Balancing Tile Drainage for Crop Production

and Environmental Impact

Kevin W. King USDA-ARS

Columbus, OH

FarmSmart Agricultural Conference Rozanski Hall, University of Guelph

January 18, 2014

Rationale

Data from Heidelberg University (Baker and Richards)

Increased DRP concentrations and loads since mid 1990s Fertilizer usage has remained constant or slightly decreasing TP loads and concentrations have remained steady during the same time period

Annu

al L

oad

(kg/

ha)

0

1

2

3 channelizedunchannelized

Soluble Phosphorus Total Phosphorus

Con

cent

ratio

n (m

g/L)

0.0

0.1

0.2

0.3

0.4

0.5 channelizedunchannelized

2005-2010 phosphorus concentration and annual loading for channelized and unchannelized streams in UBWC watershed, Ohio.

Total = 14,467 Miles of Streams Steve Davis, USDA-NRCS

Excerpts from Farm Drainage, by Henry French published in 1860:

•"The agriculture of Ohio can make no farther marked progress until

a good system of under-drainage  has  been  adopted.“ - John H. Klippart,

Esq., the learned Secretary of the Ohio Board of Agriculture

•"One of two things must be done by us here. Clay predominates in

our soil, and we must under-drain  our  land,  or  sell  and  move  west.“

- A writer in the Country Gentleman, from Ashtabula County, Ohio

Necessity of Tile Drainage 25% of cropland in US and Canada could

not be farmed without tile drainage (Skaggs et

al., 1994)

- soils with the greatest inherent production

potential

Tile Drainage (Fausey et al., 1987):

- provides trafficable conditions for field

operations

- promotes root development by preventing

exposure of plants to excess water

Why Drainage is Required - glacially derived, fine

textured soils - low gradient (flat)

J F M A M J J A S O N D

Volu

met

ric D

epth

(mm

)

0

20

40

60

80

100

120

140

160

180Precip > PETPET2005-2010 Precip

Effect of Drainage on Discharge

Solid = undrained Dashed = drainage

(Robinson and Beven, 1983)

Sandusky, Ohio (Schwab et al., 1963) Discharge from replicated 0.23 ha plots, Toledo silty clay

(March to September) Surface drained only (81 mm) Surface and subsurface (88 mm)

Tota

l Dis

char

ge (m

m/h

r)

Physical/Drainage Design Factors soils (12) depth (4) spacing (2) surface inlets (2) Management Impacts tillage (12) cropping (23) fertilizer source (24) STP and application rates (17) placement (5) Hydrologic and Climatic Effects baseflow vs. event flow) (7) seasonality (8)

Factors Affecting Phosphorus Transport to Tile Drainage

0

10

20

30

40

50

Canada Europe New Zealand United States

012345678

Current Research

Sampling Methodology • Flumes and weirs with automated sampling

equipment

• Hydrology recorded on a 10 minute interval

• Samples taken every 6 hours and composited on a weekly basis

Watershed0.0 0.2 0.4 0.6 0.8 1.0

Sum

med

Tile

0.0

0.2

0.4

0.6

0.8

1.0discharge (y = 0.47 x)dissolved phosphorus (y = 0.42 x)nitrate-nitrogen (y = 0.60 x)1:1 line

2005-2010 UBWC watershed

Edge-of-Field (EOF) Assessment Objective: 1) Elucidate and quantify the surface and subsurface hydrology and water

quality impacts of innovative conservation management practices 2) develop a suite of practices to address and mitigate offsite phosphorus

delivery 3) Use edge-of-field data to enhance Ohio P-index and other quantitative models

• Before/After Control Impact Design

• 34 fields (17 pair) representative of Ohio crop production agriculture (8 pair in WLEB, 4 pair in Upper Wabash, 4 pair in Upper Scioto)

¾ Surface and subsurface combination when

possible

Approach:

Instrumentation (EOF) • Designed for 10 year recurrence interval

• H-flumes for surface runoff

• Thel-mar compound weirs and Isco area

velocity sensors for tile

Data Collection and Analysis • Hydrology recorded on a 10 minute interval

• EOF study ¾ surface - composited on event basis

¾ tile - collected every six hours and

composited on daily basis

• Flow injection analysis using Lachat Instruments QuikChem 8000 FIA Automated Ion Analyzer)

¾ NO3-N, TN, DRP, TP

DR

P co

ncen

tratio

n (m

g/L)

0.0

0.5

1.0

1.5

2.0

Surface Tile

Chemograph of soluble P concentration in tile flow and graph of Mehlich 3 STP from the same field Positive correlation between peaks in concentrations and tile discharge indicate fast flow processes (preferential flow) and connection to surface sources

Time

6/10/11 6/14/11 6/18/11 6/22/11 6/26/11

Flow

Rat

e (l/

s)

0

5

10

15

20

25

solu

ble

P co

ncen

trat

ion

in th

e til

e (m

g/L)flow rate

soluble P

Soil Test P concentration (mg/kg)

0 100 200 300 400

Soil

Dep

th (c

m)

0

10

20

30

40

Positive correlation between peaks in P concentrations and tile discharge indicate fast flow processes (preferential flow) and connection to surface sources

Mar Apr May

Flow

dep

th (f

t)

0.00.20.40.60.81.01.2

DR

P c

onc.

(mg/

L)

0.00.20.40.60.81.0

flow depth concentration

Soil Test Phosphorus 0-2" (mg/kg)0 100 200 300 400 500 600

DRP

conc

entra

tion

(mg/

L)

0.0

0.5

1.0

1.5

2.0

DRP concentration rangesite median

Relationship between soil test phosphorus and dissolved phosphorus concentration in tile discharge (UBWC and Upper Wabash watersheds)

Quantifying hydrology and nutrient transport in matrix and preferential flow paths

Sample collection

Suction cup

Tile

Zero-tension lysimeter

Separate storm hydrographs into components using end-member mixing models to determine preferential flow

Potential Practices to Investigate • Cover crops • Banding vs broadcast • Spring vs fall vs split application • Incorporation (shallow vs deep injection • Tillage vs no-till • Tri state recommendation vs reduced rate • Manure vs commercial fertilizers • Controlled traffic and variable rate application • Surface amendments (gypsum) • Other (innovative ideas)

Upland Management (4 Rs)

Mitigation Strategies

Strategies for Addressing Agricultural Induced Phosphorus Transport

Upland Management 4Rs Interruption of connection to surface Structural Hydrologic Control Drainage water management blind inlets Filtration End-of-tile and in-stream Enhanced bioreactors Edge-of-field Buffers wetlands Ditch Design and Management Two stage, natural, and over-wide ditches Dredging Vegetated channels

What Determines Watershed Condition and Response? How Do We Measure and Monitor?

How Do Watersheds Function to Transport and Process Pollutants?

Uniqueness - Landscape and geomorphology

(drainage density, shape factors) - Management - Soils and geological deposits - Climate - Hydrologic alteration (drainage, impoundments) Complexity - Lag time - Seasonality - Land use change - Riparian function and processes - Interacting cycles of water, carbon, and nutrients

Upland/In-field Edge-of-field Downstream %

Red

uctio

n in

Pol

luta

nt T

rans

port

4-R approach

Scale What is the most effective scale to address water quality?

How do we avoid tradeoffs among pollutants? How does it depend on the ecoregion? How do we convince landowners to look at their individual fields

in a larger environmental context?

• More frequent, lower rates of fertilizer result in less loss

• Longer rotations lose less P

• No-till may result in > SP loss, but must balance that with < TP loss

• More P lost with corn

• Tillage increases buffering capacity and disrupts macropores

Management (Cropping, Tillage, & Time)

Provided by Doug Smith USDA-ARS, West Lafayette, IN

SP Load by Management

No-Till Rotation Till Conv Till/8yr Rot

SP L

oad

(g h

a-1)

0

100

200

300

400

500

TP Load by Management

No-Till Rotation Till Conv Till/8yr Rot

TP L

oad

(g h

a-1)

0

200

400

600

800

1000

1200

1400

Management (Fertilizer Source) Randall et al. (2000):

• 4 year plot study, 1993-1997, Waseca,

• Webster clay loam soil, with continuous corn

• Dairy manure and urea at two N rates in fall

• TP and DRP in tile drainage were not different

between treatments

• Max DRP conc.: urea 0.02 mg/L; manure 0.06 mg/L

• Max TP conc.: urea 0.08 mg/L; manure 0.12 mg/L

Kinley et al. (2007):

• 39 agricultural fields in Nova Scotia, CA,

2002 - 2003

• Use of poultry and swine manure

generally had greater proportions of

DRP when compared with dairy manure

and commercial fertilizers.

Kinley et al. (2007)

Management (Fertilizer Rates) Ball-Coelho et al., 2007:

• Huron silt loam soil Ontario, Canada.

• Liquid swine manure injected and topdressed

• high injection application rates (>56 m3/ha) contaminant transport to tiles was immediate with 3 year DRP concentrations 2.5 mg DRP/L

Algoazany et al (2007):

• LVW in Illinois – 4 subsurface sites, 1994-2000

• average annual DRP conc. 0.102, 0.099, 0.194, and 0.086 mg/L

• greater application rate and applying p after crop harvest had greater soluble p transport in tile

• Site where P was applied every 2 years had greater P concentration in tile drainage

Kinley et al. (2007)

Management (Fertilizer Placement) Ball-Coelho et al., 2007:

• Huron silt loam soil Ontario, Canada.

• Liquid swine manure injected and sidedressed

• Surface banding apps did not reach tile immediately but were transported in rains that fell 3d after application (0.1 mg DRP/L)

• Greater concentrations during and following sidedressing

Turtola and Jaakkola (1995):

• southwest Finland

• 16 plots (33m x 33m), 3 years (1980-1982)

• Broadcast vs. banding

• Conc. range 0.01 mg DRP/L to 1.45 mg DRP/L

• DRP conc. peaks occurred after broadcast appl.

Promote soil biological diversity

• Soil organisms control transformation between inorganic and organic P forms (Frossard et al., 2000; Illmer et al., 1995)

• Addition of microbial energy sources increased mobility of P by 38

times (Hannapel et al., 1963)

• Mobilization of P by microbial population was most important

factor in P transport (Hannapel et al., 1963)

Structural Hydrologic Control

Drainage Water Management Blind Inlets

Drainage Water Management (DWM)

• Conventional Subsurface Drainage

• Controlled Drainage

• Subirrigation

- reduced total phosphorus losses in NC by 35% (Evans et al., 1990)

- DRP losses reduced by 63% and TP losses by 50% in MN (Feser et al., 2010)

- 85% reduction in TP losses from small plots in Sweden (Wesstrom et al., 2001)

- 18% reduction in median DRP concentrations in Ohio from 8 paired fields (unpublished data from Norm Fausey)

From Ghane et al 2012

DWM Effects on Crop Yield

Blind Inlets Must be a practice farmers will implement

• Reduce sediment & phosphorus loads

• Minimize loss of productive land

• Allow farm traffic (avoid farming around risers)

• Minimal/easy maintenance

• Approved for cost share

• Effectively drain landscape

April 2010 Hydrology

4/25/10 12:00:00

4/25/10 18:00:00

4/26/10 0:00:00

4/26/10 6:00:00

4/26/10 12:00:00

4/26/10 18:00:00

4/27/10 0:00:00

Disc

harg

e (L

s-1)

0

1

2

3

4

5

Precipitation Intensity (mm

hr -1)

0

5

10

15

20

25

Blind InletTile RiserRF Intensity

33.5 mm precip total

Soluble P Concentrations During April 2010 Storm

4/25/10 1

2:00:0

0

4/25/10 1

8:00:0

0

4/26/10 0

:00:00

4/26/10 6

:00:00

4/26/10 1

2:00:0

0

4/26/10 1

8:00:0

0

4/27/10 0

:00:00

Solu

ble

P Co

ncen

tratio

n (m

g L-1

)

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

Blind InletTile Riser

Tile Riser = 3.12 g ha-1

Blind Inlet = 0.88 g ha-1

2009 2010 Soluble P 64 72 Total P 52 78

Percent Reduction (blind inlet vs. tile riser)

Provided by Doug Smith USDA-ARS, West Lafayette, IN

Filtration End-of-tile and in-stream Enhanced bioreactors

End-of-Tile Filters Laboratory and Field Results - 50% reduction in DRP concentrations and loads across 3 flow rates Constraint(s) - Flow Rates

Flow Rate (L/s)0.0 0.5 1.0 1.5 2.0

% R

educ

tion

0

20

40

60

80

100

Dissolved ReactivePhosporus

Ditch filters

Provided by Peter Kleinman USDA-ARS, State College, PA

FGD Gypsum Ditch Filter P removal efficiency

0

20

40

60

80

100

0 2000 4000 6000 8000 10000

P re

mov

al E

ffici

ency

%

Flow Rate Through Gypsum Filter L Day-1 m2

Maximum flow rate

Denitrifying Bioreactors

Provided by Ehsan Ghani Ohio State Univerisity

Potential to Enhance with P sorbent (steel slag)

- approximately 70% reduction in DRP over 2 years in New Zealand (McDowell et al., 2008) - > 70% of DRP in milkhouse wastes removed with steel slag (Bird and Drizo, 2010) - DRP concentrations reduced by 50 to 99% using in-stream gypsum (Penn et al., 2010) - 50% reduction in DRP concentrations and loads using end-of-tile filters (King et al., 2010) - bioreactors enriched with steel slag (Brown et al., in progress) - flow rate is limiting factor both in-stream and end-of-tile systems

In-stream or end-of-tile treatment summary

Edge-of-field Buffers wetlands

Forested Buffer

Ditch - No Buffer

Ditch - Buffer

Mea

n an

nual

con

cent

ratio

n (m

g/L)

0.0

0.1

0.2

0.3

Buffer Type

Dissolved Reactive Phosphorus Total Phosphorus

a b b a ab b

NB GB FB NB GB FB

Mean annual (2006-2010) stream side nutrient concentrations for dissolved reactive phosphorus and total phosphorus for streams with no buffers (NB), grassed buffers (GB), and forested buffers (FB). (unpublished data from UBWC provided by Kevin King).

Buffers

Wetlands Braskerud et al. (2005):

•17 constructed wetlands

•Sweden, Norway, Finland,

Switzerland, USA (Illinois)

•Surface area

(400 to 10000 m2)

• Limited in addressing large

flows

Braskerud et al. (2005)

Ditch Design and Management Two stage, natural, and over-wide ditches Dredging Vegetated channels

Agricultural Ditch Design Approaches

Trapezoidal Design

Two-Stage Design

Self-Forming Design

Provided by Jon Witter Ohio State University

• DRP not as variable; reduced in 3 streams

• Using paired data: two-stage reduced SRP concentrations (paired t-test, p=0.04)

Tank, Davis et al. unpublished data University of Notre Dame

Two-stage reduced SRP concentrations, but site dependent

Two-Stage vs. Trapezoidal Design

B Ditch Dipping

April 2004

January 2005

Not Dipped

Provided by Doug Smith USDA-ARS, West Lafayette, IN Year

2003 2004 2005 2006 2007 2008

Stud

y R

each

Sol

uble

P L

oad

(kg)

-20

0

20

40

60

80

100

120

ABC

In-situ P loads for monitored study reaches

Dipping

VEGETATED DRAINAGE DITCHES

Water / nutrient / sediment mixture amended flow: 600 gallons/minute for 7 hr

Load Reduction (%) Vegetated DIP 99 TP 86

Provided by Robbie Kroger (MSU)

Can We Avoid Unintended Consequences? Agricultural Systems are Leaky!

Farmers provide society with food, feed, fiber, and fuel with economical efficiency! How do we balance economic efficiency and ecological impact? Requires a shift in scale: More efficient agronomics need to be better combined with practices that provide healthier soils and approaches that effectively manage LANDSCAPES and their natural variability. Integrate knowledge about landscape variability, hydrology, and ecosystem processes into production agriculture. Accept that agricultural systems do leak – incorporate upland, EOF, and downstream approaches to minimize the impact of agriculture.

Why did it happen? What has changed? • Change in weather patterns (amounts, intensity, timing) ? • Increase in tile density (more surface connection/ macropores) ? • Change in tillage approach (macropores) ? • Herbicide monoculture (glyphosate) ? • Less small grains in rotation ? • GMOs ? • Change in soil biology ?

Kevin W. King Research Agricultural Engineer

USDA-ARS, Soil Drainage Research Unit Columbus, OH 43210

kevin.king@ars.usda.gov

(614) 292-3550 (office)

Contact Information