Multimodal data helps in identifying spatio-temporal patterns and habitat associations of Aquila chrysaetos (Golden Eagle) in Finland
Data files
Oct 12, 2025 version files 2.19 GB
-
README.md
2.75 KB
-
Stage_I_II.zip
2.19 GB
Abstract
Understanding the spatial distribution of individuals is essential for effective species conservation. We investigated the spatio-temporal distribution of Aquila chrysaetos (golden eagle) in Finland using nest surveys, citizen science observations, and environmental data from 1982 to 2021. We extended a popular N-mixture model to estimate population abundance while accounting for the number of recently hatched nestlings, the spatial distribution, and correcting for biases in the observation process, all integrated into a two-stage Bayesian hierarchical model. Results of our model aligned with the previously known north-oriented distribution of A. chrysaetos in Finland, supporting its applicability to A. chrysaetos and other species with similar ecological requirements and life-history traits. Furthermore, we observed the highest densities of successful nests near open landscapes, such as marshes and peat bogs, whereas dense forests, port areas, and areas of high human population density had a strong negative effect. While the temporal trend of nest abundance has remained constant over the years, the overall population abundance showed an increasing trend until 2007, after which it stabilized. Finally, we detected a higher amount of A. chrysaetos movements in southern and southwestern Finland than elsewhere in Finland, opening up new possible areas for conservation targets.
Dataset DOI: 10.5061/dryad.41ns1rnsg
Description of the data and file structure
This repository contains all data, code, and results for the two-stage simulation process described in our study. Each stage is organized into a separate folder with standardized file formats to ensure reproducibility and clarity.
Files and variables
All datasets are structured as matrices, where rows represent spatial sites and columns represent time steps (years). The dataset consists of Stage_I_II.zip folder.
Stage I - Nest counting model
- CountNest.Rdata: Matrix containing the number of nests observed in survey data at each site i and time t.
- Data_1.RData: Matrix of environmental covariates extracted from the CORINE Land Cover map at each site i.
- FMI_Cov.RData: Average temperature data from the Finnish Meteorological Institute for each site i and time t.
- Stage_I_Code.R: R script to:
- Apply the alr (additive log-ratio) transformation to environmental covariates
- Structure the data
- Run MCMC sampling using the NIMBLE package
- Stage_I_MCMC_results.Rdata: An R list object containing:
- My samples: posterior samples from the MCMC
- Code 2: the NIMBLE model specification
Stage II - Population level modelling
- Cov_Data.Rdata: Contains the Hex. G object, used for constructing maps and hexagonal grid structures for Finland.
- FMI_Cov.RData: Average temperature during the breeding season at each site i and time t (reused from Stage I).
- Obs.Data_YearCount_1980-2022.RData: Aggregated citizen science bird observations per site i and time t, spanning 1980–2022.
- Pop_Data.RData. Aggregated human population data from Statistics Finland and the location of observation towers, for each site i.
- Prob_Nest_ArlQ.Rdata: Posterior distribution of nest counts derived from Stage I modeling.
- Stage_II_Code.R: R script to:
- Structure the data
- Build the adjacency matrix for the CAR model
- Run MCMC sampling using the NIMBLE package
- Stage_II_MCMC_results.Rdata: An R list object containing:
- My samples: posterior samples from the MCMC
- Code 3: the NIMBLE model specification
Code/software
To run the code, you will need:
- R (version ≥ 4.0.0)
- The following R package (install as needed): nimble
Access information
Other publicly accessible locations of the data:
- National Land Survey of Finland
- Statistics Finland
- Finnish Meteorological Institute
Data was derived from the following sources:
- The Finnish Biodiversity Information Facility (FinBIF)
