Supporting information for "Data aggregation blurs inferred temporal trends in bird sampling”

This Supporting Information accompanies our Matters Arising and provides detailed methods, results, diagnostics, and figure outputs underlying the analyses reported in the main text.

When referring to or reusing these materials, please cite the corresponding preprint and this repository ¹, which contains all code, data, and outputs used in this replication.

Content

Supplement with code

original_paper/ - Contains the original file structure provided by the authors, including Data and Code subfolders downloade from the authors' repository. Due to copyright restrictions, these subfolders are not redistributed here and need to be obtained directly from the authors’ repository. The authors' article and its supplements are available at: https://www.nature.com/articles/s41562-023-01688-5

Data - Stores additiona datasets in subfolders:

• provided/ with data shared by the authors upon request via the Institute for Replication.
• MaPe/ with datasets generated by our scripts: files prefixe DAT_glmmTMB store fitted glmmTMB models, DAT_obs_year_polygon.zip containing DAT_obs_year_polygon.csv (annual observation counts per polygon), and DAT_obs_year_polygon_source.csv providing the same data partitioned by observation source. The latter two datasets were generated uisng rev_Dat_temporal_trend_v1.R. Note, the ZIP needs unpacking for dependent scripts to run correctly. Files prefixed DAT_bam store fitted bam models and are hosted at Open Science Framework repository due to GitHub file size limits (100MB) and stored in the MaPe folder for the scripts to run; these must be downloaded and placed in the MaPe folder for full reproducibility.

R - scripts used in the analysis:

• _runRmarkdown.R generates the HTML Supporting Information from the following R-scripts:
• rep_SI_v5.R contains all code used to generate the paper and Supporting Information outputs.
• rev_Dat_temporal_trend_v1.R is a modified authors' script that extracts polygon-level yearly observation counts. #' Outputs - separate files of all outputs used in our manuscript, including the HTML Supporting Information and Model_ass subfolder with diagnostic outputs for all fitted models.

LICENSE - terms of reuse.

---

(1) Martin Bulla & Peter Mikula (2026). Supporting information for 'Data aggregation blurs inferred temporal trends in bird sampling', GitHub, https://martinbulla.github.io/MA_NHB/;