Skip to main content
Dryad logo

Three-dimensional movement of the beak during seed processing in domestic canaries

Citation

Mielke, Maja; Van Wassenbergh, Sam (2022), Three-dimensional movement of the beak during seed processing in domestic canaries, Dryad, Dataset, https://doi.org/10.5061/dryad.zw3r2289z

Abstract

Many songbird species rely on seeds as a primary food source and the process of picking up, positioning, cracking, dehusking, and swallowing seeds is one of the most sophisticated tasks of the beak. Still, we lack understanding about how granivorous songbirds move their beak during the different phases of seed processing. In this study, we used multi-view high speed imaging to analyze the three-dimensional movement of the beak in feeding domestic canaries. Our analysis focuses on correlation of upper and lower beak, frequency of mandibulation, and direction of mandible movement in 3D space. We show that the correlation of maxilla and mandible movement differs among the phases of seed processing. Furthermore, we found that the beak moves at extremely high frequencies, up to 25 Hz, which resembles previously reported maximal syllable rates in singing canaries. Finally, we report that canaries use specific 3D mandible movements during the different phases of seed processing. Kinematic parameters do not differ between male and female canaries. Our findings provide an important biomechanical basis for better understanding the beak as a functional tool.

Methods

FileD1_BeakMovementData.csv

These data were obtained via digitisation of multi-view high speed (500 fps) videos of domestic canaries feeding on hemp seeds.
The upper (maxilla) and lower (mandible) beak and head of the birds were marked with black dots and these markers were tracked in XMALab to extract 3D coordinates.

Data clouds have been rotated in a way that the coordinate axes correspond to the anatomical axes of the head (x = medio-lateral axis, y = anteroposterior axis, z = dorso-ventral axis).

Head movement of the bird has been eliminated by aligning the head markers of each frame with a common reference. Consequently, all remaining movement is only due to maxilla- and mandible movement.  Marker coordinates were merged per rigid body (maxilla & mandible) via averaging to receive one data point per rigid body per frame. The columns ΔX, ΔY, and ΔZ of maxilla and mandible describe the displacement of upper and lower beak relative to the closed-beak condition.

For more detailed information, please consult the Materials & Methods section of the main article or the README file.

 

FileD2_KinematicsData.csv:

These are kinematic parameters (correlation coefficient, mandible frequency, and percental direction of mandible movement) derived from the raw data in FileD1_BeakMovementData.csv. Parmeters were calculated for each individual phase ('positioning', 'biting', 'dehusking', and 'swallowing') of the feeding events.

This is the input file for the statistical analysis script (FileD3_statistics.py).

For more detailed information, please consult the Materials & Methods section of the main article or the README file.

 

FileD3_statistics.py

This is a statistics script used for the statistical analysis of the kinematic parameters (FileD2_KinematicsData.csv). It contains the implementation, sampling, and comparison of the statistical models defined in Table 1 of the main article. The output of this file is the result of the model comparison.

Usage Notes

FileD1_BeakMovementData.csv and FileD2_KinematicsData.csv:

Can be opened with any text editor (e.g., Sublime Text) or spreadsheet program (e.g., LibreOffice Calc).

FileD3_statistics.py:

Script can be opened with any text editor (e.g., Sublime Text). To run the script, Python 3 needs to be installed as well as several (free) libraries specified at the beginning of the script.

 

Funding

Fonds Wetenschappelijk Onderzoek, Award: 1113521N

Fonds Wetenschappelijk Onderzoek, Award: 1505819N

Universiteit Antwerpen, Award: SEP BOF FFB190380

Universiteit Antwerpen, Award: FFB180316