Can we assume insignificant temporal changes in spatial variations of sap flux for year-round individual tree transpiration estimates? A case study on Cryptomeria japonica in central Taiwan

Abstract

Although spatial variations in sap flux changed seasonally in cedar trees, neglecting these seasonal changes generally resulted in less than 10% errors in year-round tree transpiration estimates. Spatial and temporal variations in sap flux (F d ) affect the accuracy of year-round individual tree water use (Q) estimates. We aimed to determine seasonal changes in radial and azimuthal variations in F d and to evaluate their potential impacts on year-round Q estimates for humid subtropical forest trees. We measured F d using 38 probes in eight Cryptomeria japonica trees in Taiwan from July 2010 to May 2011. During the study period, the ratio of inner F d to outermost F d (R r ) ranged between 0.29 and 0.68. The ratio of mean outermost F d to that at one direction (R a ) ranged between 0.36 and 1.89. The seasonal patterns were inconsistent among individuals. We compared year-round Q estimates derived from multi-sensor measurements (Q m ) with those derived from measurements at a certain position with constant correction factors for radial (Q r ) or azimuthal (Q a ) patterns of F d determined from the summer measurements. This simple exercise revealed consistent relationships between Q m and Q r year-round (<10% error), suggesting that seasonal changes in the radial profiles of F d had little effect on Q estimates. Although using a constant correction factor for azimuthal variations led to 20–40% errors in Q a of some individuals with large azimuthal variations (i.e., CV > 0.49), five out of eight individuals had less than 10% errors in their Q a estimates. These results suggested that constant correction factors can be used for radial variations in year-round Q estimates, but are not applicable for azimuthal variations in individuals showing large azimuthal variations of F d .