Re-Evaluating a Stable Isotope (δ18O) Approach for Estimating the Temperature of Photosynthesis

Some studies suggest tropical forests operate near a leaf temperature threshold, above which reduced photosynthesis will cause dieback and conversion to new biomes. Other studies show a homeostasis of leaf temperatures, suggesting plants may be buffered against temperature increases. Assessing the global generality of these results is challenging, however, because we lack methods for measuring physiologically-relevant leaf temperatures across macroecological scales of time and space. One promising approach uses a cellulosic δ¹⁸O model with data for climate and δ¹⁸O for plant cellulose and source water to yield a time-integrated estimate of the average temperature at which photosynthesis is most productive. The approach suggests most photosynthesis occurs at ~21 °C across latitude from subtropical to boreal forests. However, the approach has been debated for its treatment of post-photosynthesis oxygen exchange processes. Here we re-evaluate the approach. First, we quantify effects of oxygen exchange on temperature estimates using data for branches and leaves of 8 tree species spanning a 11 °C temperature gradient in Biosphere 2, the world's hottest tropical rainforest. Second, we examine the macroecological implications of oxygen exchange using data for alpine plants spanning an elevational gradient in Colorado, and trees spanning a latitudinal gradient from Panama to Oregon. In Biosphere 2, we observed substantial differences between temperatures estimated using leaf and branch δ¹⁸O. Temperatures estimated from branch data were invariant with air temperature, while those from leaf data varied more but were still buffered relative to air. Similar results were obtained for the Colorado and Panama-to-Oregon gradients. Thus, leaf cellulose is more suitable than wood cellulose for δ¹⁸O estimates of photosynthesis temperature. Together, these results suggest the earlier global value of 21 °C for photosynthesis reflects not only a relative homeostasis of plant temperatures, but also post-photosynthesis oxygen exchange that makes wood cellulose δ¹⁸O more similar to source water. This general decoupling of plant and air temperatures is consistent with the homeostasis hypothesis and may constitute a thermal refuge in the face of a changing climate.