Can the Ecological Characteristics of Plant Species Be Predicted from Their Chlorophyll a Fluorescence Signature?

Chlorophyll a fluorescence provides an integrated measure of photosynthetic functioning, but the extent to which species fluorescence signatures reflect evolutionary history and ecological differentiation remains poorly understood. Chlorophyll a fluorescence was analysed using 2667 OJIP measurements from 263 vascular plant species to quantify interspecific differentiation, assess phylogenetic and ecological structuring of OJIP parameters, and test whether species ecological characteristics can be predicted from multivariate fluorescence signatures. Interspecific differentiation was substantial across all 17 OJIP parameters, with intraclass correlation coefficients ranging from 0.47 to 0.84. Similarly strong differentiation arose from contrasting sources: ET0/RC, VJ, φEo, ET0/CS0, and K were predominantly phylogenetically structured, whereas ABS/CS0, TR0/CS0, DI0/CS0, and RC/CS0 showed stronger ecological structuring. Repeated out-of-sample validation across four regression approaches demonstrated substantial differences in predictability among ecological characteristics. Prediction was most successful for light (median out-of-sample R2 = 0.44–0.48 across regression approaches; interquartile bounds = 0.38–0.53). Lower predictive performance was observed for topographic moisture (median R2 = 0.25–0.31; interquartile bounds = 0.19–0.37), temperature (0.21–0.30; 0.18–0.34), and continentality (0.18–0.28; 0.11–0.32). Hemeroby exhibited comparable median predictive performance (R2 = 0.25–0.33; interquartile bounds = 0.15–0.40), with considerable variability across validation splits, including negative predictive R2 values in the lower tails of the distributions. These results were based on R2 calculated relative to the training-set mean; using the test-set mean produced similar median predictive performance and did not alter the overall ranking of the principal ecological indicators. Phylogenetic-exclusion validation further showed that predictive information for light and continentality remained comparatively transferable across phylogenetic separation, whereas prediction of salinity and naturalness declined markedly. OJIP traits provided additional predictive information beyond that captured by phylogenetic predictors and species-level ecological light preferences across most ecological indicators and modelling approaches. After controlling for training-set size, phylogenetic exclusion had a relatively small additional effect on most ecological predictions, with the greatest reductions in predictive performance observed for salinity and, under strong phylogenetic exclusion, light. Chlorophyll a fluorescence signatures contain differential information about species ecology, varying in both predictive strength and phylogenetic generality.