Tree rings can record the past climatic conditions which allows for retrospective analyses of climate-growth relationships. Generally, it is assumed that the relationships between tree growth and climate conditions are age-independent, as long as the biological growth trends related to tree age are removed from the tree-ring data through detrending. However, if trees of different ages respond differently to climatic conditions, a dendroclimatic analysis based on even-aged samples may be biased in capturing climatic variability throughout the length of the chronology. To evaluate the importance of this effect, in the present study we analyzed the role of tree age in affecting radial growth response to climate in Korean pine (Pinus koraiensis).
In the north slope of Changbai Mountain, Northeast China, we sampled 49 trees of Korean pine, which were grouped into two age classes: the younger 50-90 years old trees (young cambial age group, YCA), and the trees older than 130 years (large cambial age group, LCA). Standard and residual chronologies of both YCA and LCA were developed and analyzed in response function and correlation analyses. Statistically significant differences existed between these chronologies and the growth response to climate variables varied between the two age classes. Standard LCA chronology had higher mean sensitivity, standard deviation, and signal-to-noise ratio as compared to respective YCA chronology, suggesting higher amount of climate-related information contained in LCA chronology.
Correlation analysis showed that the radial growth of YCA was negatively affected by mean monthly temperature of current January and February, mean monthly minimum temperature of previous September and current March and September, and it was positively correlated with mean monthly maximum temperature of previous November and of the current May. The radial growth of LCA was positively correlated with mean monthly maximum temperature of current January, February, April, June, July and September, and negatively correlated with mean monthly minimum temperature of current April and September. Growth of LCA trees was positively correlated with monthly precipitation of both previous and current May. Response analysis revealed that radial growth in YCA group negatively responded to mean monthly temperature of current January, while LCA trees positively responded to mean monthly maximum temperature during current May and September.The tree-ring growth of YCA was largely correlated with monthly temperature, and the radial growth of LCA was affected by both temperature and precipitation.
Furthermore, we checked the tree-ring width annually during the period of 1982 to 2007 and noticed that in 1985, the tree-ring increment in YCA increased sharply, while the increment of LCA did not show any obvious increase in that year. In 1988, the tree-ring width in LCA trees was clearly below long-term average, while no such pattern was present in YCA trees. Examination of instrumental climate data revealed that in 1985, the temperature was below long-term average in January and March, and clearly above the average in May. In 1988, the temperature in current February was below average significantly. These climatic anomalies were likely responsible for tree ring growth depressions in above-mentioned years. These observations revealed that climate conditions might affect tree-ring growth differently between trees possessed different ages. The assumption of age-independent climate-growth relationship was probably invalid for Korean pine in Changbai Mountain. Physiological processes and hydraulic constraints related to tree age could possibly be the main causes of these age-dependent effects, which required consideration during development of sampling strategies.