Late-Season Prediction of Wheat Grain Yield and Protein

1
Late-Season Prediction of Wheat Grain Yield and
Protein
K.W. Freeman, W.E. Thomason, E.V.Lukina, G.V.
Johnson, K.J. Wynn, J.B. Solie, M.L. Stone, and
W.R. Raun
Oklahoma State University Department of Plant and
Soil Sciences
2
Introduction

• Pre-harvest prediction of wheat yield will assist
  producers
• Provide more reliable field maps
• Assist in pre-harvest marketing
• Pre-harvest prediction of grain protein
• Strong correlation between plant N and NDVI
  (Stone, 1996)
• Determine whether or not to apply late-season N

3
Introduction

• In order to describe the variability encountered
  in the field, soil, plant, and indirect
  measurements should be made at the meter or
  submeter level (Solie et al., 1999).
• Field element size area that provides the most
  precise measure of the available nutrient where
  the level of that nutrient changes with distance
  (Solie et al. 1996).
• Willis (1999) defined yield maps as tools used by
  producers to look for general patterns and
  trends, and that yield monitor data could be
  corrected using remotely sensed data.

4
Objectives

• To determine the relationship between spectral
  measurements taken from Feekes growth stages 9 to
  physiological maturity and grain yield and grain
  protein.

5
Ripening Stage
Heading
Stem Extension
Growth Stages in Cereals
Tillering
6
Materials and Methods

• Seven experimental sites
• Stillwater, Lahoma, Hennessey, Perkins and
  Haskell, OK
• Experimental design
• 4 experiments in long-term fertility trials, 2
  anhydrous ammonia NUE trials and a Sewage Sludge
  loading experiment
• 2 x 2 m subplots placed in existing experiments
  with differing N rates
• Spectral reflectance readings taken with
  photodiode-based sensor with interference filters
  for red at 6716 and near infrared (NIR) at 7806
  nm wavelengths

7
Materials and Methods

• Experimental design (cont)
• Readings were taken at Feekes growth stages 9,
  10.5,
• 11.2, and 11.4
• Spectral indices were calculated for each subplot
  
• at all growth stages
• Grain Production
• Harvest of 2 x 2 m area with a self-propelled
• combine
• Grain samples were ground to pass 120-mesh screen
  and analyzed for total N using Carlo-Erba 1500
  dry combustion analyzer (Schepers et al., 1989)

8
Materials and Methods

• Formulas for Spectral Indices
• NDVI NIR ref red ref / NIR ref red ref
• INSEY NDVI (each date) / days from planting
• RI DM yield of highest yielding plots / DM
  yield of check
• ISRI Highest NDVI / NDVI from check

9
Feekes 9
6000
5000
2
y 2188.5x
778.72x 711.53
2
R
0.46
4000
Yield kg ha-1
3000
2000
1000
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
NDVI
10
Feekes 9
R.I.lt1.5
R.I.gt1.5
6000
5000
2
2
y 1628.6x
1731.1x 443.84
y -2530.4x
8204.1x - 2054.1
2
2
R
0.69
R
0.12
4000
Yield kg ha-1
3000
2000
1000
0
0
0.2
0.4
0.6
0.8
1
0
0.2
0.4
0.6
0.8
1
NDVI
NDVI
11
Feekes 10.5
6000
5000
2
- 2310.5x 1504.5
y 5379.6x
2
R
0.5943
4000
Yield kg ha-1
3000
2000
1000
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
NDVI
12
Feekes 10.5
R.I.lt1.5
R.I.gt1.5
6000
6000
5000
5000
2
y 2344x
1844.5x 270.99
2
R
0.23
2
y 2188x
4000
4000
1336.7x 708.19
2
R
0.72
Yield kg ha-1
3000
3000
2000
2000
1000
1000
0
0
0
0.2
0.4
0.6
0.8
1
0
0.2
0.4
0.6
0.8
1
NDVI
NDVI
13
NDVI and INSEY vs. Yield, Feekes 10.5
6000
6000
5000
5000
1.815x
y 700.86e
365.08x
y 696.59e
2
R
0.6039
2
R
0.6009
4000
4000
Yield kg ha-1
3000
3000
2000
2000
1000
1000
0
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0
0.0005
0.001
0.0015
0.002
0.0025
0.003
0.0035
0.004
0.0045
0.005
NDVI
INSEY
14
Response Indices
Feekes 9
2
4
y -0.0403x
1.1861x - 0.0877

• Strong correlation between ISRI and RI determined
  at harvest
• Accurately predict the crops ability to respond
  to N
• RI can refine whether or not N should be applied,
  how much, and expected NUE

3.5
R
2
0.98
3
2.5
RI
2
1.5
1
0.5
0
0
0.5
1
1.5
2
2.5
3
3.5
4
ISRI
Feekes 10.5
2
4
y -0.1467x
1.5816x - 0.4018
3.5
2
R
0.93
3
2.5
RI
2
1.5
1
0.5
0
0
1
2
3
4
5
ISRI
15
Conclusion

• Grain yield was highly correlated with NDVI and
  INSEY
• Grain yield could be accurately predicted using
  NDVI readings at Feekes growth stages 9 and 10.5
• ISRI can accurately predict response to N
• NDVI readings (Feekes 9 and 10.5) at locations
  with high RI (gt1.5) showed higher correlation
  with grain yield than those with low RI (lt1.5)