Data and R code for "Negative effects of wind on plant hydraulics at the global scale"

To minimize ontogenetic and methodological variation, we only included trait data that met the following criteria: (a) plants were grown in natural ecosystems, excluding greenhouse and common garden experiments; (b) measurements were made on adult plants and not on seedlings; (c) hydraulic traits were measured on terminal stem or branch segments in the sapwood at the crown; and (d) trait data were calculated as the mean value for each species at the same site when data were from multiple sources.

Climate data were obtained either from the original reports or from WorldClim version 2 (http://worldclim.org/version2) if the original data were not available. The following variables were extracted from WorldClim: mean annual wind speed, mean annual precipitation, mean annual temperature, precipitation seasonality, temperature seasonality, precipitation of driest month, and minimum temperature of coldest month. The VPD data was extracted from the TerraClimate dataset (http://www.climatologylab.org/terraclimate.html). Annual PET (potential evapotranspiration) data were extracted from the CGIAR-CSI consortium (http://www.cgiar-csi.org/data). The moisture index (MI) is the ratio of precipitation to PET.

Simple linear regression was used to examine the relationships between two variables, utilizing the 'lm' function in R software. Partial regression analysis was conducted using the R package VISREG to investigate the relationships between wind speed and plant hydraulics while controlling for other variables. This analysis helped to illustrate the independent effect of wind on plant hydraulics. The Random Forest machine-learning algorithm (implemented using the R package randomForest) was utilized to assess the relative importance of environmental variables for each plant hydraulic trait. The Mean Decrease in Gini was calculated as the average of a variable's total decrease in node impurity, taking into account the proportion of samples that reach that node in each individual decision tree in the random forest. This provides a measure of a variable's importance in estimating the value of the target variable across all of the trees in the forest. A higher Mean Decrease in Gini value indicates greater importance of the variable. Multiple regression analyses were performed to develop predictive equations for plant hydraulic traits using environmental variables. To test for hydraulic traits-wind speed slope directions and differences among species groups in different climatic regions, we used standardized major axis (SMA) analyses. The R package SMATR was employed for these analyses. We considered p < 0.05 as the threshold for statistical significance in all models.