Recent findings based on the oxygen isotope ratios of tree trunk cellulose indicate that the temperature of biomass production in biomes ranging from boreal to subtropical forests converge to an average leaf temperature of 21.4°C. The above conclusion has been drawn under the assumption that biochemically related isotopic fractionations during cellulose synthesis are not affected by temperature. Here we test the above assumption by heterotrophically generating cellulose at different temperatures and measuring the proportion of carbohydrate oxygen that exchange with water during cellulose synthesis and the average biochemical fractionation associated with this exchange. We observed no variation in the proportion of oxygen that exchange with different temperatures, which averaged 0.42 as it has been observed in other studies. On the other hand, the biochemical oxygen isotope fractionation during cellulose synthesis is affected by temperature and can be described by a 2nd order polynomial equation. The biochemical fractionation changes little between temperatures of 20 and 30°C averaging 26‰ but increases at lower temperatures to values of 31‰. This temperature sensitive biochemical fractionation explains the pattern of cellulose oxygen isotope ratios of aquatic plants encompassing several latitudes. The observed temperature sensitive biochemical fractionation also indicates that divergent biochemical fractionation and not convergent leaf temperature explains the increase in oxygen isotope enrichment of cellulose across several biomes.

Styrax caporum is a native shrub from the Brazilian savanna. Most of its leaves are diaheliotropic, whereas some are paraheliotropic, mainly at noon. A previous study of this species revealed higher stomatal conductance (gs) and transpiration rates (E) in para-compared to diaheliotropic leaves, and a rise in CO 2 assimilation rates (A) with an increase of irradiance for paraheliotropic leaves. We hypothesized that this species exploits the paraheliotropism to enhance the light use efficiency, and that it is detected only if gas exchange is measured with light interception by both leaf surfaces. Gas exchange was measured with devices that enabled light interception on only one of the leaf surfaces and with devices that enabled light interception by both leaf surfaces. Water relations, relative reflected light intensity, leaf temperature (T l), and leaf anatomical analyses were also performed. When both leaf surfaces were illuminated, a higher A, E, and gs were observed in para-compared to diaheliotropic leaves; however, A did not depend on gs, which did not influence CO 2 accumulation in the stomatal cavity (Ci). When only the adaxial leaf surface was illuminated, a greater A was detected for para-than for diaheliotropic leaves only at 11:00 h; no differences in T l were observed between leaf types. Light curves revealed that under non-saturating light the adaxial side of paraheliotropic leaves had higher A than the abaxial side, but they showed similar values under saturating light. Although the abaxial leaf side was highly reflective, both surfaces presented the same response pattern for green light reflection, which can be explained by the compact spongy parenchyma observed in the leaves, increasing light use efficiency in terms of CO 2 consumption for paraheliotropic leaves. We propose that paraheliotropism in S. camporum is not related to leaf heat avoidance or photoprotection.

No abstract