Predicting tropical plant physiology from leaf and canopy spectroscopy

Area of Science:

• Ecology
• Physiology
• Remote Sensing

Background:

• Tropical forest photosynthesis is crucial for regional carbon cycles but difficult to study due to canopy access and species diversity.
• Previous efforts to model photosynthesis have been limited by these challenges.

Purpose of the Study:

• To develop an empirical model predicting sunlit, light-saturated tropical leaf photosynthesis.
• To assess the utility of leaf and canopy spectral data for this prediction.
• To explore the potential for remote sensing of tropical forest canopy photosynthesis.

Main Methods:

• Utilized partial least squares (PLS) analysis on three tropical forest datasets (159 species).
• Measured leaf-level photosynthetic rates (A, A(max), R), spectral properties (400-2,500 nm), and biochemical traits (nitrogen, chlorophylls, carotenoids, LMA).
• Combined leaf spectra with a canopy radiative transfer model to simulate canopy spectra.

Main Results:

• The empirical model accurately predicted light-saturated photosynthesis (A) (r(2) = 0.74).
• Respiration (R) and light/CO2-saturated photosynthesis (A(max)) were also predicted (r(2) = 0.48 and 0.47, respectively).
• Simulated canopy spectra provided better predictions of A (RMSE = 2.5 ± 0.07 μmol m(-2) s(-1)) than leaf spectra.

Conclusions:

• Spectral data, particularly canopy spectra, can effectively predict tropical leaf photosynthesis.
• This approach holds significant potential for remote sensing of tropical forest canopy photosynthesis using imaging spectrometers.
• Advances in remote sensing can overcome traditional challenges in studying tropical forest ecosystems.