Insect Herbivory Strongly Modifies Mountain Birch Volatile Emissions

Abstract

Insect herbivory is known to augment emissions of biogenic volatile organic compounds (BVOCs). Yet few studies have quantified BVOC responses to insect herbivory in natural populations in pan-Arctic regions. Here, we assess how quantitative and qualitative BVOC emissions change with increasing herbivore feeding intensity in the Subarctic mountain birch (Betula pubescens var pumila (L.)) forest. We conducted three field experiments in which we manipulated the larval density of geometrid moths (Operophtera brumata and Epirrita autumnata), on branches of mountain birch and measured BVOC emissions using the branch enclosure method and gas chromatography-mass spectrometry. Our study showed that herbivory significantly increased BVOC emissions from the branches damaged by larvae. BVOC emissions increased due to insect herbivory at relatively low larvae densities, causing up to 10% of leaf area loss. Insect herbivory also changed the blend composition of BVOCs, with damaged plants producing less intercorrelated BVOC blends than undamaged ones. Our results provide a quantitative understanding of the relationship between the severity of insect herbivore damage and emissions of BVOCs at larvae densities corresponding to background herbivory levels in the Subarctic mountain birch. The results have important and practical implications for modeling induced and constitutive BVOC emissions and their feedbacks to atmospheric chemistry.

Keywords: arctic; biotic stress; geometrid moth; insect herbivory; mountain birch; stress severity; volatile organic compounds.

FIGURE 1

Effects of larval density on total BVOC emissions in three experiments; (A) Abisko-1, (B) Abisko-2, and (C) Tromsø. The total BVOC emissions, expressed as ng cm^–2 leaf area h^–1, are presented stacked for the following compound groups: Benzenoids; GLV, green leaf volatiles; HT, homoterpene DMNT; MT, monoterpenes; Other; SQT, sesquiterpenes. Larval density represents the number of larvae added per branch. The bars show mean, n = 4 (Abisko) and n = 5 (Tromsø), and the error bars show ± standard error of the total BVOC emissions. Within an experiment, bars labeled with different letters are significantly different from each other (Tukey’s post hoc test; Supplementary Table S4).

FIGURE 2

Effects of increased larval density (0, 5, 15, 30 or 50 larvae per branch) on BVOC blends in experiments Abisko-1, Abisko-2, and Tromsø. (A) Changes in relative proportions of different BVOC groups: Benzenoids; GLV, green leaf volatiles; HT, homoterpene DMNT; MT, monoterpenes; Other; SQT, sesquiterpenes. (B) Principal component analysis biplots on BVOC group composition. The score values for the measured branches are shown colored for different larval densities and sized based on the leaf area eaten. Loadings for each BVOC group are presented with arrows. The variance explained by each principal component (PC) is shown in parentheses.

FIGURE 3

The relationship between the leaf area eaten and BVOC emission rates for Abisko-1, Abisko-2, and Tromsø experiments. Scatter plots and linear fits are shown for total BVOCs (A–C) and the compound groups: SQT, sesquiterpenes (D–F); HT, homoterpene DMNT (G–I); MT, monoterpenes (J–L); GLV, green leaf volatiles (M–O). Emission rates are expressed as ng cm^–2 leaf area h^–1 and leaf area eaten is expressed as percentage. For each of the experiments, n = 20.