Search

Climate warming at high latitudes has contributed to the growing interest in shrub tree-ring analysis. Shrub architecture presents new challenges for dendrochronology, such as the seemingly lower and inconsistent climatic sensitivity of stems vs. root collars. Shrub stems may thus be considered as sub-optimal to study climate–growth relationships. In this paper, we propose that the lower climatic sensitivity of stems could be caused by the use of unsuitable detrending methods for chronologies spanning decades rather than centuries. We hypothesize that the conversion of the ring width (RW) to basal area increment (BAI) is better suited than traditional detrending methods to removing age/size-related trends without removing multi-decadal climate signals. Using stem and root collar samples collected from three sites in the forest–tundra ecotone of eastern Canada, we compared the climate–growth relationships of these two approaches for stems and root collars using mixed-effects models. The climate sensitivity was, on average, 4.9 and 2.7 times higher with BAI than with detrended (mean-centered) RW chronologies for stems and root collars, respectively. The climatic drivers of radial growth were identical for stems and root collars when using BAI (July temperature and March precipitation), but were inconsistent when using detrended RW series (root collars: July temperature and March precipitation at all sites; stems: April and June temperature, depending on the site). Although the use of BAI showed promising results for studying long-term climate signals in shrub growth chronologies, further studies focusing on different species and locations are needed before the use of BAI can become broadly used in shrub dendrochronology.