Multi‐variable approach pinpoints origin of oak wood with higher precision

Abstract

Aim Spatial variations of environmental conditions translate into biogeographical patterns of tree growth. This fact is used to identify the origin of timber by means of dendroprovenancing. Yet, dendroprovenancing attempts are commonly only based on ring‐width measurements, and largely neglect additional tree–ring variables. We explore the potential of using wood anatomy as a dendroprovenancing tool, and investigate whether it increases the precision of identifying the origin of oak wood. Since different tree–ring variables hold different information on environmental conditions prevailing at specific times of the growing season—which vary between source regions—we hypothesize that their inclusion allows more precise dendroprovenancing.

Location Europe, Spain.

Taxon Quercus robur L., Quercus petraea (Matt.) Liebl., Quercus faginea Lam., Quercus pyrenaica Willd.

Methods We sampled four oak species across Northern Spain, i.e. from the Basque country and Cantabria and—in the Basque country—from low to high elevation (topographic/latitudinal gradient). We measured multiple tree–ring variables to (a) extract complementary variables; (b) present statistical relations among them; (c) analyse region‐specific variation in their patterns based on time–series of individual trees; and (d) determine underlying climate–growth relationships. Leave‐one‐out analysis was used to test whether a combination of selected variables allowed dendroprovenancing of a randomly selected tree within the area.

Results A combination of latewood width (LW) and earlywood vessel size was used to pinpoint the origin of oak wood with higher precision than ring width or LW only. Variation in LW pinpointed the wood to east and west areas, whereas variation in vessels assigned wood to locations along a latitudinal/topographic gradient. The climatic triggers behind these gradients are respectively an east–west gradient in June–July temperature and a north–south gradient in winter/spring temperatures. The leave‐one‐out analyses supported the robustness of these results.

Main conclusions Integration of multiple wood–xylem anatomical variables analysed with multivariate techniques leads to higher precision in the dendroprovenancing of ring‐porous oak species.