U.S. patent application number 11/423631 was filed with the patent office on 2006-12-14 for method and system for use of environmental classification in precision farming.
This patent application is currently assigned to PIONEER HI-BRED INTERNATIONAL, INC.. Invention is credited to DONALD P. AVEY, PHILLIP L. BAX, RICHARD GLENN BROOKE, DAVID S. ERTL, JOSEPH K. GOGERTY, DAVID J. HARWOOD, MICHAEL J. LAUER, TERRY EuCLAIRE MEYER, TODD A. PETERSON.
Application Number | 20060282228 11/423631 |
Document ID | / |
Family ID | 37532891 |
Filed Date | 2006-12-14 |
United States Patent
Application |
20060282228 |
Kind Code |
A1 |
AVEY; DONALD P. ; et
al. |
December 14, 2006 |
METHOD AND SYSTEM FOR USE OF ENVIRONMENTAL CLASSIFICATION IN
PRECISION FARMING
Abstract
Methods and software for selecting seed products or other
agricultural inputs for planting within an associated land base
include dividing the field into regions, providing an environmental
profile for each of regions, determining a recommendation of a seed
product to plant within each of the regions based on the
environmental profile and performance of the genotype of each of
the seed products in the environmental profile of the regions, and
providing an output identifying the seed product to plant within
each of the regions.
Inventors: |
AVEY; DONALD P.; (Ankeny,
IA) ; BAX; PHILLIP L.; (Johnston, IA) ;
BROOKE; RICHARD GLENN; (Johnston, IA) ; ERTL; DAVID
S.; (Waukee, IA) ; GOGERTY; JOSEPH K.;
(Algona, IA) ; HARWOOD; DAVID J.; (Johnston,
IA) ; LAUER; MICHAEL J.; (Des Moines, IA) ;
MEYER; TERRY EuCLAIRE; (Urbandale, IA) ; PETERSON;
TODD A.; (Johnston, IA) |
Correspondence
Address: |
MCKEE, VOORHEES & SEASE, P.L.C.;ATTN: PIONEER HI-BRED
801 GRAND AVENUE, SUITE 3200
DES MOINES
IA
50309-2721
US
|
Assignee: |
PIONEER HI-BRED INTERNATIONAL,
INC.
7250 NW 62nd Avenue P. O. Box 1014
Johnston
IA
|
Family ID: |
37532891 |
Appl. No.: |
11/423631 |
Filed: |
June 12, 2006 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60689716 |
Jun 10, 2005 |
|
|
|
Current U.S.
Class: |
702/81 |
Current CPC
Class: |
G06Q 30/00 20130101;
Y02A 40/22 20180101; G06Q 30/0256 20130101; G06Q 30/0282 20130101;
G06Q 50/02 20130101; A01B 79/005 20130101; Y02A 40/12 20180101;
Y02A 40/222 20180101; G06Q 40/06 20130101; G06Q 20/10 20130101;
A01C 1/00 20130101; G06Q 40/00 20130101; A01H 1/04 20130101; G06Q
30/0631 20130101; Y02A 40/10 20180101; G06Q 40/025 20130101; G06Q
40/08 20130101; A01C 21/00 20130101 |
Class at
Publication: |
702/081 |
International
Class: |
G01N 37/00 20060101
G01N037/00 |
Claims
1. A computer-assisted method of selecting seed products for
planting by a crop producer associated with a field, each of the
seed products having a genotype, the method comprising: dividing
the field into a plurality of regions; providing an environmental
profile for each of the regions; determining a recommendation of a
seed product for each of the regions based on the environmental
profile and performance of the genotype of each of the seed
products in the environmental profile of the regions; providing an
output comprising identification of the seed product to plant
within each of the regions.
2. The computer-assisted method of claim 1 wherein the
environmental profile includes an environmental classification
associated with each of the regions.
3. The computer-assisted method of claim 1 wherein the step of
determining the recommendation is based at least partially on
genotype-by-environment interactions.
4. The computer-assisted method of claim 1 wherein the step of
determining the recommendation is based at least partially on
genotype-by-environment-by-management interactions.
5. The computer-assisted method of claim 1 wherein the
environmental profile includes agronomic information.
6. The computer-assisted method of claim 1 wherein the
environmental profile includes data selected from the set
consisting of wind data, temperature data, solar radiation data,
precipitation data, soil type data, soil pH data, planting and
harvesting dates, irrigation data, tiled area data, previous crop
data, fertilizer data, nitrogen level data, phosphorous level data,
potassium level data, insecticide data, herbicide data, and biotic
data.
7. The computer-assisted method of claim 1 wherein the output is
displayed on a display.
8. The computer-assisted method of claim 1 further comprising
supplying the seed products.
9. The computer-assisted method of claim 1 further comprising
planting the seed products.
10. The computer-assisted method of claim 1 further comprising
wherein the output further comprises at least one management
recommendation for each of the regions.
11. The computer-assisted method of claim 10 wherein the management
recommendation includes herbicide recommendations.
12. The computer-assisted method of claim 10 wherein the management
recommendation includes pesticide recommendations.
13. The computer-assisted method of claim 10 wherein the management
recommendation includes fertilizer type recommendations.
14. The computer-assisted method of claim 10 wherein the management
recommendation includes application rate recommendations.
15. The computer-assisted method of claim 10 wherein the management
recommendation includes application timing recommendations.
16. The computer-assisted method of claim 10 wherein the management
recommendation includes irrigation recommendations.
17. The computer-assisted method of claim 1 further comprising
wherein the output further comprises a prescription map based on
the environmental profile for each of the regions.
18. The computer-assisted method of claim 2 wherein the
environmental classification is selected from a set of
environmental classes, the set of environmental classes comprising
a temperate class, a temperate dry class, a temperate humid class,
a high latitude class, and a subtropical class.
19. The computer-assisted method of claim 2 wherein the
environmental classification is selected from a set of
environmental classes, the set of environmental classes comprising
biotic classifications.
20. The computer-assisted method of claim 3 wherein the
genotype-by-environment interactions are determined at least
partially based on performance data associated with the seed
products.
21. The computer-assisted method of claim 3 wherein the
genotype-by-environment interactions are determined at least
partially based on environmental classifications associated with
performance data of the seed products.
22. The computer-assisted method of claim 21, wherein said
performance data includes at least one item from the set consisting
of yield, drought resistance, grain moisture, lodging, stand
establishment, emergence, midsilk, test weight, protein, oil, and
starch percentage, relative maturity, plant height, seed size,
disease resistance genes, heading date, resistance to insects,
brittle snap, stalk breakage, resistance to fungus, seed moisture,
head shape, hullability, seedling vigor, beginning bloom date,
maturity date, seed shatter, winter survival, fiber strength, and
color grade.
23. The computer-assisted method of claim 1 wherein the performance
is expected performance.
24. The computer-assisted method of claim 1 wherein the performance
is actual performance.
25-26. (canceled)
27. A computer-assisted method of selecting seed products for
planting by a crop producer associated with a land base, each of
the seed products having a genotype, the method comprising:
dividing the land base into a plurality of regions; classifying
each of the regions with an environmental classification;
determining a recommendation of a seed product for each of the
regions based on the environmental classification; providing an
output comprising identification of the seed product to plant
within each of the regions.
28. The computer-assisted method of claim 27 wherein the output
further comprises predicted performance data for the seed
product.
29. The computer-assisted method of claim 28 further comprising
comparing the predicted performance data to actual data.
30. The computer-assisted method of claim 29 wherein the step of
determining a recommendation is based at least partially on
genotype-by-environment interactions.
31. The computer-assisted method of claim 27 wherein the step of
determining a recommendation is based at least partially on
genotype-by-environment-by-management interactions.
32. The computer-assisted method of claim 27 further comprising
supplying the seed products.
33. The computer-assisted method of claim 27 further comprising
wherein the output further comprises at least one management
recommendation for each of the regions.
34. The computer assisted method of claim 27 wherein the
recommendation provides for planting a plurality of seed products
within the field.
35. The computer assisted method of claim 27 wherein the
recommendation provides for planting a single seed product within
the field.
36. A method to assist in classifying agricultural inputs, the
method comprising: collecting crop production information from a
plurality of crop producers wherein the crop production information
comprises type of seed product, environmental profile information,
and performance information; storing the crop production
information in a data base; determining a set of environmental
classifications to describe the agricultural inputs at least
partially based upon the environmental profile information and
performance information.
37. The method of claim 36 wherein the crop production information
further comprises genotype-by-environment interactions.
38. The method of claim 36 wherein the agricultural inputs include
land base.
39. The method of claim 36 wherein the agricultural inputs include
regions of a land base.
40. The method of claim 36 wherein the agricultural inputs include
the type of seed product.
41. The method of claim 36 wherein the crop production information
further comprises location information.
42. The method of claim 36 further comprising the crop production
information excludes location information and producer
identity.
43. The method of claim 41 wherein the location information
includes historic information.
44. The method of claim 43 wherein the historic information
includes soil information data.
45. The method of claim 43 wherein the historic information
includes yield information data.
46. The method of claim 43 wherein the historic information
includes chemical application information.
47. The method of claim 41 wherein the location information
includes environment and production information.
48. The method of claim 47 wherein the environment and production
information includes data selected from the set consisting of
maturity days, input traits, output traits, seed treatment, no till
practice information, planting population, nitrogen utilization,
drought impact based on environmental classification, drought
frequency information, and soil type.
49. The method of claim 36 wherein the crop production information
further comprises genotype-by-environment-by-management
interactions
50. The method of claim 36 wherein the environmental profile
information includes agronomic information.
51. The method of claim 36 wherein the environmental profile
information includes data selected from the set consisting of wind
data, temperature data, solar radiation data, precipitation data,
soil type data, soil pH data, planting and harvesting dates,
irrigation data, tiled area data, previous crop data, fertilizer
data, nitrogen level data, phosphorous level data, potassium level
data, insecticide data, herbicide data, and biotic data.
52. The method of claim 36 wherein the environmental
classifications are selected from a set of environmental classes,
the set of environmental classes comprising a temperate class, a
temperate dry class, a temperate humid class, a high latitude
class, and a subtropical class.
53. The method of claim 36 wherein the environmental
classifications are selected from a set of environmental classes,
the set of environmental classes comprising biotic
classifications.
54. The method of claim 37 wherein the genotype-by-environment
interactions are determined at least partially based on performance
data associated with the seed products.
55. The method of claim 37 wherein the genotype-by-environment
interactions are determined at least partially based on
environmental classifications associated with performance data of
the seed products.
56. The method of claim 36, wherein the performance information
includes at least one item from the set consisting of yield,
drought resistance, grain moisture, lodging, stand establishment,
emergence, midsilk, test weight, protein, oil, and starch
percentage, relative maturity, plant height, seed size, disease
resistance genes, heading date, resistance to insects, brittle
snap, stalk breakage, resistance to fungus, seed moisture, head
shape, hullability, seedling vigor, beginning bloom date, maturity
date, seed shatter, winter survival, fiber strength, and color
grade.
57. The method of claim 36 wherein the performance information is
expected performance information.
58. The method of claim 36 wherein the performance information is
actual performance information.
59. The method of claim 36 wherein the crop production information
further comprises disease information.
60. The method of claim 36 wherein the crop production information
further comprises agricultural pest information.
61. The method of claim 36 wherein determining a set of
environmental classifications to describe the agricultural inputs
is at least partially based upon the environmental profile
information, performance information and location information.
62. A method for making a recommendation for a seed product to
plant within a land base, the method comprising: collecting crop
production information from a plurality of crop producers over a
plurality of geographically diverse locations; determining
environmental classifications using the crop production
information; classifying the land base with at least one
environmental classification; determining the recommendation for a
seed product to plant within the land base based on the at least
one environmental classification.
63. The method of claim 62 further comprising supplying the seed
product.
64. The method of claim 62, wherein the wherein the crop production
information comprises type of seed product, environmental profile
information, and performance information.
65. The method of claim 62 wherein the crop production information
further comprises genotype-by-environment interactions.
66. The method of claim 62 wherein the land base comprises a
plurality of regions.
67. The method of claim 66 wherein each region is classified with
an environmental classification.
68. The method of claim 62 further comprising storing the crop
production information in a data base.
69. The method of claim 62 further comprising providing an output
comprising identification of the seed product to plant within the
land base.
70. A method to assist in classifying a seed product, the method
comprising: collecting crop production information from a plurality
of crop producers wherein the crop production information comprises
type of seed product, environmental profile information, and
performance information; storing the crop production information in
a data base; determining a set of environmental classifications to
associate with the seed product at least partially based upon the
environmental profile information and performance information.
71. A method to assist in classifying agricultural inputs, the
method comprising: collecting crop production information from a
plurality of crop producers wherein the crop production information
comprises type of seed product, environmental profile information,
performance information and location information; storing the crop
production information in a data base; determining a set of
environmental classifications to associate with the agricultural
inputs at least partially based upon the environmental profile
information and performance information; using the set of
environmental classifications associated with the agricultural
inputs to recommend a subset of the agricultural inputs.
72. The computer assisted method of claim 1 wherein the
recommendation provides for planting a plurality of seed products
within the field.
73. The computer assisted method of claim 1 wherein the
recommendation provides for planting a single seed product within
the field.
74. The computer-assisted method of claim 1 further comprising
determining an overall environmental profile for the field based on
the environmental profile for each of the regions.
75. The computer-assisted method of claim 74 wherein the
environmental profile includes an environmental classification
associated with each of the regions.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims priority under 35 U.S.C. .sctn. 119
of a provisional application Ser. No. 60/689,716 filed Jun. 10,
2005, which application is hereby incorporated by reference in its
entirety.
BACKGROUND OF THE INVENTION
[0002] The present invention provides for computer-implemented
methods and related methods to assist a crop producer or
agricultural input supplier in selecting and/or classifying
agricultural inputs, such as seed products to use in one or more
fields, utilizing environmental classification information as well
as precision agriculture information and methodologies.
[0003] The problem is generally described in the context where the
seed products are corn hybrids. The current industry-wide approach
to delivering product performance information for use in hybrid
selection by producers is to use numerous comparative yield
measurements from recent years (primarily the most recent year) and
the geography considered relevant to the producer. Use of recent
product performance data in selection of hybrids is not completely
indicative of future hybrid performance as environmental and biotic
factors vary from year to year, including extreme weather events,
such as drought or flooding and pest or disease prevalence.
Moreover, this approach does not fully take into account
environmental and biotic factors important to a hybrid's
performance. Furthermore, this approach lacks the full assessment
of the relevance of the information generated by the trials to the
relative performance of cultivars (genotype by environment
interaction), for example genetic correlations. It assumes recent
experience is the best predictor of future relative hybrid
performance, regardless of how representative the recent experience
may or may not be of the long-term environmental profile of the
producer's land base.
[0004] In addition, this selection approach does not take into
account a producer's objectives for productivity, nor does it allow
for objective and specific recommendations of seed products or
other crop product considerations, for example, fertilizer types or
irrigation needs, for a particular land base so that producers may
minimize their risk of unexpected performance occurrences. Although
risk is uncertain, it is manageable.
[0005] What is needed is a method for product selection that is
useful in characterizing relative performance of different seed
products so that risk can be managed.
SUMMARY OF THE INVENTION
[0006] Therefore it is a primary object, feature, or advantage of
the present invention to improve over the state of the art.
[0007] Another object, feature, or advantage of the present
invention is to provide a method to assist customers, including
producers, in managing risks associated with crop production.
[0008] Yet another object, feature, or advantage of the present
invention is to assist customers and others in understanding
relative performance of different agricultural inputs, including
seed products, under the same or similar environmental
conditions.
[0009] A still further object, feature, or advantage of the present
invention is to assist customers and others in understanding
relative performance of an agricultural input, such as a seed
product, under a range of environmental conditions.
[0010] Another objective, feature, or advantage of the present
invention is to assist producers in selecting the best seed product
for a particular location or regions within a land base.
[0011] It is a further object, feature, or advantage of the present
invention to describe genotype-by-environment interactions that may
affect performance of a seed product.
[0012] It is a further object, feature, or advantage of the present
invention to improve product selection decisions for regions within
a land base.
[0013] A further object, feature, or advantage of the present
invention is to increase the likelihood of high product
performance.
[0014] A further object, feature, or advantage of the present
invention is to increase yield advantage of a product.
[0015] Another object, feature, or advantage of the present
invention is to assist customers in selection of these products
most adapted to regions within their land base.
[0016] A further object, feature, or advantage of the present
invention is to use genotype by environment information to capture
more data from a broader area to use for a localized area.
[0017] Yet another object, feature, or advantage of the present
invention is to increase a producer's confidence in planning
recommendations.
[0018] Yet another object, feature, or advantage of the present
invention is to use genotype by environment interactions to
categorize particular land bases into different environmental
classifications.
[0019] A still further object, feature, or advantage of the present
invention is to allow for the creation of an environmental profile
for all or part of a particular land base.
[0020] A still further object, feature, or advantage of the present
invention is to use precision agriculture data obtained from the
producer to assist in providing an environmental classification for
different agricultural inputs, including seed products.
[0021] A still further object, feature, or advantage of the present
invention is to use precision agriculture data obtained from the
producer to assist in providing an environmental classification for
different agricultural inputs, without the need for location
information or producer information.
[0022] It is to be understood that the present invention has a
number of different aspects, each of which may demonstrate one or
more of these and/or other objects, features, or advantages of the
present invention as will become apparent from the specification
that follows.
[0023] The present invention has numerous aspects that build upon
the application of environment classification and information
extracted from hybrids. These various aspects are often described
herein from the perspective of a seed company and a crop producer
and when a specific crop is used as an example, the exemplary crop
is usually corn. Of course, aspects of the present invention are
applicable to many different types of companies or individuals and
many different types of agricultural inputs and/or products. Also,
the present invention is of use not just to crop producers but
others who have interest in comparing relative performance of
agricultural inputs under different conditions. This could include,
for example, downstream users of agricultural products, such as
agricultural input suppliers such as equipment manufacturers,
chemical producers, landlords, or others who have interests related
to agricultural production.
[0024] The present invention relates to improved understanding of
genotype-by-environment interactions and applications of those
methods in a variety of contexts for a variety of purposes.
[0025] According to one aspect of the present invention a
computer-assisted method of selecting seed products for planting by
a crop producer associated with a field is provided. Each of the
seed products has a genotype. The method includes dividing the
field into regions, providing an environmental profile for each of
regions, determining a recommendation of a seed product to plant
within each of the regions based on the environmental profile and
performance of the genotype of each of the seed products in the
environmental profile of the regions, and providing an output
identifying the seed product to plant within each of the regions.
The environmental profile can include an environmental
classification associated with each of the regions.
[0026] Determination of a recommendation is preferably at least
partially based on genotype-by-environment and/or
genotype-by-environment-by-management interactions. The
recommendation may provide for planting a plurality of seed
products or a single seed product within the field. The method may
also include determining an overall environmental profile for the
field based on the environmental profile for each of the
regions.
[0027] According to another aspect of the present invention, a
computer-assisted method of selecting seed products for planting by
a crop producer associated with a land base is provided. The method
includes dividing the land base into a plurality of regions,
classifying each of the regions with an environmental
classification, determining a recommendation of a seed product to
plant within each of the regions based on the environmental
classification, and providing an output comprising identification
of the at least one seed product to plant within each of the
regions.
[0028] According to another aspect of the present invention, a
method to assist in classifying agricultural inputs is provided.
The method includes collecting crop production information from
multiple crop producers. The crop production information includes
the type of seed product, environmental profile information, and
performance information. The crop production information is stored
in a data base and a set of environmental classifications is
determined to describe the agricultural inputs at least partially
based on the environmental profile information and performance
information. The crop production information may also include
genotype-by-environmental information as well as location
information. Location information and identity of the producer may
also be excluded from the crop production information.
[0029] According to another aspect of the invention, a method for
making a recommendation for a seed product to plant within a land
base is provided. The method includes collecting crop production
information from a plurality of crop producers over a plurality of
geographically diverse locations, determining environmental
classifications using the crop production information, classifying
the land base with at least one environmental classification, and
determining the recommendation for a seed product to plant within
the field based on the environmental classification.
BRIEF DESCRIPTION OF THE FIGURES
[0030] FIG. 1 is a flow diagram illustrating one process for
determining genotype-by-environment interactions and using that
information in categorizing land bases into different environmental
classifications.
[0031] FIG. 2A to FIG. 2C provide an example of genotype by
environment interactions and cross-over interactions between two
different varieties in four different environmental classes.
[0032] FIG. 3 illustrates environment-standardized GGE biplot of
grain yield of 18 maize hybrids (H1-H18) grown in 266 environments
over three years stratified by state.
[0033] FIG. 4 illustrates environment-standardized GGE biplot of
grain yield of 18 maize hybrids (H1-H18) grown in 266 environments
over three years stratified by environmental class.
[0034] FIG. 5 illustrates one possible schematic for categorizing
different land bases into environmental classifications based on
temperatures, solar radiation, and length of photoperiod.
[0035] FIG. 6 is a bar graph representation of the frequency of
various environmental classes among target population of
environments (TPEs) or multi-environment trials (METs).
[0036] FIG. 7 illustrates potential categories of environmental
classes identified throughout the United States in 1988 and their
locations; these include temperate, temperate dry, temperate humid,
high latitude, and subtropical classes.
[0037] FIG. 8 is a flow diagram illustrating information flow from
an environmental profile and a producer profile to providing
recommendations to a producer according to one embodiment of the
present invention.
[0038] FIG. 9 is block diagram illustrating a system for
determining product recommendations according to one embodiment of
the present invention.
[0039] FIG. 10 is block diagram illustrating a system for
determining product recommendations according to one embodiment of
the present invention.
[0040] FIG. 11 is a screen display according to one embodiment of
the present invention.
[0041] FIG. 12 is a screen display showing a product portfolio
according to one embodiment of the present invention.
[0042] FIG. 13 is a flow diagram illustrating information flow for
one embodiment of a precision farming application of the present
invention.
[0043] FIG. 14 is a screen display for one embodiment of an
application of the present invention.
[0044] FIG. 15 is a screen display for one embodiment of an
application of the present invention.
[0045] FIG. 16 is a screen display for one embodiment of the
present invention showing field-by-field product
recommendations.
[0046] FIG. 17 is a flow diagram illustrating information flow
according to one embodiment of the present invention.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
[0047] The present invention provides methods which can be used to
assist customers, including producers or others in managing risks
related to crop production. Managing risks can be performed by
understanding the relative performance of different agricultural
inputs, including seed products, under the same or similar
environmental conditions as well as understanding variations in the
performance of the same agricultural input over a range of
environmental conditions. By being able to describe and understand
these variations in performance, decisions can be made which are
consistent with overall business and/or production objectives and
limit risk associated with variations in environmental conditions.
These decisions can include what seed products or combination of
seed products to plant, where to plant different seed products,
what other agricultural inputs to use, and what crop management
practices to apply. Utilizing precision agriculture techniques,
these decisions can be made for multiple regions or locations
within a land base.
[0048] One method to manage risks associated with crop production
uses knowledge of genotype-by-environment interactions to assist a
producer or other customer in selecting seed products to plant in
one or more regions within a field. A "genotype" is generally
defined as a cultivar, genetically homogenous (lines, clones), a
hybrid of two or more parents, or heterogeneous (open-pollinated
populations). An "environment" is generally defined as a set of
conditions, such as climatic conditions, soil conditions, biotic
factors (such as, without limitation, pests and diseases) and/or
other conditions that impact genotype productivity. "Management" as
used in this context generally refers to production management
decisions, such as, but not limited to crop production practices.
In addition, the present invention allows for the use of
environmental characterizations to assist in describing
genotype-by-environment interactions. It is to be understood that
the term "genotype-by-environment" (G.times.E) is to encompass what
is sometimes known or referred to as "genotype-by-environment-by
management" (G.times.E.times.M) as the environment associated with
a plant may include management practices which affect the
environment (for example, irrigation may be considered a management
practice, but use of irrigation affects the growing
environment).
[0049] Following, is an exemplary description regarding the use of
G.times.E interactions and environmental classification.
G.times.E and Environmental Classification
[0050] Genetic manipulation alone does not ensure that a plant will
perform well in a specific environment or for that matter a wide
range of environments year after year. In other words, there is no
one genotype that is likely to performance best in all environments
or under all management practices. The performance or phenotype
results from an interaction between the plant's genotype and the
environment and the management practices used.
[0051] It is to be understood that there are some inherent
difficulties in understanding such interactions. An environment at
a given location changes over the years making multi-environment
trials (METs) performed in the same location limited as to
inferences about future crop performance. Furthermore, inferences
about a crop's future performance in different locations depend on
whether the target population of environments (TPEs) is well
sampled since the environment varies between different locations in
one year.
[0052] To assist in analyzing such interactions, the present
invention preferably uses environmental classification techniques.
The environmental classification techniques are used, preferably
with a large set of data to relate performance of different
genotypes to different environments. Environmental classification
is then used when selecting the best seed products for a particular
land base. Thus, for example, a producer can use environmental
classification to select the best seed products for their land base
based on the expected environmental conditions. Alternatively, the
producer may diversify and select a combination of seed products
based on a range of expected environmental conditions to thereby
manage risks associated with environmental variability. Of course,
environmental classification can be used by not just producers but
others having interest in agricultural production.
[0053] FIG. 1 provides an overview of one G.times.E paradigm where
G.times.E knowledge 12 is used in planning and positioning 18.
G.times.E knowledge 12 can be applied to crop modeling 14. Crop
modeling 14 and G.times.E knowledge 12 may either alone or together
be used to classify environments. The G.times.E knowledge 12 and
classified environments may be used in facilitating the positioning
and/or planning 18 strategies, such as characterization of
products, resource efficiency, risk management, product positions,
and product selection.
[0054] Subsequent to positioning and planning, the producer will
grow the selected products 26 and measure the performance results
24. The producer may also collect environmental and physiological
landmark data 28 and in conjunction with performance results 24 use
it in analysis 20. Analysis of environmental and physiological
landmark data 28 and performance results 24 may undergo analysis 20
using G.times.E analysis tools or period-of-years database 22.
Building an Environmental Classification System
[0055] The effectiveness of a product evaluation system for
genotype performance largely depends on the genetic correlation
between multi-environment trials (MET) and the target population of
environments (TPE) (Comstock, R. E. 1977. `Proceedings of the
International Conference on Quantitative Genetics, Aug. 16-21,
1976` pp. 705-18. Iowa State University Press. Ames, USA.). For
example, previous characterizations of maize environments relied
mainly on climatic and soil data (e.g. Hartkamp, A. D., J. W.
White, A. Rodriguez Aguilar, M. Banziger, G. Srinivasan, G.
Granados, and J. Crossa. 2000. Maize Production Environments
Revisited: A GIS-based Approach. Mexico, D. F. CIMMYT.; Pollak, L.
M., and J. D. Corbett. 1993. Agron. J. 85:1133-1139; Runge, E. C.
A. 1968. Agron. J. 60:503-507.). While useful to describe
environmental variables affecting crop productivity, these efforts
did not quantify the impact of these variables on the genetic
correlations among testing sites. Consequently, plant breeders have
more extensively used characterizations of environments based on
similarity of product discrimination in product evaluation trials
(e.g. Cooper, M., D. E. Byth, and I. H. DeLacy. 1993. Field Crops
Res. 35:63-74.). However, these efforts frequently fail to provide
a long-term assessment of the target population of environments
(TPE), mainly due to the cost and impracticality of collecting
empirical performance data for widespread and long-term
studies.
[0056] The present invention provides a modern approach of product
evaluation where a TPE is described. The description of a TPE
includes classifying the land base into an environmental class and
assessing the frequency of occurrence of the range of environments
experienced at a given location. The present inventors contemplate
that areas of adaption (AOA) could also be evaluated. As used
herein AOA refers to a location with the environmental conditions
that would be well suited for a crop or specific genotype. Area of
adaption is based on a number of factors, including, but not
limited to, days to maturity, insect resistance, disease
resistance, and drought resistance. Area of adaptability does not
indicate that the crop will grow in every location or every growing
season within the area of adaption or that it will not grow outside
the area. Rather it defines a generally higher probability of
success for a crop or genotype within as opposed to outside that
area of adaptation.
[0057] The environmental information collected may be used to
develop an environmental database for research locations.
Initially, multiple environment trials are performed by planting
different genotypes available from a variety of sources, e.g.
germplasm, inbreds, hybrids, varieties in multiple environments.
These trials aid the determination of whether the TPEs are
homogenous or should be categorized into different environmental
classifications. The performance data of these genotypes and
environmental and/or physiological landmark data from the MET are
collected and entered into a data set. For example, performance
data collected for a genotype of corn may include any of the
following: yield, grain moisture, stalk lodging, stand
establishment, emergence, midsilk, test weight, protein, oil, and
starch. Yield refers to bushels of grain per acre. Grain moisture
refers to a moisture determination made from each plot at harvest
time, using an instrument such as an electrical conductance
moisture meter. Stalk lodging refers to the determination of the
number of broken stalks in each plot prior to harvest. Stand
establishment refers to the differences between the desired
planting rate for each hybrid and the final stand. Emergence refers
to an emergence count made on each plot after plant emergence where
emergence percentage may be computed based on the number of plants
and the number of kernels planted. The mid silk date is the Julian
day of the year in which 50% of the plants show silks at one site
in a region. The test weights are typically reported as pounds per
bushel on grain samples at field moisture. Protein, oil and starch
are typically reported as a percent protein, oil, and starch
content at a designated percent grain moisture on dried samples
using standard methods, for example, a near infrared transmittance
whole grain analyzer.
[0058] One skilled in the art would be familiar with performance
data collected for other crops, for example, soybeans, wheat,
sunflowers, canola, rice and cotton. Performance data for soybeans
include, without limitation, relative maturity, plant height,
lodging score, seed size, protein and oil percentage, Phytophthora
resistance genes, Phytophthora partial resistance, Sclerotinia
rating, and yield. Relative maturity refers to a determination that
is designed to account for factors, such as soybean variety,
planting date, weather, latitude and disease that affect maturity
date and number of days from planting to maturity. Plant height
refers to a determination of the soybean plant's height, usually
determined prior to harvest. Lodging, traditionally, the vertical
orientation of the plant, i.e. the degree to which the plant is
erect. The lodging of a soybean plant is traditionally rated by
researchers using a scale of 1 to 9 as follows: 1.0=almost all
plants erect, 3.0=either all plants leaning slightly, or a few
plants down, 5.0=either all plants leaning moderately (45 O angle),
or 25-50% down, 7.0=either all plants leaning considerably, or
50-80% down, 9.0=all plants prostrate. The seed size of a soybean
plant typically refers to thousands of seeds per pound. Protein and
oil percentage analysis may be determined using near infrared
transmittance technology and reported at 13% moisture. Phytophthora
resistance genes may be determined using a hypocotyl inoculation
test with several races of Phytophthora to determine the presence
or absence of a particular Rps gene in a soybean plant. Soybeans
may also be evaluated for phytophthora partial resistance using a
ratings system, where ratings of 3.0 to 3.9 are considered high
levels of partial resistance, ratings of 4.0 to 5.9 are considered
moderate, ratings over 6.0 indicate very little partial resistance
or protection against Phytophthora. Soybeans may also be evaluated
for partial resistance to Sclerotinia. Yield refers to bushels per
acre at 13 percent moisture.
[0059] Typical performance data for wheat includes, without
limitation, test weight, protein percent, seed size, percent
lodging, plant height, heading date, powdery mildew, leaf blotch
complex (LBC), Fusarium head scab (FHS), flour yield, and flour
softness. Test weight refers to a determination of pounds/bushell
using harvest grain moisture. Seed size refers to thousands of
harvested seeds per pound. Percent lodging as described previously
refers to a rating system used to estimate the percent of plants
that are not erect or lean more than 45 degrees from vertical.
Plant height refers to the distance from the soil surface to the
top of the heads. Heading date refers to the average calendar day
of the year on which 50 percent of the heads are completely
emerged. Wheat infected with powdery mildew (PM) may be determined
using a scale system where each plot is rated based on a 0 to 10
scale where: 0=0 to trace % leaf area covered; 1=leaf 4 with trace
-50%; 2=leaf 3 with 1-5%; 3=leaf 3 with 5-15%; 4=leaf 3 with
>15%; 5=leaf 2 with 1-5%; 6=leaf 2 with 5-15%; 7=leaf 2 with
>15%; 8=leaf 1 with 1-5%; 9=leaf 1 with 5-15%; and 10=leaf 1
with >15% leaf area covered (leaf 1=flag leaf). This scale takes
into account the percentage leaf area affected and the progress of
the disease upward on the plants. Leaf blotch complex (LBC) caused
by Stagonospora nodorum, Pyrenophora tritici-repentis and Bipolaris
sorokiniana for example may be determined when most varieties are
in the soft dough growth stage and rated based on the percentage of
flag leaf area covered by leaf blotches. Fusarium head scab (FHS)
caused by Fusarium graminearum for example may be determined when
plants are in the late milk to soft dough growth stage and each
plot is rated based on a disease severity estimate as the average
percentage of spikelets affected per head. Flour yield refers to
the percent flour yield from milled whole grain. Flour softness
refers to the percent of fine-granular milled flour. Values higher
than approximately 50 indicate kernel textures that are appropriate
for soft wheat. Generally, high values are more desirable for
milling and baking.
[0060] Typical performance data for sunflower includes, without
limitation, resistance to aphids, neck breakage, brittle snap,
stalk breakage, resistance to downy mildew (Plasmopara halstedii),
height of the head at harvest, seed moisture, head shape,
hullability, resistance to the sunflower midge, Contarinia schulzi,
percentage of oil content, seed size, yield, seedling vigor, and
test weight. Resistance to aphids refers to a visual ratings system
indicating resistance to aphids based on a scale of 1-9 where
higher scores indicate higher levels of resistance. Neck breakage
refers a visual ratings system indicating the level of neck
breakage, typically on a scale from 1 to 9 where the higher the
score signifies that less breakage occurs. Brittle snap refers to a
visual rating system indicating the amount of brittle snap (stalk
breakage) that typically occurs in the early season due to high
winds. The ratings system is based on a scale, usually ranging from
1-9, with a higher score denoting the occurrence of less breakage.
A sunflower's resistance to Downy Mildew (Plasmopara halstedii) may
be determined using a visual ratings scaled system with 9 being the
highest and 1 the lowest. A higher score indicates greater
resistance. Height of the head at harvest refers to the height of
the head at harvest, measured in decimeters. Seed moisture refers
to a determination of seed moisture taken at harvest time, usually
measured as a percentage of moisture to seed weight. Head shape of
a sunflower is measured visually using a scale system where each
plot is rated based on a 1 to 9 scale where: 1=closed "midge" ball;
2=trumpet; 3=clam; 4=concave; 5=cone; 6=reflex; 7=distorted;
8=convex; 9=flat. Hullability refers to the ability of a hulling
machine to remove seed hulls from the kernel, typically measured on
a 1-9 scale where a higher score reflects better hullability.
Resistance to the sunflower midge, Contarinia schulzi, is
determined based on head deformation which is rated on a 1-9 scale
where: 9=no head deformation (fully resistant), 5=moderate head
deformation, 1=severe head deformation (fully susceptible). The
percentage of oil content from the harvested grain is measured and
adjusted to a 10% moisture level. The oil content of a sunflower
seed may be measured for various components, including palmitic
acid, stearic acid, oleic acid, and linoleic acid, using a gas
chromatograph. Seed size refers to the percentage of grain that
passes over a certain size screen, usually "size 13." Seedling
vigor refers to the early growth of a seedling and is often times
measured via a visual ratings system, from 1-9, with higher scores
indicate more seedling vigor. Yield is measured as quintals per
hectare, while test weight of seed is measured as kilograms per
hectoliter.
[0061] Typical performance data for canola includes, without
limitation, yield, oil content, beginning bloom date, maturity
date, plant height, lodging, seed shatter, winter survival, and
disease resistance. Yield refers to pounds per acre at 8.5%
moisture. Oil content is a determination of the typical percentage
by weight oil present in the mature whole dried seeds. Beginning
bloom date refers to the date at which at least one flower is on
the plant. If a flower is showing on half the plants, then canola
field is in 50% bloom. Maturity date refers to the number of days
observed from planting to maturity, with maturity referring to the
plant stage when pods with seed color change, occurring from green
to brown or black, on the bottom third of the pod bearing area of
the main stem. Plant height refers to the overall plant height at
the end of flowering. The concept of measuring lodging using a
scale of 1 (weak) to 9 (strong) is as previously described. Seed
shatter refers to a resistance to silique shattering at canola seed
maturity and is expressed on a scale of 1 (poor) to 9 (excellent).
Winter survival refers to the ability to withstand winter
temperatures at a typical growing area. Winter survival is
evaluated and is expressed on a scale of 1 to 5, with 1 being poor
and 5 being excellent. Disease resistance is evaluated and
expressed on a scale of 0 to 5 where: 0=highly resistant, 5=highly
susceptible. The Western Canadian Canola/Rapeseed Recommending
Committee (WCC/RRC) blackleg classification is based on percent
severity index described as follows: 0-30% =Resistant,
30%-50%=Moderately Resistant, 50%-70%=Moderately Susceptible,
70%-90%=Susceptible, and >90%=Highly susceptible.
[0062] Typical performance data for cotton includes, without
limitation, yield, turnout, micronaire, length, fiber strength of
cotton and color grade. Yield is measured as pounds per acre.
Turnout refers to lint and seed turnout which is calculated as the
percentage of lint and seed on a weight basis as a result of
ginning the sub sample from each treatment. Micronaire refers to
fiber fineness and maturity and are measured using air flow
instrument tests in terms of micronaire readings in accordance with
established procedures. Fiber length is reported in 1/32 of an inch
or decimal equivalents. Fiber strength is measured in grams per tex
and represents the force in grams to break a bundle of fibers one
tex unit in size. Color grade for cotton takes into consideration
the color, fiber color and whiteness of cotton leaves. Color grade
may be determined using a two digit scale. The two digit number is
an indication of the fiber color and whiteness (i.e. 13, 51, or
84). The first digit can range from 1 to 8 representing overall
color with 1 being the best color and 8 representing below grade
colors. The second digit represent a fiber whiteness score. This
number ranges from 1 to 5, with 1 representing good white color and
5 representing yellow stained. The second number in the overall
color grade represents the leaf score and represents leaf content
in the sample.
[0063] Typical performance data for rice includes, without
limitation, yield, straw strength, 50% Heading, plant height, and
total milling, and total milling. Yield is measured as bushels per
acre at 12% moisture. Straw Strength refers to lodging resistance
at maturity and is measured using a numerical rating from 1 to 9
where 1=Strong (no lodging); 3=Moderately strong (most plants
leaning but no lodging); 5=Intermediate (most plants moderately
lodged); 7=Weak (most plants nearly flat); and 9=Very weak (all
plants flat). 50% heading refers to the number of days from
emergence until 50% of the panicles are visibly emerged from the
boot. Plant height is the average distance from the soil surface to
the tip of erect panicle. Total milling refers to the total milled
rice as a percentage of rough rice. Whole milling refers to rice
grains of 3/4 length or more expressed as a percentage of rough
rice.
[0064] The environmental and physiological landmark data may be
historical using historical meteorological information along with
soils and other agronomic information or collected using National
Oceanic and Atmospheric Association and/or other public or private
sources of weather and soil data. Potential environmental and
physiological landmark data that may be collected includes but is
not limited to wind, drought, temperature, solar radiation,
precipitation, soil type, soil pH, planting and harvesting dates,
irrigation, tiled area, previous crop, fertilizer including
nitrogen, phosphorous, and potassium levels, insecticide,
herbicide, and biotic data, for example, insects and disease. The
environmental and physiological landmark data may then be analyzed
in light of genotype performance data to determine G.times.E
interactions.
Models
[0065] Several models for determining G.times.E interactions exist.
Base models group or classify the locations used to test the
hybrids, include several variance components, and stratify the
hybrids, for example, according to locations among station-year
combinations, locations, or other chosen variances.
[0066] For example, as shown in Table 1, one base model Year
Station (YS) groups the locations by year-stations where a
year-station designates a unique site or location by year. Other
variances include blocks within locations within year-stations,
hybrids, hybrids by year-station divided by the sum of hybrids by
locations within year station locations as well as a residual. The
YS model is disadvantageous in that a given location's environment
will vary over time so that the G.times.E information gleaned from
the model may not be relevant for predicting hybrids that will
perform well in the same location next year.
[0067] Another model for determining G.times.E interactions
disclosed in Table 1, groups different sites by location. Other
variances for the G.times.E model include blocks within locations,
hybrids, hybrids by locations, as well as a residual. However, the
G.times.E model is disadvantageous in that a genotype grown in
locations with differing environmental conditions may have similar
performance results, complicating the analysis of the specific
environmental conditions that play a role in contributing to
genotype performance and reducing the certainty of predicting
product performance.
[0068] Unlike the previous models mentioned, the present inventors
contemplate determining G.times.E interactions using a model
referred to herein as Environmental Classification that groups
locations by environmental classifications. Thus, variances for
this model include locations within environmental classifications,
blocks within locations within environmental classifications,
hybrids, hybrids by environmental classifications divided by
hybrids by locations within environmental classifications and a
residual. TABLE-US-00001 TABLE 1 Models for determining G .times. E
interactions Environmental Model Year-Station G .times. E
Classification Variance for Location within Location Location
within location year-station environmental classification Variance
for blocks within blocks within blocks within location locations
within locations locations within year-station environmental
classifications Variance for hybrids hybrids hybrids hybrids
Stratifications hybrid by year- hybrid by hybrid by station/hybrids
by locations environmental locations within classifications/
locations hybrid by locations within environmental
classifications
[0069] Burdon has shown that genetic correlation between G.times.E
interactions can be estimated. (Burdon, R. D. 1977. Silvae Genet.,
26: 168-175.). G.times.E analysis may be performed in numerous
ways. G.times.E interactions may be analyzed qualitatively, e.g.
phenotype plasticity, or quantitatively using, for example, an
analysis of variance approach. (Schlichting, C. D. 1986. Annual
Review of Ecology and Systematics 17: 667-693.). Statistical
analysis of whether a G.times.E interaction is significant and
whether environmental changes influence certain traits, such as
yield performance, of the genotypes evaluated may be performed
using any number of statistical methods including but not limited
to, rank correlation, analysis of variances, and stability.
Rank Correlation
[0070] The most basic categorization of G.times.E interaction is to
evaluate G.times.E interactions by performing a rank correlation
according to standardized tests, for example, Spearman. The
Spearman rank correlation may be performed to examine the
relationships among genotypes in different environments, for
example, crossover interactions that occur when two genotypes
change in rank order of performance when evaluated in different
environments. FIG. 2 illustrates an example of G.times.E
interactions and cross-over interactions (COI) between two
different varieties, Var A and Var B, in four different
environmental classes, Env 1, Env 2, Env 3 and Env 4. FIG. 2A shows
that Var A and Var B out-perform each other in different
environments indicating the occurrence of both G.times.E and COI.
FIG. 2B shows that Var A performed better than Var B in each
environment, indicating G.times.E interactions but no COI. In
contrast, FIG. 2C shows that Var A and Var B each performed
consistently with respect to each other in all four environments,
indicating lack of G x E interactions.
Analysis of Variance (ANOVA)
[0071] Alternately, G.times.E interactions may be analyzed using an
analysis of variance method (ANOVA) (Steel, R. G. D and J. H.
Torrie. 1980. Principles and Procedures of Statistics, 2nd edition)
over environments to determine the significance of genotypes,
environments and G.times.E interactions. G.times.E interactions may
also be analyzed using ASREML (Gilmour, A. R., Cullis, B. R.,
Welham, S. J. and Thompson, R. 2002 ASReml Reference Manual 2nd
edition, Release 1.0 NSW Agriculture Biometrical Bulletin 3, NSW
Agriculture, Locked Bag, Orange, NSW 2800, Australia.) for the
computation of variance components, and the generation of GGE
biplots (Cooper, M., and I. H. DeLacy. 1994. Theor. Appl. Genet.
88:561-572; Yan, W. and M. S. Kang. 2003. GGE Biplot Analysis: A
Graphical Tool for Breeders Geneticists, and Agronomists. CRC
Press. Boca Raton, Fla.). FIG. 3 and FIG. 4 illustrate
environment-standardized GGE biplot of grain yield of 18 maize
hybrids (H1-H18) grown in 266 environments over three years,
stratified by state or by environmental class respectively.
Stability
[0072] Once certain genotypes are identified that perform well in a
target environment they may be analyzed to determine which hybrids
are more stable in yield or other metrics using various methods.
One method uses a regression of genotypic performance on an
environmental index. In general, the environmental index is the
deviation of the mean phenotype at environment from the overall
mean phenotype of all environments. Thus, the phenotype of an
individual genotype with each environment is regressed on the
environmental index, as described in Bernardo R. 2002. Quantitative
Traits in Plants. Stemma Press, Woodbury, Minn. to generate a slope
(b-value) for each genotype/cultivar evaluated. Other methods
include the joint regression analysis method proposed by Perkins,
J. M. and Jinks, J. L. 1968. Heredity. 23: 339-359, Finlay, K. W.
and Wilkinson, G. N. 1963. Aust. J. Res. 14: 742-754 and Eberhart,
S. A. and Russell, W. A. 1966. Crop Sci. 6:36-40 to calculate the
regression coefficient (b), S.E. and variance due to deviation from
regression (S2d) as a parameter of stability and adaptability. The
model described by Eberhart and Russell has the following formula:
P.sub.ij=.mu.'g.sub.i+b.sub.it.sub.j+.delta..sub.ij+e.sub.ij where
P.sub.ij is the mean phenotype of genotype or cultivar i in
location j, [0073] .mu. is the grand mean across the whole
experiment for all genotypes and locations, [0074] g.sub.i is the
effect of genotype i across all locations [0075] b.sub.i is the
linear regression of P.sub.ij on t.sub.j, [0076] t.sub.j is the
environmental index, that is the effect of environment j across all
genotypes), [0077] .delta..sub.ij is the deviation P.sub.ij from
the linear regression value for a given t.sub.j and [0078] e.sub.ij
is the within environment error. Categorization of Land Bases into
Environmental Classes
[0079] Using the information collected for or from G.times.E
analysis, the land bases may be categorized into environmental
classifications. FIG. 5 illustrates one possible schematic for
categorizing different land bases into environmental
classifications. With reference to FIG. 5, one method of
categorizing environmental classifications is illustrated as a flow
chart. If all maximum temperatures are greater than 28.degree.
Celsius 42, then the land base may be categorized as either
Temperate Dry 54, Temperate Humid 52, Temperate 56, or Subtropical
48. If all maximum temperatures are greater to or equal to
30.degree. Celsius and solar radiation is greater than 24 and 21 at
a given crop development stage, e.g. v7-R1, R3-R6 40, then the land
base is characterized as Temperate Dry 54. If the maximum
temperature is not greater than or equal to 30.degree. Celsius and
solar radiation is not greater than 24 at a given crop development
stage, e.g. V7-R1 and 21 for R3-R6 respectively 40, then the land
base is characterized as Temperate 56. However, if the maximum
temperature is less than 30.degree. Celsius and solar radiation is
greater than 24 and 21 at a given crop development stage 50, then
the land base is characterized as Temperate Humid 52. If the
maximum temperature is not less than 30.degree. Celsius and solar
radiation is not greater than 24 and 21 at a given crop development
stage 50, then the land base is characterized as Temperate 56. If
all maximum temperatures 42 for the land base are less then
28.degree. Celsius than the land base is characterized as High
Latitude 44. In contrast, if all maximum temperatures 42 for the
land base are not less then 28.degree. Celsius and the land base
has a photoperiod less than 13.4 hours/day 46, then the land base
is Subtropical 48.
[0080] Categorizing land bases into environmental classifications
has several advantages. First, environmental classifications can
bring an understanding of the various environments under which
crops are produced. Second, occurrence probabilities for each
environmental category can be assigned to each geographic location
and the frequency of the classifications determined using routine
methods. FIG. 6 is a bar graph representation of the frequency of
various environmental classes among TPEs or METs. The frequency for
each environmental class, e.g. temperate, temperate dry, temperte
humid, high latitude, and subtropical, is given as a percent of the
total TPE or MET tested in given year or across years. FIG. 7
illustrates potential categories of environmental classes
identified throughout the United States in 1988 and their
locations; these include temperate, temperate dry, temperate humid,
high latitude, and subtropical classes. It will be apparent to one
skilled in the art that other environmental classifications may
added as identified or deemed relevant to G.times.E interactions
for various crops.
[0081] Some of the environmental classification may be defined
using general characteristics of climates. For example, temperate
may be used to refer to regions in which the climate undergoes
seasonal change in temperature and moisture; typically these
regions lie between the Tropic of Capricorn and Antarctic circle in
the Southern Hemisphere and between the Tropic of Capricorn and the
Arctic circle in the Northern Hemisphere. Temperate humid may refer
to regions in which the climate undergoes seasonal change in
temperature and moisture and has more humidity than a temperate
environment. High latitude as an environmental class may refer to
regions that have a longer photoperiod than and is typically north
of a particular latitude. A subtropical class may refer to regions
enjoying four distinct seasons usually with hot tropical summers
and non-tropical winters with a shorter photoperiod/day length;
typically these regions lie between the ranges 23.5-40.degree. N
and 23.5-40.degree. S latitude. The environmental classes may also
be defined by biotic factors, such as diseases, insects, and/or
characteristic of a plant. For example, an ECB class may refer to
regions having European Corn Borers (ECB) or the suspected presence
of ECB as evidenced by preflowering leaf feeding, tunneling in the
plant's stalk, post flowering degree of stalk breakage and/or other
evidence of feeding. The environmental class Brittle may be used to
refers to regions where stalk breakage of corn occurs or is apt to
occur near the time of pollination and is indicative of whether a
hybrid or inbred would snap or break near the time of flowering
under severe winds.
[0082] It is to be understood that the environmental
classifications may be used and defined differently for different
crops/genotypes and that these definitions may vary from year to
year, even for the same crops or genotypes. For example, in
2000-2003, trials conducted studying G.times.E interactions among
Comparative Relative Maturity (CRM) hybrids of CRM 103-113 in
different environments identified seven different environmental
classes--temperate, temperate dry, temperate humid, high latitude,
subtropical, ECB, and brittle. For the study purposes, temperate
was identified/defined as having a low level of abiotic stresses, a
growing season adequate for CRM 103-113, and found to be frequent
in Iowa and Illinois. Temperate dry was defined as temperate with
some level of water and/or temperature stress and found to be
frequent in Nebraska, Kansas, and South Dakota. Temperate Humid was
defined as similar to the temperate environmental class but had a
complex of biotic factors, such as leaf disease that may
differentially affect product performance. Temperate humid was also
characterized as having a temperature and solar radiation lower
than that identified in the temperate environmental class and found
to be frequent in Indiana, Ohio, and Pennsylvania. The High
Latitude environmental class was found to grow corn CRM 103 and
earlier (growing hybrids) and experienced colder temperatures than
the Temperate environmental class but with longer day-length. This
environmental class was found to be frequent in Canada, North
Dakota, Minnesota, Michigan, and Wisconsin. The fifth environmental
class, Subtropical, was characterized as warm and humid with a
short day-length and found frequently in the Deep South of the
United States. Another environmental class identified was European
Corn Borers (ECB) and defined as having Bacillus thuringiensis (Bt)
hybrids that outyielded base genetics by at least 10%. The last
environmental class Brittle defined areas with significant brittle
damage with differential effect on products.
[0083] Once areas of land are categorized as environmental classes,
these areas may be used in METs. Ultimately, the observed genotype
performances in METs can be linked by the environmental class to
the TPE. By evaluating product performance in a target environment,
rather than merely performance differences in METs, genotype
performance data from multiple test environments can be correlated
to a target environment and used to predict product performance.
This correlation between a genotype's performance and the target
environment or environmental classification will lead to more
precise product placement since the genotype performance is
characterized within an environmental class in which it is adapted
and most likely to experience after commercialization, consequently
resulting in improved and more predictable product performance. The
analysis of G.times.E interactions facilitates the selection and
adoption of genotypes that have positive interactions with its
location and its prevailing environmental conditions (exploitation
of areas of specific adaption). G.times.E analysis also aids in the
identification of genotypes with low frequency of poor yield or
other performance issues in certain environments. Therefore,
G.times.E analysis will help in understanding the type and size of
G.times.E interactions expected in a given region. The present
inventors contemplate that proper selection of hybrids for a
particular land base will improve agricultural potential of certain
geographic areas by maximizing the occurrence of crop performance
through the use of the environmental classification. In addition,
this approach allows the use of statistical and probability based
analysis to quantify the risk of product success/failure according
to the frequency of environment classes and the relative
performance of genotypes within each environment class. This early
identification and selection of hybrids would enable seed producers
to start seed production and accelerate the development of hybrids
in winter nurseries in warmer southern climates.
[0084] Moreover, environmental classification allows for the
creation of an environmental profile for all or any part of the
land base classified. Environmental classifications can be
determined for each producer's land base. Similarly, the
environmental performance profile of cultivars/hybrids can be
determined through field experimentation or predicted using
G.times.E analysis. In combining environmental classification
frequencies for a particular land base and product performance by
environmental classification, performance measurements are given
the appropriate amount of relevance or weight for the land base in
question. For example, the data are weighted based on long-term
frequencies to compute a prediction of hybrid performance.
Use of G.times.E in Producer's Selection
[0085] According to another aspect of the present invention, a
method of using information that documents the environmental
profile over time of a crop producer's land base, the environmental
performance profile of crop cultivars, and the producer's
objectives to select a portfolio of cultivars that maximizes and
quantifies the probability that the producer's objectives for
productivity will be met. Environmental classification can be used
to assist in this process.
[0086] Environmental classification can be used to determine the
primary environmental drivers of G.times.E interaction in crops
such as corn. That is, what are the primary environmental factors
that cause change in the relative performance of hybrids. With this
knowledge, crop production areas can be categorized into
environmental frequency classes. Within these classes, hybrids tend
to perform (as measured by yield) relatively similar to one
another. Across these classes, the relative performance of hybrids
tends to be significantly different. Using historical
meteorological information along with soils, pests, and other
agronomic information, the frequency of these environments can be
determined. This allows the creation of an environmental profile
for all or any part of the geography classified. That is, a
frequency distribution of the occurrence of the key Environment
Classes. This can be done for each crop producer's land base.
[0087] Similarly, the environmental performance profile of crop
cultivars can be determined through field experimentation. That is,
a description of relative performance of cultivars can be
determined in each of the key environment classes. In combining
Environment class frequencies for a particular land area and
product performance by Environment Class, performance measurements
are given an appropriate amount of relevance or weight for the land
area in question Thus, this aspect of the invention involves
combining of this information at the producer's level to optimize
crop productivity in such a way that it maximizes the probability
of the producer's business operation reaching its productivity
goals. The present invention contemplates that information can be
used from any number of classification schemes to the selection of
cultivars with the objective of maximizing the probability of
attainment of the productivity and business goals of a crop
producer's operation.
[0088] The approach of this aspect of the present invention does so
by using compiled long term geo-referenced weather, soils, and
agronomic data including biotic factors for the producer's land
base to categorize the land base in terms of how frequently annual
environmental variation occurs to a degree that is likely to impact
relative hybrid performance. In addition, it can incorporate the
producer's business objectives including, but not limited to
preparedness to take risk. The present invention is able to combine
environmental variability with producer business information to
create a producer profile. Product performance information
stratified by the same criteria is used to define the producer's
environmental profile (for example, environmental classes) which is
then integrated with the producer's profile.
[0089] The relative hybrid performance information that is relevant
to the producer's land base is used regardless of when and where it
was generated. The present inventors are first to predict future
performance of genotypes and quantify probability/risk associated
with that performance using data from environments that are
considered to be substantially equivalent in terms of relative
hybrid response. The result is a more robust and predictive data
set thus allowing more informed product selection decisions that,
over time will result in a higher probability of a producer
operation meeting business objectives for productivity.
[0090] FIG. 9 illustrates information flow according to one
embodiment of the present invention. In FIG. 9 there is an
environmental profile 100. The environmental profile can be based
on one or more inputs such as environment classes 102,
meteorological information 104, agronomic information 106, or field
experimentation 108. In FIG. 1 there is also a producer profile
110. The producer profile 110 is based on one or more inputs such
as risk tolerance 112 of the producer, business goals 114 of the
producer, productivity goals 116, financing 118 considerations,
third party needs 119, for example a landlord, or insurance/risk
management and marketing 120 considerations. The environmental
profile 100 and the producer profile 110 are combined in order to
produce recommendations 122. The recommendations 122 can include
risk management tools, a recommended seed product, a recommended
mix of seed products, production practice recommendations, such as
chemical application information, or any number of other specific
recommendations as may be appropriate based on the particular
environmental profile 100 and producer profile 110.
[0091] FIG. 10 illustrates one embodiment of a system 124 for
producing product recommendations. In FIG. 9, a processor 126
accesses information associated with a producer profile 110, an
environmental profile 100, and a genotype by environment database
132. There is an input device 128, a recommendation output 129, and
a display 130 operatively connected to the processor. The present
invention contemplates that the processor 126 can be associated
with a computer such as handheld computer as may be convenient for
a dealer or sales agent. The present invention also contemplates
that the producer profile 110, environmental profile 100, and
genotype by environment database 132 may be accessible over a
network, including a wide-area network such as the Internet.
[0092] Using the information in the producer profile 110,
environmental profile 100, and genotype-by-environment database
132, the processor applies one or more of a product selection
algorithm module 134, a product comparator 136, a production
practice module and a risk comparator 138, and a product portfolio
module 140. These and/or other modules are collectively the
recommendation logic 142. In a simple case, the product selection
algorithm module 134 would take information from the environmental
profile 100, such as an environmental classification ("Temperate",
for example) in addition to information from the producer profile
110, such as a producer objective ("Maximize Yield", "Risk
Minimization", "Low Harvest Moisture" for example) and match these
criteria to products in the genotype-by-environment database 132.
Of course, more specific criteria could be examined as would be the
case with more complex environmental profile information and more
complex producer profile information.
[0093] FIG. 10 illustrates one embodiment of a screen display 144
of a software application the present invention. In FIG. 10, a user
is given the choice of selecting "DEFINE ENVIRONMENTAL PROFILE"
146, "DEFINE PRODUCER PROFILE" 148, and "VIEW RECOMMENDATIONS" 150.
Of course, the present invention contemplates that software and its
accompanying user interface can be implemented in any number of
ways.
[0094] FIG. 11 illustrates one embodiment of a screen display 152
of a software application of the present invention. In FIG. 11, a
recommendation is given which n A includes a plurality of products
154, an associated number of acres 156 associated with each of the
products, a risk/probability assessment 157, and a recommended crop
revenue assurance 158. The present invention provides for
decreasing the amount of risk associated with selection of a
particular seed product by instead selecting multiple products with
different G.times.E interactions in order to reduce risk associated
with environmental variations. The resulting selection, is somewhat
akin to selection of stocks in a stock portfolio.
[0095] FIG. 12 and FIG. 13 illustrate embodiments of user
interfaces to use in precision farming applications. In FIG. 12,
the user interface 170 includes site-specific information
associated with location information 172. The present invention
contemplates that other site-specific information or historical
information is accessible based on the location information 172 and
may be used in product selections. In addition, environment and
production information is collected. Examples of such information
includes maturity days 176, input traits 178, output traits 180,
seed treatment 182, whether no till practices 174 are used, the
planting population 184, nitrogen utilization 186, and drought
impact based on environmental classification drought frequency
information 187 and soil type. Based on this information and
information associated with the location 172, a recommendation 188
of at least one hybrid seed product is made. Where multiple
recommendations are made, the recommendations can be ranked as well
as a risk assessment 189-such as shown.
[0096] FIG. 13 illustrates another embodiment of a user interface
200 that can be used in crop production applications. Site specific
information is collected such as location 172, soil type 174, and
number of acres 202. In addition, there is the option to import
precision farming data 204 as well as import environment of
frequency data 205. There are also the options to set production
practices, set environmental assumptions, set risk levels, and set
the maximum number of hybrids 212. Based on the inputs, a portfolio
is created that includes a plurality of products 214, an associated
number of acres 216 to plant for each product, a recommendation 217
of at least one hybrid seed product, a risk assessment 218, and
revenue assurance 219. Where multiple recommendations are made, the
recommendations can be ranked. There is also an option to generate
precision farming information 220 based on this information, such
as a prescription map. The present invention contemplates that the
precision farming information may indicate which acres to plant
with which hybrids, give specific production practice application
(such as chemical application rates), or other recommendations.
[0097] FIG. 14 illustrates one example of a field-by-field analysis
showing product recommendations for a land base of a producer. As
shown in FIG. 14, different land areas within a producer's land
base have different hybrids associated with them. The present
invention contemplates producing such a map or field-by-field
recommendations where multiple products are recommended. It should
further be understood that a single producer or other user may have
operations in a number of geographically diverse locations, and not
necessarily the nearby fields illustrated in FIG. 14.
[0098] It should also be appreciated that the use of environmental
classification and G x E interactions should be effectively
communicated to customers. The effectiveness of the environmental
classification process is based in part on its ability to use
historical data from many locations so that all available data is
used. This aspect of environmental classification would seem
counter-intuitive to a customer who primarily relies upon personal
knowledge in the local area. The customer's confidence in firsthand
production knowledge can be used to assist in increasing confidence
in environmental classification.
[0099] FIG. 15 illustrates one example of the methodology of this
aspect of the invention to assist in explaining these concepts to a
producer. In step 300 site-specific data collection for a land base
is performed. Based on this site-specific data collection, in step
302, the land base is given an environmental classification. In
addition to this information, the type of hybrid selected in the
previous year and its performance is provided by the producer in
step 304. In step 306, a prediction is made as to the previous
year's production based on environmental classification. In step
308, the predicted results are compared with the actual results.
The present invention also contemplates not requiring performance
results from the producer until after the previous year's results
have been predicted in case the producer is not confident that an
independent prediction is made.
[0100] FIG. 16 illustrates one example of a screen display showing
such comparisons. In FIG. 16, performance predictions (yield) are
made for a number of different hybrids for both the previous year
and the current year. In addition, a risk assessment for each
hybrid may also be provided. The producer can compare the
prediction for the previous year with the actual performance for
that year in order to understand how well the environmental
classification method can predict a result. If the producer is
confident in the method's ability to correctly predict a result,
the producer will be more inclined to use the prediction made for
the coming year. The present invention contemplates that the same
or similar information can be presented in any number of ways. It
should further be understood that such a demonstrate assists in
illustrating the accuracy of the system in predicting relative
performance differences between seed products. Due to the number of
potential variables and difficulty in controlling such variable
accurate prediction of absolute performance is generally not a
reasonable goal. However by selecting appropriate environmental
classifications, useful insight into relative performance can be
provided.
Precision Agriculture and Environmental Classification
[0101] Precision farming or site specific farming recognizes that
increased information and precise control over crop production
processes can result in more efficient use of resources. The use of
Global Positioning System (GPS) and variable rates of production
inputs and outputs are often closely associated with precision
farming. Concepts in precision farming recognize variations
throughout a land base, including throughout the same field.
Variability can include variable outputs (i.e. yield) as well as
variability in inputs obtained from the farmer (i.e. seeds,
chemicals, etc.) as well as variations in the soil properties or
variations in other environmental characteristics. Thus, examples
of site specific data can include data relating to soil type, soil
pH, irrigation related information, tiled area information,
previous crop information, fertilizer including nitrogen,
phosphorous, and potassium levels, insecticide, herbicide, and
biotic data, drainage topography (soil moisture and stresses), crop
variety (disease resistance, root systems, ability to adapt to
extreme conditions), insect or weed information, crop rotation
information, tillage practices (type, timing, wet/dry soil),
compaction, pH (extreme pH variability (<5.5 or >7.5),
herbicides (misapplication, drift, phytotoxic effects), subsoil
condition (acid or alkaline subsoil, clay layer, fragipan, etc),
fertility placement (ridge-till, no-till, etc), fertility level,
plant population, etc.
[0102] Some agricultural enterprises have historic information
which may include soil information, yield information, chemical
application information and other types of information that may be
associated with or collected through precision farming techniques.
This information is usually geo-referenced with GPS
information.
[0103] The present invention allows for information obtained
directly from a customer such as a crop producer to be used in
conjunction with environmental classification information to
predict performance of seed products under various combinations of
conditions. Moreover, the present invention provides for specific
recommendations of seed products or combinations of seed products
for the crop producer to reduce risk. These recommendations are
also not necessarily limited to the selection of seed products but
also recommendations regarding other crop management selections
including use of particular chemicals.
[0104] The use of precision agriculture-type information is
advantageous as this additional information enhances historical
data and can improve future predictions of appropriate products.
The use of precision agriculture-type information is also
beneficial in building a relationship with producers and in gaining
acceptance of the use of environment classification information and
models. In particular, it may seem unintuitive to crop producers to
use information from far off locations or otherwise geographically
diverse locations in predicting the appropriate products for their
use as is preferably performed using environmental classification
of the present invention. In fact, use of such information runs
completely contrary to the notion of "site specific" farming. The
combining of this information with site specific crop producer
experienced and/or verified information may assist in crop producer
acceptance of the adoption of environmental classification and
related products and services.
[0105] The present invention further provides for incorporating the
environmental classification information into a crop producer's
precision farming program. In particular, high resolution
environmental classification could be used to make different
selections of seed products and rate of planting for different
areas of the same field based on the known variables such as soil
properties and accordingly taking into account genetic by
environment interaction. Environmental classifications can be taken
at different levels. Although a preferred level is the township
level, the level may be of a higher resolution including by field
or by grid area within a field. and could be incorporated at a
higher resolution field, with a preferred level being a township.
Planters such as that described in U.S. Pat. No. 5,915,313 to
Bender et al, and assigned on its face to Case Corporation, herein
incorporated by reference in its entirety, allow for planting
multiple types of seeds within the same field in a single pass
according to a prescription type map. The environmental
classification of the present invention could be used to provide
for a means to create such as a prescription type map as it makes
available appropriate information about particular seed products
and provides an improved perspective on genetic by environment
interactions. Where environmental classification using appropriate
classes is used, the present invention provides for selecting the
appropriate seed product for each grid or location. Thus the
environmental classification of the present invention could also be
used to assist in precision farming. It should also be understood
that the present invention is not limited to making appropriate
seed product recommendations, but can also include providing other
recommendations as to production practices which may be appropriate
based on predicted genotype-by-environment interactions. Examples
of other types of recommendations include herbicide, pesticide,
fertilizer types, application rate, and application timing,
irrigation recommendations, and any number of other recommendations
as to the production techniques to be employed.
[0106] FIG. 13 illustrates information flow according to one
embodiment of the present invention. Environmental classification
information 160 is used to provide genotype-by-environment
information and incorporates information that is non site-specific.
The environmental classification information 160 could be used in
conjunction with precision farming information 162 which is
site-specific. In addition site specific environmental information
on drought risk, for example, soils, environmental classification,
drought frequency, is incorporated. This combination of information
is used to provide recommendation information 164. The
recommendation information 164 can take on numerous forms and may
include product recommendations as well as production practice
recommendations. Examples of such information include prescription
maps with different seed products, different seed product
recommendations for different land bases, product recommendations
with associated risk scenarios and related predictions.
[0107] Information is then obtained based on the planting of the
products, preferably based on the product recommendations. This
experiential data 166 can then be used to assist in the future
development of environmental classification 160 as well as being
incorporated into the site specific precision farming information
162.
[0108] Information may also be obtained from the producer to assist
an agricultural input supplier in forming environmental
classifications for a variety of agricultural inputs. FIG. 17
illustrates information flow according to one embodiment of this
invention. Crop producers 230 provide crop production information
232 to agricultural input suppliers 234. The crop production
information 232 may include, but is not limited to, seed product
information 236, environmental profile information 238, performance
information 240, and genotype by environment information 242. While
the crop production information may additionally include location
information, it is important to note that the crop production
information may be provided without the disclosure of location
information or producer identity. The crop production information
232 is stored in a data base 242. Utilizing the crop production
information 232 stored in the data base 242, an environmental
classification 244 can be associated with various agricultural
inputs 246. These agricultural inputs may include a land base,
regions of a land base, or a seed product.
[0109] FIG. 14 and FIG. 15 illustrate embodiments of user
interfaces to use in precision farming applications. In FIG. 14,
the user interface 170 includes site-specific information
associated with location information 172. The present invention
contemplates that other site-specific information or historical
information is accessible based on the location information 172 and
may be used in product selections. In addition, environment and
production information is collected. Examples of such information
includes maturity days 176, input traits 178, output traits 180,
seed treatment 182, whether no till practices 174 are used, the
planting population 184, nitrogen utilization 186, and drought
impact based on environmental classification drought frequency
information 187 and soil type. Based on this information and
information associated with the location 172, a recommendation 188
of at least one hybrid seed product is made. Where multiple
recommendations are made, the recommendations can be ranked as well
as a risk assessment 189-such as shown.
[0110] FIG. 15 illustrates another embodiment of a user interface
200 that can be used in precision farming applications. Site
specific information is collected such as location 172, soil type
174, and number of acres 202. In addition, there is the option to
import precision farming data 204 as well as import environment of
frequency data 205. There are also the option to set production
practices, set environmental assumptions, set risk levels, and set
the maximum number of hybrids 212. Based on the inputs, a portfolio
is created that includes a plurality of products 214, an associated
number of acres 216 to plant for each product, a recommendation 217
of at least one hybrid seed product, a risk assessment 218, and
revenue assurance 219. Where multiple recommendations are made, the
recommendations can be ranked. There is also an option to generate
precision farming information 220 based on this information, such
as a prescription map. The present invention contemplates that the
precision farming information may indicate which acres to plant
with which hybrids, give specific production practice application
(such as chemical application rates), or other recommendations.
[0111] FIG. 16 illustrates one example of a field-by-field analysis
showing product recommendations for a land base of a producer. As
shown in FIG. 16, different land areas within a producer's land
base have different hybrids associated with them. The present
invention contemplates producing such a map or field-by-field
recommendations where multiple products are recommended.
[0112] The present invention contemplates numerous variations from
the specific embodiments provided herein. These include variations
in the environmental classifications, performance characteristics,
software or hardware where used and other variations.
[0113] All publications, patents and patent applications mentioned
in the specification are indicative of the level of those skilled
in the art to which this invention pertains. All such publications,
patents and patent applications are incorporated by reference
herein for the purpose cited to the same extent as if each was
specifically and individually indicated to be incorporated by
reference herein.
* * * * *