Emission of isoprene and carbonyl compounds from a boreal forest and wetland in Sweden

Emission of isoprene and carbonyl compounds from

a boreal forest and wetland in Sweden

Robert Jansona,*, Claes De Servesb, Rodrigo Romerob

aDepartment of Meteorology, Stockholm University, S-106 91 Stockholm, SwedenbAir Pollution Laboratory, Institute for Applied Environmental Research, Stockholm University, S-106 91, Stockholm, Sweden

Abstract

Emission measurements of light hydrocarbons, light carbonyl compounds, and monoterpenes have been made with canister,

DNPH, and Tenax samples, respectively, on Norway spruce (Picea abies), Scots pine (Pinus sylvestris), the ¯oor of a mixed

pine and spruce forest, and a Sphagnum fen at the NOPEX site (Northern hemisphere climate Processes and land-surface

Experiment) in the southern boreal zone of Sweden, in 1995. The branch and ground measurements were made with the ¯ow-

through enclosure technique and static chamber technique, respectively. Norway spruce was found to emit signi®cant amounts

of isoprene, the normalised emission rate (308C, 1000 mmol mÿ2 sÿ1) averaging 24 � 18 nmol gdwÿ1 hÿ1

(1.4 mgC gdwÿ1 hÿ1), and of carbonyls with normalised rates ranging from 10 to 150 nmol gdwÿ1 hÿ1 (0.3±

4.6 mgC gdwÿ1 hÿ1). Acetone/acrolein and acetaldehyde dominated the carbonyl ¯ux with 61 and 27%, respectively. The

normalised monoterpene emission rate (308C) varied from 17 to 60 nmol gdwÿ1 hÿ1 (2±7 mgC gdwÿ1 hÿ1), with a-pinene

accounting for 34% of the ¯ux. The emission from Scots pine included only traces of isoprene, while the emission rate of

acetone/acrolein was comparable to that of the monoterpenes. The BVOC ¯ux from the forest ¯oor made up only a few

percent of the total forest ¯ux and included ethene and propane at several tens of nmol mÿ2 hÿ1, and the monoterpenes at rates

reaching 380 nmol mÿ2 hÿ1 (50 mgC mÿ2 hÿ1), dominated by a-pinene. A Sphagnum fen emitted isoprene at rates fully

comparable to the areal ¯ux of isoprene from the boreal spruce forest. Highest emission rates were observed from the low and

wet micro-sites, as compared to the higher and drier hummocks. The average ¯ux in June was 912 � 750 nmol mÿ2 hÿ1

(55 � 45 mgC mÿ2 hÿ1) and in August 6800 � 4000 nmol mÿ2 hÿ1 (408 � 240 mgC mÿ2 hÿ1). Monoterpene ¯uxes were

160 � 80 nmol mÿ2 hÿ1 (19 � 9 mgC mÿ2 hÿ1) in June and 760 � 480 nmol mÿ2 hÿ1 (90 � 60 mgC mÿ2 hÿ1) in August.

Isoprene from Norway spruce and Sphagnum wetlands, as well as acetone/acrolein from Norway spruce and Scots pine, are

shown here to be important components of the boreal emission of BVOC. More diurnal and seasonal data is needed to

correctly evaluate the seasonal ¯ux. # 1999 Elsevier Science B.V. All rights reserved.

Keywords: Isoprene; Carbonyl; Monoterpenes; Spruce; Pine; Wetland; Emissions

1. Introduction

The biogenic volatile organic compounds (BVOC)

consist of hydrocarbons and oxygenated hydrocarbons

released to the atmosphere from natural sources,

mainly vegetation. They have been the subject of

study by atmospheric chemists for 2±3 decades, par-

ticularly isoprene (C5H8) and the monoterpenes

(C10H16), because of their in¯uence on the chemistry

of the lower atmosphere, the troposphere (e.g., Feh-

Agricultural and Forest Meteorology 98±99 (1999) 671±681

* Corresponding author.

0168-1923/99/$ ± see front matter # 1999 Elsevier Science B.V. All rights reserved.

PII: S 0 1 6 8 - 1 9 2 3 ( 9 9 ) 0 0 1 3 4 - 3

senfeld et al., 1992). Estimates of the global ¯ux are on

the order of 1000 Tg C per year, methane excluded

(Guenther et al., 1995). In the atmosphere, the BVOC

are oxidised by the OH radical and some by ozone,

thus in¯uencing the oxidative capacity of the tropo-

sphere and affecting the atmospheric lifetimes of other

gases. They contribute to the production of tropo-

spheric ozone and to the carbon monoxide budget

as well as produce organic acids which in turn can

acidify the atmospheric aerosol and precipitation.

They can also contribute to aerosol formation with

possible implications for climate. As the natural VOC

are biogenic, there is an obvious relation and possible

feedback between climate change and both the com-

position and magnitude of BVOC ¯uxes.

The general rule of thumb has been that conifers

emit mainly monoterpenes while deciduous species

emit mainly isoprene. The boreal zone being a con-

iferous zone, its contribution to atmospheric chemistry

has been thought to be mainly by the emission of

monoterpenes, and the few studies to date have also

concerned monoterpenes (Isidorov et al., 1985; Jan-

son, 1992, 1993). A number of recent works have

demonstrated, however, that some conifers, in parti-

cular Norway spruce and Sikta spruce, emit signi®cant

amounts of isoprene (Steinbrecher and Rabong, 1994;

Street et al., 1996; Steinbrecher et al., 1997). The work

of Kesselmeier et al., 1997 and Steinbrecher et al.

(1993) have also shown that carbonyl compounds such

as organic acids and aldehydes are emitted at signi®-

cant rates by some conifers. Indeed, recent model

calculations of regional and global BVOC ¯uxes by

Simpson et al. (1995) and Guenther et al. (1995) have

estimated the oxygenated species, primarily alde-

hydes, ketones, organic acids, and alcohols, to be a

considerable part of the total ¯ux, although there exist

very little experimental data. As for the boreal zone,

no measurements of isoprene or carbonyl compound

emissions from Norway spruce or Scots pine, the most

common coniferous species of the European boreal

zone, have previously been done.

The forest ¯oor, i.e. soil and ®eld vegetation, also

emits VOC to the atmosphere, although the few

measurements to date are not conclusive as to the

signi®cance of these ¯uxes (Isidorov et al., 1985;

Janson, 1993; Steinbrecher et al., 1997). Lastly, while

boreal wetlands have been of interest for CH4 and CO2

budgets, only Klinger et al. (1994) have reported

measurements of isoprene and monoterpene emissions

from that ecosystem in Canada.

Here we report the ®rst measurements of isoprene

and carbonyl compound emissions, alongside mea-

surements of monoterpene emissions, from a conifer-

ous forest and a wetland in the southern boreal zone of

northern Europe. The purpose of the measurements

was to identify the natural sources of these compounds

in the European boreal zone, and to evaluate the

relative importance of those sources.

2. Methods

2.1. Sampling site

The forest experiments were performed 300 m from

the Central Tower of the NOPEX site (Northern hemi-

sphere climate Processes and land-surface Experi-

ment) at 608050N and 178300E in the southern

boreal zone in Sweden (Halldin et al., 1999). The

forest was one of mixed Norway spruce (Picea abies)

and Scots pine (Pinus sylvestris), about 50 years of age

and 20 m high. The soil is podsolic on moraine and is

covered with mosses (mainly Hylocomium spendens

and Pleurzium schreberi) and areas of blueberry

(Vaccinium myrtillus). The wetland measurements

were made on Ryggmossen, a Sphagnum fen of about

30 ha. The central parts of the fen are open with

occasional stunted pine, which at the fringes become

taller and more dominant. The fen is surrounded

mainly by Scots pine forest.

2.2. Branch emissions

Emission measurements were made on sun-side

branches of Norway spruce and Scots pine at 15 m.

height with the dynamic branch enclosure method,

Fig. 1. The chamber was made of 0.05 mm transparent

FEP-Te¯on ®lm which is tied around a 20±30 cm

branch segment (Janson, 1993). Care was taken to

avoid contact between the branch and the walls of the

chamber in order to avoid mechanical abrasion which

can temporarily affect emission rates (Juuti et al.,

1990). The volume was about 10 l and a fan ensured

mixing. A metal bellows pump supplied the chamber

with a ¯ow of ambient air maintained at 6 l minÿ1 by a

mass ¯ow controller. The water content of the inlet

672 R. Janson et al. / Agricultural and Forest Meteorology 98±99 (1999) 671±681

air¯ow was reduced with a Peltier cooling element in

order to avoid condensation in the mass ¯ow controller

and to avoid excessive humidity in the chamber. A KI

annular denuder removed ozone, which is suspected of

interfering with Tenax, canister, and DNPH sampling

(Roberts et al., 1983; Janson and Kristensson, 1991;

Hoffmann et al., 1993). Ambient and chamber tem-

peratures and relative humidities were measured con-

tinuously during all experiments with Rotronic MP-

100 sensors. PAR was measured with a LiCOR LI-

190SA quantum sensor mounted above the chamber.

Measurements for light hydrocarbons, the mono-

terpenes, and light carbonyl compounds were taken

consecutively, a full sampling series taking about 3 h.

For each emission measurement, simultaneous con-

centration samples were taken from the inlet and outlet

air¯ows. The branch emission is determined from the

difference between the concentrations at the outlet

(cout) and inlet (cin) times the air¯ow through the

chamber (Fch) and normalized to the branch needle

dry weight

En � �coutÿcin� � Fch gdwÿ1

The monoterpenes were sampled with stainless

steel tubes containing 200 mg Tenax TA at an air ¯ow

rate of 100 ml minÿ1 for 12 min. For the branch

measurements, the ¯ows were monitored with Hon-

eywell microbridge mass air¯ow sensors and for the

wetland studies they were checked against a calibrated

rotometer, noting also the air temperature and atmo-

spheric pressure. After sampling, the tubes were stored

dry until analysis by GC-FID. Compound identi®ca-

tion and quanti®cation was accomplished with terpene

standards in methanol. The standards included tricy-

clene, a-pinene, camphene, sabinene, b-pinene, myr-

cene, �3-carene, limonene, and trans-ocimene.

The light carbonyl compounds were sampled on

SepPakTM DNPH cartridges at a ¯ow rate of

0.8 l minÿ1 for 2 h. All samples were stored at 8±

108C. In the laboratory, they were eluated with acet-

onitrile, and analysed on HPLC. Acetone and acrolein

(propenal) could not be separated by our method

(Microspher C18 4.6 � 100 mm column) and the

compound has, therefore, been identi®ed as acet-

one/acrolein in our results.

The light hydrocarbons were sampled with whole

air samples taken in 3 l elektropolished stainless steel

canisters. The canisters are evacuated to 10ÿ6 bar at

1508C for 6 h, which gives a non detectable isoprene

blank signal when ®lled with isoprene free gas. The

sampling ¯ow rate was regulated to 0.8 l minÿ1 and

the canister ®lled to a ®nal overpressure of 2 bars with

an TFE-Te¯on membrane pump (BRC, Model FC-

1121). Sampling takes 10±15 min. The samples were

stored at room temperature and analysed by GC-FID.

Identi®cation was done with authentic standards and

Fig. 1. Experimental setup for the branch enclosure experiments. The dryer is a Peltier element, the O3 trap a KI annular denuder, and the

pump for the inlet airflow a metal bellows pump. All tubing is Teflon 1/400. Air inflow rate is 6 l minÿ1.

R. Janson et al. / Agricultural and Forest Meteorology 98±99 (1999) 671±681 673

quanti®cation with a propane standard, for which the

FID carbon response is 5% higher than that for iso-

prene (Apel et al., 1994).

Laboratory studies were performed in order to

evaluate eventual losses or artefacts in measurements

of isoprene from the branch chamber. In addition to

the set up as in Fig. 1, an active charcoal scrubber was

introduced to remove organics from the inlet air¯ow as

well as a humidi®er to regulate air humidity. A

3.0 ppbv isoprene standard gas was introduced to

the chamber at relative humidities of 11, 49, and

81%. Samples were collected consecutively at the

chamber inlet and outlet in stainless steel canisters

at a ¯ow rate of 0.5 l minÿ1. The canisters were ®lled

directly without being ¯ushed with the sample air.

Samples were analysed the same day.

The effect of storage time on the recovery of the

isoprene sampled at different relative humidities was

also investigated. Three groups of 4 canisters each

were ®lled at 14, 49, and 86% relative humidity. The

canisters were stored at room temperature and the

isoprene concentrations determined at 1, 6, 14, 36, and

77 days after sampling.

2.3. Forest floor emissions

Forest ¯oor emissions were measured with the static

chamber technique in June, July, and September. Two

stainless steel frames, 60 cm � 60 cm and 20 cm high,

lined on the inside with Te¯on ®lm, were placed on the

ground about 10 m apart at the beginning of the

season. Plot F1 was covered with mosses (Pleurzium

schreberi), while Plot F2 was covered with mosses

(Hylocomium splendens), 15±30 cm high wild blue-

berry (Vaccinium myrtillus), and a few grasses and

herbs (Melampyrum sylvaticum). Both plots had pine

and spruce trees within a 10 m radius. The bottom

edge of the frame was pressed a few centimeters into

the underlying mosses a week in advance of the ®rst

measurements. Depressing the frame into the soil by

cutting a furrow has previously been found to sig-

ni®cantly alter the emission pattern of the forest ¯oor

(Janson, 1993). Care was, therefore, taken not to break

or disturb the roots of the ground vegetation, i.e.

blueberry bushes, and surrounding trees.

The chamber was a 60 cm � 60 cm � 50 cm stain-

less steel frame covered on the inside with transparent

Te¯on ®lm. The chamber volume was about 235 l

(250 l empty) and the area covered was 0.36 m2. A fan

mounted from the top of the chamber with the motor

situated on the outside ensured mixing of the chamber

air. Chamber temperature and relative humidity were

measured with a Rotronic MP-100 sensor, and PAR

with a LiCOR LI-190SA sensor situated on top of the

chamber. At the time of a measurement, the chamber

was placed in a small gutter at the top of the ground

frame, lined with foamed plastic covered with Te¯on

®lm. The VOC ¯ux was determined from the differ-

ence in concentration between the chamber air

sampled after 20 min, and the ambient air, sampled

at the beginning of an experiment. For the carbonyl

compounds, simultaneous ambient and chamber mea-

surements were made for 1 h from the beginning of the

experiment. The ground plot was left uncovered at

least 30 min between experiments.

The ground chamber was tested for artifacts and

losses by sampling from an empty chamber, i.e. the

bottom frame closed on the bottom with Te¯on ®lm,

under ®eld conditions.

2.4. Sphagnum fen emissions

The above described static chamber technique was

used to measure ¯uxes from a Sphagnum fen. Two

frames were placed about 10 m apart and about 30 m

from the border zone of dwarfed Scots pine. The one,

Plot B1, was placed on a plot of Sphagnum moss at

which the water table was within 10 cm of the surface.

The other, Plot B2 was drier, being on a small hum-

mock a couple decimeters above the level of Plot B1,

and consisted of Sphagnum moss, grasses and heather

(Calluna vulgaris). Emission measurements were made

during one day in June and one day in August 1996.

3. Results

3.1. Laboratory experiments

The difference in isoprene concentration between

outlet and inlet air samples of the controlled laboratory

experiments with the Te¯on chamber, was less than

1% for all three relative humidities. The ®eld samples

taken from the empty branch chamber and ground

chamber also demonstrated neither isoprene nor

monoterpene artefacts. The monoterpene results are


in keeping with previous results reported by Janson

and Kristensson (1991). A branch chamber with a few

polyethylene details (fan, in- and outlet pieces)

showed a formaldehyde artefact and was therefore

not used for emission experiments. After exchanging

the polyethylene parts to Te¯on, the empty chamber

showed no carbonyl artefact.

In the storage experiment, the isoprene concentra-

tion in the canisters showed a 13±14% loss after 77

days of storage in all three groups of relative humid-

ities. Within a 5% analytical uncertainty, the concen-

tration did not change during the ®rst 40 days, with the

exception of the samples taken at 14% relative humid-

ity (RH). This is in good agreement with the results of

Pate et al. (1992) who investigated the stability of 10

polar organic compounds in Summa-polished and

unpolished stainless steel canisters at 3 bars initial

pressure with and without addition of 200 ml water.

They reported an excellent recovery of isoprene

(100 � 10%) from humid canisters over a 31-day

storage period, whereas the dry canisters showed a

gradual loss of more than 70% of the initial 6.1 ppbv

isoprene concentration. This severe loss of isoprene in

the dry canisters observed by Pate et al. (1992) was,

however, not reproduced in our study. Good recovery

of isoprene from humid canisters has also been

reported by Brymer et al. (1996) who examined

storage stability for 194 VOC in Summa-polished

stainless steel canisters at 2 bars and 70% RH. After

30 days, the isoprene concentration showed a loss of

7% to the initial which, again, is in very good accor-

dance with our results.

It has been suggested that the improved recovery of

many VOC, including isoprene, from humidi®ed can-

isters may be due to passivisation of active sites at the

canister surface by water (Gholson et al., 1990; Pate

et al., 1992). It has also been argued that high canister

pressure is bene®cial for recovery of many VOC

(Coutant, 1992). It is, therefore, within reason to

believe that the recovery of isoprene during normal

procedures, i.e. storage without loss of canister pres-

sure, would be even better than what our results from

this experiment indicate. We have taken samples for

analysis from the same canister 4-times prior to the

last analysis at Day 77. For each analysis, the canister

pressure decreases and was as low as 0.7 bar at Day

77. This argument ®nds support in the results of two

canisters (49% RH) that were saved untouched until

the end of the storage period. At Day 77, these two

canisters had an isoprene concentration that was

within 5% of the initial concentration, which can be

compared to the 13% loss in the 49% RH canister

group that had been analyzed 4-times.

3.2. Branch emissions

3.2.1. Norway spruce

Afternoon emission rates were measured from four

branches of Norway spruce on 20 May, 2, 4 and 20

June; 8 and 9 July, and 20 September, 1995, Fig. 2, and

Fig. 2. Afternoon emission rates (nmol gdwÿ1 hÿ1) of isoprene, monoterpenes and carbonyls from Norway spruce (Picea Abies) at the

Norunda site, 1995. Given are the mid-time, average PAR (mmol mÿ2 sÿ1), and average chamber air temperature (8C) for each measurement.


the daily variation of the ¯ux on 2 and 20 June, and 20

September, Fig. 3. The results from 2 June must be

taken with some caution as slight damage to the tip of

one side branch was observed after the experiment.

Mechanical abrasion is known to affect monoterpene

emissions, but it is unsure how it will affect the rates of

isoprene or carbonyl compounds.

Isoprene was found to be emitted at rates varying

from less than 5 in May and September to

32 nmol gdwÿ1 hÿ1 (0.06±2 mgC gdwÿ1 hÿ1) on 8

July. Traces of ethene, propane and propene were also

observed in June, generally at the pmol level except

for on 20 June, when emissions of ethene were as high

as several nmol gdwÿ1 hÿ1.

The emission of carbonyls consisted of formalde-

hyde, acetaldehyde, and acetone/acrolein at relative

rates of 10 � 16, 26 � 19, 64 � 20%, respectively.

The highest observed ¯ux was 22 nmol gdwÿ1 hÿ1

(0.67 mgC gdwÿ1 hÿ1) on 20 June.

The ¯ux of monoterpenes varied from 1 to

40 nmol gdwÿ1 hÿ1 (0.1±5 mgC gdwÿ1 hÿ1) over a

temperature and PAR range of 9±338C and 0±

1610 mE mÿ2 sÿ1, respectively. The ¯ux was domi-

nated by a-pinene, accounting on the average for

34 � 10% of the total monoterpene ¯ux, Table 1. This

is about half the percentage previously observed for

Norway spruce in another Swedish forest (Janson,

1993), and slightly lower than that reported for the

Bavarian Forest in Germany by Steinbrecher et al.

(1997). The other terpenes of importance in the ¯uxes

were myrcene, b-pinene, �3-carene, and limonene.

It is of interest to compare the emission of isoprene

and the carbonyl compounds with that of the mono-

terpenes at standard conditions, as well as to compare

the results of one period to the other. To this end, the

emission rates for each set of measurements are

normalised in accordance with current emission rate

algorithms. The isoprene data can be normalised to

temperature and photosynthetic active radiation (PAR)

according to the procedure outlined by Guenther et al.

(1995)

Is � I

�CL � CT�where Is is the isoprene emission factor at a standard

temperature and PAR flux (usually 308C and

1000 mE mÿ2 sÿ1), I is the emission rate at tempera-

ture T and PAR flux L, CL is the light dependence

factor, and CT the temperature dependence factor. The

monoterpene emissions are usually fairly well

described, at least on the short time scale, by a

temperature algorithm (e.g., Lamb et al., 1987;

Janson, 1993, and Guenther et al., 1995)

Ms � M

exp ��TÿTs��

Fig. 3. Daytime variation of the isoprene, monoterpene, and

carbonyl emission rates (nmol gdwÿ1 hÿ1) from Norway spruce

(Picca abies) at the Norunda site en 2 and 20 June, and 20

September 1995.

Table 1

The average relative composition of the monoterpene emission

from Norway spruce in a mixed spruce and pine forest at the

Norunda site in June, July, and September

a-pin cam myr b-pin �3-c lim

Average 34% 7% 24% 21% 14% 17%

SD 10% 6% 16% 13% 3% 6%


where Ms is the emission rate at a standard tempera-

ture Ts, usually chosen to be 308C, M is the mono-

terpene emission rate at temperature T, and � is an

empirical coefficient. Here we use � � 0.09, in accor-

dance with the discussion by Guenther et al. (1995)

who found that most of the estimates reported in the

literature fall within the range 0.09 � 0.025 Kÿ1.

Lastly, in light of the fact that very little data has

been reported on the emission of carbonyl compounds

and the factors which regulate those emissions, we

will tentatively apply the isoprene emission algorithm

also for those emissions.

The normalised data include 3±4 daytime

(PAR>100 mE mÿ2 sÿ1) measurements for each

period and are shown in Table 2. It is seen that, on

a molar basis, the isoprene and carbonyl emissions

are entirely comparable to and sometimes higher

than the emission rate of the monoterpenes. For

isoprene, the emission algorithm seems to describe

emission rate variability well, both diurnally and

over the season, with the exception of the early June

measurements. Averaging all the data yields a

standard emission factor of 24 � 18 nmol gdwÿ1 hÿ1,

(1.4 mgC gdwÿ1 hÿ1), which agrees quite well with

results reported earlier for Norway spruce in Germany

(Steinbrecher and Rabong, 1994). For both the mono-

terpenes and the carbonyl compounds, the standard

emission rates vary considerably over the summer

(Table 2).

In the afternoon of 9 July, branches from two

adjacent spruce trees were sampled within a 3 h

period, Fig. 4. It can be seen that the relative composi-

tion of the total emission was very similar, 0.25/0.40/

0.35 for both branches, while the absolute emission

rates of Branch 2 were 30±40% lower than those of

Branch 1.

3.2.2. Scots pine

Nine emission measurements were made from

branches of Scots pine on 18±19 August and 6±7

September 1995. With emission rates all below

1 nmol gdwÿ1 hÿ1 (0.06 mgC gdwÿ1 hÿ1), isoprene

was not an important part of the emission, while

the carbonyl compounds were emitted at rates com-

parable, on a molar basis, to those of the monoter-

penes, Fig. 5. Daytime rates ranged from 3 to

6 nmol gdwÿ1 hÿ1 (0.1±0.2 mgC gdwÿ1 hÿ1) and 2

to 14 nmol gdwÿ1 hÿ1 (0.2±1.7 mgC gdwÿ1 hÿ1), for

the carbonyl compounds and monoterpenes, respec-

tively. As with Norway spruce, the carbonyls were

dominated by acetone/acrolein, accounting for

79 � 13% of the total carbonyl emission. The rates

Table 2

Average normalised emission rates � 1 SD (number of samples)

for daytime measurements from Norway spruce, 1995

Isoprene

(nmol

gdwÿ1 hÿ1)

Terpenes

(nmol

gdwÿ1 hÿ1)

Carbonyls

(nmol

gdwÿ1 hÿ1)

2 and 4 33 � 27 (4) 17 � 8 (5)

20 June 16 � 1 (3) 38 � 4 (3) 42 � 31 (4)

8 and 9 July 22 � 6 (2) 17 � 2 (3) 11 � 3 (3)

20 September 23 � 2 (2) 61 � 45 (4) 148 � 53 (2)

Fig. 4. Emission rates (nmol gdwÿ1 hÿ1) of isoprene, monoterpenes, and carbonyls from sun branches of two adjacent trees, measured at the

same height. Measurements were taken on 9 July 1995, between 13:30 and 16:30 hours. Given is the average values for PAR (mmol m2 sÿ1),

and chamber air temperature, as well as the mid-time for each measurement.


show poor correlation to temperature and the normal-

ised rates a large scatter, 50 � 47 nmol gdwÿ1 hÿ1.

The terpenes were dominated by a-pinene and �3-

carene at 46 � 20 and 33 � 6%, respectively, and the

normalised rate was 6.3 � 3 nmol gdwÿ1 hÿ1

(0.8 � 0.4 mgC gdwÿ1 hÿ1). Both the percentual com-

position and the standardised rate for the monoterpene

emissions are in good agreement with earlier results

(Janson, 1993).

3.3. Forest floor emissions

Six daytime measurements were made from Plot F2

(mosses and blueberry) and three from Plot F1

(mosses) on 28±30 June, 7 July, and 19±20 September.

At the start of an experiment, the relative humidity

inside the chamber would increase quickly to 80±90%,

while the temperature usually remained the same as

ambient temperature. PAR was usually low, a few

100 mE or less due to the effects of canopy shading. In

a separate experiment, ozone was found to disappear

by dry deposition within 1 min of chamber closure.

No emission of isoprene was observed from either

of the two ground plots, while a small emission

of some light hydrocarbons, particularly ethene and

propane at several 10 nmol mÿ2 hÿ1, was observed,

Fig. 6. The total monoterpene ¯ux varied up to

380 nmol mÿ2 hÿ1 (50 mgC mÿ2 hÿ1), and was domi-

nated by a-pinene. The terpene emissions from Plot

F1 were considerably higher than from Plot F2 in June

and July, but not in September. The highest ¯ux was

observed from the all moss plot (F1) on 30 June. The

terpene ¯ux at this site is considerably lower than the

¯ux from a pine forest ¯oor reported earlier (Janson,

1993). Indeed, assuming a needle biomass density of

about 1000 gdw mÿ2 for this forest, the combined

terpene and light hydrocarbon ¯ux from the forest

¯oor account for only a few percent of the total

daytime ¯ux of BVOC from this forest in June and

July.

3.4. Sphagnum fen emissions

During all of the wetland experiments, condensa-

tion of water on the walls of the chamber would begin

about 10 min after the start of the experiment. The

Fig. 5. Afternoon emission rates (nmol gdwÿ1 hÿ1) of isoprene, monoterpenes and carbonyls from Scots pine (Pinus sylvestris) at the

Norunda site, 1995. Given are the mid-time, average PAR (mmol mÿ2 sÿ1), and average chamber temperature (8C) for each measurement.

Fig. 6. Average daytime forest floor emission rates

(nmol mÿ2 hÿ1) of ethene, propane, and monoterpenes in a mature

forest of Norway spruce and Scots pine at the Norunda site. The

error bars are �1 SD for three measurements from Plot FI, mainly

mosses, and from six measurements from Plot F2, mosses and

blueberry, see text. Air temperatures ranged from 258C in June and

July to 128C in September.


method is, therefore, not appropriate for water soluble

gases like the carbonyl compounds, while chamber

losses are likely to be small for less water soluble

gases like isoprene and the monoterpenes.

The weather in June was partly cloudy with air

temperatures between 15 and 188C, and ground tem-

peratures (at 5 cm) ranging from 9.58C in the morning

to a maximum of 13.58C at 17:30 hours in the after-

noon. In August it was sunny with air temperatures

around 268C and ground temperatures between 13.58Cin the morning and 17.58C at 18:00 hours.

Isoprene was found to be emitted at rates ranging

from several hundred to about 12000 nmol mÿ2 hÿ1

(720 mgC mÿ2 hÿ1) and clearly higher than the rates

of the monoterpene emission, which at the most was

1200 nmol mÿ2 hÿ1 (144 mgC mÿ2 hÿ1) in August,

Fig. 7 (a, b). With the exception of the morning

experiment of 21 August, the isoprene emission rate

was higher from the wet plot R1 (Sphagnum moss) as

compared to the drier plot R2 (mosses and heather).

The average isoprene emission rate for both the wet

and dry micro-sites was 912 � 750 nmol mÿ2 hÿ1 in

June and 6757 � 3924 nmol mÿ2 hÿ1 in August,

while for the monoterpenes the average was

161 � 79 and 755 � 478 nmol mÿ2 hÿ1, respectively.

Beta-pinene comprised 32 � 11% of the ¯ux, b-phel-

landrene 25 � 19%, and limonene 20 � 8%, while

sabinene, myrcene, and a-pinene made up less than

10% each. While the total ¯ux of monoterpenes

observed in this study is similar to the ¯ux reported

for bogs and fens by Klinger et al. (1994), the isoprene

¯ux was at times considerably higher.

Using the isoprene emission rates reported above,

we ®nd that a spruce forest with a needle biomass

density on the order of 1000 gdw mÿ2 emits about

6000 nmol isoprene mÿ2 hÿ1 at 208C and

Fig. 7. Emission rates (nmol mÿ2 hÿ1) of isoprene and monoterpenes from a wet plot of Sphagnummoss (Plot R1) and a relative dry hummock

of Sphagnum and heather (Plot R2), on a Sphagnum fen (Ryggmossen), 14 June and 21 August 1996. Given are the times and ground

temperatures (5 cm) for each measurement.


1000 mmol mÿ1 sÿ1. It is thus obvious that the iso-

prene ¯uxes from Sphagnum fens can be as high as, if

not higher than the isoprene ¯uxes from forests of

Norway spruce. The implications this has for the total

isoprene and BVOC ¯ux from the boreal zone depends

on the diurnality and seasonality of the emission rates,

neither of which have yet been resolved, as well as the

total area of Sphagnum wetland and percentage iso-

prene emitting micro-sites.

4. Conclusions

Isoprene and acetone/acrolein (64 � 20% of the

total carbonyl ¯ux) have been found to make up a

major part of the BVOC emission from Norway spruce

(Picea abies) forests in the boreal zone. Acetaldehyde

and formaldehyde make up another 26 � 19 and

10 � 16% of the carbonyl ¯ux, respectively. The

Guenther et al. (1995) algorithm describes the diurnal

and seasonal variation of the isoprene emission rate

fairly well and yields a standard emission rate (308C,

1000 mE mÿ2 sÿ1) of 24 � 18 nmol gdwÿ1 hÿ1

(1.4 mgC gdwÿ1 hÿ1). This agrees well with the rates

reported for Norway spruce in Germany. Using the

same algorithm to normalise the rates of carbonyl

compound emissions, results in a large scatter for this

data set. The normalised rates range from 10 to

150 nmol gdwÿ1 hÿ1 (0.3±4.6 mgC gdwÿ1 hÿ1). Also

the normalised rates of the monoterpene emission

show a scatter, ranging from 17 to 60 nmol gdwÿ1 hÿ1

(2±7 mgC gdwÿ1 hÿ1). More data is needed in order to

improve the emission algorithm for these groups of

compounds.

Scots pine (Pinus sylvestris) is con®rmed here to be

a non-isoprene emitter. It does emit signi®cant quan-

tities of acetone/acrolein, which make up 79 � 13% of

the total carbonyl ¯ux, The other carbonyls are for-

maldehyde (10 � 12%) and acetaldehyde (11 � 8%).

The emission rate of monoterpenes observed here

agree well with earlier reports.

The BVOC ¯ux from the ¯oor of a mixed forest of

Norway spruce and Scots pine consists mainly of

monoterpenes, traces of ethene and propane, no iso-

prene, and comprises only a few percent of the total

forest BVOC ¯ux.

Isoprene is emitted from Sphagnum wetlands at

rates comparable to or greater than the isoprene ¯ux

from boreal spruce forests. More data is needed con-

cerning the spatial and temporal variability of these

¯uxes in order to ascertain their signi®cance to the

total BVOC ¯ux from the boreal zone. The wetlands

also emit monoterpenes, but the ¯ux observed in this

study was less than 10% of the areal ¯ux from a boreal

spruce forest.

This study con®rms that isoprene from Norway

spruce and Sphagnum wetlands, as well as acetone/

acrolein from Norway spruce and Scots pine, are

important components of the boreal emission of

BVOC.

Acknowledgements

We thank Eric Kellner, the NOPEX Central Of®ce,

and Leif BaÈcklin of MISU for all their willingful

help and co-operation in the ®eld. This project

was funded by the Swedish National Board of

Environmental Protection and The Environmental

Foundation of the Swedish Association of Graduate

Engineers.

References

Apel, E.C., Calvert, J.G., Fehsenfeld, F.C., 1994. The nonmethane

hydrocarbon intercomparison experiment (NOMHICE) Tasks 1

and 2. J. Geophys. Res. 99, 16651±16664.

Brymer, D.A., Ogle, L.D., Jones, C.J., Lewis, D.L., 1996. Viability

of using Summa-polished canisters for the collection and

storage of parts per billion by volume level volatile organics.

Environ. Sci. Technol. 30, 188±195.

Coutant, R.W., 1992. Theoretical evaluation of stability of volatile

organic chemicals and polar volatile organic chemicals in

canisters. EPA/600/R-92/055, Research Triangle Park, NC,

USA.

Fehsenfeld, F., Calvert, J., Fall, R., Goldan, P., Guenther, A.,

Hewitt, C.N., Lamb, B., Liu, S., Trainer, M., Westbert, H.,

Zimmerman, P., 1992. Emissions of volatile organic com-

pounds from vegetation and the implications for atmospheric

chemistry. Global Biogeochemical Cycles 6, 389±430.

Gholson, A.R., Jayanty, R.K.M., Storm, J.F., 1990. Evaluation of

aluminum canisters for the collection and storage of air toxics.

Anal. Chem. 62, 1899±1902.

Guenther, A., Hewitt, C.N., Erickson, D., Fall, R., Geron, C.,

Graedel, T., Harley, P., Klinger, L., Lerdau, M., McKay, W.A.,

Pierce, T., Scholes, B., Steinbrecher, R., Tallamraju, R., Taylor,

J., Zimmerman, P., 1995. A global model of natural volatile

organic compound emissions. J. Geophys. Res. 100, 8873±

8892.


Halldin, S., Gottschalk, L., Gryning, S.E., Jochum, A., Lundin,

L.C., Van de Griend, A.A., 1999, Energy, water and carbon

exchange in a boreal forest ± NOPEX experiences. Agric. For.

Meteorol. 98±99, 5±29.

Hoffmann, T., Jacob, P., Linscheid, M., Klockow, D., 1993.

Measurements of biogenic hydrocarbons and their atmospheric

degradation in forest.. Intern J. Environ. Anal. Chem. 52,

29±37.

Isidorov, V., Zenkevich, I., Ioffe, B., 1985. Volatile organic

compounds in the atmosphere of forests. Atmos. Environ. 19,

1±8.

Janson, R., Kristensson, J., 1991. Sampling and analysis of

atmospheric monoterpenes, Report CM-79. Department of

Meteorology, Stockholm University, Stockhlm, Sweden.

Janson, R., 1992. Monoterpene concentrations in and above a forest

of Scots pine. J. Atm. Chem. 14, 385±394.

Janson, R., 1993. Monoterpene emissions from Scots pine and

Norwegian spruce. J. Geophys. Res. 98, 2839±2850.

Juuti, S., Arey, J., Atkinson, R., 1990. Monoterpene emission rate

measurements from a monterey pine. J. Geophys. Res. 95,

7515±7519.

Kesselmeier, J., Bode, K., Hofmann, U., MuÈller, H., SchaÈfer, L.,

Wolf, A., Ciccioli, P., Brancaleoni, E., Cecinato, A., Frattoni,

M., Foster, P., Ferrari, C., Jacob, V., Fugit, J.L., Dutaur, L.,

Simon, V., Torres, L., 1997. The BEMA project: emission of

short chained organic acids, aldehydes and monoterpenes from

Quercus ilex L. Pinus pinea L. in relation to physiological

activities, carbon budget and emission algorithms. Atmos.

Environ., 31, 119±133.

Klinger, L.F., Zimmerman, P.R., Greenberg, J.P., Heidt, L.E.,

Guenther, A.B., 1994. Carbon trace gas fluxes along a

successional gradient in the Hudson Bay lowland. J. Geophys.

Res. 99, 1469±1494.

Lamb, B., Guenther, A., Gay, D., Westberg, H., 1987. A national

inventory of biogenic hydrocarbon emissions. Atmos. Environ.

21, 1695±1705.

Pate, B., Jayanty, R.K.M., Peterson, M.R., Evans, G.F., 1992.

Temporal stability of polar organic compounds in stainless steel

canisters. J. Air Waste Manage. Assoc. 42, 460±462.

Roberts, J., Fehsenfeld, F., Albritton, D., Sievers, F., 1983.

Measurement of monoterpene hydrocarbons at Niwot Ridge.

Colorado. J. Geophys. Res. 88, 10667±10678.

Simpson, D., Guenther, A., Hewitt, C.N., Steinbrecher, R., 1995.

Biogenic emissions in Europe 1. Estimates and uncertainties. J.

Geophys. Res. 100, 22875±22890.

Steinbrecher, R., Rabong, R., 1994. Emission of terpenoids and

C2±C6 hydrocarbons by Norway Spruce: Phytotron experi-

ments and field studies. In: Borrell, P.M., P. Borrell, T. CvitasÏ,

W. Swiler (Eds.), Proceedings of the Symposium on EURO-

TRAC, 1994. SPB Academic Publishers, The Hague, The

Netherlands, pp. 502±505.

Steinbrecher, R., Jocher, M., Wiedemann, M., Slemr, J., Hahn, J.,

1993. Low molecular weight compounds above a natural

Norway spruce forest, In: Borrell, P.M. (Ed.), Proceedings of

the Symposium on EUROTRAC 1992. SPB Academic Publish-

ers, The Hague, The Netherlands, pp. 299±302.

Steinbrecher, R., Ziegler, H., EichstaÈdter, G., Fehsenfeld, U.,

Gabriel, R., Kolb, C., Rabong, R., SchoÈnwitz, R., SchuÈrmann,

W., 1997. Monoterpene and isoprene emission in Norway

spruce forests. In: Slanina, S. (Ed.), Biosphere-Atmosphere

Exchange of Pollutants and Trace Substances. Springer±Verlag,

Berlin, pp. 352±365.

Street, R.A., Duckham, S.C., Hewitt, C.N., 1996. Laboratory and

field studies of biogenic volatile organic compounds from Sitka

spruce (Picea sitchensis Bong.) in the United Kingdom. J.

Geophys. Res. 101, 22799±22806.


Emission of isoprene and carbonyl compounds from a boreal forest and wetland in Sweden

Documents