Variation in Isoprene Emission from Quercus rubra: Sources, Causes, and Consequences for Estimating Fluxes

Isoprene is the dominant volatile organic compound produced in many forest systems. Uncertainty in estimates of leaf level isoprene emission rate stems from an insufficient understanding of the patterns and processes controlling isoprene emission capacity in plant leaves. Previous studies suggest that variation in isoprene emission capacity is substantial; however, it is not known at what scale emission capacity is the most variable. Identifying the sources of variation in emission capacity has implications for conducting measurements and for model development, which will ultimately improve emission estimates and models of tropospheric chemistry. In addition, understanding the sources of variation will help to develop a comprehensive understanding of the physiological controls over isoprene emission. This study applied a variance partitioning approach to identify the major sources of variation in isoprene emission capacity from two populations of northern red oak (Quercus rubra) over three growing seasons. Specifically, we evaluated variation due to climate, populations, trees, branches, leaves, seasons, and years. Overall, the dominant source of variation was the effect of a moderate drought event. In the years without drought events, variation among individual trees (intraspecific) explained approximately 60% of the total variance. Within the midseason, isoprene emission capacity of sun leaves varied by a factor of 2 among trees. During the third year a moderate 20-day drought event caused isoprene emission capacity to decrease fourfold, and the relative importance of intraspecific variation was reduced to 24% of total variance. Overall, ambient temperature, light, and a drought index were poor predictors of isoprene emission capacity over a 0 to 14-day period across growing seasons. The drought event captured in this study emphasizes the need to incorporate environmental influences into leaf level emission models