The stationary and non-stationary character of the silver fir, black pine and Scots pine tree-growth-climate relationships

Abstract

Tree-growth-climate relationships are usually assumed to have a stationary character, i.e., continuous and/or time-independent, along the lifetime of the trees. The fact that non-stationarity, i.e., discontinuous and/or time-variable, is more likely to actually be their general rule, has been often neglected in dendrochronology. Nine silver fir, black pine and Scots pine residual ring-width index chronologies (RWIresidual) and five precipitation- and temperature-derived seasonal climatic variables, covering the 20th century and the beginning of the 21st one, were used in this study. Heat map analyses based on rolling window correlations, using corrected p-values in order to deal with the type I errors (i.e., the multiple testing or comparison problem) and reduce them, were conducted to evaluate the evolution and stability of tree-growth-climate relationships along the lifetime of the trees, i.e., their stationary and/or non-stationary character. The obtained results showed that stationary tree-growth-climate relationships were well conserved within trees belonging to a given genus: positive effects, both at young and mature stages, of Twinter (winter temperature) on the Abies trees and of Psprsum(t) (spring-summer precipitation of the current-to-growth year) on the Pinus trees. Non-stationary tree-growth-climate relationships were instead species- and site-dependent and stopped in the 1970s/1980s/1990s. Growth decoupling from seasonal climatic variables was linked in many cases with climatic anomalies but the obtained results did not yield a general rule in this regard. Heat map analyses based on rolling window correlations proved to be a powerful statistical tool in disentangling between the stationary and/or non-stationary character of the tree-growth-climate relationships. Summarizing, this study puts into perspective the critical aspect of looking at the stationary and/or non-stationary character of the tree-growth-climate relationships if we want to better predict the impact of climate change on the future forest tree growth and dynamics based on past tree-growth-climate relationships