Data from: Warm or dry springs (still) boost the reproduction of most temperate songbirds
Data files
Oct 29, 2025 version files 15.09 MB
-
20250428_43species.nex
2.10 MB
-
df_mod_final_select.csv
12.97 MB
-
pop_growth_rates_2001-2022.csv
2.79 KB
-
README.md
5.10 KB
-
THV_all_sp.csv
5.79 KB
Abstract
Nonlinearities are ubiquitous features of ecosystem dynamics in climate change ecology. The changing climate is expected to reveal hump-shaped patterns in ecosystem responses, delineating weather optima and constraints. In temperate mid-latitudes, both cold and warm spring constraints are reported to limit songbird breeding productivity. However, with many studies focusing on specifically declining functional groups, such as long-distance migrants, our understanding of the overall influence of increasingly warmer and drier conditions on songbird productivity remains limited. Here, we modelled changes in songbird productivity in relation to temperature and water balance anomalies during the breeding season, aiming to identify key weather constraints—whether life-history dependent or shared across species—in a warming temperate world. Using standardised capture data of 300,031 birds across 68 species, we quantified changes in songbird productivity along gradients of spring weather anomalies and extremes using generalized linear mixed models. We then conducted interspecific analyses to explore how life-history traits (e.g., migratory strategy, thermal and hydric affinities) modulate species’ sensitivity to weather. Songbird productivity increased along gradients from cold to warm and from wet to dry anomalies. Nonetheless, warm-related constraints also emerged: in early spring, particularly at already warm sites and most strongly for long-distance migrants; and in late spring, especially for cold-adapted species. Warmer or drier springs continue to enhance songbird productivity in temperate France, reaffirming the detrimental influence of cold and wet snaps. Beyond the well-known benefits of earlier breeding phenologies, these effects are likely driven by the impacts of such weather conditions on ecosystem productivity and resource availability. Non-linear patterns and early signs of negative effects of late spring temperatures, however, suggest that productivity gains are likely to fade, if not turn negative in already warm regions, as warming intensifies.
https://doi.org/10.5061/dryad.vq83bk43t
Description of the data and file structure
The dataset includes songbird productivity and environmental variables used in the study. It is structured into four files:
- df_mod_final_select.csv
- THV_all_sp.csv
- pop_growth_rates_2001-2022.csv
- 20250428_43species.nex
Files and variables
File: df_mod_final_select.csv
Description: Main dataset used in the study, combining: (1) Songbird productivity data from the French Constant Effort Site (CES) ringing program (1991–2022), provided by the Centre de Recherches sur la Biologie des Populations d'Oiseaux (CRBPO); (2) Environmental, demographic, and weather variables used in Multi-Species Models.
Variables
- ID_PROG: IDs of the 365 French CES sites
- YEAR: Year of capture session (1991–2022)
- ESPECE: Three-letter species code
- Nom_sc: Scientific name of species
- AD: Number of adults captured per year, species, and site
- JUV: Number of juveniles captured per year, species, and site
- n_birds: Total number of individuals captured per year, species, and site
- Prod: Percentage of juveniles among total birds per year, species, and site
- HABITAT_SP: Primary habitat of species (aquatic vs terrestrial)
- MIGRATION: Migratory strategy (short vs long-distance migrants)
- site_mean_Temp_spring: Mean spring site temperature (°C) over the reference period (1991-2022)
- cat: Categorysite-by-sitesite temperature (cold, medium, warm)
- catYeCombinationison of 'cat' and 'YEAR' columns for figures
- aTemp_early, aTemp_late, aTemp_spring: Temperature anomalies (°C) for early and late spring, and for the whole spring
- aTemp_slid_early, aTemp_slid_late: Temperature anomalies (°C) for early and late spring with windows adjusted for the phenology gradient dependent on site temperature (See Supporting Information)
- winter_SPEI, early_SPEI_standaTemp, late_SPEI_standaTemp, ninemonth_SPEI_standaTemp: Water balance anomalies (SPEI) for winter, early spring, late spring, ring and nine-month periods. "standaTemp" indicates a variable that is standardised to temperature anomalies (see Supporting Information)
- aNDVI_posopt: NDVI anomalies for the post-optimum period
- nbECE_3cons_early, nbECE_3cons_late: Number of days within sequences of at least three consecutive days during which daily temperature anomalies exceeded the 95th percentile for early and late spring
- nb_ECE_SPEI_early_cat, nb_ECE_SPEI_late_cat: Binary variables, classifying years as 'dry' when they included at least one week of severe drought—characterised by a SPEI value < -1.5—or 'normal' in the absence of such events
- a_density_scale_corrected: Annual anomalies in adult population size per species and site
- a_pheno: Annual anomalies in fledging phenology (days) per species at the national level
- a_doy_tot_sc: Annual anomalies in the mean date of capture sessions (in days) per year and site
- sp_rangesitetemperature: Species included in the robustness analysis, considering the 37 species experiencing a broad temperature range across their French distribution
- sp_singlespeciesmodels: Species included in Single-Species Models (42 species)
File: THV_all_sp.csv
Description: Life-history traits used for Single-Species Models.
Variables
- Species: Scientific names of species
- ESPECE: Three-letter species code
- nom_fr: French species name
- HABITAT_SP: Species' hydric affinity (aquatic vs terrestrial)
- MIGRATION: Migratory strategy (short vs long-distance migrants)
- Broods per year: Average number of broods per species and year
- thermal_niche_range: Difference (°C) between the thermal maximum and minimum experienced across each species’ breeding range
- STI: Species Temperature Index (STI), the thermal centroid of the breeding range (°C)
File: pop_growth_rates_2001-2022.csv
Description: Species-specific population growth rates for Trait-Based Models.
Variables
- ESPECE: Three-letter species code
- estimate: Species-specific population growth rate estimates
- se: Standard error of the Species-specific population growth rate estimate
- perc_STOC: Percent change relative to the growth rate over the period (from "start" year to 2022)
- start: Year at start of the growth rate analysis
File: 20250428_43species.nex
Description: Phylogenetic tree of the 42 species for Trait-Based Models, generated with BirdTree.
Code/software
The model was created using R version 4.2.2
Two scripts are provided (uploaded to Zenodo) and contain all the information to walk you through the datasets and analysis:
- Script 1: Code for Multi-Species Models
- Script 2: Code for Single-Species Models and Trait-Based Model
Access information
NDVI data are accessible at https://appeears.earthdatacloud.nasa.gov/
Weather data are accessible at https://meteo.data.gouv.fr/
