Evaluating Wetland Impacts on Nutrient Loads within Watershed

Systems Using AnnAGNPS

• Ronald L. Bingner, USDA-ARS, Oxford, MS• Jill Kostel, The Wetland Initiative, Chicago, IL• Jim Monchak, The Wetland Initiative, Chicago, IL• Yongping Yuan, USEPA, Las Vegas, NV• Henrique Momm, Middle TN State Univ.,

Murfreesboro, TN

Introduction

Nutrient losses to surface waters are of great concern on both national and regional scales.

Large areas of hypoxia in the northern Gulf of Mexico are due to excessive nutrients derived primarily from agricultural runoff via the Mississippi River.

Excessive pollutant loading is also responsible for algal blooms and associated water quality problems in lakes and rivers.

Nutrient Management Approaches

Fertilizer management include improved timing of N application at appropriate rates, using soil tests and plant monitoring to split applications, diversifying crop rotations, using cover crops (Dinnes et al., 2002).

Creation of wetlands, riparian buffers or other biofilters (Mitsch et al., 2001; Crumpton et al., 2007); and integrated wetland riparian buffers.

Improved drainage management practices (Skaggs et al., 2005).

Integrated drainage wetland systems.Etc.

Implement wetland & riparian buffer technology within watershed models for

an integrated watershed system conservation management approach

Integrating Wetland & Buffer Components within Watershed Models

Cell FCell E

Cell D

Cell CCell ACell B

watershed outlet

feedlot

gully

A wetland is located on reach 2, buffer in reach 1, more wetlands can be constructed on other reaches such as reach 1 and 3

0.136 0.082

1.411 1.661Dormant season

Max farm effluent load

0.04 0.05

0.55 1.08

N Pkg/ha

Outlet loadN Pkg/ha

Growing season

2004 – Ditch reduction capacity

N: 71% decreaseP: 40% decrease

N: 61% decreaseP: 35% decrease

• Modification of natural wetland receiving agricultural runoff by adding water control structures

Comprehensive sampling of vertebrates using these habitats revealed that sites with:

large area, woody vegetation, and permanent pool yielded 65 to 82 vertebrate species

smaller sites yielded only 11 to 20 species

Objective

Develop technology describing wetland and riparian buffer processes to evaluate their effectiveness towards reducing non-point source pollution leading to evaluations of integrated conservation practice within watershed systems.

AnnAGNPS(Annualized Agricultural Non‐Point Source) pollution model 

Agricultural Research Service

A partnering effort between the USDA:

andNatural Resource Conservation Service

AnnAGNPS• Developed as a tool to evaluate the integrated 

effect of all conservation practices within ungaged watersheds, particularly for USDA‐NRCS recommended practices, and identify sources of water quality pollutants for: 

• Water Quantity• Sheet & rill erosion• Gully erosion• Channel erosion• Chemical water quality• Buffers and wetlands

Simulation of Wetland Nitrogen Removal

Identify critical physical and chemical processes.

Develop appropriate algorithms for simulating these processes.

Defining needed parameters for model simulation.

Apply a mass balance approach is used to simulate hydrologic and water quality processes in the wetland.

Hydrology Balance

Where: Vi= amount of water per unit area in the wetland at the end of the day

(mm), Vi-1= amount of water per unit area in the wetland at the beginning of

the day (mm), Qinflow = amount of water added to the wetland during the day (mm), Qoutflow = amount of water released from the wetland during the day

(mm), P = precipitation on current day, from climate data information (mm), ET = daily evaporation or evapotranspiration (mm), I=daily infiltration, or seepage to the groundwater after inundation

(mm),

IETPQQVV outflowlowii inf1 )1(

Wetland Component Development

Nitrogen Loss Calculation

Where: J = N loss rate (g m-2 day-1), K20 = the area based first order loss rate coefficient (m/day),• C = the concentration of N in wetland (mg/L or g/m3),• θ = the temperature coefficient for N loss, and• T = water temperature (oC).

)20(20 ** TCkJ

A temperature dependent first-order model developed by Crumpton et al., (1997) is used:

)3(

Where:

• S = N loss from the wetland on a given day (kg),• J = N loss rate (g m-2 day-1), and• Awetland = wetland surface area (m2)

1000* wetlandAJS

Nitrogen Loss Calculation

)4(

Simulation Sequence

1) Simulate the amount of water and pollutant in the wetland at the beginning of a day.

2) Determine N loss during the day based on equations 3 & 4.

)20(20 ** TCkJ

1000* wetlandAJS )4(

)3(

Simulation Sequence 3) Determine the hydraulic head (H) is calculated

as a function of the water in the wetland:

)5(weirHVH 1000

Where:• H = hydraulic head on a given day (m), • V = amount of water per unit area in a wetland on a

given day (mm), and• Hweir = the height of a weir (m).

Simulation Sequence

4) Determine the amount of flow out of the wetland and the amount of N released with water is calculated using equations 6 & 7.

)6(5.1_ ** HLBQ outflowvolume

1000* _ outflowvolumeoutflow

outflow

QCM )7(

Where:• Qvolume_outflow = total volume of water flow out of the

wetland on a given day (m3),• B = the weir coefficient and it is determined by a user,• L = width of the weir opening (m), and

Simulation Sequence 5) the amount of water in the wetland is updated

using equation 1; and the amount of N in the wetland is updated using equation 2.

)2(

)1(IETPQQVV outflowlowii inf1

SMMMM outflowlowii inf)1(

6) The steps 1-5 are repeated for each day.

Large wetlands relative to the contributing drainage area can be most effective in improving water quality.

Highly sensitive to temperature and retention time impacted by flow.

Greatest nitrogen removal efficiency can be achieved during low flow conditions with a retention time of at least one-two weeks.

Additional water quality benefits can be obtained when wetlands are incorporated with riparian buffers.

Wetland Nitrogen Removal Efficiency

March 2013 Watershed‐Scale Characterization of Riparian Vegetation as Potential Filter Strips using Multi‐Source Remote Sensing  22

In‐field and edge‐of‐field buffer types 

Integration of Riparian Zones Within AGNPS

Riparian Buffer System

Stream

FieldDrainage with

transport of sediment & chemicals

Influence of natural and/or planted riparian (streamside) vegetative buffer strips

The amount of sediment/chemicals trapped by buffers is affected by:• Width of the riparian buffer• Type of vegetation cover• Amount and intensity of rainfall• Runoff characteristics• Local terrain slope

Trapping Efficiency of Grass Buffers < 5% slope

Trapping Efficiency of Grass Buffers > 5% slope

Trapping Efficiency of Forest Buffers

Buffer length needed for mitigation

Chlorpyrifos 184-230 mAtrazine 101-281 mMetolachlor 100-400 m

Methyl Parathion (plants) 18.8 mMethyl Parathion (no plants) 62.9 m

AGBUFAGNPS‐GIS Buffer Utility Feature

Provides buffer characterization for AnnAGNPS input parameters

AGBUFRiparian Buffer GIS representation

AnnAGNPS cells & reaches Buffer

AGBUF (continued)TE Estimation in AnnAGNPS Cells includes the effect of concentrated flow paths through buffers (short‐circuits)

A

0

0

0

1 2

3 4

5 6 TE = 97%

Potential short‐circuits

TE = 90%

TE = 47%

Varying widths & lengths

BIG BUREAU CREEK WATERSHED

Land Cover 92% agriculture

77% cultivated crops

BIG BUREAU CREEK WATERSHED

NUTRIENT REDUCTION

Nutrient Management Practices Nutrient management plans Conservation tillage practices Cover crops Prescribed grazing Fencing Grassed waterways and control structures

Nutrient Reduction Practices Riparian buffers Drainage water management systems Constructed wetlands

Sour

ce: U

SDA

ARS

Sour

ce: U

SDA

ARS

Constructed Wetland for Tile Drainage Wetland size and positioning key to nutrient removal Positioned to capture high nutrient loads Sized to allow adequate residence time (0.5 – 5%)

BBC POTENTIAL WETLAND SITES

Total Number of Wetland Complexes = 80

Total Wetland Area = 225 ha

Range of Wetland Sizes = 0.14 - 38.7 ha

Total Wetland & Buffer Area = 351 ha

Range of Wetland & Buffer Area = 0.54 – 46.5 ha

Area : Total BBC Watershed Area = 0.3%

% of Runoff of BBC Watershed = 23%

Model Input (Data) Climate – 30 year daily simulated weather

Topographic – 10 m DEM

Land use – 2009 Illinois Cropland Data Layer

Soil type – USDA Soil Survey Geographic database (SSURGO)

Point sources – Average Design Flow rates and estimated effluent concentrations

ANNAGNPS MODEL

Baseline Results:

Nutrients TN Load = 2,542,400 kg

TN Loss = 20.4 kg/ha

NO3-N = 72%

TP Load = 60,150 kg

TP Loss = 3.7 kg/ha

Attached P = 56%

Dissolved Inorganic P = 14%

TSS = 100,446 Mg

Point Sources TN = 0.70% increase

TP = 0.66% increase

Seasonal Effects Spring and summer

seasons higher loadings

ANNUAL AVERAGE LOAD DELIVERED

The 351 ha of wetlands plus buffer would reduce the entire BBC watershed TN and TP load by an average annual of 14% and 11%, respectively.

NO3-N accounted for 84% of the TN removed by potential wetlands plus buffer.

Dissolved inorganic P accounted for only 14-17% of TP removed by the potential wetlands plus buffer attached.

ANNAGNPSWETLAND & BUFFER FEATURE

ANNUAL LOAD DELIVERED-NITROGEN

_ g0 - 8.288

8.288 - 18.74

18.74 - 25.655

25.655 - 36.675

36.675 - 93.317

With Buffers and Wetlands Without Buffers and Wetlands

kg/ha/yr

Summary

• Technology has been developed within AnnAGNPS to evaluate wetland and riparian buffer effects on pollutant loads at a watershed‐scale 

• AnnAGNPS incorporates source load technology to assess wetland and buffer capability to reduce pollutant loads at any location within a watershed

Questions