By Thompson, M.J., Martin, J.G.A, Biard, C., Bleu, J., Branston, C. J., Capilla-Lasheras, P., Dingemanse, N.J., Dominoni, D.M., Eens, M., Eeva, T., Evans, K.L., Isaksson, C., Liker, A., Massemin, S., Matthysen, E., Mouchet, A., Perret, S., Senar, J.C., Seress, G., Szulkin, M., Vincze, E., Watson, H., Réale, D., Charmantier, A.

Description of the data and file structure

Raw data to generate the formatted data below used for statistical models can be accessed through SPI-Birds (https://spibirds.org) or by contacting data owners directly. Data processing steps for each dataset are described below.

Data list
1."ThompsonEA_UrbanVp_AdultTarsus.csv" / "ThompsonEA_UrbanVp_AdultTarsus_RanObs.csv"
2."ThompsonEA_UrbanVp_NstTarsus.csv"
3."ThompsonEA_UrbanVp_layDate.csv"

1."ThompsonEA_UrbanVp_AdultTarsus.csv": Dataset on individual-level tarsus length of Adult great tits and blue tits, resulting from trans-European collaboration with long-term data from multiple populations along urbanization gradients across Europe. Data "ThompsonEA_UrbanVp_AdultTarsus_RanObs.csv" has identical structure aside from the column "Tarsus_mean" (see below). Data processing included the i) conversion of tarsus measurements from the Paris study system using equation S2 in supplementary, ii) removal of four outliers of tarsus length from raw data: 10.4mm (Barcelona ID BH8112), 10.5mm (Barcelona ID BS2981), 11.73mm (Malmo ID 2KT47318), 29.2mm (Malmo ID 2KS92199), iii) retention of only one observation per individual by taking the mean tarsus length of a given individual across observations in the dataset (or random observation for the "_RanObs" version of the dataset), iv) assignment of nest boxes into cluster IDs using a clustering algorithm, and v) modification of ID codes. Please see methods in main text and supplementary materials for more details concerning these processing steps.

Column: Description
*City : City code pertaining to study system where ANT = Antwerp, BUD = Budapest, GLA = Glasgow, GOT = Göteborg, HAR = Harjavalta, HEL = Helsingborg, Malmo = Malmö, MON = Montpellier, MUN = Munich, PAR = Paris, STR = Strasbourg, VES = Veszprém, WAR = Warsaw
*BreedingSeason : Year data collected in yyyy format
*ID_clu : Cluster (=subpopulation) ID code
*Habitat : Habitat type category urban vs. forest
*ID : Individual ID code (note only one observation per individual in dataset)
*Tarsus_mean : Mean tarsus length in millimeters (Other dataset uses Tarsus_mod instead which is just a random observation of tarsus length per individual instead)
*Species : Species where great tit = PARMAJ or blue tit = CYACAE
*Sex : Sex were F = female and M = male
*mean_propurb_1000: Mean proportion urbanization of each cluster measured as impervious surface area within 1000m radius circular buffers of each nest box
*mean_propurb_100 : Mean proportion urbanization of each cluster measured as impervious surface area within 100m radius circular buffers of each nest box
*mean_shdi_1000 : Mean Shannon diversity index of each cluster measured using number of land cover types within 1000m radius circular buffers of each nest box
*mean_shdi_100 : Mean Shannon diversity index of each cluster measured using number of land cover types within 1000m radius circular buffers of each nest box
*mean_lat : Mean latitude of each cluster
*area : Area in squared meters of each cluster
*years_clu : Number of years of data collection for each cluster

2."ThompsonEA_UrbanVp_NstTarsus.csv": Dataset on individual-level tarsus length of nestling great tits and blue tits, resulting from trans-European collaboration with long-term data from multiple populations along urbanization gradients across Europe. Data processing included the i) conversion of tarsus measurements from the Paris study system using equation S2 in supplementary, ii) removal of nine outliers of tarsus length from raw data: 4.6mm (Munich ID 18307), 5.3mm (Munich ID 1871), 5.7mm (Munich ID 611), 4.5mm (Munich ID 614), 5.3mm (Munich ID 846), 2.1mm (Paris ID 8949165), 2.53mm (Malmo ID 2KR97129), 3.30mm (Malmo ID 2KX12867), 29.7mm (Malmo ID 2KS91854), iii) including only observations of nestlings that were measured between 13 - 17 days old, iv) the retention of only one random nestling observation for each brood (observation per nest box in a given year) in the dataset, v) assignment of nest boxes into cluster IDs using a clustering algorithm, and vi) modification of ID codes. Please see methods in main text and supplementary materials for more details concerning these processing steps.

Column: Description
*City : City code pertaining to study system where ANT = Antwerp, BUD = Budapest, GLA = Glasgow, GOT = Göteborg, HAR = Harjavalta, HEL = Helsingborg, Malmo = Malmö, MON = Montpellier, MUN = Munich, PAR = Paris, STR = Strasbourg, VES = Veszprém, WAR = Warsaw
*BreedingSeason : Year data collected in yyyy format
*ID_clu : Cluster (=subpopulation) ID code
*Habitat : Habitat type category urban vs. forest
*ID : Individual ID code (note only one observation per individual in dataset)
*Tarsus_mod : Tarsus length in millimeters taken from one random nestling per each brood
*Species : Species where great tit = PARMAJ or blue tit = CYACAE
*ChickAge_comb : Age of chick in days when measured
*mean_propurb_1000: Mean proportion urbanization of each cluster measured as impervious surface area within 1000m radius circular buffers of each nest box
*mean_propurb_100 : Mean proportion urbanization of each cluster measured as impervious surface area within 100m radius circular buffers of each nest box
*mean_shdi_1000 : Mean Shannon diversity index of each cluster measured using number of land cover types within 1000m radius circular buffers of each nest box
*mean_shdi_100 : Mean Shannon diversity index of each cluster measured using number of land cover types within 1000m radius circular buffers of each nest box
*mean_lat : Mean latitude of each cluster
*area : Area in squared meters of each cluster
*years_clu : Number of years of data collection for each cluster

3."ThompsonEA_UrbanVp_layDate.csv": Dataset on individual-level lay date of female great tits and blue tits, resulting from trans-European collaboration with long-term data from multiple populations along urbanization gradients across Europe. Data processing included the i) assignment of nest boxes into cluster IDs using a clustering algorithm , ii) inclusion of first clutches only determined by selecting observations that were within 30 days of the first lay date within a given year, cluster, and species, iii) retention of only one observation of each known individual by taking the first observation of that individual that appeared in the dataset, and iv) modification of ID codes. Please see methods in main text and supplementary materials for more details concerning these processing steps.

Column: Description
*City : City code pertaining to study system where ANT = Antwerp, BUD = Budapest, GLA = Glasgow, GOT = Göteborg, HAR = Harjavalta, HEL = Helsingborg, Malmo = Malmö, MON = Montpellier, MUN = Munich, PAR = Paris, STR = Strasbourg, VES = Veszprém, WAR = Warsaw
*BreedingSeason : Year data collected in yyyy format
*ID_clu : Cluster (=subpopulation) ID code
*Habitat : Habitat type category urban vs. forest
*ID : Individual ID code (note only one observation per individual in dataset)
*LayDate_julian : Laying date in julian days (days since Jan 1) from the first observation of each individual in the dataset
*Species : Species where great tit = PARMAJ or blue tit = CYACAE
*Age : Age category where Adult = Ad and Yearling / Juvenile = Juv
*mean_propurb_1000: Mean proportion urbanization of each cluster measured as impervious surface area within 1000m radius circular buffers of each nest box
*mean_propurb_100 : Mean proportion urbanization of each cluster measured as impervious surface area within 100m radius circular buffers of each nest box
*mean_shdi_1000 : Mean Shannon diversity index of each cluster measured using number of land cover types within 1000m radius circular buffers of each nest box
*mean_shdi_100 : Mean Shannon diversity index of each cluster measured using number of land cover types within 1000m radius circular buffers of each nest box
*mean_lat : Mean latitude of each cluster
*area : Area in squared meters of each cluster
*years_clu : Number of years of data collection for each cluster

Model object .rds files

Model objects (22 total) include output from Bayesian statistical models fitted using the package "brms" (https://cran.r-project.org/web/packages/brms/index.html) and are provided to reduce processing time to obtain the study's results. Model object file names include information on the type of model fitted where models have been run separately for each trait (tarsAd = Adult tarsus length, tarsNst = Nestling tarsus length, LD = Lay Date) and species (GT = great tit, BT = blue tit). Each of these trait-species model combinations further evaluate three different effects of urbanization (Hab = habitat type urban vs forest, 1000 = continuous impervious surface area at 1000m scale, 100 = continuous impervious surface area at 100m scale). See further descriptions below and where model results are reported in main text and supplementary. Note that saved model objects (.rds files) are called in the scripts and can be downloaded here in the file list.

CompositeModel_tarsAdGT_Hab_pub.rds / CompositeModel_tarsAdBT_Hab_pub.rds
CompositeModel_tarsNstGT_Hab_pub.rds / CompositeModel_tarsNstBT_Hab_pub.rds
CompositeModel_LDAdGT_Hab_pub.rds / CompositeModel_LDAdBT_Hab_pub.rds: Model results for effect of urbanization via habitat type (Hab) shown in Table 1 of main text for adult tarsus length (tarsAd), nestling tarsus length (tarsNst), and lay date (LD) for each species (great tit models on left of dash and blue tit models on right of dash above).

CompositeModel_tarsAdGT_Hab_RanObs_pub.rds / CompositeModel_tarsAdBT_Hab_RanObs_pub.rds: Model results for effect of urbanization via habitat type (Hab) shown in Table S4 of supplementary for adult tarsus length (tarsAd) for each species (great tit models on left of dash and blue tit models on right of dash above) when instead selecting a random observation per individual rather than their mean.

CompositeModel_tarsAdGT_1000_pub.rds / CompositeModel_tarsAdBT_1000_pub.rds
CompositeModel_tarsNstGT_1000_pub.rds / CompositeModel_tarsNstBT_1000_pub.rds
CompositeModel_LDAdGT_1000_pub.rds / CompositeModel_LDAdBT_1000_pub.rds: Model results for effect of urbanization via impervious surface area at 1000m scale (1000) shown in Table S5 of supplementary for adult tarsus length (tarsAd), nestling tarsus length (tarsNst), and lay date (LD) for each species (great tit models on left of dash and blue tit models on right of dash above).

CompositeModel_tarsAdGT_100_pub.rds / CompositeModel_tarsAdBT_100_pub.rds
CompositeModel_tarsNstGT_100_pub.rds / CompositeModel_tarsNstBT_100_pub.rds
CompositeModel_LDAdGT_100_pub.rds / CompositeModel_LDAdBT_100_pub.rds: Model results for effect of urbanization via impervious surface area at 100m scale (100) shown in Table S6 of supplementary for adult tarsus length (tarsAd), nestling tarsus length (tarsNst), and lay date (LD) for each species (great tit models on left of dash and blue tit models on right of dash above).

CompositeModel_tarsAdGT_Hab_MUC_pub.rds / CompositeModel_tarsNstGT_Hab_MUC_pub.rds: Model results for effect of urbanization via habitat type (Hab) shown in Table S8 of supplementary for adult (tarsAd) and nestling tarsus length (tarsNst) for great tits when excluding data from the larger Munich cluster.

Code/Software

Data compiled and analyzed using R script: "ThompsonEA_UrbanVp_Script.R" with free open software R (https://www.r-project.org/).

License

Data and code provided are available under a Creative Commons Zero v1.0 Universal License (CC0 1.0; https://choosealicense.com/licenses/cc0-1.0/).