The Applicability Of Thermal Dissipation Probe Method In Six Tree Species

There are conflicting reports on the accuracy of thermal dissipation probe (TDP, theGranier method) measurement which is widely used to estimate transpiration of individualtrees. In this article, six woody species, representing three wood types, were used to studypossible association between TDP measurement accuracy and anatomical characteristics ofdiameter and density distribution, sap flow density, environmental temperature, as well as sapsolutions. In addition, we investigated the effects of different orientations and probes on theTDP measurement. Our results showed that the thermal dissipation probe technique withGranier's original equation underestimated to various degrees the sap flux density in sixspecies. We found our calibration results for two conifers, with small diameter and high densityof tracheids, were relatively consistent with the Granier's calibration. For Platycladusorientalis and P. bungeana, a is0.137and0.129and b is1.195and1.197while a is0.119andb is1.231in Granier's formula. However, with larger diameter and lower density of vessels inthe two diffuse-porous species, the underestimation of sap flow by the original calibrationbecame bigger. Compared with the above two wood types, two ring-porous species had thelargest diameter and lowest density of vessels. Particularly, the R. psedoacacia possessedshallowest sap wood depth, less than the probe length. In coinciding with the wood anatomicalcharacteristics, our calibration for the ring-porous species, especially R. psedoacacia, deviatedfar from the original calibration which most underestimated the sap flow. The underestimationkept at80%. For the same species, the underestimation of TDP under different sap flow isdifferent. The underestimation will be larger if the sap flow is too small or large. Theunderestimation degree was well associated with sap wood depth, radial diameter and densitydistribution of conduits and sap flux density.We found that temperature difference in a range between25to0°C did not changemeasurement of ΔTmvalues, testifying the ΔT_m measured in night is validated as a reference toΔT in daytime in terms of temperature difference. Furthermore, additions of KCl and ethanol into the fluid did not alter the measurement accuracy with TDP, suggesting the technique isindependent of fluid ionic or organic concentrations in a certain range. Through thecomparation of TDP1(TDP-10, Dynamax HT, USA) and TDP2(TDP-10, Yugen, Beijing,China), we can confirm that they both have the same accuracy and time lag. The maximal sapflux density of Platanus orientalis L. measured by20mm probes is double than10mm. Thisilluminates that the sap flux density mesasured by different probes for the same tree is alsoconfliciting. So the probe length must be considered in the measurement. Therefore, wesuggested to consider the sap wood depth, radial diameter and density distribution of conduits,sap flux density and length of probes to obtain the more accurate sap flow.