| Recent studies have shown how human induced nitrogen (N) and phosphorous (P) imbalances
affect essential ecosystem processes, and might be particularly important in water-limited
ecosystems. Hyperspectral information can be used to directly infer nutrient-induced
variation in structural and functional changes of vegetation under different nutrient
availability. However, several uncertainties still hamper the direct link between
photosynthetic CO2 uptake (gross primary productivity, GPP) and hyperspectral reflectance.
Sun-induced fluorescence (SIF) provides a new non-invasive measurement approach that has
the potential to quantify dynamic changes in light use efficiency and photosynthetic CO2
uptake.
In this contribution we will present an experiment conducted in a Mediterranean
grassland, where 16 plots of 8x8 meters were manipulated by adding nutrient (N, P, and NP).
Almost simultaneous estimates of canopy scale GPP and SIF were conducted with
transparent transient-state canopy chambers and high resolution spectrometers, respectively.
We investigated the response of GPP and SIF to different nutrient availability and plant
stoichiometry. The second objective was to identify how structural (LAI, leaf angle
distribution, and biodiversity) and canopy biochemical properties (e.g. N and chlorophyll
content - Chl) control the functional relationship between GPP and SIF. To test the different
hypotheses the SCOPE radiative transfer model was used. We ran a factorial experiment with
SCOPE to disentangle the main drivers (structure vs biochemistry) of the relationship
GPP-SIF.
The results showed significant differences in GPP values between N and without N
addition plots. We also found that vegetation indices sensitive to pigment variations
and physiology (such as photochemical reflectance index PRI) and SIF showed
differences between different treatments. SCOPE showed very good agreement with the
observed data (R2=0.71). The observed variability in SIF was mainly related to
changes in functional traits of the vegetation (changes in N and P content and Chl).
However, beside changes in functional traits, changes in canopy structure (and in
particular variation in plant forms abundance after fertilization) controlled the GPP-SIF
relationship. According to these results, plant N/P stoichiometry and structure should be
considered when modelling GPP assuming a linear relationship with SIF at grasslands
sites. |