"Drought legacy effects (DLE) in radial tree growth (RTG) have been extensively studied over the last decade and are found to critically influence carbon sequestration in woody biomass. However, the statistical significance of DLE depends on our definition of expected vs. unexpected growth variability, a definition that has not received sufficient scrutiny.
Here, we revisit popular DLE analyses using the ITRDB and employ a synthetic data simulation to disentangle four key factors influencing the magnitude of DLE. We show that DLE can be explained by the auto-correlation of RTG, depend on climate-growth cross-correlation, are directly proportional to the year-to-year variability of RTG, and scale with the chosen extreme event threshold. Using this simulation, we can reproduce the magnitude of observed pattern, meaning DLE cannot be distinguished from biological memory. We further find that the interpretation of DLE following isolated drought events at individual sites is challenged by high stochasticity, and show that the commonly perceived stronger DLE for conifers are a result of higher auto-correlation compared to deciduous broadleaves.
We present two pathways to improve the future assessment and interpretation of DLE: First, we provide a simulation algorithm to a posteriori account for auto-correlated residuals of the initial regression model between growth and climate, thereby retrospectively adjusting expectations for the statistical null model. The second pathway is to a priori include lagged climate parameters in the regression model. Doing so heavily reduces the magnitude of observed DLE and thus challenges us to consider the full spectrum of expected variability when evaluating drought-induced growth deviations."