Quantifying antecedent climatic drivers of tree growth in the Southwestern US

Variation in antecedent (past) climate conditions is likely to govern tree growth over long periods of time. Antecedent conditions are rarely considered in models of tree growth, representing a weakness in quantitative understanding of forest responses to climate variations. We applied the stochastic antecedent modelling (SAM) framework to 367 International Tree Ring Data Bank chronologies in the southwestern US (“Southwest”) representing eight conifer species. To better understand climatic and physiologic controls on tree growth, we quantify the effects of antecedent precipitation, temperature and Palmer Drought Severity Index (PDSI) over 60 months preceding and including the year of ring formation. In Pinus edulis, Pinus ponderosa and Pseudotsuga menziesii, growth responded primarily to recent precipitation and temperature conditions (43%–49% of the response was driven by conditions during the year of ring formation), but to less recent PDSI conditions (>50% of response driven by conditions 13–48 months prior to the year of ring formation), though PDSI significantly affected growth at only 21% of sites. Combining extensive tree-ring data with monthly resolution climate data also reveals key climatic events, such as the effect of monsoon arrival date on growth, especially in P. menziesii, highlighting the ability of the SAM framework to identify climate effects at multiple time-scales. Sensitivity to antecedent climate, baseline growth at average climate conditions and the strength of first order autocorrelation varied spatially, suggesting variation in mean growing conditions, non-structural carbohydrate storage, and/or seasonal precipitation contribution of the North American Monsoon may drive differences in growth sensitivities across species’ ranges. Synthesis. Our findings provide further evidence for multi-year legacy effects of climate conditions, particularly drought metrics, on tree growth. Antecedent climate and especially drought are key drivers of growth in these species, and associated climatic sensitivities and growth indices vary spatially. We argue such factors should be considered in modelling efforts. The spatial variability in antecedent climate sensitivities points to key differences in how different populations within a species range may respond to climate change, particularly if timing of weather events, such as monsoon arrival date, or annual precipitation amounts undergoes significant changes.