Assessment and restoration of riparian
processes in urban watersheds
Peter M. Groffman
Cary Institute of Ecosystem Studies
and
The Scientists of the NSF funded Baltimore
Long-Term Ecological Research Project, the
Baltimore Ecosystem Study (BES).
Outline:
• Urbanization effects on riparian zones.
• Are detention basins the new riparian zone?
• Does stream and riparian restoration have a
nitrogen benefit.
• Evaluating riparian zones at the watershed scale.
RIPARIAN ZONES:
• Critical interface between terrestrial and
aquatic components of a watershed.
• Demonstrated ability to prevent pollutant
movement from upland land uses into
streams.
• Most work on groundwater nitrate, in
agricultural watersheds.
• Urban stream
syndrome:
Natural Channel
Water Table
Stream
Channel with Incision
Due to Increased
Runoff
– High storm flows.
– Incised channels.
– Drier riparian zones
with lower water
tables.
•Channel Erosion
•Nonfunctional Floodplain
•Dry Riparian Soils
Urban stream syndrome:
Urban stream syndrome results in drier soils and
lower water table in riparian zone:
0
Jan-00
Jan-02
Jan-04
Water table depth (mm)
-200
-400
-600
-800
-1000
Suburban (Gwynnbrook)
-1200
Forested (Pond Branch)
Jan-06
Jan-08
Jan-10
Jan-12
Urban stream syndrome results in higher
groundwater nitrate in riparian zone:
Nitrate (mg N/L)
2.4
1.6
0.8
0
Urban (Cahill)
Suburban
(Glyndon)
Suburban
(Gwynnbrook)
Forested (Pond
Branch)
Denitrification
NO3-  NO2-  NO  N2O  N2
- Anaerobic
- Heterotrophic (requires organic C)
• Expect high rates in wetland soils.
• Key component of the water quality
maintenance function of riparian zones.
Nitrification
NH4+  NO2-  NO3- Aerobic
- Substrate driven (need high NH4+)
• Important internal source of nitrate in an
ecosystem.
• Symptom of N richness or N saturation.
Nitrate (mg N kg-1)
0
2
4
6
8
10
0
10
20
Depth
30
40
50
60
70
80
90
Source: Groffman et al. (2002)
Urban
Forested reference
Suburban (downstream)
Suburban (headwaters)
Nitrification (mg N kg-1 d-1)
-0.2
0
0.2
0.4
0
10
20
Depth
30
40
50
60
70
80
90
Source: Groffman et al. (2002)
0.6
0.8
1
1.2
1.4
Are detention basins the new riparian zone?
• Stormwater
structures
(detention basins)
are engineered to
mitigate impact of
impervious surfaces
on stream discharge.
• Their impact on
nutrients (especially
nitrogen) in unclear.
Can you find the detention basins in this
suburban landscape?
½ mile x ¾ mile area
Source: Neil Bettez
Can you find the detention basins in this
suburban landscape?
½ mile x ¾ mile area
Source: Neil Bettez
Storm water Management Structures
in the Gwynns Falls
Type
Structure Type
(TN removal efficiency)
Count
Percent
of total
Percent of
Drainage
Area
Area
weighted
removal
A
Wet Ponds and Wetlands (30%)
23
3%
1
0.33
Type of Structure
shallow marsh
Retention Pond
Bay Separator
B
Dry Detention & Hydrodynamic
structure (5%)
Oil & Grit Separator
275
33%
10
0.48
Still Basin
Underground storage
Detention Pond
C
Dry Extended Detention (30%)
272
33%
8
2.39
Dry Ext Detention Pond
Ext Det Pond
Porous Pavement
D
Infiltration Practices (50%)
90
11%
0.34
0.14
Swale
Infiltration Trench
Infiltration basin
Dry Well
E
Filtering Practices (40%)
167
20%
1.79
0.72
BIO-Retention
Sand filter
Total
827
100%
21% (?)
4.06%
(Stormwater workgroup: BMP Pollutant Removal Efficiencies.PDF )
Denitrification potential higher in detention basins
than in natural riparian zones:
Denitrification Potential (mg N kg-1 hr-1)
2.5
2.0
1.5
1.0
0.5
0.0
A
B
C
D
E
SM
DP
EDSD
IB
SF
Bettez and Groffman (2012)
Herb
Fors
Riparian
Is there a
nitrogen benefit
to stream and
riparian
restoration?
Restored Urban Stream: Early Stage
Restored Urban Stream: Developed Stage
Stable, unconnected floodplain
“ConnectedFloodplain
Push
Pull
Soil Surface
Water Table
Push-Pull Method
1.
Pump ground water
2.
Amend with 15NO3- ,Br-, SF6
3.
Lower DO to ambient levels.
4.
Push into mini-piezometer
5.
Incubate for 4 hours
6.
Pull from mini-piezometer
Plume
Addy et al. (2002)
Denitrification Rate (μg/N/kg
soil/day)
Denitrification higher in restored, connected
riparian zones:
300
250
200
Unrestored High Bank
Unrestored Low Bank
Restored High Un-connected Bank
Restored Low Connected Bank
150
100
50
0
June 2003
Kaushal et al. (2008)
November 2003
June 2004
Nitrate concentrations lower in restored hyporheic
zones:
Nitrate Concentration (mg/L)
Hyporheic Ground Water
2
1.6
1.2
0.8
0.4
0
June
November
Site
Kaushal et al. (2008)
Unrestored
Restored
Mass Removal of Nitrate-N (μg/L)
Nitrate dynamics improved in restored hyporheic
zones:
Unrestored (Coinciding Measurements)
Restored (Scenario)
8000
6000
R2 = 0.87
4000
R2 = 0.70
2000
0
0
2
4
6
8
10
-2000
Groundwater Residence Time (Days)
Kaushal et al. (2008)
Can stream restoration create
denitrification “hotspots” on
the floodplain?
• Floodplain wetlands can be created
deliberately or incidentally during
stream restoration projects.
• These wetlands have the potential to
serve as floodplain NO3- sinks if:
• They process a significant amount of
water.
• With significant residence time.
• With high denitrification.
• With high plant (algae, macrophyte)
uptake.
Minebank Run
oxbow wetlands:
Oxbow 1, Seepage flow wetland
• Created incidentally
during stream
restoration.
• Vary in nature and extent
of connectivity to the
stream.
• Measurements of
hydrology, denitrification,
15N uptake by plants and
algae, storm dynamics.
Harrison et al. (2011, 2012a,b, 2014)
Oxbow 2, Surface flow wetland
Can we assess
riparian condition
at the watershed
scale:
Rte 165
N
30m buffer
T1
Right bank
T2
T3
• Ground-truth map: 34 drainage classes
• Does SSURGO reflect
this complexity?
PD
VPD
VPD
Stream flow
PD
Rte 165
SPD
SPD
MWD
Left bank
Source: Rosenblatt et al. (2001)
Original Scale of GIS Data Base Can Alter
Inputs & Outputs to Watershed Models
5%
Proportion of stream length bordered by hydric soils
Source: Rosenblatt et al. (2001)
75%
Scale/type of spatial data
can mask or display
pathways and sinks
(data: Kingston Quad-RI)
Streams (1:24,000)
Ponds (1:24,000)
Forest / Open Space
Agriculture
Residential (low density)
Residential (med density)
Res. (med high density)
Institutional
Gravel pits
Source: Groffman et al. (2009)
National Wetland
Inventory (1:24,000)
displays potential sinks
Streams (1:24,000)
Ponds (1:24,000)
Forest / Open Space
Agriculture
Residential (low density)
Residential (med density)
Res. (med high density)
Institutional
Gravel pits
NWI Wetlands (1:24K)
Source: Groffman et al. (2009)
SSURGO Hydric Soils
suggest wetlands and
zero order streams
connect source to
stream
Streams (1:24,000)
Ponds (1:24,000)
Forest / Open Space
Agriculture
Residential (low density)
Residential (med density)
Res. (med high density)
Institutional
Gravel pits
NWI Wetlands
Source: Groffman et al. (2009)
Hydric Soils (SSURGO)
(1:15,840)
High resolution stream
data and hydric soils
display an active
biogeochemical
landscape
Streams (1:5,000)
Ponds (1:5,000)
Forest / Open Space
Agriculture
Residential (low density)
Residential (med density)
Res. (med high density)
Institutional
Gravel pits
NWI Wetlands
Source: Groffman et al. (2009)
Hydric Soils (SSURGO)
(1:15,840)
Detecting riparian and stream condition
with LiDAR:
Source: Andy Miller
Key issues going forward:
• Urbanization disrupts riparian processing,
converting riparian zones from sinks to sources of
nitrogen.
• Stormwater control measures may restore the
“lost” riparian effect.
• Stream and riparian restoration, especially those
that “reconnect” streams and riparian zones
should have a nitrogen benefit.
• Riparian evaluations at the watershed scale:
– New tools such as SSSURGO and LiDAR.
– Need monitoring at the watershed scale.
• Surface water has higher
NH4+ (mg/L)
nitrate concentrations
than groundwater and
streamwater – suggests
oxbows are “sinks” for
nitrate.
A
NO3- (mg/L)
Minebank Run oxbow
wetlands:
• Surface water has lower
PO4- (μg/L)
phosphate
concentrations than
groundwater or
streamwater, suggests
oxbows are “sources” of
Harrison et al. (2012)
phosphate.
GW
WC
SSW
GW
WC
SSW
GW
WC
SSW
B
C
Denitrification in urban floodplain
wetlands:
• Measured in-situ denitrification summer and winter
2008 using the 15N-push pull method.
• Analyzed gas samples for N2O and N2
Constructed storm
water wetlands
Oxbow “relict”
wetlands
Denitrification in urban floodplain
wetlands:
• Denitrification
rates were high –
a significant sink
for nitrate.
• Rates were higher
in winter than in
summer.
• Denitrification
similar across all
wetland types.
Harrison et al. (2011)
Oxbow
“Relict”
Constructed Reference
N2O yield during denitrification in urban
floodplain wetlands: Oxbow “Relict” Constructed Reference
A
• N2 was the dominant
end-product at all
sites.
• N2O yield (N2O:N2
ratio) was low at all
sites.
Harrison et al. (2011)
B
15N
mass-balance approach in
oxbow “relict” wetlands:
• Determine the fate of 15N added to
wetlands:
• Algae
• Macrophyes
• Sediments
• Denitrification (unaccounted for
N)
• Two additions:
• Mid growing season
• Very early (almost dormant)
growing season
15N
mass-balance approach in oxbow
“relict” wetlands: Methods
• Water column samples, algae, macrophytes,
and sediment were collected before and after
the six-day experiment to determine 15N
natural abundance.
•
15N-KNO 3
and Br- solution added summer
2009 and spring 2010.
• Water samples were collected six consecutive
days in six discrete locations (inlet to outlet) in
each oxbow.
15N
200
• Complete mixing of
15N
Bromide
160
14
12
µg/L
10
120
8
15N
80
6
• All added 15N was
4
40
2
0
processed by day 3.
0
0
1
2
3
Days since 15N additions
20
Oxbow 2
Bromide
15
7
6
5
10
4
15N
µg/L
8
15N
3
5
2
1
0
0
0
1
Days since
2
15N
3
additions
Harrison et al.
Bromide (mg/L)
added 15N by day 2.
Oxbow 1
Bromide (mg/L)
mass-balance approach in oxbow
“relict” wetlands: Results
15N
mass-balance approach in oxbow
“relict” wetlands: Fate of 15N
23
57
19
97
13
38
39
24
87