A new modelling approach for estimating abiotic and biotic stress-induced de novo emissions of biogenic volatile organic compounds from plants

The emission of biogenic volatile organic compounds (BVOCs) is usually thought to depend on species-specific emission capacities that vary with seasonal and phenological conditions. Actual—so called constitutive—emissions are then calculated from prevailing temperature and radiation. However, various abiotic and biotic stressors such as ozone, extreme radiation and temperature conditions, as well as wounding e.g., from insect feeding, can lead to de-novo emissions of stress-induced BVOCs (sBVOCs) that may excel constitutive emissions by more than an order of magnitude. These emissions often have a considerable different compound composition and are short-lived but can prolong under continuous stress for quite some time. Thus, they may easily have a large impact on overall regional BVOC emissions. However, sBVOCs are generally not considered in models since up to date no consistent mechanism has been proposed. This manuscript suggests a new mechanism based on the finding that sBVOCs originate from a handful of biosynthetic pathways which synthesize compounds in the groups of monoterpenes, sesquiterpenes, and green leave volatiles, as well as methyl salicylate, ethanol/acetaldehyde, methanol/formaldehyde, and acetone. ... mehrIsoprene is also considered but since it is often constitutively emitted, the specific role of stress induction is difficult to determine for this compound. A function is proposed that describes the production of all de-novo sBVOCs sufficiently well and scales with stress intensity. It is hypothesized that the response delay and the form of the function is specific for the production pathway and valid for ozone as well as wounding (herbivory) induced stress. Model parameters are then derived from pooled literature data based on a meta-analysis of suitable induction-response measurements of different plant species. The overall emission amount derives from the intensity and frequency of the stress impulse. We present a number of literature studies that are used to parameterize the new model as well as a selection of evaluations for single- and multiple-stress inductions. Furthermore, coupling and interaction with constitutive emission models as well as limitations and possible further developments are discussed.