Ozone induced emissions of biogenic VOC from tobacco: relationships between ozone uptake and emission of LOX products

Plant, Cell and Environment

(2005)

28

, 1334–1343

© 2005 Blackwell Publishing Ltd

1334

Blackwell Science, LtdOxford, UKPCEPlant, Cell and Environment0016-8025Blackwell Publishing Ltd 2005? 2005

28••13341343Original Article

Ozone induced emissions of biogenic VOC from tobaccoJ. Beauchamp

et al.

Correspondence: Jürgen Wildt. Tel.

+

49 (0) 2461 616784; fax:

+

49 (0) 2461 612492; e-mail: [email protected]

Ozone induced emissions of biogenic VOC from tobacco: relationships between ozone uptake and emission of LOX products

JONATHAN BEAUCHAMP

1

, ARMIN WISTHALER

1

, ARMIN HANSEL

1

, EINHARD KLEIST

2

, MARCO MIEBACH

2

, ÜLO NIINEMETS

3

, ULI SCHURR

2

& JÜRGEN WILDT

2

1

Institute of Ion Physics, Leopold-Franzens-University, A-6020 Innsbruck, Austria,

2

Research Centre Juelich GmbH, Institute Phytosphere (ICG-III), D-52425 Juelich, Germany and

3

Department of Plant Physiology, University of Tartu, Riia 23, 51011 Tartu, Estonia

ABSTRACT

Volatile organic compound (VOC) emissions from tobacco(

Nicotiana tabacum

L. var. Bel W3) plants exposed toozone (O

3

) were investigated using proton-transfer-reaction mass-spectrometry (PTR-MS) and gas chromatog-raphy mass-spectrometry (GC-MS) to find a quantitativereference for plants’ responses to O

3

stress. O

3

exposuresto illuminated plants induced post-exposure VOC emissionbursts. The lag time for the onset of volatile C

6

emissionsproduced within the octadecanoid pathway was found to beinversely proportional to O

3

uptake, or more precisely, tothe O

3

flux density into the plants. In cases of short O

3

pulses of identical duration the total amount of these emit-ted C

6

VOC was related to the O

3

flux density into theplants, and not to ozone concentrations or dose–responserelationships such as AOT 40 values. Approximately one C

6

product was emitted per five O

3

molecules taken up by theplant. A threshold flux density of O

3

inducing emissions ofC

6

products was found to be (1.6

±±±±

0.7)

¥¥¥¥

10----

8

mol m----

2

s----

1

.

Key-words

:

Nicotiana tabacum

; biogenic volatile organiccompounds (VOC); LOX products; octadecanoid pathway;ozone; proton-transfer-reaction mass-spectrometry (PTR-MS).

INTRODUCTION

Emissions of volatile organic compounds (VOC) fromplants are interesting from many aspects. Plant-generatedVOC have a strong impact on chemical processes in theatmosphere, such as ozone (O

3

) production and formationof aerosols (e.g. Kavouras, Mihalopoulos & Stephanou1998; Guenther

et al

. 2000). They also play a key role inplant–environment interactions. Floral VOC can act asattractants for species-specific pollinators (e.g. Pichersky &

Gershenzon 2002), and other VOC can act as constitutiveand induced defences (e.g. Price

et al

. 1980).VOC emissions commonly depend on light and tempera-

ture (e.g. Guenther

et al

. 2000; Niinemets, Loreto & Reich-stein 2004 for review). However, stress conditions such aswounding (Loreto & Sharkey 1993), water (Sharkey &Loreto 1993), temperature stress (Staudt & Bertin 1998),or ozone (Loreto & Velikova 2001; Loreto

et al

. 2004) resultin altered VOC emissions that may increase by up toseveral orders of magnitudes compared with those of non-stressed plants. This strong stress sensitivity of VOC emis-sions suggests that plant-generated VOC emissions provideimportant information on plant responses to environmentalstresses.

The aim of the experiments described herein was to testthe hypothesis that VOC emissions from plants may beused to find quantitative information on the impact of stresson the plants. Therefore, VOC emissions from O

3

exposedtobacco (

Nicotiana tabacum

L. var. Bel W3) weremeasured.

Ozone was used as elicitor for VOC emissions becausethe tobacco plant’s response to O

3

is similar to a hypersen-sitive response and O

3

exposure has been established as amodel system for studying the downstream components ofsignalling pathways in cell death regulation (Kangasjärvi

et al

. 1994; Rao & Davis 2001; Langebartels

et al

. 2002a,b;Pellinen

et al

. 2002; Overmyer, Brosche & Kangasjärvi2003). Furthermore, using O

3

exposure as a stress agent isadvantageous due to a number of reasons: O

3

exposure canbe conducted under well-defined conditions; the experi-ments may be repeated using the same amount of O

3

andthus applying the same amount of stress to the plants; theamount of stress caused by O

3

exposure can be varied overa wide range, allowing investigation of the plants’ responsesin relation to the degree of stress.

Tobacco plants (

Nicotiana tabacum

var. Bel W3) wereused for these experiments due to their particular sensitiv-ity to O

3

and since their internal responses to O

3

havealready been studied extensively (e.g. Schraudner

et al

.1998; Rao, Koch & Davis 2000; Rosetti & Bonatti Mede-ghini 2001).

Ozone induced emissions of biogenic VOC from tobacco

1335

© 2005 Blackwell Publishing Ltd,

Plant, Cell and Environment,

28,

1334–1343

MATERIALS AND METHODS

Plant material

Seeds of tobacco plants (

Nicotiana tabacum

var. Bel W3)were germinated in a commercial mixture of soil, peat andcompost (Pikiererde, Plantaflor, Vechta, Germany). Plantswere grown at a temperature of 28

∞

C, with 60–85% relativehumidity, 14 h-photoperiod, and photosynthetic active radi-ation (PAR) of 300

m

mol m

-

2

s

-

1

. After germination theplants were cultivated individually in standard substrate(Einheitserde, type ED 73) under the same conditions.Nine- to 10-week-old plants were used in all experiments.

Plant chambers

The continuously stirred tank reactors (CSTR) used for theexperiments have been described in detail by Wildt

et al

.(1997). In short, the glass chambers were mounted in tem-perature-controlled housings and were equipped with sev-eral connections to introduce temperature and lightintensity sensors, and to connect the tubes for gas-phaseanalysis and air supply. In order to minimize wall losses alltubes were either Teflon (PFA or PTFE) or glass. HQI400 W/D lamps (Osram, Munich, Germany) were used forillumination. Filters (Type IR3; Prinz Optics GmbH, Strom-berg, Germany) reflecting wavelengths between 750 and1050 nm were used as heat shields to avoid overheating ofthe plants by infrared radiation.

Ambient air purified by a palladium catalyst (450

∞

C) andan adsorptive drying device (KEA 70; Zander Aufberei-tungstechnik GmbH & Co. KG, Essen, Germany) waspumped through the chambers. The air flow through thechambers was kept constant by mass flow controllers.Teflon fans were used to destroy the boundary layer resis-tance at the surfaces of the plants and to homogeneouslymix the air within the chambers. Mixing ratios of O

3

, carbondioxide (CO

2

), and water vapour (H

2

O) were measured asdescribed in detail by Schuh

et al

. (1997). Differences inmixing ratios between chamber inlet and outlet were mea-sured by infrared absorption for both H

2

O and CO

2

(Binos100 4P, Emerson Process Management GmbH & Co. OHG,Hasselroth, Germany). Absolute H

2

O concentrations weredetermined with the aid of dew point mirrors (MTS-MK1;Walz, Effeltrich, Germany) and absolute CO

2

concentra-tions were measured by infrared absorption (Uras 3G,Hartmann & Braun AG, Frankfurt, Germany). O

3

concen-

trations were measured by UV-absorption (Model 49;Thermo Electron GmbH, Dreieich, Germany).

In most cases the experiments were conducted with sin-gle plants. For these experiments a plant chamber with avolume of 164 L was used. The air flow was set to 25 Lmin

-

1

, resulting in a residence time of air in the chamber ofabout 6.5 min. All plants were investigated at a leaf tem-perature of 28

∞

C and PAR during illumination was800

m

mol m

-

2

s

-

1

. For some control experiments a chamberwith a volume of 1100 L was used. In these experimentsfour plants were introduced in the chamber and investi-gated together at a temperature of 28

∞

C and a PAR of480

m

mol m

-

2

s

-

1

.

Exposures

O

3

was produced in a glass cell outside the CSTR by pho-tolysis of oxygen with a UV light source (Pen-Ray, UVP,Inc., Upland, CA, USA

l

=

189 nm). The O

3

-containingflow from the O

3

generator was mixed with the air enteringthe chamber. Plants were exposed to O

3

in short pulses. Thedurations of individual exposures varied between 1.0 and8.8 h and the mixing ratios of O

3

used for the individualexposure experiments ranged from 80 to 1700 ppbv. In 21experiments the plants were kept under illumination bothduring and after O

3

exposure. In other experiments O

3

exposure was conducted during illumination and thereafterthe lamps were turned off. Two further experiments wereconducted with continuous illumination until about 4 hafter O

3

exposure, and then the lamps were switched off. Infour experiments (two experiments with single plants andtwo experiments each with four plants together) the plantswere exposed during darkness. In these experiments lampswere turned on again directly after ending O

3

exposures inorder to exclude possible problems from light-dependentVOC synthesis. Table 1 lists the exposure experiments.

VOC measurements

VOC measurements were conducted simultaneously with aproton-transfer-reaction mass-spectrometer (PTR-MS) todetermine the emission kinetics, and an online gas chroma-tography mass-spectrometry (GC-MS) system for com-pound identification.

Detailed descriptions of the PTR-MS system can befound in Hansel

et al

. (1995) and Lindinger, Hansel & Jor-

Table 1.

Overview of exposure experiments

PAR-D(

m

mol m

-

2

s

-

1

)PAR-A(

m

mol m

-

2

s

-

1

)

c

(O

3

) range(ppbv)

Exposure duration (h)

Number of experiments/plants

800 800, left under illumination 80–1770 1–7.3 21/21800 Set to 0 directly after exposure, left in darkness 290–560 1.5–2.3 3/3800 800, switched to 0 at 4 h after ending the exposure 120–133 7.6–8.8 2/2

0 Set to 800 directly after exposure and left under illumination thereafter 573–606 1–2 2/20 Set to 480 directly after exposure and left under illumination thereafter 439–550 1 2/8

PAR-D, PAR during exposure; PAR-A, PAR after exposure.

1336

J. Beauchamp

et al

.

© 2005 Blackwell Publishing Ltd,

Plant, Cell and Environment,

28,

1334–1343

dan (1998a, b). The instrument was calibrated using a cali-bration gas mixture prepared by Apel-RiemerEnvironmental Inc. (Denver, CO, USA) containing VOCin the ppmv range per compound, which was dynamicallydiluted to 0.5–10 ppbv in humidified VOC-free air. Theaccuracy of the measurements is estimated to be

±

15%.A detailed description of the GC-MS system used for

VOC analysis has been given by Heiden

et al

. (1999). Cal-ibration of the GC-MS system and an additional calibrationof the PTR-MS instrument were performed using a perme-ation source containing pure chemicals in individual vialsin combination with a dynamic dilution system. Concentra-tions of the compounds released from the calibrationsource were determined from mass loss rates of the indi-vidual compounds and their dilution flows. VOC mixingratios were in the lower ppbv to pptv range. Identificationby GC-MS was based on mass spectra and retention timesof pure chemicals (Sigma-Aldrich Chemie GmbH, Munich,Germany, purity

>

93%).Identification by PTR-MS was based on mass spectra also

obtained from pure chemicals: (E)-2-hexanal was predom-inantly (approximately 85%) found at mass 99 amu (M99);1-hexanol was found at M85; the isomers hexanal and (E)-3-hexenol were found at M83 and M101. (Z)-3-hexenylacetate was predominantly (approximately 80%) found atM83. The total signal from emissions of VOC producedwithin the octadecanoid pathway (LOX products) pre-sented in this study is therefore represented by the sum ofthese individual mass signals (M83

+

M85

+

M99

+

M101).The plants’ responses within the octadecanoid pathway

up to the production of the first volatile LOX product ((Z)-3-hexenal) were investigated. Accordingly, for interpreta-tions herein, the sum of the flux densities of the individualvolatile LOX products was used. By considering only thesum of emissions, all information regarding the transforma-tion steps after the formation of (Z)-3-hexenal were notconsidered. Nevertheless, the kinetics of the sum of emis-sions are not affected by the time needed for these trans-formation steps (Heiden

et al

. 2003).Data obtained with PTR-MS showed the sum of emis-

sions of LOX products from linoleic acid (18 : 2) and lino-lenic acid (18 : 3). Data gathered by the GC-MS system,however, clearly showed that the emissions of C

6

com-pounds produced from 18 : 2 contributed less than 6% tothe emissions of LOX products. Since the sensitivity of thePTR-MS measurements was found to be nearly identicalfor LOX products from 18 : 2 and 18 : 3, respectively, thetemporal shapes of the flux densities (PTR-MS data) maybe taken to be predominantly from the unsaturated LOXproducts synthesized from 18 : 3.

Determination of flux densities

Flux densities of compound X,

F

X

, were determinedaccording to the usual procedure (e.g. Neubert

et al

. 1993)using mixing ratios of the individual compounds measuredat the chamber outlet

c

o

(X) and inlet

c

i

(X), respectively.The flux densities were calculated according to Eqn 1:

(1)

where

F

is the air flow through the chamber (m

3

s

-

1

) and

A

is the total (one-sided) leaf area (m

2

) of the investigatedplant. In SI units, the value of

c

(X) would be given inmol m

-

3

. However, it is common practice to describe air-borne abundance of ozone or other trace gases as mixingratios rather than concentrations, namely in units of partsper billion of volume (ppbv) or equivalently, in nL L

-

1

. Inthis case, the preferred (mol m

-

2

s

-

1

) units for flux densitycan be re-established simply by applying the air flow inunits of mol s

-

1

.The effects of losses to the chamber walls for the deter-

mination of flux densities have previously been shown tobe negligible (Schuh

et al. 1997; Heiden et al. 1999, 2003).In the case of VOC or H2O the concentrations at the cham-ber outlet were higher than those at the chamber inlet andthus the flux densities were positive. For O3 during expo-sure, and for CO2, the concentrations measured at thechamber outlet during periods of illumination were lowerthan those measured at the chamber inlet. In such cases thedetermined flux densities were negative. However, to allowfor comparison, absolute magnitudes for O3 and CO2 fluxdensities were used.

In order to find a suitable alternative reference quantityfor comparison of the data sets, integrated flux densitieswere taken. These represent the total amount of VOCemissions after O3 exposure, or the total amount of O3

taken up. Data were obtained by multiplying flux densi-ties measured during given time intervals with the respec-tive time interval duration. The resulting data weresummed:

(2)

is the amount of compound X taken up or emittedover a certain time by 1 m2 leaf area (units of :mol m-2), Dt is the time interval duration of the measure-ment, and FX is the flux density of X measured duringthis time interval. The time intervals were 10 min for themeasurements of O3, H2O and CO2, and 4 min for VOCconcentrations. In the case of O3 uptake, summationswere performed over the whole exposure durations; incases of VOC emissions summations were performedover the entire time periods during which induced emis-sions were observable. The total amount of emitted LOXproducts, , was used here as a quantity to describethe absolute amount of the plant’s response. Fromhereon, is defined as the absolute response of theplant.

Another common reference for the impacts of O3 onplants is the AOT 40 value (accumulated O3 above aconcentration of 40 ppbv). AOT 40 values for O3 werecalculated by taking the mean hourly O3 concentrationabove 40 ppbv and adding the data for the duration ofO3 exposure. Table 2 provides an overview of the quanti-ties used as a reference to characterize the impacts ofozone.

FX o iFA

X X= ◊ ( ) - ( )( )c c

F D FXtot

X= ◊ tt

FXtot

FXtot

FLOXtot

FLOXtot

Ozone induced emissions of biogenic VOC from tobacco 1337

© 2005 Blackwell Publishing Ltd, Plant, Cell and Environment, 28, 1334–1343

Determination and quantitative definition of lag times

The temporal shapes of LOX emissions showed sigmoidincreases. Thus, these shapes were analysed quantitativelyusing fits to a formula describing such sigmoid increases(Eqn 3).

(3)

F is the emission rate at time t (t = 0 at start of O3 exposure)and Fmax is the maximum VOC emission rate measured. Theparameter k accounts for different curvatures of theincreases. D is defined as the lag time and is described inthe following.

The most well-defined point of a sigmoid curve is theinflection point. Therefore, this point was used as a refer-ence for a quantitative definition of the lag time D. D isthe time between the beginning of O3 exposure and theinflection point of LOX product emissions. Data fitsenabled values of D to be determined from the beginningof O3 exposures until Fmax. The inverse of D (1/D) wasdefined as the induction rate. For clarification, Fig. 1 showsa typical example of an increase of VOC emissions such asobserved during these experiments, as well as a fit to thedata.

FF

=- ◊ -{ }( )

max

1+ exp k D t

RESULTS

Control plants not exposed to ozone emitted methanol andsmall amounts of isoprene and sesquiterpenes (SQT).Emissions of VOC produced within the octadecanoid path-way (LOX products) and methyl salicylate (MeSA) werenot detectable (mixing ratios below 1 pptv). The samebehaviour was found for all tobacco plants before theywere exposed to O3.

Exposures at high O3 concentrations (c(O3) between 127and 1700 ppbv) and during periods of illumination causedemissions of LOX products and MeSA, and also increasedmethanol and SQT emissions. Often, these conditions alsoled to decreases in the rates of transpiration and net pho-tosynthesis, implying reductions in stomatal aperture. SinceO3 is taken up through the stomata (e.g. Neubert et al.1993), these reductions in stomatal aperture also led todecreases of . To demonstrate this behaviour, Fig. 2shows results from measurements conducted with the high-est O3 concentrations, in which this effect was most pro-nounced. In the cases where stomatal closure in response

FO3

Table 2. List of quantities used to find the best reference for the impacts of ozone

Abbreviation Quantity (unit)

c(O3) Ozone concentration, here mixing ratio of O3 (ppbv). 1 ppbv ª 4.1 ¥ 10-8 mol m-3 at a temperature of 25 ∞C andatmospheric pressure

Flux density of ozone (mol m-2 s-1)Total flux density of ozone = sum of ¥ Dt over the duration of exposure with Dt = time intervals of measurements,

in (mol m-2)AOT 40 Accumulated ozone above a threshold of 40 ppbv (1 ppbv = 1 nL L-1)

FO3

FOtot

3FO3

FOtot

3

Figure 1. Definition of lag time, D. According to Eqn 3, D is the time from the start of O3 exposure (time t = 0) until the time where the inflection point occurs (t at F = 1/2 Fmax).

0 1 2 3 4 5

0

2x10–9

4x10–9

6x10–9F S LOX products

data fit Fmax

1/2 Fmax

D

F(S

LOX

pro

duct

s) (

mol

m–2

s–1)

Time after starting ozone exposure (h)

Figure 2. Flux densities normalized to their respective maxima. Water vapour (open circles, left scale), CO2 (black squares, left scale), and O3 (open diamonds, right scale). Absolute data at the respective maxima were H2O, 2.4 ¥ 10-3 mol m-2 s-1; CO2, 5.6 ¥10-6 mol m-2 s-1; and O3, 1.2 ¥ 10-7 mol m-2 s-1. PAR, 800 mmolm-2 s-1; temperature, 28 ∞C, O3 exposure for about 1 h. During exposure the ozone concentration at chamber inlet was constant. As a consequence of stomatal closure, ozone uptake lessened and the mixing ratio of ozone at the chamber outlet increased from 1040 to 1700 ppbv.

–1 0 1 20.0

0.5

1.0

1.5

Nor

mal

ized

H2O

and

CO

2 flu

x de

nsiti

es

Time after starting O3 exposure (h)

0.0

0.4

0.8

Nor

mal

ized

O3

flux

dens

ity

1338 J. Beauchamp et al.

© 2005 Blackwell Publishing Ltd, Plant, Cell and Environment, 28, 1334–1343

to O3 caused temporal variation of values given for relate to the mean O3 flux density during the experiment.

In all cases where O3 exposures induced VOC emissionsthe necrotic spots on the older leaves became visible to thenaked eye. At c(O3) below 120 ppbv and with exposures forperiods of less than 2 h, neither increased VOC emissions,reductions in stomatal conductivity, decreases in the netrates of photosynthesis, nor visible symptoms of injury wereobserved.

The VOC emissions induced by O3 exposure were tran-sient. Emissions did not start directly with O3 exposure butafter variable lag times ranging from 30 min to 16 h,depending on and type of VOC. Methanol emissionsincreased first, followed by those of MeSA and LOX prod-ucts. SQT emissions increased in some cases before and inother cases simultaneously to those of MeSA and LOXproducts. As an example, Fig. 3 shows temporal shapes asmeasured for methanol, MeSA, and the sum of LOXproducts.

In all cases when O3 exposures induced additional emis-sions the pattern of VOC was similar. Methanol emissionswere dominant, followed by those of LOX products. MeSAand SQT were emitted in lower amounts. The focus of thisstudy was on the emissions of LOX products. Details of theother emissions will be described elsewhere.

In particular in this study, the phase of increasing emis-sions of LOX products was focused on. This increase had asigmoid shape, which was initially exponential, crossed aninflection point and thereafter lessened until a maximumwas reached. For all experiments in which significant emis-sions of LOX products were observed, the description ofthe emission kinetics by Eqn 3 was excellent. Thus, the datafor the lag times, D, or the respective induction rates, 1/D,were determinable in a reliable way.

It was found that the higher , the earlier the O3-induced emissions of LOX products started, indicating that

had a significant impact on the timing of the plants’

FO3 FO3

FO3

FO3

FO3

responses. A good relationship was found between 1/D and (see Fig. 4a). Linear regression analysis yielded:

The relationship between 1/D and the respective AOT 40value was worse (Fig. 4b). The correlation coefficientobtained from linear regression of the data shown in Fig. 4bwas R2 = 0.13 and analysis of plots of 1/D versus (notshown) yielded R2 = 0.12.

FO3

10 072 0 052 2 7 0 4 10

0 74

6

D

R

= ±( )[ ]+ ±( ) ◊◊◊

ÈÎÍ

˘˚̇

◊

=

-. . . . ,

. .

hm s

mol h with

12

O

2

3F

F Otot

3

Figure 3. Time course of VOC emissions from O3-exposed tobacco. Solid line: sum of LOX products; open circles: MeSA, multiplied by 30; filled squares: methanol, divided by 2. PAR: 800 mmol m-2 s-1; temperature: 28 ∞C; O3 exposure to 500 ppbv for 1 h as indicated by the top left bar; = 5.2 ¥ 10-8 mol m-2 s-1; s( ) = 6.7 ¥ 10-9 mol m-2 s-1.

0 5 10 15 200

2x10–9

4x10–9

6x10–9

8x10–9

1x10–8

F S LOX productsF Methanol, divided by 2F MeSA, multiplied by 30

F (

mol

m–2

s–1)

Time after starting ozone exposure (h)

FO3

FO3

Figure 4 (a) Induction rates (1/D) for emissions of the sum of LOX products plotted versus for the respective experiments. Open diamonds, data obtained from experiments conducted at permanent illumination; closed squares, data obtained from experiments in which the lamps were turned off after O3 exposure; open circles, data obtained from experiments where the lamps were turned off about 4 h after ending the O3 exposure. Since often varied due to stomatal closure it is represented by the mean measured during the exposure. The bars indicate standard deviations (1 s) for .(b) Induction rates (1/D) for emissions of the sum of LOX products plotted versus AOT 40. The symbols have the same denotation as those shown in Fig. 4a. Errors for AOT 40 values were estimated from the accuracy of the O3 analyser (4%).

0.0 5.0x10–8 1.0x10–7 1.5x10–70.0

0.1

0.2

0.3

0.4

0.5(a)

(b)

permanent illumination light turned off directly after O

3 exposure

light turned off 4 h after O3 exposure

1/D

(h–1

)

F(O3) (mol m–2 s–1)

0 200 400 600 800 1000 1200 1400 16000.0

0.1

0.2

0.3

0.4

0.5 permanent illumination light turned off directly after O

3 exposure

light turned 4 h after O3 exposure

1/D

(h–1

)

AOT 40 (ppbv h)

FO3

FO3

FO3

Ozone induced emissions of biogenic VOC from tobacco 1339

© 2005 Blackwell Publishing Ltd, Plant, Cell and Environment, 28, 1334–1343

It should be noted that Fig. 4a includes data from exper-iments conducted with different light regimes after O3

exposure. The data obtained from these different experi-ments show no significant deviations.

A further possible reference for the impacts of O3 on theplants might have been the O3 concentrations during expo-sures. However, at a fixed stomatal aperture was pro-portional to c(O3) in the air. In the experiments describedso far, plants were exposed to O3 at the same PAR and atthe same temperature. Thus, stomatal aperture was similarfor the investigated plants and it was not possible to distin-guish between or c(O3) being the better reference.

In order to distinguish between or c(O3) being themost suitable reference, tobacco plants were exposed tohigh c(O3) during periods of darkness. During darknesswhen stomata were nearly closed, O3 flux densities into theplants were quite low despite O3 concentrations in the airbeing very high. Neither emissions of LOX products orMeSA, nor increases of methanol or SQT emissions wereobserved, even in experiments in which plants wereexposed to 600 ppbv O3 for periods of 2 h.

The same individuals were exposed to O3 during periodsof illumination 1–3 d after the first exposures. Althoughc(O3) used for these second exposures were less than thoseused during the first the typical VOC emissions wereobserved, indicating that there is no direct relation betweenc(O3) and induced VOC emissions. Figure 5 shows an exam-ple from such an experiment.

Total emissions

So far only the temporal behaviour of the plants’ responseshas been looked at. The total amount of emitted LOXproducts, , is now considered here as an absoluteresponse of the plants. As in the case of the temporal behav-

FO3

FO3

FO3

FLOXtot

iour of the plants’ responses, the best reference to describethe impacts of O3 exposure on was investigated.

Data obtained for were plotted versus , versus, and versus AOT 40 with no good relationships being

found. The relation was much better when restricted onlyto data points from experiments where plants were exposedto O3 for about 1 h only. In these cases, a good relation wasfound between and , and and ,respectively.

Using these data only, linear regression analysis of versus led to

The slope shows that approximately 1 volatile LOX prod-uct is emitted per 5 O3 molecules taken up. Linear regres-sion analysis of versus led to:

The intercept at the x-axis was assumed to be the thresholdflux density of O3 inducing emissions of LOX products{ = (1.6 ± 0.7) ¥ 10-8 mol m-2 s-1}.

In order to check why no good relationship was obtainedwhen considering all data points, data determined fromexposures for more than 7 h and at minimally abovethe threshold were compared with those obtained for 1 honly. Figure 6 shows a plot of versus for thesedata.

As is obvious from Fig. 6, the amount of emitted LOXproducts after exposure durations longer than 1 h was

FLOXtot

FLOXtot FO3

F Otot

3

FLOXtot FO3 FLOX

tot F Otot

3

FLOXtot

F Otot

3

F FLOXtot 2

Otot

2

mol m 3= - ±( ) ◊ ◊[ ]+ ±( ) ◊=

- -1 39 0 53 10 0 21 0 02

0 90

5. . . . ,

. .R

FLOXtot FO3

F FLOXtot 2

O

2

mol m 3= - ±( )◊ ◊[ ]+ ±( )[ ]◊=

- -1 34 0 59 10 836 93

0 87

5. . ,

. .

s

R

F Otot

3

FO3

FLOXtot F O

tot3

Figure 5. Time course of emissions of LOX products (open circles) after exposing a plant to 600 ppbv O3 during darkness (1st ‘exposure’ bar) and to 350 ppbv during illumination (2nd ‘exposure’ bar). Illumination at 800 mmol m-2 s-1 is indicated by the upper bars. Despite the lower O3 concentrations during the second exposure emissions of LOX products as well as necrotic spots on the plants’ leaves were only observable after this second exposure.

–20 0 20 40 60 80 1000

1x10–9

2x10–9 F S LOX products

FS

LOX

-pro

duct

s (m

ol m

–2s–1

)

Time relative to starting the first ozone exposure (h)

0

200

400

600

800

2nd exposure

1st exposure

PA

R (mm

ol m

–2s–1

); c

(O3)

(ppb

v)

Figure 6. Total amount of emitted LOX products versus total amount of O3 taken up . Open squares: data points determined from experiments with 1 h O3 exposure. The line shows the result from linear regression for these data. Filled circles, data from experiments with long exposure durations (7.3–8.8 h) and low ozone flux densities (1.2 ¥ 10-9 to 2.47 ¥ 10-9 mol m-2 s-1). Error bars for are estimated from the error of the O3 analyser (accuracy 4%) and errors in the measurements of air flows (accuracy 1%). From calibrations of both instruments measuring VOC emissions it is known that the absolute values of the emission rates include an error of about 15%.

0 2x10–4 4x10–4 6x10–4 8x10–4 1x10–3

0

5x10–5

1x10–4

O3 exposure for 1 hO3 exposure longer than 7 h

Fto

tLO

X (

mol

m–2

)

FtotO3

(mol m–2)

FLOXtot

FOtot

3

FOtot

3

1340 J. Beauchamp et al.

© 2005 Blackwell Publishing Ltd, Plant, Cell and Environment, 28, 1334–1343

mostly lower than might have been expected from a linearrelationship between and .

DISCUSSION

Timing of plant responses

Following O3 exposures to high concentrations and duringperiods of illumination typical LOX products were emittedfrom tobacco var. Bel W3. The processes leading to suchresponses are well known; they are reminiscent of a hyper-sensitive response and commonly referred to as acuteresponses. One of these acute responses is the induction ofactivity within the octadecanoid pathway and the subse-quent emission of volatile LOX products as outlined inFig. 7.

In this study, relationships between the first step and thelast step shown in Fig. 7 were found. One of these relation-

FLOXtot F O

tot3

ships was between the induction rates (1/D) for emissionsof LOX products and the O3 flux densities into the plants( ). Relationships between induction rates and AOT 40values, or between induction rates and the amount of O3

taken up, respectively, were not good. In particular, theresults from experiments where plants were exposed tohigh c(O3) during periods of darkness imply that the acuteresponses of tobacco plants are related to . From theresults of these experiments it is concluded that the timingsof the plants’ acute responses are determined neither byc(O3) in the air, nor by dose–response relationships such asAOT 40 values. These data suggest that with respect to thetiming of the stress response the amount of O3 entering theplants through the stomata per unit time is a better refer-ence.

The use of the flux density of O3 as the most suitablereference to describe O3 impacts on plants has already beensuggested by Chameides (1989). Chameides calculated the

FO3

FO3

Figure 7. Schematic of the processes linking O3 uptake and emissions of LOX products. O3 uptake results in the formation of reactive oxygen species (ROS). ROS accumulation acts as a signal for the formation of free fatty acids. These free fatty acids are oxidized by lipoxygenases. The destruction of the hydroperoxides formed by lipoxygenases leads to the formation of volatile LOX products. The schematics of the octadecanoid pathway are drawn according to Croft, Jüttner & Slusarenko (1993), Hatanaka (1993) and Gardner (1995).

O

OH

O

OH

OOH

O

OH

O

O

OH

OH

linolenic acid 18:3

13-(S)-hydroperoxylinolenic acid

lipoxygenaseso2

(Z)-3-hexenal

(E)-3-hexenal

(E)-2-hexenal

(Z)-3-hexenol

(E)-3-hexenol

(E)-2-hexenol

ADH

ADH

ADH

hydroperoxide lyases

iso

mer

izat

ion

reactive oxygen species ( )ROS

ozone

C -VOC6

Ozone induced emissions of biogenic VOC from tobacco 1341

© 2005 Blackwell Publishing Ltd, Plant, Cell and Environment, 28, 1334–1343

destruction of O3 in the apoplastic water through reactionswith an antioxidant. For such a reaction O3 concentrationsin the plant are important and not O3 concentrations in theair. According to this approach, O3 uptake should impactthe plants’ health if exceeds the amount of antioxidantsdelivered to the apoplast. If is lower than the amountof antioxidants transported to the reaction site, O3 will reactwith antioxidants already before reaching a cell and indu-cing acute responses in the plants. Data presented hereinare consistent with this model.

Measurements in this investigation were made with Nic-otiana tabacum var. Bel W3. Ozone enters Nicotianatabacum var. Bel W3 nearly exclusively through the sto-mata and losses of O3 caused by diffusion through the cutic-ula or reactions on the cuticula do not significantlycontribute to (Neubert et al. 1993). In cases of plantspecies where significant losses of O3 on the cuticula occuror where internal resistances against O3 uptake exist may not be a good reference to the plants’ responses. How-ever, especially with respect to the timing of the plants’responses after O3 exposures it is obvious that neither O3

concentrations nor dose–response relationships can beused as a reference.

With the experimental conditions used in this study, theinduction rate varied by more than a factor of 4, showingclearly that the plants’ acute responses do not at all appearwith a fixed internal time course. With respect to such a highdynamics of the plants’ timing there are still no quantitativedata published in literature.

Furthermore, using as the reference for the timingof these responses, no significant deviations between thedata points obtained from experiments conducted with dif-ferent light regimes after O3 exposure were found. Thisimplies that O3 exposure triggers a sequence of processeswhich, with respect to induction rates, is not significantlyinfluenced by light intensity. Once the sequence of plantinternal processes is triggered by O3, the time needed fromthe trigger to the induction of VOC emissions is mainlydependent on . With respect to the timing of the plants’responses, may therefore be seen as a reference for anamount of stress.

Total emissions

Total emissions of LOX products, , were regarded asan absolute amount of the plants’ responses. A good corre-lation was found between and when restrictedto data with 1 h O3 exposure duration. However, especiallyin cases where plants were exposed for longer durationsand with flux densities only minimally above the threshold,the absolute responses of the plants were lower than mightbe expected from a relationship between and (see Fig. 6, filled circles).

A response that is lower than expected from such a linearrelationship can be explained using information given inthe literature. Pasqualini et al. (2001, 2002) have shown thatthe O3-sensitive Bel W3 tobacco cultivar apparentlyresponds to oxidative stress by increasing ascorbic acid

FO3

FO3

FO3

FO3

FO3

FO3

FO3

FLOXtot

FLOXtot F O

tot3

FLOXtot F O

tot3

transport into the apoplast. If, as a response to the oxidativestress, endogenous levels of antioxidants are increased atthe reaction sites, the impact of O3 uptake will be lowered.Only in cases where such abating feedback processes arenegligible or inhibited may good relations betweenthe amount of the elicitor and the plants’ responses beobservable.

Such a good relationship was found here for the experi-ments with the short durations of exposure. Obviously, theinfluences of abating feedback processes on the impact ofstress were negligible. Either the duration of O3 exposurewas too short to allow for significant reactions in the plants,or the reaction due to the identical exposure times was verysimilar. In any case, for short and identical exposure dura-tions, the amount of O3 taken up by the plant may be usedalso as a quantitative reference for the absolute responsesto the stress caused by O3.

Taken together, short O3 exposures are a quantitativeelicitor for stress. The O3 quantities such as flux densities orthe amount of O3 taken up can be used as an externalreference for the plants’ reactions with respect to stress-induced emissions of LOX products. Both the temporalbehaviour and the absolute amounts of emitted compoundsare related to these quantities.

All results described here are understandable on thebasis of the effects described for O3 impacts on plant-internal processes. However, in the present study no plant-internal processes were measured; instead VOC emissionswere used as a reference for the plants’ responses, leadingto quantitative relationships. These relationships betweenO3 uptake and emissions of LOX products must be linkedthrough the plant-internal processes, as outlined in Fig. 7.

CONCLUSIONS

There is ample evidence that O3 exposure is a good tool forinitiating stress to plants, such as programmed cell death.Statements regarding the impacts of O3 exposure to plants’acute responses have been generally quite qualitative. Thisinvestigation, however, shows quantitative relationshipsbetween O3 uptake into tobacco plants and their responses.

As a result of this study it follows that AOT 40 values orjust O3 concentrations are not good references for O3 stressimpacts on plants. The data shown here clearly indicate thatboth references are flawed. For example, care must betaken if O3 sensitivities of different plant species are to becompared. In such cases, certainly must be used asreference. Care must also be taken if interference of O3

with other environmental variables are to be analysed. Allenvironmental processes leading to changes in stomatalconductance will impact . Using AOT values or O3 con-centrations as a reference for the plants’ responses will leadto misinterpretations.

The data presented here indicate that these responses toO3 stress are not at all static: The plants responses are highlydynamic and linearly related to the external quantity .Most probably, the plants’ responses to other stresses arealso highly dynamic. Timing of the responses as well as the

FO3

FO3

FO3

1342 J. Beauchamp et al.

© 2005 Blackwell Publishing Ltd, Plant, Cell and Environment, 28, 1334–1343

absoluteness of the responses may depend on the amountof stress. However, apart from ozone flux density there isstill no external and quantitative reference for stressimpacts.

Results herein furthermore suggest close relationshipsbetween the plant-internal processes within the octade-canoid pathway and the emissions of VOC produced in thispathway. Assuming these relationships to be close in timeas well as magnitude, emissions of LOX products seem tobe an excellent tool to study also the kinetics of plant-internal processes.

ACKNOWLEDGMENTS

This work was financially supported by the German Minis-try for Education, Science, Research and Technology withinthe project AFO 2000, subproject ECHO (Emission andchemical transformation of biogenic volatile organic com-pounds: Investigations in and above a mixed forest stand);Grant number: 07ATF47. A.W. thanks the Verein zurFörderung der wissenschaftlichen Ausbildung undTätigkeit von Südtirolern an der Landesuniversität Inns-bruck for postdoctoral support.

REFERENCES

Chameides W.L. (1989) The chemistry of ozone deposition toplant leaves. Environmental Science and Technology 23, 595–600.

Croft K.P.C., Jüttner F. & Slusarenko A.J. (1993) Volatile prod-ucts of the lipoxygenase pathway evolved from Phaseolus vul-garis (L.) leaves inoculated with Pseudomonas syringae pv.phaseolicola. Plant Physiology 101, 13–24.

Gardner H.W. (1995) Biological roles and biochemistry of thelipoxygenase pathway. Horticultural Science 30, 197–205.

Guenther A., Geron C., Pierce T., Lamb B., Harley P. & Fall R.(2000) Natural emissions of non-methane volatile organic com-pounds, carbon monoxide, and oxides of nitrogen from NorthAmerica. Atmospheric Environment 34, 2205–2230.

Hansel A., Jordan A., Holzinger R., Prazeller P., Vogel W. &Lindinger W. (1995) Proton transfer reaction mass spectrome-try: on-line trace gas analysis at the ppb level. InternationalJournal of Mass Spectrometry and Ion Processes 149/150, 609–619.

Hatanaka A. (1993) The biogeneration of green odour by greenleaves. Phytochemistry 34, 1201–1218.

Heiden A.C., Hoffmann T., Kahl J., et al. (1999) Emission ofvolatile organic compounds from ozone-exposed plants. Ecolog-ical Applications 94, 1160–1167.

Heiden A.C., Kobel K., Langebartels C., Schuh-Thomas G. &Wildt J. (2003) Emissions of oxygenated volatile organic com-pounds from plants, part I: Emissions from lipoxygenase activ-ity. Journal of Atmospheric Chemistry 45, 143–172.

Kangasjärvi J., Talvinen J., Utriainen M. & Karjalainen R. (1994)Plant defence systems induced by ozone. Plant, Cell and Envi-ronment 17, 783–794.

Kavouras I.G., Mihalopoulos N. & Stephanou E.G. (1998) Forma-tion of atmospheric particles from organic acids produced byforests. Nature 395, 683–686.

Langebartels C., Schraudner M., Heller W., Ernst D. & Sander-mann H. (2002b) Oxidative stress and defense reactions inplants exposed to air pollutants and UV-B-radiation. In Oxida-

tive Stress in Plants (eds D. Inzé & M. Van Montagu), pp. 105–135. Taylor & Francis, London, UK.

Langebartels C., Wohlgemuth H., Kschieschan S., Grün S. &Sandermann H. (2002a) Oxidative burst and cell death inozone exposed plants. Plant Physiology and Biochemistry 40,567–575.

Lindinger W.A., Hansel A. & Jordan A. (1998a) Proton-transfer-reaction mass spectrometry (PTR-MS): on-line monitoring ofvolatile organic compounds at pptv levels. Chemical SocietyReviews 27, 347–354.

Lindinger W.A., Hansel A. & Jordan A. (1998b) On-line monitor-ing of volatile organic compounds at pptv levels by means ofProton-Transfer-Reaction Mass Spectrometry (PTR-MS): med-ical applications, food control and environmental research.International Journal of Mass Spectrometry and Ion Processes173, 191–241.

Loreto F., Pinelli P., Manes F. & Kollist H. (2004) Impact of ozoneon monoterpene emission and evidence for an isoprene likeantioxidant action of monoterpenes emitted by Quercus ilexleaves. Tree Physiology 24, 361–367.

Loreto F. & Sharkey T.D. (1993) Isoprene emission by plants isaffected by transmissible wound signals. Plant, Cell and Envi-ronment 16, 563–570.

Loreto F. & Velikova V. (2001) Isoprene produced by leaves pro-tects the photosynthetic apparatus against ozone damage,quenches ozone products and reduces lipid peroxidation of cel-lular membranes. Plant Physiology 127, 1781–1787.

Neubert A., Kley D., Wildt J., Segschneider H.J. & Förstel H.(1993) Uptake of NO, NO2 and O3 by sunflower (Helianthusannuus L.) and tobacco plants (Nicotiana tabacum L.): depen-dence on stomatal conductivity. Atmospheric Environment 27A,2137–2145.

Niinemets Ü., Loreto F. & Reichstein M. (2004) Physiological andphysico-chemical controls on foliar volatile organic compoundemissions. Trends in Plant Science 9, 180–186.

Overmyer K., Brosche M. & Kangasjärvi J. (2003) Reactive oxy-gen species and hormonal control of cell death. Trends in PlantScience 8, 335–342.

Pasqualini S., Batini P., Ederli L., Porceddu A., Piccioni F., DeMarchis F. & Antonielli M. (2001) Effects of short-term ozonefumigation on tobacco plants: response of the scavenging systemand expression of the glutathione reductase. Plant, Cell andEnvironment 24, 245–252.

Pasqualini S., Della Torre G., Ferranti F., Ederli L., Piccioni C.,Reale L. & Antonielli M. (2002) Salicylic acid modulates ozone-induced hypersensitive cell death in tobacco plants. PhysiologiaPlantarum 115, 204–212.

Pellinen R.I., Korhonen M.-S., Tauriainen A.A., Tapio Palva E. &Kangasjärvi J. (2002) Hydrogen peroxide activates cell deathand defense gene expression in Birch. Plant Physiology 130,549–560.

Pichersky E. & Gershenzon J. (2002) The formation and functionof plant volatiles: perfumes for pollinator attraction and defense.Current Opinion in Plant Biology 5, 237–243.

Price P.W., Bouton C.E., Gross P., McPheron B.A., ThompsonJ.N. & Weiss A.E. (1980) Interaction among three trophic levels:influence of plants on interactions between insect herbivoresand natural enemies. Annual Review of Ecology, Evolution, andSystematics 11, 41–65.

Rao M.V. & Davis K.R. (2001) The physiology of ozone-inducedcell death. Planta 213, 682–690.

Rao M.V., Koch J.R. & Davis K.R. (2000) Ozone: a tool forprobing programmed cell death in plants. Plant Molecular Biol-ogy 44, 345–358.

Rosetti S. & Bonatti Medeghini P. (2001) In situ histochemicalmonitoring of ozone- and TMV-induced reactive oxygen species

Ozone induced emissions of biogenic VOC from tobacco 1343

© 2005 Blackwell Publishing Ltd, Plant, Cell and Environment, 28, 1334–1343

in tobacco leaves. Plant Physiology and Biochemistry 39, 433–442.

Schraudner M., Moeder W., Wiese C., Van Camp W., Inzé D.,Langebartels C. & Sandermann H. Jr. (1998) Ozone-inducedoxidative burst in the ozone biomonitor plant, tobacco Bel W3.Plant Journal 16, 235–245.

Schuh G., Heiden A.C., Hoffmann T., Kahl J., Rudolph J. & WildtJ. (1997) Emission of volatile organic compounds from sun-flower and beech: dependence on temperature and light inten-sity. Journal of Atmospheric Chemistry 27, 291–318.

Sharkey T.D. & Loreto F. (1993) Water stress, temperature, and

light effects on the capacity for isoprene emission and photosyn-thesis of kudzu leaves. Oecologia 95, 328–333.

Staudt M. & Bertin N. (1998) Light and temperature dependenceof the emission of cyclic and acyclic monoterpenes from holmoak (Quercus ilex L.) leaves. Plant, Cell and Environment 21,385–395.

Wildt J., Kley D., Rockel A., Rockel P. & Segschneider H.J. (1997)Emission of NO from several higher plant species. Journal ofGeophysical Research 102, 5919–5927.

Received 24 March 2005; accepted for publication 12 April 2005