GCP, Beijing, 15-18 November 2004. Regional Carbon Budgets: from methodologies to quantification

Carbon fluxes and sequestrationopportunities

in grassland ecosystems

Jean-Francois Soussana

INRA, Grassland Ecosystem Research Group, Clermont-Ferrand, France

Overview

Introduction.C cycling and Net Biome Productivity in grasslands

• Soil C and land management• CO2 fluxes• Farm gate budget and mitigation options• Upscaling to regional scale

Conclusions

Net Biome Productivity of a grazed grassland(tC ha-1 yr-1)

Végétation

Soil

Root turnoverRhizodeposition

Litter

CO2

9.2

Gross primaryproductivity

Below-groundrespiration

Vegetation0 Shoot

respiration

9

3 7

0.7 Animalexcreta

Net Biome Productivity: +0.5

Herbivorerespiration

CH4

0.2

Herbivore+0.05

Grazing

192.1

DOC, DIC?

SownSown grasslandgrassland withwith intensive intensive grazinggrazing ((SoussanaSoussana et et alal., ., SoilSoil Use Use andand Management, 2004Management, 2004))

1. Grassland C stocks and land management

Land use changes from or to grasslands(Guo & Gifford, 2002)

(-59%)

(+8%)

(+19%)

Land use change effects on soil carbon stocks

-40

-30

-20

-10

0

10

20

30

40

0 20 40 60 80 100 120Years after start of policy measure

Carb

onsto

cks (

tC/ha

)

arable -> forestarable -> grasslandforest -> arable

grassland -> arable

Land use change: carbon storage is slower than carbon relase(After INRA, 2002)

Soil C and grassland management

Improved (medium/high input)FertilizersMore Productive

GrassesIrrigationLimingSeeding Legumes

Nominally-ManagedNative VegetationPasture/Rangeland

with no grazingproblems or inputs

DegradedOvergrazingLess productive

VegetationDegraded Sites

IPCC Grassland Management Categories

Changes in grassland C stocks• Rates of land use and grassland management changes

– Need to develop improved systems for the collection of statistics on types and timing of land use change and of agricultural management events (e.g. manuring, cutting, extent of grazing etc.)

- Need to know past history of land use and grassland management

• Soil organic carbon change per unit area:– a) non-linear, more rapid during the early years after adopting a

practice. – b) asymmetric: accumulation is slower than release – Any estimate of soil C storage must refer to a given time period

and both to the previous and current management.– Interrupting stock-enhancing practices usually results in a rapid

release of carbon to the atmosphere.

2. CO2 fluxes

EC ‘GreenGrass’ project FP5 - EESD (2002-2004)

Contributing to CarboEurope

France (Co-ordination) Scotland Hungary Switzerland ItalyIreland DenmarkThe Netherlands

CNR-IATAA

Activity 1.6. Grasslands & Wetlands

28sites

Main Grass.

Main Wet.Anc. Grass.

Anc. Wet.

Components of the Grassland Carbon BudgetComponents of the Grassland Carbon Budget

∆∆C/C/∆∆tt

manure

?

CO2 exchange (NEE) harvest

sequestration: ∆C/∆t = NEE + C imports- C exports

Global warming potential: N2O and CH4 trade-offs with CO2

N2O: automated staticchambers and TDL SF6 pill

SF6

CH4

CH4 , SF6

CH4: in-situ SF6 tracer method

Field scale flux measurements

• In terms of NEE, managed grasslands are large apparent sinks of C• C imports from manures and C exports from cuts should be taken into

account to estimate NBP. • Uncertainty associated to NBP is high• The NBP was found to be a carbon sink, with approximately half of the

sink activity resulting from imports of C from manures• Emissions of N2O and CH4 resulted in a 40 % offset of the NEE • The attributed emission balance, including indirect emissions of CO2

and CH4 from the cut herbage, was on average neutral.

3. Upscaling C budgets at the farm scale

PASIM

CO2, N2O,CH4

IPCC

N2O,CH4

Imports /Exports

Module 7:Feeding systems

Module 7:Feeding systems

Module 1:Farm structure

Module 1:Farm structure

Module 2:Herd

Module 2:Herd

Module 3:Grazing systems

Module 3:Grazing systems

Module 5:Fertilis. & Cutting

Module 5:Fertilis. & Cutting

Module 6:Crops management

Module 6:Crops management

Module 4:Housing

Module 4:Housing

Module 9:Energy consump.

Module 9:Energy consump.

Module 8:Waste managem.

Module 8:Waste managem.

FARMSYSTEM

Structure of the farm model FARMSIM

PaSim model (FAL, LSCE, CEH, INRA)

photosynthesis

substrate

allocation

lamina stem earCarbon and Nitrogen BIOMASS

mortality

Carbon and Nitrogen litter(structural / metabolic)

decomposition

soil carbon and nitrogen

cutting

yield

grazing

animals

CH4

autotrophicrespiration

active slow passive

heterotrophicrespiration

CO2

root

N uptake

N fertiliser

N deposition

nitri/denitrification

volatilization

leaching

N2O NO3-NH3

N2O N2

N fixation

urinedung

Nitrogen outputs Nitrogen Inputs Carbon Inputs CO2 emissions CH4emissions

yielddecomposition

Farm scale C and GHG balance

• All simulated cattle farms were sources of GHG• Indoors emissions are a major component of the farm budget• Non CO2 trace gases play a major role in the GHG budget• Rate of emission per unit land area increased from extensive to

intensive farming systems. • Possibility to upscale budgets per region according to farm type. This

allows to account for changes in farming practices.• Pre-chain emissions may lead to much higher emission than farm gate

budgets.

4. Upscaling grassland C budgets at the regional scale

Ecosystem Similarity Concept

Gilmanov, Wylie, et al. 2004

Grasslands in Europe

% land cover (CORINE – PELCOM data)

Upscale Pasim at a regional scale

PaSim

Hourly values of • irradiance• temperature• pressure• humidity. • wind speedClimatology from ECMWF

Grassland fractional coverage

Combined CORINE. PELCOM

• Soil texture (Zoebler)• Water content parameters(FAO)

Soil data

dates of harvest. animal stocking rate and grazing periods. dates of

application and amount of N-fertilizers

Management drivers

Equilibrium run at a spatial resolution of 1 degree

• GPP• Respirations• N2O emissions• CH4 emissions• …

Model outputs

Automatic optimal management module

Climate drivers Land cover map

(Vuichard et al.)

Upscaling C budgets at regional scale

• Phenomenological models (with NDVI and GIS)– Light-response functions method works with grasslands– Provides estimates compatible with ecological theory and data– There are significant relationships between ecosystem scale

characteristics and remotely sensed and on-site factors

• Process-based simulation models– CO2, N2O and CH4 fluxes can be predicted by a process driven model

at the European scale. It takes into account management– Optimal management scenario predicts

• realistic values of yield • realistic dependencies with temperature and precipitation

– Refine runs by using “real” management data– Develop “mitigation” runs