Jan 2008 www.GeoAgris.com

Goals www.GeoAgris.com

•Test AgroFresh product under real field conditions

•Use PA (Precision Agriculture) to

•Increase operation capacity by getting data automatically

•Filter noise by taking into account field spatial heterogeneity

•Get yield response information from different stress conditions (environments) within the same paddock

Background www.GeoAgris.com

Typical Strip design

Applied AppliedNot

AppliedNot

Applied

Background www.GeoAgris.com

54.28

59.32

57.59

60.25

Typical Strip Design

Paddock

Background www.GeoAgris.com

Without PA technology: -strip result may be influenced by spatial

heterogeneity (noise)

-different responses under different stress conditions may be hidden because strip yield result is the average of different stress conditions along strip

PA can:

-Filter spatial heterogeneity noise

-Obtain yield response information for different stress conditions

Methods www.GeoAgris.com

Paddock selection

Spatial Heterogeneity Analysis with previous years data

Strip design crossing (perpendicular to) heterogeneity limits

Product application and application control

Combine calibration

Harvest and combine control

Data Analysis

Methods www.GeoAgris.com

Paddock selection

Results www.GeoAgris.com

La Gambeta 2007-2008 Trigo

Results www.GeoAgris.com

Calibration Trial Harvest

Max Error: 1 %

combine yield monitorvs.

standard scale

Results www.GeoAgris.com

La Gambeta 2007-2008 Trigo (QQ/ha)

REPET APP TEST DiferenciaR1 66,14 62,86 3,28R2 62,33 63,66 -1,33R3 63,38 62,26 1,12R4 64,46 63,98 0,48R5 65,99 65,39 0,59Total general 64,46 63,63 0,83

P-Value 0,1625

Results www.GeoAgris.com

Analysis by environments of different yield potential levels

Low Yield Area 52 QQ/haMid Yield 62 QQ/ha

High Yield Area 70 QQ/ha

Results www.GeoAgris.com

Analysis by environments of different yield potential levels

Agregar cuadro con el rendimiento promedio de cada ambiente

Ambiente Repetición APP TEST Diferencia Diferencias P-ValueR1 55.11 54.87 0.24R2 50.93 54.13 -3.20R3 50.21 52.05 -1.83R4 51.41 49.24 2.17R1 63.93 60.08 3.85R2 63.26 61.48 1.79R3 63.84 62.79 1.05R4 61.19 64.59 -3.40R5 64.13 62.06 2.07R1 72.30 69.01 3.30R2 71.95 70.34 1.60R3 71.38 70.78 0.60R4 70.89 68.95 1.94R5 68.70 70.46 -1.75

1.14

0.3082

0.2126

0.1240

-0.66

1.07

Bajo potencial de

rend.

Potencial de rend.

intermedio

Alto potencial de

rend.

Trial Sensibility: 2.3 QQ/ha (3.5% yield)n = 5significance level α = 0.05 (β=0.5)

Results www.GeoAgris.com

Four field results (QQ/ha)

- - -

Differences P-Value Differences P-Value Differences P-Value Differences P-Value

Whole Paddock -0.3 0.08 0.8 0.16 0.5 0.40 -0.3 0.41

Low Yield Area 0.1 0.35 -0.7 0.31 0.6 0.34 0.2 0.42

Mid Yield -0.5 0.12 1.1 0.21 -0.1 0.48 -0.8 0.36

High Yield Area 0.2 0.33 1.1 0.12 0.4 0.42 -0.6 0.37

La Carmen La Gambeta La Invernada Maria Eugenia

Results www.GeoAgris.com

Sensitivity for α = 0.05 and β=0.5

Sensitivity

0123456789

10111213141516

2 3 4 5 6 7 8 9 10 11

N (repetitions)

Pro

du

ct E

ffec

t Q

Q/h

a.

Average case

Worst case

if 4 repetitions and product effect is 4 QQ/ha then we will obtain 0.05 pvalue in 50% of the trials (the worst case)