This repository contains data and code relating to the study:
Nature (2023). https://doi.org/10.1038/s41586-023-06400-1
Haase, P., D.E. Bowler, N.J. Baker, N. Bonada, S. Domisch, J.R. Garcia Marquez, J. Heino, D. Hering, S.C. Jähnig, A. Schmidt-Kloiber, R. Stubbington, F. Altermatt, M. Álvarez-Cabria, G. Amatulli, D. Angeler, G. Archambaud-Suard, I. Arrate Jorrín, T. Aspin, I. Azpiroz, I. Bañares, J. Barquín Ortiz, C.L. Bodin, L. Bonacina, R. Bottarin, M. Cañedo-Argüelles, Z. Csabai, T. Datry, E. de Eyto, A. Dohet, G. Dörflinger, E. Drohan, K.A. Eikland, J. England, T.E. Eriksen, V. Evtimova, M.J. Feio, M. Ferréol, M. Floury, M. Forcellini, M.A. Forio, R. Fornaroli, N. Friberg, J.-F. Fruget, G. Georgieva, P. Goethals, M.A.S. Graça, W. Graf, A. House, K.-L. Huttunen, T.C. Jensen, R.K. Johnson, J.I. Jones, J. Kiesel,, L. Kuglerová, A. Larrañaga, P. Leitner, L. L'Hoste, M.-H. Lizée, A.W. Lorenz, A. Maire, J.A. Manzanos Arnaiz, B. Mckie, A. Millán, D. Monteith, T. Muotka, J.F. Murphy, D. Ozolins, R. Paavola, P. Paril, F.J. Peñas, F. Pilotto, M. Polasek, J.J. Rasmussen, M. Rubio, D. Sánchez-Fernández, L. Sandin, R.B. Schäfer, A. Scotti, L.Q. Shen, A. Skuja, S. Stoll, M. Straka, H. Timm, V.G. Tyufekchieva, I. Tziortzis, Y. Uzunov, G.H. van der Lee, R. Vannevel, E. Varadinova, G. Várbíró, G. Velle, P.F.M. Verdonschot, R.C.M. Verdonschot, Y. Vidinova, P. Wiberg-Larsen, E.A.R. Welti
- Contains information on data providers, site characteristics, and the number of sites from each country
- Two tabs within this metadata spreedsheet contain metadata for the two csv files (in the outputs folder):
- All_indices_benthicMacroInverts_AllYears_alienzeros.csv (metadata tab: metadata_siteyear)
- All_siteLevel.csv (metadata tab: metadata_sites)
-
RawData_metadata.docx : Provides metadata regarding raw data files
-
csvs_each_dataset folder: Contains raw taxonomic composition data for all included studies
-
All_indices_benthicMacroInverts_AllYears_alienzeros.csv : Contains data for each site and year on:
- taxonomic diversity metrics
-
All_siteLevel.csv : Contains data for each site on slopes of functional and taxonomic diversity indices and environmental drivers
-
outputs_metaAnaylsis folder: Trend meta-analysis model outputs
-
outputs_movingWindow folder: Moving window analysis outputs
-
outputs_driver folder: Driver analysis outputs
-
outputs_sensitivity folder: Sensitivity analyses outputs including:
- high threshold moving window analysis (HTMW2)
- one-country removal effects (Jackknife)
- meta-analysis model comparisions (metaanalysisModelComparison)
- site level high threshold moving window analysis (siteLevel_HTMW)
- driver and moving window analyses only for sites with taxonomic resolution to species level (SppLevel_outputs)
- taxonomic resolution and seasonality sensitivity analyses outputs (TaxonomicSeason)
-
TaskIDs folder: Task IDs for running models and model estimates
-
ClimateModel_stan folder: Contains outputs of models calculating temperature and precipitation trends
-
Initial_Biodiversity_FuncTrait_and_climate_calcs folder: Contains scripts to calculate taxonomic and functional trait metrics for each site year and calculate temperature and precipitation trends
-
stan_models folder: Contains scripts of all stan models used in analyses
-
trend_metaAnalysis folder: Contains the following scripts:
- HPC_macroinverts_stanmodels : Models to calculate trends for each biodiversity metric and each of the 1816 sites
- HPC_Meta_analysis : Meta-analysis models to synthesize the site-level data using the output of the previous step and get overall trend for each metric (Trend ~ 1 + (1|study_ID) + (1|Country)
- HPC_modelchecking : Examining meta-analysis model fit parameters and calculating probability increasing/decreasing for each biodiversity metric
-
MovingWindow folder: Contains the following scripts:
- HPC_MovingWindow_sites : Models to calculate trends for each site within each window and biodiversity metric in moving window analysis
- HPC_Meta_analysis_movingwindow : Models to calculate overall estimates of trends within each window and biodiverisity metric
- MetaMeta_movingwindowEsts_Yr : Models to calculate overall linear year effects on biodiversity trajectories in moving window analyses
-
Driver folder: Contains the following scripts:
- HPC_Meta_analysis_drivers : Models to caluculate driver effects on biodiversity metrics using site-level data
- Modelchecking_Drivers_horseshoePriors : Examining driver model fit parameters
-
Sensitivity folder: Contains scripts for sensitivity checking models including:
- effects of trends within countries (Country_effects)
- high threshold moving window analysis (HTMW)
- using a one stage model rather than the two stage meta-analysis model for trend estimates (OneStage_models)
- effects of taxonomic resolution for the moving window analysis and claculating the proportion of pos/neg trends/ window (checking_MovingWindow_parameters)
- effects of taxonomic resolution and seasonality on trend estimates (HPC_Sensitivity_analysis)
- examing driver model outputs for sites with taxa IDed to species level only (Modelchecking_Drivers_sppLevel_horseshoe)
-
HPC_combine_site_trends script: Concatenates outputs for many analyses including:
- biodiverity trends
- meta-analysis
- moving window analysis
- driver analysis
- sensitivity analyses
-
Online Figures.docx : Includes all additional online figures not included in main text or Extended Data
-
Supplementary Table 2 : Contains information on time series locations, durations, and site characteristics
-
Fig1 folder: Contains scripts, data, and icons used to make Figure 1
-
Fig2_DensityPlots folder : Contains scripts and plots related to meta-analysis trend results
-
Fig3_movingWindow folder: Contains scripts and plots related to the moving window analyses
-
Fig4_drivers folder: Contains scripts and plots related to driver analyses
-
descriptive_plots folder: Contains scripts and plots descriping data distributions and correlations
-
Sensitivity folder: Contains scripts and plots for sensitivity analyses regarding:
- one-country removal effects (Jackknife)
- meta-analysis model comparisions (one stage, two stage weighted and unweighed)
- effects of sampling years & start year
- effects of seasonality
- effects of taxonomic resolution