Combination of different techniques and multi-scale approach to understand CO2 budget in a temperate beech forest

1

Bernard Longdoz1_{, D. Epron}1_{, S. Goffin}2_{, A. Granier}1_{, E. Granier}1_{, P. Gross}1_{, N. Heydt}1_, J. Ngao3_{, C. Plain}1_{, N. Marron}1_{, F. Parent}1_{, M. Schrumpf}4

Combination of different techniques and multi

Combination of different techniques and multi

-

-

scale

scale

approach to understand CO

approach to understand CO

budget in a temperate

budget in a temperate

beech forest

beech forest

2 _{Liège University, Gembloux Agro-Bio Tech, Belgium} 3 _{Paris-Sud University, CNRS, UMR 8079, Orsay, France}

1 _{INRA, Centre INRA de Nancy, UMR1137, 54280 Champenoux, France}

Nancy University, INRA, UMR 1137, Vandoeuvre les Nancy, France

4 _{MPI Biogeochemistry, Jena, Germany}

Context

Quantification C storage Understanding C processes

European Beech (Fagus Sylvatica), French forest

N E W

S

Nancy

Nancy HesseHesse

Hesse site

48°40’N, 7°05’E

48°40’N, 7°05’E

Mean annual air Temp. : 10°C10°C

Mean annual Precip. : 950 mm950 mm Temperate

Temperate

65 km from Nancy (East)

Climate

3 200 m Hesse 1 N 90% beech 42 years old Height 19 m

Soil type: stagnic luvisolstagnic luvisol

5 horizons : A, S1, Sg2, II Cg, III C C_soil 9.5 kg m-2 LAI : 7.5 -5 Density ≈ 3000 n ha-1 Aerial biomass ≈ 100 t ha-1

Hesse site

Automatic measurements :

• Net Ecosystem Exchange NEE (Eddy Cov., 30 min) • Micro-climate (T°, radiation,

humidity, precipitations) • Soil T° and water content • Trunk circumferences

(dendrometers) ⇨ C biomass

Measurement campaigns :

• Aerial (Stems, leaves, fruits,…) and below ground (roots) Biomass

• Soil composition & characteristics (density, C & N contents…

• δ13_{C of sampled materials (IRMS) and gaz (TDLS} + IRMS) • Soil Respiration (Rs) • LAI Tower 22m Sonic anemometer IRGA (CO2)

Material

5 NEE GPP TER -30 -20 -10 0 10 20 14/7 15/7 16/7 17/7 Time NEE (µ m ol m -2 s -1 ) GPP GPP+TER+TER TER TER

NEE measurements

GPP

GPP & TERTER have difference in response to changes in environnemental conditions

1st _approach:

1. TERTER (t°, SWC) determination from night and leafless data

2. TERTER (t°, SWC) extrapolation to leafy daytime

3. Daytime : GPP

GPP = NEENEE (data) – TERTER (extrapolation)

2nd _{approach: Combining 2 equations}

1. NEENEE = GPPGPP + TERTER

2. NEENEE *δ13_C

NEE = GPPGPP *δ13Catm + TERTER *δ13CTER

NEE partitioning

7 Improvement Rs model a 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 180 240 300 360 420 480 540 600 660 720 780 Jour (2003-2004) RS (µmo lCO 2 m -2 s -1 ) H1A H1B H1C H1D H1E H1F H1G H2B H2I Footprint fluctuations (WD, WS, u*)

+

TER (Rs) spatial variability

Low turbulence (u*<<) periods

Sous-estimation des flux de nuits

0 1 2 3 4 5 6 7 8 9 0.0 0.2 0.4 0.6 0.8 1.0 1.2 u* (ms-1) FL UX ( µ molm-2s-1) R10 ETE 1997 R10 ETE 1998 R10 ETE 1999 R10 ETE 2003 EC anomalies TER (Ts₁₀, SWC₁₀) validity 0 1 2 3 4 5 6 7 8 9 10 9 10 11 12 13 14 15 16 17 18 19 20 21 Ts10 (°C) T ER (µmo l m -2 s -1 )

1

Approach Uncertainties

Large uncertainty on TER (t°, SWC) TER (Ts₁₀, SWC₁₀): R² < 0.1

(Longdoz et al., 2008)

N

E

W

S

Hydromorphic levellevel

BOR BO Bre

Hesse1

Hesse1

Rs spatial variability

soil bulk density of the A layer (C/N) (C/N)_AA ratio of carbon to nitrogen content of A layer &

Above & belowground characteristics

(

)

(

exp

a

b

SWC

)

exp

Q

R

R

⋅

−

−

⋅

⋅

=

9 (R2 _{= 0.87)} R2 _{= 0.6602} 0.0 0.5 1.0 1.5 2.0 2.5 0.4 0.6 0.8 1.0 1.2 Densité (g cm-3) R1 0 (µ m o lCO 2 m -2 s -1 ) R2 _{= 0.7258} 0.0 0.5 1.0 1.5 2.0 2.5 8 9 10 11 12 C/N A0 R1 0 (µ m o lCO 2 m -2 s -1 ) (C/N)_A (ρ_S)_A (g/cm3)

( )

( )

1

.

47

0

.

19

C

_N

R

=

−

⋅

ρ

+

⋅

Soil sampling for ((ρρ_SS))_AAand (C/N)(C/N)_AA mapping _{Soil sampling points (>100) in footprint area}

Rs spatial variability

Rs spatial variability

( )

( )

1

.

47

0

.

19

C

_N

R

=

−

⋅

ρ

+

⋅

11 Rs spatial variability

Confirmed by first analyse of EC-TER data

ANOVA with 45° Sectors significant difference (p=0.029) “Hot area” at West-South-West

“Cold area” at South-West

T°s & SWC R₁₀ map Rs map Rs spatial variability Rs contributions to TER Estimations of TER temporal fluctuations due to Rs spatial variability N N sparse vegetation mature deciduous young deciduous coniferous agriculture sparse vegetation mature deciduous young deciduous coniferous agriculture

Distance East - West [m]

-600 -300 0 300 600 900 D is ta nce N orth - So uth [m] -900 -600 -300 0 300 600 + 5 10 20 N

Distance East - West [m]

-600 -300 0 300 600 900 D is ta nce N orth - So uth [m] -900 -600 -300 0 300 600 + 5 10 20 N N Footprint model

13 a 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 180 240 300 360 420 480 540 600 660 720 780 Jour (2003-2004) RS (µmo lCO 2 m -2 s -1 ) H1A H1B H1C H1D H1E H1F H1G H2B H2I Improvement Rs model

Large uncertainty on TER (t°, SWC) TER (Ts₁₀, SWC₁₀): R² < 0.1 (Longdoz et al., 2008) EC anomalies Low turbulence (u*<<) periods TER (Ts₁₀, SWC₁₀) validity

Sous-estimation des flux de nuits

0 1 2 3 4 5 6 7 8 9 0.0 0.2 0.4 0.6 0.8 1.0 1.2 u* (ms-1) FL UX ( µ molm-2s-1) R10 ETE 1997 R10 ETE 1998 R10 ETE 1999 R10 ETE 2003 0 1 2 3 4 5 6 7 8 9 10 9 10 11 12 13 14 15 16 17 18 19 20 21 Ts10 (°C) T ER (µmo l m -2 s -1 )

1

Approach Uncertainties

Footprint fluctuations (WD, WS, u*) +

TER (Rs) spatial variability

Improvement of Rs model

Differentiate

Soil CO2 efflux

Production & Transport

Better description of the Transport processes

Not enough information

Important vertical

profile

Need for CO

Need for CO₂₂ concentration concentration

measurements in the different

measurements in the different

layers

layers

∆ [CO₂]

∆ [CO₂]

∆ [CO₂]

Better description of the Production processes

Differentiate sources and their behaviour response to environmental

conditions Also necessary to have better understanding

15 Improvement of Rs model

To be continued in the Caroline

To be continued in the Caroline

Plain Talk Plain Talk

+ CO

+ CO

Transport

Transport

model

model

2nd _{approach: Combining 2 equations} 1. NEENEE = GPPGPP + TERTER

2. NEENEE *δ13_C

NEE = GPPGPP *δ13Catm + TERTER *δ13CTER

2

Approach Uncertainties

Footprint fluctuations +

δ13_C

TER (δ13CRs ) spatial variability

δ13_C

TER (δ13CRs ) temporal variability

Large uncertainty on δ13_C

17

Soil Keeling Plot

Soil Keeling Plot

LICOR Li LICOR Li- -6200 6200 Sampling chamber Sampling chamber IRGA IRGA Li Li--62006200 Respiration Respiration chamber chamber Mass spectrometer Mass spectrometer Keeling plot Keeling plot

δ

_C

spatio-temporal variability

Air

Soil

Sample Pump

Pump Reference gas (10% CO2)

Laser Sample cell Sample detector Reference detector Dewar δ13_{C temporal variability} Ref : Daniel

Ref : Daniel EpronEprontalk, talk, MarronMarronet et al. 2009 al. 2009 TDLS (δ C of gas) Soil chambers (δ13_{C out -} δ13_{C) ⇒}δ13_C Rs

19 Marron et al., 2009 Subplots Seasonal variation ≈3‰ Daily variation ≈2‰ Small scale spatial variation ≈2‰ Large scale spatial variation > 1‰ δ

δ13_C Rs Spatio-temporal variations from –25‰ to -30 ‰ TER TER uncertainty of 20% Complex climatic determinism (SWC 4 to 5 days before) Use of δ13_{C as tracer : C transfer,}

allocation-source partitioning (Daniel Epron talk)

More knowledge about 13_C

discriminations

(δ13_{C Production &} δ13_CTransport)

Understanding

δ13_{C Production &}

δ13_{CTransport determinism}

Improvement of δ13_C

21

Improvement of

δ

C

model

Better description of the δ13_{C Production}

processes

Differentiate sources and their behaviour Differentiate

Important vertical

profile

Need for

Need for δδCO2 concentration CO2 concentration measurements in the different

measurements in the different

layers

layers

Soil δCO2 efflux

δ13_{C Production &} δ13_{C Transport}

response to environmental

conditions

Not enough information Not enough information

Better description of the δTransport processes

∆ [δCO₂]

∆ [δCO₂]

∆ [δCO₂]

Improvement of δ13_{C model}

To be continued in the Caroline

To be continued in the Caroline

Plain Talk Plain Talk

+

+

δ

δ

CO

CO

Transport

Transport

model

model

Pump Reference gas (10% CO2)

Laser Sample cell Sample detector Reference detector Dewar

23

あり

がと

う

Rauto = 69% de Rs

Corrections: - trenched roots decomposition

- higher SWC in trenched plot

NPP = NEE – Rs hetero

R

R

R

R

R

R

Rs autotrophic-heterotrophic

partitioning: Trenched plots

NPP from Eddy covariance (NPPec)

Li-6252

Rs : soil chambers

25 Rauto 52.0% tree growth 29.8% leaves 11.0% fine roots 4.8% fruits 0.2% mortality_2.2%

Pourcentage from GPP = 1404 gC m

NPP from biomass measurements (NPPbio)

0 200 400 600 800 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 N P P (g C m -2 y r -1 ) NPPec NPPbio

On 10 years : difference of 1%

General good reproduction of inter-annual variability but large

divergences unexplained (1998 & 2001)