Data from: Beak dimensions affect feeding performance within a granivorous songbird species
Data files
Feb 25, 2025 version files 272.37 MB
-
analysis.zip
60.97 KB
-
image_data.zip
272.14 MB
-
landmarks_curve.csv
41.38 KB
-
landmarks_regular.csv
31.98 KB
-
PC_scores_curve.csv
31.74 KB
-
PC_scores_regular.csv
22 KB
-
README.md
5.24 KB
-
shape_summary.csv
31.42 KB
Abstract
Beaks of granivorous songbirds are adapted to dehusk seeds fast and efficiently. This is reflected in the large variety of beak shapes and sizes among species specialized in different seed types. Generally, larger beaks improve the dehusking of larger seeds by transmitting and withstanding higher bite forces. Meanwhile, smaller beaks are better suited for processing smaller seeds by allowing faster beak movements and better seed handling dexterity. These patterns are presumably the result of a trade-off between force and velocity inherent to lever systems. Since beak shape also varies among individuals of the same species, we investigated whether beak shape relates to variation in feeding performance and beak kinematics in the Domestic Canary (Serinus canaria). We analysed beak morphology of 87 individuals through both traditional size measurements and 3D-landmark analysis to capture metrics such as beak depth, length, width and curvature. We related these metrics of morphology to data on feeding performance and beak kinematics during feeding on smaller canary seeds and larger, tougher hemp seeds. We found that individuals with larger absolute beak depths were faster at dehusking the large seeds. Despite that individuals with shallow or long beaks displayed higher beak opening-closing frequencies, this did not result in a significantly shorter processing time of the smaller seeds. Our data are therefore in line with the presence of a force-velocity trade-off within a species, but without a velocity-related drawback of beak-size adaptations for increased bite force on the handling performance of a smaller and easier-to-crack seed.
https://doi.org/10.5061/dryad.d51c5b0dm
Description of the data and file structure
This dataset contains all the necessary data for reproducing the results of the research, as described in the manuscript with the name: 'Beak dimensions affect feeding performance within a granivorous songbird species'
The data includes beak morphology data obtained from landmark analysis and size measurmeents on calibrated images and data on feeding performance, skills and beak kinematics obtained from Andries et al. (2023).
Files and variables
This dataset includes 5 .csv files with landmark and measurement data, one R-file containing the code for the landmark analysis, one R-file containing the code for statistical analysis and associated .txt files (see below; code/software) and the raw image data used in the anaylses.
image_data.zip: File folder containing the original images used for landmark placement and size measurements, including calibration images (_calib.jpg). Data is ordered per 1 or 2 individuals (often 2 individuals share the same calibration images due to being filmed on the same day). Files are structured to be ready-to-use in XMAlab. Also includes an extra folder with the necessary files required to do the calibration (framespec_obj4_mm.csv & ref_obj4.ref).
landmarks_regular.csv: This file contains the raw coordinates (x,y,z) of the regular landmarks per individual. The first column shows the identification number of the individual.
landmarks_curve.csv: This file contains the raw coordinates (x,y,z) of the curve semi-landmarks, analogous to the landmarks_regular.csv file.
PC_scores_regular.csv: This file contains the PC scores output of the regular landmark analysis (see R-file 'landmark_analysis') per individual.The first column shows the identification number of the individual.
PC_scores_curve.csv: This file contains the PC scores output of the curve semi-landmark analysis, analogous to the PC-scores_regular.csv file.
shape_summary.csv: This file contains the relevant PC scores (from landmark analyses) and measurement data that were used for further analyses, as well as metrics on feeding performance and beak kinematics taken from Andries et al. (2023). Missing values are indicated by 'NA'.
Terminology:
- bird: identification number of the individual
- sex: indicates whether the individual is male (M) or female (F)
- birth_year: the year in which the individual was born
- age_group: indicates whether the individual is a juvenile (young) or adult (adult).
- weight: body mass of the individual at the time of the experiment (described in Andries et al. 2023)
- seed: indicates whether total_duration, success_rate, closing_velocity and frequency are for feeding on either canary seed (canseed) or hemp seed (hemp).
- PC1/PC2/PC3: PC scores of the first, second and third principal components respectively of the regular landmark analysis
- curve_PC1: PC scores of the first principal component of the curve semi-landmarkanalysis
- beak_length: absolute length of the beak (in mm)
- beak_height: absolute height (in the manuscript referred to as 'beak depth') of the beak (in mm)
- beak_width: absolute width of the beak (in mm)
- upper_angle: average angle of the upper beak (in °)
- lower angle: average angle of the lower beak (in °)
- total_duration: Average duration of the entire feeding act (in s)
- success_rate: proportion of seed husking attempts that are successful (in %)
- closing_velocity: average maximum beak closing velocity (in mm/s)
- frequency: average opening-closing frequency of the beak (in Hz)
analysis.zip: contains all the necessary files for the landmark and statistical analyses conducted in R:
- landmarks.txt: Text file containing raw landmark coordinates, ready to use in the R-script for the regular landmark analysis.
- landmarks_curve.txt: Text file containing raw semi-landmark coordinates, ready to use in the R-script for the curve semi-landmark analysis.
- shape_summary.txt: Text file containing the data from shape_summary.csv, ready to use in the R-script for the statistical analyses.
- landmark_analysis.R: R-file detailing the regular and curve semi-landmark analyses.
- shape.R: R-file detailing the mixed linear modelling and construction of graphs used in the manuscript.
Code/software
All statistical analyses were conducted in R (version 4.3.3).
This dataset includes two R-files with the full script for 1) the regular and curve semi-landmark analyses and 2) all statistical analyses and generation of graphs. Additionally, .txt-files of the relevant data are included in this dataset, ready-to-use in the R-file. All these files can be found in the analysis.zip folder.
The required packages are: 'geomorph' for the landmark analyses, 'lme4' and 'lmerTest' for linear mixed modelling, and 'ggplot2' and 'plotly' for graph generation.
Access information
Other publicly accessible locations of the data:
- Not applicable
Data was derived from the following sources:
Image data of birds with their beaks held in a closed resting position were selected from the recordings of Andries et al. (2023).
Image data was calibrated using synchronized images of a calibration object with known dimensions in XMAlab.
Landmarks were annotated on the calibrated images in XMAlab.
3D-coordinates of the landmarks were extracted and used to calculate beak size metrics in Microsoft Excel.
To analyze beak curvature, 3D-coordinates of the landmarks on the upper beak's curve were further processed and analyzed in R (version 4.3.3.).
Morphological data were related to feeding performance, beak kinematics and seed handling skill metrics, directly taken from Andries et al. (2023).