Xylem Embolism in Ring Porous, Diffuse Porous, and Coniferous Trees of Northern Utah and Interior Alaska

Xylem embolism was measured in nine tree species for one or more years. Species were ring—porous (Quercus sp.), diffuse—porous (Alnus, Betula, Populus spp.) or coniferous (Picea, Larix, Abies spp.). Intraspecific (Populus tremuloides) and intrageneric (Betula, Alnus) comparisons were made between sites in northern Utah and interior Alaska. Most embolism, \u3e90% in some dicot species, occurred in winter. Within sites, dicot trees embolized more than conifers. Between sites, Alaskan dicot trees embolized less than their Utah counterparts. Differences were explained by vulnerability to embolism caused by freeze—thaw cycles. Most conifers were entirely resistant, whereas dicot trees were vulnerable. Less embolism in Alaskan dicot trees was associated with fewer freeze—thaw events in Alaska vs. Utah. Vulnerability was positively correlated with conduit volume and hydraulic conductance per unit xylem area (ks). Tracheids were superior to vessels in avoiding freeze—thaw—induced embolism, and had lower ks. At the other extreme, ring—porous xylem had the highest ks but lost \u3e90% of hydraulic conductance after a single freeze—thaw event. Vulnerability to water—stress—induced cavitation was not correlated with conduit volume or ks. Dicot species either reversed winter embolism by refilling vessels with positive root pressures during spring (Betula, Alnus spp.), or tolerated it and relied on new xylem production to restore hydraulic conductance (Quercus sp.). Conifers reversed embolism by refilling tracheids in the absence of positive pressure. Populus species behaved inconsistently, showing some reversal one year but none the next. Even without embolism reversal, Populus species had hydraulic conductances per unit leaf area equal to other diffuse—porous species