6:15 PM - 7:30 PM
[ACG30-P06] Seasonal change in photosynthetic function of cedar and cypress forest estimated by tower-based chlorophyll fluorescence
Keywords:Remote sensing, Light use efficiency, Eddy covariance, Carbon cycle
Chlorophyll fluorescence is emitted from the chlorophyll a and b to release the excess sun-light energy, and has been used for detecting the photosynthesis stress in numerous ecophysiological studies so far. Recently, the chlorophyll fluorescence has been utilized to represent the gross primary production (GPP), ecosystem photosynthesical activity, by the satellite remote sensing studies (e.g. Frankenberg et al., 2011, GRL). However, small number of chlorophyll fluorescence observation on the ecosystem scale at the ground reduces its availability despite its highly expected usefulness.
The aim of this study is to clarify the relationships between chlorophyll fluorescence, and photosynthesis and light use efficiency (LUE) by the ground based measurement in coniferous evergreen forest. The observations were carried out in the plantation forest consisting of mature Japanese cedar and Japanese cypress in Takayama, Japan, from 2007 to 2008. Downward and upward spectral radiances were measured with hemispherical spectroradiometer (MS-700, Eko Instruments, Japan) mounted at 30m-high above the ground surface. We calculated the Sun-Induced fluorescence (FS) around the O2-A band from the spectral data with the Fraunhofer Line Depth method. The GPP was calculated from the carbon fluxes measured with eddy covariance at the top of the tower (Nagai et al., 2012). Both FS and GPP were averaged for 30 minutes.
FS showed the strong correlation to GPP linearly in the diurnal course (sunny day (2007/08/10): r2 = 0.81, cloudy day (2007/07/24): r2 = 0.87) and logarithmically in the seasonal change (2008 (half-hour): r2 = 0.68, 2008 (daily): r2 = 0.87). The GPP was fitted against the FS for each month by the following rectangular hyperbolic curve:
GPP = αGPPSATFS / (GPPSAT + αFS),
where α is the initial slope of the FS-GPP curve and GPPSAT is the saturated GPP at high FS emission. These two parameters showed the clear seasonal change. The root-mean-square error (RMSE) calculated for 2008 was 4.56 μmolCO2 m-2 s-1. We also investigated the relationship between FS and LUE in daily averages. The FS-LUE relationship could be regressed by logarithm curve for each month (r2 = 0.46 ~ 0.95). The seasonal changes in the regression coefficients for FS-GPP and FS-LUE curves were thought to be induced by the seasonal variation in the temperature-dependency of photosynthesis and the phenology.
We conclude that FS can be utilized to estimate GPP and LUE in evergreen forest, and that relationship between FS and GPP is influenced by environmental factors such as air temperature.