Data and code from: Consistent intra-individual seed preferences, yet no dietary conservatism in a granivorous songbird
Data files
Jul 23, 2026 version files 23.45 KB
-
2855_tab.txt
366 B
-
3024_tab.txt
357 B
-
3058_tab.txt
354 B
-
3113_tab.txt
417 B
-
analysis.R
9.08 KB
-
consistency.csv
4.45 KB
-
novel_seed.csv
3.26 KB
-
README.md
5.16 KB
Abstract
Birds, like most other animals, face challenges in finding food in ever-changing habitats where food availability can vary greatly. Given the fluctuating quantities and varieties of accessible food sources, the ability to explore and exploit alternative, perhaps completely new, food sources would be of great benefit. Despite the benefits of such generalist strategies, there is often foraging specialization, which may allow greater efficiency in accessing and processing a specific food source, and it helps reduce competition. However, how specialisation arises and whether it implies that specialists avoid novel foods (dietary conservatism), even though these new foods may provide benefits, is poorly studied. This study therefore examines whether domestic canaries are consistent in their seed choice and whether they tend to avoid a novel seed type, in this case, white millet. To this end, food preferences were investigated in a series of individual-based seed preference tests. Most individuals exhibited consistent seed preference patterns over time. Seed choice appeared to be a repeatable behaviour, and seeds picked are hence not chosen at random. Subsequently, when exposed to a novel seed type (white millet), individuals did not show any dietary conservatism. On the contrary, individuals familiar with white millet picked a significantly lower amount of that seed than individuals who were not familiar with the seed. This suggests that individuals with more experience with this seed type may have learned that its foraging is not very efficient. Taken together, our results show that being consistent in dietary choices does not necessarily relate to dietary conservatism.
Dataset DOI: 10.5061/dryad.jsxksn0s2
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:
'Consistent intra-individual seed preferences, yet no dietary conservatism in a granivorous songbird'.
The data includes seed count data obtained from manual video analysis of seed preference tests, as well as the code for statistical analysis. In summary, this dataset includes two .csv files of seed count data, one for the consistency test and one for the novel seed test, an R-file with the code for statistical analyses and additional associated .txt files (see below; code/software).
Files and variables
consistency.csv:
This file contains seed count data of the seed consistency test.
Terminology:
- Bird: identification number of the individual
- Rep: Indicates the trial number for each individual’s repeated seed preference test (1 = first test, 2 = second test, 3 = third test). Each individual appears three times in the dataset, once for each replication
- Sex: sex of the individual (M = male, F = female)
- Date: date of the day the experiment was conducted on
- Canary: number of canary seeds chosen during the respective trial (in N)
- Hemp: number of hemp seeds chosen during the respective trial (in N)
- Flax: number of flaxseeds chosen during the respective trial (in N)
- Rape: number of bird's rapeseeds chosen during the respective trial (in N)
- Niger: number of nigerseeds chosen during the respective trial (in N)
- Oat: number of peeled oat seeds chosen during the respective trial (in N)
- Total_Amount: total number of seeds chosen during the respective trial (in N)
novel_seed.csv:
This file contains seed count data of the novel seed test.
Terminology:
- Date: date of the day the experiment was conducted on
- Bird: identification number of the individual
- Group: indicates whether the individual belongs to the control groups which received the standard diet (c), or the experimental 'millet' group which received a diet complemented with white millet seed (Mi).
- Film: Indicates the trial number for each individual (1 = first test, 2 = second test).
- GroupFilm: Combined indicator of experimental group and test sequence (Sc1 = short control group, first test; Sc2 = short control group, second test; Lc1 = long control group, first test; Mi1 = 'millet' group, first test; Mi2 = 'millet' group, second test)
- Sex: sex of the individual (M = male, F = female)
- Canary_seed: amount of canary seeds chosen as proportion of the total number of seeds chosen during the respective trial
- Hemp: amount of hemp seeds chosen as proportion of the total number of seeds chosen during the respective trial
- Flaxseed: amount of flaxseeds chosen as proportion of the total number of seeds chosen during the respective trial
- Bird's_rapeseed: amount of bird's rapeseeds chosen as proportion of the total number of seeds chosen during the respective trial
- Nigerseed: amount of nigerseeds chosen as proportion of the total number of seeds chosen during the respective trial
- Oat: amount of peeled oat seeds chosen as proportion of the total number of seeds chosen during the respective trial
- Millet: amount of white millet seeds chosen as proportion of the total number of seeds chosen during the respective trial
- Total_Amount: total number of seeds chosen during the respective trial (in N)
2855_tab.txt, 3024_tab.txt, 3058_tab.txt, 3113_tab.txt:
Terminology:
- Bird: identification number of the individual
- Rep: Indicates the trial number for the repeated seed preference test (1 = first test, 2 = second test, 3 = third test).
- Seed: Indicates one of the six seed types used in the preference test.
- Value: Proportion of the respective seed chosen in the respective trial (value between 0 and 1)
analysis.R:
Code/software
All statistical analyses were conducted in R (version 4.3.3).
This dataset includes an R-file with the full script for the statistical analyses and generation of graphs. Additionally, the .csv/.txt-files of the relevant data included in this dataset are ready-to-use in the R-file.
The required packages are:
- 'rptR' for calculating repeatabilities
- 'DHARMa' for checking model assumptions
- 'ggplot2' and 'viridisite' for graph generation
- 'lme4' and 'lmerTest' for generalized linear mixed modelling
Access information
All data is extracted from video recordings which are not publicly available due to their excessive combined file size. These videos can be accessed upon request by contacting the corresponding author.