Linking xylem diameter variations with sap flow measurements

Measurements of variation in the diameter of tree stems provide a rapid response, high resolution tool for detecting changes in water tension inside the xylem. Water movement inside the xylem is caused by changes in the water tension and theoretically, the sap flow rate should be directly proportional to the water tension gradient and, therefore, also linearly linked to the xylem diameter variations. The coefficient of proportionality describes the water conductivity and elasticity of the conducting tissue. Xylem diameter variation measurements could thus provide an alternative approach for estimating sap flow rates, but currently we lack means for calibration. On the other hand, xylem diameter variation measurements could also be used as a tool for studying xylem structure and function. If we knew both the water tension in the xylem and the sap flow rate, xylem conductivity and/or elasticity could be calculated from the slope of their relationship. In this study we measured diurnal xylem diameter variation simultaneously with sap flow rates (Granier-type thermal method) in six deciduous species (Acer rubrum L., Alnus glutinosa Miller, Betula lenta L., Fagus Sylvatica L. Quercus rubra L., and Tilia vulgaris L.) for 7-91 day periods during summers 2003, 2005 and 2006 and analyzed the relationship between these two measurements. We found that in all species xylem diameter variations and sap flow rate were linearly related in daily scale (daily average R-2 = 0.61-0.87) but there was a significant variation in the daily slopes of the linear regressions. The largest variance in the slopes, however, was found between species, which is encouraging for finding a species specific calibration method for measuring sap flow rates using xylem diameter variations. At a daily timescale, xylem diameter variation and sap flow rate were related to each other via a hysteresis loop. The slopes during the morning and afternoon did not differ statistically significantly from each other, indicating no overall change in the conductivity. Because of the variance in the daily slopes, we tested three different data averaging methods to obtain calibration coefficients. The performance of the averaging methods depended on the source of variance in the data set and none of them performed best for all species. The best estimates of instantaneous sap flow rates were also given by different averaging methods than the best estimates of total daily water use. Using the linear relationship of sap flow rate and xylem diameter variations we calculated the conductance and specific conductivity of the soil-xylem-atmosphere water pathway. The conductance were of the order of magnitude 10(-5) kg s(-1) MPa-1 for all species, which compares well with measured water fluxes from broadleaved forests. Interestingly, because of the large sap wood area the conductance of Betula was approximately 10 times larger than in other species.