DEPARTMENT OF ENVIRONMENTAL SCIENCE & TECHNOLOGY

Laboratory for Agriculture and Environmental Studies

Sensor Based Variable Rate N Management in Mid-

Atlantic Cropping Systems

Joshua M. McGrath, UMD

Wade Thomason, VT

Steve Phillips, IPNI

Background on the Mid-Atlantic

Nitrogen Management Basics(since early 1990s in MD)

Corn:• Apply N based on

realistic yield goal – 1 lb N/bu expected yield

• Split-apply N fertilizer on corn– apply 30 – 40 lbs/A at

planting– remaining N as sidedress

• When utilizing manure or following leguminous cover crops use the

Pre-Sidedress Nitrate Test

Wheat:• Apply a total 70 – 120 lbs

of N• 20 – 40 lbs at planting• Split top-dress 80 – 100

– Green-up– Feekes 5 - 6

What makes us special?

• >90% side-dress N in Corn on Delmarva

• Yields vary widely 120 – 220 bu/ac

• >90% no-till corn• 50 – 60% irrigated on the

lower shore• In Maryland farmers cannot exceed University recommendations

VIRGINIA APPROACHVIRGINIA APPROACH

Generated calibration models for wheat and corn grown in the Mid-Atlantic

Developed N fertilization algorithms for both crops

Validated these rate equations for performance

VIRGINIA APPROACHVIRGINIA APPROACH

Generated calibration models for wheat and corn grown in the Mid-Atlantic

0

1

2

3

4

5

6

7

0 0.001 0.002 0.003 0.004 0.005 0.006 0.007

Predicted Yield Index

Yie

ld,

Mg

ha

-1

0

1

2

3

4

5

6

7

0 0.001 0.002 0.003 0.004 0.005 0.006 0.007

VIRGINIA APPROACHVIRGINIA APPROACH

Generated calibration models for wheat and corn grown in the Mid-Atlantic

y = 6.7825e367.41x

R2 = 0.9752

y = 8.5528e244.11x

R2 = 0.95

0

20

40

60

80

100

120

140

160

180

0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009

Yie

ld,

bu

/ac

YPI

Wheat N Rate Algorithm

Rate (kg/ha) vs. NDVI

0.0000

20.0000

40.0000

60.0000

80.0000

100.0000

120.0000

140.0000

0.00 0.20 0.40 0.60 0.80 1.00

N Rate, kg/ha

Constrained NRate, kg/ha

VIRGINIA APPROACH - WheatVIRGINIA APPROACH - WheatTARGET NDVI

LOW REFERENCE HIGH REFERENCE GS 25 N DAP MAX YIELD MAX N NUE N RATE

0.60 0.50 0.80 50 150 85 80 50 80lbs/N dap bushels max constraint% lb/ac

87.1361 N fertilizer required23.5681 N required for additional grain46.0133 Bushel/A expected yield increase due to fertilizer

coefficients Define these units SI units ?ka 0.74076kb 577.66kc 1000kd

NUE 50%CONVERSION 0.0149

0.220.230.240.250.260.270.28

Rate (LB/ac) vs. NDVI

0.0000

20.0000

40.0000

60.0000

80.0000

100.0000

120.0000

140.0000

0.00 0.20 0.40 0.60 0.80 1.00

N Rate

Constrained

VIRGINIA APPROACH - CornVIRGINIA APPROACH - CornNDVI Low_Ref Hi_Ref N_preplant DFP MAX YIELD MAX N N RATE

0.82 0.50 0.90 60 63 175 100 85lbs/N dap bushels max constraintgal/acre

kg/ha85.27

coefficients Define these units SI units ?ka 56.479kb 67.626kc 56kd

NUE 60%Pn 0.0125

0.220.230.240.250.260.270.28

Rate (gal/ac) vs. NDVI

0.0

20.0

40.0

60.0

80.0

100.0

120.0

140.0

0.00 0.20 0.40 0.60 0.80 1.00

N Rate

Constrained

What’s Needed

• Reference Strips– Difference between:

• High N reference (best possible with more than adequate N)

• Low N reference (0 N applied)

And• Area currently in sensor’s view.

In-season N management

1st topdress based on estimated tiller #

2nd topdress based on Tissue N concentration

30 to 40 lb N/ac preplant

Zadoks stage, in parenthesis

In-season N timing

Days From Planting

1st topdress based on estimated tiller #

2nd topdress based on NDVI and DFP

Validation in VA, MD, DE

• ~60 sites across Virginia since 2000

• 12 sites in DE and MD (2008, 2009)

• Irrigated and non-irrigated

• Conventional and no till• Various rotations,

hybrids, varieties, and soil types

• Wide range of pre-plant, starter, and in-season N rates (including VR)

• Collected spectral measurements and an assortment of plant physical and chemical characteristics at various growth stages

• Determine grain yield

Validation in VA, MD, and DE

• RCBD; 3-8 reps• Plot size: 60 by 250

to 500 ft• In-season N:

– 4-5 Fixed Rates– 1 Standard Rate– 1 Variable Rate

Spra Coupe 220 w/ 60 ft boom and

GreenSeeker RT200

Raven 440 flow rate controller

Recommended vs. Applied Rates

0

5

10

15

20

25

30

1 201 401 601 801 1001 1201 1401 1601

gal/a

cre

Prescribed

As Applied

• 55% are within 2 gallons of prescribed rate• 88% within 5 gallons

Wheat: Applied N GreenSeeker vs. Tissue

0102030405060708090

100

ST

DG

SS

TD

GS

ST

DG

SS

TD

GS

ST

DG

SS

TD

GS

ST

DG

SS

TD

GS

ST

DG

SS

TD

GS

ST

DG

SS

TD

GS

ST

DG

SS

TD

GS

ST

DG

SS

TD

GS

ST

DG

S

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 AVG

N r

ate,

kg

ha-1

Average Wheat Performance

Farmer Practice

Farmer Practice

GS

GS

0102030405060708090

100

Grain Yield, bu/ac GS 30 N Rate Applied, lb/ac

NS

7% less N with GS

Corn Algorithm Performance

0

20

40

60

80

100

120

1 2 3 4 5 6 7 8 9 10 11

N R

ate,

lb

/ac

STD

GS

0

20

40

60

80

100

120

140

160

STD

GS

Average Corn Performance

Grain Yield, bu/ac Sidedress N, lb/ac

NS

21 % (21 lb/ac) less N with GS

Limitations and Challenges

• Corn: V-7 versus V-4• Wheat: Herbicide in

second spring top-dress?• Check strips• How to tell if the system

was “right” – Skepticism• When parameters are

entered the application graph often needs tweaking by faking parameters

Honest Answers Don’t Always Work!

Can’t just plug and chug (spray)

Farmer would apply 46 gal/acre

N_pre = 0This has been corrected in the software

Part of the Adaptive Management Toolbox

• Systems approach– Manage soil N

• Focused on in-season N management– PSNT– Active Crop sensors– LCM

• End of season assessment– CSNT– Yield monitoring

The Future in the Mid-Atlantic

• The Chesapeake Bay and cost of fertilizer have us poised for rapid adoption– A few major farmers and ag service companies– Environmental Groups

• Likely to be covered through various cost share mechanisms soon

• Need to continue to refine algorithm to address– Accuracy of input values– Expand side-dress window– Straightforward user interface

Thanks to our partners