Development and recovery from winter embolism in silver birch:: seasonal patterns and relationships with the phenological cycle in oceanic Scotland

Silver birch (Betula pendula Roth) is increasingly used in the United Kingdom for reforestation. However, recent evidence indicates that, under some circumstances, planted birch can suffer serious and repeated mortality of the apical leaders and branches, with consequent loss of apical dominance and the formation of a contorted stem. Plants from 37 seed sources of silver birch from Scotland and northern England planted at two sites were compared for several characteristics related to hydraulic architecture, vulnerability to freeze-thaw cycle induced embolism and spring recovery from winter embolism during the period 2000-2002. Phenological rhythms were also monitored in late winter-early spring to document relationships between phenology and water relations parameters. Significant differences were found across seed sources in stage of bud flushing for four dates in spring. Early flushing seed sources differed by about 1 to 2 weeks from late-flushing seed sources across the two sites. Wintertime xylem embolism in stems reached a peak of about 50 to 70% loss of xylem hydraulic conductivity, depending on the size and position of the sample shoots in the canopy. Small apical shoots were significantly more embolized than large basal shoots. Development of winter embolism was coupled to the occurrence of frost events. As percent loss of hydraulic conductivity increased during the winter, wood relative water content declined. Embolism reversal occurred rapidly in spring at the time of development of positive root pressure. No significant differences in the degree of winter embolism in 2001 were found among the three seed sources examined. The investigation was expanded in the winter-spring of 2002 to include 10 seed sources across both sites. Significant differences were found in degree of winter embolism across sites, dates and seed sources. For each date, there was a significant relationship between flushing scores and wood relative water contents across the two sites and all seed sources, suggesting that differences in time of flushing across sites and seed sources were likely caused by differences in the time of occurrence of root pressure, a necessary precondition to flushing.