What have we learned from forest tower flux data following

disturbance?Brian Amiro, A. Barr, J. Barr, T.A. Black, R. Bracho, M. Brown, J. Chen, K. Clark, K. Davis, A. Desai, S. Dore, V.

Engel, J. Fuentes, A. Goldstein, M. Goulden, T. Kolb, M. Lavigne, B. Law, H. Margolis, T. Martin, J.H. McCaughey, M. Montes-Helu, A. Noormets, J. Randerson, G. Starr, J. Xiao

Disturbances in North American Forests

(total area 7 million km2)• Fire: 40,000 km2/year• Harvest: 50,000 km2/year• Insects: larger• Storms/hurricanes: > 17,000 km2/year• Disease: ???

Disturbance Sequence Studies

• Our goal: Synthesize flux tower measurements of CO2 exchange for experiments designed to understand the effect of disturbance

• Chronosequence approach: forests of different ages

• Time series approach: follow a single site through a disturbance cycle

JGR 115: G00K02, doi:10.1029/2010JG001390, 2010

Acknowledgements: Large host of contributors and funders

Disturbance Chronosequenceswith Annual NEP measured by eddy covariance

Fire = 4Harvest = 7+Insects = 3Hurricane Wilma

Fire Sites: Boreal and Arizona

Forest Tent Caterpillar

Mountain Pine BeetleGypsy Moth

INSECTS

Harvesting: Practices highly variable. Clearcut, thinning…

Hurricane Wilma in October 2005; destroyed 30% of trees at a Florida mangrove site

Trajectory of hurricane Wilma during October 2005

Annual Carbon Flux from Eddy Covariance Towers

• Measure CO2 directly using fast-response anemometer and gas analyser

• Get 30-min mean fluxes• Calculate annual NEP (Net

Ecosystem Production)• Calculate GPP (Gross

Photosynthetic Production) and ER (Ecosystem Respiration) from tower data

• Networks: Ameriflux, Fluxnet-Canada (Canadian Carbon Program)

Age (years)

0 20 40 60 80 100

NE

P (

g C

m-2

y-1

)

-200

-100

100

200

300

0

Saskatchewan: PineManitoba: SpruceAlaska SpruceArizona: Pine

Annual NEP for fire chronosequences: boreal sites gaining carbon after 10 years

Fire: GPP and ER

GP

P (

g C

m-2

y-1 )

0

200

400

600

800

1000

Age (years)

0 20 40 60 80 100

ER

(g

C m

-2 y

-1)

0

200

400

600

800

1000

1200

SaskatchewanManitobaAlaskaArizona

ER relatively flat with age at boreal sites

Time to recovery from fire?Curve fit of NEP data from boreal fire sites

Boreal Forest Sites

Time Since Fire (years)

0 5 10 15 20 25 30

Cu

mu

lati

ve N

EP

(g

C m

-2)

-800

-400

400

800

1200

1600

0

2000

Typical range of mean carbon lost through combustion

Harvest Chronosequences: NEP

Age (years)

0 20 40 60 80 100

NE

P (

g C

m-2

y-1

)

-800

-600

-400

-200

200

400

600

800

0

1000

Saskatchewan: pineQuebec: spruceNew Brunwick: firVancouver Island: firOregon: pineFlorida: pineWisconsin: hardwoods

Harvest GPP and ER

Age (years)

0 20 40 60 80 100

ER

(g

C m

-2 y

-1)

0

500

1000

1500

2000

2500

GP

P (

g C

m-2

y-1

)

0

500

1000

1500

2000

2500

3000

SaskatchewanQuebecNew BrunswickVancouver IslandOregonFloridaWisconsin

Are there any relationships that apply with age following stand-replacing disturbance?

• Leaf area index (LAI) development?• What about GPP and ER relationships?

Ratio GPP/ER with age

Age (years)

0 20 40 60 80 100

GP

P/E

R

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

0.0

1.0

2.0

FireHarvest

Asymptote = 1.23

Ra = 0.55*GPP

Heterotrophic Respiration (normalized to mean for individual chronosequences)

assuming Ra = 0.55*GPP

Age (years)

0 20 40 60 80 100

Het

ero

tro

ph

ic R

esp

irat

ion

No

rmal

ized

0.0

0.5

1.0

1.5

2.0

2.5

FireHarvest

No statistically significant pattern with age

NEP Thinning: Harvest and Hurricane

Time since disturbance (years)

-2 -1 1 2 3 4 50

NE

P (

g C

m-2

y-1

)

-200

200

400

600

800

1000

1200

1400

0Florida HurricaneArizona ThinningNew Brunswick ThinningCalifornia Thinning

Insect sequences: NEP

Time since disturbance (years)

-2 -1 1 2 3 4 50

NE

P (

g C

m-2

y-1

)

-200

200

400

600

0

Mountain Pine BeetleForest Tent CaterpillarGypsy Moth

CFS web page

What have we observed?

• Forests become carbon sinks within 5 to 20 years following stand-replacing events

• GEP/ER ratio becomes relatively constant by 15 years

• Heterotrophic respiration variable with no strong age function

• Quick recovery in NEP following thinning• Insect effects variable: depends on nature of

infestation, but NEP becomes positive quickly

What should we do; where are the gaps?

• All sites were disturbed at some time, so when does the disturbance history become less important than inter-annual variability and other site factors?

• Do young sites respond to climate and stressors differently from more mature sites?

• Do we need more chronosequences? Where? For how long? Can we compare severity/management practices?