Atmospheric transformation of plant volatiles disrupts host plant finding

Abstract

Plant-emitted volatile organic compounds (VOCs) play important roles in plant-insect interactions. Atmospheric pollutants such as ozone (O3) can react with VOCs and affect the dynamics and fidelity of these interactions. However, the effects of atmospheric degradation of plant VOCs on plant-insect interactions remains understudied. We used a system comprising Brassica oleracea subsp. capitata (cabbage) and the specialist herbivore Plutella xylostella to test whether O3-triggered VOC degradation disturbs larval host orientation, and to investigate the underlying mechanisms. Larvae oriented towards both constitutive and larva-induced cabbage VOC blends, the latter being the more attractive. Such behaviour was, however, dramatically reduced in O3-polluted environments. Mechanistically, O3 rapidly degraded VOCs with the magnitude of degradation increasing with O3 levels. Furthermore, we used Teflon filters to collect VOCs and their reaction products, which were used as odour sources in behavioural tests. Larvae avoided filters exposed to O3-transformed VOCs and spent less time searching on them compared to filters exposed to original VOCs, which suggests that some degradation products may have repellent properties. Our study clearly demonstrates that oxidizing pollutants in the atmosphere can interfere with insect host location, and highlights the need to address their broader impacts when evaluating the ecological significance of VOC-mediated interactions.

Figure 1. Experimental setup for the demonstration of chemical reaction between plant VOCs and O₃, and dual-choice olfactory bioassay.

SC: odour source chamber; RC: reaction chamber; OG: ozone generator; OS: ozone scrubber (KI coated copper tube); Y: Y-tube olfactometer; L: P. xylostella larva.

Figure 2

Orientation (percentage ± se) of P. xylostella larvae to VOCs from (a) healthy cabbage plants vs clean air, (b) larva-infested vs healthy cabbage plants, and (c) a synthetic bouquet resembling the larva-induced blend vs a hexane control in Y-tube olfactometers. The number of responding larvae/the total number of larvae tested is 110/128 (a), 123/134 (b), and 65/132 (c). Asterisks indicate a preference which is significantly different from an equal distribution within a choice test: *P ≤ 0.05; **P ≤ 0.001.

Figure 3

Degradation of P. xylostella-induced cabbage VOCs by O₃: (a) relative reduction (mean ± se) in VOC concentrations and (b) changes in relative blend compositions after exposure to different O₃ concentrations. Asterisks indicate a significant reduction at each O₃ level (na: no statistical analysis because of zero-inflated data distribution). (E)-DMNT: (E)-4,8-dimethyl-1,3,7-nonatriene. *P ≤ 0.05.

Figure 4. Attraction (percentage ± se) of P. xylostella larvae to P. xylostella-induced cabbage VOC blends mixed with O₃ (50 or 100 ppb) or clean air in Y-tube olfactometers.

The number of responding larvae/the total number of tested larvae in (a–c) is 85/98, 109/133 and 104/117, respectively. Asterisks indicate a preference which is significantly different from an equal distribution within a choice test: **P ≤ 0.001.

Figure 5. Orientation of P. xylostella larvae in Petri dishes towards Teflon filters that had previously been exposed to cabbage VOCs (iVOC: P. xylostella-induced VOC; cVOC: constitutive VOC) mixed with O₃ (grey bars; 50 or 100 ppb) or clean air (white bars).

(a–c) Percentage (mean ± se) of larvae choosing filters. The number of responding larvae/the total number of tested larvae in (a–c) is 80/120, 73/121 and 77/121, respectively. Asterisks indicate a significant preference based on two-sided binomial tests. (d–f) Mean time spent on filters (±se). Data in (d–f) correspond to observations in (a–c), respectively; the number of larvae observed (grey|white) is 22|26, 12|32 and 23|31, respectively. Asterisks denote a significant difference according to Mann-Whitney U tests. *P ≤ 0.05; **P ≤ 0.001.

Figure 6. Orientation of P. xylostella larvae in Petri dishes towards Teflon filters that had been exposed to cabbage VOCs (while bars; iVOC: P. xylostella-induced VOC; cVOC: constitutive VOC) or clean air (grey bars).

(a,b) Percentage (mean ± se) of larvae choosing filters. The number of responding larvae/the total number of tested larvae in (a,b) is 64/160 and 65/112, respectively. Asterisks indicate a significant preference based on two-sided binomial tests: **P ≤ 0.001. (c,d) Mean time spent on filters (±se). Data in (c; grey|white = 12|16) and (d; grey|white = 17|47) were derived from a subset of observations in (a,b), respectively.