Dominant temperate and subalpine Japanese trees have variable photosynthetic thermal optima according to site mean annual temperature

Aim: Global flux data analyses have shown a significant positive and linear relationship between site-scale photosynthetic optimum temperature (T_opt-s) and averaged temperature variables. However, as existing studies have not fully considered species composition, it remains unclear to what extent the change in T_opt-s is derived from intraspecific plasticity or from a difference in species with a consistent species-specific T_opt-s. We tested these two hypotheses using the satellite-derived enhanced vegetation index (EVI). Location: Subalpine and temperate forests in Japan. Time period: 2001–2020. Major taxa studied: Thirty-three tree species. Methods: Based on an in situ vegetation survey dataset, we identified 1814 EVI cells (250 m × 250 m) that were characteristic of the spectral reflectance of a single dominant species. For each cell, we pooled EVIs from 2001 to 2020 and estimated the temperature at the time of EVI peak as a species-specific T_opt-s. We tested our hypotheses by fitting linear regression models to the T_opt-s and mean annual temperature (MAT). A positive slope indicated support for the intraspecific plasticity hypothesis. Results: For 32 of the 33 tree species, the T_opt-s increased with an increase in MAT. A linear mixed-effect model provided a good explanation of the variation in the T_opt-s based on the MAT (intercept = 14.63, slope = 0.58, marginal R² =.78), with minimal effect of species difference (conditional R² =.75). Main conclusions: Our results support the intraspecific plasticity hypothesis. At least in East Asian subalpine and temperate forests, the T_opt-s can be estimated from the MAT, irrespective of the dominant canopy species. The observed MAT–T_opt-s relationship is consistent with the findings of previous photosynthetic studies and has important implications for the establishment of new algorithms for estimating gross primary productivity that can account explicitly for spatial and temporal differences in temperature-dependent photosynthetic responses.