ISOPRENOID FLUXES AND PHOTOSYNTHETIZED CARBON MEASURED OVER THE TROPICAL RAINFOREST NEAR MANAUS DURI

1
ISOPRENOID FLUXES AND PHOTOSYNTHETIZED CARBON
MEASURED OVER THE TROPICAL RAINFOREST NEAR MANAUS
DURING THE DRY SEASON 2001 

•  
• P. Stefani 1, A.C. de Araujo2, A. D. Nobre 2, P.
  Ciccioli 3 , E. Brancaleoni 3, M. Frattoni 3, U.
  Kuhn4 , J. Kesselmeier4 , T. Dindorf 4, C.
  Corradi 1, R. Valentini 1
• 1-Dipartimento di Scienze dellAmbiente Forestale
  e delle sue Risorse, Università della Tuscia,
  Viterbo ITALY
• 2 -Instituto Nacional de Pesquisas da Amazonia,
  Manaus, Amazonas, BRASIL
• 3-Istituto sullInquinamento Atmosferico del
  CNR,Monterotondo Scalo ITALY
• 4-Max Planck Institute for Chemistry,
  Biogeochemistry Dept.,Mainz GERMANY

2
Tasks in LBA

• CARBONSINK / CARBONCYCLE Project
• To quantify the amount of reduced carbon emission
  in relation to the Net Ecosystem Exchange (NEE)
  and the Gross Primary Productivity of CO2 (GPP)
  in the tropical rainforest of Manaus.

3
Within this project, particular emphasis was
given to seasonality aspects of isoprenoid
emissions in order to relate observed changes
with those occurring in the hydrogeochemical
carbon cycle of the area investigated. In this
context, the main focus was to quantify, and
possibly describe, fluxes of isoprenoids and
their seasonal variations, through suitable
algorithms. This work was essentially made with
the research groups working in CO2 exchange.
4

• CLAIRE 2001 Project
• to measure the emission and deposition fluxes of
  VOC (both biogenic and anthropogenic) under
  different oxidizing conditions to assess the role
  of biogenic VOC in the formation of photochemical
  oxidants and secondary organic aerosols.

5
In this project, data should have been collected
in an intensive campaign to be held in July 2001
(dry season) and Isoprenoid fluxes determined in
parallel with those of other precursors and
products of photochemical smog formation.
Ground based data were complemented with
airborne and boat measurements. The results of
airborne flights will be presented here by other
participants to the same campaign.
6
THE LOCATION OF THE TOWER
TOWERS
C14
K34
MANAUS
7
FLUX DETERMINATIONS WERE PERFORMED USING THE REA
TECHNIQUE
8
THE CANOPY STRUCTURE
9
A REVERSED GEOMETRY WAS USED TO COLLECT THE UP
AND DOWN SAMPLES ON ADSORPTION TRAPS
Dummy
Up
Down
PC for data acquisition and valves actuation
Aspirating pump
10
1 Isoprene 2 a-Pinene 3 Sabinene 4
b-Pinene 5 Myrcene 6 Limonene 7
1,8-Cineole
11
RESULTS OBTAINED 1) NET ISOPRENOID ECOSYSTEM
EXCHANGE (NIEE)
July 2001
20
19
10
11
8
25
18
15
23
27
26
24
17
12
Data collected were quite representative of the
dry season situation because in most of the days
the dominant wind was blowing from the prevalent
direction (East to West) measured in this season.
13
THE LARGEST CONTRIBUTION TO THE NIEE WAS GIVEN BY
ISOPRENE
14
Values of isoprene varied from 50 to 100 of
NIEE. The average contribution was 72. The
remaining 28 was given by monoterpenes.
15

• Data obtained showed that no large variations
  were observed in the maximum values of isoprene
  fluxes measured in the various seasons.
• 3.7 mg C m-2 h-1 were measured at the beginning
  of the wet season (december 1999-january 2000)
• 4.5 mg C m-2 h-1 were measured during the dry
  season (july 2001)
• 2.9 mg C m-2 h-1 were measured during the wet
  season (april 2002)

16
ALTHOUGH SMALL, THE CONTRIBUTION OF
MONOTERPENES WAS MORE THAN TWICE AS THAT
MEASURED IN THE WET SEASON
17
The term NMEE was used here to highlight the fact
that each monoterpene was contributing to a
different extent to the total flux. While some
were emitted, others were deposited.
Particularly different was the behaviour of
a-pinene and limonene. The former started to be
emitted earlier in the morning, the latter was
contributing most to the NIEE in the afternoon.
18
Differences in monoterpene fluxes reflected well
the boundary layer concentrations that were
measured with airborne flights made in the late
morning and in the late afternoon.
Average composition between 200 and 2000 m
19
2) CONSISTENCY OF ISOPRENOID FLUXES WITH THE
PHYSIOLOGICAL RESPONSE OF VEGETATION
Although isoprenoid fluxes showed the expected
exponential dependence from the temperature of
the canopy, attempts to simply fit the isoprene
fluxes with the original light and temperature
algorithm developed by Guenther et al. (1993)
failed.
20
According to the Guenther algorithm (G93) the
isoprene emission is given by E Eo CL CT where
Eo is the isoprene emission measured at 30C (Ts)
and at values of the photosynthetic active
radiation (PARQ) of 1000 mmol of photons m-2
s-1. CL and CT are empirical terms describing the
light and temperature dependence of E .
21

• By knowing that
• F E r LAI e
• where
• F the isoprene flux in mg m-2 h-1,
• E the isoprene emission mg m-2 (leaf surface)
  h-1 measured with enclosures,
• LAI leaf area index
• r the portion of the canopy emitting isoprene
• e a transport term
• we can write that
• F F CL CT
• In which F is the flux measured when the canopy
  reaches 30C under light saturation conditions.

22
This simplified approach worked quite well in the
temperate forest of Castelporziano where the main
vegetation species (Quercus ilex L.) emitted
monoterpenes with the same light and temperature
algorithm followed by isoprene.
23
It did not work in the tropical forest of Manaus
where it was impossible to predict the isoprene
canopy exchanges using a fixed value of the basal
canopy emission (F).
24
Plots of F/CTCL showed, indeed, that F was not
constant with values varying from 1 to 4.
25
By plotting vs. T the values of F/CT CL measured
at canopy temperatures higher than 30C and under
light saturation conditions, it was found that F
increased with the temperature according to the
following equation
F F e 0.38 (T-Ts)
Where F 0.55 mg m-2 s-1
26
These observations suggested that the G93
algorithm could have been modified to account
for the change of the F with the temperature of
the canopy. F F' CC CT CL Where F
basal flux of fully exposed leaves CC the
correction term accounting for the change of F .
27
F r LAI F r LAI
Full sun adapted leaves
Partly adapted leaves
Dark growing leaves
E
F r LAI F r LAI F r LAI
t
28
The corrected G93 algorithm provided a better
description of the isoprene fluxes than the
original one.
29
It was able to pick up better the basic features
of daily variations of isoprene fluxes as a
function of the environmental parameters of the
canopy.
30
The same approach was used to describe the
fraction of monoterpenes showing an early onset
of emission (mainly a-pinene, b-pinene, camphene
and myrcene). In this case, the following
correction term (CC) was used for F CC F e
0.07 (T-Ts) with F 0.25 mg m-2 h-1 The
modified G 93 algorithm was better representing
the fact that no substantial monoterpene emission
was observed in the early morning hours and in
the late afternoon.
31
(No Transcript)
32
To describe the late appearence of some
monoterpenes (such as limonene and sabinene), a
temperature dependent component was added to the
light and temperature monoterpene emissions. The
following algorithm was used F F e
0.1(T-Ts) where F 0.04 mg m-2 h-1 The
fraction of temperature dependent monoterpenes
was selected on the basis of the percent
increase in emission measured with REA fluxes.
33
(No Transcript)
34
3) Modeling the Net Isoprenoid Ecosystem Exchange

• Using the empirical relations derived from the
  REA fluxes, the NIEE was modeled as a function of
  the PAR intensity and temperature variations.

35
Modeled data seemed to fit fairly well with
observations
THEY SHOWED THAT ISOPRENE CONTRIBUTED TO THE
NIEE BY 66 AND MONTERPENES BY 34.
36
Using these data, the fraction of NEE emitted as
isoprenoids was calculated for the Manaus
site. It was found that isoprenoid emission
accounted for ca. 1.9 of the NEE, a value 30
lower than that given by the temperate oak forest
of Castelporziano during summer (3).
37
By considering that -strong isoprenoid
emitters produce ca. 4 mg C m-2 (leaf area)
h-1 -our ecosystem has an average LAI of ca.
6 -the forest canopy in Manaus is rather
close Such fractions can only be explained by
assuming that only ca. 30-40 of canopy fraction
irradiated by sunlight is emitting isoprenoids at
high rates.
38
By considering the contribution coming from the
total respiration we can estimate that total
isoprenoid emission accounts for a fraction of
the GPP ranging from 0.8 to 1.2, a value not
much different from that estimated by
Kesselemeier et al. (J. Biogeochemical Cycles,
2002, in press).
39
As far as the reactivity is concerned, REA and
modelling data indicate that isoprene emission
contributes to the NIEE by a fraction ranging
from 66 to 72 whereas the average concentrations
in the boundary layer go from
Just before noon
Late afternoon
This suggests that in the middle of the day
monoterpenes are removed by photochemical
reactions at a rate which is twice as faster as
that of isoprene.
40
CONCLUSIONS

• Isoprene emission in the Manaus forest site
  accounts for the largest portion of the NIEE (66
  to 72). Little seasonality is observed in the
  emission.
• Monoterpenes seem to show some seasonality as
  they increase their emission going from the wet
  to the dry season.
• Isoprenoid emission cannot be described using
  simple algorithms due to the substantial impact
  of the portion of the canopy not fully adapted to
  sunlight on the emission.
• The fraction of NEE allocated to isoprenoid
  emission is of the order of 2. This indicates
  that a substantial portion of vegetation is
  represented by low or non isoprenoid emitters.
• Monoterpenes show midday removal rates by
  photochemical reactions that are twice as faster
  as that of isoprene. The composition is
  dominated by a- and b-pinene that accounts by
  more than 70 of the entire fraction.