Estimating photosynthesis with high resolution field spectroscopy in a Mediterranean grassland under different nutrient availability

Abstract:

Recent studies have shown how human induced N:P imbalances are affecting essential processes (e.g. photosynthesis, plant growth rate) that lead to important changes in ecosystem structure and function. In this regard, the accuracy of the approaches based on remotely-sensed data for monitoring and modeling gross primary production (GPP) relies on the ability of vegetation indices (VIs) to track the dynamics of vegetation physiological and biophysical properties/variables. Promising results have been recently obtained when Chlorophyll-sensitive VIs and Chlorophyll fluorescence are combined with structural indices in the framework of the Monteith’s light use efficiency (LUE) model. However, further ground-based experiments are required to validate LUE model performances, and their capability to be generalized under different nutrient availability conditions.

In this study, the overall objective was to investigate the sensitivity of VIs to track short- and long-term GPP variations in a Mediterranean grassland under different N and P fertilization treatments. Spectral VIs were acquired manually using high resolution spectrometers (HR4000, OceanOptics, USA) along a phenological cycle. The VIs examined included photochemical reflectance index (PRI), MERIS terrestrial-chlorophyll index (MTCI) and normalized difference vegetation index (NDVI). Solar-induced chlorophyll fluorescence calculated at the oxygen absorption band O2-A (F760) using spectral fitting methods was also used. Simultaneously, measurements of GPP and environmental variables were conducted using a transient-state canopy chamber.

Overall, GPP, F760 and VIs showed a clear seasonal time-trend in all treatments, which was driven by the phenological development of the grassland. Results showed significant differences (p<0.05) in midday GPP values between N and without N addition plots, in particular at the peak of the growing season during the flowering stage and at the end of the season during senescence. While NDVI did not show any significant difference between treatments, VIs sensitive to pigment variations and physiology (PRI, MTCI) and F760 behaved as GPP. Model performance test indicated that VIs related to physiology and fluorescence are key to account for nutrient availability in LUE models and to better predict GPP.