identification of suitable sites for constructed wetlands ... · 3. wetland is on cropland crep...

CREP Wetland Final Report August 31, 2009

1

Identification of Suitable Sites for

Constructed Wetlands to Remove Nitrate

Final Project Report to the Indiana State Department of Agriculture

Dr. Jane Frankenberger, Agricultural and Biological Engineering, Purdue University

Dr. Indrajeet Chaubey, Agricultural and Biological Engineering, Purdue University

Dr. Eileen Kladivko, Agronomy, Purdue University

Margaret McCahon, Graduate Research Assistant, Purdue University


2

Contents Introduction .................................................................................................................................................. 3

Criteria Used for Wetland Siting and Design ........................................................................................ 5

Data Layers .................................................................................................................................................... 6

Locating Suitable Sites .................................................................................................................................. 8

1. Sufficient Watershed Area from Tile-drained Land ....................................................................... 8

2. Wetland may be Located at the Interface between Closed and Open Drains ............................ 10

3. Wetland is on Cropland ............................................................................................................... 10

4. Topography Lends itself to Wetland Placement ......................................................................... 11

Wetland Design ........................................................................................................................................... 13

5. Desired Wetland Size is 0.5-2% of its Watershed Area ............................................................... 13

6. No More than 25% of the Wetland is Greater than Three Feet Deep ......................................... 14

7. The Surrounding Buffer Must Extend Four Feet above Wetland Surface, and Should Not Exceed

Four Times the Size of the Wetland .................................................................................................... 14

An Example Wetland Design (Site 8) ................................................................................................... 14

Applying Wetland Design Criteria to Wetland Designs ...................................................................... 15

Nitrate Removal Estimation ........................................................................................................................ 17

Conclusions ................................................................................................................................................. 21

References .................................................................................................................................................. 22

Appendix ..................................................................................................................................................... 23

I. Locations where Suitable Wetlands are Designed ...................................................................... 23

II. Wetland Design Characteristics and Nitrate Removal Efficiencies ............................................. 24

III. Wetland Designs at Each Site ...................................................................................................... 25

CREP Wetland Final Report

Intensification of agricultural practices in the Midwest has led

runoff and subsurface drainage, potentially impacting the water quality downstream and hypoxia in the

Gulf of Mexico. In heavily tile drained land,

losses in the form of nitrate. The 200

thirty percent of its five-year running average by 2015

reduction will require significant nutrient load reduct

Midwest.

Constructed wetlands have been shown to be an effective practice to reduce nitrate load

Midwestern crop land. Strategically targeting sites that intercept high nitrate loads and sizing th

wetlands according to the characteristics of their watersheds can maximize wetland efficiency while

minimizing costs and maintaining productive agriculture (Crumpton, 2001).

Determining suitable wetland sites for effective nitrate removal begins as a cla

combining various GIS layers to determine potentially suitable sites. An innovative aspect of this GIS

analysis is calculating the watershed area

which ensures that these wetlands intercept large flows and maximize nitrate r

landscape. This is the concept that forms the basis for Iowa’s CREP program (USDA, 2001; Tomer et al.,

2003).

Figure 1: Wetlands funded by Iowa’s CREP program treat drainage water from

nitrate removal, (Photo from http://www.fsa.usda.gov/Internet/FSA_File/crepphotogallery.ppt

August 31, 2009

3

Introduction

Intensification of agricultural practices in the Midwest has led to increased nutrient losses in surface


Gulf of Mexico. In heavily tile drained land, characteristic of parts of Indiana, there are great nitrogen

s in the form of nitrate. The 2008 EPA action plan calls for a reduction of the hypoxic zone to

year running average by 2015 (Gulf Hypoxia Action Plan, 2008)

reduction will require significant nutrient load reductions from large contributing areas throughout the

Constructed wetlands have been shown to be an effective practice to reduce nitrate load

Midwestern crop land. Strategically targeting sites that intercept high nitrate loads and sizing th


minimizing costs and maintaining productive agriculture (Crumpton, 2001).

Determining suitable wetland sites for effective nitrate removal begins as a classic GIS problem,


analysis is calculating the watershed area (also called the contributing area) draining to each location,

lands intercept large flows and maximize nitrate reduction

This is the concept that forms the basis for Iowa’s CREP program (USDA, 2001; Tomer et al.,

Figure 1: Wetlands funded by Iowa’s CREP program treat drainage water from at least 500 acres, to maximize

http://www.fsa.usda.gov/Internet/FSA_File/crepphotogallery.ppt)

August 31, 2009

to increased nutrient losses in surface


there are great nitrogen

EPA action plan calls for a reduction of the hypoxic zone to about

(Gulf Hypoxia Action Plan, 2008). Such a

ions from large contributing areas throughout the

Constructed wetlands have been shown to be an effective practice to reduce nitrate loads leaving

Midwestern crop land. Strategically targeting sites that intercept high nitrate loads and sizing the


ssic GIS problem,


to each location,

eduction in the

This is the concept that forms the basis for Iowa’s CREP program (USDA, 2001; Tomer et al.,

at least 500 acres, to maximize

)


4

The watershed used for this analysis was the Middle Wabash-Little Vermillion 8-digit HUC, limiting

suitable sites to within Indiana (see Figure 2). The goals of this work were to (1) complete a GIS analysis

to target suitable locations for constructed wetlands, (2) perform a site-specific analysis to create

preliminary designs at each suitable site and estimate the nitrate removal achieved for each design, and

(3) document the methodology clearly and completely so that it may be applied in other watersheds.

Figure 2: Middle Wabash – Little Vermillion Watershed (HUC 05120100) used for the GIS analysis.


5

Criteria Used for Wetland Siting and Design

We used a number of criteria in the siting and placement of wetlands (see box). We developed these

criteria based upon the Iowa CREP criteria, knowledge of Indiana agricultural practices, and discussions

with the following people:

� Staff of the Indiana State Department of Agriculture, Division of Soil Conservation, at a meeting

in Indianapolis on March 9, 2009

� Several people involved in wetland siting for the Iowa CREP program from Iowa State University

and the Iowa Department of Agriculture and Land Stewardship

� The State CREP Steering Committee and Technical Committees, at a meeting in Indianapolis on

June 1, 2009

Wetland Siting Criteria

1. Wetland has sufficient watershed area from tile-drained land (500-2000 acres of tile-

drained land, exclude streams to reduce permitting needed)

2. Wetland may be located at the interface between closed and open drains

3. Wetland is on cropland

4. Topography lends itself to wetland placement

Wetland Design Criteria

5. Desired wetland size is 0.5-2% of its watershed area

6. No more than 25% of the wetland is more than 3 feet deep (“deep wetland”)

7. The surrounding buffer must extend four feet above wetland surface, and should not

exceed 4 times the size of the wetland


6

Data Layers

Publicly available data layers were used in this analysis, so that the procedure could be continued on all

other Indiana watersheds. These datasets are described in Table 1 and shown for the 8-digit watershed

in Figure 3. An ArcInfo license of ArcGIS (ESRI, 1999-2006) and Arc Hydro tools (ESRI, 2007) were used

for the GIS analysis.

Table 1: Datasets used in the analysis

Data layer Source dataset How the layer was used in the analysis

Streams National Hydrography Dataset

(NHD), high resolution streams,

downloaded June 2009

Used to eliminate streams as suitable locations for

placing wetlands, and to approximate the location

where closed drains empty into open drains

Elevation National Elevation Dataset (NED),

with one-third arc second

resolution

Used to find watersheds, describe how water

flows through the landscape, and create contours

for wetland design

Roads TIGER 2008 Census data, “edges” Used to break up separate crop fields and also for

locating different wetland sites

Cropland National Land Cover Dataset

(NLCD), selecting the field

“cultivated crops”

Used for locating suitable sites for wetland

placement and also to approximate tile-drained

land in the landscape

Tile drained

land

(estimate)

From SSURGO soils and Cropland

(poorly, very poorly, and somewhat

poorly drained cropland)

Used to locate sites that drain a large watershed of

tile-drained land

Hydric soils Soil Survey Geographic (SSURGO)

Database

To give a general idea of areas that were formerly

wetland

Orthophotos Indiana Framework Data, streamed

by University of Indiana

Used for visualization of sites and determination of

locations where closed drains empty into open

ones


7

Figure 3: Data layers used in the identification of suitable locations for placement of constructed wetlands

in the 8-digit watershed.


8

Locating Suitable Sites

1. Sufficient Watershed Area from Tile-drained Land

To make the greatest impact in the landscape, each wetland should be placed strategically so that it

intercepts a large flow of water from tile-drained land, which is the source of much of the nitrate in

Indiana landscapes. We targeted these locations through the following three steps: (1) finding locations

that have watersheds between 500 and 2000 acres in size; (2) narrowing these locations to only those

that drain at least 500 acres of tiled land; and (3) excluding those locations that are in streams, rivers, or

open ditches to reduce permitting issues.

Step (1) was completed by processing 30-foot grid elevation data in Arc Hydro to determine flow

accumulation, or the paths where water should flow in a landscape. Locations with very large

watersheds appear like line segments or paths because water flow is channelized. Step (2) was similar

to step (1), except a weighted flow accumulation was used that only takes into account flow originating

from tile drained land. In Step (3) the NHD Streams dataset was buffered 100 meters to account for

discrepancies between stream and flow accumulation datasets, and all locations within this region were

excluded from the analysis. The results of each step are shown in Figure 4.

After Step (1), there were 1029 paths that had watersheds of 500-2000 acres. Step (2) narrowed the

sites considerably, with only 431 paths remaining, generally located near headwaters at the edges of the

watershed, and not in the Wabash River valley. The portion of the 8-digit watershed in Illinois was not

considered for wetland placement. Step (3) narrowed the number of sites to 105, about 10% of those

from Step (1), which were located almost exclusively in the northern portion of the watershed. It

appears that in the south (Parke and adjacent counties) the stream network is denser, perhaps because

of hillier terrain or decreased tile drainage of this land.


9

(a): Step (1) (b): Step (2)

(c): Step (3)

Figure 4: Determining

locations that have sufficient

watershed area from tile

drained land. (a) locations

with large (500-2000 acre)

watersheds, (b) use of an

estimate of tile drained land to

narrow these locations to

target high nitrate flows, and

(c) use of the NHD streams

data to eliminate locations

that are found in open

waterways.


10

2. Wetland may be Located at the Interface between Closed and Open Drains

For each flow path found in the previous step there is a location where the path meets an open

waterway. Results of our analysis in the pilot watershed in Tippecanoe County using the county’s

Regulated Drains data layer showed that this is typically where a large underground tile main (“closed

drain”) empties into a stream or open ditch (“open drain”). Placing wetlands at the interface between

closed and open drains is the preferred location, rather than interrupting a closed drain, since the water

leaving the wetland is a combination of flows above and below ground and is potentially difficult to

route back through a closed tile. Locating the wetland downstream in an open drain (ditch or stream)

would require permitting from various regulatory agencies and therefore is not desired.

3. Wetland is on Cropland

CREP constructed wetlands must be placed on cropland. To determine suitable cropped fields for

wetland placement, we delineated the watersheds of desirable outlet points and intersected these

watersheds with the cropland. Then we selected those fields that contain a location with sufficient

watershed area. The final output looks like the example shown in Figure 5, where the bright green field

is the only site where a wetland could be placed. A total of 113 crop fields were found, though not all

were truly suitable and were narrowed further in the next steps.

Figure 5: The output of the methods for siting wetlands. First an outlet point is found where the locations with

sufficient watershed area (red line) empty into an open stream (blue line). Next the watershed for this outlet

point is delineated (blue and dark green areas). Then the watershed and cropland layers are combined, and any

crop field that contains a location with sufficient watershed area is selected as potentially suitable for wetland

placement.


11

Note: The three steps described above were completely automated, using ArcGIS and Arc Hydro tools.

The remaining steps required human judgement, and thus cannot be completely automated.

4. Topography Lends itself to Wetland Placement

We looked carefully at each potentially suitable cropland field to determine, qualitatively, if topography

lends itself to wetland placement. We created 1-foot elevation contours and used them to visualize

wetland designs. Orthophotos and the streams data layer were used to more accurately determine the

interface between closed and open drains. The keys to good topography are a small dam, indicated by

dense contours, and a large and steep topographic rise of over 4 feet for buffer placement. Another

way to think of it is that the wetland and buffer should form a bowl-shape, where the wetland is

relatively flat and the surrounding buffer is steep. An example of a good site is shown in Figure 6.

Figure 6: An example of topography that lends itself to wetland placement. The orthophotos are used to

confirm that land is indeed cropland, and also to find the actual interface between closed and open drains. The

wetland and buffer shapes are determined by dam placement and topography.

We ranked each of the 113 sites qualitatively for wetland suitability, using the ranking scheme shown in

Table 2. We also noted whether or not the wetland could be placed at the interface between closed

and open drains. Results of the suitability ranking are shown in Figure 7. There were 11 sites given rank

5, 21 given rank 4, and 17 were ranked 3. The 49 sites ranked 3-5 were taken to the next stage of

placing preliminary wetland designs where possible.

Dam location


12

Table 2: Ranking Cropland Fields for Wetland Suitability

Wetland Suitability

Ranking

Qualitative Definition

5: Quite suitable Topography easily defines wetland and buffer with a relatively small dam, where

buffer does not appear more than 4 times the size of the wetland

4 A dam can be placed that allows for wetland and buffer features

3 Some element of topography or surrounding features make wetland placement

difficult, but not impossible

2 Wetland placement is impossible without considerable terrain alteration

1: Least suitable There is absolutely no possibility of a wetland at this location (probably due to

small size of field or surrounding features)

0: Invalid This plot of land does not meet wetland siting criteria, and was selected

erroneously in the automated method

Figure 7: Suitability Rankings on all potential sites found in part 3 of the analysis.


13

Wetland Design

We created preliminary wetland designs on all sites with suitability rankings of 3-5 where wetland

placement was possible. For each site, we created hypothetical “dams” to indicate the outlet of the

wetland, and then used contours to create the shape of wetland features for various heights of the dam.

A design was not considered valid unless the wetland and buffer contours intersect the dam to form

closed polygons. Topography and other features limited the number of feasible designs to one or two at

most sites. We made designs at a total of 31 sites. After creating all possible designs for each site, we

compared these designs quantitatively based on the wetland placement criteria discussed below, and

narrowed the suitable locations to 19 sites (Figure 8; Appendix).

Figure 8: Left: Suitable wetland designs were created at 19 sites. See Appendix for numbered sites. Right: The

watersheds of each potential wetland shows the drainage area that would be treated by the wetlands, which

account for three percent of the entire tile-drained portion of the 8-digit watershed.

5. Desired Wetland Size is 0.5-2% of its Watershed Area

Wetlands must be large enough to treat the volume of water they receive, but larger wetlands have

higher costs. The size of the wetland depends on the size of its watershed, the amount of drainage

water, and the degree of treatment desired. In Iowa, the desired size for these wetlands is 0.5-2% of its

watershed area. Yet Indiana generally has higher rainfall, which could produce higher flows in Indiana,

and wetlands should be designed accordingly. For this analysis we aimed for the desired wetland range

of 0.5-2% of its watershed area. This means a wetland with a 1,000 acre watershed should be 5 to 20

acres in size, with a buffer surrounding it up to 80 acres. However, designs with a higher percentage are

included as well for they may still hold value as they would result in a higher level of treatment

(although at a greater cost).


14

6. No More than 25% of the Wetland is Greater than Three Feet Deep

The primary process of nitrate reduction in a wetland is usually denitrification by bacteria, which takes

place most efficiently under oxygen depleted conditions (usually less than 3 ft deep). At the same time

water should be shallow enough in parts of the wetland for the wetland plants to establish. Therefore,

most of a wetland should be less than three feet deep to achieve nitrate treatment efficiency. We used

the criteria that no more than 25% of the pond should be greater than three feet deep, the same as in

the Iowa CREP program. As described above, this means that the ideal site would be fairly flat in the

specific area where the wetland pool is located, but sloping fairly steeply just outside the wetland pool

to minimize buffer size.

7. The Surrounding Buffer Must Extend Four Feet above Wetland Surface, and Should Not

Exceed Four Times the Size of the Wetland

When a wetland is installed, the local water table level will be the wetland surface elevation. If the

surface of a wetland is above a nearby tile then the tile will be submerged, hindering proper drainage.

Therefore the land surrounding the wetland surface should not be farmed, but rather converted to a

buffer of vegetation that is tolerant to wet conditions. In Indiana, tiles are usually located about three

feet below grade, so we aimed to design wetlands that are about four feet below the surrounding land.

This means that the buffer strip extends from the edge of the wetland to the farmable land, which is a

four-foot rise.

A four-foot rise in elevation around a wetland basin is a rather large topographic feature in the

landscape in this part of Indiana. Many locations do not even have such a rise, and others may rise so

gradually that the buffer strip is excessively large relative to the size of the wetland. In Iowa the desired

size of a buffer strip was no more than four times the size of the wetland, and this is also the criterion

we used. Note that if the farmer wishes to square off the buffer strip to give it a regular and more

farmable shape, then the buffer will be larger than the minimum necessary buffer region that we

estimated in these designs.

An Example Wetland Design (Site 8)

The goal of the wetland design process was to determine ways to place the wetland in the landscape to

meet the three design criteria of (5) wetland size of 0.5 to 2% of watershed area, (6) no more than 25%

greater than 3 feet deep, and (7) 4-foot elevation buffer with area no more than 4 times the wetland

size. We created each design by selecting the contour that defines the wetland surface, then selecting

the contour four feet above this which is the necessary buffer, and finally selecting the contour three

feet below the wetland (if one exists) that will become deep wetland (see Figure 9). The first contour

considered was one of the lowest contours that intersected the dam. This contour probably would not

allow for a large enough wetland, and in this case we move on to the next lowest contour (Figure 9).

The design is considered “valid” if the contour four feet above the wetland wraps around the wetland

and intersects the dam on both sides on the outlet. The next design we create has the wetland

extending to the contour one foot above the previous design, and if a buffer can be made, this design is

considered valid as well. After a few designs there will be a point where the buffer no longer surrounds

the wetland and a wetland of the proposed size can be not be placed at the site (see Figure 9, Figure

10).


15

All valid designs were compared against the wetland design criteria. We have documented those

designs that met (or nearly met) the design criteria (see Appendix).

Figure 9: The process of creating preliminary wetland designs based on 1-foot contours and a hypothetical dam

location.

Applying Wetland Design Criteria to Wetland Designs

The design criteria were then applied to the example wetland designs (Table 3). All but the smallest

wetland are within the desired range of wetland size; all wetlands have a small enough portion of deep

wetland, and all have a reasonable buffer. The dam length and the associated cost increases with the

larger designs. This shows that this example wetland truly has some flexibility in its design, which is not

characteristic of most wetland sites, which have one or two possible designs. Therefore, this location is

one of the most suitable for wetland placement in the entire watershed. A similar analysis for each of

the 19 suitable locations for wetland placement is presented in the Appendix, along with maps of each

design.

“Wetland” (usual water level)

“Deep wetland” (>3 feet deep)

“Buffer” (4 feet above wetland)

“Dam” (used to create the polygons)

Locations with sufficient

watershed area

Orthophotos

1-foot contours

Lowest contour would

not make large

enough wetland

This is the highest

contour for the buffer

with this particular

dam, because it

intersects the dam on

both sides of the

wetland outlet


16

Figure 10: All possible designs for this example site. The large design is only possible when the dam is adjusted

to intercept the buffer contour.

Table 3: Applying the Wetland Placement Criteria to the example wetland, designed in Figure 10.

Component Desired Range Smallest Small Medium Large

Wetland 0.5-2% of watershed 0.4% of

watershed

0.8% of

watershed

1.2% of

watershed

1.3% of

watershed

Deep wetland < 25% of wetland area 22% 14% 2% 0%

Buffer Less than 4:1 ratio

buffer : wetland

2:1 ratio 2:1 ratio 2:1 ratio 4:1 ratio

Dam As small as possible 1200 ft 950 ft 825 ft 775 ft

Notes

Smallest Small Medium Large

0.4% of watershed 0.8% of watershed 1.2% of watershed 1.3% of watershed

4 acres 8 acres 11 acres 12 acres


17

Nitrate Removal Estimation

Nitrate removal takes place by plant uptake and microbial processes in oxygen-poor environments such

as wetlands. A number of factors affect the rate of nitrate removal, including hydraulic loading

rate/hydraulic retention time, the concentration of nitrate in the inflow water, the temperature of the

water, soil conditions, vegetation processes, and flow characteristics in the wetland.

Because so many variables affect nitrate removal in a wetland, we chose to use two wetland nitrate

removal models developed in climates and conditions similar to Indiana. The first model is an annual

model developed on Iowa CREP wetlands (Crumpton et al., 2006), in which nitrate removal depends on

annual hydrulic loading rate and the average concentration of nitrate in the wetland. This regression

was shown to predict nitrate removal quite well in Iowa CREP wetlands. The second model was

developed in North Carolina to predict how large constructed wetlands should be to achieve a certain

nitrate reduction (Burchell et al., 2007). We used this model to predict nitrate reduction on a monthly

timescale, and then determined annual reduction. The model estimates effluent nitrate concentration

as a function of inflow concentration, surface area of the wetland, an empirically determined nitrate

removal constant that varies with water temperature, the depth of water in the wetland, the wetland

sediment porosity, and the volume of inflow. See equations in box below.

Annual Model (Crumpton et al., 2006): � �� .��

N Removed = annual percent of nitrate removed from wetland (%)

HLR = annual hydraulic loading rate to the wetland (m/yr)

Monthly Model (Burchell et al., 2007): �� !�� " �

#

Ce = effluent nitrate concentration (mg/L)

Co = influent nitrate concentration (mg/L)

As = surface area of wetland (m2)

Knitrate = nitrate removal rate constant (1/month)

We used -0.45 in winter and -5.7 in summer

y = depth of the water in wetland (m)

n = porosity of wetland (0.65-0.75)

We used 0.70

Q = inflow (m3/month)


18

Applying these models to hypothetical wetland designs required estimates of annual and monthly flow

and average nitrate concentrations of the flow entering the wetland. Actual flow and nitrate

concentration data from these locations are not available. We used data from 2000 to 2002 in Hoagland

Ditch, located in White County, Indiana, which has similar land use and drainage characteristics to the

wetland watersheds analyzed in this project. Monthly data on nitrate concentration and ditch flow are

shown in Figure 11. The annual flow weighted average nitrate concentration in Hoagland Ditch was

approximately 10 ppm, so the results shown are for this concentration. It should be noted that the

actual nitrate removal efficiency will be different by these wetlands as some wetlands may receive a

higher nitrate concentration than 10 ppm, and in this case they will show greater efficiency than those

reported here.

The estimated annual wetland efficiency for all suitable wetland designs is shown in Figure 12, and the

estimated monthly effluent nitrate concentration is shown in Table 4. All designed wetlands are

predicted to remove approximately 25 to 45 percent of the nitrate they receive each year. Yet nitrate

removal depends heavily on the time of year, with high removal rates in the warm months (April

through September) and low removal rates in the winter. Therefore, the removal efficiency will vary

based on timing of nutrient application as well as timing of flow events in the watershed.

Figure 11: Nitrate concentration and flow data from Hoagland Ditch used in nitrate removal estimates for each

wetland design.

0.00

5.00

10.00

15.00

20.00

1 2 3 4 5 6 7 8 9 10 11 12

Month

Nitrate Concentration (mg/L)

Ditch Flow (cm)


19

Figure 12: Estimated nitrate removal efficiency for each wetland design using the annual model. Efficiencies

range between about 25 and 45%. Bars are colored only to show the difference between adjacent wetland

design sites.

0.00 10.00 20.00 30.00 40.00 50.00

01a

02a

03a

03b

04a

05a

06a

07a

07b

08a

08b

08c

08d

09a

10a

10b

10c

10d

10e

11a

11b

11c

12a

13a

14a

14b

15a

16a

17a

18a

19a

19b

19c

19d

Annual Percent Nitrate Removal (%)

We

tla

nd

De

sig

n


20

Table 4: Estimated monthly nitrate inflow and outflow concentrations for each wetland design.

Inflow 11.6 12.8 15.2 13.6 15.1 17.3 11.8 0.9 0.7 3.4 7.4 11.1

Wetland

No. Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

01a 10.8 12.5 14.6 9.5 7.4 10.5 0.1 0.0 0.4 3.4 6.6 10.7

02a 10.8 12.6 14.7 9.8 7.8 10.9 0.1 0.0 0.4 3.4 6.6 10.7

03a 11.0 12.6 14.8 10.5 9.0 12.1 0.3 0.0 0.4 3.4 6.8 10.8

03b 11.2 12.7 14.9 11.5 10.8 13.7 1.0 0.0 0.5 3.4 7.0 10.9

04a 10.8 12.6 14.6 9.7 7.7 10.8 0.1 0.0 0.4 3.4 6.6 10.7

05a 10.9 12.6 14.7 10.3 8.5 11.6 0.2 0.0 0.4 3.4 6.7 10.8

06a 10.8 12.5 14.6 9.6 7.5 10.6 0.1 0.0 0.4 3.4 6.6 10.7

07a 11.2 12.7 15.0 11.8 11.2 14.1 1.4 0.0 0.5 3.4 7.0 10.9

07b 10.7 12.5 14.5 9.0 6.6 9.7 0.0 0.0 0.3 3.4 6.5 10.6

08a 10.5 12.5 14.4 8.2 5.4 8.5 0.0 0.0 0.3 3.4 6.3 10.5

08b 10.7 12.5 14.6 9.3 6.9 10.1 0.0 0.0 0.3 3.4 6.5 10.7

08c 11.1 12.6 14.9 11.3 10.3 13.2 0.7 0.0 0.5 3.4 6.9 10.9

08d 11.4 12.7 15.1 12.7 13.0 15.6 4.1 0.1 0.6 3.4 7.2 11.0

09a 11.0 12.6 14.8 10.7 9.3 12.3 0.3 0.0 0.4 3.4 6.8 10.8

10a 9.0 12.0 13.2 3.9 1.2 3.0 0.0 0.0 0.1 3.2 4.9 9.8

10b 9.9 12.3 13.9 6.1 3.0 5.6 0.0 0.0 0.2 3.3 5.7 10.2

10c 10.6 12.5 14.5 8.6 6.0 9.1 0.0 0.0 0.3 3.4 6.4 10.6

10d 11.1 12.6 14.8 10.8 9.5 12.5 0.4 0.0 0.5 3.4 6.8 10.8

10e 11.4 12.7 15.1 12.5 12.6 15.2 3.2 0.1 0.6 3.4 7.2 11.0

11a 11.1 12.6 14.9 11.1 10.1 13.0 0.6 0.0 0.5 3.4 6.9 10.9

11b 10.0 12.3 14.0 6.4 3.3 6.0 0.0 0.0 0.2 3.3 5.8 10.3

11c 10.6 12.5 14.5 8.9 6.4 9.5 0.0 0.0 0.3 3.4 6.4 10.6

12a 11.3 12.7 15.0 11.9 11.5 14.3 1.6 0.0 0.5 3.4 7.1 10.9

13a 11.1 12.6 14.8 11.0 9.8 12.8 0.5 0.0 0.5 3.4 6.9 10.9

14a 11.2 12.6 14.9 11.3 10.3 13.3 0.8 0.0 0.5 3.4 6.9 10.9

14b 11.4 12.7 15.1 12.7 13.1 15.6 4.1 0.1 0.6 3.4 7.2 11.0

15a 10.7 12.5 14.5 9.0 6.5 9.6 0.0 0.0 0.3 3.4 6.5 10.6

16a 11.3 12.7 15.0 11.8 11.4 14.2 1.5 0.0 0.5 3.4 7.0 10.9

17a 11.4 12.7 15.1 12.8 13.3 15.8 4.6 0.2 0.6 3.4 7.2 11.0

18a 11.2 12.7 14.9 11.5 10.7 13.6 0.9 0.0 0.5 3.4 7.0 10.9

19a 11.5 12.7 15.1 13.1 13.9 16.3 6.6 0.3 0.6 3.4 7.3 11.1

19b 11.0 12.6 14.8 10.6 9.0 12.1 0.3 0.0 0.4 3.4 6.8 10.8

19c 10.2 12.4 14.2 7.2 4.2 7.1 0.0 0.0 0.2 3.3 6.0 10.4

19d 9.2 12.1 13.4 4.2 1.4 3.4 0.0 0.0 0.1 3.2 5.1 9.9


21

Conclusions

Using publicly available data and GIS analysis, we found 31 locations in the 8-digit watershed where

wetlands may be placed to intercept high nitrate loads. In all, 64 designs were created at these 31 sites.

Of these 31 locations, 19 have at least one reasonable wetland design based on the wetland placement

criteria, and 34 wetlands are designed at these 19 sites. Many of these sites are located at the interface

between open and closed drainage.

All suitable sites are located in the northern part of the watershed, which is flatter and more extensively

tiled than the south, which has a denser stream network. The best locations for wetlands are in the

headwaters of streams, often near the outer edges of the 8-digit watershed. Topography greatly limits

wetland placement on a given site, because the four-foot elevation drop in the buffer is difficult to find

in the relatively flat landscapes.

The total watershed of these 19 locations is 21,650 acres, with 16,081 acres estimated to be tile drained

land. If wetlands are placed at all of these 19 locations, they will intercept three percent of the entire

tile drained portion of the watershed.

The estimated nitrate removal efficiency of these 34 wetlands varies from about 25 to 45 percent, with

an average of 35 percent. Efficiencies may be greater if the average incoming nitrate concentration is

higher than 10 ppm, which is likely in heavily tile drained watersheds.

The most critical step in this analysis is to target locations with a large watershed area draining into the

wetland, to maximize nitrate reduction. Any GIS analysis must consider wetland placement from a

watershed-scale perspective, and the use of GIS in the placement of these wetlands is an important part

of a strategy to reduce nitrate loads in the agricultural Midwest.


22

References

Burchell, M.R., Skaggs, R.W., Lee, C.R., Broome, S., Chescheir, G.M., Osborne, J. (2007). Substrate

organic matter to improve nitrate removal in surface-flow constructed wetlands. Journal of

Environmental Quality, 36:194-207.

Crumpton, W.G. (2001). Using wetlands for water quality improvement in agricultural watersheds: The

importance of a watershed scale approach. Water Science and Technology, 44:559-564.

Crumpton, W.G., Stenback, G.A., Miller, B.A., Helmers, M.J. (2006). Potential benefits of wetland filters

for tile drainage systems: Impacts on nitrate loads to Mississippi River subbasins. Report to USDA.

ESRI (1999-2006). ArcGIS 9.2. ArcInfo license.

ESRI (2007). Arc Hydro for ArcGIS 9 Version 1.2. Environmental Systems Research Institute, Redlands,

California, USA.

Mississippi River/Gulf of Mexico Watershed Nutrient Task Force (2008). Gulf Hypoxia Action Plan 2008

for Reducing, Mitigating, and Controlling Hypoxia in the Northern Gulf of Mexico and Improving Water

Quality in the Mississippi River Basin. Washington, DC.

Tomer, M.D., D.E. James, and T.M. Isenhart (2003). Optimizing the placement of riparian practices in a

watershed using terrain analysis. J. Soil Water Conservation, 58(4): 198-206.

USDA, 2001. Conversation Reserve Enhancement Program: Iowa enhancement program. Available at

http://www.agriculture.state.ia.us/waterResources/pdf/LandownerGuide.pdf.


23

Appendix

I. Locations where Suitable Wetlands are Designed

Potentially suitable wetland designs were created at the sites shown above. Designs for each of these

sites are shown in the following pages.


24

II. Wetland Design Characteristics and Nitrate Removal Efficiencies

Wetland

No. Watershed Wetland Buffer Deep Dam

Average

Depth Wetland

Buffer:

Wetlan

d

Deep

Wetland

NO3

Removal

acres acres acres acres ft ft % of

Watershed ratio

% of

Wetland

% NO3

removal

01a 619 8.32 18.84 1.10 502 1.4 1.34 2.3 13.2 31

02a 1640 26.90 116.65 0.00 1646 1.1 1.64 4.3 0.0 45

03a 1604 9.32 32.14 3.34 637 2.4 0.58 3.4 35.8 32

03b 1604 6.47 26.63 2.16 617 2.2 0.40 4.1 33.4 28

04a 1119 21.60 41.95 0.00 508 0.9 1.93 1.9 0.0 42

05a 788 9.01 26.63 0.00 554 1.3 1.14 3.0 0.0 31

06a 889 16.99 88.14 0.00 835 1.0 1.91 5.2 0.0 39

07a 1056 8.81 28.18 0.00 1381 0.9 0.83 3.2 0.0 31

07b 993 27.67 74.31 0.00 1688 0.8 2.79 2.7 0.0 45

08a 944 12.60 30.49 2.78 1182 2.0 1.33 2.4 22.1 35

08b 944 11.26 23.72 1.56 946 1.7 1.19 2.1 13.8 34

08c 944 7.97 17.97 0.19 825 1.2 0.84 2.3 2.4 30

08d 944 3.97 15.66 0.00 775 0.9 0.42 3.9 0.0 24

09a 1407 15.96 87.45 0.00 1212 1.1 1.13 5.5 0.0 38

10a 694 21.84 34.52 6.76 1169 2.1 3.15 1.6 30.9 42

10b 694 14.42 32.83 3.88 1102 2.1 2.08 2.3 26.9 37

10c 694 10.51 28.14 1.27 1045 1.6 1.52 2.7 12.1 33

10d 694 6.76 21.56 0.00 967 1.3 0.97 3.2 0.0 29

10e 694 3.88 17.96 0.00 895 0.9 0.56 4.6 0.0 24

11a 1191 11.03 48.60 0.00 752 1.2 0.93 4.4 0.0 34

11b 1191 26.30 91.23 5.21 925 1.8 2.21 3.5 19.8 45

11c 1191 17.61 64.59 1.98 814 1.5 1.48 3.7 11.2 39

12a 1023 6.29 20.30 0.00 666 1.2 0.61 3.2 0.0 28

13a 931 9.90 41.22 0.00 1959 1.1 1.06 4.2 0.0 32

14a 1590 12.55 41.49 0.00 814 1.3 0.79 3.3 0.0 35

14b 1590 7.40 34.63 0.00 726 0.8 0.47 4.7 0.0 29

15a 672 26.01 96.48 0.00 2146 0.6 3.87 3.7 0.0 45

16a 1251 10.18 41.28 0.00 647 0.9 0.81 4.1 0.0 33

17a 1551 9.07 34.40 0.00 1268 0.6 0.58 3.8 0.0 31

18a 1033 17.05 66.53 0.00 2470 0.6 1.65 3.9 0.0 39

19a 654 4.81 21.59 0.00 383 0.3 0.73 4.5 0.0 25

19b 654 11.14 21.42 0.00 469 0.8 1.70 1.9 0.0 34

19c 654 15.53 24.27 0.00 532 1.4 2.37 1.6 0.0 38

19d 654 21.37 26.69 4.81 605 1.9 3.27 1.2 22.5 42


25

III. Wetland Designs at Each Site

Site 1 Wetland Designs

Watershed of the Site 1

wetland

Site 1 is located in Fountain

County.

95% of the watershed is

estimated to be tile

drained.

Site of wetland placement

The best location for a

wetland is shown by the

black proposed dam. This

location DOES appear to be

located near the interface

between closed and open

drains. The interface of

closed and open drains lies to

the north. The dam could be

placed anywhere along the

red line or the blue line to

alter the designs.


26

Component Desired Range Design A

Wetland 0.5-2% of watershed 1.3% of watershed

Deep wetland < 25% of wetland area 13%


buffer : wetland

2.3:1 ratio

Dam As small as possible 500 ft

Notes Satisfies all desired criteria

Design 1A:

Wetland is 1.3%

of watershed


27



wetland

Site 2 is located in Fountain

County.



drained.








drains. The dam could be


red line or along the blue

line between the red line

and the dam to alter the

designs.


28





buffer : wetland

4.3:1 ratio


Notes Buffer is a bit large, and dam

is fairly long.

Design 2A:

Wetland is 1.6%

of watershed


29



wetland

This site is located in

Vermillion County.



drained.





location appears to be



drains, but intercepts open

drainage. The dam could

be placed anywhere along

the red line or along the

blue line between the red

line and the dam to alter

the designs.


30

Component Desired Range Design A Design B

Wetland 0.5-2% of watershed 0.58% of watershed 0.40% of watershed

Deep wetland < 25% of wetland area 36% 35%


buffer : wetland

3.4:1 ratio 4.1:1 ratio

Dam As small as possible 637 ft 617 ft

Notes Wetland is a bit too deep Wetland is smaller than

desired, and a bit too deep

Design 3A:

Wetland is 0.58%

of watershed

Design 3B:

Wetland is 0.40%

of watershed


31



wetland


Montgomery County.



drained.










red line to alter the designs.

The topography lends itself

to wetland placement.


32





buffer : wetland

1.9:1 ratio


Notes This site appears to be quite

suitable for wetland

placement

Design 4A:

Wetland is 1.9%

of watershed


33



wetland


Benton County.



drained.





location DOES NOT appear

to be located near the

interface between closed

and open drains. The dam

could be placed anywhere

along the red line to alter

the designs.

The topography lends itself

to wetland placement.


34





buffer : wetland

3:1 ratio


Notes This site appears to be quite

suitable for wetland

placement

Design 5A:

Wetland is 1.1%

of watershed


35



wetland


Tippecanoe County.



drained.












36





buffer : wetland

5.2:1 ratio


Notes The buffer is larger than

desired

Design 6A:

Wetland is 1.9%

of watershed


37



wetland


Fountain County, though

the watershed extends into

Tippecanoe.



drained.


The best locations for a

wetland are shown by the

black proposed dam. These

locations DO appear to be





red line or along the blue

line between the red line

and the dam on the left to

alter the designs.


38





buffer : wetland



Notes Long dam Long dam, wetland may be

larger than desired

Design 7A:

Wetland is 0.83%

of watershed

Design 7B:

Wetland is 2.8%

of watershed


39



wetland


Tippecanoe County.



drained.


The best locations for a

wetland are shown by the

black proposed dams.

These locations DO appear

to be near the interface




red line.

Topography lends itself to

wetland placement. This

may be one of the most

promising sites.


40

Design 8A:

Wetland is 1.3%

of watershed

Design 8B:

Wetland is 1.2%

of watershed

Design 8C:

Wetland is 0.84%

of watershed

Design 8D:

Wetland is 0.42%

of watershed


41





buffer : wetland



Notes

Component Desired Range Design C Design D




buffer : wetland



Notes Wetland is smaller than

desired


42



wetland


Tippecanoe County.



drained.












43





buffer : wetland

5.5:1 ratio


Notes The buffer is larger than

desired

Design 9A:

Wetland is 1.1%

of watershed


44



wetland


Montgomery County.



drained.


The best locations for a wetland

are shown by the black proposed

dams. These locations DO appear

to be near the interface between

closed and open drains. There

may be no other possible dam

location.

There may be features preventing

the placement of wetlands. A site

visit may be necessary. Otherwise

this is a promising site.


45

Design 10A:

Wetland is 3.1%

of watershed

Design 10B:

Wetland is 2.1%

of watershed

Design 10C:

Wetland is 1.5%

of watershed

Design10D:

Wetland is 1.0%

of watershed


46





buffer : wetland



Notes Wetland is larger than

desired, with more deep

wetland than desired

Component Desired Range Design C Design D Design E

Wetland 0.5-2% of watershed 1.5% of watershed 1.0% of watershed 0.6% of watershed

Deep wetland < 25% of wetland area 12% 0% 0%


buffer : wetland

2.7:1 ratio 3.2:1 ratio 4.6:1 ratio

Dam As small as possible 1045 ft 967 ft 895 ft

Notes Buffer is larger

than desired

Design 10E:

Wetland is 0.6%

of watershed


47


Watershed of the Site 11 wetland

This site is located in Tippecanoe

County.


estimated to be tile drained.




dams. These locations DO NOT

appear to be near the interface

between closed and open drains.

The dam could be placed

anywhere along the red line.


48

Component Desired Range Design A Design B Design C

Wetland 0.5-2% of watershed 0.9% of watershed 2.2% of watershed 1.5% of watershed

Deep wetland < 25% of wetland area 0% 20% 11%


buffer : wetland

4.4:1 ratio 3.5:1 ratio 3.7:1 ratio

Dam As small as possible 752 ft 925 ft 814 ft

Notes Buffer is larger

than desired

Design 11A:

Wetland is 0.9%

of watershed

Design 11B:

Wetland is 2.2%

of watershed

Design 11C:

Wetland is 1.5%

of watershed


49



wetland


Tippecanoe County.



drained.





location DOES NOT appear

to be located at the


and open drains, but

instead intercepts closed

drainage. The dam could

be placed anywhere along

the red line to alter the

designs.


50





buffer : wetland

3.9:1 ratio


Notes Wetland is larger than

desired

Design 12 A:

Wetland is 3.5%

of watershed


51



wetland


Tippecanoe County, though

the watershed enters

Montgomery County.



drained.





location appears to be

somewhat near the


and open drains, which is

to the north of the site.


anywhere along the red line

to alter the designs.


52





buffer : wetland

4.2:1 ratio


Notes Buffer is larger than desired,

dam is quite long, a patch of

trees interferes with

preliminary design. This site

may be more promising than

it appears from this design.

Design 13 A:

Wetland is 1.1%

of watershed


53




County.






dams. This location DOES appear


closed and open drains. The dam

could be placed anywhere along

the red line.


54





buffer : wetland



Notes Buffer is a bit

large

Design 14A:

Wetland is 0.8%

of watershed

Design 14B:

Wetland is 0.5%

of watershed


55




County.






dams. This location DOES NOT






56





buffer : wetland

3.7:1 ratio


Notes Wetland is much

larger than

desired, and dam

is quite large

Design 15A:

Wetland is 3.9%

of watershed


57




County.












58





buffer : wetland

4.1:1 ratio


Notes

Design 16A:

Wetland is 0.8%

of watershed


59




County.










the red line.


60





buffer : wetland

3.8:1 ratio


Notes

Design 17A:

Wetland is 0.6%

of watershed


61



This site is located in Montgomery

County.










the red line.


62





buffer : wetland

3.9:1 ratio


Notes Dam is quite large,

and topography

makes wetland

design difficult.

Design 18A:

Wetland is 1.6%

of watershed


63



wetland


Tippecanoe County.



drained.









There may be features preventing

the placement of wetlands. A site

visit may be necessary. Otherwise

this is a promising site.


64

Design 19A:

Wetland is 0.7%

of watershed

Design 19B:

Wetland is 1.7%

of watershed

Design 19C:

Wetland is 2.4%

of watershed

Design19D:

Wetland is 3.3%

of watershed


65





buffer : wetland



Notes Channel would need to be

dug to empty wetland

Channel would need to be


Component Desired Range Design C Design D




buffer : wetland



Notes Channel would need to be


identification of suitable sites for constructed wetlands ... · 3. wetland is on cropland crep...

Documents