Chlorophyll fluorescence can serve as a proxy of photosynthesis in boreal forests. When sustained non-photochemical quenching (NPQS) relaxes towards summer, leaf chlorophyll fluorescence (ChlF) emission increases along with photosynthesis. Yet, other physical and physiological factors can also leave a measurable imprint on the fluorescence emission spectra, and disrupt this relationship.

We measured leaf-level spectral ChlF of Scots pine, Norway spruce and lingonberry exposed to contrasting light environments throughout the spring recovery of photosynthesis, simultaneously with a series of photosynthetic, biochemical and morphological traits. Correlations between traits and ChlF spectral components were analyzed to identify the mechanisms underlying both the spatial variation found between species and light environments, and the temporal variation found across the study period.

Spatially, we found evidence of baseline differences in leaf-level ChlF magnitude, which we attribute to species- and light environment-specific changes in leaf morphology. Temporally, ChlF magnitude followed the relaxation of NPQS  towards summer, but only in upper canopy foliage and lingonberry, suggesting a seasonal compensation effect between sustained photochemical quenching (PQS) and NPQS, potentially decoupling the seasonal relationship between ChlF and photosynthesis in shaded foliage. Finally, we show subtle changes in the shape of the ChlF spectra that took place independently of chlorophyll concentration dynamics, pointing to the complexity of NPQs which can involve structural rearrangements in the thylakoids and changes in the relative contribution of PSI to emitted ChlF.

We conclude that the diversity of species and light environments found within an ecosystem generates a baseline level of variation in leaf spectral ChlF as well as contrasting seasonal photosynthetic acclimation patterns. These sources of variability should be taken into account when developing quantitative models for the interpretation of ChlF data, in particular for applications involving high resolution SIF imaging systems capable of resolving different plant individuals and their parts.

The dataset was collected during a measuring campaign carried out at Station for Measuring Ecosystem-Atmosphere Relation (SMEAR-II) in Hyytiälä, southern Finland (61⁰50’ N, 24⁰17’ E) from February to July 2017. The data available here are raw or pre-processed and if it concerns, the pre-processing procedure is described in the readme files available in each variable-dedicated folder. The chlorophyll fluorescence dataset (folder: spectral_fluorescence) has been collected with an Optical Chamber protocol as described in the methods section of the associated manuscript referenced above. From the dataset available in this depository, the chlorophyll fluorescence data was processed by a series of Singular Value Decomposition (SVD) analyses to evaluate the relative role of vectors in explaining the spatial and temporal variation of chlorophyll fluorescence spectra. The remaining datasets (folders: PAM fluorescence, gas exchange, pigments, light environment, and SMEAR, where environmental variables from the station are available) have been used to test their correlations with the SVD vectors in order to investigate the relative roles that different variables have in controlling the spatial and temporal variation of chlorophyll fluorescence spectra. Results of these analyses have been presented in the publication titled "The photosynthetic response of spectral chlorophyll fluorescence differs across species and light environments in a boreal forest ecosystem" in the Agricultural and Forest Meteorology Journal.

The dataset is divided into folders for each of the measured variables. In the SMEAR folder, three variables regarding environmental variables of the SMEAR-II measuring station are included. In the spectral fluorescence folder, separate files for each analyzed case (species and canopy position) are included. The readme file in each folder contains an explanation of each of the variables in the dataset, its measurement device, measurement units, file layout, and pre-processing (if concerns). Information on how the measurements were done can be found in the associated manuscript referenced above.