Modeling patterns between drought and tree biomass growth from dendrochronological data: A multilevel approach

We established a new approach for incorporating dynamic tree biomass variables into dendrochronological studies. A multi-level algorithm was developed for modeling biomass growth from tree-ring chronologies, and for establishing temporal relationships between biomass dynamics and drought. The algorithm, BIOdry, integrates conventional procedures for modeling patterns between tree biomass growth and drought, using dendrochronological data from Mediterranean ecosystems, accounting for multiple sources of variation from sample design, and comparing patterns from contrasting climatic portions of study site during last 50 years. The modeling approach was validated by exploring principal factors affecting annual biomass increments and annual aridity indexes in two Mediterranean pine forests: Pinus pinaster and P. sylvestris. BIOdry corrected several statistical problems related with dendrochronological sampling and modeling of biomass growth. The most important factors affecting relationships between the series were time span of the analysis, regional variations and adaptive strategies of species. Series from trees in the Atlantic region were not sensitive to increasing drought while responses of trees in the Mediterranean regions depended on species. Series from P. pinaster were more sensitive to drought oscillations than series of P. sylvestris. Both species growing in southern Spain have experienced increasing sensitivity to drought during last 50 years. Southern forests are little adapted to decreasing precipitations of autumns and winters. BIOdry can be used for assessing spatial and temporal vulnerabilities of growth attributes derived from tree-ring chronologies to increasing drought.