Volatile organic compounds (VOCs) emitted by forests strongly affect the
chemical composition of the atmosphere. While the emission of isoprenoids
has been largely characterized, forests also exchange many oxygenated VOCs
(oVOCs), including methanol, acetone, methyl ethyl ketone (MEK), and
acetaldehyde, which are less well understood. We monitored total
branch-level exchange of VOCs of a strong isoprene emitter (
Global budgets indicate that biogenic emissions of volatile organic compounds (VOCs) are about an order of magnitude larger than those from anthropogenic sources (Benkovitz et al., 2004; Guenther et al., 1995). VOCs emitted from forest environments account for approximately half of the reactive carbon introduced globally into the atmosphere (Guenther, 2002), with isoprene alone contributing about a third of the worldwide VOC emissions (Guenther et al., 2006). Biogenic VOCs are typically more reactive than anthropogenic ones and more readily contribute to the formation of secondary organic aerosols (Atkinson and Arey, 1998; Carlton et al., 2009; Claeys et al., 2004; Griffin et al., 1999; Helmig et al., 2006; Kanakidou et al., 2005; Kavouras et al., 1998). Oxygenated VOCs (oVOCs) constitute an abundant class of atmospheric VOCs. oVOCs can be emitted from biogenic and anthropogenic sources and are often produced by secondary reactions (mainly VOC oxidation) in the atmosphere (Galbally and Kirstine, 2002; Jacob et al., 2002). The most abundant biogenic oVOC in the troposphere is methanol, typically followed by acetone, formaldehyde, and acetaldehyde. A vast number of other oVOCs are also present in the atmosphere (Goldstein and Galbally, 2007). In particular, oxidation products of isoprene, such as methyl vinyl ketone (MVK), methacrolein (MACR), isoprene hydroperoxides (ISOPOOHs), and isoprene epoxydiols, play a critical role in reactive carbon cycling, SOA formation, and modulation of the atmospheric oxidation capacity (Rivera-Rios et al., 2014). Atmospheric oVOCs are removed via SOA formation, photooxidation, or dry and wet deposition. Their atmospheric lifetimes vary widely depending on functionality, from several hours for MVK, MACR, and ISOPOOH, to 10 days for methanol and 1 month for acetone.
Forests not only are major sources of biogenic VOCs but also can act as efficient oVOC sinks. oVOCs can be lost by dry deposition onto plant surfaces or by uptake into the plant through the stomata. Recently, plant uptake has been shown to be a more significant mechanism for oVOC removal than previously assumed (Karl et al., 2010). Plant uptake of oVOCs requires an in vivo sink, such as enzymatic consumption (Cojocariu et al., 2004). Despite many studies documenting plant uptake, the mechanisms of these sinks, which vary across compounds, are not yet fully understood. Uptake of methanol and acetaldehyde by plants has been reported (Jardine et al., 2008; Laffineur et al., 2012), and mechanisms for their metabolism within plants have been proposed (Gout, 2000; MacDonald and Kimmerer, 1993). As both biogenic sources and sinks of these compounds exist, bidirectional exchange can be observed (Jardine et al., 2008; Karl et al., 2010; Laffineur et al., 2012; Misztal et al., 2011; Schade et al., 2011). Similarly, there are many reports of negative acetone fluxes at ecosystem level (Karl et al., 2004, 2005), suggesting deposition or plant uptake. However, evidence for acetone uptake at plant level is inconclusive (Cojocariu et al., 2004; Tani and Hewitt, 2009), and no mechanism driving acetone uptake by plants is known (Cojocariu et al., 2004).
A plant sink of isoprene oxidation products MACR and MVK has also been demonstrated (Andreae et al., 2002; Karl et al., 2004, 2005, 2010), suggesting a metabolic consumption of these compounds, especially given that MVK and MACR are toxic. For MACR, detoxification mechanisms have already been elucidated (Muramoto et al., 2015), whereas knowledge of metabolic pathways consuming MVK is scant. At the same time, some studies have suggested that isoprene plays an antioxidant role within plants, reacting with reactive oxygen species (ROS) to produce MVK and MACR, which are then emitted (Fares et al., 2015; Jardine et al., 2012). Production and emission of isoprene oxidation products by plants is still matter of debate, but if it does occur, it could explain why a detoxification mechanism for MVK and MACR also exists. Moreover, very little is known about plant emission and uptake of other common biogenic oVOCs, such as methyl ethyl ketone (MEK). MEK can lead to ozone and PAN production in the atmosphere (Pinho et al., 2005) and photochemical odd-hydrogen formation in the upper troposphere (Atkinson, 2000; Baeza Romero et al., 2005). Recent studies point out the great importance of biogenic MEK sources (Yáñez-Serrano et al., 2016); however, the origin of biogenic MEK remains unclear.
Among observed oVOCs, several belong to the class of benzenoid compounds, which are believed to be synthesized and emitted by vegetation under stress conditions and as chemical signals, although direct observations of such compounds are still very limited (Bouvier-Brown et al., 2009; Heiden et al., 1999; Jardine et al., 2010; Kim et al., 2010; Leone and Seinfeld, 1984; Misztal et al., 2010, 2015; Owen et al., 2002; White et al., 2009). Uptake of benzenoid compounds such as phenol and benzaldehyde has been seen in houseplants but has not been reported in forest environments so far (Kondo et al., 1996; Tani and Hewitt, 2009). In general, the impact of vegetation on the atmospheric concentrations of benzenoid compounds has usually been overlooked. Only recently, the first suggestion emerged that plant emissions of benzenoid compounds might be comparable to those from anthropogenic sources (Misztal et al., 2015).
In general, direct plant-level field observations of the balance between emissions and deposition of oVOCs are scarce. The present study aims to directly investigate exchange of oVOCs at branch level in a forest environment. The compounds include methanol, acetone, acetaldehyde, MEK, the isoprene oxidation products MACR, MVK, and ISOPOOH, and the aromatic compound benzaldehyde. Isoprene and monoterpenes are also included as references. Emissions and/or uptake of each compound compared with the literature reports and with recent canopy-scale flux measurements at a nearby site (McKinney et al., 2011). Moreover, observations were corroborated with ancillary fumigation experiments with oVOC standards. The results (i) advance the current understanding of oVOC exchange, suggesting plant detoxification of MVK and a mechanistic explanation for the emission of MEK; (ii) provide direct evidence of oVOC exchange at branch level, testing in particular the presence of bidirectional exchange of benzenoids; (iii) suggest avenues for further mechanistic studies on oVOC exchange.
Experiments were performed at Harvard Forest, a New England mixed forest
located in Petersham, MA (42.54
The forest stand is 85–120 years old and has been mainly undisturbed over the past 70 years (Urbanski et al., 2007). The canopy is about 22.5 m high. Measurements of forest leaf area index are routinely made as well as meteorological data (Boose, 2001). Branch enclosure measurements were made using canopy-top branches of red oaks located on the south or east side of the tower.
Air samples were drawn from several heights on the tower (7.5, 15, 22.5, 30 m)
through FEP Teflon tubing (0.635 cm outer diameter
Background measurements were made by adding zero air to the inlet line at
tower top through a third FEP Teflon line. Zero air was generated by
compressing ambient air and delivering it to a catalytic converter
consisting of palladium coated alumina heated to 290
Red oak branches accessible from the walk-up tower were selected for branch enclosure measurements. Three cylindrically shaped PFA Teflon branch enclosures (ca. 5 L) were used to encase three unshaded upper-canopy branches (22.5 m above ground). At the end of the experiments leaves were collected and leaf areas were measured via the software ImageJ (version 1.47; Schneider et al., 2012). A fourth, empty enclosure was placed close to the others to serve as a background.
Ambient air was continuously supplied to each enclosure via FEP Teflon
tubing (0.635 cm OD
Average measured emission and deposition rates at branch scale in
the upper canopy (14–25 August). Data are reported as mean
Average measured emission and deposition rates at branch scale in
the upper canopy (25 August–1 September). Data are reported as mean
A PTR/SRI-ToF-MS instrument was continuously drawing 200 sccm of the sample flow for
analysis. VOC exchange rates were computed from differences
between the concentration in the branch enclosure and those in the empty
enclosure, converted in nmol m
On 2 September ancillary fumigation experiments were carried out. Gas
cylinders containing known amounts of the target VOCs (Scott Specialty
Gases, Inc.) were diluted with zero air using mass flow controllers (MKS
Instruments) and delivered to the branch enclosures. Tested VOCs included
isoprene (80 ppm
VOC measurements were performed by a PTR/SRI-ToF-MS 8000 (Ionicon Analytik
GmbH, Innsbruck Austria) equipped with a switchable reagent ion system
(Jordan et
al., 2009), allowing either NO
Details of the instrument operation are reported in the Supplement.
Averaged rates of exchange of the most important oVOCs and volatile isoprenoids measured in the present experiments at Harvard Forest are reported in Tables 1 and 2 and Fig. 1. The reported compounds were selected based on canopy-scale flux measurements previously performed at a nearby site (McKinney et al., 2011) with the addition of benzaldehyde. The results differ strongly among compounds, and each compound is discussed separately below. The main focus is investigating factors affecting bidirectional exchange of oVOCs with the forest upper canopy. Vertical concentration profiles and corollary fumigation experiments are used to corroborate conclusions.
Diel average measured mixing ratios and fluxes in enclosures containing upper-canopy red oak branches at Harvard Forest. Mixing ratios of the inflow ambient air and of the outflow air are shown in red and black, respectively. Diel average temperature and PAR during the measurement period are also shown. Error bars represent standard errors of the data points in each time interval.
As expected, isoprene emission showed a diel cycle peaking in the middle of
the day and dropping at night (Fig. 1). Isoprene emission during daytime
averaged to ca.
Average daytime and nighttime vertical concentration profiles. Measured VOC concentrations at various heights within and above canopy are reported. Horizontal bars represent standard deviations.
Summary of the ancillary experiments: Fumigation of forest red oak upper-canopy branches with VOCs.
Literature findings are in agreement with these results.
Sharkey et al. (1996) reported emissions
of isoprene of ca.
The major atmospheric source of MVK, MACR, and ISOPOOH is gas-phase isoprene
oxidation. If there is a plant or surface sink of these compounds, a
negative flux would be expected. The existence of a sink is supported by
many canopy-level studies, which show deposition of MVK Previous studies reporting fluxes of MVK
Branch enclosure measurements of isoprene oxidation product fluxes from high
isoprene emitters such as red oak are challenging due to possible
interferences from the high levels of isoprene present in the enclosures
during daylight. These interferences can arise from gas-phase oxidation of
isoprene in the sampling system or in the PTR-MS drift tube. As a result,
only measurements during darkness are reported for isoprene oxidation
products. For daytime, a constraint of 38 nmol m
Example of branch-level VOC flux measured as a function of MVK
fumigation mixing ratio during daytime. The concomitant MVK uptake and
emission of MVK reduction products (namely MEK, 3-buten-2-ol, and 2-butanol)
implies plant detoxification of MVK. Linear fit parameters are the
following: MEK (
The absence of uptake saturation (Fig. 3) suggests the existence of a degradation mechanism for isoprene oxidation products inside leaves. Isoprene oxidation products are thought to be cytotoxic and their rapid removal may increase plant survival under stress conditions (Oikawa and Lerdau, 2013; Vollenweider et al., 2000). Muramoto et al. (2015) studied the uptake of MACR by tomato plants and proved that MACR concentrations up to more than 100 ppmv are efficiently degraded by glutathionylation and enzymatic reduction within the leaves. Glutathionylation may be a more efficient mechanism for MACR removal than other hypothesized pathways, e.g., enzymatic degradation by aldehyde dehydrogenase (Karl et al., 2010; Kirch et al., 2004). Muramoto et al. (2015) identified release of isobutyraldehyde (5.8 % of the fumigated MACR) produced by reduction of MACR. Reduction of carbonyl compounds can therefore be a mechanism of detoxification of oVOC deposited to vegetation.
Contrary to MACR, the fate of MVK taken up by leaves is unknown.
Interestingly, in fumigation experiments, the observed uptake of MVK was
matched by an emission of MEK, 2-butanol, and 3-buten-2-ol of similar
magnitude in total (see, e.g., Fig. 3). Specifically, emission of MEK
accounted for 82
Branch-level emission of MEK in red oak upper canopy ranged between 0 and 460 nmol m
Very little is known about MEK emission and deposition in forest
environments. Kim
et al. (2010) detected significant amounts of MEK over a Ponderosa pine
forest in western Colorado during both days and nights in summer, and
Karl et al. (2005)
measured positive fluxes of MEK over a loblolly pine plantation.
Brilli et al. (2016) reported positive
and negative MEK fluxes at canopy level over a poplar plantation. In
laboratory experiments,
Holzinger et al. (2000) measured emission of MEK from leaves of young
McKinney et al. (2011) reported slightly negative
MEK fluxes during nighttime, which were not seen in our experiment on
branches. Perhaps at night MEK sinks other than leaf uptake are present. Our
nighttime vertical profiles of MEK (Fig. 2) show a decreasing trend with
height, with MEK concentrations lower under the canopy (0.12 ppbv on
average) than within (0.18 ppbv) and above canopy (0.2 ppbv). This may be
consistent with a ground sink causing an ecosystem-scale negative flux of
MEK during nighttime.
Karl et al. (2005),
measuring vertical MEK profiles within a loblolly pine plantation, also
deduced a ground sink. MEK uptake in houseplants
(Tani and Hewitt, 2009) and in
The metabolic pathway leading to the emission of MEK by plants is still
unclear. As discussed in Sect. 3.2, MEK might be produced by MVK
detoxification via enzymatic reduction. During daytime, the sources of MVK
that might undergo reduction in the leaf are twofold. First, there is uptake
of atmospheric MVK. Second, given the fact that
Scatter plot of branch-level MEK fluxes versus
MACR
Due to unreliable estimates of daytime MVK uptake, a plot with daytime MEK
emission (similar to Fig. 4) cannot be produced. However, maximum MVK uptake
rates can be estimated from measured MVK concentrations based on the
ancillary MVK fumigation experiments. For example, maximum MVK uptake rates
of ca. 4 nmol m
Both emission and deposition of benzaldehyde by upper-canopy branches were
detected (Fig. 1). On average the net daily benzaldehyde exchange was
positive in the period of 14–25 August, with a mean daily flux of 0.4 nmol m
Example of benzaldehyde daily pattern for a day when benzaldehyde emission was detected. Measurements correspond to three different red oak upper-canopy branches.
Anthropogenic sources, such as evaporation or combustion of fossil fuels, typically have been considered the main benzenoid sources in the atmosphere (Rasmussen and Khalil, 1983). It is also formed in the atmosphere through the photochemical degradation of toluene (Atkinson et al., 1980; Leone and Seinfeld, 1984) or from other precursors such as styrene and methylstyrene (Grosjean, 1985). Misztal et al. (2015) recently estimated that biogenic emissions of benzenoid compounds including benzaldehyde rival these other sources.
Benzaldehyde is an aromatic benzenoid compound naturally produced as a plant
volatile (Graedel, 1978) and component of floral scents
(Knudsen et al., 2006;
Pichersky et al., 2006). Benzaldehyde in plants is generally synthesized
from phenylalanine (Boatright et
al., 2004; Dudareva et al., 2006), which is produced via the shikimate
pathway (Herrmann, 1995). Emissions of benzaldehyde have been
reported from
A linear relationship between benzaldehyde deposition flux and benzaldehyde concentration in the branch enclosures (Fig. 6) indicates benzaldehyde exchange is regulated according to Fick's law (Karl et al., 2010; Omasa et al., 2002). Uptake of benzaldehyde by stomata has been suggested in houseplants (Tani and Hewitt, 2009). Daytime uptake rates of benzaldehyde suggest the existence of a compensation point (Fig. 6) of around 0.02–0.025 ppbv. Further studies in controlled conditions would be needed to confirm the presence of a compensation point and to elucidate its dependence on environmental and physiological parameters. Further investigations are also required to explain why benzaldehyde is deposited. Early studies demonstrated that benzenoid compounds such as benzene and toluene may be assimilated by crop plants grown under sterile conditions via aromatic ring cleavage and successive incorporation of their carbon atoms into nonvolatile organic acids or amino acids (Durmishidze, 1975; Ugrekhelidze et al., 1997; Ugrekhelidze, 1976).
Flux measured as a function of mixing ratio for benzaldehyde
during daytime of the whole branch-level measurement campaign. Black
triangles represent the period of 25 August–1 September, when deposition
dominated, and grey dots correspond to the period of 14–25 August, where mostly
emission was measured. Linear fit parameters are the following:
Figure 6 also shows that benzaldehyde emissions were detected during the first measurement period when deposition would be expected on the basis of its gas-phase mixing ratio. As mentioned, benzaldehyde emissions from leaves are typically associated with stress conditions (Misztal et al., 2015). This might have been the case, even though it was not possible to determine whether a stress was the driving factor.
Bidirectional exchange of acetaldehyde was observed (Fig. 1, Tables 1, 2),
with the strongest emissions (up to 170 nmol m
Acetaldehyde uptake by plant canopies from the atmosphere occurs primarily via stomata (Karl et al., 2005; Rottenberger et al., 2004) whereas surface deposition is minor (Jardine et al., 2008). Kondo et al. (1998) demonstrated that a biological sink for acetaldehyde should exist within plants since they are capable of continuous uptake for as long as 8 h even under ambient acetaldehyde concentrations exceeding 200 ppbv.
Jardine et al. (2008) demonstrated that acetaldehyde exchange rates are controlled by ambient acetaldehyde concentrations, stomata resistance to acetaldehyde, and acetaldehyde compensation point, which suggests a plant source as well as a sink. The cause of acetaldehyde emissions from plants remains elusive. Acetaldehyde might be formed in leaves from ethanol translocated from roots, especially when anoxic conditions prevent root respiration (Kreuzwieser et al., 1999) or from cytosolic pyruvate, e.g., after darkening (Karl et al., 2002a) or wounding (Loreto et al., 2006). In the period of study, we report net acetaldehyde emission from the upper-canopy branches, which may be due to any of these sources.
The measured bidirectional flux of acetaldehyde implies a compensation point in the range of 0–3 ppbv. During transitions from emission to uptake (e.g., on 25–28 August) the value was better constrained to 1–2 ppbv. The acetaldehyde compensation point seems to increase with increasing light intensity as might be deduced comparing light intensity, acetaldehyde mixing ratios, and fluxes (see, e.g., Fig. 1, 25–28 August). This might imply a light control on acetaldehyde exchange in the field, in line with the findings of Jardine et al. (2008) in laboratory experiments.
A great variety of plants have constitutive emissions of acetone
(Bracho-Nunez
et al., 2013; Isidorov et al., 1985). High acetone concentrations have been
reported above several forests (e.g.,
Geron
et al., 2002; Helmig et al., 1998; Karl et al., 2003; Müller et al.,
2006; Pöschl et al., 2001), tree plantations
(Brilli et al., 2014b), and after
biomass fires (Brilli
et al., 2014a). Red oaks also emit acetone
(Steinbrecher et al., 2009). In our
branch-scale measurements acetone emission reached 340 nmol m
The branch-scale measurements can be scaled up to daytime average
canopy-scale fluxes of about ca. 20–40
Dependence of biogenic acetone emission on temperature has been consistently
reported
(Bracho-Nunez
et al., 2013; Cojocariu et al., 2004; Grabmer et al., 2006; Janson and de
Serves, 2001; Schade and Goldstein, 2001), whereas the relationships with
light and stomata conductance are less clear
(Bracho-Nunez
et al., 2013; Cojocariu et al., 2004; Filella et al., 2009). However, relative humidity is believed to have a negative influence on acetone
canopy fluxes (McKinney et al., 2011). Indications
in the same direction have been found by
Cojocariu et al. (2004) by studying acetone
emissions from
McKinney et al. (2011) reported several events of
negative fluxes of acetone (up to
Methanol production by vegetation is known to be related to growth and tissue repair processes. It is produced from the demethylation of membrane pectins during formation of cell walls (Fall, 2003; Fall and Benson, 1996). Long-term methanol monitoring (Schade and Goldstein, 2006) shows methanol emission peaks during spring in conjunction with the rapid growing of leaves. Hence, growth rate is a key factor controlling plant methanol emission as it strongly influences internal production rate within the plant (Harley et al., 2007; Hüve et al., 2007). Additional drivers of methanol emissions include temperature and stomatal conductance (Harley et al., 2007; Hüve et al., 2007). While in the case of acetaldehyde stomata exert long-term control on emission rates (Jardine et al., 2008), stomatal limitations on methanol emissions are short-term. As methanol (as well as other oVOCs) is more soluble in water than non-oxygenated VOCs, it builds up liquid foliar pools that result in a constant flux out of the leaf despite stomatal movements (Niinemets and Reichstein, 2003).
During the campaign, methanol emissions from red oak upper-canopy branches
were up to 3000 nmol m
Example of enclosure flux measured as a function of fumigation mixing ratio for methanol in the field.
No methanol deposition was observed at branch scale during the campaign. The ancillary experiments (Table 3, Fig. 7) show that for the same branches methanol deposition occurs when ambient concentration is rather high. Methanol compensation points in our data were ca. 15–20 ppbv at night and ca. 25–45 ppbv during the day (Fig. 7). However, throughout the campaign ambient methanol mixing ratios did not exceed 10 ppbv (Fig. 1), which explains why no uptake was observed under ambient conditions. McKinney et al. (2011) reported near-zero methanol fluxes at Harvard Forest during night, although methanol uptake events were detected, often in the early morning. Vertical profiles of methanol concentrations (Fig. 2) are consistent with a methanol source within the canopy during daytime while at nighttime the lowest layer displays a significantly lower concentration than the upper layers. This might indicate a lower-canopy/ground nighttime methanol sink.
In the fumigation experiments, methanol deposition is not transient and a compensation point is present (Fig. 7), suggesting the existence of a methanol degradation mechanism. Likewise, many studies have reported significant canopy-scale methanol deposition to vegetation or soil and evidence of bidirectional exchange (Wohlfahrt et al., 2012), but a holistic mechanistic explanation is lacking. Other studies suggest that methanol is taken up by stomata. For example, Gout (2000) reported stomatal uptake followed by oxidation of methanol producing formaldehyde. Given the similar uptake rates found during day and night (Fig. 7), this would imply that stomata were still (partially) open at night (Niinemets and Reichstein, 2003). As already mentioned, this is corroborated by the fact that the water vapor emissions remain positive at night, although lower than daytime emissions (data not shown). Another possibility is the consumption of methanol by microorganisms that are commonly found on leaves (Holland and Polacco, 1994) and are able to degrade it via enzymatic reactions, e.g., methylotrophic bacteria (Duine and Frank, 1980). A further hypothesis is chemical transformation of methanol dissolved in water films on leaf surfaces, e.g., via reaction with OH radicals (Elliot and McCracken, 1989).
Although isoprene is by far the most abundant isoprenoid emitted by red
oaks, monoterpene emission may also be found. In previous branch enclosure
studies, Kim et al. (2011) and
Helmig et al. (1999) both detected
monoterpene emissions from red oaks amounting to about 1.1 % of total
isoprenoids, though with slightly different monoterpene compositions. In our
experiment, the daytime monoterpenes
In the literature, plant species are considered to emit either isoprene or
monoterpenes (Harrison et al., 2013), which
indicates a possible trade-off of available carbon among isoprenoids
(Dani et al., 2015). However, there are important
exceptions. First, there are families that actually emit both isoprenoids,
although only monoterpenes are stored in permanent pools
(Loreto and Fineschi, 2015). Second, emissions of
isoprene and monoterpenes in the same plants may occur at different
developmental stages. In general, monoterpenes are emitted by young,
expanding leaves that upon maturation produce and emit only isoprene. For
example,
Oaks emitting large amounts of monoterpenes have been described and are
generally evergreen species living in Mediterranean conditions
(Loreto et al., 1998, 2009). These
oaks do not store monoterpenes in permanent pools and therefore emit
monoterpenes in an isoprene-like fashion, with similar light and temperature
dependence (Loreto et al., 1996). We measured
daytime emission of monoterpenes with a diurnal cycle very similar to that
of isoprene (Fig. 1), and a highly significant correlation is present
between daytime isoprene and monoterpene emissions (Pearson correlation
coefficient 0.88,
Finally, monoterpene deposition was found on only one out of the three branches sampled (Table 2). By analogy with the much larger emissions of isoprene where deposition was not observed, it may be suggested that monoterpene deposition may be rare or absent.
The observations presented herein provide new information about the bidirectional exchange of oVOCs between forest plants and the atmosphere. In the predominantly isoprene-emitting red oaks that were the subject of our investigation, we found a link between exchanges of MVK and MEK, which provides a new framework for the understanding of MEK exchange, based on a possible in-plant source from isoprene oxidation products. This link supports the hypothesis of plant stress defense by isoprene via reaction with ROS. Production of MEK in excess of MVK uptake and in correlation with isoprene emission suggests within-plant oxidation of isoprene to MVK followed by detoxification and the eventual release of volatile products such as MEK, 3-buten-2-ol and 2-butanol. Further studies are needed to confirm this link. We also found small emissions of monoterpenes, which might be a marker of juvenility in the canopy. We report bidirectional exchange of benzaldehyde between biosphere and the atmosphere, describing a so far unknown (to our knowledge) foliar deposition of benzenoid compounds. More investigations on benzenoid bidirectional exchange are needed to improve global budgets of the biogenic and anthropogenic sources of these volatiles. Emissions and fluxes of other important oVOCs such as acetaldehyde, methanol, and acetone have been confirmed.
The full dataset is available from the authors upon request (email). Data will also be available on the Harvard Forest archive.
L. Cappellin and K. McKinney designed this research; field work was carried out by L. Cappellin, A. Algarra Alarcon, I. Herdlinger-Blatt, J. Sanchez, and K. McKinney. Laboratory analyses were performed by A. Algarra Alarcon and L. Cappellin. L. Cappellin, F. Biasioli, S. Martin, F. Loreto, and K. McKinney wrote the paper.
The authors declare that they have no conflict of interest.
Luca Cappellin acknowledges funding from H2020-EU.1.3.2 (grant agreement n. 659315). The authors thank Evan Goldman and J. William Munger for providing LAI data and Mark Vanscoy for providing meteorological data and for managing the measurement site. Edited by: J. Rinne Reviewed by: two anonymous referees