Comparison of tissue heat balance- and thermal dissipation-derived sap flow measurements in ring-porous oaks and a pine

Sap flow measurements have become integral in many physiological and ecological investigations. A number of methods are used to estimate sap flow rates in trees, but probably the most popular is the thermal dissipation (TD) method because of its affordability, relatively low power consumption, and ease of use. However, there have been questions about the use of this method in ring-porous species and whether individual species and site calibrations are needed. We made concurrent measurements of sap flow rates using TD sensors and the tissue heat balance (THB) method in two oak species (Quercus prinus Willd. and Quercus velutina Lam.) and one pine (Pinus echinata Mill.). We also made concurrent measurements of sap flow rates using both 1 and 2-cm long TD sensors in both oak species. We found that both the TD and THB systems tended to match well in the pine individual, but sap flow rates were underestimated by 2-cm long TD sensors in five individuals of the two ring-porous oak species. Underestimations of 20-35% occurred in Q. prinus even when a "Clearwater" correction was applied to account for the shallowness of the sapwood depth relative to the sensor length and flow rates were underestimated by up to 50% in Q. velutina. Two centimeter long TD sensors also underestimated flow rates compared with 1-cm long sensors in Q. prinus, but only at large flow rates. When 2-cm long sensor data in Q. prinus were scaled using the regression with 1-cm long data, daily flow rates matched well with the rates measured by the THB system. Daily plot level transpiration estimated using TD sap flow rates and scaled 1 cm sensor data averaged about 15% lower than those estimated by the THB method. Therefore, these results suggest that 1-cm long sensors are appropriate in species with shallow sapwood, however more corrections may be necessary in ring-porous species.