Estimation of annual forest evapotranspiration from a coniferous plantation watershed in Japan (1): Water use components in Japanese cedar stands

To increase the ability to control forest ecosystem water and carbon cycles using forest management, we estimated watershed-scale evapotranspiration (ET) and its components, i.e., upper-canopy stand transpiration (E_UC), sub-canopy vegetation transpiration (E_SC), and canopy interception (I_C), in a Japanese cedar (Cryptomeria japonica D. Don.) plantation over a whole year. For E_UC, xylem sap flux density was measured in three plots: an upper (UP) and lower (LP) plot on a northeast-facing, and one on a south-facing slope (SP). Mean stand sap flux density (J_S) in the UP, LP, and SP was similar despite differences among plots in tree density and size, implying that J_S measured in a partial stand within the watershed is a reasonable estimator of the values of other stands, and that stand sapwood area is a strong determinant of the E_UC. Prior information on annual variations in ET and its components was insufficient and urgently needed in Japan. Using a combination of observations and modeling, we obtained reliable estimations of E_SC and I_C, and thus, of annual variations in ET and its components (911.4, 359.3, 126.9, and 425.2mm/year for ET, E_UC, E_SC and I_C, respectively). We found a conservative ratio of I_C to rainfall (P) (I_C/P) throughout the year, a significant contribution of I_C/P to the ratio of ET to P (ET/P) during heavy rainfall conditions, and an increase in I_C and E_SC when E_UC was decreasing, resulting in a constant monthly ET/P in the growing season and winter. These support the idea of the conservative process of forest water use in that P mainly controls ET on a monthly and longer time scale.