Seasonal variability in leaf reflectance, the photochemical reflectance index, and their response to light stress in a mixed temperate deciduous forest in Ontario, Canada

Variations in leaf reflectance can be correlated with measurements of plant physiology, phenological status and leaf functional traits that are important inputs to ecosystem models. Remote sensing enables assessment of how these traits vary with spatial continuity not possible using data collected in situ. However, it remains difficult to generalize relationships between reflectance spectra and leaf functional traits due to variability in these relationships throughout the season and across species.

We measured top of canopy leaf reflectance and key leaf functional traits on three deciduous broadleaf species (Populus grandidentata, Fraxinus pennsylvanica, Acer rubrum), at the Borden Forest Research Station (44°19'N, 79°56'W), on 19 dates over the 2018 growing season. We sampled hyperspectral leaf reflectance using an ASD FieldSpec spectrometer and contact probe. We calculated several vegetation indices including PRI, NDVI, and chlorophyll related indices. Reflectance factors were compared to gas exchange, CO₂ response curves (used to determine Vcmax), and PAM fluorescence measurements using a LI-6400xt portable photosynthesis system and laboratory-analyzed leaf chlorophyll and carotenoid content. To assess the ability of PRI to respond to light stress throughout the season, reflectance was also measured on dark adapted leaves that were suddenly exposed to light. The results showed:

- seasonal patterns in vegetation indices differed between species, particularly in spring and fall

- during a period of drought neither reflectance at 531nm nor PRI decreased on dark adapted leaves exposed to a dark-light transition

- the sensitivity of PRI to light stress reversed in senescent leaves (i.e. the value of PRI increased rather than decreased when we induced light stress on senescent leaves), and

- the seasonal timing of this senescence-specific PRI response differed for each species studied.

Our findings illuminate the seasonal and species-specific variability in how PRI relates to light stress. Sub-pixel species variability and phenological differences between species impact interpretation of PRI and other indices. The remote sensing community will need to account for this seasonal and species-specific variability in order to meaningfully interpret vegetation indices measured from space.