Numerous studies have investigated reflectance-based estimations of physico-chemical leaf traits such as the contents of light absorbing pigments, leaf mass per area, or carbon and nitrogen contents. Only few studies, however, attempted to estimate leaf traits that are more directly linked to photosynthesis. We tested the feasibility of estimating two important photosynthesis traits, the maximum carboxylation capacity (V_cmax,25) and the maximum electron transport rate (J_max,25), from in-situ leaf reflectance spectra using approaches that are applicable also on larger spatial scales. We conducted measurements of reflectance, CO₂ response curves, leaf mass per area (LMA), and nitrogen content per area (N_a) for 37 temperate deciduous tree species and a total of 242 leaves from widely differing light environments. Partial least squares (PLS) regression was used to estimate V_cmax , 25, J_max , 25, LMA, and N_a from reflectance spectra.

The results showed that both V_cmax , 25 and J_max , 25 can be estimated from leaf reflectance measurements with good accuracy (R² = 0.64 for V_cmax , 25, R² = 0.70 for J_max , 25) even for a large number of tree species and varying light environments. Detailed analysis of reflectance-based PLS and linear regression models with regard to prediction performances and regression coefficients led to the conclusion that the correlation to N_a was the dominating mechanism on which the V_cmax , 25 and J_max , 25 PLS models were based. The PLS regression coefficients of N_a, V_cmax , 25 and J_max , 25 showed clear signatures of nitrogen-related absorption features.

The finding that V_cmax , 25 and J_max , 25 estimations from leaf reflectance are predominantly based on their relationships to N_a has important implications for large scale estimations of these photosynthesis parameters. We suggest that future studies should focus more on large scale estimation of N_a from remote sensing and estimate V_cmax , 25 and J_max , 25 in a separate step using their respective relationships to N_a.