Constructing Probability Surfaces of Ecological Changes in ... · Dennis Whigham, Karin Kettering, Melissa McCormick, Smithsonian Environmental Research Center Kirk Havens, Virginia

1

Constructing Probability Constructing Probability Surfaces of Ecological Surfaces of Ecological

Changes in Coastal Aquatic Changes in Coastal Aquatic Systems Through Systems Through

Retrospective Analysis of Retrospective Analysis of Phragmites Phragmites australisaustralis

Invasion and ExpansionInvasion and Expansion

Denice Heller Wardrop, Jessica PetersonDenice Heller Wardrop, Jessica Peterson--Smith, Mary Easterling, Smith, Mary Easterling, Hannah Ingram, Hannah Ingram, MuraliMurali HuranHuran and G.P. and G.P. PatilPatil Penn StatePenn StateDennis Whigham, Karin Kettering, Melissa McCormick, Dennis Whigham, Karin Kettering, Melissa McCormick,

Smithsonian Environmental Research CenterSmithsonian Environmental Research CenterKirk Havens, Virginia Kirk Havens, Virginia InsitutueInsitutue of Marine Scienceof Marine Science

2

Estuarine IndicatorsEstuarine Indicators

The Chesapeake Bay is highly monitored. While there are many indicators of changes in the Bay, it is difficult to link these indicators to what is occurring on land. This project is working to link these indicators to the land. The hope is that the researchers will be able to identify linkages and detail how those linkages occur.

3

The researchers sampled the Chesapeake Bay from north to south, focusing on estuarine segments. An estuarine segment is a watershed that is large enough to have a perennial stream flowing into the subestuaries of the Chesapeake Bay. A broad range of indicators that responded to land use patterns was identified. In particular, there appeared to be a strong association between an invasive plant species called Phragmites and land use.

The dots on the map represent the places where the researchers sampled the water for Phragmites. Estuarine segments were chosen using GIS data to represent the three dominant land use types: development, agriculture, and forest (reference condition).

4

r2 = 0.61, P < 0.001

% Developed Land in Watershed

0 20 40 60 80

Phra

gmite

sab

unda

nce

estim

ate

0

1

2

3

4

5

6

7

DevelopedMixed-DevAgriculturalMixed-AgForested

r2 = 0.61, P < 0.001r2 = 0.61, P < 0.001

% Developed Land in Watershed

0 20 40 60 80

Phra

gmite

sab

unda

nce

estim

ate

0

1

2

3

4

5

6

7

DevelopedMixed-DevAgriculturalMixed-AgForested

King et al. (2007) Estuaries and Coasts

There was a significant correlation between the amount of development on the watershed and the abundance of Phragmites.

5

0 10 20 30 40 50 60 70 80 90

0

1

2

3

4

5

6

7

Phra

gmite

s Ab

unda

nce

Inde

x

0.0

0.2

0.4

0.6

0.8

1.0

Cum

ulat

ive

Thre

shol

d Pr

obab

ility

41000004150000

42000004250000

43000004350000

0

1

2

3

4

5

6

7

Phra

gmite

s Ab

unda

nce

Inde

x

0.0

0.2

0.4

0.6

0.8

1.0

Cum

ulat

ive

Thre

shol

d Pr

obab

ility

% IDW Developed Land (Local)

25 35 45 55 65 75

0

1

2

3

4

5

6

7

Phra

gmite

sAb

unda

nce

Inde

x

0.0

0.2

0.4

0.6

0.8

1.0

Cum

ulat

ive

Thre

shol

d P

roba

bilit

y

% IDW Forested Land (Watershed)Northing (m)

PhragmitesAbundance Index

Model r2 = 73.4%

The three most important variables affecting Phragmites are: developed land; northing (location of the subestuary in the Chesapeake Bay), which has shown that the invasion front is moving from the north to the south in the Chesapeake Bay; and forested land.

The researchers identified a threshold using these indicators. The data show that only a small amount of development is needed to reach the threshold and, once a certain point is reached, change occurs very quickly.

Developed land is the most important variable related to the abundance of Phragmites. Although the amount of developed land along the watershed is important, where the developed land is occurring along the watershed also is very important. The closer the disturbance to the water, the stronger the relationship.

6

Luzula campestris (2.5)Centaurea jacea (4.33)

Several metres

Upper limit several metres under stormy conditions

Selective release mechanismsWind ballists

Cirsium dissectum (0.38)Eupatorium cannabinum (0.4)Narthecium ossifragum (0.6)Senecio paludosus (0.6)Valeriana dioica (0.6)Leontodon autumnalis (0.9)Holcus lanatus (1.1)Anthoxanthum odoratum (1.4)Filipendula ulmaria (1.7)Molinia caerulea (1.8)Cardamine pratensis (1.9)Succisa pratensis (2.14)

Several tens of metres

Upper limit ranges from at least several kilometres to at least several tens of metres under stormy conditions.

Low terminal velocity seeds (0.3 m/s < terminal velocity < ± 2 m/s) and relatively high release heights, often in combination with selective release mechanisms

Species with low terminal velocity seeds

Typha latifolia (0.1)Phragmites australis (0.1)Epipactis palustris (0.2)Liparis loeselii (0.2)Epilobium hirsutum (0.2)Epilobium palustre (0.2)Eriophorum angustifolium (0.2)Dactylorhiza species (0.3)Cirsium palustre (0.3) Aster tripolium (0.3)

Several kilometresUpper limit is at

least several kilometres under highly convective or stormy conditions

Low terminal velocity seeds (terminal velocity ≤ 0.3 m/s) and relatively high release heights, often in combination with selective release mechanisms

Species with low terminal velocity seeds

Representative species(terminal velocity in m/s)Dispersal distancesAdaptationsCategory

Source: Sooms (Applied Veg. Sci.)

Phragmites can spread rapidly and colonize a large area. In fact, their seeds can disperse up to several kilometers.

7

% IDW Developed Land (Watershed)0 15 30 45 60 75 90

2.00

2.25

2.50

2.75

3.00

3.25

3.50

Leaf

-tiss

ue %

N

0.0

0.2

0.4

0.6

0.8

1.0

Cum

ulat

ive

Thre

shol

d P

roba

bilit

y

>14.3%≤14.3%r2 = 65.3%

p=0.0178

The concentration of nitrogen in the leaves of the plant showed the same pattern. As more development close to the wetland occurs, the percentage of nitrogen in the leaves begins to increase rapidly. Thus, there may be a second important factor: the nutrient status of the ecosystem of the subestuaries.

The researchers initially hypothesized that there were two important factors in the establishment and spread of Phragmites: (1) a disturbance was needed to establish Phragmites in a new area, and (2) nutrient enrichment in the system was needed to allow Phragmites to spread.

8

Patterns of canopy-air CO2 concentration in a brackish wetland: analysis of a decade of measurements and the simulated effects on the vegetation.

Daniel P. Rasse, Stavroula Stolaki, Gary Peresta, Bert G. Drake Agricultural and Forest Meteorology 114 (2002) 59–73

9

ObjectivesObjectivesChoose an aquatic ecosystem with clearly identifiable alternative states, and define a limited number of variables that are considered to be the driving factors in state changesEstablish the database of explanatory and response variables over both a spatial and temporal extent. A retrospective analysis is the most powerful if performed over a truly temporal extent, instead of a “space for time” experimental design.Construct a probability surface of state change over the n-dimensional space of selected explanatory variablesDescribe thresholds in terms of the probability surface

10

Selection of Anne Arundel Selection of Anne Arundel County, MDCounty, MD

Abundance of Abundance of historical datahistorical dataRapid developmentRapid developmentInvaded/Invaded/uninvadeduninvadedmarshesmarshesExisting management Existing management infrastructureinfrastructure

11

Prioritized Slow VariablesPrioritized Slow VariablesAccumulating potential for an alternativeAccumulating potential for an alternative

ecosystemecosystem1. Flooding2. Organic vs. mineral soil3. Presence of non-native haplotype4. Genetic variation in population5. Drought6. Developed Land Cover7. Agricultural Land Cover8. Forested Land Cover9. Nitrogen from septic systems10. Nitrogen from point sources11. Hydrologic connectivity to Bay12. Salinity13. Atmospheric CO2

14. Atmospheric nitrogen deposition15. Phosphorus16. Precipitation and Climate

12

Prioritized Fast VariablesPrioritized Fast VariablesEvents which precipitate a Events which precipitate a

reorganization of the plant communityreorganization of the plant community1. Aerial distance to nearest Phragmites2. Water distance to nearest Phragmites3. Sedimentation4. Road/Bridge5. Tidal Restrictions/ Bulkhead6. Upland Fills/Construction7. Marsh modification (Dock, Boardwalk)8. Marsh surface water input9. Storms10. Fetch11. Species composition of marsh12. Ditches/Tidal Creek

13

Data SourcesData Sources

Complete inventory, classification, and mapping of Complete inventory, classification, and mapping of MD tidal marshes, 1970s (training set)MD tidal marshes, 1970s (training set)Aerial photographs for 1970, 1977, 1984, 1992, 2001 Aerial photographs for 1970, 1977, 1984, 1992, 2001 (preliminary analysis for (preliminary analysis for PhragPhrag expansion)expansion)Remote sensing data for 1984, 1992, 2001Remote sensing data for 1984, 1992, 2001Shoreline Survey for Maryland, completed in 2005 Shoreline Survey for Maryland, completed in 2005 (1441 shoreline miles surveyed)(1441 shoreline miles surveyed)

Immediate riparian zoneImmediate riparian zoneBankBankShorelineShoreline

14

PhragmitesPhragmites Invasion in South River Invasion in South River MarshesMarshes

1.89 ha

0.07 ha

• 1970: Present only in two marshes; < 2 ha 2.90 ha

0.11 ha

0.07 ha

0.14 ha

• 1977: Invades two additional marshes; expands to 3.2 ha

3.32 ha

0.04 ha 0.11 ha

0.33 ha

0.17 ha

0.04 ha

• 1984: Present in all six marshes; area increases to 4.1 ha

• 2000: Large increases in all marshes (38%-800% over 1984 areas); total area = 7.1 ha

4.60 ha

0.41 ha 0.26 ha

0.58 ha

0.98 ha

0.23 ha

15

Shoreline Situation ReportsShoreline Situation Reports

Immediate riparian area Immediate riparian area for land usefor land useHeight, stability, and Height, stability, and natural protection of natural protection of bankbankRecreational and access Recreational and access structures on shorelinestructures on shoreline

Red represents sites where shoreline surveys have been completed.

16

Rhode River. 1-km buffer in yellow; SERC sites in red; DNR wetland polygons in blue;VIMS Phrag in bright pink; VIMS unknown in pale pink: VIMS no Phrag in gray

17

Expansion of temporal/spatial Expansion of temporal/spatial data setdata set

Decision to Decision to ““recreaterecreate”” inventory, land cover, inventory, land cover, and shoreline survey for all available time and shoreline survey for all available time periods, proceeding from 2005 periods, proceeding from 2005 ““backwardsbackwards””Increases temporal Increases temporal datapointsdatapoints from 5 to 15from 5 to 15

18

CompleteCompleteComplete2005






IncompleteIncompleteIncomplete1988


IncompleteIncompleteIncomplete1980

CompleteIncompleteComplete1977

IncompleteCompleteComplete1970

IncompleteCompleteComplete1962/3


CompleteIncompleteIncomplete1952

CompleteIncompleteX1943

Curtis BaySouth RiverRhode RiverYear

Availability of Aerial PhotographyAvailability of Aerial Photography

19

Explanatory Variables from Explanatory Variables from Aerial PhotographyAerial Photography

Aerial distance to nearest Aerial distance to nearest PhragmitesPhragmitespopulationpopulationWater distance to nearest Water distance to nearest PhragmitesPhragmitespopulationpopulationLand coverLand coverShoreline disturbancesShoreline disturbancesSpecies composition of marshSpecies composition of marsh

20

pixels whichinclude marsh1 measure (1970’s)

% of species (plant community type) in pixel

Salinity, Flooding, Anoxia tolerance; Response to N;Aggressiveness

Species Composition ofMarsh

1) within 1 km of marsh, 2) within200 m of marsh, 3)within 10m ofmarsh

1) period of the aerial photo, 2) previous photos (periodof 10, 20, 50 years ago?)

1) % of land within acertain radius; 2) N load,3) Sediment load (typical load determined fromliterature)

nitrogen in surface H2O, sedimentation, flooding, salinity

Land Cover: Developed, Agricultural, Forested

pixels whichinclude marsh:in or within 10m(30m?) ofmarsh

aerial photo time step;time steps beforelinear m in pixel

sedimentation, flooding, salinity, dispersal

Road/Bridge, Upland Fill/Construction, Marsh Modification (Dock, Boardwalk)

pixels whichinclude marsh

aerial photo time step;time steps before

1) closest pixel withPhragmites, 2) % ofpixel in Phragmitescoverage

Increase dispersal; within marsh: decrease anoxia andflooding (clonalIntegration);

Aerial Distance to Nearest PhragmitesPopulation

Spatial ScaleTemporal PeriodMetric of MeasureVariables ImpactedVariable

Historical Photo Interpretation: Excerpt of Explanatory Variable Metrics, Temporal and Spatial Scale

21

Additional Historical DataAdditional Historical Data

SedimentationSedimentationSalinitySalinityPhosphorousPhosphorousDroughtDroughtPrecipitationPrecipitationClimateClimateFloodingFloodingMean Sea Level RiseMean Sea Level RiseMetonicMetonic cyclescyclesNitrogen Nitrogen -- Atmospheric deposition, point sources, Atmospheric deposition, point sources, septic systems, surface waterseptic systems, surface water

22

Curtis Bay and South RiverCurtis Bay and South RiverMarley Creek Curtis Bay

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

1970 1977 1984 1990 2000

Are

a of

Phr

agm

ites

(m^2

)

Furnace Creek Curtis Bay

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

1977 1984 1990 2000

Are

a of

Phr

agm

ites

(m^2

)

South River 1

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

1970 1977 1984 1990 2000

Are

a of

Phr

agm

ites

(m^2

)

South River 2

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

1 2 3 4 5

Are

a of

Phr

agm

ites

(m^2

)

23

Rhode RiverRhode RiverRhode River 11A

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

1 2 3 4 5

Are

a of

Phr

agm

ites

(m^2

)

Rhode River 11B

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

1 2 3 4 5

Are

a of

Phr

agm

ites

(m^2

)

Rhode River 12

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

1 2 3 4 5

Are

a of

Phr

agm

ites

(m^2

)

24

Curtis Bay sub-estuary, with the original SERC sites outlined in red, and the 1 km buffer of the shoreline outlined in yellow.

The researchers hypothesized that there was a correlation between nutrient increase and Phragmites increase, but they do not appear to be correlated.

25

Airport: Land designated or used for air traffic.Airport: Land designated or used for air traffic.

Commercial: Retail and office uses.Commercial: Retail and office uses.

Industrial: Industrial and industrial parks.Industrial: Industrial and industrial parks.

UtilityUtility: : Corridors defined by electrical line paths. May be represented bCorridors defined by electrical line paths. May be represented by cleary clear--cuts in noncuts in non--urban areas or cleared urban areas or cleared paths through urban areas.paths through urban areas.

Residential 1/8Residential 1/8--acreacre: : Single or Single or MulitMulit--Family Residential or Townhouses Family Residential or Townhouses -- 1/8 acre lot size.1/8 acre lot size.

Residential 1/4Residential 1/4--acreacre: : Single Family Residential Single Family Residential -- 1/4 acre lot size.1/4 acre lot size.

TransportationTransportation: : Highway, road and railroad right of way.Highway, road and railroad right of way.

Residential 1/2Residential 1/2--acreacre: : Single Family Residential Single Family Residential -- 1/2 acre lot size.1/2 acre lot size.

Residential 1Residential 1--acreacre: : Single Family Residential 1 Single Family Residential 1 -- acre lot size.acre lot size.

Residential 2Residential 2--acreacre: : Single Family Residential Single Family Residential -- 2 acre lot size.2 acre lot size.

Open SpaceOpen Space: : Open, Recreational, or vacant space maintained in turf.Open, Recreational, or vacant space maintained in turf.

Pasture/HayPasture/Hay: : Cultivated land used for pasture or hay.Cultivated land used for pasture or hay.

Row CropsRow Crops: : Cultivated land used for crops. Includes orchards.Cultivated land used for crops. Includes orchards.

WoodsWoods: : All upland forested areas.All upland forested areas.

WaterWater: : Open or standing water.Open or standing water.

Forested WetlandForested Wetland: : Lowland forest.Lowland forest.

Open Wetland: Emergent, floating aquatic or shrub wetlands.Open Wetland: Emergent, floating aquatic or shrub wetlands.

AA Co. Land Cover Categories

26

Land Cover Change in Ann Arundel Co.

-100-90-80-70-60-50-40-30-20-10

0102030405060708090

100

1950 1960 1970 1980 1990 2000 2010

Year

Perc

ent C

hang

e R

elat

ive

to 1

95

Commercial

Industrial

Open Space

Open Wetland

Pasture/Hay

Residential

Row Crops

Transportation

Water

Woods

27

RHODE RIVER

28

Land Cover Change in Ann Arundel Co.

0

100

200

300

400

500

600

700

800

900

1000

1950 1960 1970 1980 1990 2000 2010

Year

Hec

tare

s

Commercial

Industrial

Open Space

Open Wetland

Pasture/Hay

Residential

Row Crops

Transportation

Water

Woods

Land Cover Change in Rhode R. Sub-estuary

29

CLASSNAME2Transportation

Industrial

Commercial

Residential

Pasture/Hay

Row Crops

Open Space

Woods

Open Wetland

Water1952 2005

Landcover in Rhode River Sub-estuary

30

Total Length of Roads (km)

40.0

42.0

44.0

46.0

48.0

50.0

52.0

54.0

56.0

1950 1960 1970 1980 1990 2000 2010

Year

km o

f roa

ds

31

1952 2005

Roads in Rhode River Sub-estuary

32

Number of Shoreline Structures

0

50

100

150

200

250

300

350

400

1950 1960 1970 1980 1990 2000 2010

Year

No.

of s

truc

ture

s

33

1952 2005

Shoreline Structures in Rhode River Sub-estuary

34

Curtis Creek

35

Land Cover Change in Curtis Creek Sub-estuary

0

100

200

300

400

500

600

700

800

900

1000

1930 1940 1950 1960 1970 1980 1990 2000 2010

Year

hect

ares

Commercial

Industrial

Open Space

Open Wetland

Pasture/Hay

Residential

Row Crops

Transportation

Utility

Water

Woods

36

Percent Change in Land Cover in Curtis Creek Sub-estuary Relative to 1943

-150.0

-100.0

-50.0

0.0

50.0

100.0

150.0

200.0

250.0

1930 1940 1950 1960 1970 1980 1990 2000 2010

Year

Perc

en

Commercial

Industrial

Open Space

Open Wetland

Pasture/Hay

Residential

Row Crops

Transportation

Utility

Water

Woods

37

20051952

Curtis Creek Sub-estuary:Land Use Change 1952 vs. 2005

38

1943 1952

Curtis Creek Sub-estuary:Land Use Change During Initial 9-year Period of Record

39

Roads in Curtis Creek Sub-estuary

0.000

20.000

40.000

60.000

80.000

100.000

120.000

140.000

1930 1940 1950 1960 1970 1980 1990 2000 2010

Year

kilo

met

ers

Other roads

State & Federal roads

40

Roads – 1952 vs. 2005

Black roads were present in 1952; red roads have been built since that time

41

1943 1952

Curtis Creek Sub-estuary:Road Change During Initial 9-year Period of Record

42

Number of Shoreline Structures: 1943-2005Curtis Bay Sub-estuary

0

50

100

150

200

250

1930 1940 1950 1960 1970 1980 1990 2000 2010

Year

No.

of s

hore

line

stru

ctur

e

43

Shoreline Structures: 1952 vs. 2005Blue structures were present in 1952; red structures have been constructed since that time

44

0

10

20

30

40

50

60

70

1940 1950 1960 1970 1980 1990 2000 2010

Year

Perc

ent o

f lan

d ar

ea CC-Urban

CC-Ag

CC-Forest, Open, Wetl.

RR-Urban

RR-Ag

RR-Forest, Open, Wetl.

Land Use Comparison: Rhode River vs. Curtis Creek

45

Road DensityCurtis Creek vs. Rhode River Sub-estuaries

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

1940 1950 1960 1970 1980 1990 2000 2010

Year

km ro

ad p

er h

ecta

re

Rhode River

Curtis Creek

46

0

2

4

6

8

10

12

1940 1950 1960 1970 1980 1990 2000 2010

Year

# st

ruct

ures

/ km

sho

relin

e

RR-Structures/km

CC-Structures/km

Shoreline Structures: Rhode River vs. Curtis Creek

47

Annapolis and Baltimore Tidal Data:Annapolis and Baltimore Tidal Data:Data will need to be matched to elevation data to determine flooData will need to be matched to elevation data to determine flooding for marshesding for marshes

Baltimore and Annapolis Yearly Average MSL

-0.35-0.3

-0.25-0.2

-0.15-0.1

-0.050

0.050.1

0.15

1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Year

met

ers

rela

tive

to M

SL D

atum

AnnapolisBaltimore

MSL = mean sea level

48

Historic Water Quality Data, Yearly Averages:Historic Water Quality Data, Yearly Averages:Salinity and Sedimentation for Curtis Bay (Patapsco River), RhodSalinity and Sedimentation for Curtis Bay (Patapsco River), Rhode and South Rivere and South River

Salinity Yearly Averages

0

2

4

6

8

10

12

14

16

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Year

Salin

ity (p

pt)

PatapscoRhodeSouth

Total Suspended Solids Yearly Averages

0

5

10

15

20

25

30

35

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006Year

TSS

(mg/

L) PatapscoRhodeSouth

49

Yearly Averages TN with SE bars

0

0.5

1

1.5

2

2.5

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Year

TN (m

g/L) Patapsco

RhodeSouth

Yearly Averages TDP

0

0.02

0.04

0.06

0.08

0.1

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Year

TDP

(mg/

L) PatapscoRhodeSouth

Historic Water Quality Data, Yearly Averages:Historic Water Quality Data, Yearly Averages:Total nitrogen and total dissolved phosphorus for Curtis Bay (PaTotal nitrogen and total dissolved phosphorus for Curtis Bay (Patapsco River), Rhode and tapsco River), Rhode and

South RiverSouth River

50

Nitrogen Additions from Point Source DataNitrogen Additions from Point Source DataCurtis Bay and Patapsco River Point Source TN Discharges

0

2000

4000

6000

8000

10000

12000

14000

Jan-

84

Jan-

85

Jan-

86

Jan-

87

Jan-

88

Jan-

89

Jan-

90

Jan-

91

Jan-

92

Jan-

93

Jan-

94

Jan-

95

Jan-

96

Jan-

97

Jan-

98

Jan-

99

Jan-

00

Jan-

01

Jan-

02

Jan-

03

Jan-

04

Jan-

05

TN (K

g/D

ay)

COX CREEKERACHEM COMILOGFT. MCHENRY TUNNELISG SPARROWS PT.LEBANON CHEM.PATAPSCOPIT. D.M. STEELSEALAND SRVS.US GYPSUM COUSCG CURTIS CREEKW R GRACE

Rhode River Point Source TN Discharge

0

10

20

30

40

50

60

70

Jan-

84

Jan-

86

Jan-

88

Jan-

90

Jan-

92

Jan-

94

Jan-

96

Jan-

98

Jan-

00

Jan-

02

Jan-

04

TN (K

g/D

ay)

MAYO LARGE COMMUNALRIVERBOAT MOTEL

51

The area (m^2) of Phragmites at Rhode River site # 10 (Fox Creek). The logarithmic growth equation is used to interpolate between points. The points in yellow are actual data points, with the annual averages for TN for Rhode River (*1000).

Fox Creek Phragmites with Scaled TN

0500

1000150020002500300035004000

1980 1985 1990 1995 2000 2005 2010

Time

Are

a an

d TN

*100

00Scaled TNInterpolated PhragActual Phrag

52

South River Graphs TKNW: All South River Graphs TKNW: All StationsStations

South River TKNW

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

12/14/67 12/13/68 12/13/69 12/13/70 12/13/71 12/12/72 12/12/73 12/12/74 12/12/75 12/11/76 12/11/77 12/11/78 12/11/79 12/10/80 12/10/81

mg/

L

XGE6187

XGE6778

XGE7167

XGE7551

XGE7554

XGE7747

XGE8338

XGF1905

XGF4706

XGF4805

53

Patapsco River Graphs TKNW: All Patapsco River Graphs TKNW: All StationsStations

Nitrogen load is higher than in the South River area; Still no continuous records from any station

Patapsco TKNW

0

0.25

0.5

0.75

1

1.25

1.5

1.75

2

2.25

2.5

2.75

3

3.25

3.5

3.75

4

12/14/67 12/13/68 12/13/69 12/13/70 12/13/71 12/12/72 12/12/73 12/12/74 12/12/75 12/11/76 12/11/77 12/11/78 12/11/79 12/10/80 12/10/81

mg/

L

PAT0176

XIE2293XIE2885

XIE3380

XIE5260

XIE5343

XIE6747

XIF1126

54

CART-based Probability Surface

X1<50

X1<15 X2<7

X2<10

X1<40

E FA

B

C D

Y N

Y

Y

Y

Y NN

N

N 15 50

10

7

40

A

B

C D

E

F

X1

X2

55

Susceptibility versus DispersalSusceptibility versus Dispersal

Statistical model is for susceptibilityStatistical model is for susceptibilityAssumed dispersal was not limitingAssumed dispersal was not limitingDispersal and spread mechanisms investigated Dispersal and spread mechanisms investigated separatelyseparately

Karin Kettering; CO2 and nutrient enrichment, Karin Kettering; CO2 and nutrient enrichment, seed production and viabilityseed production and viabilityMelissa McCormick, genetic fingerprinting of Melissa McCormick, genetic fingerprinting of populationspopulations

56

Preliminary observations Preliminary observations

Land cover does not adequately describe Land cover does not adequately describe disturbances relevant to disturbances relevant to PhragmitesPhragmites invasioninvasionSignificant interactions and lag times in a Significant interactions and lag times in a multivariate settingmultivariate settingChange in susceptibility state may have Change in susceptibility state may have occurred previouslyoccurred previously

57

Future StepsFuture Steps

Historical photoHistorical photo--interpretation of land cover, interpretation of land cover, roads, and shoreline structures completed by roads, and shoreline structures completed by midmid--JuneJuneHistorical photoHistorical photo--interpretation of interpretation of PhragmitesPhragmitesexpansion by midexpansion by mid--AugustAugustStatistical Statistical modellingmodelling through summer, through summer, completed by 2008completed by 2008

58

DiscussionDiscussion

A participant asked for confirmation that he had correctly interpreted the graph showing a one-third meter rise in sea level over the past few decades. Dr. Whigham confirmed that this was correct.

One participant asked if Dr. Whigham and his colleagues had examined the percent land cover as a potential indicator. In the participant’s research, he found that percent land cover was important with respect to the amount of freshwater entering the watershed. Dr. Whigam replied that he and his colleagues have found that the non-native genotype Phragmites flourish in brackish conditions; one of the reasons Phragmites are becoming established and spreading is that they do not require freshwater.

A participant asked if the Phragmites could be eradicated. Dr. Whigham responded that in Maryland there are no restrictions on the eradication of Phragmites.

One of the participants asked Dr. Whigham to define metonic cycle. According to Dr. Whigham, a metonic cycle is a drought cycle. It occurs approximately every 10 to 15 years in the area. Dr. Whigham and his colleagues have not yet been able to link this cycle to any of the patterns they have seen.

59

Discussion (Continued)Discussion (Continued)

A participant asked if there was concern about insects as predators in the development and spread of Phragmites. Dr. Whigham responded that he did not know how insects might affect Phragmites development and spread. He did not know of any research on this topic.

One participant asked if the graphs on the first slides were on a watershed scale or a 1-kilometer scale. Dr. Whigham responded that those slides were on the scale of the entire estuarine segment. The participant asked if the later data shown were on a scale of 1 kilometer by 1 kilometer. Dr. Whigham confirmed that it was and explained that the two are not very different. The strongest predictor the researchers have seen to date is land use within 500 meters of the watershed.

Constructing Probability Surfaces of Ecological Changes in ... · Dennis Whigham, Karin Kettering, Melissa McCormick, Smithsonian Environmental Research Center Kirk Havens, Virginia

Documents