| The progress of the present domestic research on the digital camera-based phenology is very slow, especially for the analysis of plant physiological mechanisms, which are the major driving factors for phenology. In this study, we combine with the manual observation, measurements of leaf physiological and biochemical indices, and digital carmera-based observation to analyze the pheological characteristics of an oak forest, located in Xiashu town in Jiangsu during 2014. Our results, which document the seasonal variation of the oak canopy pheology and the corresponding leaf physiological properties, indicate a relationship between a phenological color-based index (i.e. Green Chromatic Coordinate, Gcc) and leaf physiology, and demonstrate the applicability of the digital camera-based imagery in canopy phenology research. Specifically, the results show that:(1) Growing season phenology of an oak forest could be divided based on seasonal changes of camera-based phenological color indices. The early green-up date started on the 80th day of the year (Day-of-Year, DOY) in 2014, and ended at DOY 86; between DOY 86 to 104, Gcc, LAI and SPAD all increased gradually, and Gcc reached a seasonal peak at DOY 104. From DOY 104-160, Gcc began to decline, while LAI and SPAD continued to increase to their seasonal peak. From DOY 160-278, oak leaves matured and color was stable. Gcc, Leaf Area Index (LAI) and Soil Plant Analysis Development (SPAD) all were stable; leaves began to fall at DOY 278 and by DOY 335, all leaves had fallen from the canopy. The whole season length was 255 days around.(2) The seasonal trajectories of leaf chlorophyll concentration which based leaf area (chla), net photosynthetic rate (Pn) and Leaf Mass per Area (LMA) were similar:as leaves developed, leaf trait values increased to a seasonal peak, stabilized through early autumn, thereafter chla and Pn declined as leaves senesced. The seasonal trajectories of leaf total nitrogen concentration (TN) and equivalent water thickness (EWT) followed a similar pattern:they declined gradually in spring, were stable in summer, and continued to decrease in autumn. We also divided the whole canopy into upper, middle and lower layer, and found that the value of Pn, TN and LMA of upper layer was greatest, followed by the middle layer and lower layers. However, the chlorophyll concentration which based leaf dry mass showed the converse pattern.(3) In both spring or in autumn, we found the camera-based Gcc had a significant correlation with leaf biochemical and physical properties, except for Pn during senescence. In spring, chla and LMA increased at the slope of 591.87 and 829.28, respectively, when Gcc changed (R2=0.856 and 0.930), while the relationship of Gcc and Pn, N and EWT followed a polynomial function (R2=0.700,0.789 and 0.569, respectively). For data in collected in autumn, we fit an exponential function of Gcc and chla, Pn and LMA (R2=0.866,0.682 and 0.794) and polynomial function of Gcc and LMA and EWT(R2=0.794 and 0.930). Our finds suggested that after corrected, camera-based Gcc could be used as a proxy of leaf biochemical and biophysical properties within a certain range.(4) In general, the seasonal trajectory of Gcc tracks those of chla, Pn and LMA, while the seasonal peak of Gcc was earlier than chla, Pn and LMA. Its spring trajectory showed leaf function was altered, and the day when Gcc came to its seasonal peak is corresponds to its function shifting from increasing size to thickness.(5) Our results show that camera-based phenological observation is an effective technique to quantify the dynamic variation of canopy and monitor the phenological change of deciduous forest. |
PDF Full Text Request