Stability of growth-climate relationships at treeline populations of mountain birch in Arctic Sweden

Background/Question/Methods

Tree-ring data are commonly used to interpret past climate conditions.  The usefulness of this method is based on the assumption that climate-growth relationships remain stable over time.  In this study, we use an extensive tree-ring database from 327 trees at 25 treeline sites in northern Sweden to investigate the stability of the climate-growth relationship over the 20^th century for mountain birch (Betula pubescens ssp. czerepanovii).  Partial correlations are calculated between monthly summer temperatures for different successive 30 year periods throughout the period of record. 

Results/Conclusions

We find that the correlation between temperature and growth varied in magnitude over the 20^th century,  but the sign of the correlations did not. The strength of the correlation is heavily influenced by ecological factors, such as periodic geometrid moth (e.g. Epirrita autumnata) outbreaks.  Moth outbreaks reduce the correlation between growth and climate variables, and short duration outbreaks can influence the observed correlations for long periods.  In some cases, the relative strength of correlations among months shifted over the 20^th century.  We show that although the strengths of the correlations vary over the period of instrumental record past climate reconstructions are generally adequate because the signs of the correlations do not change.