Assessing the Consequences of Land Use Change in the Upper Potomac

1
Assessing the Consequences of Land Use Change in
the Upper Potomac

• Robert H. Gardner
• with
• Jason Julian, Andrew J. Elmore,
• Todd R. Lookingbill, Marcella Suarez-Rubio

Appalachian Laboratory University of Maryland
Center for Environmental Science
2
The Appalachian Laboratory

• To determine the effects of natural and
  human-induced changes on organisms, landscapes,
  and biogeochemical and hydrological cycles.

3
Outline

• Importance of land-use and land-cover (LULC)
  change in the Potomac River Basin
• The challenge of determining effects
• An integrated approach for prediction

4
1. Importance of LULC change

• LULC is accelerating and is global in extent
• Directly linked with declines in
• Biodiversity
• Water quality and availability
• Ecosystem productivity (especially economically
  important species)
• LULC may also
• Accelerate climate change
• Enhance the spread of disease (new pandemics)

5
The Potomac River Basin

• Basin 38,000 km2
• Mainstem
• 617 km (170 km tidal)
• 6 physiographic provinces
• Climate boundary

6
Key attributes of the PRB

• Located in one of most rapidly urbanizing areas
  in the US
• 5.3 million people w/n basin
• Coal mining affects Appalachians
• Agriculture in Ridge and Valley
• Piedmont and Coastal Plain continue to be
  urbanized
• The 617 km river main stem has relatively
  unregulated flows

7
History of LULC change

• Not glaciated but glacial runoff produced
  coastal plain Chesapeake Bay
• Frontier stage (17th century)
• natural resource use, local deforestation
• Agricultural Expansion (18th century)
• Pops of 380,000
• 20-30 of forests cleared
• Sediment accumulation in Bay affect navigation

http//www.chesapeakebay.net/history.htm
8
More history

• Industrialization (late 18th century)
• Urban corridor formed
• Population of 2.5 million, raw sewerage in Bay
• Railroads consume 15-20 million acres of Eastern
  Deciduous Forest
• Population expansion (19th century)
• Beginning of environmental legislation and
  control (Clean Air Act, etc.)

9
Effects of LULC within the Potomac

• Hardened surfaces result in buried streams with
  increased throughput
• Nutrient retention declines, export increases
• Population growth increases water demands
• From Upper Potomac to Lower
• Ecosystem recovery from wide variety of
  disturbances remains unknown
• New invasives impact terrestrial and aquatic
  habitats

10
The problem of buried streams
11
The gradient of population density
12
Lower Potomac gtgt Upper Potomac
13
Low-flow correlated with high demand
112 year record shows 13 of years have extremely
low flows
Lookingbill et al., in press
14
Invasive organisms are here to stay

• Gypsy moth
• Hemlock wooly adelgid
• Chestnut blight
• New (potential)
• Emerald ash borer has been found in MD
• Sirex noctilio wasp (horntail) kills pines
• Sudden oak death
• Asian long horned beetle (in MD)

15
Potomac River Ecosystem has not been adequately
studied
River ISI References ISI References
Columbia Columbia 3,263
Mississippi Mississippi 2,921
Colorado Colorado 2,195
Hudson Hudson 1,193
Missouri Missouri 826
Potomac Potomac 309
16
2. Determining effects

• Landscapes are composed of many elements
  including
• roads
• agricultural units
• forests of diverse types and ages
• urban suburban development
• And diverse economic conditions

17
We know that the spatial arrangement of
elements is critical

• Riparian buffers effectively reduce sediment and
  nutrient export
• While development selectively removes headwaters
  ecosystems
• No single sub-watershed is representative of the
  Potomac
• Small critical areas (wetlands) are most
  effective nutrient and sediment filters

Elmore and Kaushal, 2008
18
Effects of LULC are not additive

• If linear then effects of change are additive
• we can extrapolate using mean value(s)
• landscape assessment can be produced by simple
  summation (spread sheet)
• or by sampling extremes (boundaries) and
  interpolating for each set of unique conditions

19
The problem of scale
Many challenges remain in extending our
understanding of how hydrologic processes within
small catchments scale to larger river basins.
20
Critical thresholds Brook trout density and
impervious cover
Stranko et al. 2008
21
Disturbance induces time lags

• Disturbances are not simple transient events?
• History of change is important
• We may not be able to predict the future from the
  past
• Forest harvesting has altered age and species
  distribution of flora
• Decline (possibly permanent declines) in oak and
  pine abundance

22
Significant effects on nutrient cycling
Eshleman et al. 2005
23
Permanent effects of coal mining
AMD
24
3. An integrated, predictive approach

• Understanding -gt prediction
• But this requires
• Spatial and temporal characterization of weather
  patterns
• Determination of trends in land use change
• A process-based representation considers
  interactive effects of multiple changes
• Estimation of unknowns and uncertainties

25
Interactive effects are important?

• Flood potential is a combined effect of LULC
  and climate change
• Denitrification depends on the location of
  critical habitat placement
• Sources and sinks
• Effectiveness of restoration
• Meeting water quality demands
• A moving target growth, development, LULC and
  climate change

26
SLEUTH a model of land-use change

• USGS sponsored development
• Slope, Land use, Exclusion, Urban extent,
  Transportation, Hillshade Clark (1998)
• Being explored and widely used w/n Chesapeake
  Watershed
• A pattern-based model
• Uses a fine-scale, gridded landscape
• Projects urban growth

27
Historical records are necessary

• Required GIS layers
• Urban growth (3-4 layers)
• Roads (2 layers)
• Exclusion (1 layer) protected lands
• Hillshade (1 layer)
• Slope (1 layer)
• Land use (1 layer) current

28
Empirical (best fit) of 5 growth parameters
govern probability of urbanization

• Spontaneous dispersion formation of new urban
  locations
• Growth (increase in size) of new urban locations
• Growth of old (established) locations
• Road gravity increased growth rates near roads
• Slope resistance decreased growth with
  increasing slope

Dietzel (2007)
29
Calibration first

• Brute force calibration (inefficient)
• Parameters varied over broad range
• Monte Carlo techniques applied
• Subset (best fit) determined by spatial
  comparison to history of change

30
Prediction requires

• Current land use maps for initialization
• Urban extent
• Transportation network
• Exclusion layer
• Future scenarios performed by varying
• Exclusion layer (e.g., streams, etc.)
• And exclusion rules
• Constraints on transportation network

31
Baltimore-Washington projections (Jantz et al.
2003)

• Three scenarios for piedmont coastal plain of
  Maryland and Northern Virginia
• Variable exclusion layers developed
• By state and land use type
• Scenarios
• A. Current trends
• B. Managed growth
• C. Ecological preservation

32
Results
Change (km2 / y) Change (km2 / y) Change (km2 / y)
Scenario Urban Forest Agriculture
A. Current trends 110 -43 -51
B. Managed growth 41 -15 -15
C. Ecological preservation 28 10 -9
33
Summary

• Moderate exclusions have large effects on
  patterns of LULC change
• Population trends continue to drive change
• Model improvements always desirable
• Local policies not yet implemented
• Linkage of land use projections with ecosystem
  models urgently needed
• Water, nutrients, sediments as a function of
  land-use change
• Biotic effects of land-use change

34
Our efforts for the Potomac

• We have spent 1 year on data acquisition,
  verification
• Calibration has been performed
• Simulations begun on development scenarios in the
  Upper Potomac
• Focus on effects of habitat change on bird
  community (Ph.D. thesis)

35
Next steps

• This summer
• Plans for a workshop at Appalachian Lab
• Include relevant parties using SLEUTH w/n
  Chesapeake
• Share mutual resources (data layers)
• Apply uniform methods for calibration and
  prediction for cross-comparisons
• Shared effort data enhancement, model
  improvement

36
Special thanks
Sujay Kaushal Walter Boynton Tom Fisher Larry
Sanford Jeff Cornwell Bill Dennison Clair Jantz