Water, biology, and climate in northern Wisconsin: Carbon dioxide fluxes at a drying wetland Benjamin Sulman Department Seminar Department of Atmospheric and Oceanic Sciences University of Wisconsin-Madison Advisor: Ankur Desai

What to expect

• Introduction: A meeting of disciplines • Motivation: The wetland carbon pool and future hydrology • Methods: What the heck is eddy covariance?

• Results: How did carbon dioxide fluxes and energy balance change at a drying wetland?

• Conclusions: What does this mean for climate change feedbacks?

A meeting of disciplines

Ecology Climate science Land-atmosphere interactions Boundary layer meteorology This talk: Something for everyone!

Hydrology

Definitions

• Wetland: terrestrial ecosystem that is inundated with water for at least part of the year, with characteristic plants – Fen: fed by groundwater, nutrient rich, high productivity – Bog: fed by rain, nutrient poor, low productivity, different plants • Peatland: terrestrial ecosystem with a thick layer of rich, poorly decomposed, organic soil • Water table (WT): depth where soil is saturated with groundwater – Negative numbers indicate depth below the soil surface – Positive numbers indicate standing water above the surface • Fluxes: exchanges between the land and atmosphere. Positive numbers are upward, from the land to the atmosphere – Sensible heat flux: direct heat transfer from land to atmosphere – Latent heat flux: transfer of energy contained in water vapor. This is equivalent to evapotranspiration (ET)

Motivation: How do wetlands interact with climate?

• Biogeochemical interactions: How do wetlands impact the carbon budget of the atmosphere?

• Biophysical interactions: How do wetlands impact the heat and moisture budgets of the atmosphere?

• How will climate change affect these interactions?

The record of atmospheric CO

2

and warming

Images created by Robert A. Rohde / Global Warming Art Data published by NOAA Data from NASA Goddard institute http://www.globalwarmingart.com/wiki/Image:Instrumental_Temperature_Record_png http://www.globalwarmingart.com/wiki/Image:Mauna_Loa_Carbon_Dioxide_png

Projected climate changes affecting wetlands:

• Higher temperatures, more precipitation • Net drying due to more evaporation • Poleward shift of biomes Our question: how will predicted hydrological changes affect interactions between wetlands and climate?

Biogeochemical interactions: Only about half of CO 2 emissions stay in the atmosphere IPCC AR4, WG1 (2007)

Biogeochemical interactions: Future land carbon uptake is not well characterized Friedlingstein et al., 2005,

J. Clim

Biogeochemical interactions: The wetland carbon pool is large Boreal and subarctic wetlands contain an estimated 455 Pg soil carbon.

This is up to 1/3 of total global soil carbon pool (Gorham, 1991) Mitra et al, 2005, Curr. Sci.

Biogeochemical interactions: Peatland carbon balance is sensitive to hydrology CH 4 CO 2 CO 2 CH 4 Underwater (anoxic; anaerobic bacteria) Above water (oxygenated; aerobic bacteria)

Biogeochemical interactions: Existing literature is contradictory • Modeling studies have identified a climate feedback (e.g. Ise et al. 2008) • Observations have mixed results – Lowering water table increased CO 2 emission or changed wetlands from a carbon sink to a source: (Silvola et al. 1996, Alm et al. 1999, Bubier at al. 2003) – No correlation between water table and CO 2 emission: (Updegraff et al. 2001, Lafleur et al. 2005)

Biophysical interactions: Energy and moisture budgets

An example: the biophysical effects of deforestation Foley et al. 2003

Biophysical interactions: How does drying a wetland affect the energy and moisture budgets of the atmosphere?

Case 1: High water table High latent heat loss Low sensible heat loss Case 2: Low water table High sensible heat loss Low latent heat loss

The context of this observational study: Global distribution of wetlands Forested bog Nonforested bog Forested Swamp Nonforested swamp Alluvial Formations Other land Water body Matthews and Fung, 1987, GBC

Our study: Observations on a wetland in northern Wisconsin during a decline in water table

• This region is representative of boreal wetland regions • The declining water table is representative of changes in hydrology predicted by climate simulations

Study sites are located in the Northern Highland region of WI

Wisconsin Online http://www.wisconline.com/wisconsin/geoprovinces/northernhighland.html

LEAF - Wisconsin Center for environmental education

The Northern Highlands landscape 15 ± 3 7 ± 4 Percent of area Percent of carbon pool 10 ± 3 Forests 62% 36% 111 ± 37 Wetlands 20% 55% 0.05

± 0.02

30 ± 25 Lakes 13% 9%

Estimated pool sizes (kg-C m -2 )

Estimates and diagram from Ishi Buffam (post-doc in UW Zoology Dept.)

Study site locations

Legend MODIS IGBP 1km landcover

Our Sites: Lost Creek

• Alder-willow fen • Fed by stream and groundwater • Covered in woody shrubs • Six years of flux data (2001-2006)

Our sites: Willow Creek

• Upland hardwood forest • Mature forest, about 60 80 years old • This is the period of maximum carbon uptake for a typical forest • Seven years of flux data (2000-2006)

Our sites: Sylvania

• Old-growth hemlock-hardwood forest • More than 300 years old • Six years of flux data (2001-2006)

Our sites: South Fork and Wilson Flowage

• Wetland sites • SF: Ericaceous bog • WF: Grass-sedge-shrub fen • Three years of growing season flux data with roving tower • Switched between sites every two weeks • Much less data than LC, WC, and Sylvania

Methods: Eddy Covariance

Turbulent flux Equipment: • 3D sonic anemometer • Open or closed path gas analyzer • 10Hz temporal resolution • Multiple level CO 2 profiler Storage

CO 2 w

Turbulent data: CO

2

and w

Time (s)

Carbon flux data products

• Net Ecosystem Exchange (NEE) – Total net carbon flux (measured) • Ecosystem Respiration (ER) – Carbon released to atmosphere – Calculated based on nighttime NEE • Gross Ecosystem Production (GEP) – Carbon absorbed from atmosphere – Calculated based on NEE - ER

Other data

• Water table (WT, height above soil surface) • Precipitation • Air and soil temperature • Photosynthetically active radiation (PAR) • Latent and sensible heat fluxes • Net radiation

Results

• Regional climate and trends • Decline in water table • Biogeochemical: Changes in Lost Creek carbon balance • Biophysical: Changes in Lost Creek energy balance • Comparison with nearby forests and wetlands

Results: Regional climate

T Precip WT

Regional drying trend

Stow et. al. 2008 Precip LC water table

Precipitation was well correlated with Lost Creek water table

r 2 = 0.79

ER NEE GEP Lost Creek Wilson Flowage Willow Creek (forest)

Results: Carbon fluxes

Biogeochemical interactions: Ecosystem respiration and water table CO 2 CO 2 Underwater (anoxic; anaerobic bacteria) Above water (oxygenated; aerobic bacteria)

Respiration results: Wetland and upland site comparison • Wetland ER increased as water table declined • Upland ER was independent of wetland water table    Lost Creek (wetland) Sylvania (old growth) Willow Creek (mature)

Respiration results: Wetland site comparisons

• Lost Creek and Wilson Flowage show similar patterns • South Fork (bog) ER is less correlated with WT

Biogeochemical interactions: GEP and water table

• Lowering of water table dries upper level soil • Oxygen and nutrients become available to plants, leading to increased growth • The result is an increase in GEP

GEP results: Wetland and upland site comparison    Lost Creek (wetland) Sylvania (old growth) Willow Creek (mature) • Wetland GEP increased as water table declined • Aboveground biomass also increased • Upland GEP was independent of wetland water table

GEP results: Wetland site comparisons

• Lost Creek and Wilson Flowage show similar patterns • South Fork, a bog, shows increased GEP with higher WT

Biogeochemical interactions: No correlation between water table and NEE Increases in both ER and GEP offset: Wetland NEE was not correlated with water table

Biophysical interactions

Case 1: High water table High latent heat loss Low sensible heat loss Case 2: Low water table High sensible heat loss Low latent heat loss

Biophysical interactions: Evapotranspiration and water table ET declined along with water table Water table ET

Biophysical interactions: Changes in energy balance

• Net radiation increased due to increasing incoming radiation • Sensible heat flux increased relative to latent heat flux • If these changes occur on a large scale, they can have significant effects on regional climate (Sampaio et al. 2007, Foley et al. 2003) Latent heat flux Sensible heat flux Net radiation Heat flux into ground

Biophysical interactions: Vegetation can affect regional climate Foley et al. 2003

Conclusions:

• Biogeochemical interactions: – Both ER and GEP increased at Lost Creek as water table declined – Net ecosystem CO 2 table exchange was independent of water – This supports literature arguing against a strong north temperate wetland water table-carbon feedback • Biophysical interactions: – ET declined and energy balance shifted in favor of higher sensible heat flux – This could potentially be a positive feedback to warming and drying trends

Discussion and further work

• We did not address methane fluxes or dissolved carbon export • Lost Creek, a shrub wetland, may not represent responses at other types of wetlands – Bogs appear to respond differently -- these differences are not typically included in GCMs – Studies in arctic tundra have observed net emissions of carbon with warming • The observed ecological changes suggest that future successional changes may be important over the long term • We plan to incorporate these results into an ecosystem model for upscaling

Acknowledgements

• My advisor, Ankur Desai • Jonathan Thom, Shelley Knuth • Bruce Cook (UMN) • Pete Pokrandt • Fellow grad students • AOS faculty and staff This research was sponsored by the Department of Energy (DOE) Office of Biological and Environmental Research (BER) National Institute for Climatic Change Research (NICCR) Midwestern Region Subagreement 050516Z19, and by a NASA Carbon Cycle grant.

References

Alm et al., Carbon balance of a boreal bog during a year with an exceptionally dry summer, Ecology, 80(1), 161-174, 1999.

Bubier et al., Spatial and temporal variability in growing-season net ecosystem carbon dioxide exchange at a large peatland in Ontario, Canada, Ecosystems, 6, 353-367, 2003.

Foley et al., Green surprise? How terrestrial ecosystems could affect earth’s climate, Fr. Ecol. Env., 1(1), 38 44, 2003.

Friedlingstein et al., Climate-carbon cycle feedback analysis: Results from the C 4 MIP model intercomparison, Journal of Climate, 19(4), 3337-3353, 2006.

Ise et al., High sensitivity of peat decomposition to climate change through water-table feedback, Nature Geoscience, 2008.

Lafleur et al. Ecosystem Respiration in a Cool Temperate Bog Depends on Peat Temperature But Not Water Table. Ecosystems, 8, 619-629, 2005.

Matthews and Fung, Methane emission from natural wetlands: Global distribution, area, and environmental characteristics of sources, Global Biogeochemical Cycles, 1 (1), 61-86, 1987.

Mitra et al., An appraisal of global wetland area and its organic carbon stock, Current Science, 88 (1), 25-35, 2005.

Sampaio et al., Regional climate change over eastern Amazonia caused by pasture and soybean cropland expansion, Geophys. Res. Lett., 34(L17709), 2007.

Silvola et al., CO2 fluxes from peat in boreal mires under varying temperature and moisture conditions, Ecology, 84, 219-228, 1996.

Updegraff et al., Response of CO2 and CH4 emissions from peatlands to warming and water table manipulation. Ecological Applications, 11(2), 311-326, 2001.