Transmission of maternal aggression in canaries requires direct mother-daughter interaction
Description
This dataset supports a study on the transgenerational transmission of maternal aggression (MA) in domestic canaries (Serinus canaria). MA was scored as a binary, nest-level phenotype: a female was classified as aggressive (yes) if she plucked the feathers of at least one nestling during the standard observation period, and non-aggressive (no) otherwise. The data were collected at the University of Antwerp (Belgium) across four consecutive breeding seasons (2022–2025). The study combined breeding under natural (non-cross-fostered) conditions with a cross-fostering design, in which complete clutches were swapped between nests shortly after clutch completion so that biological and foster mothers could be assigned independent MA phenotypes. This design separates the biological-mother pathway (genetic inheritance and prenatal maternal effects) from the foster-mother pathway (the postnatal rearing environment), allowing their respective contributions to daughter MA to be disentangled. The dataset (GarciaCo_2026_MA_TT_data.xlsx) contains three sheets: - CF_daughters (n = 44): one row per cross-fostered daughter, with the MA phenotypes of her biological and foster mothers, her own MA phenotype when she later bred, whether she was herself plucked as a nestling, and covariates (brood size, breeding year). Primary dataset for the cross-fostered transmission analysis and the concordant subset of the comparison analysis. - NonCF_daughters (n = 30): one row per non-cross-fostered daughter, for whom the biological and rearing mother are the same individual. Includes the mother's MA phenotype, the daughter's own MA phenotype, her origin (2022 baseline cohort or repeated-breeding cohort), and covariates. Primary dataset for the non-cross-fostered transmission analysis. - Consistency_females (n = 32): one row per female recorded across two consecutive breeding seasons (a cross-fostered season followed by a non-cross-fostered season), used to assess within-female consistency of MA. Full variable definitions, sample-size derivations, and analysis notes are provided in the accompanying README file (GarciaCo_2026_MA_TT_README.txt). Missing values (NA) indicate variables that were not applicable or could not be recorded; rows for which MA could not be assessed (no nestling surviving to day 14) were excluded from the analysis datasets. The annotated analysis code (GarciaCo_2026_MA_TT_analysis.Rmd) reproduces every statistical result reported in the associated manuscript from this data file.
Files
Steps to reproduce
All statistical results reported in the manuscript can be reproduced from the data file using the annotated R Markdown script provided. Requirements: - R (version 4.0 or later). - R packages: readxl, dplyr, broom. - The data file GarciaCo_2026_MA_TT_data.xlsx and the script GarciaCo_2026_MA_TT_analysis.Rmd, placed in the same working directory. Procedure: 1. Download GarciaCo_2026_MA_TT_data.xlsx and GarciaCo_2026_MA_TT_analysis.Rmd into a single folder. 2. Open the .Rmd file in RStudio (or run it with R), and set the working directory to that folder. 3. Install the required packages if needed: install.packages(c("readxl","dplyr","broom")). 4. Run the script (Knit, or source it chunk by chunk). It reads the three sheets of the Excel file and codes all yes/no phenotypes as factors (reference level "no") and breeding year as a factor. The script reproduces, in order: - The cross-fostered transmission model (additive effects of biological- and foster-mother MA phenotype on daughter MA), together with the interaction model, which is reported as non-estimable due to complete separation. - The non-cross-fostered transmission model (effect of mother MA on daughter MA), and the observed proportions of aggressive daughters by mother phenotype. - The comparison of non-cross-fostered and concordant cross-fostered daughters (n = 65), including the likelihood-ratio test of the mother-MA-by-group interaction reported in the manuscript. - The within-female consistency model across the two breeding seasons. - The supplementary individual-level exposure models, in which the foster mother's nest-level phenotype is replaced by whether each daughter was herself plucked as a nestling. Each code chunk is annotated with a short description of what it does and which result it corresponds to. Model coefficients, standard errors, z- and p-values printed by the script match the values reported in the manuscript's tables and text.
Institutions
- University of AntwerpFlanders, Antwerp
- Ghent UniversityFlanders, Ghent