Light and temperature dependence of the emission of cyclic and acyclic monoterpenes from holm oak (Quercus ilex L.) leaves


N. Bertin INRA Unité de Bioclimatologie, Site Agroparc, F-84914 Avignon Cedex 9, France. Fax: (33) 04 90 89 98 10; e-mail:


In a laboratory study, we investigated the monoterpene emissions from Quercus ilex, an evergreen sclerophyllous Mediterranean oak species whose emissions are light dependent. We examined the light and temperature responses of individual monoterpenes emitted from leaves under various conditions, the effect of heat stress on emissions, and the emission-onset during leaf development. Emission rate increased 10-fold during leaf growth, with slight changes in the composition. At 30 °C and saturating light, the monoterpene emission rate from mature leaves averaged 4·1 nmol m–2 s–1, of which α-pinene, sabinene and β-pinene accounted for 85%. The light dependence of emission was similar for all monoterpenes: it resembled the light saturation curve of CO2 assimilation, although monoterpene emission continued in the dark. Temperature dependence differed among emitted compounds: most of them exhibited an exponential increase up to 35 °C, a maximum at 42 °C, and a slight decline at higher temperatures. However, the two acyclic isomers cis-β-ocimene and trans-β-ocimene were hardly detected below 35 °C, but their emission rates increased above this temperature as the emission rates of other compounds fell, so that total emission of monoterpenes exponentially increased from 5 to 45 °C. The ratio between ocimene isomers and other compounds increased with both absolute temperature and time of heat exposure. The light dependence of emission was insensitive to the temperature at which it was measured, and vice versa the temperature dependence was insensitive to the light regime. The results demonstrated that none of the models currently applied to simulate isoprene or monoterpene emissions correctly predicts the short-term effects of light and temperature on Q. ilex emissions. The percentage of fixed carbon lost immediately as monoterpenes ranged between 0·1 and 6·0% depending on temperature, but rose up to 20% when leaves were continuously exposed to temperatures between 40 and 45 °C.


Monoterpenes and isoprene are isoprenoid compounds synthesized in plants from a common precursor (i.e. isopentenyl-diphosphate) and emitted into the atmosphere by terrestrial vegetation. These natural emissions constitute a considerable source of reactive carbon and, therefore, they may have a crucial impact on the formation and lifetime of air pollutants and greenhouse gases, such as ozone and carbon monoxide (Brasseur & Chatfield 1991; Fehsenfeld et al. 1992). A lot of multidisciplinary investigations have been carried out to attain realistic predictions of annual emissions from vegetation at national or global scale, but the inaccuracy of so-called emission inventories is still high (Hewitt & Street 1992; Lamb et al. 1993). This partly results from an insufficient understanding of the short- and long-term influences of environmental factors on the amount and composition of emissions. In the current inventories, the influence of environmental factors is modelled by a short-term positive response to temperature of both isoprene and monoterpene emissions, and by a short-term positive response to light of isoprene emission only (e.g. Guenther et al. 1995). This relies on many experimental observations and hypotheses concerning the control of isoprenoid emissions by plants (Sanadze 1964; Rasmussen & Jones 1973; Tyson, Dement & Mooney 1974; Tingey et al. 1980; Tingey, Evans & Gumpertz 1981). Actually, important storage pools of isoprene have never been found in isoprene emitting plants and emission follows isoprene biosynthesis. The light dependence of emission is ascribed to the need of photosynthetic products for isoprene biosynthesis (Monson, Guenther & Fall 1991; Sharkey, Loreto & Delwiche 1991; Loreto & Sharkey 1993), whereas the temperature dependence of emission would result from the temperature dependence of the isoprene synthase, a chloroplastic enzyme that ultimately produces isoprene from dimethylallyl-diphosphate (Monson et al. 1992; Silver & Fall 1995; Schnitzler et al. 1996; Wildermuth & Fall 1996). Monoterpene synthesis is also light dependent, but contrary to isoprene, large storage pools in plant organs, such as glandular trichomes or resin ducts, constitute a permanent emission source that act as a buffer against short-term climatic changes that affect the synthesis rate. Therefore, monoterpene emission is assumed to be principally a volatilization from storage organs and, thus, only the effects of temperature on the gas vapour pressure in plant tissue and on the resistance along the emission pathway are considered (Tingey, Turner & Weber 1991; Monson et al. 1995).

This concept about the difference between the control of monoterpene and isoprene emissions has been questioned by some observations of a light effect on monoterpene emission from coniferous trees (Yokouchi & Ambe 1984; Schürmann et al. 1993; Staudt et al. 1997), and by recent findings on Quercus ilex. This oak species emits large amounts of monoterpenes, despite the absence of a storage pool, and its emission rate was found to be under the short-term control of light (Staudt et al. 1993; Seufert et al. 1995). Thus, for this species, monoterpene emission would follow their synthesis immediately as for isoprene (Staudt & Seufert 1995; Loreto et al. 1996a). This result was further confirmed by field observations (Kesselmeier et al. 1996; Bertin et al. 1997; Street et al. 1997) and was definitely proved by labelling experiments from Loreto et al. (1996b,c). Quercus ilex is an evergreen sclerophyllous oak of great abundance in the thermophilous vegetation of the Mediterranean basin. Originally, the area of Q. ilex forests was restricted to particular edaphic niches and to some semiarid regions, but since neolithicum it has been largely expanded by human activity, replacing forests dominated by deciduous oak species (Reille 1992). Because of its abundance and its strong emissions, Q. ilex is of great interest for emission inventories in the Mediterranean area (Pio, Nunes & Brito 1993).

As monoterpene emissions from Q. ilex are light dependent, the model of Guenther et al. (1993) describing the short-term responses of isoprene emission to light and temperature has been applied to simulate the variations of monoterpene emissions from Q. ilex in the field (Bertin et al. 1997; Kesselmeier et al. 1997). These works demonstrated a fairly good adequacy of the isoprene model, except in summer at high temperatures (Bertin et al. 1997). This inaccuracy may result from two points not yet investigated for monoterpene emissions from Q. ilex: (i) the isoprene model of Guenther et al. (1993) assumes that the influences of light and temperature on emissions are independent of each other, which has never been reported for any monoterpene emissions; (ii) the other uncertainty concerns the response of emissions to temperatures beyond 35 °C, which is particularly relevant since the temperature responses of isoprene and monoterpene emissions are supposed to differ only in the higher range of temperature. Up to ≈ 35 °C, both monoterpene and isoprene emissions exponentially increase with increasing temperature; at higher temperatures, monoterpene emissions usually continue to increase in an exponential mode while the increase of isoprene emission slows down and finally drops after a temperature optimum. The temperature optimum typically lies around 40°C (Guenther et al. 1993), but may depend on the rate of temperature increase (Singsaas & Sharkey 1997).

This paper describes the light and temperature dependencies of individual compounds emitted from Q. ilex leaves. It further reports on the effect of heat stress on monoterpene release and on the onset of emission during leaf development. Our results bring new light into the debate on the similarity between isoprene emission from isoprene emitters and monoterpene emissions from Q. ilex, and question the adequacy of the isoprene model of Guenther et al. (1993) to simulate Q. ilex emissions under Mediterranean summer conditions.


Plant material

The experiment was carried out on a 10-year-old potted Holm Oak (Q. ilex) which originated from Tuscany, Italy. The tree was placed in a greenhouse 1 month before the experiment started and was regularly irrigated. A high pressure sodium lamp provided an artificial light from 0830 to 1700 h. Photosynthetically active radiation (PAR) was ≈ 1500 μmol photon m–2 s–1 at the top of the tree and the room temperature ranged from 10 to 15 °C at night and from 20 to 25 °C during the day. Measurements were conducted on eight different groups of leaves from the same tree (L1–L8). At the end of the experiment, leaf areas were measured with an optical area meter (Delta-T device, Burwell, UK) and leaf dry weights were determined after 4 d in a ventilated oven at 80 °C (Table 1).

Table 1.  . Projected area (cm2), dry weight (d.w. g) and specific leaf weight (SLW g m–2) of measured leaves Thumbnail image of

Plant enclosure system and gas exchange measurements

All measurements were performed in an open gas exchange system (minicuvette CMS400, Walz, Effeltrich, Germany) equipped with a 0·5 dm3 round chamber of Plexiglas (KK02, Walz). Air entering the chamber was pumped from outside and successively passed through a desiccant (silicagel plus charcoal) and a filter (CaSO4-anhydride) to remove water, monoterpenes and ozone. Air of 50–70% relative humidity was generated by bubbling part of the air stream in distilled water. Chamber flow was 1 dm3 min–1 for all experiments and air inside the chamber was kept in motion by an electric fan.

The chamber temperature was monitored by a temperature controller (RSV41, Walz) which allowed a steady (± 0·2°C) temperature between 5 and 45 °C to be reached in a few minutes. Leaf temperature was measured on the abaxial leaf side by a thermocouple. Different light intensities from 0 to 2000 μmol photon m–2 s–1 were supplied by an external light source and diverse neutral filters (LR4, Walz) inserted in the light path. Light intensity was measured inside the chamber at leaf level with a PAR sensor (Licor Quantum sensor, Lincoln, USA). CO2 and H2O concentration differences between inlet and outlet air were measured by a differential infrared gas analyser (BINOS 100, Leybold Hanau, Germany). Leaf net photosynthesis, transpiration rates and leaf water vapour conductance were calculated according to von Caemmerer & Farquhar (1981).

Monoterpene sampling and analysis

Monoterpenes were trapped on adsorption glass tubes (Chrompack, Middleburg, The Netherlands, 15 cm long and 3 mm inner diameter) filled with ≈ 125 mg Tenax TA (Aldrich, 20–35 mesh), plugged at both ends with small pieces of silanized glass wool and placed in a cooled sampling device. Defined volumes of air were sampled at the inlet and outlet ports of the chamber by means of a pump and two mass flow controllers (Brooks, Veenendaal, The Netherlands) connected to a timer (2 dm3 of air at a rate of 200 cm3 min–1). A prepurging time of 5–10 min on a bypass line was run before each sampling. A low concentration (10 ppb) of an internal monoterpene standard (β-citronellene, Fluka, Buchs, Switzerland) generated by an open tube diffusion system was introduced into the chamber inlet air in order to check sampling efficiency, terpene sinks and memory effects within the chamber (Staudt et al. 1995). Every day, measurements were made at the inlet and outlet ports of the empty chamber to detect eventual contamination by the diverse materials of the system.

Air samples were analysed by a gas chromatograph (GC CP9001, Chrompack) fitted with a flame ionization detector (FID) and a thermal desorption and cold trap injector (TCT/PTI CP4001, Chrompack). Compounds were separated on a fused silica capillary column (25 m × 0·32 mm, d.f.: 1·2 μm CP-Sil 8 CB, Chrompack) with helium as the carrier gas and nitrogen as the make-up gas. Desorption and analysis ran with the following temperature program: 3 min precooling to – 100 °C, 10 min desorption to 200 °C, 1 min injection at 200 °C. Gas chromatograph oven: 4 min at 65 °C, 2·5 °C min–1 to 80 °C, 2·0 °C min–1 to 100 °C and 20 °C min–1 to 240 °C. Peak identification and quantification were based on terpene solutions in methanol (liquid standards) prepared from commercial authentic monoterpenes of high purity (Fluka, Aldrich, 95–99% purity). The whole sampling and analysing system has been described in detail by Staudt et al. (1995).

The monoterpene emission rate was calculated as the difference between the air concentration in the chamber enclosing the leaf and the concentration measured in the empty chamber, multiplied by air flow. It was expressed both on the basis of projected leaf area (nmol m–2 s–1) and leaf dry weight (μg g–1 h–1).

Experimental protocol

All measurements were made between 0900 and 1600 h. During the night, leaves were removed from the chamber, except for the final experiment (see below). In the morning, the leaf was clamped in the chamber and measurements were started after 1 h equilibration to temperature and light set points. Each time PAR or temperature was varied in the chamber, an equilibration time of 45–60 min was required before air sampling, in order that emission rate stabilized. At high temperature (> 35 °C), this delay was reduced to 30 min in order to limit leaf stress and damage.

Three independent groups of measurements were carried out on different leaves. (i) The influence of leaf age was studied on six single terminal leaves, three of the current year (L1, L2 and L3) and three 1-year-old leaves (L4, L5 and L6). Among the youngest leaves, only L3 was totally expanded. Monoterpene emission was measured at 2000 μmol m–2 s–1 and 30 °C (two replications for each leaf, L1–L6). (ii) Response to light and temperature was studied on one fully expanded leaf of the current year (L7) and on a group of four terminal 1-year-old leaves (L8). The influence of light from 0 to 2000 μmol m–2 s–1 was measured at a leaf temperature of 16, 25, 30 and 42 °C. The response to leaf temperature was observed under three light intensities: from 5 to 45 °C at 2000 μmol m–2 s–1 and from 10 to 45 °C at 315 and 615 μmol m–2 s–1. Increments or decrements of 5–10 °C and 300–1000 (at high intensity) μmol m–2 s–1 were applied between two successive measurements. Many repetitions were performed on several days for each combination and both leaves (L7 and L8). (iii) The response to a 3 d continuous exposure at 45 °C and 2000 μmol m–2 s–1 was observed on a group of four 1-year-old leaves (L9). In the morning of the first day, leaves were clamped in the chamber and measurements started as soon as the temperature reached the set point. For 3 d, air was sampled every 60–90 min during the day and once in the night. The lamp was switched off from 1800 to 0830 h. During this period, leaves remained enclosed in the chamber and the leaf temperature decreased to 42 °C on the first night and to 39 °C on the second. Air humidity fell to 35–45% during the day. Leaves were still alive after the 3 d exposure.

Application of light and temperature models

In emission inventories, models typically use Eqn 1 to simulate the temperature dependence of monoterpene emission (Tingey et al. 1991; Guenther et al. 1993)

inline image

where M is the emission rate at leaf temperature T, MS is the emission rate at standard temperature TS (303 K or 30 °C) and β is an empirical parameter ranging from 0·057 to 0·144 K–1, with a typical average value of 0·09 K–1 (Guenther et al. 1993).

Isoprene emission is modelled as:

inline image

where E is isoprene emission rate at temperature T and PAR flux L, ES is the emission rate measured at standard temperature TS and PAR (1000 μmol m–2 s–1), and the scaling factors CT and CL are defined by:

inline image

inline image

where R is the gas constant (8·134 J K–1 mol–1) and CT1 (95000 J mol–1), CT2 (230000 J mol–1), Tm (314 K or 41 °C), α (0·0027) and CL1 (1·066) are empirical parameters derived from measurements made on four isoprene emitting plant species (Guenther et al. 1993). These five parameters were derived from our measurements of monoterpene emissions from Q. ilex by non-linear least square regression according to Marquardt (1963).


Amount and composition of leaf emissions at different developmental stages

Among the detected compounds, α-pinene, sabinene and β-pinene accounted for ≈ 85% of total emission under standard conditions. Myrcene, limonene, para-cymene, camphene, cis-β-ocimene, trans-β-ocimene, linalool and borneol were emitted in small quantities, as well as the sesquiterpene β-caryophyllene. Interestingly, at high temperatures (> 38 °C), emissions of cis-β-ocimene and trans-β-ocimene remarkably increased, and cis-β-ocimene was even the main compound between 40 and 45 °C (Fig. 1). The sum of monoterpenes emitted from fully expanded leaves at 30 °C and 1000 μmol photon m–2 s–1 (MS in Eqn 1 or ES in Eqn 2) was, on average, 4·1 nmol m–2 s–1[standard deviation (SD) = 0·75] or 15·1 μg g–1 h–1 (SD = 2·4). Under the same conditions, the average net CO2 assimilation rate was 6·85 μmol m–2 s–1 (SD = 1·2).

Figure 1.

. Emission rates (upper panel) and fractions (lower panel) of individual compounds emitted from Quercus ilex leaves as a function of temperature under saturating light intensity [photosynthetically active radiation (PAR) 2000 μmol m–2 s–1].

During early leaf development, the total emission rate rose by a factor 10 on a leaf dry weight basis and by a factor 7·5 on a leaf area basis, but remained rather stable after full leaf expansion (Fig. 2). The net CO2 assimilation rate of L1 was ≈ 0·75 μmol m–2 s–1, while it ranged between 3·5 and 6·5 μmol m–2 s–1 for the other five leaves. L1, the youngest expanding leaf emitted principally α-pinene, β-pinene, sabinene, linalool and para-cymene in similar proportions. All other leaves, whatever their age, emitted 80–90% of α-pinene, sabinene and β-pinene in constant proportions (43% α-pinene, 29% sabinene and 28% β-pinene). Among the other compounds, cis-β-ocimene and trans-β-ocimene were emitted by the 1-year-old leaves only, whereas the linalool emission rate was much higher for the younger leaves. A certain balance was observed between the main compounds and cis-β-ocimene and trans-β-ocimene: the more main compounds, the less cis-β-ocimene plus trans-β-ocimene.

Figure 2.

. Sum of monoterpene emissions from six different leaves of Quercus ilex calculated on a leaf area basis (black columns nmol m–2 s–1) and on a leaf dry weight basis (grey columns μg g–1 h–1). L1 and L2 are two young expanding leaves, L3 is a full expanded leaf of the current year, and L4, L5 and L6 are 1 year old. All measurements were made at a temperature of 30 °C and a photosynthetically active radiation (PAR) of 2000 μmol m–2 s–1.

Light and temperature dependence of monoterpene emissions

Responses of individual compounds to light and temperature indicated that cis-β-ocimene and trans-β-ocimene behaved differently than other monoterpenes, as mentioned above. Figure 3 shows the light responses measured at four different temperatures (left panels) and the temperature responses measured at three different light levels (right panels), for emissions of cyclic monoterpenes (Fig. 3a & e), ocimene forms (Fig. 3b & f) and the sum of all compounds (Fig. 3c & g). Emission rates were normalized to standard light or temperature in order to compare the responses of different compounds under various light and temperature regimes. The CO2 assimilation rate is depicted in Fig. 3(d & h) for the same experiment. Isoprene and monoterpene models are compared on this figure with their original parameters and with parameters adjusted on our data sets. The light response of ocimene emission rates was too variable to assume a single mathematical relationship (Fig. 3b).

Figure 3.

. Light (left panels) and temperature (right panels) dependencies of emissions of α-pinene + sabinene + β-pinene (a, b), cis-β-ocimene + trans-β-ocimene (b, f), total monoterpene emissions (c, g), and net CO2 assimilation A (d, h) of Quercus ilex leaves. Responses to light were measured at a leaf temperature of 16 °C (○), 25 °C (◊), 30 °C (□) and 42 °C (▵), and responses to temperature were measured at a photosynthetically active radiation (PAR) of 315 μmol m–2 s–1 (closed symbols), 615 μmol m–2 s–1 (grey symbols) and 2000 μmol m–2 s–1 (open symbols). Ratios ET/E30 °C, ET/E35 °C and EPAR/E1000μmol correspond, respectively, to the temperature and light scaling factors CT and CL of Eqns 3 and 4 or the ratio M/Ms in Eqn 1. Dashed lines represent the isoprene (a, c, e) and monoterpene (g) models with the parameters suggested by Guenther et al. (1993). Solid lines show the same models with parameters estimated from Q. ilex data by non-linear least square regressions.

As expected, the emission of each compound was stimulated by light. Apart from the large scattering of ocimene data, all emission rates, as well as the CO2 assimilation rate, exhibited a similar response to light, i.e. a steep linear increase with light at low intensity followed by a plateau between 600 and 900 μmol photon m–2 s–1. At the same time, the transpiration rate slightly increased from 1·0 to 3·0 mmol m–2 s–1 when light intensity increased from 200 to 2000 μmol m–2 s–1 and it was ≈ 0·3 mmol m–2 s–1 in the dark (data not shown). No obvious influence of temperature on the light responses was seen, although the emission rates measured at 42 °C tended to decrease at the higher light intensities, as did the CO2 assimilation rate. Fitted parameters of the light function (α and CL1 in Eqn 4) were not significantly different between the sum of the main cyclic compounds (α = 0·0041; CL1 = 1·040) and the sum of all compounds (α = 0·0046; CL1 = 0·998), and they were close to the values suggested for isoprene emissions (α = 0·0027; CL1 = 1·066; Fig. 3a & c). Compared with isoprene emission, monoterpene emission from Q. ilex increased with light more at low intensity and seemed to saturate at a lower light intensity.

The response to temperature was more variable, in particular at temperatures higher than 35 °C. From 5 to ≈ 35 °C, emissions of α-pinene, sabinene, and β-pinene increased exponentially with temperature, while cis-β-ocimene and trans-β-ocimene were not or hardly detected in this range (Fig. 3e & f). At higher temperatures, emissions of α-pinene, sabinene, and β-pinene slowed down from 35 to 40 °C and levelled off or dropped above 40 °C. On the contrary, cis-β-ocimene and trans-β-ocimene emission rates exhibited a deep increase from 35 to 45 °C, so that the sum of emissions increased exponentially over the whole range of temperatures (Fig. 3g). In the dark, emissions were very low, but still temperature dependent (not shown). In contrast to the light dependence, there was no similarity between the response of emissions to temperature and that of CO2 assimilation, which exhibited a broad optimum range between 16 and 31 °C (Fig. 3h). At higher temperatures, CO2 assimilation diminished but was still positive, even at 45 °C. Transpiration data (not shown) were largely scattered over the whole range of temperatures, but tentatively decreased at higher temperatures.

Although the emission rates were rather dispersed between 35 and 45 °C, the response of cyclic compounds suggested a temperature optimum as established for the temperature dependence of isoprene emission (Fig. 3e). Best fit values (± 95% confidence interval, n = 74) of the parameters of the temperature model (Eqn 3) were: CT1 = 87620 J mol–1 (± 2636), CT2 = 188200 J mol–1 (± 7428), Tm = 317 K (± 0·4). Until ≈ 35 °C, emissions increased exponentially with a slope (β in Eqn 1) of 0·121 °C–1 (R2 = 0·96, n = 47) close to that of the isoprene function (0·13 °C–1). At high temperatures, the curve adjusted on our data presented a higher optimum (≈ 42 °C) than the isoprene curve and a slight decrease above this threshold.

The temperature dependence observed for cis-β-ocimene and trans-β-ocimene emission (Fig. 3f) could be adjusted by an exponential function (Eqn 1) as for the sum of emission (Fig. 3g). Values of β estimated for the whole range of temperatures (5–45 °C) were 0·317 (R2 = 0·61), but 0·436 (R2 = 0·56) when low emission data under 30 °C were excluded. Finally, the coefficient β deduced for the sum of emissions from 5 to 45 °C equalled 0·108 (R2 = 0·96) which is close to the global value of 0·09 recommended by Guenther et al. (1993) (Fig. 3g).

Data plotted on 3Fig. 3e–g clearly show that there was no obvious interaction between light and temperature up to ≈ 35 °C, as normalized values (ET/E30 °C) measured under diverse light conditions followed a single curve. In the high temperature range, the emission decrease of cyclic compounds, and in reverse the emission increase of acyclic compounds, were more pronounced at 2000 μmol m–2 s–1 than at 600 μmol m–2 s–1, but again more pronounced at 300 μmol m–2 s–1 than at 2000 μmol m–2 s–1. Because the response of emission also strongly varied within single treatments (for example, at 300 μmol m–2 s–1), we tried to relate the emission variability observed at high temperature to other factors than light intensity, such as leaf water vapour conductance, CO2 assimilation rate, transpiration rate or hour of the day. We found no explanation for this variability, but for a given leaf (L7 or L8), the respective decrease and increase of cyclic and acyclic monoterpene emission were more stressed at the end of the experiment, which suggests a cumulative effect of the time of exposure (data not shown), which was examined in the following experiment.

Monoterpene emissions during a 3 d exposure at 45°C

During the 3 d exposure, α-pinene, sabinene and β-pinene on the one hand, and cis-β-ocimene and trans-β-ocimene on the other, behaved exactly in the same way. 4Figure 4a shows one representative compound of each group. Cis-β-ocimene was the main emitted compound with an initial absolute emission rate (≈ 7 nmol m–2 s–1) two times higher than that of α-pinene, and trans-β-ocimene was emitted in similar amounts to sabinene and β-pinene. The total emission rate (cyclic plus acyclic compounds) decreased over the 3 d from 17·6 to 6·6 nmol m–2 s–1. It was constant during the first day, because the emission of cyclic compounds dropped by ≈ 28%, whereas the emission of cis-β-ocimene and trans-β-ocimene increased by 30%. During the second and third days, α-pinene, sabinene and β-pinene emissions decreased during the day, while cis-β-ocimene and trans-β-ocimene increased, although a small drop occurred in the afternoon, which became earlier the longer the exposure lasted. The evolution of cyclic and acyclic compounds was exactly symmetric only on the first day.

Figure 4.

. Effect of a 3 d continuous heat exposure on (a) relative leaf emission rates of α-pinene (◊) and cis-β-ocimene (□) and (b) on net CO2 assimilation rate (A, ○) and leaf water vapour conductance (gH2O, ▵). Simultaneous variations of leaf temperature (Leaf T, +) and relative air humidity in the chamber (Air Rh, ×) are plotted on (a).

Proportional to total emission, cis-β-ocimene plus trans-β-ocimene increased during the day, while α-pinene, sabinene and β-pinene decreased. Nevertheless, over the whole period, relative proportions of these five compounds were maintained, as initial proportions were recovered after each night (data not shown). Leaf water vapour conductance and CO2 assimilation rates were stable during the first day, but dropped by 50–60% from the first to the second day and remained in the same range on the third day (Fig. 4b). For the last 36 h, CO2 assimilation continuously decreased during the day, contrary to leaf water vapour conductance.

Loss of photosynthetically fixed carbon by monoterpene emission

As all, or almost all, of the carbon included in monoterpenes is derived from carbon recently fixed by photosynthesis (Loreto et al. 1996b,c), the net photosynthesis and monoterpene emission rates were used to calculate the percentage loss of fixed carbon by monoterpene release from Q. ilex leaves. Because of the differences in light and temperature responses of monoterpene emission and CO2 assimilation (Fig. 3), a fairly steady carbon loss was found over the whole range of light, but a 50-fold increase was observed when the temperature rose from 10 to 45 °C. Carbon loss was ≈ 0·5% at 30 °C and ranged from 2 to 6% between 40 and 45 °C. The highest values of ≈ 20% were observed during the continuous heat exposure on the second and third days. Theoretically, the carbon consumption by monoterpene emission is even higher, taking into account the possible formation and release of CO2 within the monoterpene biosynthesis. The number of CO2 units formed per unit monoterpene depends on the pathway used for isopentenyl-diphosphate synthesis. If, for instance, monoterpenes are exclusively synthesized from pyruvate via the common mevalonate pathway, then the carbon consumption for one monoterpene could increase from 10 to 18 (Sharkey et al. 1991). Instead carbon consumption would increase to only 12 if all monoterpenes are synthesized from pyruvate and glyceraldehyde-3-phosphate via the novel mevalonate-independent pathway (Lichtenthaler et al. 1997). However, this additional carbon loss as CO2 would not affect the above-mentioned percentage losses referring to net CO2 assimilation rate, but it could have contributed to lessen the net CO2 assimilation rate of Q. ilex leaves measured, especially at high temperatures: from 25 to 45 °C, CO2 assimilation under saturating light decreased from ≈ 6 to 3 μmol m–2 s–1 (Fig. 3h), while monoterpene increased from 1·5 to 12 nmol m–2 s–1 at the same time. Assuming that all monoterpenes were synthesized from pyruvate via mevalonate, the additional CO2 formation would be 84 nmol m–2 s–1 which is ≈ 3% of the observed decrease in CO2 assimilation.


The emission factor Es, defined as the emission rate under standard conditions (30 °C and saturating light) is the most important factor for biogenic emission inventories. It determines the overall emission capacity of vegetation (Harley, Guenther & Zimmerman 1996; Sharkey et al. 1996) and is crucial for an accurate estimation of emission model parameters that are used to predict isoprenoid emissions according to temporal and geographical climatic fluctuations (Monson et al. 1995; Bertin et al. 1997). In this study, Es values measured on 1-year-old leaves ranged between 3·5 and 5·0 nmol m–2 s–1 on a leaf area basis and between 12 and 18 μg g–1 h–1 on a dry weight basis for the sum of monoterpenes. This is at the lower limit within the range of emission factors previously reported for Q. ilex (Staudt & Seufert 1995; Bertin & Staudt 1996; Loreto et al. 1996a; Bertin et al. 1997; Kesselmeier et al. 1997; Street et al. 1997). Variations in Es, which are assumed to account for long-term adaptations (Monson et al. 1995) may result from many factors, such as genotype, phenology, leaf and plant age or growth conditions (Tingey et al. 1991; Pio et al. 1993; Street et al. 1997). It has been found, for instance, that within a natural Q. ilex canopy, Es is substantially higher for sun leaves than for shade leaves (Bertin et al. 1997), as reported for several isoprene emitters (Harley et al. 1996; Sharkey et al. 1996).

During early leaf development (i.e. before full expansion, Fig. 2), few qualitative but strong quantitative changes occurred in the emission under standard temperature and light. This phenomena has been scarcely mentioned in the literature for monoterpenes. A 13C-labelling experiment of spruce seedlings showed that the appearance of individual monoterpenes is staggered during needle development (Jüttner & Bufler 1988). For isoprene emitters, the onset of emission during leaf ontogeny has been related to isoprene synthase activation (Harley et al. 1994; Monson et al. 1994). Kuzma & Fall (1993) reported a 100-fold increase of both isoprene emission rate and enzyme activity from leaf emergence to 14 d old for velvet bean. Likewise, the 10-fold increase of monoterpene emission and the change in the emission pattern during leaf expansion observed in this work could relate to the activation of a specific enzyme within the monoterpene biosynthesis path.

Quercus ilex leaves emitted large amounts of monoterpenes only under illumination. Dark emissions were low, ≈ 20–40 times lower than light emissions at the same temperature. The light dependence was observed for all emitted compounds and was similar to that of CO2 assimilation and isoprene emission from temperate plant species (Guenther et al. 1993). Recent labelling experiments and gas exchange studies on Q. ilex leaves demonstrated that the light dependence of monoterpene emissions results from its direct link to a photosynthesis-dependent synthesis, as for isoprene (Loreto et al. 1996b, c). Our findings are consistent with these studies, but give evidence that, unlike isoprene synthesis, monoterpene synthesis never completely ceases when CO2 assimilation stops. This has also been observed during a drought stress period (Bertin & Staudt 1996).

The response to temperature of almost all compounds was an exponential increase of emissions between 5 and 35 °C, as already acknowledged (Staudt & Seufert 1995; Loreto et al. 1996a), but beyond 35 °C, single compounds behaved differently. While emissions of cyclic compounds such as α-pinene, sabinene and β-pinene progressively declined, as typically observed for isoprene (Guenther et al. 1993), emissions of the two isomers, trans-β-ocimene and cis-β-ocimene, increased up to 45 °C. These opposite responses suggest the replacement of the main cyclic compounds, α-pinene, sabinene and β-pinene, by the two acyclic compounds, trans-β-ocimene and cis-β-ocimene. To our knowledge, such dissimilarities in the temperature response of individual monoterpenes emitted from the same plant species have never been reported in the literature, although a recent field study indicated that analogous processes are responsible for the seasonal appearance of trans-β-ocimene in the emissions from Pinus pinea (Staudt et al. 1998). Loreto et al. (1996c) also observed an apparent competition between cis-β-ocimene and α-pinene emissions from Q. ilex leaves, but in response to relative air humidity. They proposed that a higher transpiration rate at low humidity would favour the emission of more water-soluble terpenes, like ocimene. Our results did not support this theory, as we could not find any correlation between leaf transpiration and emission of individual compounds at high temperature, neither in the short term nor the long term (data not shown).

We suggest that changes in the emission pattern result from an alteration of the product pattern of monoterpene biosynthesis, where cyclic and acyclic compounds are formed at the same site and within the same enzymatic step, that is the conversion of the monoterpene alcohol diphosphate ester geranyl pyrophosphate (GPP) to monoterpene hydrocarbons by monoterpene cyclases inside plastids (Gershenzon & Croteau 1990). α-Pinene and cis-β-ocimene are both likely to be synthesized from photosynthetic carbon inside the chloroplast of Q. ilex leaves (Loreto et al. 1996c), and several isoenzymes may coexist in plants, each of them able to produce diverse cyclic and acyclic compounds (Gleizes et al. 1982; Croteau et al. 1988; Colby et al. 1993). For instance, pinene cyclases I and II, isolated from sage, produce minor quantities (5–12%) of ocimenes beside the main pinene compounds, but large amounts when the unnatural linalyl pyrophosphate is supplied as the substrate instead of the natural precursor GPP (Croteau et al. 1988). One could speculate that climatic conditions induce some interconversions of the precursor forms in Q. ilex leaves as assumed for rose and mint plants (Akhila & Thakur 1989). More simply, we believe that high temperature impedes the enzymatic cyclization of GPP, promoting ocimene output, or that different isoenzymes with different temperature optima coexist and compete for the same substrate.

As could be seen during continuous heat exposure (Fig. 4), the alteration of the emission pattern did not depend only on absolute temperature, but also on the time of exposure. Especially during the first day, ocimene emission clearly increased at the expense of cyclic compounds. The subsequent recovery during the night is likely due to the lower leaf temperature (≈3°C). Generally, the reversibility of thermal inhibition of many biological processes extends up to 8 °C above the temperature inactivation threshold, but largely depends on the time of exposure (Sharpe & De Michelle 1977). In the longer term (second and third days), additional heat effects were involved, as the sum of emission decreased from day to day and the compensation between cyclic and acyclic compounds became less obvious. The drop of total emission may result from progressive heat damage within the monoterpene biosynthesis pathway, and/or from a lack of ATP, reduction power or carbon substrates supplied by photosynthesis, which strongly declined. Nevertheless, Q. ilex seems well adapted to cope with heat stress, as leaf CO2 assimilation and stomatal conductance, as well as total monoterpene emission, remained high for at least 8 h at 45°C, and dropped only when heat lasted over night. Moreover, leaves were not irreversibly damaged after the 3 d heat exposure. Quercus ilex is known as a thermo- and drought-tolerant tree species, able to support low water potentials and temporary heat stress up to 50 °C without losing its photosynthetic capacity (Pigott & Pigott 1993; Trabaud & Methy 1994;). Thus, leaves can maintain relatively low but steady gas exchange under Mediterranean summer conditions, when the maximum air temperature often exceeds 40 °C and the leaf temperature may be several degrees higher (Trabaud & Methy 1994). However, under these conditions the low photosynthetic carbon gain will be further reduced by the increase of monoterpene emissions. From our study, the percentage of fixed carbon lost as monoterpenes was as high as that reported for isoprene emitters (Harley et al. 1996; Sharkey et al. 1996).

Previous studies strongly suggested that factors and mechanisms controlling monoterpene emissions from Q. ilex are the same as for isoprene emission, and that the isoprene model suggested by Guenther et al. (1993) is also appropriate to simulate emissions from Q. ilex. Our results did not confirm this totally. The light dependence was indeed similar to that of the isoprene model and no obvious interrelationship between the short-term effects of light and temperature could be observed as assumed by the model. However, we could clearly establish that high temperature exposure strongly influences the emission composition, and that individual compounds exhibited different responses to temperature, among which only the response of cyclic monoterpenes was comparable to that of isoprene. It was not the case for the sum of emissions, which increased exponentially up to at least 45 °C, as typically observed for monoterpene emissions from other plants. Consequently, the appropriate temperature model for monoterpene emissions from Q. ilex would be that given by Eqn 1 or by another function like the Arrhenius equation (Singsaas & Sharkey 1997), but not that given by Eqn 3. This can be quite relevant to assess the role of vegetation emissions in the formation of tropospheric ozone in the Mediterranean basin, where Q. ilex is the dominant species in various shrub lands and forest formations. Contrary to what was hitherto expected, the release of unsaturated hydrocarbons from this vegetation will not level off during the hottest hours in summer, but will further increase with a high proportion of compounds even more reactive than those dominating at moderate temperatures (Atkinson, Hasegawa & Aschmann 1990).


This work was realized at the Environment Institute of the Joint Research Centre in Ispra, Italy in the frame of the European BEMA (Biogenic Emission in the Mediterranean Area) project. It was financially supported by a postdoctoral grant from the EEC. We thank Dr G. Seufert and Dr B. Versino who supported the realization of this work.