Supplementary data from: Current and past climate co-shape community-level plant species richness in the Western Siberian Arctic

The Arctic ecosystems and their species are exposed to amplified climate warming and, in some regions, to rapidly developing economic activities. We used macroecological modeling to estimate the community-level species richness across the Western Siberian tundra, with climate variables and anthropogenic influence identified as main explanatory factors. Our results reveal complex spatial patterns of community-level species richness in the Western Siberian Arctic. We show that climatic factors such as temperature (including paleotemperature) and precipitation are the main drivers of plant species richness in this area, and the role of relief is clearly secondary.
Here we present a supplementing dataset to the analysis of our paper “Current and past climate co-shape community-level plant species richness in the Western Siberian Arctic” (https://doi.org/10.1002/ece3.11140). Our research is based on the Western Siberian part of the Russian Arctic Vegetation Archive (AVA-RUS, http://avarus.sp..., The dataset consists of R scripts we used for the analysis as well as the training data in .csv format.
The R script is separated into five .R files:

YANAO_PGF_final_PP_test_paleo. The script contains predictive power and autocorrelation test;

GLM_GBM_random_forest_GAM_fitting_paleoclimate. The script includes fitting of four models (GLM, GAM, GBM, Random forest) used for the analysis;

cross-validation. Cross-validation of the model;

spatial_projections_paleo. Spatial projections of the models.

The data includes six csv files:

YANAO_PGF_full_with_paleoclimate (contains all the sampled predictors we used for testing);

YANAO_final_predictors_paleo (contains only predictors selected for model’s fitting)

Cryo_db_main (sampled raw paleoclimatic data from CHELSA-TraCE21k dataset)

paleoprecip_data_new (paleoprecipitation data prepared for tests)

paleotemperature_data_new (paleotemperature data prepared for tests)

distance_to_land_ice_data_new (distance to land ice data prepared fo..., The scripts were written and used in R programming environment, version 4.1.2. Microsoft Excel can be used to view csv files.
The scripts and data are free for non-commercial use. We kindly request to cite the original publication while referencing them.
Zemlianskii, V., Brun, P., Zimmermann, N. E., Ermokhina, K., Khitun, O., Koroleva, N., & Schaepman-Strub, G. (2024). Current and past climate co-shape community-level plant species richness in the Western Siberian Arctic. Ecology and Evolution, 14, e11140. https://doi.org/10.1002/ece3.11140, # Supplementary data from: Current and past climate co-shape community-level plant species richness in the Western Siberian Arctic

The dataset consists of R scripts we used for the analysis as well as the training data in .csv format.

The R script is separated into four .R files:

1. **YANAO_PGF_final_PP_test_paleo.R** The script performing a test of univariate predictive performance and limited collinearity for the predictors.
2. **GLM_GBM_random_forest_GAM_fitting_paleoclimate.R**\
The script includes fitting of four models (GLM, GAM, GBM, Random\
forest) used for the analysis. We modeled species richness as a function of non-anthropogenic predictors using four different model algorithms: random forest, gradient boosting machine, generalized linear model, and generalized additive model. For RF, we fitted 500 regression trees, considering three predictors for each tree. For GBMs, we set the number of trees to 80, the minimum number of data points per leaf to 10, the learning ra...