Temporal and spatial variability in tree growth response to climate

A study was conducted using tree-ring data collected by the Tennessee Valley Authority from a system of permanent forest inventory plots. The purpose of the study was to characterize temporal and spatial variability in the growth response ot four major hardwood species to climatic fluctuations, and to evaluate the role of environmental factors associated with differences in response among individuals. Standardized series of annual radial increment for individual trees were grouped using cluster analysis. The relationship between average monthly temperature and total monthly precipitation values for the current growing season and values lagged one year was modelled using a time-variant response function. Resulting models were evaluated using graphical analysis and estimated confidence intervals for parameter values. Environmental variables associated with each core tree were determined using the permanent plot remeasurements, detailed site descriptions, and site-specific soils descriptions. Discriminant analysis and canonical discriminant analysis were used to reclassify cluster groupings of core trees using the environmental variables, and to assess which variables were most strongly correlated with important in producing differences in growth response. Results of the dendroclimatic modelling indicated that there is a great deal of temporal variability in the climate/growth relationship for the species and sites included in this study. Only three groups out of 54 showed no variation over time in parameter coefficients. Trends in growth generally showed a decrease in sensitivity to climate occurring in the tree-ring series. Statistical tests indicate that these trends are significantly different from random. However, a majority of the groups also showed a slight increase in sensitivity to climate in recent years. Changes in tree sensitivity do not appear to coincide with changes in climate. Environmental factors were effective in explaining groupings of tree core series as long as interactions between variables were considered. Overall, the most important variables appeared to be relative frequency and importance value of a species, and the relative frequency and importance value of competing species. Other important environmental variables included soil depths and textures. These factors indicate that differences in growth response to climate may be a function of changes in stand structure, competitive interactions and soil moisture holding capacity.