, body size, body mass, sex, reproductive condition, locality, caudal autotomy, and regenerated tail length.

Lizards were collected from two Cerrado localities in Brazil: Reserva Ecológica do IBGE, Brasília, Distrito Federal, and Parque do Bacaba, Nova Xavantina, Mato Grosso. Individuals were measured morphologically and then tested repeatedly under laboratory conditions across a range of body temperatures. Each row in the data file represents a single sprint trial performed by an individual at a specific body temperature.

The data file deposited here is a harmonized version of the original locality-specific data files. The two original files were merged into a single tab-delimited file, the column order was standardized, and categorical codes were converted to explicit values to improve reusability.

Files included

1. Micrablepharus_atticolus_locomotor_performance_data.tsv

Merged, harmonized tab-delimited data file containing all locomotor performance observations.

• Number of rows: 481 sprint trials
• Number of columns: 13
• Number of individuals: 103 lizards
• Number of localities: 2

Rows by locality:

• Distrito_Federal: 170 sprint trials from 39 individuals
• Nova_Xavantina: 311 sprint trials from 64 individuals

2. Micrablepharus_atticolus_locomotor_performance_analysis.R

R script used to validate the merged dataset, recreate the derived variables, fit the generalized additive mixed models (GAMMs), conduct model selection and model averaging, and export optional summary tables and figures.

3. README.md

This README file.

Data structure

Each row represents one locomotor performance trial.

Each individual appears in multiple rows because each lizard was tested repeatedly at different body temperatures. The variable Identity identifies repeated observations belonging to the same individual and should be used as the grouping variable in repeated-measures or mixed-effects analyses.

Most individuals have five observations. Some individuals have three or four observations because not all planned sprint trials yielded usable records.

Variable dictionary

Detailed variable notes

Identity

Unique identifier for each individual. The same identity can occur in multiple rows because each lizard was tested repeatedly at different body temperatures.

Sex

Sex of the individual, coded as Female or Male.

Gravid

Reproductive condition, coded as No or Yes.

In the original locality-specific files, reproductive condition was coded inconsistently: the Distrito Federal file used N, and the Nova Xavantina file used n for non-gravid and s for gravid. In the merged data file, these values were standardized to No and Yes.

Males are coded as No.

SVL

Snout-vent length in millimeters. This is the standard measure of lizard body size from the tip of the snout to the anterior margin of the cloaca.

CC

Total tail length in millimeters. For individuals with regenerated tails, this variable includes the remaining original portion and the regenerated portion.

Mass

Body mass in grams.

Temperature

Body temperature in degrees Celsius immediately before the sprint trial. Individuals were tested across a range of body temperatures to estimate thermal performance curves.

Speed1

Maximum sprint speed in meters per second for the trial.

Values equal to zero correspond to trials in which the individual did not run at body temperatures near critical thermal limits. These observations were retained because they define the lower and upper ends of thermal performance curves.

RC

Length of the regenerated portion of the tail in millimeters.

Individuals with intact tails have RC = 0. Some autotomized individuals also have RC = 0, indicating that the tail was autotomized but the regenerated portion was absent or too small to measure.

AU

Caudal autotomy status.

• Intact: no evidence of caudal autotomy.
• Autotomized: naturally autotomized tail, with or without measurable regeneration.

All autotomy recorded in this dataset was naturally occurring. Tails were not experimentally removed.

CCr

Relative tail length, calculated as:

CC / SVL

This variable is dimensionless.

RCr

Relative regenerated tail length, calculated as:

RC / SVL

This variable is dimensionless.

Local

Collection locality.

• Distrito_Federal: Reserva Ecológica do IBGE, Brasília, Distrito Federal, Brazil.
• Nova_Xavantina: Parque do Bacaba, Nova Xavantina, Mato Grosso, Brazil.

File format

The data file is a UTF-8-encoded tab-delimited text file with a header row.

Recommended import commands:

In R:

dat <- read.table("Micrablepharus_atticolus_locomotor_performance_data.tsv",
                  header = TRUE, sep = "\t")

In Python:

import pandas as pd
dat = pd.read_csv("Micrablepharus_atticolus_locomotor_performance_data.tsv", sep="\t")

Missing values

The merged data file contains no missing values.

If missing values are introduced in future versions, they should be coded as NA.

Quality-control checks applied to the merged file

The following checks were performed before deposition:

1. The two locality-specific source files were read using their column names, avoiding errors due to their different column order.
2. The merged data file was written with a single standardized column order.
3. The Gravid variable was standardized to No and Yes.
4. The AU variable was standardized to Intact and Autotomized.
5. The Local variable was standardized to Distrito_Federal and Nova_Xavantina.
6. The merged file was checked for missing values; none were present.
7. The derived ratios CCr = CC / SVL and RCr = RC / SVL were checked against the archived values, allowing only negligible rounding differences.
8. The final file was checked for the expected 481 rows, 13 columns, and 103 unique individuals.

Software and code

The accompanying R script requires R and the following packages:

• ggplot2
• mgcv
• MuMIn
• nlme
• psych
• visreg

The script does not install packages automatically. Users should install any missing packages before running the script.

The script performs the following steps:

1. Reads Micrablepharus_atticolus_locomotor_performance_data.tsv.
2. Validates column names, categorical values, missing data, and ratio calculations.
3. Converts categorical variables to factors.
4. Creates a derived variable called Tail_condition for descriptive summaries.
5. Produces summary statistics.
6. Fits a null GAMM and a full GAMM.
7. Computes approximate deviance-based R-squared and adjusted R-squared.
8. Performs a likelihood-ratio test comparing the full and null models.
9. Performs AICc-based model selection and model averaging.
10. Optionally exports figures and results tables.

Derived variables in the R script

The R script creates one derived variable that is not included in the deposited data file:

Tail_condition

This variable has three categories:

• Intact: AU = Intact
• Autotomized_not_regenerated: AU = Autotomized and RC = 0
• Autotomized_regenerated: AU = Autotomized and RC > 0

This variable is used only for descriptive summaries. The primary statistical models use AU and RCr, following the analytical structure of the associated publication.

Relationship to the associated publication

These data were used to evaluate whether caudal autotomy affects sprint performance in Micrablepharus atticolus while accounting for body temperature, reproductive condition, body mass, body size, locality, and regenerated tail length.

The associated publication reports that locomotor performance was strongly related to body temperature, pregnancy, body mass, locality, and the relative length of the regenerated tail, whereas caudal autotomy status itself did not substantially reduce sprint speed.

Sharing and access information

These data are intended for public reuse through Dryad.

The data were not derived from another public dataset.