Wheat Special Report No.

Management and Use of

International Trial Data
for Improving Breeding Efficiency
Ciudad Obregon, Sonora, Mexico
January 21-22, 1992

. . .... .. ...
~

.
. ......
.. .. . .
.'
.

Wheat Special Report No.8

Management and Use of .

International Trial Data
for Improving Breeding Efficiency
Ciudad Obrygon, Sonora, Mexico
January 21-22, 1992

P.N. Fox and G.P. Hettel

Editors

Correct Citation: Fox, P.N., and G.P. Hettel, eds. 1992. Management and Use of
International Trial Data for Improving Breeding Efficiency. Wheat Special Report No.8.
Mexico, D.F.: CIMMYT.

Contents
iii

Preface

iii

Acronyms/Abbreviations Used in this Report

iv

Workshop Participants

Initial Discussion

Shaping the Second Revolution--an Historical Perspective on International Trials

Statistical Analysis of International Yield Trial Data in ICARDA's

Cereal Program

AMMI Analysis of Yield Trials

13

Long-Term Similarity of Environments in the ISWYN

15

Using the Shifted Multiplicative Model (SHMM) to Identify Subsets of

Environments without Genotypic Rank Change: II. Clustering Method

23

Analysis of International Nursery Data--Results and Implications for International

Nursery Design

29

Update on the Data Managment System (DMS) for the Wheat Program

33

Use of Relational Data Bases in Wheat Breeding

37

Yield Trial System!lnternational Nursery Logistics

39

General Discussion/Conclusions/Actions to Be Taken

42

Appendix I--The Wheat Program's Data Management Sysem (DMS)

46
46
48

50

Appendix
A.
B.
C.

53

Appendix 3--1989 Position Paper on CIMMYT International Wheat Nurseries

62

Appendix 4--SuIVey of CIMMYT Staff on International Nurseries Reporting

66

66
70

Appendix 5--Cooperating Stations Involved with CIMMYf International

Nurseries
A. Country Abbreviations from the International Standards Organization
B. CIMMYf Wheat Program Cooperating Locations

84

Appendix 6--Genotypes Used in ISWYNs

2--Proposed New Versions of E, GE, and G Data

General Notes to Be Taken (E Data)
New Trait List (GE Data)
PMS Characteristics (G Data)

ii

Preface
The Germplasm Improvement Subprogram organized a workshop and think-tank on
management and use of international trial data for increasing breeding efficiency at
Ciudad Obregon, Mexico on January 21-22, 1992. Participants included scientists from
INIFAP (Mexican National Institute of Forestry, Agriculture, and Livestock Research),
Cornell University, the University of Queensland (Australia), India, ICARDA, and
CIMMYT.
The workshop explicitly recognized that data associated with germplasm are valuable
resources that enhance the worth of germplasm for both plant scientists and farmers of
the developing world. The Wheat Program's planned Data Management System
embraces this principle and will organize information around the germplasm to which it
refers, rather than by the source of the information (e.g., a specific trial).
Selected papers presented during the workshop and included in this Special Report
provide a point of departure or bench mark for our future breeding strategies and
planning as we strive to improve both CIMMYT and national wheat breeding programs.
Selected discussion notes are included after each paper. Appendices 1-4 provide
additional background and details about CIMMYT International Wheat Nurseries and the
Data Management System that may not be detailed in the selected papers. Appendices 5
and 6 provide listings of cooperating locations for all nurseries and ISWYN genotypes of
which we would appreciate assistance in making corrections.
This Wheat Special Report on the management and use of international trial data
emphasizes the importance of shared knowledge. But please keep in mind the
information in this report is disseminated with the understanding that it is not published
in the sense of a refereed journal.

S. Rajaram
Leader,
Germplasm Improvement Subprogram
Acronyms/Abbreviations Used in this Report
AMMI--Additive Main Effects and Multiplicative Interaction.
ANOVA--Analysis of variance.
CD-ROM--Compact disk with read only memory.
CIANO--Centro de Investigaciones Agricolas del Noroeste (Northwestern Agricultural
Research Center), Mexico.
CIMMYT--International Maize and Wheat Improvement Center.
CV --Coefficient of Variation.
DBMS--Database management system.
DD--Data dictionary.
DMS--Data Management System (proposed by CIMMYT).
EDYT--Elite Durum Yield Trial.
ESWYT--Elite Selection Wheat Yield Trial.
ICARDA--International Center for Agricultural Research in the Dry Areas.
INIFAP--Instituto Nacional de Investigaciones Forestales y Agropecularias (Mexican
National Institute of Forestry, Agriculture, and Livestock Research).
ISWYN--International Spring Wheat Yield Nursery.
ISO--International Standards Organization.

iii

Acronyms/Abbreviations Continued
ITYN--International Triticale Yield Nursery.
ITSN--International Triticale Screening Nursery.
LISA-Laboratory for Information Science in Agriculture (Colorado State University).
LSD--Least significant difference.
NARS--National agriculture research system.
PCA--Principal component analysis.
RAPID--Rapid Analysis Program for International Data.
RDB--Relational database.
SCS--Systems and Computing Services of CIMMYT.
SGYS--Small Grains Yield System.
SGSS--Small Grains Screening System.
SHMM--Shifted Multiplicative Model.
WANA--West Asia and North Africa.
WGB--Wheat Germplasm Bank.
WINS--Wheat International Nursery System.
WPMS--Pedigree Management System of CIMMYT.
Workshop Participants
Osman Abdalla, Head, Durum Wheat Section
Max Alcala, CIMMYT Wheat International Nurseries
Miguel Camacho e., INIFAP, Mexico
John Corbett, CIMMYT Geographer
Jose Crossa, CIMMYT Biometrician
Ian DeLacy, University of Queensland, Australia
R.A. Fischer, Director, CIMMYT Wheat Program
Paul Fox, CIMMYT Wheat International Nurseries
Guillermo Fuentes, CIMMYT Pathologist
Hugh Gauch, Cornell University, USA
Maarten van Ginkel, CIMMYT Bread Wheat Breeder
Arturo Hernandez J., INIFAP, Mexico
Gene P. Hettel, CIMMYT Information Services
K.B.L. Jain, IARI, India
Jesus Martinez S., INIFAP, Mexico
M.e. Jesus Naro S., INIFAP, Mexico
Wolfgang Pfeiffer, Head, Triticale Section, CIMMYT
Sanjaya Rajaram, Leader, Germplasm Improvement
Subprogram and Head, Bread Wheat Section, CIMMYT
Juan M. Ramirez D., INIFAP, Mexico
Rodrigo Rascon, CIMMYT Experiment Stations
P. Roger Rowe, CIMMYT Deputy Director General, Research
Mario Salazar G., INIFAP, Mexico
Hector Sanchez, CIMMYT Systems and Computing Services
Ken Sayre, CIMMYT Wheat Agronomist
Henrik Schou, CIMMYT Systems and Computing Services
Ravi P. Singh, CIMMYT Wheat GeneticistlPathologist
Bent Skovmand, Head, CIMMYT Wheat Germplasm Bank
S.K. Yau, ICARDA International Nurseries

iv

Initial Discussion
Rajaram: From a breeders' perspective, analysis of yield data is certainly required, but
how much is too much? Historically, ingredients of the CIMMYT bread wheat breeding
recipe have been:
Shuttle breeding.
Knowledge of parents, adaptability, disease resistance, quality, etc.
Stability of performance over time and space.
Combining ability.
Average Coefficient of Infection for rust diseases.
Intuition.
Simple experimental designs and quick analyses.

Fischer: What has been the influence of notes coming back from the international
screening nurseries and yield trials on breeding decisions across mega-environments?
Rajaram: The yield nurseries have been the best. The ISWYN, for example, has
educated us on where our lines stand. In recent years the ESWYT has been helpful, but
without the ISWYN, we would have no idea of what breeders would need in Zimbabwe,
for example.
Fischer: Have the international trials been more or less important over time?
Rajaram: Importance has remained about the same. Information related to yield has
always been very important. The major importance of the screening nurseries has been
the distribution of gennplasm to national programs.
Abdalla: How do you feel about the mega-environment concept?
Rajaram: It is a natural evolution. That is not to say that gennplasm will not cross ME
boundaries.

Shaping the Second Revolution--an Historical Perspective on International Trials

P.N. Fox
Head, International Wheat Nurseries
CIMMYT

Summary
A revolution in wheat breeding occurred with germplasm exchange. The second
revolution will exchange information related to germplasm, not only adding value to the
germplasm, but also strengthening bonds between institutions and between scientists. We
are on threshold of dramatic advances, limited only by our imaginations in using them.
Positive dynamic feedback between genetics and environmental information will provide
unprecedented insights into crop adaptation.

Introduction
Norman Borlaug spoke of a revolution in wheat breeding brought about by germplasm
exchange. I foreshadow a second revolution driven by exchange of information relating
to germplasm. The international nursery system will provide the cornerstone of this
revolution, which will increase breeding efficiency, with the major beneficiaries being
farmers in developing countries. Further expansion of the nursery network is unlikely,
but major advances in efficiency must occur.
The international nursery system is user-driven with CIMMYT managing, on behalf of
cooperators, the associated data that are freely available and will become increasingly
accessible with CD-ROM and other technologies.

Small Leaps Forward

Mechanization, computerization, and robotics
Although this workshop focuses on data exploration, it would be unsound to forget the
related areas of experimental design and mechanization of nursery preparation, including
robotics, in the overall international nursery scenario.
International feedback loop
International nursery data provide feedback to CIMMYT breeders. There is an element of
recurrent selection in the process of distributing elite germplasm from Mexico and recrossing or discarding this material in subsequent cycles, on the basis of international
performance. Anyone who has planned crosses in Obregon will attest to the fact that
overall yields in international testing and less formal data, such as notes from outreach
staff, are used extensively. One of the challenges before us is to make the feedback loop,
to CIMMYT and to others, more efficient. This will occur through:
Better data exploration.
Quicker turnaround of data.
Better access to raw data, results of analyses and interpretative summaries.

Data exploration and breeding

An extensive battery of techniques is available for data exploration and speakers will
detail some. Such techniques complement but in no way replace the intimate field
experience breeders have for their germplasm. I have a profound respect for breeders'
knowledge of their germplasm, including an uncanny ability to identify superior crosses.
In tandem to these skills, newer statistical methods are useful in identifying subtle

differences among sister lines and in rapid assessments, from wide testing, of adaptation
of new germplasm. Seri 82 appears the most broadly-adapted, high-yielding spring bread
wheat and international data exploration forewarns breeders of the Achilles' Heel of such
outstanding widely sown germplasm in a way that is impossible with national testing. For
example, Seri 82's leaf rust resistance depends on Lr23 and Lr26 and its superiority is
tending to decline in locations with heavy infection of BYD, septoria tritici blotch and
Helminthosporium species. This warning allows timely initiation of corrective
backcrossing and other measures.
Because different major stresses with time may be detected from the results of
international nurseries, as a result of many random processes in play, we must be ready to
use what surfaces. For example, identification of a cohesive group of locations subject to
heat stress in ESWYT-9 allowed inferences about superior germplasm for such
conditions, but different stresses emerged in the next year.

Time-lag problem
The ESWYT-lO bulletin provides some of the flavors to come in the area of data
exploration. These analyses, however, do not attack the time-lag problem, but a joint
ITYN-21 and -22 bulletin is planned for this year to bring the dissemination of the
second year's analysis forward, while, at the same time, providing a prototype for future
desk-top publishing initiatives for international nurseries.
Short-term versus long-term analyses
To date, the lead in newer analyses in the ESWYTs has given year-specific results, but
moving from short-term considerations to long-term ones is the theme of a major analysis
of ISWYN data, to be presented by another speaker. Long-term relationships among
locations will be important in refining the definitions of mega-environments.
Analysis or pathological data
Multivariate exploration of disease data has been largely ignored and correction of foliar
scores for height and maturity to better identify genetic sources of resistance has been
infrequent.
Adding value to seed through data
Where possible, CIMMYT should distribute existing data with nurseries to aid in the
selection efficiency of clients. For example, there is much information on bread making
quality generated before germplasm is distributed in an ESWYT.
Impact and singing for our supper
CIMMYT will increasingly have to sing for its supper. To this end, the next impact study
for germplasm will be facilitated by arranging our proposed databases to store the origins
of wheats (especially with respect to CIMMYT's contribution) according to the system
used in the recent impact study.
Great Leaps Forward
I have spoken of small improvements of an evolutionary nature in nursery analysis and
reporting. However, relational database technology heralds the great leap on which the
Wheat Pedigree Management System (WPMS) is based and on which the Data
Management System (OMS) will follow. I refer to other speakers and Appendix 1, which
outlines OMS and its query structure.

Removing barriers to association

WPMS overcame barriers to association by uniquely identifying wheat germplasm. The
line Siete Cerros was represented no less than 14 different ways in ISWYN bulletins and
one had to know a lot about wheat (and a little Spanish) to realize that all the associated
data related to the same wheat.

Crossing data frontiers

WPMS overcomes such ambiguities and redundancies and lays the foundation for DMS,
which will integrate information from different sources around the germplasm to which it
pertains. DMS will make possible many powerful associations between genetic
information and performance data. This interface was seldom crossed because of
problems in association of data from different sources. Detailed genetic information
generated in laboratory studies was seldom coupled to field performance data for several
reasons. One was a different scale of operations. Thousands of lines might be evaluated
in the field with the resultant data eventually finding their way to a dingy end on a series
of flat files on magnetic tapes. The results from the more intensive laboratory test on a
limited number of genotypes might be committed to paper. Even if the researchers
involved in the separate efforts were aware of each other's work, combining the data was
never considered because there would have been so many gaps in the intensive laboratory
information if combined with the field data in a flat file. Relational databases make cross
referencing these types of information feasible and efficient. Today, decisions on which
types data should be stored are less critical. As data storage becomes increasingly
cheaper relative to data generation, the question becomes: on which characteristics do
we require rapid querying?
The scope of DMS will ultimately embrace data from international trials, national trials,
the Wheat Germplasm Bank, quality and pathology laboratories, and research in
molecular biology, as well as hopefully interfacing to a recent Canadian initiative for a
directory of elite germplasm for studies of plant mineral nutrition.

Envi ron mental

Data

G Data
(genotypic adaptation, genetic
information, genes)

Figure 1. Environmental data and genotypic

performance data will provide dynamic positive
feedback.

Genes and the environment: dynamic feedback

Here is an example of how data integration might function. If we knew the genetics of
boron tolerance of only one or two entries in our database, but could extract international
performance data on them, implications about the distribution of boron related problems
could be made. Conversely, better knowledge of the environment facilitates typing
germplasm (Figure 1).
Integration of isolated rust research
Other powerful new associations may be possible by enhanced integration of rust data.
The existing international nursery data system accommodates Modified Cobb Scales and
for this reason has under-utilized a wealth of data generated on seedling reactions, many
of which relate to known races. McIntosh in Australia and Smit and co-workers in South
Africa regularly return such meticulously recorded information and they are generated by
CIMMYT's own pathologists. While imputing rust genes from field scores may be
possible by multivariate analysis of field scores alone, by extrapolation from responses of
known genotypes to responses of uncharacterized germplasm, integrating such trends
with seedling reactions will strengthen the process.
Discussion Notes
van Ginkel: Is it possible to identify 10 key locations so that we can get data back
quickly?
Fox: We should not be getting data back from key sites only. We are not bogged down
necessarily on the return of data, but on how CIMMYT processes it.
Rajaram: We can't blame delays on the cooperators. We take 2 years to process the
data.
Fox: I agree, that's what I'm saying. We are going to desktop publish some key nurseries
as a prototype and thus circumvent the old bulletin generation system.
van Ginkel: The best cooperators will send the best data early. What do we gain by
waiting?
Fox: The Southern Hemisphere is 11 months behind anyway.
Abdalla: Bulletins are not the only means of sending information to cooperators. There
are other ways, such as visitations, etc. Maybe we need to consider other means.
Fox: Send information, such as quality data, with the seed.
Yau: How can we go about getting better disease data? Cooperators use different scales
and even then can one trust the data?
Fox: This is not much of a problem anymore--at least for CIMMYT--since our Outreach
staff has been working hard to gain standardization across national programs.

Statistical Analysis of International Yield Trial Data in ICARDA's Cereal Program

S.K. Yau
International Center for Agricultural Research
in the Dry Areas (ICARDA)
This presentation provides: 1) a brief introduction about ICARDA's
international/regional cereal nurseries, especially yield trials, and 2) an explanation of the
approach and procedures used by ICARDA to extract information on entry response and
adaptability and site similarity.
When ICARDA was established in 1977, an international nursery system based on
CIMMYT's experience was already in place. Initially, there were segregating
populations, observation nurseries, and yield trials. In 1978, the crossing blocks were
added. Recently, trait-specific nurseries and gerrnplasm pools for disease resistance were
established. These six types of nurseries cover the requirements of most national
programs.
In 1978, there were only three yield trials: 1) barley, 2) bread wheat, and 3) bread and
durum wheats for rainfed areas. A yield trial for durum wheat was added in 1979.
ICARDA now has a joint mandate with CIMMYT for wheat improvement in West Asia
and North Africa (WANA). As experience has accumulated, it became apparent that the
WANA is a heterogeneous region. So for the benefits of the national programs, the yield
trials were gradually divided and targeted to one of the region's three mega-environments
(Le., high altitude, lowland with low rainfall, and lowland with moderate rainfall). In
1991, there were four barley, three durum wheat, and two bread wheat yield trials.
About two-thirds (more for wheat, less for barley) of the yield trials are sent upon request
to cooperators in the WANA. Asia and Mediterranean Europe receive some trials, while
the rest of the world receives a few. Following the recent agreement between CIMMYT
and ICARDA, the joint wheat yield trials will not be sent outside WANA in the coming
season. The total number of sets sent to national programs increased from 99 in 1978 to
572 in 1986. Since then the numbers have declined to 443 in 1991 as seed requests are
now better scrutinized to ensure that a nursery is appropriate for the area making the
request.
All of the yield trials were/are of the same design--lattice (RCB before 1990), three
replicates (four before 1988), six rows of 2.5 m spaced 30 em apart, and 24 entries
(including a local check, a long-term check, and two or three other checks). Many entries
and growing sites are different each year. There is no individual randomization for each
site.
Although the international nursery system was established in 1977, it was not until 1984
that a scientist was placed in charge of the operation full-time. The main emphasis of that
first scientist was to computerize the annual reporting and the development of computer
programs. When I acquired the position in the summer of 1986, I felt that my challenge
would be to provide better statistical analysis for each year's nurseries/trials and to
conduct in-depth analysis of the vast amount of data collected over the years.
Since 1987, the following analyses have been conducted and reported on in the annual
nursery reports: 1) nonparametric adaptability, 2) relative yield, 3) regression, and 4)
clustering.

.For nonparametric adaptability analysis, we use a visual display consisting of locations

(rows) and entries (columns). In 'each row, entries with a rank of five or less for grain
yield are highlighted. The total number of times the entry is among the top five yielders
is shown at the bottom of the display and provides an indication of the adaptability of the
entry. Since the locations are grouped by regions, one can also observe whether an entry
has specific adaptation to a particular region or not.
The exercise is repeated to show which entries are significantly higher yielding than the
long-term, improved and local checks at each location. The comparison with the local
check is of particular importance. It draws attention to those locations where the local
check performs very well or very poorly. Investigation of these locations is suggested.
The use of entry mean yield across locations with contrasting yields often gives bias to
entries that perform well in high-yielding location. However, using relative yield, i.e., site
mean coding by expressing entry yield relative to location mean yield give equal weight
to each location and converts the simple variance of entry yield across locations from a
biological to an agronomic type of stability measure (Yau 1991). So in the nursery
reports, mean relative yields and standard deviations (based on relative yields) of entries
are calculated for each region. High relative mean yields and low standard deviations
indicate good adaptability.
The use of nonparametric analysis and relative yield is simple and fast for indicating
adaptability of the test entries. It would be useful for routine analyses of multienvironment trials conducted by plant breeders.
Although the joint regression analysis is carried out for each trial, the usefulness of the
method for highly selected entries is doubtful. Analyses over four seasons of the
Regional Bread Wheat Yield Trial on rainfed locations showed that heterogeneity of
regression accounted for less than 3% of the interaction variance and was not significant,
but deviation from regression was highly significant.
For better understanding of the GxE interaction, a multi-variate technique is needed.
After reviewing different multi-variate procedures, I chose cluster analysis. Since 1987,
we have clustered trial entries and locations on mean entry yields for each location.
Yields are standardized within a location before cluster analysis is done. Centroid linkage
was used in the early years, but the group-average linkage is being used at the moment.
Examples of the usefulness of cluster analysis were presented on entry as well as location
grouping in Yau et al. (1989, 1991).
The Additive Main-Effect and Multiplicative Interaction (AMMI) model has been
suggested to be a powerful tool for analyzing GxE interaction. The postdictive model has
been tried on some data sets, but the results were not encouraging. The predictive mode
will be tried as soon as we obtain the program.
Besides the routine analyses carried out for each year's yield trials, greater use of the
international nurseries as a research tool for more penetrating analyses of GxE interaction
is being attempted. The possibilities to work collaboratively with CIMMYT, advanced
institutions and scientists working on agro-ecological characterization are being explored.

References Cited
Yau, S.K. 1991. Variance of relative yield as an agronomic type of stability measure. In
pages 297-306, Proceedings of the Eighth Meeting of the EUCARPIA Section on
Biometrics on Plant Breeding, July 1-6, 1991. Brno, Czechoslovakia.
Yau, S.K., G. Ortiz-Ferrara, and J.P. Srivastava. 1989. Cluster analysis of wheat lines
based on grain yield from different sites. Rachis 8(2):31-35.
Yau, S.K., G. Ortiz-Ferrara, and J.P. Srivastava. 1991. Classification of diverse bread
wheat growing environments based on differential yield responses Crop Sci. 31(3):571576.

Discussion Notes
Schou: What program do you use for analysis?
Yau: We have one program for analysis and another for management, but the software is
old and not adequate for the new analyses we want to do.
Fox: What is your rate of return of data from sets sent to cooperators?
Yau: We get an initial 60-70% return; we send out a letter to obtain an additional 10%.
van Ginkel: How do ICARDA breeders use the data?
Yau: They use the data to help confirm their own subjective ideas.
Fischer: At least one genotype (Nesser) coming out of the CIMMYT/ICARDA bread
wheat program does well in Syria. What does your data say about it?
Yau: The data are being analyzed
Fischer: How do you go about setting checks?
Yau: The long-term check never changes; the second check changes over time.
Rajaram: I think we need to compare relevant crosses simultaneously in Mexico and
Aleppo.

AMMI Analysis of Yield Trials

Hugh G. Gauch, Jr.
Cornell University
Ithaca, New York
In order for Additive Main Effects and Multiplicative Interaction (AMMI) to be
applicable to a given data set, two structural requirements must be met:
The data must be organized in a two-way table, with or without replication.
The data matrix must contain one kind of measurement and units, such as kg/ha.
The usual application in agriculture is a regional yield trial organized as a genotypes by
environments table, where the "environments" are actually site-year combinations. Basic
AMMI modeling is possible without replication, but rigorous assessments of predictive
accuracy (using data splitting and validation) and of postdictive accuracy (using F-tests)
require replication. An expectation maximization algorithm, EM-AMMI, can tolerate and
impute missing data. Also, the smallest allowable matrix size is 3x3, but genuinely
interesting results require a larger size of at least 5x5 or preferably 5xlO. It is not allowed
for different matrix rows (or columns) to contain different units, such as soil nutrient
concentrations, average monthly temperatures, and average annual rainfall. Also the
management must be quantitative, not qualitative (such as the flower colors white, pink,
and purple).
In order for AMMI to be truly useful, two additional conditions are required:
The data structure must conform, to a substantial degree, with the AMMI model.
When data exhibit significant main effects and significant interaction effects-which is the most common case encountered in yield trials--then AMMI is usually
effective.
Research purposes must call for parameters, estimates, error reduction, tables,
graphs, or insights of the sort provided by AMMI. This condition is almost always
met because AMMI serves a remarkably rich variety of purposes.
Violation of one more of the four conditions discussed above means that AMMI is not
appropriate and some other statistical model should be used. AMMI is not applicable to
data in a one-way factorial, nor in a three-way or higher factorial design. For example, a
set of genotypes tested in one site for 1 year constitutes a one-way factorial design.
However, if these genotypes have an underlying two-way design arising from, say, a
diallel cross, then AMMI is applicable. Likewise, a single trial could have an underlying
factorial structure in its environmental component, such as a single genotype tested in a
single site and year, but with fertilizers formulated in a two-way factorial design having
five levels of nitrogen and five levels of phosphorus. On the other hand, an experiment
most naturally regarded as a three-way or higher factorial can often be structured or
decomposed into one or more two-way subproblems. The most common instance is an
experiment most naturally regarded as a genotypes by sites by years three-way factorial
design, structured as a genotypes by environments two-way design, where environments
are site and year combinations. Especially when experimental sites change from year to
year, this can be a fruitful approach because the original three-way perspective would
generate an enormous number of missing cells.

For a data matrix with different units in each row (or column), the additive portion of
AMMI is meaningless, so other analyses are required. Popular candidates involve first
standardizing each matrix row to mean zero and variance one, and then applying
Principal Component Analysis (PCA), factor analysis, or some other clustering procedure
such as incremental sums of squares. More fundamentally, agricultural experiments often
generate data structures other than a single two-way matrix. Sometimes multiple
observations are made for each cell of a two-way design, such as a genotypes by
environments design with measurements of yield, height, and insect infestation. AMMI
can analyze each measurement separately, but exploration of their joint structure requires
other methods. Sometimes, agricultural experiments generate two or more two-way
matrices, most commonly a genotypes by environments matrix of yields and a factors by
environments matrix of environmental factors (such as soil nutrient concentrations,
temperatures, and rainfall). The joint structure between the two matrices must be
explored by other multivariate methods, such as canonical correlation analysis and
redundancy analysis. However, AMMI may be useful to summarize the matrix of
environmental data in terms of a few vectors, and likewise standardized PCA may
summarize the matrix of environmental data. Then simple methods like multiple
regression can explore relationships between these two sets of vectors.
The AMMI literature has emphasized the usefulness of the AMMI biplot for model
diagnosis (Bradu and Gabriel 1978; Technometrics 20:47-68). If only the additive effects
are significant or sizeable, then the analysis of variance (ANOVA) submodel is
diagnosed. Likewise, if only the interaction is significant, then the PCA submodel is
diagnosed. If the interaction has some special, simple structure--such as FinlayWilkinson linear regressions, or an even simpler joint regression (concurrence)-inspection of the biplot makes the diagnosis evident. If the main effects or the interaction
of both has some cluster structure, this will also be apparent. Hence, even when AMMI is
not the most appropriate model, nevertheless an initial AMMI analysis may offer the
easiest means for diagnosing the appropriate other model or submodel.
If the best research purpose is to classify the sites (or genotypes) into, say, five groups,
then it may be best to use a classification method instead of AMMI. Nevertheless,
particularly when the inherent variability of the material is continuous rather than
naturally clustered, an informal subdivision of the material based upon an AMMI biplot
may work about as well as anything.
Although invented in 1952, the AMMI analysis is relatively new to agriculturists. It will
take time to assess its usefulness for various kinds of yield trials and various research
purposes, particularly in comparison to other available analyses. Nevertheless, a few
comments seem in order.
Yield trials with relatively elite genotypes tested in relatively favorable (highly
productive) environments typically have most of the sum of squares in environments,
then interaction, and finally genotypes. More diverse material tested in more diverse
environments tends to have the interaction predominate. In either situation, both of the
main effects and also the interaction are all significant, so AMMI is usually effective.
Interaction and main effects present the breeder with very different problems and
decisions. Exactly for this reason, there is an implicit wisdom in the venerable tradition
of distinguishing and separating main effects from interaction. Interaction disturbs
genotype rankings from environment to environment; main effects do not. Hence, it is
helpful to have an AMMI biplot show main effects along one axis, and interactions along
another axis, thereby showing whether any two genotypes (or two environments) differ in
main effects, interaction effects, both, or neither. In other words, the usual decomposition

10

of the data into genotypes, environments, and genotype-environment interaction is not

only a nice statistical model, but also matches up with distinguishable agricultural
problems and opportunities. Of course, a slight modification of many other statistical
analyses could allow them to make this same distinction. For example, besides applying a
clustering alogrithm to the original yield data, one might also analyze the interactions.
Increasing selection gains and containing research costs both demand experimental
efficiency. AMMI often gives free extra replications of 1 to 3 times the actual physical
replications, whereas blocking typically gives 0.2 to 0.4, so this neglected opportunity to
partition the treatment design into a model and residual is often more effective than is the
usual analysis of the experimental design. However, since the treatment and experimental
designs are orthogonal, both strategies can be employed. These offer remarkably costeffective means for gaining accuracy, which in turn improves selection gain.
A deeper understanding of interactions can clarify the target mega-environments, reveal
the implications of interaction for breeding gains, and show the consequences (intended
or not) of past breeding policies for the present populations.
The entire subject of heritability, from definitions to analyses, is based upon simplistic
additive or linear models. Something new might emerge from revisiting this subject with
thinking that better handles interaction.
Without interaction, a single wheat or maize variety would flourish planet-wide. In many
ways, interaction is the problem that breeders face. In retrospect, it may seem that
breeders' greatest successes (apart from disease resistances) largely concern favorable
situations (mega-environment one) in which interaction is mostly sidestepped. But
further progress may have more to do with handling interaction head-on.

Summary
Agronomists and breeders frequently collect yield data for a number of genotypes in a
number of environments (site-years), resulting in a two-way data table--usually
replicated, but sometimes not. The AMMI model combines regular ANOVA for additive
main effects with peA for multiplicative structure within the interaction (i.e., within the
residual from ANOVA).
AMMI is effective for:
Understanding genotype x environment (GxE) interaction, particularly with a
biplot graph showing both main effects and interaction effects for both genotypes
and environments.
Improving the accuracy of yield estimates, offering an orthogonal and additional
error-control strategy to the usual strategy of replicating and blocking.
Increasing the probability of successfully selecting genotypes with the highest
yields, as quantified through order statistics calculations.
Imputing missing data with the expectation maximization algorithm, EMAMMI.
Increasing the flexibility and efficiency of experimental designs.

11

Ultimately, these advantages imply larger selection gains in breeding research and more
reliable recommendations in agronomic research. AMMI is ordinarily the statistical
method of choice when main effects and interaction are both important, which is the most
common case encountered in yield trial research.
The calculation for AMMI can be done conveniently by the MATMODEL program
(which resides in the CIMMYT VAX computer; see Dr. Jose Crossa for details). For
information on this program as well as citations of recent literature, see Gauch and
Furnas (1991, Agronomy Journal 83:916-920).

Discussion Notes
Fox: How do we divide resources between better design and better analysis?
Gauch: Generally, one should have fewer reps and more treatments. In an extreme
example, with CIMMYr's connections to a few large centers and many smaller sites, 510% of the trial could get you 80-90% of the information.
Pfeiffer: We gain more if we consider the environment and have the appropriate number
of reps accordingly...also change the size of the plot.
Gauch: It is true that we gain accuracy by enlarging plot size or increasing reps, however
AMMI is an extraordinarily cost-effective way to gain accuracy with fewer reps and
smaller plot sizes.
Fischer: A lot of our yield performance is based on information from one site in 1 year.
We are confident that performance at CIANO is representative, but do we know that it is
the best predictor of the overall ISWYN mean? In other words, is it the best site?
Gauch: You should be able to use AMMI to find out.
Fischer: Perhaps we should do yield tests at other sites as advanced yield trials, for
example, two reps at CIANO and two reps somewhere else in Mexico--would this mean
anything to the rest of the world?
Gauch: The key word is perhaps.
Fischer: Noise is more than error, in fact it can be biologically useful. For example, say
we have 60 sites, but one site has a unique boron problem--this would re-rank the
genotypes and AMMI would drop the atypical site out.

12

Long-Term Similarity of Environments in the ISWYN

Ian DeLacy
University of Queensland
Australia
Table 1 presents preliminary results of a manuscript in preparation for Euphytica. Here
the locations are divided into 15 relatively homogeneous groups, based on their longtenn genotypic performance. The final study will be conducted on standardized
deviations from location mean yield, such that each location-year combination has a
mean of zero and phenotypic standard deviation of one. Such a data manipulation ensures
that clustering depends largely on genotypic rank changes among environments.
Within each of the 15 groups, locations have been re-ordered in numerical order with
respect to their new proposed identifiers. The alternative would have been to leave all
location identifiers in their order from the complete dendrogram, i.e., the untruncated
dendrogram extended to the level at which each of the locations was a single member
group. Either way, advantages of the new location numbering system become apparent.
For example, the 23 locations in group 265 are all from Africa, as shown by the fact that
their identifiers commence with the digit one. Similarly, the six members of group 313
are from Europe (including the former USSR), indicated by the first digit being a three.
Group 339, however, comprises members from Africa, Asia (start with 2), Europe, the
North American Continent (4), Oceania (5) and South America (6).

Table 1. Membership at the 15 group level for the classification of the 178 locations
used three or more times in the first 26 ISWYNs based on deviations from location
mean yield of the entries grown in them.
Gr~
NlIIle

NUllber
in
group

95
265

23

308

313

316
325
330
333
334

336

21

337

24

338

36

339

23

340
341

6
3
8
4

"'-.berl of

314101
100021
122071
128091
194021
301041
204011
411051
601281
190041
212061
171021
241051
635081
202021
283011
610051
192011
222101
410051
560101
190031
333011
550081
207011
191071

gr~

100031
122091
140011
202031
312011
315051
421031
610011
191011
216021
190011
250021

100041
123011
147011
209011
316021
354041
601021
632011
191021
225041
191061
283021

212051
311011
610091
194011
241081
410071
601031
191031
375011
600011
216011
410021

220011
317011
612021
200021
270241
410081
610041
193031
400031
635011
226031

100071 100121 105021 111011 111031 122031 122061

123021 127011 127021 127031 127041 128021 128031
212011
351011
453011
601041

212041 610021 610031 614011

353011 374011
635031
630021 630051

192071 202011 203012 203041 612041

226011
220021 222071 222091 222141 224031 225011 226071
285011 354031 371021 400011 411031 601011 612031
222061
317021
631011
202051
317041
411021
612011
200031
401021

222081 226021 226041 226081 270061 281011

330011 332011 350011 354131 371011 560111
205021
334101
420021
615011
203011
410061

207021
354051
420101
635021
205011
410091

210051 211021 212021 222051

354091 400021 401041 410031
421011 421211 550011 550061
210021 211011 211031 316011
411011 413071 420042 550021

334051 354081 401011 410011 420041

13

All location numbers except one in Table 1 terminate with the digit "1", which represents
standard or default management conditions. The terminating digit "2" signifies that more
than one management system was used at the location indicated.
Readers are referred to Appendix 5, which details all locations which have participated in
the ISWYNs, organized according to the proposed system. From this appendix, the
specific identities (along with latitude, longitude and elevation) of the 178 locations in
Table 1 can be determined. Comments on the proposed system are solicited and readers
are especially urged to inform the first named editor of this working paper of any errors
detected. Please send the complete record as it appears and the complete corrected
record. Your assistance will be greatly appreciated.

Discussion Notes
DeLacy: The correlation between the yield of a set of adapted lines in a single yield trial
and the long-term performance in the farmer's field is essentially "zero"--no matter what
you do.
Singh: Should we do analysis on a regular basis or just every 10 years or so?
DeLacy: Add new data as they come in to the historical set and see where they fit.
Fischer: There is something about the optimum environment that gives good
predictability (Obregon, Giza, Israel, etc.). We were really lucky that we were making all
of our selections in Obregon.

14

Using the Shifted Multiplicative Model (SHMM) to Identify Subsets of

Environments without Genotypic Rank Change:
II. Clustering Method
J. Crossa, CIMMYT; P.L. Cornelius, University of Kentucky;
M. Seyedsadr, Wayne State University; and P. Byrne, CIMMYT
Note: Based on paper presented at the 1991 ASA Meetings in Denver, Colorado.

Introduction
In plant breeding, genotype x environment interaction is critical only if it involves
significant crossover interaction (COl) (Baker 1990), i.e., significant reversal in
genotypic rank across environments. Gregorius and Namkoong (1986) defined
"separability of genotypic effects from environmental effects". Such separability implies
absence of significant genotypic rank-change. The shifted multiplicative model (SHMM),
developed by Seyedsadr and Cornelius (1991b) provides a powerful analytical tool for
studying COL The general relationships between the multiplicative terms of the SHMM
model and Gregorius and Namkoong's definition of separability are described by
Cornelius and Seyedsadr (1991). Conditions considered sufficient for the absence of
statistically significant COl were:
SHMM1 is an adequate model for fitting the data.
"Primary effects" of environments have the same sign.
Cornelius and Seyedsadr (1991) used an exploratory method for identifying subsets of
environments that possess separability of genotypic effects.
In most crop breeding trials, genotype x environment interaction is complex and
separability of genotypic effects may not occur. In this situation, it would be useful to
have a methodology for finding subsets of environments in which separability of
genotypic effects holds. The objectives of this study were to present a cluster
methodology for finding subsets of environments with separability of genotypic effects,
and to examine the success of the cluster method in reducing the number of significant
COl within each final group of environments.

Materials and Methods

Yield data (kglha) from an international maize variety trial conducted in 1987 were
analyzed. The trial had eight genotypes grown in 59 locations.

SHMM statistical analysis

The SHMM model with t multiplicative terms (SHMMt) is:
Yij = S + ~tk=l "'kaikYjk + eij (Seyedsadr and Cornelius 1991b)
where

Yij is the yield of the ith genotype in the jth environment,

S is the shift parameter,
Ak is the singular value for the axis k,

15

ai 1 and Yj 1 are the "primary effects" of the ith genotype and the jth environment,
ai2 and Yj2 are their "secondary effects", and eij is a random error.
When the SHMM1 values are plotted against primary effects of environments or
genotypes, the diagram shows a set of concurrent regression lines which intersect at yield
level y = 13. If SHMM1 is an adequate model and the phenotypic values are located either
all to the left or all to the right of the point of concurrence, then genotypic effects are
"separable" from environmental effects. Otherwise, there is rank-change on opposite
sides of the concurrence point and effects are "nonseparable". If SHMM1 is inadequate,
statistically significant secondary effects are an almost sure indicator of nonseparability
even if primary effects have the desired pattern.

Finding subsets of sites for which SHMMt fits adequately:

clustering method, distance measures, and dichotomous split procedure
Distance measure (d) between two sites is defined as:
d = residual sum of squares, RSS(SHMM 1) if the fitted
SHMM1 does not show genotypic crossover and otherwise
d= min[~gi=1(Yi1-Y.1)2, ~gi=1(Yi2 - Y.2)2].
The latter chooses the RSS of the better of the two possible SHMM1 solutions
constrained to not show rank-change, namely, either 13 = Y.1 or 13 = Y.2.
Once d is obtained, clustering is done by the complete linkage method. SHMM analyses
are then done on the last two cluster groups (end of the dendrogram) to determine if
SHMM 1 without genotypic crossover is an adequate model. If this fails for either cluster,
this cluster is dichotomously split into two smaller clusters suggested by the next lower
branch of the tree in its portion of the dendrogram. This process continues until
satisfactory clusters are identified.

Assessing the adequacy of SHMMt

Adequacy of SHMM1 for modelling a subset of data was assessed using the following
tests:
Approximate tests, F1 and FGHl, against a pooled error (Cornelius et al. 1992).
The Seyedsadr and Cornelius (1991a) (SC) test.
The Seyedsadr and Cornelius/Schott-Marasinghe (SCISM) test (Cornelius et al.
1992).
A test constructed by analogy to the test of Yochmowitz and Cornell (1978)
(YC) for Additive Main effects and Multiplicative Interaction (AMMI) models.

Statistical tests for assessing significance of COl

The following tests (Cornelius et al. 1992, Crossa et al. 1992) were used to assess COl
among and within subsets:
Test 1, which requires calulating differences between all possible pairs of
genotypes in all possible pairs of environments.
Test 2, which uses a t-test with a comparison-wise error rate (a) of 0.05.

16

 Test 3, which is a t-test with a defined interaction-wise Type I error rate, i.e.,
error rate per 2 x 2 subtable tested for COls.

Results and Discussion

Grouping sites with separable genotypic effects
SHMMI is not an adequate model for the entire data set because secondary and tertiary
effects were significant (Table 1), so we proceed to look for satisfactory subsets of sites.
SHMM1 is also an inadequate model in both of the two major groups of sites (A and B)
identified in the dendrogram (Figure 1), since secondary and tertiary effects are large and
significant in both of these groups (Table 2). The next split fonns groups A1, A2, B1, and
B2. SHHM1 is adequate for Al (nonsignificant secondary effect) but not for A2, which is
then subdivided into A3 and A4 (Figure 1). SHMM1 is adequate for group A3, but not
for group A4 where mean squares owing to secondary and tertiary effects (172 and 141,
respectively) are significant by the F1 test. Because site 33 joins the rest of the sites late
in the dendrogram, its exclusion from A4 to form group AS improves the fit of SHMM1
(mean square of secondary effects is 111 and not significant). Site 33 is left as an
ungrouped site. SHMM1 is satisfactory for fitting group B1 due to the large value of the
pooled error variance (301; mean square error of sites 6 and 17 were 263 and 339,
respectively). SHMM1 is satisfactory for group B2 (only primary effects are significant).
The cluster analysis has identified five final groups of sites (A1, A3, AS, Bl, and B2) for
which a SHMMI that does not display genotypic crossover (Figure 2) is an adequate
model.

Significant COl among and within groups

Table 3 gives all possible 2 x 2 interactions (I), crossover interactions (COl), and
significant COls obtained by Test 1, Test 2, and Test 3 for sites in different groups and
for sites in the same final group. The number of all possible interactions between any two
genotypes with any two environments is 47,908; 41% of them (19,713) had a COl
pattern, Le., with genotypic rank change (Table 3). Of the possible within-group
interactions, 34% (4,339) had a COl pattern, while 44% (15,374) of the possible amonggroup interactions had a COl pattern (Table 3).
The percentage of COl that was significant by Test 3 (the most liberal one) was 8% (352)
within groups and 16% (2,482) among groups. As percentages of all possible
interactions, these are 2.8% and 7.1 %, respectively. These results indicate the
effectiveness of the distance measure used in the cluster analysis in reducing the number
of significant COls\ with genotypic rank changes within the final groups of sites.

Conclusions
The cluster analysis based on the proposed distance measure appears to do an excellent
job of allowing the user to identify groups of environments in which genotypic rankchange interactions are negligible relative to those between groups. This procedure can
be effectively used for grouping environments with genotypic separability.

References
Azzalini, A., and D.R. Cox. 1984. Two new tests associated with the analysis of variance.
J. Royal Statist. Soc. 46:335-343.

17

Baker, R.J. 1990. Crossover genotype-environmental interaction. In: Genotype-ByEnvironment Interaction and Plant Breeding (Ed. Manjit S. Kang). Louisiana State
University Agricultural Center, Baton Rouge.
Cornelius, PL., and M. Seyedsadr. 1991. Using the Shifted Multiplicative model
(SHMM) to identify subsets of environments without genotypic rank change. I
Exploratory method. Agronomy Abstracts.
Cornelius, P.L., M. Seyedsadr, and J. Crossa. 1992. Using the shifted multiplicative
model to search for separability in crop cultivar trials. Theor. Appl. Genet. (in press).
Crossa, J., P.L. Cornelius, M. Seyedsaur, and P. Byrne. 1992. Multiplicative model
cluster analysis for grouping environments without genotypic rank change. Theor. App.
Genet. (in press).
Gregorius, H.R., and G. Namkoong. 1986. Joint analysis of genotypic and environmental
effects. Theor. Appl. Genet. 72:413-422.
Seyedsadr, M., and P.L. Cornelius. 1991a. Hypothesis testing for components of the
shifted multiplicative model for a nonadditive two-way table. University of Kentucky
Dept. of Statistics Tech. Rept. 315.
Seyedsadr, M., and P.L. Cornelius. 1991b. Shifted multiplicative models for nonadditive
two-way tables. Communications in Statistics (in press).
Yochmowitz, M.G., and R.G. Cornell. 1978. Stepwise tests for multiplicative
components of interaction. Technometrics 20:79-84.

Discussion Notes
Crossa: SHMM analysis can be used for any number of genotypes and sites.
Corbett: CIMMYT sends out nurseries on request so we have a large group of similar
environments. If subsetting environments, how do too many similar environments present
a bias? What is the viability of doing a cluster anaysis on a subset?
PfeitTer: The difference in genotypes can be very small.
Fischer: Should we run an AMMI first before doing a SHMM or is that redundant?
DeLacy: I think SHMM has potential--It is the crossover that matters.

18

----------------,

~l

Figure 1. Dendrogram resulting from cluster analysis of 59 sites.

Final groups are marked with an arrow.

8000
7000
y
I

6000

E
L

5000

I
N

K 400J
G

H
A

3000
2000

1000

~___r_--r--........-___.,r---....,...............,..-..,...._____r_-~--,--___r-_r_----.--.,.___,______,-....,._____';_JJ

o. 195

0.200

o 205
PRI~ARY

o. 2 10

EFFECT OF 51 TE

2 15

0.220

(SHM~I)

Figure 2. SHMMl fitted model to final group AI. Numbers

refer to individual genotypes.

19

0.225

Table 1. Analysis of variance (ANOVA) and SHMM analysis for

grain yield (kg/ha).

Source of variation

df

MS

ANOVA
Model's corrected total
Genotype (G)
Site (S)

471
7

58

406
1239

G X S

Pooled error

2194
2947 a
16324 a
162 b
111

SHMM analysis
Model's corrected total
Primary effects
Secondary effects
Tertiary effects
Remainder

471
98.20
82.23
70.45
220.10

2194
9862 1 2 3 4 5
196 4 5
160 5
170

a Significant (P<0.05) when tested against the interaction.

b Significant (P<O.OS) when tested against the pooled error.
1 Significant (P<0.05) by the YC test.

2 Significant (P<0.05) by the SCiSM test.

3 Significant (P<0.05) by the SC test.

4 Significant (P<O.05) by the FGHI test.

5 Significant (P<O.OS) by the Fl test.

20

Table 2. Mean squares of SHMM analyses for various site

groups.
Group A+

Group Al

Group A2

Component 1
8476 1 2 3 4 5 6385 123 4 5 7775 1 2 3 4 5
Component 2
156 4 5
133
202 4 5
5
Component 3
168
135
227 4 5
Remainder
122
147
118
Error
109
106
111
126
170
SHMM1 residuals 153
Group A3

Group A4

Group AS

Component 1
4620 1 2 3 4 5 5496 1 2 3 4 5 4628 1 2 3 4 5
Component 2
103
172 5
111
Component 3
170
141 5
116
Remainder
118
134
13
Error
121
102
98
147
115
SHMM1 residuals 115
Group Bl

Group B

Component 1
Component 2
Component 3
Remainder
Error
SHMM1 residuals

7761 1 2 3 4 5
255 1 2 3 4 5
170 5
133
115
179

878 3 4 5
312
301
312

Group B2

6189 1 2 3 4 5
112
121
129
91
122

1 Significant (P<0.05) by the YC test.

2 Significant (P<0.05) by the SCISM test.
3 Significant (P<0.05) by the SC test.
4 Significant (P<0.05) by the FGHl test.
5 Significant (P<0.05) by the FI test.

+ Site composition of the final groups:

Group A1 = 1,2,3,4,7,8,10,14,16,19,20,21,22,23,24,25,28,35
48,54,55,56,59;
Group A3 = 9,15,30,40;
Group A5 = 11,12,13,29,31,36,37,42,45,47,50,53,57,58;
Group B1 = 6, 17;
Group B2 = 5,18,26,27,32,34,38,39,41,43,44,46,49,51,52.

21

Table 3. All possible 2 x 2 interactions (I), crossover

interactions (COl), and significant crossover interactions
obtained by test 1, test 2, and test 3 for sites in
different final groups and for sites in the same final
group.

Allocation of sites

Test 1+

cor

Test 2

Test 3

Total

47980

19713

282

2834

Among final groups

35140

15374

276

2482

within final groups

12768

4339

352

Group Al

7084

2550

213

Group A3

168

50

Group AS

2548

677

71

Group B1

28

Group B2

2940

1053

66

+ Test 1 is the Azzalini and cox test; Test 2 is the simple

t-test; Test 3 is the t-test with the joint 0.05 significant
protection.

22

Analysis of International Nursery Data-Results and Implications for International Nursery Design
Wolfgang H. Pfeiffer
CIMMYT Wheat Program

Introduction
"... virtually all phenotypic effects are not related to the gene in any simple way. Rather
they result from a chain of physico-chemical reactions and interactions initiated by genes
but leading through complex chains of events, controlled or modified by other genes and
the external environment, to the final phenotype" (Allard 1960).
The nearly infinite number of these genotype x environment (GE) interactions constitutes
a major challenge for plant breeders; selection of high yielding germplasm with
consistent performance is based largely on phenotypic traits, without knowledge of the
causal biochemical, physiological and morphological relationships. Identification of elite
gerrnplasm is further complicated by the dependency of genotypic and environmental
effects, and scientist are "... plagued ... by the disturbing feeling that each genotype may
have its own characteristic response" (Sprague 1955).
Although fluctuations in performance caused by GE are "masking" genetic breeding
values and force breeders to evaluate germplasm over a range environments, the
capitalization on specific GE interactions and resulting response pattern in crop
improvement is a major source of genetic progress, particularly if the underlying
biological factors of GE can be identified. Consequently, the critical importance of GE in
crop enhancement has been emphasized. However, opinions have widely differed
regarding the methods to evaluate GE, the level of knowledge on biochemical and
physiological "pathways" and associated traits required to make "real" progress, and the
application of trait/marker oriented versus empirical approaches in breeding programs. A
wider acceptance of integrated crop improvement strategies that consider biological and
practical, empirical approaches as complementary, rather than alternative, reflects a new
trend in breeding and increased knowledge and availability of tools for both approaches.
Since CIMMYT's cereal breeding programs have gradually shifted towards integrated
approaches, the role of CIMMYT's International Nurseries in this altered scenario needs
to be revised.

Objectives
This paper discusses different types of international yield nursery analysis in the context
of applied breeding and the integration of different types of analysis in CIMMYT's
breeding concept and strategy.

GE-Interaction Concepts and Approaches

CIMMYT's breeding operations focus on the development of high-yielding, widelyadapted gerrnplasm with built-in spatial, temporal, and system-independent yield
stability. Wide adaptation is a key issue in breeding efforts, since this prerequisite
facilitates the access and transfer of genetic systems to local germplasm or a direct
release. In this context, adaptation can be defined as the ability of a variety to produce
consistent high yields over a range of environments (wide adaptation), or over a specific,
narrowly defined environment (specific adaptation). In CIMMYT's breeding hypothesis,
adaptation is considered as a quantitative character related with yield per se; yield

23

stability can be defined for spatial, temporal, and system dependent yield fluctuations of
adapted genotypes.
The most important prerequisite of adaptation on the genotype level is yield per se and its
components. Change in components related with yield per se will, in general, not alter
adaptation. Other components of adaptation such as tolerance or resistance to biotic (e.g.,
resistance to single/multiple pathogens and races of a pathogen) and abiotic stresses (e.g.,
tolerance to drought and thermal stress) affect environmental yield potential, but also
adaptation and yield stability. Incorporation or modification of such a buffering
mechanism can increase the realized environmental yield potential as well as yield
stability and the range of adaptation. Hence, increasing specific adaptation can increase
wide adaptation and vice versa. In contrast to photoperiod insensitivity, which
contributes to wide adaptation, photoperiod sensitivity contributes to both yield potential
and specific adaptation, but decreases the range of adaptation.
Hence, the information required by breeders involves both genotypes and environments:
the evaluation and quantification of genetic yield potential, adaptation, and yield stability
of genotypes, as well a characterization of similarities among test sites as defined megaenvironments. Once such mega-environments are defined in terms of GE, the underlying
biotic and abiotic determinants should be identified. With stresses recognized, traitoriented selection could reinforce breeding for both broad and specific adaptation.
Based on this concept, data from multilocational international yield nurseries have been
analyzed in the past to generate information for the various themes. However, the
nurseries were not specifically designed for research purposes and served as germplasm
distribution tools.

ISWYN Analysis--Results and Implications

Genotypes
The early, more common statistical approach entailed an evaluation of the "total"
environment, by analysis of variance of yield data, without identifying the specific
environmental factors which determine GE. This approach provided relevant results on
adaptive pattern, spatial, and temporal yield stability, genotypic response to increasing
levels of productivity, similarities among genotypes and was used to compare locally
developed and so-called high yielding germplasm (Pfeiffer 1984, Pfeiffer and Braun
1989). Further, the results permitted conclusions on underlying biological factors of
germplasm performance particularly if biotic stresses were involved. The conditionality
of the results, Le., their association with the particular set of genotypes and test sites
included in the nursery, caused a certain bias and difficulties in their interpretation.
Estimates for regression response, stability, repeatability of stability parameters,
heritability estimates, and components of variance in the stability analysis are likely to be
overestimated due to the extreme diversity of the genotypes and test sites included in the
ISWYN. The linear approach largely failed to differentiate among genetically closely
related germplasm, e.g., sister lines from the same cross. Problems of a more statistical
nature, such as very low portions of "explainable" GE, overestimation of deviations from
the regression due to similar or identical site means, etc. are discussed in detail by
Freeman (1973) and Pfeiffer (1984).
Test sites
Other objectives of the ISWYN analysis were:
The description and quantification of the selection ability (SA) of test sites--test
sites with high SA allow discrimination among genotypes are representative for a

24

given target area, and repeatable in different growing cycles in terms of their
environmental conditions.
The grouping of the test sites into agroecological zones.
The identification of key sites with high SA on a global level and within
agroecological zones--based on the assumption that selection at those locations
should maximize genetic gains.
Identification of the underlying biological factors of SA and agroecological
groupings.
The results and methods of this study have been published by Braun (1983) and Braun et
a1. (1992). The regression technique used to estimate genotypic response was used in a
reverse manner: the regression of the entry grain yields at a test site on the entry means
across test sites can be used to describe the discrimination among entry mean yields--the
higher the regression coefficient, the better the discrimination. The deviations from this
regression can be used to describe the accuracy and should be small. Other parameters
used to describe SA are the heritability of a test site, and the correlation between the
entry grain yields at a test site and the entry mean yields across test sites for all test sites
included or subgroups of test sites. All these parameters can be combined in one measure
(Vtz 1972, Wricke 1971). Since this analysis was based on the same linear model and
data as the analysis to estimate parameters for genotypes, the restrictions, problems, and
limitations remained the same. The estimates provided a fairly rough picture of SA,
clusters of test sites, and underlying biological factors, however, lack of orthogonality
and the small number of test sites within the different clusters allowed detailed analysis
only for a limited number of major groupings. In general, it was not possible to obtain
applicable results for groups of marginal abiotic stress environments with low grain
yields in absolute terms, low variability among entry grain yields, low heritabilities, and
high error variances. The generation of data with a certain accuracy at such environments
may have required a different experimental design and larger plot size; however, the
problem of similar site means would have remained. Another major problem encountered
was the grouping of locations itself. By using several parameters--correlations among test
sites, site mean yields, regression slopes of test sites, disease scores, and heritability,
variance and error estimates--a meaningful grouping of the test sites was possible. The
procedure and the interpretation of the results are too time consuming for general
application--e.g., correlation matrices for a single year can reach dimensions of 80 x 80,
with 3160 correlation coefficients to be grouped in an unknown number of clusters or
predefined number of clusters.

IWWPN Analysis--Results and implications

Based on the experience with the ISWYN data, multivariate analysis appeared to be the
more appropriate technique to analyze 17 years of IWWPN (International Winter Wheat
Performance Nursery) data (Peterson and Pfeiffer 1989). Multivariate analysis, such as
cluster and factor analysis, consider genotypic response as multidimensional, with each
dimension representing a test environment. Conversely, they can characterize similarities
among environments in terms of performance of genotypes grown. Thus, both genotypes
and environments can be described in terms of GE (Peterson and Pfeiffer 1989, Pfeiffer
and Fox 1991).
A parallel study using cluster analysis resulted in very similar groupings. Further, the
grouping of test sites into regional and subregional divisions was facilitated by the
relatively high orthogonality of test sites over years.

25

Discussion
CIMMYT's Wheat Program distributes different types of nurseries with different
objectives:

Germplasm distribution nurseries for early segregating (F2) and intermediate

products (F3, F4) without request to return data.
Unreplicated screening nurseries primarily for germplasm dissemination with a
secondary function as germplasm screening mechanism (e.g., IBWSN, IDSN,
ITSN).
Replicated yield nurseries primarily to quantify adaptive pattern of genotypes
with a secondary germplasm dissemination function (e.g., ISWYN, ESWYT,
IDYN,ITYN).
Specially designed, replicated investigative yield nurseries, such as the
International Drought Trial have been distributed occasionally on a irregular base.
Some of the suggestions that resulted from the ISWYN analysis have been implemented:
1) the shift from RCB to lattice design for international yield trials; 2) the use of
improved, new methods for analysis such as AMMI, cluster, and factor analysis, which
are easily interpretable by breeders and provide additional information on environmental
similarities and adaptive pattern of the genotypes; and 3) focus on mega-environments by
separate screening and yield nurseries for different agroecological zones with a set of
genotypes adapted to the environmental conditions of the mega-environment and
standards common to different nurseries.
However, the information potential of international nurseries has not been exploited. The
inclusion of differentials for particular abiotic (e.g., trace element deficiency or toxicity)
or biotic stress (e.g., nematodes) in screening nurseries with specific instructions for note
taking could be used to detect actual or potential production constraints. Once such
environmental constraints have been identified, they can be quantified in specifically
designed investigative nurseries with plot size, for example, adjusted to environmental
variability. The International Drought Trial may have yielded more information for
moisture stress situations than several years of ISWYN data: it was distributed to
relatively few selected cooperators who recorded additional information on genotypes
and environmental parameters; additional information was generated in line source
testing and breeders trials. Hence, this allows scientists at CIMMYT base to evaluate the
representativity of their selection and testing environments, modify testing environments
and methodologies accordingly, and target basic research. Further, expanded multilocational yield testing during intermediate and/or final stages of germplasm
development in Mexico maYI result in higher selection gains and substitute for part of the
international yield testing with higher overall resource efficiency. Past experience has
shown that international cooperators adopt CIMMYT's methodologies and procedures,
e.g., new experimental designs. Hence, investigative nurseries can play an important role
in human resource development in developing countries. Further, efforts from CIMMYT
to distribute germplasm together with data and information, e.g., on milling and baking
quality or leaf rust genes, will facilitate selection and increase efficiency in NARCs.

26

References
Allard, R.W. 1960. Principles of Plant Breeding. John Wiley and Sons, Chichester.
Braun, H.-J. 1983. Untersuchungen ueber die Selektionseignung von Orten fuer die
Zuechtung von Sommerweizen im Tropisch- Subtropischen Bereich. Dissertation
Universitaet Hohenheim, Federal Republic of Germany.
Braun, H.-J., W.H. Pfeiffer, and W.G. Pollmer. 1992. Environments for selecting widely
adapted spring wheats. Crop Sci. (in press).
Comstock, R.E., and R.H. Moll. 1963. Genotype-environment interactions. In pages 164196, Statistical Genetics and Plant Breeding, Nat. Acad. Sci.--Nat. Res. Council Pub!.
982.
Freeman, G.H. 1973. Statistical methods for the analysis of genotype-environment
interactions. Heredity 31:339-354.
Peterson, c.J. and W.H. Pfeiffer. 1989. International winter wheat evaluation:
Relationship among test sites based on cultivar performance. Crop Sci. 29:276-282.
Pfeiffer, W.H. 1984. Ertragsleistung, Ertragsstabilitaet und Adaptation von
Sommerweizen auf regionaler und globaler Ebene --Analyse einer Serie von
internationalen Sortenversuchen ueber 15 Jahre und 973 Umwelten. Dissertation
Universitaet Hohenheim, Federal Republic of Germany.
Pfeiffer, W.H., and H.-J. Braun. 1989. Yield stability in bread wheat. In pages 157-174,
J.R. Anderson and P.B. Hazell, eds., Variability in Grain Yields: Implications for
Agricultural Research and Policy in Developing Countries.
Pfeiffer, W.H., and P.N. Fox. 1991. Adaptation of triticale. In pages 54-59, Proceedings
of 2nd International Triticale Symposium, Passo Fundo, Rio Grande do SuI, Brazil, 1-5
Oct. 1990.
Sprague, G.F. 1966. Quantitative genetics in plant improvement. In pages 315-354, KJ.
Frey, ed., Plant Breeding--a symposium held at Iowa State University, Ames.
Utz, H.F. 1972. Die Zerlegung der Genotyp x Umwelt-Interaktionen. EDV in Medizin
und Biologie 3:52-59.
Wricke, G. 1971. Eine orthogonale Aufteilung der Interaktion fuer ein eingeschraenktes.
Model. Rundschr. Arb-Gem Bioun. DLG 1.
Discussion Notes

Rajaram: In the screening nurseries, maybe we should reduce the load on cooperators by
asking them to provide information only on lines that are useful to them.
Pfeiffer: Cooperators should take notes for all entries. They should report what lines they
selected as soon as possible...Perhaps we should supply data with seed.
Singh: Ifwe follow through on Rajaram's suggestion, we would not know why a
particular line was not selected--it would be interesting to know if a disease was
involved.

27

Gauch: When is GLY significant statistically?

Pfeiffer: Almost always.
DeLacy: There is no year effect when testing in many locations around the world. On
average, the worldwide year mean is the same each year. Similarly GY is usually not
significant under these testing conditions.
Pfeiffer: I agree.
Fischer: Is the correlation still there when looking at highly selected material--as
opposed to the early years when there was a lot of junk in the nurseries?

28

Update on The Data Management System (DMS)

for the Wheat Program
Henrik Schou and Hector Sanchez
Systems and Computing Services

Note: For additional background on the Wheat Program's Data Management System, See
Appendices 1 and 2.
This paper reports on the findings of a DMS survey conducted by SCS concerning
problems and deficiencies involving the management of evaluation data in the Wheat
Program. The aim of this study was to identify:
Current user procedures.
Deficiencies to overcome.
Goals of the new system.
Scope of the DMS project.
Acceptable scenarios for completing the project.
Suggestions for a possible project plan.

Current User Procedures

In this section, we list current user procedures, i.e., the daily operations performed by
Wheat Program staff. The aim of analyzing these procedures is not to get a complete
picture of the Wheat Program's way of working, but to get a feel for the work flow so
that appropriate decisions can be made relative to setting up a functional and userfriendly data management system.
Below is a list of the day-to-day operations done in various sections of the Wheat
Program. The codes mean the use of the following:
S = Special package.
WPFBS = The field book system.
M = No system.
WINS = The WINS system.
WGB = The wheat germplasm bank system.
WSGSS/WSGYS = WSGSS and WSGYS.
FIBOS = The FIBOS system.
EXTRACT =The extract program.
* = Operations to be considered by the DMS project.
1) Scientist designs a field book. FIBOS
2) Scientist requests a field book. The Field Book System generates the book containing
crosses and selection histories. WPFBS, FIBOS

* Scientist collects evaluation data. These data are recorded on a sheet. M

4) * Wheat technician captures evaluation data. The recorded evaluation data are

3)

converted to an ASCII format. M

29

5) * Scientist loads the converted evaluation data into LOTUS or MSTAT. A data
analysis is extracted. S
6) Secretary drafts an international nursery trial. A draft book with crosses and selection
histories is produced. M
7) A nursery sheet is drafted, which is then converted to an international nursery sheet
with the same information. M
8) Secretary sends an international trial to a cooperator. WINS
9) Secretary sends a list of available international trials to the CIMMYT cooperators.
WINS
10) Secretary receives request from cooperator for an international trial. WINS
11) A decision is made on how to distribute the nurseries. Secretary sends international
trials requested by cooperators through the mail. WINS
12) Secretary receives a filled-in nursery sheet from a cooperator. M
13) * Secretary checks nursery field book for errors, which if found, an error list is
mailed to the cooperator stating the error(s) and requesting the cooperator to correct the
error(s). M
14) * Production department filters the nursery data. Values are checked to see that they
all are within tolerated ranges; if they are not, a decision is made on how to correct the
error(s). WSGSS/WSGYS
15) * Data on returned nursery sheets are converted to ASCII format. A nursery file is
generated. M
16) * An international nursery bulletin is produced. The converted nursery sheet is used
as input to a program which extracts the International nursery bulletin. WSGSS/WSGYS

* The ASCII nursery file is transformed to special file format.

18) * A scientist analyzes international nursery data. The special file is loaded into a

17)

standard package and reports are extracted and then combined with WSGSS/WSGYS
output. EXTRACT
19) Head of Wheat Germplasm Bank enters passport data. Passport data about new
collections are entered into the Bank. WGB
20) Head of Wheat Germplasm Bank prints field books from the Bank. Accessions with
the desired properties are extracted from the Bank and used as entries in the field book.
WGB
21) Head of Wheat Germplasm Bank transfers regeneration data to the Bank. The Bank
accessions are updated with the regeneration information and addresses are allocated in
the Bank. WGB

30

Data Management System

In this section current deficiencies and new system goals for operations 3, 4, 5, 13, 14,
15, 16, 17, and 18 above are discussed.

Current deficiencies to overcome

Deficiencies of the current system expressed by members of the user community in the
survey are listed below:.
1) Evaluation data are not always stored.
2) It is not possible to retrieve information.
3) In order to facilitate research on wheat data, a lot of "donkey work" to organize
the data must be done.
4) New traits can not be analyzed.
5) It is tedious to make a LOTUS or MSTAT report because the data needed are
only available on tape.

New system goals

Operations 3, 4, 5, 13, 14, 15, 16, 17, and 18 are supposed to be affected by DMS and the
new system goals include the following:
1) All the wheat data should be accessible.
2) Evaluation data must be stored in a format that facilitates researching.
3) It should be possible to add traits to evaluation data.

Acceptable scenarios
A brief description is presented here of two possible systems that satisfy the new goals-VAX-PC integration and a VAX-PC network. It is important to realize that goal #1
implies a connection of the three Vax databases--the Pedigree Management System
(PMS), Wheat Germplasm Bank (WGB), and the Wheat Program Field Book System
(WPFBS)--and the evaluation database that has to be constructed as part of goal #2.
The VAX-PC integrated solution. The three systems--PMS, WGB, and WFPBS--that
reside on the VAX and would be interconnected. The Evaluation database would also be
VAX-based and connected to the others. Facilities to export data to the PCS for use in
MSTAT, Lotus, or other packages would be available. Data would be entered on the
VAX or on a PC and would reside in the common data repository on the VAX. Filtering
would take place on the VAX.
The VAX-PC network solution. The three systems residing on the VAX would be
interconnected, but the evaluation database would be PC-based and not physically
connected to the VAX databases. The link to the VAX-based systems would maintained
through cross and selection history.
Development, operational, and maintenance costs
For either scenario, it is estimated that the whole project will take 4 programmer/analyst
years. The Wheat Program will need to allocate resources for the acceptance test,

31

acceptance test generation, and conversion of evaluation data. For the network solution,
an additional network database package would have to be licensed and installed.
Operational costs are estimated to be less than those currently, but will include entry of
evaluation data from about 1500 books per year, hardware maintenance, and costs of
using standard software packages.
The network solution would have additional costs of keeping two separate databases
consistent and of operating a network database.
Maintenance costs will involve fine tuning the system, including correction of minor
errors.

Suggested Development Plan

SCS suggests that the project be broken down into units with each unit consisting of
analysis, design, and implementation. Each product has to be accepted by the Wheat
Program and upon delivery be able to generate acceptance tests and execute acceptance
test cases.

Conclusion
The scope of the DMS project is limited to evaluation data from field books and
international nurseries. A large part of the project will be the interconnection of PMS,
WPFBS, and WGB. The other part concerns the design of a new database containing the
evaluation data.
We suggest that a steering committee consisting of wheat scientists and SCS
analyst/programmers agree on the steps of the development plan. Ideally, the project
should be broken down into smaller units and as each unit is delivered, it should be
approved and accepted by the Wheat Program.

Discussion Notes
Sanchez: The PMS conversion is the bottleneck--the Bank system cannot be linked to
PMS until the PMS conversion is completed.
Fox: Certain tapes in SCS should not be put on the system--either too old or too
expensive to convert.
Fischer: There is not much need to convert historical data. Conversion of new data will
be an accomplishment.
DeLacy: Data for potential inclusion in the system should be treated on a case-by-case
basis.
van Ginkel: Is it possible to include data from previous years in the fieldbooks?
Schou: As long as the data are there, yes.
Fischer: Even if you have all the data at your fingertips, is it humanly possible to use it?
It boggles the mind!

32

Use of Relational Databases in Wheat Breeding

John D. Corbett
CIMMYT Geographer
"Data is a resource in much the same way that employees, products, natural resources,
finances, and other materials are resources" (Relational Databases: Concepts, Design, and
Administration, 1991, McGraw Hill)
Wheat breeders generate volumes of data for which the major limitation to use is access.
Information access and information exchange offer mechanisms to enhance the efficiency
of breeding efforts, if only through the better management of the data a breeder
generates. There are two parts to this paper: First, some general characteristics of
relational databases (RDB) and how such data management systems might directly affect
CIMMYT. Second, I will present a practical or prototype example of how a relational
database works (the interface) using data from the International Spring Wheat Yield
Nursery (ISWYN). This example will be summarized through a simple listing of the
example queries.

Introduction
For breeders, better efficiency can be found in expediting the access to annual trials and
in access to historical data as well as germplasm data from other sources (namely PMS
and the germplasm bank). One way in which CIMMYT contributes to its cooperators is
in the form of seed (nurseries). We need these data. We also need to make the feedback
to cooperators faster, more thorough, and more convenient and, thus, these data should be
managed in a relational database environment. One possible effect of more timely
feedback is the improvement of trial data returned from cooperators.
An untapped source of information is the academic community worldwide. If supplied
with our data, the contributions from these off-CIMMYT sources may prove valuable.
A relational database means access to your data and ease of query of these same data.

Goals of the Relational Data Base:

Access to "all" data--international trials, the germplasm bank, Pedigree
Management System, genetic information, environmental data (climate, soils,
meteorology), output from crop models, etc. Because there is no positional
dependency between the relations, requests do not have to reflect any preferred
structure and therefore can be nonprocedural.
Simplicity: The end user is presented with a simple data model. Requests are
formulated in terms of the information content and do not reflect any complexities
caused by system-oriented aspects. A relational data model is what the users see,
but it is not necessarily what will be implemented physically.
Application program independence. Data which are accessed by a relational
database can be extracted so that it is in a form suitable for many applications
(e.g., SAS and other statistical package input files, ASCII text, etc.).

33

 Consistency through use of same data by all researchers.

Reduction in application development, storage, and processing costs.
CIMMYT can be thought of as an information-driven enterprise. Access to and use of our
data is paramount to the success of our enterprise.
Potential users must know what data exists! Thus CIMMYT is actively seeking ways to
publish our data in such ways as to maximize utility. CD-ROMs (Read Only Access)
offer great potential.

Part 1: The Relational Database

RDBs link "entities": An entity is a uniquely identifiable thing such as a person, place,
genotype, or concept about which we desire to record information.
An entity is a logical object, whereas a table is a relational object that can be mapped by
a database into physical datasets stored on disk.
An entity can be thought of as a ROWand each entity has some unique identifier
(primary key) and attributes (columns, also called fields) and through the keys, some
relation with other data.
The attributes of the logical entities will be implemented in relational tables as
COLUMNS. Each column must be identified. One or more columns in each table will
consist of the unique identifier, or primary key. Relationships and the corresponding
foreign keys will also be implemented as columns in tables. Primary and foreign keys
should be the only columns that appear in more than one table because they represent the
only attributes that appear with more than one entity.
An attribute (also called a field and which appear as columns) is a data element.
A relation is a group of attributes.
Attributes are related through a key and describe some entity of interest.
The RDB begins with the translation of each entity into a relational table.
The RDB can be defined as a collection of interrelated data items that can be processed
by one or more application systems. The RDB permits common data to be integrated and
shared between corporate functional units and provides flexibility of data organization. It
facilitates the addition of data to an existing ROB without modification of existing
application programs. This data independence is achieved by removing the direct
association between the application program and physical storage of data.

Final comments
Our data become an asset that we can make available to NARSs and universities.
CIMMYT will have a distributed database system. Each group has direct control over its
data (more efficient data processing =increased data integrity). A network allows for the
sharing of such data and the relational database management system (DBMS) allows
access to the data.

34

Note that for calculated fields, an ROB will not store the calculated field (Lotus and other
spreadsheets do) as the numeric fields that created it are sufficient and are created at
retrieval time. Scripts can be created which store the algorithm and these are easily
accessed.
The data dictionary (DO) is the second most important tool in the ROB environment
(after the DBMS itself). The DD is used to record facts about objects or events in the
database environment in order to facilitate communication and provide a permanent
record. The DO is based on giving information about the database itself, its contents, and
its structure. The DO focuses on data related components: data elements or attributes;
data groups, rows, tables; data structure; databases. CIMMYT will need to put resources
into the development and design of the data dictionary.
In an ROB environment, the DO is a very important tool for systems design,
documentation, data management, and data security. It can be used as a repository to
collect the user requirements which are used as the primary input to the logical design
phase of database development.

A note on data entry. The relational database and the DBMS use forms to assist with the
entry of data. In this way, each bit of information entered into the database is checked for
a range of possible values. Thus, yields would be flagged if entered as 10000 kg/ha.

Part 2: An example of possible queries: ISWYN

1) Select all data (germplasm type, rep, yield, disease scores) for all trials in which
rainfall was less than 200 mm and yields were greater than 5000 kg/ha.
2) Select all data for cases where disease scores on stripe rust and head rust are high.
3) Select for all cases where Nand P fertilizer were used and yields were less than 3000
kg/ha.
4) Use the cross number for Bluebird (II23584) and find all occurrences of this cross and
when grown in an ISWYN nursery.
5) Calculate the average yield of the repetitions for a given site in each ISWYN where it
occurs.
6) Count the number of times a select germplasm is used.
7) Select stations used more than eight times and which occur in South America. Provide
the latitude, longitude, and elevation of these stations.

Conclusion
The development of a relational database system at CIMMYT will require effort to clean
historical data and to change the way in which we view the data we generate. In
Appendices 5 and 6, we ask for your input to this process. Appendix 5 lists cooperating
stations. Please update any errors you can identify. Appendix 6 provides the common
name, a unique identifier, the cross and selection 10 (when known) for all germplasm
that has been grown in the ISWYN trials. Please send any corrections to the first named
editor. Provide a complete copy of the data in error along with a complete note on your
corrections.

35

Discussion Notes
Corbett: Access to our own data should not be an issue...The Paradox package is very
user-friendly. We will be able to come up with a computer screen that looks like the
cooperators' forms. This will speed-up data input by secretaries and enable them to get
new data into the system as quickly as it is sent in by cooperators.
Fischer: That is the key. The breeders need all the new information (up to yesterday) at
crossing time--50% of the demand will be during this period.
Sanchez: What about security?
Corbett: No problem, Paradox allows the files to be easily locked.

36

Discussion on the Yield Trial System/

Inte~national Nursery Logistics
B. Skovmand
Head, Wheat Germplasm Bank
The present system for managing yield trials, ESCAM, is now outmoded and needs to be
replaced. However, development of a new yield trial system is not foreseen until the
development of the Data Management System (DMS) is completed. The yield trial
system is inflexible in the sense that it is Randomized Complete Block Design (RCB)
with number of entries and replications fixed. Further, the system cannot be integrated
with our Wheat Pedigree Management System (WPMS), with the result that breeding
programs cannot be totally converted.
A small change in program management of the yield trials can minimize the effect of not
being able to connect the ESCAM to WPMS. Two types of fieldbooks are being
produced in relation to yield trials:
The actual randomized yield trial.
A Parcella Chica (PC or small plot) book.
Traditionally, the yield trial books are produced first and the PC books generated from
this list, which means that we have to get a list from WPMS, transfer this to the old
fieldbook system (FlBGS) and in the end convert the PC list to WPMS. This is
cumbersome and time consuming. If the breeding programs were to generate the PC list
first and then use this list as input for producing the yield trial list, we could avoid the
conversion back to WPMS and ESCAM could be utilized for the time being without
causi ng extra work.

Discussion Notes
Rajaram: In bread wheat, there would be a problem to generate the PC list first.
Fox: Regarding new experimental designs, the breeders will have to push for these or
there will be no sense of urgency.
DeLacy: Beware of the fact that you may end up with a system that can't handle the new
designs that will be coming along in the next 5-6 years.
Rajaram: What is the cost of switching from RCB to alpha-lattice?
Fox: Major labor component is adding sticky labels to seed envelopes for the lattice.
Rajaram: Smaller plots mean at least 15% higher yields and their use may mean we'll
start losing information--especially if there are germination problems like this year.
Jain: In India, we prefer 5 meters because 2.5 meters requires us to hand plant.
Yau: ICARDA, after some discussion, has decided to keep rows at 2.5 meters.
DeLacy: In Australia, we custom fit seed amounts to what cooperators want. This results
in better data, however, trying to fit local designs may be a bit radical for CIMMYT.
Fox: We need a larger sample of cooperator input before we make any changes.
Fischer: Then we need to do a cooperator survey. A concern I have is: are cooperators
following our instructions to trim the edge rows?

37

Alcala: We sent 7.5 tons of seed to cooperators in 1991-92, averaging 12

grams/envelope...Seed washing for KB currently takes 3.5 months.
Abadalla: What is our current seed washing policy?
Fischer: We want to reduce KB teliospores to the lowest level possible. We have
extended the policy to Central Mexico. Until we find a location without teliospores in
Mexico or can prove that the teliospores do not present a risk, we will have to continue to
wash the seed. We are currently installing washing equipment on the International
Nurseries bodega roof at El Batan.

Rajaram: 40% of the seed weight is currently with the Yield Trials. For the total weight
to remain the same, we must reduce the number of entries and sets if we add any new
nurseries.

38

General Discussion/Conclusions/Actions to Be Taken

Discussion
Rajaram: There is some risk in selecting crossing material from the PYT and YTs ...Give
F2s to cooperators who request them, but reserve the right to regulate when necessary
what we give them.
Fischer: Is there anything to be gained by testing in two locations in the Yaqui Valley?
How well do the artificial environments fit the targeted ones?
Sayre: ENET is a potential data base for determining if more locations in Mexico are
needed.
DeLacy: Before making a change in such <Lsuccessful system, you need the data to show
that a significant improvement is likely.
Fox: As I understand it, the policy is that the SNs are our major germplasm distribution
tool; the yield nurseries (YNs) for research. The YNs are the only ones that distribute
non-CIMMYT material; could such material be distributed in the SNs?
van Ginkel: Screening nurseries (SNs) provide the major flow of information for the
crossing programs.
Rajaram: We are looking primarily at disease resistance of 250 lines; not all that
interested in yield in the SNs.
Delacy: I think that yield data in the SNs are useful and should be recorded even though
2-row unreplicated.
Singh: I also agree that yield data are useful in the SNs.
Fischer: Looks like there are a number of votes that yield in the SNs is useful.
Crossa: Suggest going to an alpha lattice design.
Fox: All five YNs will be going to this design soon.
Fischer: What are the additional costs of such a move?
Fox: We'll need some kind of machine to attach the sticky labels...The important thing is
that we must know now if we are to move to lattice for the next cycle as we prepare the
seed shipments
Fischer: Before we decide we need to involve the national programs in the decision.
We've gone long enough without any feedback from them. Perhaps delay the lattice
decision for a year during which time we do cooperator survey.
Fox: National programs are very passive--not all that many complain about any changes
we make. It is almost frightening how cooperators accept the way we do things and the
responsibility implied...However, I believe the lattice design will be an important
technology transfer mechanism for national programs to test their own varieties.
CIMMYT is often the only contact for such technology transfer. Most national programs
are not keeping up with developments in the literature.
Fischer: CIMMYT needs to develop some sort of user-friendly training manual for yield
testing.
Fox: Perhaps distribute data with seed if DMS turns out to be the efficient and rapid tool
it is purported to be.
Fischer: There is a possibility that too much data will distract our attention.
Fox: Is there a risk in going from three to two reps? We need to decide if we are going to
do it--it will cut the number of envelope by one-third.
DeLacy: More varieties and fewer reps is the way to go for selection. Newer designs
(lattice, AMMI) are more efficient and only require two reps. However, telling
cooperators you are going to two reps is not the same as convincing them.

39

Gauch: Two vs three reps regarding AMMI: take historical data using two reps and use
the third rep to validate.
Workshop Conclusions
Although we have had tremendous impact in the past with the old system, changes are
coming. Broad adaptation across several environments will not carry us too much further
into the future, especially with shrinking resources. We will be moving to specific
environments with more analysis.
Some major points:
Yield testing within Mexico--no consensus that there be any major changes at
the moment. Although we should consider using smaller plots, testing on beds to
save resources, and analyzing the ENET.
Screening nurseries--Consensus that the SNs are important for disease
resistance; no consensus on whether yield data are important. Need to look at
Australian infonnation on the subject. Need to streamline data collection and
speed up its turnaround.
Yield trials--consensus on the following:
1) Go from three to two reps--notwithstanding the problems in the national
programs.
2) Do not increase site numbers, perhaps even reduce number of sites due to
the quality of data from certain sites.
3) Upgrade the statistical analyses (to AMMI, etc.) being used.
Infonnation for NARSs--CIMMYT has important technology transfer and
training roles. Need to design guidelines for experimental design and yield
testing, e.g., include a chapter on biometrics in the Training Manual being put
together by Villareal and Rajaram and develop a training manual for yield testing.
Need to consult with NARSs on anticipated changes.
Data Management System--After seeing what can be done during the
demonstrations at this workshop, the breeders agree on its value. However, DMS
is only part of the challenge--how to extract the wisdom is the rest of the
challenge. There is a lot of infonnation on how to do it, but not so much on why
to do it.

Specific Actions to Be Taken

The following actions are planned in the near future:
In 1993-94, ISWYN, ESWYT, EDYT, IDYN, and ITYN will be distributed
internationally with two replicates instead of three. This is possible because of
gains from the alpha-lattice design. The savings may allow a modest increase in
numbers of entries in subsequent years when new protocols for seed washing
become routine and automated.
The CIMMYT Wheat Program and the Biometrics Unit will fund a consultancy
to address Spatial Analysis of field experiments and trial design.

40

 The scope of coverage of general nurseries, such as ISWYN, will not change,
but others will be more specifically targeted.
The first product of the Wheat DMS Project will focus on fast feedback to
CIMMYT breeders for making crossing decisions.

41

Appendix i--The Wheat Program's Data Management System (OMS)

P.N. Fox
International Wheat Nurseries
February 1992

"What is needed here [nursery reportsJ is the compilation of international trial data in a
desktop-computer-database file that can be provided instead of the hardcopy and allow
for query and data retrieval according to the user's whims."
A. Blum, Volcani Institute.
Introduction
Norman Borlaug once said that international testing "broke down a psychological barrier
which had tended to keep the efforts of each wheat researcher isolated. The new testing
led to an unexpected acceleration of wheat breeding around the world." This revolution
brought about by germplasm sharing will be paralleled by another revolution driven by
information exchange and lead by OMS.

Objectives of OMS
A. To provide a secure, flexible system for the storage of important data on wheat,
triticale and barley germplasm.
B. To provide an integrated set of user-friendly tools for entering, filtering,
transforming and accessing data.

Justification
Efficiency, Research, Serving NARSs and Policy Making.
A. The Wheat Program handles data each year from international trials as well as
breeders' trials in Mexico and from activities in laboratories, greenhouses and
the Germplasm Bank. Important genetic information is also generated outside of
CIMMYT operations. Because much of these data are not stored in an organized
fashion, integration of laboratory and field data, for example, is impossible. The
increased use of PCS, more decentralized research programs, and personnel
changes raise the risk of a fragmented database, part of which may become lost
or inaccessible in the future without a more systematic approach to data storage.
B.

Wheat Program scientists require a more accessible database. DMS would

furnish convenient read-only data access to program scientists, while
.
guaranteeing the integrity of the master database.

C.

International trial data coordinated by CIMMYT is under-utilized. Storing

agroclimatic and crop management information together with genetic
information and plot data will facilitate partitioning and aggregation of data in
new and potentially revealing ways.

O.

OMS is viewed as part of a larger system that will include flexible options for
trial design, statistical analysis and reporting, and graphic presentation. Together
these components will result in more timely and better quality information for
CIMMYT scientists and cooperators.

42

Some History
International nursery bulletins are produced using software written in FORTRAN in the
early 1970s, when CIMMYf wheat staff with the computing system division of the RCA
company in Mexico City and Dr. Abel Mexas from the Ford Foundation designed and
implemented RAPID (Rapid Analysis Program for International Data). Modifications and
additions were found to be extremely difficult. However, in 1978-79 LISA (Laboratory
for Infonnation Science in Agriculture) from Colorado State University reviewed the
system and made some limited modifications to accommodate new requirements of the
rapidly expanding testing program. Since that time, the software has been used without
significant changes. The resulting software systems (SGYS and SGSS) were never
designed as data management tools. Their function was solely generation of international
nursery bulletins and they only became a type of default data management system, when
the program EXTRACf was written by Neal Bredin to access data from SGYS/SGSS.
. However, data from trials conducted with advanced lines in Mexico, the stage before
lines enter international nurseries, are not included. EXTRACf is unwieldy and requires
intervention of SCS personnel.
A system for the "corporate data", maize and wheat, was assigned a high priority at a
special meeting of the MAC on 12 June 1989.
During 1990, "Data Management System: System Requirements," aimed at a common
relational database system for Maize and Wheat was produced.

It was considered by the Wheat Program that, at least initially, agronomic trials would
not be considered by DMS. We will prime DMS with international nurseries data, adding
data from other field, laboratory and Gennplasm Bank operations later (Appendix 2
covers the status of definition of traits, as well as genetic and environmental descriptors,
to be stored or accessed by DMS).
In 1991, a Memorandum of Understanding was signed by the CIMMYf Maize Program
and Michigan State University (MSU) for software development. The MSU option was
not chosen by the Wheat Program for a number of reasons:
SCS has delivered the Pedigree Management System (PMS) and has an intimate
appreciation of the Wheat Program. By contracting outside, we would lose this
considerable in-house expertise, placing all strategic Wheat systems in jeopardy.
PMS is the core of the Wheat Program's systems strategy.
Because of this expertise and the good working relations established between
Wheat and SCS, we believe we will get a better return on every dollar invested inhouse compared with outside.
Problems of communication over distance.
The Maize Program is seeking much more than a Data Management System.
The Wheat Program has many of these additional elements available through
PMS.
The Wheat and Maize Programs will remain in contact on the progress of their separate
systems. It may prove possible to share certain specific operational functions peripheral
to DMS, such as nursery shipment control, automatic generation of phytosanitary
certificates and spatial analysis of field trials.

43

Pedigree Management System (PMS)

This section draws from Neal Bredin's "PMS/NWFB Consultancy Report" of December
1990. PMS and associated functions have been developed in SYSTEM 1032 on a VAX
computer under the VMS operating system.
The benefits realized via PMS can be attributed to:
Unique identification of germplasm and
Accessible pedigree relationships.
Why is unique identification important? If one wishes to share information about
something, one has to identify it. Although this sounds simple, it is not always straight
forward. Consider how germplasm is identified at CIMMYT. A cross name and selection
history constitute primary identifiers. These identifiers are not always unique, the same
germplasm may be known by several different names (e.g., VEE#5,
KVZ/BUHO/IKAL/BB, CM33027-F-15M-500Y-OM, etc.). It is not obvious that one of
these names may be equivalent to another.
This 'failure to associate' is a barrier to effective information sharing. This barrier has
prevented the Wheat Program from developing more than a collection of isolated
software systems supporting specific tasks. SGYS/SGSS, Breeders Seed Inventory
System and the old Field book System (FIBOS) are examples of systems that are self
contained, serve a useful function and will never be able to effectively interchange the
information they manage.

Information has been managed by how and where it was collected rather than by
identifying the germplasm to which it is associated. To illustrate this consider how one
would request information about a specific li ne. The questioning procedure in a source
oriented system proceeds as follows:
Which data collection systems record the type of data in which I'm interested
(e.g., SGSS, SGYS)?
Within the confines of that system, which trials (information sources) are
potentially interesting?
Did the line in which I'm interested appear in any of those trials?
Were we to associate information directly to the germplasm to which it relates, the search
for information would proceed as follows:
In what trials (information sources) did the line occur?
Of these, which are potentially interesting?
In the first scenario, we were faced with the task of reducing the number of places to look
even before establishing that the line of interest is in one of those places. Consequently
many potential sources of information are overlooked.
In the second scenario, we started from a position of knowing what data were available;
we only had to choose those sources that were potentially interesting.

44

The data-germplasm association provides a more effective information system than a

source-data system. The key to developing a data-germplasm association is the ability to
uniquely identify the germplasm:
PMS implements a system to identify germplasm uniquely. The cross identifier (CID)
uniquely identifies each cross. The selection identifier (SID) uniquely identifies among
the progeny of a cross. Together the CID and SID uniquely identify any germplasm.
Already, the 'ability to associate' of PMS has proved a powerful and versatile tool in
unexpected areas. For example, the cytoplasmic diversity of CIMMYT bread wheats was
examined. From PMS we have the potential to develop a highly integrated data
management system, linked to the Wheat Germplasm Bank. DMS will enhance the
'ability to associate', providing the capacity to associate data on germplasm with
combinations of latitude, altitude, diseases, country, mega-environment, available
moisture and other environmental parameters.

Some examples of what DMS will and won't do

DMS could ask for:
A list of all sites where international bread wheat, durum wheat and triticale
trials have been grown in the same season with less than 350 rom of available
water.
A list of instances where variety X, to be released by a national program, occurs
in the database and a second sublist in which all instances of susceptible reactions
to five major diseases for the variety are reported.
DMS will not:
Produce balanced data sets from unbalanced ones, by estimating missing variety
by site combinations.
Provide yield data for all trials sprayed with ILOXAN.

Final Comment
CIMMYT itself must become an exemplary cooperator in the network and take the lead
in returning data.

45

Appendix 2--Proposed New Versions orE, GE, and G Data

P.N. Fox
International Wheat Nurseries
February 1992

A. General Notes to Be Taken (E Data)

We are asking for more than in the past. We must see the changes from the cooperators'
point of view and guard against undue complexity, which could prove counter-productive
in efforts to manage data more usefully.
It has been suggested that the proposed detail on water availability is out of proportion
with for instance, crop stand where we solicit "poor, fair or good". However, crop stand
is also considered as a per plot observation under GE data. Cooperators should also be
encouraged to score characters 16 to 21 (below) on a per plot basis when observations are
in the moderate to heavy range.

The question of temperature data was raised. Our thinking was that temperature, as
opposed to rainfall, is relatively stable across years. Nonetheless, extreme deviations
from the norm could be accommodated in 23, a general comment on weather. Screening
nurseries may require less ennvironmental data than replicated yield nurseries. There is
opportunity to revise these questions.
1.

COUNTRY

2.

STATE, PROVINCE OR DEPARTMENT

3.

LOCATION NAME

4.

STATION NAME

5.

LATITUDE
degrees-minutes-N/S

6.

LONGITUDE
degrees-minutes-EIW

7.

ALTITUDE ABOVE SEA LEVEL IN METERS

8.

USE OF FIELD IN YEAR PRECEDING TRIAL SOWING

crop (please specify)/natural or improved pasture/
weed-free fallow

9.

SOWING
date dd-mm-yy
germination delayed by dry seed bed?
yes/no

10.

DATE HARVESTED dd-mm-yy

46

11.

UNITS OF FERTILIZER APPLIED

kg ha- 1/other (please specify)
N-P205-K20-other (please specify)

12.

CROP STAND
poor/fair/good

13.

SCALES USED AND DATES ON WHICH AGRONOMIC AND DISEASE

NOTES WERE TAKEN including height and test weight
trait-scale used-dd-mm-yy
trait-scale used-dd-mm-yy
" 10 times

14.

YIELD SCALE USED

kg plot- 1/other (please specify)

15.

LOCAL CHECK
name-crop

16.

FOLIAR DISEASE DEVELOPMENT

none/slight/moderatelheavy

17.

ROOT DISEASE DEVELOPMENT

none/slight/moderatelheavy

18.

INSECf DAMAGE
none/sl ight/moderatelheavy

19.

WEED PROBLEM
none/slight/moderatelheavy
major species, if serious

20.

BIRD DAMAGE
none/slight/moderatelheavy

21.

HERBICIDE DAMAGE
none/slight/moderatelheavy

22.

OTHER COMMENTS AND OBSERVATIONS

23.

WEATHER
general comment

24.

WATER AVAILABLE TO CROP IN MILLIMETERS (mm)

A. estimate of stored plant-available moisture in
full root zone at sowing in mm
B. precipitation (ppn.) in mm in the 12 months
(mo) up to month of harvest
month
ppn. 11th mo before harvest
"10th
"
"9th
"
"8th
"
"7th
"
"6th
"

47

name

"
"
"
"

"

5th
4th
3rd
2nd
1st
month of harvest
total

II

II

II

11

II

II

"

"
"

"

"

C. irrigated?
yes/no If yes,
number pre-sowing irrigations
number post-sowing irrigations
estimate of total water applied by
irrigation (mm)
25.

SOIL
classification (if available)
surface texture
sandy/sandy loam/loam/clay loam/silty
clay/clay/other (please specify)
pH
unknown/<4/4-5 .5/5.5-7/>7/actual value
% organic matter
aluminum toxicity?
yes/no
if yes,
surface/subsoillboth/unknown
any root barrier?
yes/no/unknown
if yes, depth (cm)
if no, depth of root zone (cm)

26.

PLOT DIMENSIONS
Sown
No. of rows
Length of row (m)
Space btn rows (cm)
Area (m 2 ) =
(AxBxC/100)

Harvested
No. of rows
D
E
Length of row (m)
Space btn rows (cm) F
Area (m2 ) =
(DxExF/100)

A
B
C

B. Trait List (GE Data)

We do not expect all traits to be recorded by all cooperators. Currently international
nursery fieldbooks list the following traits to be recorded: the three rusts, powdery
mildew, loose smut, net blotch, scald, BYD, days to heading, height, yield and "others".
The current software, Small Grain Yield System (SGYS) and Small Grain Screening
System (SGSS) can, however, manipulate 80 fixed traits. Should we remove this biasing
factor towards given traits in fieldbooks and stress that cooperators return data on the
traits important at their sites? * DENOTES KEY TRAITS.

PATHOLOGY TRAIT

CROP

SCALE

Stem rust
Leaf rust
Stripe rust on leaf

ALL
ALL
ALL

MCobb
MCobb
MCobb

48

Septoria tritici blotch

Septoria nodorum blotch
Helminthosporium sativum
Helminthosporium tritici-repentis
Fusarium nivale on leaf
Xanthomonas stripe
Powdery mildew
Scald
Net blotch
General foliar blights
Yellow rust in spike
Covered smut
Loose smut
Fusarium scab
H. sativum in spike
S. nodorum in spike
Barley stripe
Barley yellow dwarf
Barley stripe mosaic virus
Black point (H. sativum,
Alternaria, S. nodorum)
Kamal Bunt (grain infection)
Other bunts (grain infection)
Scab (grain infection)

BWDUTC
BWDUTC
ALL

BWDUTC
BA
ALL
BA
ALL

00-99
00-99
00-99
00-99
00-99
00-99
00-99
00-99
00-99
00-99
0-9
0-9
0-9
0-9
0-9
0-9
0-9
0-9
0-9
0-9

ALL
BWDUTC
BWDUTC

0-9
0-9
0-9

QUALITY TRAIT
ALVEOGRAM W VALUE*
ALVEOGRAM PIG OR P/L VALUE*
TEST WEIGHT*
FLOUR SEDIMENTATION INDEX*
YELLOW BERRY*
LOAF VOLUME*
GRAIN PROTEIN *
FLOUR PROTEIN *
FLOUR YIELD*
FALLING NO.
FLOUR PIGMENT
SEED HARDNESS
SEED TYPE
SEMOLINA PIGMENT
SEMOLINA PROTEIN
DOUGH MIXING TIME
COOKIE SPREAD FACfOR
GLUTEN CONTENT

CROP
BWTC
BWTC
ALL
ALL
DU
BWTC
ALL
ALL
BWTC
BWTC
DU
BWTC
ALL
DU
DU
BWTC
BWTC
TC

SCALE
0-999
0.0-99.9
kglhl
c.c.

AGRONOMIC TRAIT
GRAIN YIELD*
CROP STAND
1000 GRAIN WEIGHT*
GRAINS/SPIKE
SPIKES/M2
CHECKMARK*
DAYS TO HEADING (DC 55)*
DAYS TO ANTHESIS (DC 65)*

CROP
ALL
ALL
ALL
ALL
ALL
ALL
ALL
ALL

SCALE
kglha

ALL
?

ALL
BA
BA
ALL
ALL
BA
ALL
ALL
?

49

c.c.
%
%
%

sec.
p.p.m.
%

1-4
p.p.m.
%

min.
0-3
%

g
yes/no
days
days

DAYS TO MATURITY (DC 85)*

PLANT HEIGHT*
LODGING % AREA*
LODGING ANGLE TO VERTICAL
FORAGE YIELD (FRESH WEIGHT)
ACID SOIL TOLERANCE
HARVEST INDEX
SHATTERING %
NECK BREAK
ABOVE GROUND BIOMASS
COLD TOLERANCE
STAY GREEN
EARLY VIGOR
DAMAGE * (SPECIFY TYPE)

ALL
ALL
ALL
ALL
TC
ALL
ALL
ALL
BA
ALL
ALL
ALL
ALL
ALL

days
em
%
1-3
kg/ha
1-5

%
%
%
kg/ha
1-5
1-5
1-5
%

C. PMS Characteristics (G Data)

PMSTRAIT

CROP
ALL
ALL
BW
ALL
ALL
ALL
ALL
BWDUTC
BW
BWDUTC
BWDUTC
BWDUTC
ALL
BWDUTC
BWDUTC
BWDUTC
ALL
ALL
ALL
ALL
ALL
ALL

CID
SID
1R(1B) TRANSLOCATION
OTHER TRANSLOCATIONS & SUBSTNS
ALUMINIUM TOLERANCE GENES
BORON INSENSITIVITY GENES
MANGANESE INSENSITIVITY GENES
GRAIN COLOR
HMW SUBUNITS
LRGENES
SRGENES
YRGENES
SLOW RUSTING
Pdp GENES
Vrn GENES
RhtGENES
SPROUTING TOLERANCE
LEAF WAXINESS
COMMERCIAL VARIETY
YEAR OF RELEASE
COUNTRY OF ORIGIN
ORIGIN (as in impact study)

Pathology Notes
Key pathological traits have to be defined. Also lacking are transformations for the
Modified Cobb Scale and the 00-99 scale. These are required for averaging across
replicates and over sites. The 00-99 scale is described in Septoria disease manual (Eyal et
al. 1987, page 24). The first digit 0 to 9 is the percent height of the plant, up to where
infection has occurred (flag leaf height = 9), and the second digit 0 to 9 estimates the
percentage of the foliar area covered with the disease within that height. A new version
of "Instructions for the management and reporting of results for the CIMMYT wheat
program international nurseries"--including all Latin names for each disease--is required
for implementation of DMS.

50

0-9 Scale for spike diseases and/seedborne viruses

infecting whole plant.
0=
1% infected spikes/or plants
1=
1 to 10% infected spikes/or plants
11 to 20
2=
21 to 30
3=
4=
31 to 40
41 to 50
5=
6=
51 to 60
61 to 70
7=
8=
71 to 80
"
9=
81 to 100%
"
A second digit (0-9) is suggested for H. sativum, S. nodorum, and Fusarium scab to
estimate the severity of infection on the spike.
0-9 Scale for Grain Infections due to black point (due to H. sativum, Alternaria, S.
nodorum), Kamal Bunt, other bunts and scab
0=
No visible infected grain
1=
1 to 10% infected grains
2=
11 to 20
"
"
"
3=
21 to 30
"
4=
31 to 40
"
"
41 to 50
"
"
5=
6=
51 to 60
"
"
7=
61 to 70
"
"
8=
71 to 80% infected grains
9=
81 to 100% infected grains
The following diseases and insects should be included in comments or as % damage:
S. nodorum on node
none/slight/moderatelheavy
H. sativum on node
"
Root rot
"
Eye spot
"
Insect damage
"
Aphid incidence
"
Soilborne Wheat Mosaic Virus
"

AGRONOMY NOTES
Grain yield:
Days to heading:
Days to anthesis:
Days to maturity:
Plant height:
Lodging area:

corrected to 12% or other specified

moisture
days from sowing to when 50% of culms
have half of the spike visible (DC55)
days from sowing to when 50% of spikes
have flowered i.e. at least 1 anther
dehisced (DC65)
days from sowing to physiological
maturity when 50 percent of the
peduncles are yellow (approx. DC86)
measured from ground level to tip of
the average spike, excluding awns at
maturity
as percent of effective plot area i.e. %
the harvest area

51

Harvest index:
Shattering %:
Neck break:
Early vigor:
Damage:

measured at 0% moisture
percent per plot
percent per plot
scored at five leaf stage
a generic term covering birds, worms,
root rots, aphids, snails, rodents, suni
pest, hail, frost, grazing and any other
damage a cooperator might want to
include on a % damage per plot basis.

52

Appendix 3--1989 Position Paper on CIMMYT International Wheat Nurseries

P.N. Fox
International Wheat Nurseries
We at CIP learned fast that if we do not gather relevant data to set priorities
someone else will gladly assume that role.
--Robert E. Rhoades,
Potatoes and Our Fragile Earth
Agroecological Research at CIP
... the panel recommends that:
* CIMMYT overhauls its international testing program of nurseries to take better
advantage of new technology and trial designs and thus make more relevant the
experimental and analytical components of this important effort. Greater economy
and efficiency should be possible from the testing programme if it were
concentrated in selected sites in each mega-environment. ...
--Report of the third external program
review of CIMMYT, 1988.

Aim of this Paper

This document was written in 1989 to provoke, rather than to propose definite
solutions and to stimulate thinking on the future use of International Nurseries. Flaws in
the system are mentioned only to describe where we are, before addressing where we are
going.

Summary
Readers should consider the following points:
There is danger of throwing away the baby with the bath water in a "key site"
approach because the chance of accumulating genes for undiagnosed stresses is
minimized. A key site system must complement and link to the existing nursery
network. Any initiative involving key sites should strengthen the efficiency and
increase the relevance of our major research tool--the current network.
"Aberrant sites" may be more useful than "key sites", especially when
underlying biological explanations for abnormal behavior are revealed.
Large data sets risk including unreliable sites that are, at worst, random noise.
Our experience shows that inclusion of such "unreliable" sites hardly influences
across-site analyses of yield data. Discarding such sites, however, may result in
the loss of valuable information.
International nursery yield data are notoriously noisy, but methods, especially
the Additive Main Effects and Multiplicative Interactive (AMMI) analysis, are
currently being applied to this problem at CIMMYT.
International yield nursery data are under-rated as a research tool. However,
integration of results and feedback to cooperators must be improved. The quality

53

of the tool will be improved with a data management system and current
implementation of more efficient trial designs.
Any trial design improvements for yield data analysis should have positive spinoffs for disease as well. Analysis of disease data requires in-depth revision and
more staff input.
CIMMYT could become a center of excellence for analysis of large biological
data sets. Our comparative advantage is in the huge data bases collected and
awaiting analysis, combined with in-house biological experience for
interpretation.
CIMMYT must move rapidly towards giving cooperators the choice of
receiving field books on diskette or paper and must be ready to receive data from
cooperators on diskette.
International Nurseries must be linked to the Pedigree Management System
(PMS) to take full advantage of its management capabilities and to input valuable
information to the proposed and urgently required Data Management System
(OMS).
International nursery bulletins are still produced by software from the early
19708. Bulletin formats desperately need overhauling and the best option appears
to be building an interface between OMS and an off-the-shelf statistical package.
Implementation of the recommendations discussed in this position paper will
require a strong and explicit commitment from the Wheat Program and a dynamic
Systems and Computing Services (SCS).

Sites, Sacraments, and Serendipity

A science fiction film revolved round the theme of election of the u.S. President. Instead
of the unjustifiable expense of today's electoral process, the one person who most closely
represented the feeling of the entire U.S. population was identified to individually select
the President. Perhaps the theme has elements in common with the "key site" approach to
germplasm testing.
It is beyond dispute that CIMMYT's wheat breeding efforts would benefit from more
detailed characterization of environments. Judicious selection of sites for this process
could form a subset of II key sites" to improve understanding of genotype by environment
interaction. However, such a system must complement and link to the
ISWYN/ESWYT/lDYN/EDYT/lTYN network so that we know some points in the
environmental spectrum covered by international trials with greater accuracy than the
remainder of sites. This should facilitate extrapolation, with respect to limiting
environmental factors, from key sites to other testing sites. Any "key site" initiative
should strengthen the efficiency and increase the relevance of the existing International
Nurseries network and not be an independent thrust. Similarly, special research efforts,
like the three crop stress trial distributed in 1987, should link to the major network.
Unusual results (referred to below in the section on discarding sites from analyses) from
the ISWYN21 at Quito were interpretable mainly because an acid soil indicator set of
varieties was also sown. This provides an example of the general principle that detailed
physiological or environmental observations or special trials and regular international
nurseries, should complement each other.

54

Key sites should strengthen the current system's scope and not restrict it. Against a
global current of more sites and less replication it appears somewhat ironical that forces
seem to be pushing CIMMYT in the opposite direction. Efficiency aside, the potential
political repercussions of denying more cooperators yield nurseries cannot be
overlooked.
The word "key" has become somewhat emotionally charged in the English language. The
media also love this star player of the world of scientific grantsmanship. Once the label is
attached, "key" information is superior to other information.
Analysis of scientists' conceptions of what constitutes a key site generally indicates
environmental conditions which are stable over years. The factors responsible for
discrimination of germplasm are repeatable and thus more readily identified than in
highly unpredictable sites. Examples of key sites include:
CIANO (Mexico) and some locations in the Punjab (India and Pakistan) for yield
potential and leaf rust resistance,
Quito (Ecuador) for stripe rust, and
Passo Fundo (Brazil) for acid soils.
The information may be no more accurate than from any other site, but interpretation is
neater. However, the idea that key sites will provide the same information as extensive
wide testing is an illusion in which generally large genotype by site by year interactions
are not considered. These interactions may preclude finding key locations for such
elusive traits as drought tolerance.
One thrust for moving towards a key site approach may be the fact that we did relatively
little that was innovative or adventurous in reporting the huge volume of International
Nurseries results to cooperators. At the same time a shotgun distribution approach which
served CIMMYT breeders fairly well was not coherently and rationally defended.
Inexplicably, external pressure to reduce the scale of international testing by focusing on
"key sites" seems stronger than pressure to use available data more intelligently. Maybe
this pressure stems from scientific reductionism which is an approach also commonly
used by some nuclear physicists to tackle complex problems in their field. If it is difficult
to study complex interactions between subatomic particles, they often decided to take on
only one or two dimensions of the problem at a time. As Mitchell J. Feigenbaum said in
describing the flaws of such methodology: "There's a fundamental presumption in
physics that the way you understand the world is that you understand the stuff that you
think is truly fundamental. Then you presume that the other things that you don't
understand are details. The assumption is that there are a small number of principles that
you can discern by looking at things in their pure state--this is the true analytic notion-and then somehow you put these things together in more complicated ways when you
want to solve more dirty problems. If you can. In the end you have to change gears. You
have to reassemble how you conceive of the important things that are going on." Is
genetics any different? It is relatively straightforward to interpret varietal performance
when dominated by reactions to stripe rust, but not as the function of complex
interactions between genes for dwarfing, photoperiod, vernalization, undetected soilborne diseases, several foliar diseases and abiotic stresses which occur unpredictably at
different stages of the growing cycle.

55

The international testing system is perhaps conceived of, especially with pre-computer
thinking, as too large, too random and too difficult to manage. It is true that careful
scrutiny of data from five sites is more useful than cursory review of data from 50. So
why not just look at a small manageable "key site" data set? This attitude comes through
loud and strong in Abraham Blum's book Plant Breeding for Stress Environments. Blum
affirms that "while CIMMYT may perform some 90 wheat variety trials throughout the
world in 1 year, no one will take seriously a unified biometrical analysis of all these
tests." However, it must be acknowledged that at least "Heredity" took one such analysis
seriously, with the analysis of ISWYNlO data by Byth et al. becoming a landmark article
on multivariate characterization of locations and genotypes. While it is not clear what
constitutes a unified biometrical analysis, Blum does have a point about generalization
over an extremely large environmental range. However, Byth's pattern analysis approach
addresses exactly this point in attempting to subdivide the ecological spectrum into more
homogeneous subgroups. Pattern analysis of large data sets may be seen as a tool for
generating hypotheses which can be tested, perhaps at key sites using more conventional,
and possibly more rigorous, statistics.
A strict "key site" approach to germplasm testing may unwittingly discard screening sites
which expose germplasm to unrecognized stresses. Experience with tolerance to two
nematode species which attack wheat in Australia is pertinent in this context. Heterodera
avenae, the "eelworm" nematode, whose economic implications were almost ignored
until the 19705, has probably been in southern Australia since the last century. The
variety 'Festiguay' had become popular in some areas of South Australia through its
moderate resistance to stem rust during epidemics in 1973 and 1974. When fields
cropped to Festiguay in 1974 and 1975 were resown to wheat in 1978 after rotation,
spectacular cleaning effects were recognized through failure of the eelworm to reproduce
on Festiguay during the previous wheat cycle, reports South Australian breeder Tony
Rathjen. The fortuitous resistance of Festiguay to eelworm resulted in this variety
representing 50% of grain deliveries to some silos in the early 1980s, despite inherently
low yield. It is now believed that eelworm was responsible for average production losses
of about 15% in South Australia. CIMMYT lines 'Pitic 62' and 'Mexico 120' both have a
high degree of tolerance to this pest. Data presented as examples of intractable genotype
x environment interactions were subsequently explainable by tolerance introduced
unknowingly to Australian breeding programs in 1963. By 1978 the tolerance was
common in advanced lines and is now a feature of some varieties. In the state of
Queensland, testing sites which had been regarded as "aberrant" because of their
unexplained variability were found to be heavily infested with the nematode Pratylenchus
thomei and are now considered important screening sites for the pest. CIMMYT line
'Potam' is a source of tolerance.
An "aberrant site" approach could be the other side of the "key site" coin. Maybe the next
breakthroughs will come from "aberrant" rather than "key" sites. The very aberrance may
be due to important but still unrecognized factors. It should be underlined that this
phenomenon has already occurred for CIMMYT. Before aluminium tolerance was
recognized as an important constraint, Passo Fundo appeared to be a miserable location
characterIzed by high Coefficients of Variation (CVs) and low heritability. With
accumulation of higher levels of aluminium tolerance in the gene pool, it became a key
location for screening aluminium tolerance. E.J. Kahn Jr. in The Staffs of Life describes
how Portuguese Jesuits persevered with wheat in the highly unfavorable environment of
colonial Brazil to produce communion wafers. Today's legacy is a number of key sites
and levels of aluminium tolerance which are useful in many parts of the world.

56

Discarding Sites from Analyses

In general for large data sets, the risk of including unreliable sites which are at worst
random noise is less than the risk of discarding them. Our experience shows that
inclusion of "unreliable" sites barely influences across-site analyses. However, discarding
such sites may result in loss of valuable information, once an underlying biological
explanation for the abnormal behavior is found. ISWYN21 data from Quito, Ecuador,
were nearly excluded from reports, until the differential genotypic response could be
associated with soil acidity. CIMMYT Mexico was uneasy with the pattern of yields
exemplified by 'Thornbird' and 'Siete Cerros', 5.2 tlha and 37 kg/ha, respectively. At the
opposite extreme, New Zealand data have been questioned as being atypically high for
Australasia, even though New Zealand held world spring wheat yield records in the 50s.
The examples from New Zealand and Ecuador highlight pitfalls of discarding trials on
subjective bases. However, objective methods may cause more problems. The time
honored CV, the ratio of the square root of the error mean square to the trial mean, is a
poor criterion for discarding trial data. Environments with low yields and large error
terms are unduly penalized, although often providing valuable discrimination of
germplasm. Disease and other stresses may produce high CVs--and very useful
information. For plant breeding purposes, the per experiment heritability or the statistical
significance of the genotypic variation is more meaningful than the CV.
The following table provides an extreme example of how genetic parameters may be
biased by site selection.

No. of sites

ISWYN 17
ag2
7

54

33*

age 2

17

46
37

* excluding sites from nthn. USA and

nthn. Europe
Sites from northern USA and northern Europe may differentiate germplasm in a different
manner than sites in developing countries (for which CIMMYT is primarily working).
Being generally high yielding and with fewer environmental constraints to differentiate
germplasm, these northern sites reduced the apparent genetic variance. They increased
the genotype by environment interaction largely by the interaction between those sites as
a group ve~us the rest of the world as another group. While this explanation may be
debated, the effect of the subdivision on the genetic parameters is irrefutable and
alarming.
Trial Design and Analytical Methods for Yield
International yield data are notoriously noisy, but methods, especially AMMI analysis,
are currently being applied to this problem at CIMMYT. Other statistical approaches,
including Neighbor methods, may also be useful in reducing noise. Generalized Lattices
should result in more useful information through greater trial accuracy. For this reason,
ITYN21 for 1989/90 was distributed using a Generalized Lattice design and independent
randomizations at each location. However, the inflexibility of the existing nursery
reporting system will come home to roost with a vengeance when the output from lattice
analyses has to be spliced into a bulletin. Immediate decisions have to be taken on such

57

issues and the policy determined for 1990/91 international yield nurseries. Rowand
Column and other designs may be considered in the future.
The recent initiative of ICARDA to improve across-site analysis, using cluster analysis is
to be commended. While CIMMYf is considering similar options, we are largely
hamstrung by the inflexibility of the processing software for international data. Simple,
coherent explanations of newer analytical techniques are essential to make them useful to
cooperators. For example, a well-explained dendrogram of inter-site relationships should
allow some cooperators to begin to make their own inferences over time about their
environment with respect to other regions of the world.
To increase the utility of international nursery data, the question of long term checks
should be addressed.

Analysis of Disease Data

Processing and analysis of disease data is complex and trial design improvements should
have positive spin-offs for disease screening. As disease notes are to a degree subjective,
judgements on cooperators' scoring ability are inevitable and make the issue somewhat
delicate. Our ability to analyze disease data lags behind that for yield, but indications are
that more filtering and manipulation is needed. Leon van Beuningen's studies on
correction of foliar blight scores for height and maturity provide a useful starting point.
The Global Disease Monitoring Project will complement this work. The justification for
a "key site" approach is stronger for disease screening, where unreliable data are more
influential, than for yield.

Biotechnology and Upstream International Nurseries

If biotechnology plays a role in wheat breeding, it will probably have the biggest impact
in the highly reliable conditions which characterize "key sites". For instance, known
genes to match known environmental conditions could be monitored by RFLP
techniques. If this thesis is accepted, it follows that traditional statistically oriented
testing should concentrate more on marginal and unpredictable environments at the
expense of key sites, which become increasingly the gene jockeys' domain.
In passing upstream, CIMMYf could become a center of excellence for analysis of large
biological data sets. Our competitive advantage is in having huge data bases already
collected, combined with on-tap biological experience required for interpretation.
Computing hardware facilities would generally be sufficient if the current tempo of
upgrading were maintained. Maybe the time is now ripe to consider at least Postdoctoral
positions in analysis of international wheat data.
The relationship between observed phenotypic diversity for adaptation and its underlying
genetic base is largely unexplored, except for more simple examples where adaptation is
grossly affected by disease susceptibility. The scope of the CIMMYf testing programme
combined with modem cytogenetic and transformation techniques, for single gene
insertion, might provide powerful insights.

Germplasm Distribution Mechanism, Research Tool and Catalyst

Norman Borlaug once said that international testing "broke down a psychological barrier
which had tended to keep the efforts of each wheat researcher isolated. The new testing
led to an unexpected acceleration of wheat breeding around the world."

58

It is perceived that screening nurseries have been a much more potent agent for
gennplasm distribution than replicated yield nurseries, which have provided a generally
under-rated research tool. Some of the studies which have been conducted largely in
isolation are listed below. All but the penultimate were conducted on yield nursery data.

Refereed Scientific Journals:

Byth et ale (1976)
Fox et ale (1990)
Laing & Fischer (1976)
Menz (1980)

ISWYN4
ITYN
ISWYN
ISWYN

Heredity 37:215-230
Euphytica (in press)
Euphytica 26: 129-139
Field Crops Res. 3:33-41

Abdalla & Varughese

Braun
Byth et ale
Crossa et ale
Fox & Skovmand
Fox et ale
Pfeiffer
Pfeiffer & Braun

ITYN
ISWYN
ISWYN10
ESWYT8
ITYN
ISWYN/ITYN
ISWYN
ISWYN

Phung et ale
Skovmand et ale
Skovmand et ale
Worrall et ale

ISWYN7
ITYN
ITSN
ISWYN

Eucarpia 1987
Thesis Univ. of Hohenheim
Unpubl. report to CIMMYT
submitted to TAG
ASA Washington 1983
ASA Chicago 1985
Thesis Univ. of Hohenheim
In Variability in grain
yields (J.R. Anderson &
P.B.R. Hazell)
Unpubl. report to CIMMYT
6IWGS Kyoto 1983
Eucarpia Clermont-Ferr.1984
3rd Int. Wheat Conf.
Madrid, 1980

Other Presentations:

The ISWYN10 study, which suggested a dramatic narrowing in the genetic base since
ISWYN4, was not well received by CIMMYT. The authors themselves acknowledge that
they perhaps ignored the influence of discarding very poor germplasm between ISWYNs
4 and 10. All parties involved would have benefitted from a more open discussion of the
findings.
The research function of yield nurseries will be significantly improved with improved
trial design and analyses. However, priority setting will be required to affect such
changes and tradeoffs will have to be made. Use of CIMMYT-managed international
data by others should be actively encouraged. Slow response to requests for data amounts
to passive discouragement. Independent analyses can be safeguards against CIMMYT
biases. More importantly, the more heads working on international data the better, and
the greater the chance of detecting trends with implications for breeders.

Electronic Media
CIMMYT must move rapidly towards giving cooperators the choice of receiving field
books on disk or paper and must be ready to receive data from cooperators on disk. Our
more advanced cooperators, who have the greatest capacity to return results of more
sophisticated genetic and other analyses, are frustrated by CIMMYT's adherence to the
paper medium. The International Nurseries data base should be freely available on COROM to stimulate its use. Because of the size of the data base, other fonns of distribution
are unwieldy.

59

Glasnost in Reporting
A degree of "glasnost ll in international nursery reporting should be encouraged. For
example, germination problems associated with new seed dressings could be shared
openly with cooperators, as could discussion of different sowing rates used by CIMMYT
and ICARDA in international testing. Cooperators would also benefit from guidance
from CIMMYT for issues such as in which circumstances they might expect problems
from sticky dough associated with the 1B/1R translocation.

Bulletin Formats
International nursery bulletins are produced using software written in the early 1970s,
when CIMMYT wheat staff with the computing system division of the RCA company in
Mexico City and Dr. Abel Mexas from the Ford Foundation designed and implemented
RAPID (Rapid Analysis Program for International Data). Modifications and additions
were found to be extremely difficult. However, in 1978-79 LISA (Laboratory for
Information Science in Agriculture) from Colorado State University reviewed the system
and made some limited modifications to accommodate new requirements of the rapidly
expanding testing program. Since that time, the software has been used without
significant changes. The extremely rapid growth of the nursery network in the 70s
precluded detailed attention to the finer points of reporting. The immediate issues were to
pack and despatch increasing quantities of seed and to input incoming data, whose
growth paralleled the growth in seed tonnage. In 1988, the CIMMYT wheat program
shipped 2961 sets of seed.
CIMMYT now suffers some degree of credibility problem from the reporting format.
Close cooperators may accept shortcomings, but the wider scientific community will
become increasingly critical. It is time for improvements. For example:
Sometimes, CIMMYT sites do not report sowing and harvest dates as well as
other basic information. Immediate attention should be paid to this.
LSDs and CVs are produced for sites that return unreplicated data.
Many latitudes are incorrect e.g. 28N for Cordoba, Spain.
Correlations greater than one sometimes appear.
Many small details, in themselves not major, collectively detract from the
bulletins. For instance, 'BYD', not 'BYDV', should be used for field symptoms
and many place names are misspelled.
One wonders why there is so little reaction from cooperators to less trivial errors. Their
silence is worrying and begs the question "Are the reports read?"
The international botanical system for dates, day/month/year with the day and year in
arabic numerals and the month in roman numerals, should be considered to avoid
confusion between the commonly used date conventions of Latin America and Europe on
one hand and USA and Canada on the other.
Refinements in headings of individual location tables could be made. The almost
universal "cereal staff' in the tables is not useful. Abbreviations N, P20S and K for
fertilizer applications could reduce cluttering, as would a default convention for seasonal

60

conditions (whereby II normal conditions II would disappear from most tables, with only
atypical factors reported).

Ancillary Environmental Data

It would be useful to maintain a detailed data base for sites, bringing together many
valuable ancillary data, such as soil tests. This would not be published in bulletins, but
would be available for research.

The Next Step

The Pedigree Management System (PMS) will become the core of the Wheat Program's
computerized plant breeding and germplasm management systems in Mexico.
International Nurseries must also be linked to PMS to take full advantage of its
management capabilities and to input valuable infonnation to the proposed DMS-envisaged as an extension of PMS. PMS was designed basically to manage the bookkeeping in breeding, but has also shown its worth as a powerful and versatile research
tool in unexpected areas. For example, it was determined extremely rapidly using PMS
that the majority of CIMMYT bread wheats have one of three dominant cytoplasms.
Of possible strategies for overhauling nursery bulletin format, the best appears to be
through an interface from DMS to an off-the-shelf package such as SAS or GENSTAT
for bulletin generation. The question is, does CIMMYT have available expertise to
implement this option rapidly? The most optimistic estimate is that DMS will be at least
one and a half years in development.

It would be fruitful to consider how other international centers handle reporting and
anal ysis of data.

Acknowledgment
Helpful commentary from Hans Braun was appreciated in the preparation of this position
paper.

61

Appendix 4--Survey of CIMMYT Staff on International Nurseries Reporting

In 1989, twenty-one scientific staff--less than 50% of the Wheat Program--responded,
though of these, not all responded to all questions. Eleven replies came from outreach.
Ql: Is more, or less, information about the origin of entries required, for instance
the name and type of the institution producing each entry?
Four respondents felt that no more infonnation was required, while 13 suggested
additions, the most common being credit for the breeding institution. It was also
suggested that for large countries it is useful to know the region of origin.
Q2: Is the complete pedigree for all entries required?
YES, 15; NO, 2.
'
Q3: In ISWYN21, one site reported protein and another falling number. Would
more information on quality from the CIMMYT laboratory, for instance
sedimentation values from Obregon and Toluca samples, be useful?
Two thought that more information would not be useful and fourteen thought to the
contrary. However, it was suggested that depending on the heritability of the characters
involved, some parameters might be misleading for certain locations. Two people
suggested that the earlier practice of providing a general description of quality and
agronomic characteristics had been most useful. Greater effort to obtain and compile
more quality data generated by cooperators was advocated. Pigment content, protein
percentage and microsedimentation values for durums should be considered.
Q4: For whom should CIMMYT be producing nursery reports? Cooperators? The
wider scientific community? CIMMYT staff? Other? Why?
The majority felt we should be addressing principally cooperators and CIMMYT staff.
Six expressed the need also to cover the wider scientific community. Additionally, two
people replied that for targets beyond CIMMYT staff and cooperators, a different
reporting medium should be considered.
Q5: Do you find them useful?
Thirteen indicated finding the reports useful, while two suggested limited value for
reasons including generally high CVs, inadequate synthesis of data and failure to explain
abnormal yields.
Q6: Is it necessary to repeat the variety or cross in each table?
YES, 12; NO, 6.
Q7: Is it necessary to repeat the origin of each entry in each table?
YES, 3; NO, 15.
Q8: In your opinion which of the variables being returned can be considered as
reliable data and which are unreliable?
This question elicited a strong and useful response. There was general agreement that
much useful data is returned, qualified by the idea that direct measurements were more
reliable than readings, for instance for diseases. Within the disease field, several felt that
rust scores are usually reliable, but foliar blight scores are more doubtful, depending to a
large degree on the cooperators and the representativeness of locations. The need for case
by case evaluation was a common response, with a plea from one staff member for better
communication between International Nurseries and outreach. Two suggested that the
variability of traits should guide us with respect to their utility, although this would be
difficult for unreplicated data. Interestingly, no one stated explicitly that yield data were

62

unreliable, although the opinion was expressed that the relative variety yields were more
useful than the yields themselves. One reply doubted the utility of yield data from
screening nurseries. Another took the ISWYN21 variables and rated days to maturity,
stripe rust in the head, lodging %, scab %, falling no., protein %, tan spot and BYD as
unreliable. In retrospect, this may have been an easier way for everyone to respond.
Other responses supported this reply with respect to maturity and lodging. Careful
examination of disease data was recommended as was the recording of the date of
observation. One reply suggested that yield components were often badly taken.

Q9: Should we be more discriminating with respect to which data are included in
across-site summary statistics?
YES, 12; NO, 2.
If so, how?
This had been handled to a degree in the previous question. Of three suggesting generally
more screening of data, one explicitly advocated "pre-screening " in the appropriate
outreach offices. Two suggested a minimum level of infection for disease data. LSDs and
CVs for yield data were again mentioned twice as was the idea that summaries from a
very few sites are a waste. More training of cooperators was mentioned along with tact to
avoid offence to cooperators when discarding data.

QI0: Is the information presented at the top of tables for individual locations
insufficient, adequate or excessive?
The majority opinion (13) was adequate, but some of these were qualified by "adequate if
all information supplied correctly" and the question was asked why CIANO and Toluca
are generally the worst offenders in this respect. Three suggested that the information
was insufficient, in one case in relation to soil fertility and in the others with respect to
moisture supply. The necessity for a clear statement of both precipitation and irrigation,
preferably including number of irrigations, was highlighted. Otherwise, it appears that
Kafr EI Sheikh grew 6 tons of wheat on 52 mm moisture!

Qll: Could the above information (QI0) be presented in another part of the
bulletin?
YES, 7; YES, BUT INADVISABLE, 2; YES IF COMPRISES SUMMARY
ADDITIONAL TO RESULTS PRESENTED FOR EACH SITE, 1; NO, 6.

Q12: Should we solicit data on pH, soil depth and soil texture from cooperators?
Eleven replies were in favour of seeking this additional information, with two suggesting
that the data might aid in mega-environment definition. One answer suggested that we
only solicit such information from reliable cooperators, another suggested key sites only.
Adding which chemicals are applied as well as identifying problem weeds was
suggested. It was pointed out that the information we may seek is not always available.
Three other replies recommended requesting pH, while another favoured soil texture.
Three staff felt we should not solicit these data, with one suggesting the effort would
better directed to obtaining more important meteorological data.

Q13: Would more across-site statistical analyses be useful?

YES, 14; YES BUT ONLY FOR INTERNAL USE, 1; NO, 2.
If so, which kind? Why?
Five specifically mentioned cluster analysis, usually in the context of grouping like sites
and sometimes like genotypes. Others mentioned the need to group like sites without
suggesting a method. Some suggested multivariate methods in general. The need to
analyze across sites with similar stresses was mentioned along with the necessity for
some across-year analysis. Three indicated that stability analyses should be employed.
Some felt that across-region analyses would be useful, while others expressed

63

reservations about the present regional groupings and the use of clustering was advocated
to derive biologically meaningful site groups, for which different gene pools could be
developed. One suggested weighting sites by their error terms before clustering. Parallel
analyses across regions and also across mega-environments was suggested. Identification
of key sites was given as a goal of across-site analysis. One reply specifically drew
attention to the need for more across-site analysis for disease data.

Q14: Would more within-site statistical analyses be useful?

YES, 8; NO, 7.
If so, which kind? Why?
Suggestions embrace better qualitative and quantitative description of disease
development, broad sense heritabilities, a 7 x 7 lattice design for the ISWYN, more
attention to randomization to overcome interplot interference and presentation of yields
as percent of the top yield and percent of local check.
Should the local check and check varieties of other species be included in the
analyses?
YES, 3; NO, 2; SAME SPECIES YES & OTHERS NO, 2. One reply also indicated that
other species should not be included in the analyses of variance, but yield could be
expressed as a percentage of other species when they represent the top yield. One reply
questioned the practice of including more than one species in a trial.
Q15: Are the correlations presented for each location useful and valid?
YES, 4; NO, 3; DOUBTFUL, 5; IF DATA GOOD CORRELATIONS USEFUL, 2. One
of these last two replies included the comment that correlations should not be presented
automatically and especially not when there a large number of missing or zero values.
Q16: Are the regional summary tables useful?
YES, 12; NO, 2; BETTER DEFINITION REQUIRED, 3 (2 SUGGESTING MEGAENVIRONMENTS). Many replies referred to question 13.
Q17: Should we differentiate individual location tables, regional summaries and
overall summaries using different colored paper?
YES, 7; NO, 8; IF INEXPENSIVE, 4.
Q18: With increasing mechanization, will we, or should we, have to prepare seed for
cooperators who wish to sow non-standard plot lengths?
YES, 6; NO, 10. One reply suggested that we supply the amount of seed for each entry
that the cooperator requires along with empty sets of envelopes for sowing. Another had
serious doubts concerning edge effects and our standard yield plot.
Q19: Any other specific comments on entries, layout or analyses?
* Include more basic information such as quality characteristics, Rht genes and
resistance genes for entries.
* CVs are too high. Can we use Nearest Neighbor analyses?
* CIMMYT managed locations should be exemplary, not the least well documented.
* Large tables should be laser printed for better reproduction.
* Some disease scales need modification.
* Indicate which lines are not significantly different from the top yielder.
* Implement new designs such as Nearest Neighbor and lattices.
* Plot yields of each entry across locations.
* Make more use of ranking, as in ICARDA reports.
* Produce bulletins faster to increase their utility.

64

Q20: BROADER ISSUES RAISED FOR INPUT. We suggest that there is a

fundamental conflict between producing rapid superficial analyses for mass
circulation and answering more detailed questions relating to CIMMYT's breeding
and selection methodology.
Currently international nurseries fulfill two quite separate functions, namely
to distribute germplasm worldwide and to obtain yield, agronomic and disease data
for feedback to CIMMYT. The reliability of data and the question of whether the
two functions in one trial are compatible should be addressed.
There is a trade-otT to be examined between large global data sets with little
reliable environmental data and detailed yield, disease and physical data collection
at key locations where more sophisticated experimental design may be feasible.
It was suggested that:
* analysis of key locations and the broader data set is quite feasible. Better analysis
will serve both goals as well as widen the audience range.
* The ISWYN be discontinued and the IBWSN improved with a subset of entries
maintained for three years. Base staff to brainstorm the question and advise
outreach.
* More discussion is needed.
* ISWYN could be upgraded and distributed to "good" or "useful" locations, with
germplasm dissemination left to screening nurseries.
* Data returned is better than 10 years ago.
* Questions related to CIMMYT's methodology are not relevant for bulletins.
* International Nurseries staff should know "mega-environments", stations and
statistical procedures.
* Key locations can grow special as well as general trials. More sophisticated testing
should be conducted in addition to, not instead of, routine testing.
* Rapidly disseminated data can also be more exhaustively analyzed for
methodological purposes.
* Meteorological and disease data must be improved.
* Additional data from mega-environments using (specially?) selected germplasm is
required.
* ISWYN is useful for initial observations, but special material and areas selected for
CIMMYT long range breeding strategy are needed.
* Multivariate analyses of both physical and biological data are required to
characterize test sites and mega-environments.

65

Appendix S--Cooperating Stations Involved with CIMMYT International Nurseries

A. Country Abbreviations from the International Standards Organization (ISO)
and New Numbers, March 1992
AFRICA (AF)
Southern
S. AFRICA
AF-S
BOTSWANA
AF-S
COMOROS
AF-S
LESOTHO
AF-S
AF-S
MADAGASCAR
AF-S
MALAWI
MAURITIUS
AF-S
AF-S
MOZAMBIQUE
NAMIBIA
AF-S
SWAZILAND
AF-S
ZAMBIA
AF-S
ZIMBABWE
AF-S
East
BURUNDI
AF-E
DJIBOUTI
AF-E
ETHIOPIA
AF-E
KENYA
AF-E
RWANDA
AF-E
AF-E
SEYCHELLES
SOMALIA
AF-E
SUDAN
AF-E
TANZANIA
AF-E
UGANDA
AF-E
Central
ANGOLA
AF-C
CAMEROON
AF-C
CENTRAL AFRICAN REP.
AF-C
CHAD
AF-C
CONGO
AF-C
EQUATORIAL GUINEA
AF-C
AF-C
GABON
NIGERIA
AF-C
AF-C
ZAIRE
West
BENIN
AF-W
BURKINA FASO
AF-W
COTE D'IVOIRE
AF-W
AF-W
GAMBIA
AF-W
GHANA
AF-W
GUINEA
GUINEA BISSAU
AF-W
LIBERIA
AF-W
AF-W
MALI
MAURITANIA
AF-W
AF-W
NIGER
SENEGAL
AF-W

66

ZAF
BWA
COM
LSO
MDG
MWI
MUS
MOZ
NAM
SWZ
2MB
ZWE

100
101
102
103
104
105
106
107
108
109
110
111

BDI
DJI
ETH
KEN
RWA
SYC
SOM
SDN
TZA
UGA

120
121
122
123
124
125
126
127
128
129

AGO
CMR
CAF
TCD
COG
GNQ
GAB
NGA
ZAR

140
141
142
143
144
145
146
147
148

BEN
HVO
CIV
GMB
GHA
GIN
GNB
LBR
MLI
MRT
NER
SEN

160
161
162
163
164
165
166
167
168
169
170
171

AF-W
SIERRA LEONE
AF-W
TOGO
Western Islands
AF-WI
CAPE VERDE
AF-WI
REUNION
AF-WI
SAO TOME
AF-WI
ST. HELENA
North
AF-N
ALGERIA
AF-N
EGYPT
AF-N
LIBYA
AF-N
MOROCCO
AF-N
TUNISIA

SLE
TGO

172
173

CPV
REU
STP
SHN

174
175
176

DZA
EGY
LBY
MAR
TUN

190
191

AFGHANISTAN
BAHRAIN
CYPRUS
IRAN
IRAQ
JORDAN
KUWAIT
LEBANON
OMAN
QATAR
SAUDI ARABIA
SYRIA
TURKEY
UNITED ARAB EMIRATES
YEMEN ARAB REPUBLIC
YEMEN DEMOCRATIC
ISRAEL

AFG
BHR
CYP
IRN
IRQ
JOR
KWT
LBN
OMN
QAT
SAU
SYR
TUR
ARE
YEM
YMD
ISR

200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216

BANGLADESH
BHUTAN
INDIA
MALDIVES
MYANMAR (BURMA)
NEPAL
PAKISTAN
SRI LANKA

BGD
BTN
IND
MDV
BUR
NPL
PAK
LKA

220
221
222
223
224
225
226
227

PHILIPPINES
SINGAPORE
THAILAND
VIETNAM
BRUNEI
INDONESIA
KAMPUCHEA REPUBLIC
KIRIBATI
LAOS
MALAYSIA

PHL
SGP
THA

240
241
242
243
250
251
252
253
254
255

177

192
193
194

ASIA (AS)

West
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
AS-W
South
AS-S
AS-S
AS-S
AS-S
AS-S
AS-S
AS-S
AS-S
Southeast
AS-SE
AS-SE
AS-SE
AS-SE
AS-SE
AS-SE
AS-SE
AS-SE
AS-SE
AS-SE

VNM
BRN
IDN
KHM
KIR
LAO
MYS

67

AS-SE
AS-SE
AS-SE
North
AS-N
AS-N
AS-N
AS-N
AS-N
AS-N
AS-N

NAURU

NIUE
PAPUA NEW GUINEA
P.R. CHINA
HONG KONG
JAPAN
KOREA (N)
S. KOREA
MACAU
TAIWAN

NRU
NIU
PNG

256
257
258

CHN
HKG
PRK
KOR
MAC
TWN

270
280
281
282
283
284
285

SUN
SUN
ALB
BGR
CSK
DEU
HUN
POL
ROM
YUG

300
301
310
311
312
313
314
315
316
317

AUT
BEL
GBR
FRA
DDR
IRL
LUX
NLD
CHE

330
331
332
333
334
335
336
337
338

GRC
ITA

350
351
352
353
354

JPN

EUROPE & USSR (80)

East
EU-E
USSR (WESTERN PART)
EU-E
USSR (EASTERN PART)
EU-E
ALBANIA
EU-E
BULGARIA
EU-E
CZECHOSLOVA
EU-E
GERMANY
EU-E
HUNGARY
EU-E
POLAND
EU-E
ROMANIA
EU-E
YUGOSLAVIA
West
EU-W
AUSTRIA
EU-W
BELGIUM
EU-W
ENGLAND
EU-W
FRANCE
EU-W
GERMANY
EU-W
IRELAND
EU-W
LUXEMBOURG
EU-W
NETHERLANDS
EU-W
SWITZERLAND
South
EU-S
GREECE
EU-S
ITALY
EU-S
MALTA
EU-S
PORTUGAL
EU-S
SPAIN
North
EU-N
DENMARK
EU-N
FINLAND
EU-N
GREENLAND
EU-N
ICELAND
EU-N
NORWAY
EU-N
SWEDEN

MLT

PRT
ESP

NORTH AMERICA (NA)

North
NA-N
CANADA (WESTERN PART)
NA-N
CANADA (EASTERN PART)
NA-N
U.S.A. (WESTERN PART)

68

DNK
FIN
GRL
ISL
NOR
SWE

370

CAN
CAN
USA

400
401
410

371

372
373

374
375

U.S.A. (CENTRAL PART)

NA-N
NA-N
U.S.A. (EASTERN PART)
U.S.A. (ALASKA)
NA-N
NA-N
U.S.A. (HAWAII)
NA-N
MEXICO (NORTH + 23LAT)
MEXICO (SOUTH - 23LAT)
NA-N
central America
BELIZE
NA-CA
NA-CA
COSTA RICA
NA-CA
EL SALVADOR
NA-CA
GUATEMALA
NA-CA
HONDURAS
NA-CA
NICARAGUA
NA-CA
PANAMA
Caribbean (CARB)
NA-CARB
ANTIGUA
NA-CARB
BAHAMAS
NA-CARB
BARBADOS
BERMUDA
NA-CARB
CAYMAN ISLANDS
NA-CARB
NA-CARB
CUBA
NA-CARB
DOMINICA
DOMINICA REPUBLIC
NA-CARB
GRENADA
NA-CARB
NA-CARB
GUADELOUPE
NA-CARB
HAITI
NA-CARB
JAMAICA
MARTINIQUE
NA-CARB
NA-CARB
MONTSERRAT
NA-CARB
NETHERLANDS ANTILLES
NA-CARB
PUERTO RICO
NA-CARB
ST. CHRISTOPHER & NEVIS
NA-CARB
ST. LUCIA
ST. PIERRE & MIQUELON
NA-CARB
ST. VINCENT GRENADINE
NA-CARB
TRINIDAD & TOBAGO
NA-CARB
UK VIRGIN ISLANDS
NA-CARB
US VIRGIN ISLANDS
NA-CARB

USA
USA
USA
USA
MEX
MEX

411
412
413
414
420
421

BLZ
CRI
SLV
GTM
HND
NIC
PAN

450
451
452
453
454
455
456

ATG
BHS
BRB
BMU
CYN
CUB
DMA
DOM
GRD
GLP
HTI
JAM
MTQ
MSR
ANT
PRI

470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492

SPM
VCT
TTO
VGB
VIR

OCEANIA (OC)

Islands
OC-IS
OC-IS
OC-IS
OC-IS
OC-IS
OC-IS
OC-IS
OC-IS
OC-IS
OC-IS
OC-IS
OC-IS
OC-IS

AMERICAN SAMOA
COOK ISLANDS
FIJI
FRENCH POLYNESIA
GUAM
NEW CALEDON
NORFOLK ISLAND
SAMOA
SOLOMON ISLANDS
TAHITI
TOKELAU
TONGA
TUVALU

ASM
COK
FJI
PYF
GUM
NCL
NFK
WSM
SLB
TKL
TON
TUV

69

510
511
512
513
514
515
516
517
518
519
520
521
522

OC-IS
OC-IS
West
OC-W
OC-W

VANUATU
WALLIS & FATUNA ISLANDS

VUT
WLF

523
524

AUSTRALIA
NEW ZEALAND

AUS
NZL

550
560

BRA
BRA
ARG
CHL
FLK
PRY
URY

600
601
610
612
613
614
615

BOL
COL
ECU
GUF
GUY
PER
SUR
VEN

630
631
632
633
634
635
636
637

SOUTH AMERICA (SA)

Southern Cone (SC)
SA-SC
BRAZIL (NORTH PART)
SA-SC
BRAZIL (SOUTH PART)
SA-SC
ARGENTINA
SA-SC
CHILE
SA-SC
FALKLAND ISLANDS
SA-SC
PARAGUAY
SA-SC
URUGUAY
Andean Region (AR)
SA-AR
BOLIVIA
COLOMBIA
SA-AR
SA-AR
ECUADOR
SA-AR
FRENCH GUIANA
SA-AR
GUYANA
SA-AR
PERU
SA-AR
SURINAME
SA-AR
VENEZUELA

B. CIMMYT Wheat Program Cooperating Locations, March 1992

The list of cooperating locations beginning on page 71 uses country abbreviations
accepted by the ISO system. The list represents CIMMYT's first attempts to convert the
old location system file (COOP.LIS) to a new sortable system that refers to geographic
locations alone without considering management factors and other location descriptions.
Such information, however, will be linked to location codes in DMS (e.g., see Table 1 on
page 13).
The first digit indicates one of six major world zones. Together, the first three digits
uniquely indentify countries. The fourth and fifth digits then identify locations, defined
by latitude and longitude, within countries.
Duplication and errors remain. Please advise the first named editor of errors. Please
supply the complete record as it appears in the list and then how it should appear. We
would also greatly appreciate hearing of the multiple station numbers that refer to what
should be considered one single location.

70

Cnt

Hew
Abb. stat

state

City

Lat.

ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
ZAF
BWA
BWA
BWA
BWA
LSO
LSO
MDG
MWI
MWI
MWI
MOZ
MOZ
MOZ
MOZ
2MB
2MB
2MB
2MB
2MB
2MB
2MB
ZWE
ZWE
ZWE
ZWE
ZWE
ZWE
ZWE
ZWE
ZWE
BDI
BDI
ETH
ETH
ETH
ETH
ETH
ETH
ETH
ETH
ETH
KEN

TRANSVAAL ( 2 )
OFS (2)
TRANSVAAL (1)
CAPE PROVo ( 6)
CAPE PROVo ( 6)
CAPE PROVo (6 )
CAPE PROVo ( 5 )
CAPE PROVo (9 )
CAPE PROVo ( 4 )
OFS ( 3 )
NATAL
CAPE PROVo ( 1)
TRANSVAAL (3)
OFS (1)
TRASVDAI
CAPE PROVo ( 7 )
CAPE PROVo (3 )
CAPE PROVo (2 )
CAPE PROVo (10 )
CAPE PROVo ( 8 )
RUSTENBURG
GOOD HOPE

LICHTENBURG SENSAKO
BETHLEHEM
ROODEPLAAT
STELLENBOSCH - MARIENTAL
WELGEVALLEN
ELSENBURG
WITDAM
TYGERHOEK
LANGGEWENS
5ENSAKO
MAKATINI
CRADOCK
GROBLERSDAL
BETHLEHEM
BURGES HALL
OUTENIAGUA
STELLENBOSCH - SENSAKO
STET SENSAKO
NAPIER SENSAKO
STELLENBOSCH - 5TELLENBOSCH
TOBACCO RES. I .
BAROLONG
MOGOBANE

026
028
025
033
033
033
033
034
033
028
027
032
025
028
025
033
033
033
034
033
025
025
024
020
019
029
029
020
014
015
014
026
013
013
024
015
015
008
009
015
015
014
017
021
021
017
017
018
018
019
017
003
003
008
009
006
015
008
008
009
009
008
000

10001
10002
10003
10004
10006
10016
10005
10007
10008
10009
10010
10011
10012
10013
10014
10015
10017
10018
10019
10020
10021
10101
10102
10103
10104
10301
10302
10401
10501
10502
10503
10701
10702
10703
10704
11001
11002
11003
11005
11004
11006
11007
11101
11102
11105
11103
11104
11106
11107
11108
11109
12001
12002
12201
12202
12203
12204
12205
12206
12207
12208
12209
12301

MOSHU
NAGAMILAND WEST ESTMA
MASERU
THABA TSEKA
TANANARIVE
FIFA NANOR
CENT. REG.
CHITEDJE
NTCHEU DIS.
TSANGANO
CENT. REG.
BEMBEKE
MAPUTO
UMBELUZ1
NIASSA (1)
LICHINGA
NIASSA (1)
MATAMA
GAZA
GUIJA-CHOKWE
STHN. PROVo
MAZABUKA
LUSAKA
UNIVERSITY
NTHN. PROVo
MBALA
NTHN. KATITO
MBALA
CENTRAL
MOUNT MAKULU
CENTRAL
GOLDEN VALLEY
WSTN. PROVo
KALABO
HARARE (2)
GWEBI
HARARE-(4)
CHIREDZI (UNI OF ZWE)
CHIREDZI (UNI OF ZWE)
HARARE-(4)
HARARE (1)
CHISIPITE
HARARE (3)
HARARE
NYANGA
NYANGA
NYANGA
NYANGA
MUTARE
MUTARE
RATTRAY ARN
CHISIPITE
MURAMVYA
NYAKARARO
MURAMVYA
KISOZI
ARSI ( 1 )
KULUMSA
SHOA (6)
MELKA WERER
SHOA (1)
KOKATE
ASMARA
ERITREA
SHOA (2)
HOLETTA - A
SHOA (3)
DEBRE ZEIT
SHOA (5)
HOLETTA - B
ARSI (2 )
NEGHALLIE
SHOA (4)
AMBO (SHEWA)
RIFT VALLEY (3) MOLO

71

105
12S
35S
50S
52S
51S
45S
09S
30S
18S
25S
08S
255
lOS
27S
55S
20S
185
285
565
40S
28S
55S
03S
07S
18S
315
OOS
155
l1S
lOS
03S
18S
17S
31S
51S
25S
53S
08S
32S
OOS
57S
41S
01S
01S
40S
45S
17S
17S
OOS
45S
31S
33S
OON
16N
55N
16N
01N
55N
OON
OON
57N
15S

Long.

Elev.

026
028
028
018
018
018
018
019
018
028
032
025
029
028
030
022
021
018
019
018
027
025
025
023
022
027
028
047
033
034
034
032
035
035
033
027
028
031
031
028
028
022
030
031
031
031
031
032
032
032
030
029
029
039
038
037
038
040
038
038
038
038
035

1340
1687
1164
152
152
176
100
168
91
1631
74
660
950
1631
660
204
99
200
180
91
1234
1000
1000
1000
900
1510
2250
1500
1100
1600
1560
12
1356
1200
33
1049
1282
1668
1673
1213
1200
1051
1448
430
1468
1300
1506
1830
1830
1070
1300
2200
2090
2180
750
2100
2340
1700
1860
2390
1800
2225
2804

10E
12E
21E
50E
42E
50E
25E
54E
40E
10E
10E
38E
OOE
18E
50E
25E
08E
59E
54E
51E
14E
30E
45E
12E
18E
30E
36E
OOE
39E
37E
26E
32E
14E
20E
OOE
45E
19E
22E
20E
15E
30E
42E
52E
35E
35E
14E
05E
45E
45E
45E
15E
36E
41E
09E
42E
50E
52E
07E
58E
30E
30E
07W
44E

Cnt

Hew

state

City

Lat.

RIFT VALLEY (5 )
RIFT VALLEY ( 6)
RIFT VALLEY (7 )
RIFT VALLEY ( 8)
RIFT VALLEY (2 )
RIFT VALLEY (4)
RIFT VALLEY(10)
RIFT VALLEY (1)
RIFT VALLEY (9)
LR SHABELE
GEZlRA
HIDEIBA
KASSALA
KASSALA
KASSALA
SHAMBAT
MOSHI (3 )
MOSHI (2 )
IRINGA
MBEYA (1)
MBEYA (2)
ARUSHA (2)
MOSHI ( 1 )
ARUSHA (1)
ARUSHA (3)
ENTEBBE
HUAMBO
WAS SANDE
NORTH WEST

NJORO
ELDORET
KENYA BREWERIES
NGORE
ALJORO
N.P.B.S.
ENDEBERS
TlMAU
NAROK
AFGOY
WAD MEDANI
ED DA'EIM
KHASHM ELGIRBA
KHASHM ELGIRBA (NEW HALFA)
KHASHM ELGIRBA
KHARTOUM
KILlMANJARO
LYAMUNGA
TANWAT
UYOLE
U.A.C.
MONDULI TBL
W. KILlMANJARO
TARO AG.
TARO SELIAN
KAMPALA
CHIANGA
NGAOUNDERE
IRA-BAMBUI
FORT LAMY
KANO
ZARIA
MAl DUGURI

000
000
000
001
000
000
008
000
001
002
014
017
015
015
015
015
003
003
009
008
009
003
002
003
004
000
012
007
006
015
012
011
011
12
004
001
000
010
011
010
012
013
005
007
010
012
013
014
016
036
036
036
036
036
036
035
030
030
031
025
025
026
029

Abb. stat

KEN
KEN
KEN
KEN
KEN
KEN
KEN
KEN
KEN
SOM
SDN
SDN
SDN
SDN
SDN
SDN
TZA
TZA
TZA
TZA
TZA
TZA
TZA
TZA
TZA
UGA
AGO
CMR
CMR
TCD
NGA
NGA
NGA
NGA
ZAR
ZAR
ZAR
ZAR
ZAR
ZAR
HVO
HVO
GHA
GHA
GHA
MLI
NER
SEN
SEN
DZA
DZA
DZA
DZA
DZA
DZA
DZA
EGY
EGY
EGY
EGY
EGY
EGY
EGY

12302
12303
12304
12305
12306
12307
12308
12309
12310
12601
12701
12702
12703
12705
12706
12704
12801
12802
12803
12804
12805
12806
12807
12808
12809
12901
14001
14101
14102
14301
14701
14702
14703
14704
14801
14802
14803
14804
14805
14806
16101
16102
16401
16402
16403
16801
17001
17101
17102
19001
19002
19003
19004
19007
19005
19006
19101
19102
19103
19104
19109
19105
19106

KADAWA
SAMARU
LAKE CHAD
GAMBORU
KINSHASA
KIVU (2)
KIVU (1)
SHABA (2)
SHABA (3)
SHABA (1)
MOGTEDO
SOUROU
ACCRA
KUMASI
UPPER REG.
IMRAN KOLLO
DAKAR
DU FEEUVE
ALGER
ALGIERS
GUELMA
SETIF
SETIF
CONSTANTINE
SIDI-BEL-ABBES
EL GEMMEIZA
GIZA
SAKBA
EL MATANA
U.E. (KENA)
SHANDAWELL
SIDS

KINSHASA
LOHOTU
NDHlRA
KOLWEZI
KIPUSHI HAUT
FUNGURUME
MOGTEDO
SOUROU
ACCRA
KUMASI
TONO
BAMAKO
NIAMEY
DAKAR
VILLAGE GUEDE
QUED SHAR
BARAKI
GUELMA
SETIF
SETIF
EL KHROUB
SIDI-BEL-ABBES
EL GARBIA
CAIRO
KAFR EL SHEIKH
UPPER EGYPT (KENA)
EL MATANA
SONAG
BENI-SUEF

72

25S
30N
40S
OOS
05S
20S
OON
05N
OOS
06N
24N
35N
08N
08N
08N
37N
23S
14S
17S
55S
40S
30S
55S
l1S
30S
05N
44S
05N
OON
05N
OON
01N
55N
30N
30S
01S
16S
43S
44S
04S
30N
06N
35N
OON
51N
SON
18N
SON
30N
43N
05N
30N
UN
09N
16N
UN
48N
02N
07N
30N
30N
36N
04N

Long.

Elev.

036
035
035
035
036
035
034
035
035
045
033
033
035
035
035
032
037
037
034
033
034
036
037
036
035
032
015
014
010
012
008
007
013
14
015
029
029
025
027
026
001
003
000
001

OOE
20E
56E
30E
20E
56E
45E
25E
30E
11E
29E
27E
45E
45E
45E
32E
17E
53E
46E
22E
40E
30E
41E
51E
10E
30E
50E
03E
OOE
10E
36E
30E
37E
50E
17E
20E
10E
30E
14E
04E
OOW
25W
25W
40W

2165
2100
2790
1830
2200
2166
1920
2700
1830
100
411
353
400
400
440
375
1250
1280
1980
1800
2250
1846
1950
1372
1768

008
002
017
014
003
003
007
005
005
006
000
031
031
030
032
032
031
031

OOW
21E 210
32W
10
SOW
10
84E
24
10E
10
30E 263
25E 1033
21E 1033
42E 640
37W 450
07E
8
13E
21
57E
6
35E
89
35E
89
40E
57
06E
28

1695
1400
1982
1000
686
310
280
2100
2190
1380
1310
1165
270
250
282
180

Cnt

New

stat.

city

Lat.

BAHTIM
SERW .S.
MALLAWI
FAUME
FAUME
KOMOMBO
GIZA
BAHTIM
TRIPOLI
TRIPOLI
PENDUNTE
EL MARJ ARC
FEZ ZAN
SE. DESERT
KUFRA
GAREAN
AL-ZAHERA
MISURATA
AL-ZAHERA
ZORDA BENGHAZI
DEROUA MEN.
MENARA
RABAT (3)
KENITRA
RABAT (I)
RABAT (I)
SETTAT
RABAT (2)
ARIANA
TUNIS
BEJA
BOU SALEM
MATEUR
KHAROAN
BOULIFA
FIROZ KALU
SHISHAM BAGH
DARUL AMAN

KALOBIA
DONIAT
EL MINIA
TAMIA
TAMAI
ASWAN
GIZA
BAHTIM
SIDI
TAJOURA

030
031
027
029

13N
14N
42N
07N

031
031
030
030

12E
39E
45E
06E

015
031
032
032
032
030
027
026
025
032
032
032
032
032
032
031
033
034
033
033
036
034
036
036
036
036
037
035
036
034
034
034
036
036
036
044
067
035
035
035
034
035
031

08N
OON
SON
53N
28N
30N
SON
28N
OON
ION
40N
45N
45N
OON
20N
48N
33N
13N
59N
59N
34N
59N
48N
48N
44N
35N
03N
SON
38N
24N
25N
33N
SON
42N
06N
30E
12N
06N
04N
08N
52N
06N
17N

35
030
013
013
020
020
014
021
023
013
012
012
012
023
006
008
006
005
006
006
012
006
010
010
009
009
009
000
008
062
070
069
068
131
068
036
036
033
033
033
033
032
048

45E 440
OOE
21
05E
10
24E
11
53E 210
52E 310
40E 465
22E 300
OOE 415
OOE 725
50E
50
45E
32
45E
15
OOE 350
17W 500
07W 464
24W 500
43W
84
52W
25
52W
25
52W 450
SOW
65
12E
10
12E
10
08E 150
OOE 134
42E 120
OOE 350
50E 350
13E 964
27E 552
12E 1803
SSE 403
67E 378
39E
51
49N 804
42
350
28E 153
20E 200
24E 142
36E
25
25E
50
40E
20

032
036
035
028
029
036
036
035
036
036
030
030
029

16N
SIN
SON
45N
OON
OON
55N
47N
13N
37N
53N
20N
46N

048
054
050
054
055
050
054

25E
25E
58E
33E
OOE
OOE
20E

059
058
061
050
052

40E
60E
41E
OOS
43E

Abb. stat

EGY
EGY
EGY
EGY
EGY
EGY
EGY
EGY
LBY
LBY
LBY
LBY
LBY
LBY
LBY
LBY
LBY
LBY
LBY
LBY
MAR
MAR
MAR
MAR
MAR
MAR
MAR
MAR
TUN
TUN
TUN
TUN
TUN
TUN
TUN
AFG
AFG
AFG
AFG
AFG
AFG
AFG
AFG
CYP
CYP
CYP
CYP
CYP
IRN
IRN
IRN
IRN
IRN
IRN
IRN
IRN
IRN
IRN
IRN
IRN
IRN
IRN
IRN

19107
19108
19110
19111
19112
19113
19114
19115
19201
19202
19203
19204
19205
19206
19207
19208
19209
19210
19211
19212
19301
19302
19303
19304
19305
19306
19307
19308
19401
19407
19402
19403
19404
19405
19406
20001
20002
20003
20004
20005
20006
20007
20008
20201
20202
20203
20204
20205
20301
20302
20303
20304
20305
20306
20307
20308
20309
20310
20311
20312
20313
20314
20315

NICOSIA (2)
NICOSIA (I)
NICOSIA (3)
DROMOLAXIA
PRASTIO
AHWAZ
SAFIABAD
ARAGHEE MOH
KARAJ
FARS{l)
FARS(2)
CHALOOS
GORGAN
MAZANDARAN (1)
KHORASAN (I)
KHORASAN (2)
ZABOL
GACHSARAN
ZARGAN

EL JABEL AL AKHDAR
SEBHA
EL SARIR
KUFRA
JILEALEA
ZAWIA AL-ZAHRA
TOMENA
ZAWIA AL-ZAHRA
ZORDA BENGHAZI
BENI MELLAL
MARRAKECH
MARCHOUCH
SIDI KACEM
GUlCH (DEBBAGH)
DEBBAGH (GUlCH)
SIDI EL AYDI
INRA
TUNIS
ARIANA
BEJA
KOUDIAT
MATEUR
HINDI ZITOUN
LE KEF
HERAT
JALALABAD
KABUL
KUNDUZ
MAZAR-I-SHAR
BAGHLAN
TAKHAR
BALKH
NICOSIA
LAXIA
ATHALAS SA
LARMACA
SYNGRASST
KHUZESTAN
CHEPAN
DEZFUL
GORGAN
TEHRAN
DARAB
DARAB-HASSAN-ABAD
KELARDASHT
MAZANDARAN
DASHT NAZ
MASHHAD
FARUBROMAN NISHABOO
SISTAN & BALOCHESTAN
KOHGILOOYEH & BOVAIR AHMAD
FARS

73

Long.

Elev.
21
2
39
28

83
132
1321
1100
1100
1200
5
30
985
1350
49
71
160

Cnt

Hew

state

City

Lat.

ABU'GBRAIB
BAKRAJo
NINlVEH FAR
QADISSIA
ARBIL
NINEWA
KUFA
JORDAN VALLEY
JORDAN VALLEY
AMMAN
IRBID (2)
AMMAN
IRBID (3)
RABBAH
MAAN
MADABA
MOSHAGAR
IRBID (1)
MARU
TERBOL
TEL AMARA
NTH COASTAL PL
RAWDAT HARM
WADI EL-AREIG
RIYADH (2)
AL KHARJ
JIZAN
ONAIZAH
RIYADH (1)
IZRAA A.R.S
ALEPPO (1)
ALEPPO (2)
DAMASCUS ( 1 )
JELLIN-ACSA
EL ZIERCH
DAMASCUS (2)
CUKUROVA
SAKARYA
ANKARA
E .A.R. I.
LA.G.P.B.
IZMIR (4)
IZMIR
DIYARBAKIR
TRAKYA
IZMIR (1)
ADANA (2)
KASTAMONU
SAMSUN
AL-MINJEDAH
TAIZ (1)
BODEIDAH
CENT. HIGHLAND
TAIZ (2)
TAIZ (2 )
HAORAMOUT
BADRAMOUT
STHN UPLAND
KIRYAT-GAT
BET DAGAN
BET DAGAN
NVE-YA AR
REHOVOT

BAGHDAD
SULAIMANYA
MOSUL
F.C.R.S.
ARBIL
HAMMAN AL ALILE
KUFA
DEIR ALLA
UNIV. FARM
JUBEIHA
RAMTHA
WADI YABIS
IRBID
KERAK
SHOUBAK
MOUSHAKER
MOSHAGAR
MARROW
MARU
BEKA'A VALLEY
BEKA'A VALLEY

033
035
036
031
036
036
037
032
032
030
032
035
035
035
033
031
036
032
034
033
033
038
025
025
024
024
017
026
024
032
035
036

20N
34N
20N
59N
11N
01N
OON
12N
12N
01N
34N
36N
SIN
45N
30N
34N
12N
33N
32N
55N
55N
35N
48N
48N
37N
15N
55N
55N
20N
SIN
56N
05N

044
045
043
044
044
043
040
035
035
035
036
032
032
031
032
035
039
035
035
035
035
040
051
051
046
047
055
044
046
036
037
036

24E
34
23E 750
09E 223
59E
20
OOE 414
14E 305
OOE
24
36E -224
37E -224
52E 980
OlE 520
24E 200
32E 555
16E 970
35E 1400
48E 785
OOE 785
51E 618
51E 620
28E 950
28E 950
OOE
10
18E
50
18E
50
45E 700
10E 540
43E
92
59E 724
37E 600
15E 575
11E 300
55E 282

032
035
033
036
040
039
039
038
038
038
037
041

45N
28N
34N
SON
47N
40N
46N
27N
40N
35N
55N
40N

035
036
036
035
030
032
030
027
027
027
040
026

59E 421
03E
50
37E 617
20E
20
25E
30
39E 1044
31E 801
15E
28
04E
10
05E
20
12E 660
34E
48

Abb. stat

IRQ
IRQ
IRQ
IRQ
IRQ
IRQ
IRQ
JOR
JOR
JOR
JOR
JOR
JOR
JOR
JOR
JOR
JOR
JOR
JOR
LBN
LBN
LBN
QAT
QAT
SAU
SAU
SAU
SAU
SAU
SYR
SYR
SYR
SYR
SYR
SYR
SYR
TUR
TUR
TUR
TUR
TUR
TUR
TUR
TUR
TUR
TUR
TUR
TUR
TUR
YEM
YEM
YEM
YEM
YEM
YEM
YMD
YMD
YMD
ISR
ISR
ISR
ISR
ISR

20401
20402
20403
20404
20405
20406
20407
20501
20512
20502
20503
20504
20505
20506
20507
20508
20509
20510
20511
20701
20702
108
20901
20902
21001
21002
21003
21004
21005
21101
21102
21103
21104
21105
21106
21107
21201
21202
21203
21204
21205
21207
21210
21206
21208
21209
21211
21212
21213
21401
21402
21403
21404
21405
21406
21501
21502
21503
21601
21602
21605
21603
21604

DOHA
DOHA (RAWDAT HARM)
RIYADH
AL KHARJ
HAKMA
AL GASSIM
DIRAB
DERRA MOUHAPHAZA
BREDA
TEL HADYA
KARAHTA
DARAA
LATTAKIA
DAMASCUS
ADANA
ADAPAZARI
C.A.I.
ESKISEHIR
IZMIR
EGE RARI
AEGEAN
DIYARBAKIR
EDIRNE
E.B.Z.A.
CUKUROVA UNIV.
ORZA

037 19N 035 15E

041
015
013
015
014
013
013
016
016
013
031
032
032
032
031

SANA'A
AUSSEFERA
SURDUD
DHAMAR
ARA

SEIYUN
SEIYUN
TAIZ
MIVHOR FARM
VOLCANI
STET R.E.S.
WEIZMANN INST.

74

Long.

18N
23N
OON
15N
36N
57N
58N
OON
OON
42N
37N
OON
OON
05N
90N

036
044
042
043
044
044
044
049
049
044
034
034
034
035
034

20E
13E
30E
15E
21E
11E
12E
OOE
OOE
OOE
47E
48E
17E
OOE
80E

Elev.

90
10
2350
1350
170
2330
1750
1700
600
600
1350
120
30
30
60
50

Cnt Hew
Abb. stat

state

City

Lat.

BGD
BGD
BGD
BGD
BGD
BGD
BGD
BGD
BGD
BGD
BGD
BTN
BTN
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
IND
BUR
BUR
BUR
BUR
NPL
NPL
NPL
NPL
NPL
NPL
NPL
NPL
NPL
NPL
NPL
NPL
NPL
NPL
PAK
PAK
PAK
PAK

JOYDEBPUR
PABNA
JAMALPUR (1)
JAMALPUR (2)
DHAKA
JESSORE
MYMENSINGH
DINAJPUR (2)
DINAJPUR (1)
CHITAGONG
RAJSHAHI
WANDIPHODRANG
BHUR
BIHAR ( 1)
GUJARAT( 1)
GUJARAT( 1)
HARYANA(3)
KASHMIR
M.P.(2)
MAHARASHTRA
NEW DELHI
NEW DELHI
PUNJAB
RAJASTHAN
U.P.(5)
M.P.(l)
KARNATAKA(l)
U.P.(3)
BIHAR(2)
KARNATAKA(2)
WEST BENGAL
A.P.
HARYANA( 1)
u.P.(S)
GUJARAT(2)
U.P.(4)
KARNATAKA(3)
HARYANA(2)
RAJASTHA
U.P.(2)
U.P.(I)
SHAN
PYNMANA
YE-U SAGAIN
SAGAING
RUPANDEHI(2)
DHANUSHA ( 1 )
DHANUSHA(2)
LALITPUR
DHANKUTA
CHITWAN
DOLAKHA
BARA
GANDAKI
KABRE
RUPANDEHI(l)
KHAJURA
LAMJUNG
JHAPA
NWFP
PUNJAB
FAISALABAD
FAISALABAD

BARI
ISHURDI
JAMALPUR
JAMALPUR
DHAKA
JESSORE
BAU CAMPUS
RAJABARI
NASHIPUR
HATHAZARI
RAJSHAHI
WANDIPHODRANG
BHUR
PATNA
JUNAGADH

023
024
024
024
023
023
024
025
025
022
025
027
027
025
021
021
029
032
022
020
028
028
030
026
029
022
012
026
025
015
025
017

46N
25N
25N
56N
06N
13N
46N
67N
29N
30N
OON
31N
35N
33N
30N
30N
38N
43N
44N
06N
35N
35N
56N
58N
OON
37N
22N
27N
52N
42N
02N
19N

090
089
089
089
090
089
090
088
088
091
089
089
089
085
070
070
075
074
077
074
077
077
075
075
079
075
076
080
085
076
088
078

23E
8
OOE
8
07E
8
SSE
8
04E
7
8
13E
24E
19
70E
29
25E
29
47E
13
OOE
18
49E 1300
SSE 460
09E
51
28E 137
28E
60
06E 215
54E 314
42E 299
06E 549
12E 228
12E 228
52E 247
48E 450
OOE 243
50E 600
40E 638
25E 150
48E
51
07E 638
06E
38E 542

026
023
026
016
029
025
015

47N
OON
28N
SON
ION
13N
30N

082
072
080
071
075
075
060

12E 113
OOE 125
24E 123
07E 555
46E 215
25E 258
75E 4061

020
019
023
022
027
027
026
027
027
027
026

45N
SIN
02N
02N
30N
48N
48N
SON
05N
37N
CON

090
096
095
095
083
085
085
085
087
084
086

50E
07E
28E
02
25E
58E
58E
20E
20E
24E
OOE

22001
22002
22003
22007
22004
22005
22006
22008
22009
22010
22011
22101
22102
22201
22202
22221
22203
22204
22205
22206
22207
22218
22208
22209
22210
22211
22212
22213
22214
22215
22216
22217
22219
22220
22222
22223
22224
22225
22226
22227
22228
22401
22402
22403
22404
22501
22502
22503
22504
22505
22506
22507
22508
22509
22510
22511
22512
22513
22514
22601
22602
22603
22609

JAMMU
POWARKHEDA
NIPHAD
IARI
IARI
LUDHIANA
DUNGARPUR
PANTNAGAR
INDORE
MYSORE
BANARAS
PUSA
DHARWAR
MALDA
RAJENDRANAGAR
KARNAL
FAIZABAD MASADHA
UIJAPUR
KANPUR
UGAR
HISSAR
BARKHERA KOTA
KANPUR - AZAD UNIV.
DAVRALA MEERUT
HEHO
YEZIN
PANGON
ZALOKE FARM
BHAIRAHWA
JANAKPUR 1
JANAKPUR 2
KHUMALTAR
PAKHRIBAS
RAMPUR
JIRI
PARWANIPUR (NARAYANI)
TANAHU
KABRE
BHAIRAHWA (I.A.A.S.)
NEPALJUNG
SUNDAR BAZ
TARAHARA
PESHAWAR
BAHAWALPUR
FAISALABAD (LYALLPUR)
FAISALABAD

75

Long.

Elev.

1140
100
120
24
105
90
90
1369
1933
228
1700

028 03N 084 04E 580

027 06N 086 03E 1700
027 06N 082 04E 105
182
720
026 45N 087 20E
20
032 33N 068 OOE 340
029 25N 071 40E 170
031 30N 073 10E 213
031 30N 073 10E 213

Cnt Hew
Abb. stat

state

City

Lat.

PAK
PAK
PAK
PAK
PAK
PAK
PAK
PAK
PAK
PAK
PAK
LKA
LKA
LKA
THA
THA
THA
THA
THA
THA
THA
THA
THA
THA
THA
VNM
VNM
PHL
PHL
PHL
PHL
PHL
PHL
IDN
IDN
IDN
ION
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN

FAISALABAD
SIND
ISLAMABAD
BALUCHISTAN
ISLAMABAD
NWFP
SILSIT N.
PUNJAB
NWFP
NWFP
SIND
UPPER UVA
UPPER UVA
UVA
CHAING MAl ( 3 )
CHIANG MAl (I)
NAKHON RAT.
CHIANG MAl (2 )
CHIANG MAl (3)
CHIANG RAI (I)
CHIANG RAI (I)
CHIANG MAl (3)
LAMPANG
CHIANG RAI ( 2 )
SURIN
HANOI (2)
HANOI (1)
MANILA
LAGUNA
ISABELA
ISABELA
BAGUIO
CAGAYAN
WEST JAVA ( 2 )
WEST JAVA ( 3 )
WEST JAVA ( 1 )
W SUMATRA
BEIJING
BEIJING
BEIJING
BEIJING
BEIJING
FUJIAN
YUNNAN (3)
HEILONGJIANG-3
JIANGSU (4)
JIANGSU (I)
JIANGSU (I)
GUANGDONG
ZHEJIANG
JIANGSU (2)
HEILONGJIANG-2
IN. MONGOLIA-2
SHANGHAI
HUBEI
SICHUAN (1)
HEILONGJIANG-5
QINGHAI
QINGHAI
QINGHAI
JIANGSU (3 )
SHAN XI
YUNNAN (I)

FAISALABAD (AYUB)
TANDOJAM
(C.D.R.I. )
SARIAB QUETTA
ISLAMABAD
PIRSABAK
PEKORRA
BARANI
SERAI NAURUNG (BANNU)
DERA ISMAIL KHAN
KARACHI
BANDARAWELA
BANDARAWELA
MONERAGALA
SANPATONG
PANG DA
SUWAN FARM
SAMOENG
UNIV.
CHIANG RAI
CHIANG RAI
FANG
ARTC
BOONRRAWD FARM
SURIN
VAN DIEN
HANOI
QUEZON CITY
LOS BANOS
SAN MATEO
SAN MATEO
BENGUET
CAGAYAN
BOGOR
KUNINGAN
SUKAMANDI
SUKARAMI R.I.
BEIJING
BEIJING
BEIJING
BEIJING (CAAS)
BEIJING (AG. UNIV.)
FUZHOU
KUMMING
NAN HU
SUZHOU
NANJING
NANJING
GUANGZHOU
HENGZHOU
NANJING
KESHAN
KESHI KETENGGI
SHANGHAI ACADEMY
WUHAN
CHENGDU
HEIHEI

031
025
033
030
033
034
036

30N
02N
41N
11N
39N
OON
20N

073
063
073
066
073
072
073

10E 213
38E
19
07E 526
57E 1730
05E 683
OOE 288
50E 2120

034
031
024
006
006
006
018
018
014
018
018
019
019
019
018
019
014
021
021
014
014
017
017
016
018
006
006
006
001
039
040
039
039
039
026
025
048
034
032
032
023
030
032
048
042
031
030
030
050
036
036
036
033
037
024

OON
OON
SIN
SON
SON
SIN
46N
15N
40N
17N
47N
52N
52N
57N
21N
52N
53N
OlN
OON
38N
ION
08N
08N
40N
21N
405
595
295
005
48N
OON
59N
48N
56N
OON
07N
31N
19N
OON
03N
06N
12N
25N
04N
54N
13N
38N
40N
15N
38N
45N
44N
23N
47N
27N

071
070
067
080
080
081
099
098
102
098
098
099
099
099
099
099
103
105
105
121
121
121
121
120
121
106
108
107
100
116
116
116
118
116
119
102
114
117
118
118
113
120
119
125
117
121
114
104
127
101
101
101
120
112
118

OOE
00
02E
53E
53E
20E
02E
36E
OOE
36E
58E
47E
47E
10E
36E
47E
30E
48E
OOE
03E
15E
53E
53E
60E
37E
45E
50E
39E
30E
28E
OOE
16E
24E
20E
23E
43E
OOE
22E
48E
47E
18E
19E
25E
52E
36E
19E
04E
OlE
27E
38E
38E
52E
09E
33E
24E

22611
22604
22605
22606
22607
22608
22610
22612
22613
22614
22615
22701
22703
22702
24101
24102
24103
24104
24105
24106
24107
24108
24109
24110
24111
24212
24213
25001
25002
63
25003
25004
25005
25201
25202
25203
25204
27001
27016
27019
27021
27023
27002
27003
27004
27005
27006
27029
27007
27008
27009
27010
27011
27012
27013
27014
27015
27017
27025
27026
27018
27020
27022

XINING
DA YOU SHAN
TAIYAN
FUKIN

76

Long.

Elev.

171
8
1219
1219
120
300
820
300
820
314
403
403
430
275
450
146
5
5
45
21
69
69
1330
4
545
15
925
50
50
54
31
44
5
1916
29
34
9
67
8
8
12
1234
1235
4

23
506
168
2309
2309
2560
19
800
63

Cnt

Hew

state

City

Lat.

IN. MONGOLIA-1
NING XIA
HEIBEl
JILIN
SICHUAN (2)
SICHUAN (2)
YUNNAN (2)
HEILONGJIANG-1
HIELONGJIANG-4
NING XIA
KITAMI
FUKUOKA
KANAGAWA
SHIMANE
IBARAKI
CHUNAM (1)
CHUNAM (3)
HONAM
YEONGNAM
CHUNAM (2)
GYEONGGI PROVo

HU HEHAOTE
YONGNING
HAN DAN

040
038
036
043
030
030
024
045
048
036
043
033
035
035
036
034
035
035
035
034
036
026
042
059
049
046
045
040
043
042
042
043
042
050
050
049
048
050
047
046
046
051
050
052
050
052
050
052
050
054
049
051
052
052
044
046
044
045
045
045
045
045
043

Abb. stat

CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
CHN
JPN
JPN
JPN
JPN
JPN
KOR
KOR
KOR
KOR
KOR
KOR
TWN
SUN
SUN
SUN
SUN
SUN
ALB
BGR
BGR
BGR
BGR
BGR
CSK
CSK
CSK
CSK
CSK
HUN
HUN
HUN
POL
POL
POL
POL
POL
POL
POL
POL
POL
POL
POL
POL
POL
ROM
ROM
ROM
ROM
ROM
YUG
YUG
YUG
YUG

27024
27027
27028
27030
27031
27033
27032
27034
27035
27036
28101
28102
28103
28104
28105
28301
28302
28303
28304
28305
28306
28501
30001
30102
30103
30104
30105
31001
31101
31102
31103
31104
31105
31201
31202
31203
31204
31205
31410
31411
31412
31501
31502
31503
31504
31505
31506
31507
31508
31509
31510
31511
31512
83
31601
31602
31603
31604
31605
31701
352
31702
31703

ABKHAZIAN
SFSR
UKRAINE
UKRAINE
KRUMEX
DOBRUDJA (2 )
CHIRPAN
SADOVA PLOVDIV
DOBRUDJA (1)
SOFIA
BOHEMIA (3)
BOHEMIA (2)
BOHEMIA (1)
SLOVAKIA
BOHEMIA (4)
MARTONVASAR
KECSKMET
SZEGED
LASKI
PUSTKOW
CENTRAL
ULHOWEK
WARSAW
HENRYKOW
POZNAN
KRAKOW
ELBLAG
KRAKOW
RADOM
LESZNO
HENRYKOW 2
FUNDULEA
TURDA
BUCHAREST
TIMISOARA
PRASOV
CROATIA
NOVISAD
NOVISAD
MACEDONIA

GUAN XIAN
XIN DU DAFONG SIT
ME DO
HARBIN (AC. AGR. SCS.)
JIU SAN AGR. INS.
GU YAAN DIST. AGR. INST.
HOKKAIDO
KYUSHU
N.I.A.S.
S.A.E.S.
YATABE
MOKPO
KWANGJU
CHUNBUK
MILYANG
MUAN
SUWON
TAICHUNG
ABKHAZSKAYA ASSR
LENINGRAD
POLTAVA
ODESSA
KYACREOGAP
LUSHNJE
TOLBUKHIN
CHIRPAN
I.I.P.R.
GEN. TOSHEVO
SOFIA
STUPICE
PRAHA RUZYNE
KLATOVY MIROVKA
PIESTANY
KAMENICNA
MARTONVASAR
V.C.R.I.
SZEGED
RADOM
WARKA
WARSAWA
ZAMOSC
RADZIKOW
WROCLAW
NAGRADOWICE
KRAKOW
DEBINA
GRODKOWICE
DANKOW
CHORYN
WROCLAW
CALARASI
CLUJ
BUCHAREST
TIMISOARA
PRASOV
CROATIA
VOYVODINA
VOYVODINA
MACEDONIA

77

49N
14N
36N
54N
59N
56N
32N
41N
52N
OON
47N
12N
OON
20N
01N
47N
08N
55N
29N
47N
17N
57N
22N
56N
22N
27N
OON
57N
40N
12N
42N
40N
OON
05N
03N
23N
37N
05N
21N
55N
oN
45N
55N
14N
25N
12N
39N
20N
OON
ION
59N
42N
30N
18N
27N
35N
29N
46N
42N
49N
OON
OON
OON

Long.

Elev.

111
106
114
125
103
103
100
126
125
106
143
130
139
132
140
126
126
126
028
126
126
120
057
030
033
030
038
019
028
025
023
028
028
014
014
013
017
016
018
019
020
021
016
021
023
020
017
017
020
019
020
020
017
019
026
023
026
021
025
015
019
019
022

1041
1118
57
237
600
514
1659
172
288
1753
196
10
20
10
23
25
43
8
12
53
39
15
7

41E
15E
30E
13E
40E
91E
19E
37E
17E
15E
42E
30E
OOE
43E
08E
20E
56E
57E
45E
23E
59E
32E
OOE
18E
10E
42E
55E
42E
10E
20E
18E
10E
OOE
25E
20E
17E
50E
30E
49E
43E
OOE
15E
48E
OOE
25E
38E
OlE
10E
OOE
OOE
16E
44E
30E
06E
32E
48E
07E
15E
33E
59E
OOE
OOE
ODE

104
34
37
236
172
156
562
270
350
430
162
474
150
122
80
100
200
100
270
90
220
85
220
220
155
72
64
66
428
85
496
177
84
84
250

Cnt

Rew

state

Abb. stat

YUG 31704 BOSNIA 1ST

YUG 31705 BOSNIA 2ND
AUT 33001 LOWER AUSTRIA
BEL 33101 GEMBLOUX
BEL 33102 GEMBLOUX
BEL 33103 LINTER
GBR 33201 CAMBRIDGE
GBR 33202 NORFOLK
GBR 33203 LINCOLNSHIRE
GBR 33204 WILTSHIRE
GBR 33205 LINCOLNSHIRE
GBR
92 WALES
FRA 33301 SEINE ET MARNE
FRA 33302 PUY DE DOME
FRA 33303 NORD (1)
FRA 33304 HERAULT
FRA 33305 PARIS
FRA 33306 YVELINES
FRA 33307 EURE ET LOIRE-1
FRA 33308 OISE
FRA 33309 TARN
FRA 33310 NORD (3)
FRA 33311 NEW CALEDONIA
FRA 33312 GERS
FRA 33313 EURE ET LOIRE-2
FRA 33314 ILE DE FRANCE-2
FRA 33315 ILLE ET VlLAINE
FRA 33316 MORENVAL
FRA 33317 ILE DE FRANCE-1
FRA 33318 CHER (CENTER)
FRA
71 NORD (2)
DEU 31301 BAVARIA
DDR 33401 BAVARIA
DEU 31302 LWR SAXONY (2 )
DDR 33402 LWR SAXONY (2)
DEU 31303 MAGDEBURG (1)
DDR 33403 MAGDEBURG (1)
DEU 31304 SCHWERIN
DDR 33404 SCHWERIN
DEU 31305 BADEN-WURT. (2 )
DDR 33405 BADEN-WURT (2)
DEU 31306 LWR SAXONY (1)
DDR 33406 LWR SAXONY (1)
DEU 31307 MAGDEBURG (3)
DDR 33407 MAGDEBURG (3)
DEU 31308 BADEN-WURT. ( 1)
DDR 33408 BADEN-WURT (1)
DEU 31309 MAGEDBURG (2)
DEU
106 HOLSTEIN
DEU 31310 HOLSTEIN
IRL 33501 KILDARE
IRL 33502 LYONS
NLD 33701 WAGENINGEN
NLD 33702 LELYSTAD
NLD 33703 GRONINGEN
NLD 33704 RILKAND
CHE 33801 NYON
CHE 33802 ZURICH
GRC 35001 THESSALONIKI-1
GRC 35004 THESSALONIKI-1
GRC 35002 PLATY
GRC 35003 THESSALONIKI-2
GRC 35005 THESSALONIKI-3

City

Lat.

SOKOLAC
SOKOr...p.C
GROSS ENZERSZDORF
GEMBLOUX
GEMBLOUX
NEERHESPEN
P.B.I.
MILN MARSTERS LTD.
ELSOMS SEEDS
HARTHAM PARK
NICKERSON
PLAS GOGGERDAN
MONTCEAUX
CLERMONT-FERRAND
CANADA SEMENCES
MONTPELLIER
PARIS
VERSAILLES
CARGILL
C.A.C.B.A.
LA COURTADE
FLORIMOND DESPREZ
NESSADIOU
CLAEYS-POUY
BOISSAY (TOURY)
ORGERUS YVELINES
LE RHEU
PICARDIE
BENOIST
LE CLOS LAN
PERONN
FREISING-WEI
FREISING-WEI
WETZE
WETZE
LANGENSTEIN
LANGENSTEIN
GUELZOW GUESTROW
GUELZOW GUESTROW
HOHENHEIM
HOHENHEIM
BERGEN
BERGEN
HADMERSLEBEN
HADMERSLEBEN
NURTINGEN
NURTINGEN
GATERSLEBEN
SCHLESWIG
SCHLESWIG
BACKWESTON
LYONS
WAGENINGEN
CEBECO
GEERTSEMA ZADEN
VAN DE HAVE
NYON
ZURICH
EPANOMI P.B.I.
EPANOMI P.B.I.
PLATY
AGIOS MAMAS
CEREAL INST

043
043
048
050
050
050
052
052
052
051
053
052
048
045
048
043
048
048
048
049
043
050
021
044
049
048
048
049
048

56N
56N
12N
28N
34N
47N
ION
53N
48N
25N
90N
26N
21N
47N
01N
35N
52N
46N
5 N
ION
59N
34N
30S
OON
OON
SON
04N
OON
SON

018
018
016
004
004
005
000
000
000
002
000
004
002
003
002
003
002
002
001
002
002
003
002
001
002
001
001
003
001

48E
48E
34E
53E
40E
04E
06E
40E
07W
11W
lOW
01W
57E
06E
SOW
54E
20E
09N
08E
30W
OOE
05E
45E
08
OOE
43E
44W
OOE
42

050
048
048
051
051
051
051
053
053
048
048
049
049
052
052
048
048
051
051
051
053
053
051
052
053
51
046
047
040
040
040
040
040

OON
24N
24N
08N
08N
42N
42N
48N
48N
43N
43N
OON
OON
OON
OON
39N
39N
49N
25N
25N
30N
30N
58N
30N
13N
27N
32N
29N
38N
38N
39N
14N
40N

001
011
011
009
009
011
011
012
012
009
009
013
013
011
011
009
009
011
010
010
006
006
005
005
006
4
006
008
022
022
022
023
022

OOE
44E
44E
09E
09E
OOE
OOE
12E
12E
13E
13E
OOE
OOE
OOE
OOE
23E
23E
11E
36E
36E
40W
35W
40E
30E
35E
20E
40E
32E
57E
57E
31E
20E
60E

78

Long.

Elev.
860
860
153
160
160
45
17
79
1
400
60
33
53
340
100
18
130
120
183
50
0
150
130
101
40
10
12
17
59
467
467
125
125
202
202
12
12
407
407
70
70
87
87
310
310
110
5
5
50
65
5
-4
0
-4
430
445
10
10
10
14
10

Cnt Hew
Abb. stat

state

City

Lat.

GRC
GRC
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
ITA
MLT
PRT
PRT
PRT
PRT
PRT
PRT
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
ESP
FIN
FIN
NOR
NOR

VERMIO
GIANITSA
ROME (4)

VERMIO (MOUNTAIN)
GIANITSA
CASSACCIA
FOGGIA
RIETI
PONTE GALERIA
PIACENZA
CASACCIA
RIETI
MARC HE
INVIOLATELLA
INVIOLATELLA
BADIA POLESINE
FOGGIA
MACERATA
SEHAG-AZIENDA
MATERA
MATERA
MATERA
VITERBO
UNIV. TUSCIA
VALLETTA
ELVAS
BEJA
LISBON
LU.T.A.D.
VILLAREAL
CASAS VELHAS
EL ENCIN
MAHISA
COGULLADA
LN.LA.
JEREZ LA MERCED
TOMEJIL
LAS ALFAYATAS
EL ENCINAR
LA RINCONADA
LA RINCONADA
ZAMADUENAS
MONTANANA
LA ORDEN
ALAMEDA DEL OBISPO
L URGELL
PALUAU D ANGLEFOLA
CRUZ DEL CAMPO
BERGUS-CARDONA
CAMPILLOS DE LLERENA
EL PEDROSO
BOLDU
SEMILLAS AGRIC.
FINCA LA CARRERADA
GIMENELLS
GRANADA
EL PALMAREJO
EL ROCIO
LOBERAS
GUADIX
GOMARA
EL LLANO
HELSINKI
HANKKIJA
VOLLE BEKK
VOLLE BEKK

040 37N 022 30E

35006
35007
35101
35102
35103
35104
35105
35106
35107
35108
35109
35110
35111
35112
35113
35114
35115
35116
35117
35118
35119
35220
35301
35302
35303
35304
35305
35306
35401
35402
35403
35404
35405
35406
35407
35408
35409
35425
35410
35411
35412
35413
35414
35416
35415
35417
35418
35419
35420
35421
35422
35423
35424
35426
35427
35428
35429
35430
35431
37101
37102
37401
37404

ROME (5)
ROME (1)
PIACENZA
ANCONA
ROME (3)
ENEA & ITA
ROME (2)
ROME (2)
ROVIGO
ENEA & CERM
VITERBO
POLICORO
POLICORO
POLICORO
TARQUINIA
VITERBO
ALENTEJO
BEJA
LISBON
VILLA POUCA
TRAS-OS-MONTES
ELVAS
MADRID (3)
CORDOBA (2)
ZARAGOZA (1)
MADRID (2)
CADIZ
SEVILLA ( 3 )
CORDOBA ( 1 )
CORDOBA ( 4 )
SEVILLA (4 )
SEVILLA (4)
VALLADOLID
ZARAGOZA (2)
BADAJOZ ( 3 )
CORDOBA (3)
LLEIDA (1)
LLEIDA (1)
SEVILLA (5)
BARCELONA
BADAJOZ (1)
SEVILLA (6)
LLEIDA (2)
MADRID (1)
LLEIDA (3)
LLEIDA (4)
GRANADA (2)
SEVILLA (1)
HUELVA
BADAJOZ ( 2 )
GRANADA ( 1 )
SORIA
GUADALAJARA
HYRLA
AAS
AAS

79

Long.

Elev.
140

041
041
041
041
044
043
041
043
041
041
045
041
043

54N
28N
54N
48N
56N
30N
54N
15N
53N
53N
05N
27N
15N

012
015
012
012
012
013
012
013
012
012
011
003
013

29E
20
31E
76
29E 402
41E
3
27E
79
10E
30
29E -400
30E 229
29E
75
29E
75
30E
1
02E
76
30E 190

040
040
040
042
042
035
038
038
038
041
041

ION
ION
ION
13N
24N
53N
54N
01N
43N
30N
19N

016
016
016
011
012
014
007
007
009
007
007

40E
40E
40E
46E
08E
31E
09W
52W
08W
39W
44W

750
400

040
037
041
040
036
037
037
038
037
037
041
041
038
037
041
041
037
041
038
037
041
040
041
041
037

33N
53N
38N
26N
43N
24N
SON
OON
30W
30N
42N
43N
49N
53N
33N
33N
35N
55N
28N
50N
34N
OON
35N
40N
21N

003
004
000
003
006
005
000
004
005
005
004
000
006
004
000
000
006
001
005
005
001
003
004
000
003

19W
46W
54W
51W
09W
35W
59W
20W
57W
57W
43W
50W
39W
47W
55E
55E
OOW
40E
50W
45W
OlE
40W
25W
25E
35W

600
220
202
490
20
72
200
180
20
20
700
225
200
110
250
250
12
625
520
500
250
490
360
250
650

037
038
037
041
040
060
060
060
059

08N
47N
21N
40N
ION
05N
20N
OON
40N

006
006
003
002
001
024
025
011
010

30W
43W
05W
12W
lOW
54E
02E
OOE
47E

2
30
95
998
430
38
38
90
90

5
5
5
8
40
208
272
77

Cnt Hew
Abb. stat

stat.

City

Lat.

NOR
NOR
NOR
SWE
SWE
SWE
SWE
CAN
CAN
CAN
CAN
CAN
CAN
CAN
CAN
CAN
CAN
CAN
CAN
CAN
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA

STANGE
OSTRE TOTEN
OSLO
LANDS KRONA (2)
LANDS KRONA (1)
SVALOV
BJERTOP
ALBERTA (3)
MANITOBA
SASKATCHEWAN-3
SASKATCHEWAN-2
ALBERTA (4)
ALBERTA (2)
SASKATCHEWAN-l
ALBERTA (1)
P.E.I.
QUEBEC (1)
ONTARIO (1)
ONTARIO (2)
QUEBEC (2)
ARIZONA (4)
ARIZONA (2)
CALIFORNIA (4)
CALIFORNIA (3)
COLORADO (2)
COLORADO (1)
IDAHO
MONTANA
WASHINGTON ( 1 )
CALIFORNIA ( 6)
OREGON (1)
WASHINGTON (2 )
OREGON (2)
OREGON (3)
CALIFORNIA (1)
ARIZONA (1)
ARIZONA (3)
CALIFORNIA (2)
CALIFORNIA (5)
CALIFORNIA (7)
UTAH
WASHINGTON (4 )
WASHINGTON (3)
NTH. DAKOTA (1)
NTH DAKOTA (4)
STH DAKOTA
MINNESOTA (7)
MINNESOTA (5)
INDIANA
MICHIGAN
MISSOURI
NEBRASKA ( 1 )
OHIO
NEBRASKA (2)
LOUISIANA
TEXAS (3)
TEXAS (4)
TEXAS (1)
TEXAS (2)
NTH. DAKOTA (2 )
NTH. DAKOTA (3 )
TEXAS (5 )
MINNESOTA (6)

STAUR
APELSVOLL
JORDET
SVALOV
WEIBULLSHOL
SVALOV
BJERTOP
ELLERSLIE
WINNIPEG
SASKATOON
WATROUS
BEAVER LODGE
OLDS
SWIFT CURRENT
LACOMBE
CHARLOTTETOWN
LAVAL UNIV.
GUELPH
ELORA
SAINTE FOY
MESA
YUMA
DAVIS
FRESNO
FORT COLLINS
FORT COLLINS
ABERDEEN
BOZEMAN
PULLMAN
TULELAKE
CORVALLIS
ROYAL SLOPE
MADRAS
PENDLETON
IMPERIAL VALLEY EL CENTRO
PIONEER HI-BRED
MARICOPA
ARCO SEED
WOODLAND NORTHRUP KING
STOTTENBERG
LOGAN
RITZVILLE
LIND
CASSELTON
LANGDON
BROOKINGS
MOORHEAD
ST. PAUL
BROOKSTON
EAST LANSING
COLUMBIA
LINCOLN
WOOSTER
MEAD
A.E.S.
ERICKSON
MCGREGOR
BEEVILLE
UVALDE
AMENIA
N.A.P.B.
OVERTON
MINNEAPOLIS

061
060
059
056
055
055
058
053
049
052
051
055
051
050
051
046
046
043
043
046
033
032
038
036
041
040
042
045
046
041
044
046
045
045
032
032
033
036
038
056
041
047
046
046
048
044
047
044
040
042
039
040
040
041
030
030
031
028
029
047
047
032
045

37402
37403
79
37501
37502
37503
37504
40001
40002
40003
40004
40005
40006
40007
40008
40101
40102
40103
40104
40105
41001
41002
41003
41004
41005
41006
41007
41008
41009
41010
41011
41012
41013
41014
41015
41016
41017
41018
41019
41020
41021
41022
149
41101
41102
41103
41104
41105
41106
41107
41108
41109
41110
41111
41112
41113
41114
41115
41116
41117
41118
41119
41120

80

OON
42N
05N
OON
55N
35N
16N
34N
38N
ION
SON
12N
47N
17N
47N
20N
47N
OON
39N
48N
25N
44N
32N
43N
OON
35N
56N
04N
42N
58N
34N
55N
15N
30N
SON
38N
15N
40N
41N
SIN
45N
6 N
58N
45N
45N
20N
54N
57N
25N
40N
OON
48N
47N
ION
OON
35N
23N
23N
14N
OON
30N
OON
OON

Long.

Elev.

011
010
011
013
012
013
013
113
097
106
105
119
114
107
113
063
071
080
080
071
III
114
121
119
105
105
112
III
117
121
123
119
121
118
115
114
III
121
121
120
III
118
117
097
098
096
096
093
086
084
092
096
081
096
091
096
097
097
099
097
096
095
93

200
264
58
50
5
50
84
677
235
501
576
745
914
825
838
15
80
333
380
50
375
46
18
108
1518
1575
1341
1456
768
129
68
365
1235
454
9
23
363
20
21
150
1371
488
462
208
492
591
274
260
215

OOE
52E
OOE
OOE
50E
06E
06E
31W
07W
41W
SOW
24W
OOW
SOW
40W
OOW
18W
25W
25W
18W
52W
36W
46W
48W
OOW
OOW
SOW
09W
08W
28W
12W
15W
48W
26W
34W
45W
45W
36W
48W
27W
49W
25W
08W
15W
20W
SOW
48W
05W
56W
30W
40W
41W
56W
SOW
OOW
20W
30W
46W
45W
13W
SOW
OOW
lOW

228
360
315
1170
30
110
210
28
278
293
152
110
294

Cnt

Hew

state

City

Lat.

MINNESOTA (4 )
MINNESOTA (3)
ALABAMA (2)
GEORGIA (2)
ALABAMA (1)
GEORGIA (1)
KENTUCKY
NEW YORK
ALASKA
COAHUILA (3)
NUEVO LEON ( 1 )
SINALOA
SONORA (1)
SONORA (1)
COAHUILA (2)
CHIHUAHUA
SONORA (3)
TAMAULIPAS
B.C.N. (2 )
NUEVO LEON (2 )
SONORA (2)
DURANGO (1)
B.C.N. (3 )
B.C.N. (1 )
CAESTOD
DURANGO (2)
COAHUILA (1)
MICHOACAN (4)
MICHOACAN (5)
GUANAJUATO (2)
JALISCO (2)
TOLUCA
MICHOACAN (2 )
TLAXCALA
NAYARIT
VERACRUZ
AGUASCAL.
COLIMA
GUANAJUATO ( 1 )
GUANAJUATO (3 )
GUANAJUATO (4)
JALISCO (3)
TAMAULIPAS
MICHOACAN (1)
S.L.P.
CAEVAMEX
HIDALGO
JALISCO (1)
MICHOACAN (3)
TEXCOCO
ALAJUELA
ALAJUELA (1)
CARTAGO (2)
ALAJUELA (3)
CARTAGO (1)
ALAJUELA (2)
SL SALVADOR
QUEZALTENANGO
CHIMALTENANGO-1
CHIMALTENANGO-2
SAN PEDRO SULA
TEGUCIGALPA

EDEN PRAIRIE
STANTON
ALABAMA
GEORGIA
A.M.U.
GEORGIA
LEXINGTON
ITHACA
PALMER
ZARAGOZA
MONTERREY
LOS MOCHIS
CIANO
CIANO
BUENAVISTA
DELICIAS
NAVOJOA
RIO BRAVO
MEXICALI
LA LEGANA NAVIDAD
HERMOSILLO
VALLE DEL GUADIANA
CAEMEXI
CAECOEN
CD. CONSTITUCION
FCO. I. MADERO
CAESIA
VILLA MADERO
POTZUMARAN
EL BAJIO
ZAPOPAN
ATIZAPAN
PATZCUARO
HUAMANTLA
SANTIAGO
POZA RICA
PABELLON
LA POSTA
CORTAZAR
ROQUE
SAN JOSE ITURBIDE
TEPATITLAN
LAS HUASTECAS
CUTZITAN
EBANO
CHAPINGO
LAGUNILLA
JESUS MA.
BRISENAS
EL BATAN IRRIGATED
POASITO
FABIO BAUDRIT MORENO
DURAN
FRAIJANES
ALTURA
NARANJO
EL SALVADOR
LABOR OVALLE
CHlMALTENANGO
PATZUN
SAN PEDRO SULA
TEGUCIGALPA

044
044
034
031
030

49N
29N
39N
28N
30N

93
93
086
083
088

27w
01W
46W
30W
OOW

277
278
190
370

038
45
061
028
025
025
027
027
025
028
027
026
031
025
028
024
032
031
024

OON
5 N
34N
33N
05N
48N
20N
20N
22N
11N
02N
OON
40N
40N
36N
01N
31N
35N
57N

084
76
149
100
100
109
109
109
101
105
109
098
114
100
111
104
115
116
111

30W
5 w
16W
ssw
36W
OOW
54W
54W
OOW
30W
25W
13W
45W
36W
12W
40W
26W
61W
42W

304
305
61
350
537
15
38
38
1785
1170
40

Abb. stat

USA
USA
USA
USA
USA
USA
USA
USA
USA
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
MEX
CRI
CRI
CRI
CRI
CRI
CRI
SLV
GTM
GTM
GTM
HND
HND

41121
41122
41201
41202
41203
41204
41205
41206
41307
42001
42002
42003
42004
42016
42005
42006
42007
42008
42009
42010
42011
42012
42013
42014
42015
42017
42018
42019
42020
42101
42102
42103
42104
42105
42106
42107
42108
42109
42110
42111
42112
42113
42114
42115
42116
42117
42118
42119
42120
42121
45101
45102
45103
45104
45105
45106
45201
45301
45302
45303
45401
45402

81

Long.

Elev.

25
1895
60
1889
6
720
48

025 13N 100 35W 2140

020
020
019
019
019
021
020
022
019
020
020
021
020
022
019
022
019
020
020
020
019
009
010
009
010
009
010
013
014
014
014
015
014

32N
40N
16N
30N
19N
20N
34N
09N
27N
29N
34N
OON
42N
33N
20N
12N
17N
32N
37N
15N
31N
SON
01N
56N
15N
56N
06N
35N
52N
39N
39N
31N
12N

100 49W 1765

1550
099 51W 2640
101 40W 2040
097 56W 2553
105 20W
097 26W
60
102 17W 1912
104 37W
100 58W 1750
100 SOW 1765
100 24W 1870
102 42 1960
098 31W
40
101 40 2350
098 23W
65
098 53W 2249
100 49W 1764
102 11W 2110
102 33W 1517
098 SOW 2249
084 02W 1600
084 16W 840
083 52W 2337
084 15W 1650
083 52W 2200
084 22W 900
089 lOW
091 30W 2407
090 49W 1790
090 49W 2035
088 01W
087 09W

Cnt

ew

state

City

Lat

Abb. stat

NIC
NIC
NIC
CUB
DOM
DOM
DOM
JAM
JAM
PRI
TTO
NCL
THI
AUS
AUS
AUS
AUS
AUS
AUS
AUS
AUS
AUS
NZL
NZL
NZL
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
BRA
ARG
ARG

45501
45502
45503
47501
47701
47702
47703
48101
48102
48501
49001
51501
51901
55001
55002
55003
55004
55005
55006
55007
55009
55008
56010
56011
56012
60001
60003
60002
60004
60005
60006
60007
60101
60119
60102
60112
60124
60129
60103
60104
60110
60105
60116
60106
60107
60108
60117
60109
60111
60113
60114
60115
60118
60120
60122
60121
60123
60125
60126
60127
60128
61001
61002

MANAGU
MATAGALPA
JINOTEGA
SIBONEY
SANTIAGO (2 )
SANTIAGO ( 1 )
AZU
KINGSTON
LAWRENCEFIELD

MANAGUA
ESTANCIA CORA
JINOTEGA
HAVANA
SANTIAGO
QUINIGUA
AZUA
HOPE GARDEN
LAWRENCEFIELD
MAYAGUEZ
ST. AUGUSTINE
BOURAIL
PAPARA
SA (2)
ADELAIDE
QLD
TOOWOOMBA
VIC (2)
WERRIBEE
SA (1)
ROSEWORTHY
NSW (1)
TAMWORTH
NSW (3)
WAGGA WAGGA
VIC (1)
HORSHAM
VIC (1)
HORSHAM
NSW (2)
CASTLEHILL
MANAWATU
PALMER5TON NTH
CANTERBURY
LINCOLN
SOUTHLAND
GORE
DF
BRASILIA
DF
BRASILIA
SALVADOR
BAHIA
M. GERAIS
SAO GOTARDO
SAO PAULO (2 )
GUAlRA (FAZENDA MATEIRO)
SAO PAULO (1)
CRUZ ALTA
5TA. CATAR. ( 1 ) CAMPOS NOVOS
RGS (3)
CRUZ ALTA
RGS ( 3 )
CRUZ ALTA
RGS (1)
PASSO FUNDO
RGS (1)
PASSO FUNDo
RGS (1)
PASSO FUNDO
RGS (1)
PASSO FUNDO
SAO PAULO (3)
CAMPINAS
PARANA (1)
LONDRINA
PARANA (1)
LONDRlNA
PARANA (6)
PONTA GROSSA
PARANA (6)
PONTA GROSSA
RGS (8)
BAGE
RGS (6)
PORTO ALEGRE
PARANA (4)
PALOTINA (OCEPAR)
PARANA (4)
PALOTINA
PARANA (5)
CASCAVEL (OCEPAR)
MS
DOURADOS
SAO PAULO (4)
CAPAO BONITO
RGS (4)
SAO BORJA
PARANA (2)
FAXINAL
PARANA (7)
MORRETES
RGS (2)
VACARIA
RGS (2)
VACARIA
PARANA (3)
GUAlRA
RGS (7)
DA PALMA
STA. CATAR. ( 2) MAFRA
PEDRAS ALTAS
PEDRAS ALTAS
RGS (5)
JULIO DE CASTILHO
BA (4)
BALCARCE
BA (7)
LA DULCE

82

Long.

Elev.

012
013
013
023
019
019
018
018
017
018

ION
OON
17N
08N
34N
31N
23N
OON
SON
17N

086
085
086
082
070
071
070
076
000
067

21W
OOW 200
11W 1000
21W
46W
46W 148
SOW
25
46W
77W
15
08W

021
017
034
027
037
034
031
035
036
036
033
040
043
046
015
015
013
019
020

OOS
305
58S
30S
54S
32S
05S
07S
45S
43S
44S
23S
38S
07S
45S
35S
OOS
30S
20S

167
149
138
151
144
138
151
147
142
142
150
175
172
168
047
047
038
046
048

OOE
30E
2
38E 123
58E 666
38E
23
40E
60
56E 600
22E
22
15E 138
12E 138
10E 122
37E
15
30E
11
54E 123
47W 1000
42W 1000
38W
3
OOW 1100
18W 490

000
028
028
028
028
028
028
022
023
023
025
025
031
030
024
024
024
022
024
028
024
025
028
028
024
031
026
031
031
029
037
038

34S
38S
38S
15S
15S
15S
15S
53S
22S
22S
06S
30S
20S
08S
17S
18S
56S
14S
OOS
39S
OOS
30S
30S
30S
05S
OOS
lOS
40S
41S
13S
45S
20S

051
053
053
052
052
052
052
047
051
051
050
050
054
051
053
053
053
054
048
056
051
048
050
050
054
052
049
053
053
053
058
059

13W
36W
36W
24W
24W
24W
24W
04W
lOW
lOW
lOW
20W
05W
08W
50W
SSW
26W
4'9W
20W
OOW
21W
49W
56W
56W
15W
30W
80W
35W
45W
40W
18W
40W

920
473
473
684
684
684
684
663
540
540
868
900
214
6
300
310
760
452
710
96
730
59
955
955
230
127
800
300
250
516
130
72

Cnt

Hew

state

City

Lat.

BA (7)
BA (1)
CORDOBA
TUCUMAN
BA (5)
CHACO (1)
BA (8)
PARANA
BA (3)
LA PAMPA
BA (6)
BA (9)
CHACO (2)
BA (10)
BA (2)
CAUTIN (1)
NUBLE
SANTIAGO (3 )
SANTIAGO ( 1)
SANTIAGO ( 2 )
RANCAGUA
RANCAGUA
CAUTIN (2)
VALLENAR

LA DULCE (BUCK)
PERGAMINO
MARCOS JUAREZ
SAN MIGUEL DE TUCUMAN
BORDENAVE
R.S. PENA
TRES ARROYOS
ENTRE RIOS
PLA (CRIADERO KLEIN)
ANGUIL
TRES ARROYAS (CHACRA)
NECOCHEA (CARGILL)
COL. BENITEZ
BAHIA BLANCO (ACA)
CASTELAR
TEMUCO
CHILLAN (QUlLAMAPU)
HUELQUEN
LA PLATINA
PIRQUE
GRANEROS
GRANEROS
GORBEA (BAER )
HDA LAS VENTANAS
HIDANGO
OSORNO
CAACUPE (IAN)
CAACUPE
CAP. MIRANDA
CAP. MESA (LAPACHO)
HERANDORIAS
CAMPO A.
VOLENDAM
LA ESTANZUELA
SALTO
SALTO
SAYAGO
YOUNG (LA ESTAN. )
MENAFRA
BETANZOS
SAN BENITO
ABAPO IZOGZOG (CORGEPAI)
CHINOLI
ABAPO IZOGZOG (CORGEPAI)
PAl RUMAN I
SAAVEDRA
LA VIOLETA
TIBAITATA
OBONUCO
META
MOTILONIA
lMUES
STA. CATALINA
CUENCA
EL ANGEL
STA. CATALINA (EL PUGRO)
COTOPAXI
GEORGETOWN
ANCASH
HUANCAYO
LA MOLINA
CAJAMARCA
SALCEDO

038
033
032
026
037
026
038
031
035
036
038
038
027
038
034
038
036
033
033
033
034
034
039
028
034
040
025
025
027
026

205
565
425
485
515
525
235
50S
01N
30S
lOS
29S
255
295
405
41S
31S
51S
27S
40S
03S
03S
12N
34S
07S
52S
355
24S
17S
265

059
060
062
065
063
060
060
060
060
063
058
058
058
061
058
072
071
070
070
070
070
070
072
071
071
073
057
057
055
054

OOW
33W
07W
12W
01W
27W
16W
31W
15W
59W
OOW
49W
56W
54W
26W
25W
55W
41W
38W
36W
42W
42W
80W
OOW
44W
12W
32W
06W
49W
41W

72
65
110
460
212
91
109
110
55
165
50
20
54
159
22
200
217
39
629
654
479
500
90
348
120

034
034
031
034
032
033
019
017
016
019
018
017
017
017
004
001
004
010

20S
20S
20S
45S
40S
lOS
40S
30S
30S
34N
39S
21S
14S
205
04N
13N
07N
02N

057
057
058
056
057
058
065
066
068
065
063
066
063
066
074
077
073
073

41W
41W
OOW
lOW
41W
lOW
30W
06W
08W
24W
01W
19W
lOW
33W
12W
16W
30W
13W

81
81
50
0
50
50
3450
2730

000
002
000
000

225
50S
40N
22S

078
078
077
078

33W
50W
50W
33W

3050
2370
3110
3200

006
009
012
012
007
015

46N
15S
05N
05S
08S
535

058
077
075
076
078
070

lOW
40W
08W
57W
39W
OOW

2600
3245
251
2600
3835

Abb. stat

ARG
ARG
ARG
ARG
ARG
ARG
ARG
ARG
ARG
ARG
ARG
ARG
ARG
ARG
ARG
CHL
CHL
CHL
CHL
CHL
CHL
CHL
CHL
CHL
CHL
CHL
PRY
PRY
PRY
PRY
PRY
PRY
PRY
URY
URY
URY
URY
URY
URY
BOL
BOL
BOL
BOL
BOL
BOL
BOL
BOL
COL
COL
COL
COL
COL
ECU
ECU
ECU
ECU
ECU
GUY
PER
PER
PER
PER
PER

61014
61003
61004
61005
61006
61007
61008
61009
61010
61011
61012
61013
61015
61016
61017
61201
61202
61203
61204
61205
61206
61210
61207
61208
61209
61211
61401
61403
61402
61404
61405
61406
61407
61501
61502
61503
61504
61505
61506
63001
63002
63003
63004
63005
63006
63007
63008
63101
63102
63103
63104
63105
63201
63202
63203
63204
63205
63401
63501
63502
63503
63504
63505

LA PAMPA
CORDILLERA
CORDILLERA
ITAPUA (2)
ITAPUA (1)
ALTO PARANA
CAAGUAZU
SAN PEDRO
COLONIA
SALTO (2)
SALTO (1)
MONTEVIDEO
RIO NEGRO (1)
RIO NEGRO (2)
POTOSI
COCHABAMBA ( 3 )
SANTA CRUZ (1 )
POTOSI
SANTA CRUZ ( 3 )
COCHABAMBA (2 )
SANTA CRUZ (2 )
COCHABAMBA (1)
CUNDINAMARCA
NARINO
CESAR
NARINO
PICHINCHA ( 1 )
AZUAY
CARCHI
PICHINCHA ( 2 )
PICHINCHA
KINGSTON
ANCASH
JUNIN
LIMA (2)
CAJAMARCA (2 )
PUNO

83

Long.

Elev.

228
200
294

3450
386
2584
320
2680
2550
2710
330
130

Cnt

Hew

stat.

City

Lat.

CUSCO (2)
AREQUIPA
CUSCO (1)
AYACUCHO
CAJAMARCA ( 1 )
LIMA (1)
ARAGUA (1)
ARAGUA (2)

ANDENES
SANTA RITA
TARAY

013
016
013
013

Long.

Elev.

Abb. stat

PER
PER
PER
PER
PER
PER
VEN
VEN

63506
63507
63508
63509
63510
63511
63701
63702

CAJABAMBA
HUARAL DONOSO
CAGUA

27S
30S
27S
13S

072
072
072
072

16W 3391
lOW 610
16N 2910
12W 2720

011 28S 017 14W

010 llN 067 30W
010 15N 067 36W

182
444
450

Appendix 6--Genotypes (Sorted by Common Name) Used in ISWYNs,

LISting as of February 1992
Many selection histories and crosses were unavailable from ISWYN reports. We will use WPMS
to check the list beginning on page 85 for errors and to extract additional stored information,
including synonyms and missing selection histories where possible. Abbreviations will be checked
against the approved WPMS list. However, as historical information is incomplete, we would
appreciate notification of errors and receiving missing information. Please send these to the
first named editor of this document.

84

Appendix 6Genotypes Used in ISWYNs

Uniq. Abbr.

Common HIUIl
Selection History

1267

5149

209
185
210
211
1001
1002
1003
159

11IS9
340013
521/69
715/70
77 .13
79.218
25IS21
m789

(tel)/Anza*8
6A250(tel)/ANZA*8
S149-3-15-3-5-3-5-2-0(lR)
309
3400-1-3
521/69
715/70
77-13
79-218
793-3402
7C//TOB/NAP063
7C//TOB/NAPO
CM789-21-1A-2Y-OM
MAY054E/LR64//TAC/3/LR64//TZPP/Y54
ARZ
II21419-288
ATA 81
KVZ/CUT75
SE1066-9S-1S-6S-0S-7KE-OKE
Abu-Ghraib #3
SN64/KLRE
IIl9975
Aintree
Akbar
RON/TOB
CM7705-3M-IY-2M-2Y-OY
ABN"R" /MIA
Alamos 83 (tel)
X24SS1
E4870/C306//M5392-66.5/3/BB//CC/
Almansor 1
INIA66
Alondra "S"
D6301/NAI60//WRM/3/CNO*2/CHR
CM11683
Alondra "SOIA
D6301/NAI60//WRM/3/CNO*2/CHR
CM11683-A-1Y-1M-3Y-2M-OY
Alondra 4546
D6301/NAI60//WRM/3/CNO*2/CHR
CM11683
RUFF/FG//MEXI75/3/SHWA
Altar 84 (durum)
CD22344
Altar INIAP
IAS58/IAS55//ALD/3/MRNG/4/ALD/
Alubue
IASS8.103A//ALD/5/BUC
CM74362-(1-10)M-07Y026M-SY-1B-OY
Alva
1I12300//LR64/81S6/3/NOR
Anahuac 75
II30842
Th/2*Supreza/3/Fn//Kenya58/N/7/
Angus
pembina/Fn/5*Th/6/Mida//Kenyal17A/
2*Th/3/Fn/4*Th/4/MNIII584/5/
Kenya58/N//3*Le
CNO/GLL
Antizana
II27829-19Y-2M-4Y-OM
LR/N10B//3*ANE
Anza
II8739-4R-1M-1R
K338/EDCH
Ariana 66
Arvand
908/FN*2//4160/3/YT54/N10B/4/2*C14
Aurifen
CH7817-3P-4P-1P-2P-1P
PI/CHR//SN64
Azteea 67
II199S7-18M-1Y-3M-9Y
BL 1022
BL 1044
BL 1049
BR 15
AS54*2/TOKAI80//PF6913
P73387-1P-37F-OF-OR-1F-OR
BR 16
PAT70402/ALONDRA//PAT72160/ALONDRA
B19789-H-508M-1Y-10F-701Y-1F-700Y
BT 2281
TZPP/SN64//LR64/SN64/3/
SON64A/SKE*3/AN
TACUARI/PJ62
BT 2288

87

ARZ

1114

ATA

240

AGB3

1005
1229

AINT
AKBAR

1006

tel7

1083

ALMR

1155

ALD

1157

ALD Sl

1156

ALD S2

1153

durum5

1007
1227

Altar
ALUBUC

1004
305

ALVA
ANA

1132

AGS

1259

ANT

21

ANZA

191
213
1154

ARI
ARV
AUFN

89

AZ

1008
1009
1059
1264

BL1022
BL1044
BL1049
BR15

1150

BR16

192

BT2281

193

BT2288

Cross

85

Uniq. Abbr.

COJlllDOn Rame

Cross

Selection History

1010
1219

BT501
m58766

1075

cl9268

1225

BCN

703

tc13

1221

BAU

90
139

BJ66
BAJ66S

1086
1232

BALA
BYA

1011
160

BANKS
BAN

1171

BAN83

362

BART

1159

BTU

1244

durum3

126

BB#l

157

BB#2

135

BB#3

94

BB#4

161

BB PR1

162

BB PR2

164

BB S#7

1183

BOW

1184

BOW Sl

1188

BOW S2

1185

BOW#l

199

BON

46

BZA55

47

BZA63

93

BCH

1170

BUC

BMAN

BT 501
BUC/PVN
BUC/PAVON
CM58766-l8Y-3M-5Y-2M-OY
BUCKY/TOB/CNO
BUCKY//TOB/CN067
CID19268 no selection history
Bacanora 88
JUP/BJY//URES
CM67458-4Y-1M-3Y-1M-5Y-OB
Bacum (tcl)
MAYA II/ARM
X2832-24N-3M-7N-4M-OY
Bagula
TTR/JUN
CM59123-3M-1Y-3M-1Y-3M-OY
Bajio 66
SN64//TZPP/NAI60
1I18889-4M-1Y-1M-3Y
Bajio 66 "s"
SN64//TZPP/NAI60
1I18889-4M-1Y-3Y
Balaka
E5557/H 845
Balanya 80
WE/GTO//KAL/BB
CM8288
Banks
Barani 70
PI62/GB55//C271
PK50-18A-4A-13A
Barani 83
BB/GLL/3/GTO/7C//BB/CNO
CM32347-3M-1Y-1M-1Y-1B-OA
Barkat
BB/GLL//CARP/3/PVN
CM33483-C-7M-1Y-OM
Baturira
TZPP*2/AN64//INIA66/3/CNO/JAR//KVZ
CM21335-C-9Y-3M-1Y-1Y-IY-OB
Bittern "s" (durum)
JO/AA//FG
CM9799-126M-1M-4Y-OY-OM
Bluebird #1
CNO//SN64/KLRE/3/8156
1I23584-15Y-6M-OY
Bluebird #2
CNO//SN64/KLRE/3/B156
1I23584-26Y-2M-1Y-OM
Bluebird #3
CNO//SN64/KLRE/3/8156
1I23584-26Y-2M-2Y-OM
Bluebird #4
CNO//SN64/KLRE/3/8156
1I23584-26Y-2M-3Y-2M-OY
Bluebird Pak Res1
CNO//SN64/KLRE/3/8156
1I23584-303M-OY-11A-8A-2304
Bluebird Pak Res2
CNO//SN64/KLRE/3/8156
1I23584-303M-OY-11A-1A
Bluebird Sam #7
CNO//SN64/KLRE/3/8156
1I235B4-102M-OY-6M-OY-1T-OT
Bobwhite
AU//KAL/BB/3/WOP
CM33203
Bobwhite "s"
AU//KAL/BB/3/WOP
CM33203-F-4M-4Y-1M-1Y-OM
Bobwhite "5"2
AU//KAL/BB/3/WOP
CM33203-K-9M-2Y-1M-1Y-2M-OY
Bobwhite #1
AU//KAL/BB/3/WOP
CM33203-K-IOM-7Y-3M-2Y-1M-OY
Bonanza
PI/CHR//SN64
II19957-18M-1Y-3M-(?)
Bonza 55
Y50/KT48
1I2254-2p-l1lb-4b-lb
Bonza 63
RN/2*BZA55
VI-36-2-30b-2t-1b
Brochis "5"
CNO/BB//CDL/4/7C/3/LR64/INIA66//
CM5872-C-1Y-IM-3Y-OM
INIA66/BB
Buck Buck
BY/MAYA/4/BB//HD832.5.5/0N/3/CNO/PJ
CM31678-R-4Y-2M-24Y-1M-OY
Buck Manantial
RMAG/BQQ

86

Uniq. Abbr.

COllDDOn Raae
Selection History

Cross

1074

BMAP

Buek Mapuehe

1161

BOMB

Buek Ornbu

BUCK PANGARE/3/RAFAELA MAG/BUCK

PAMPERO//BUCK RELEN
CNO/GLL/TOB/BMANI/BB/4/TXPP/
SON64//TZPP*2/AN

BPAT

CM26339-E-4Y-IY-2Y-IM
-15Y-3LD
Buek Pataeon

1076
175
231
166
167
62
63
64
1133
1012
1149

BUL
BTT
C271
C273
C306
C518
C591
C7659
C8055
CEP14

313
367
375

CGT700
CGT705
rn1221

104

rn8943

103

i30793

1234

rn83271

35
1013
1201

CT244
te15
CAETES

95
152

CAL
CAL A

1014
96

te14
CAN

31
194

CZHO
CRT

1015
39

CTAN
CTFN

1087
1182

CH74
CHAT

168
1123
65

CHB
CHENAB
CHL

215

CHRC

1279

te16

1016

CHIV

CAEREN 2.4.2/3/RAFAELA MAG/

B. PAM//BAGE/K. PETISO
PI/FROND/3/PI/MZ//MXP

Bulbul
PK2858-7A-3A-4A-OA
Butte
C230/NP165
C271
C 273
C519/C209
RGN 1974/CHZC3//2*C591/3/P19/C281
C 306
PB TYPE 8A/PB TYPE 9
C 518
PB TYPE 8B/PB TYPE 9
C 591
C 7659
UNKNOWN
C 8055
PEL72 3 8 0 / ATR71
CEP 14 Tapes
B13374-3Z-1A-6A-2A-OS
CAN/JAR66
CGT 700
TOB/CI
CGT 705
CNO//NAD63/CHR/3/S0N/KLRE//EN
CM1221
CM1221-57M-3Y-OM-KOOO
CNO/7C//CNO/INIA/3/SX
CNO/7C//CNO/INIA/3/SX
CM8943-F-1M-4Y-OM
CNO/PJ62/GALLO
CNO/PJ62/GLL
II30793-1M-2R-2M-OR
CNO/PRL
CN079/PRL
CM83271-18Y-1B-5Y-1B-OY
CT 244
Caborea 79 (tel)
Caete "s"
JUP/BJY
CM39992-8M-3Y-1M-OY
TZPP/SON64A/TZPP//ANE
Calidad
II22429-16M-IY-IM-OY
Tzpp/sON64A/TZPP//ANE
Calidad S A
II22429-11M-IY-IM-OY
Cananea (tel)
Canario "s"
HK/38MA//2*RFN/90S/3/FN/4/YR
CM2097-31M-IY-IM-4Y-OM
COLONISTA/FRONTANA
Carazinho
NAPO/TOB//S156
Carthage
II28071-7M-3Y-7M-OY
Castan
YAQUI53/N10B//2*LR
centrifen
CM7920-9R-1M-2R
FRATERNAL SELECTION
Chapingo 74
Chat
KVZ/TI//TITO
CM33090-M-4M-2Y-5M-OY
Chenab 70
C271/WILE//SN64
PK146-12A-4A-IA
PK3563/cH70
Chenab 79
LR64/HUAR
Chhoti Lerma
II15929-1M-4Y-2M
SD648.5/8156/3/CHR//SN64/KLRE/41
Chiroea
BB/CAL/5/ZBZ
CM8963-A-IM-IY-IM-5Y6M-OY
ABN"R" /MIA
Chiva (tel)
X24551-8Y-3M-IY-OM
Chivito

87

Uniq. Abbr.

99

CLI

1194

CVA

202

CHR

1017
1158
98

CHUM18
CKR
CN067

131

CN067A

200

CN067B

97

CN067C

317

CN079

100

CGN

1275

CISI

101

CLEO

114

COC

1018
369
217
48

durum2
COOK
COW
CPO

201

CR1M

1205

CROW

1193

CZI

230

HMN

228

CUC RS

1207

CUPE

1208

CUPE S

1073

CUVA

1222

CUMPAS

1282

CUMPAB

1091
2

DWR39
DTE

195
3

DGA
ELGA

Common Hame
Selection History

Cro

Choli
21931/CH53//2*LR64/3/8156/4/NAR59
1I21515-1P-lP-3P-5M-OY
Chova
GLL/YR RESEL(B)/3/AU//KAL/BB
CM34603-A-1M-3Y-3M-1Y-1M-OY
Chris
FRONTANA/3*THATCHER/3/KEN58/
NEWTHATCH//2*THATCHER
Chuan Mai 18
Chukar
PTS/3/TOB/CTFN//BB/4/BB/HD832.5.5/0N
CM20769-A-8Y-1M-2Y-5M-1Y-OM
Ciano 67
PI/CHR//SN64
I119957-18M-1Y-3M-9Y
ciano 67 "SUA
PI/CHR//SN64
1119957
Ciano 67 "SUB
PI/CHR//SN64
II19957-18M-2Y-2M-1Y-7M-1Y
Ciano 67 "s"c
PI/CHR//SN64
1I19957-18M-2Y-6M-1Y-2M-1Y
Ciano 79
BY/MAYA/4/BB//HD832.5.5/0N/3/CNO/PJ
CM31678-R-4Y-2M-21Y-OM
Ciguena
SN64*2//TZPP/TZPP/Y54/3/AN64A/4/
1I21406-6-2-300Y-301M-OY 2*FR//Y/KT
Cisne INIA
KVZ/CJ
SWM1430-4Y-3Y-OM
Cleopatra 74
BB/3/2*INIA66/NAPO//II20350/4/F2
CF258-30R-4M-OT
Cocoraque 75
I112300//LR64/8156/3/NOR
II30842-58R-1M-4Y-OM
Cocorit (durum)
RAE/4*TC60//STW63/3/AA
D27617
Cook
Cowbird
KLAT//INIA66/BB/4/NP876/PJ//CAL/3/BB
CM16716-M-3M-2Y-3M-OY
Crespo 63
FR/4/N/3/2*MT/K//BAGE/5/GB60
11l1263-3T-1B-2T-1B-1T
crim
THATCHER/3/THATCHER*3/KLEIN
T1TAN//KENYA 58/NEWTHATCH
CrownS"
FR316/3/MCM/KT//Y50/4/ZA75/S/BJY
CM40457
Cruz Alta 1NTA
KVZ/BON/7/2193/CH53//AN/3/GB/4/
CM33729-A-1M-2Y-1Y-3Y
PJ/5/S0TY/6/YR RESEL
-OM-1J-OS-?
Cuckoo "S"A
7C/ON//1NIA66/BMAN
1I28428-8Y-1M-1Y-1M
Cuckoo ReSel
7C/ON//INIA66/BMAN
II28428-8Y-1M-OM-(1-16B)
Cucurpe S 86
HIM/COC//NAC
CM41195-A-13M-2Y-3M-1Y-1M-OY
Cucurpe S 86 "S"
HIM/COC//NAC
CM41195-J-7M-1M-OM
Cukurova 86
BB/KAL
CM9160-11M-5Y-6M-OY
Cumpas T88 A
PRL/VEE#6
CM64624-2Y-1M-4Y-OM-16Y-OM
Cumpas T88 B
PRL/VEE#6
CM64624-2Y-1M-4Y-OM-97Y-OM
DWR 39
HD(M)1508/S308
Diamante INTA
SN64/TZPP//NA160/3/NAR59
II21872-9T-2B-3T-2B
Dougga
KLPE/RAF//2*8156(R)
1I23997-4Y-100M-300Y
El Gaucho
SINVALOCHO MA//RICCO/LIN CALEL MA

88

Uniq. Abbr.

107

EMU

204

ERA

1215

ESDA

242
1144

EDAK
FAN1

1273

wm4578

1211

FSD85

1218

FALKE

222

FLK SA

1238

FLK SB

1239

FLK SC

323

FLK SD

1090
23
4
1191

G162
GB
GTO
GALVEZ

1186

GARA

1175

GEN

55
56
57
58
241
1181

GZ139
GZ144
GZ150
GZ155
GLE
GLEN

221
243

GLI74
i41593

1019
66
1095

HB501
HD1941
HD2172

1092
1094
1093

HD2177
HD2236
HD2278

1096

HD2281

1097
1098
1089

HD2329
HD2402
HI977

1085
1177

HP1209
HUW206

Common Hame
Selection History

Cross

Emu
TOB/NAPO//NO/ERA/3/BB/GLL
CM8327-9M-1Y-1M-1Y-1M-OY
Era
II-50-10/4/PEMBINA/II-52-329/3/II-5338/II-58-4//II-53-5EN
Minn 11-62-64
Esmeralda 86
BUC/BJY
CM49641-9Y-1M-1Y-5Y-OM
Estanzuela Dakaru
LEE/ND34
FAN #1
-33B-OY
FLN/ACC//ANA
FLN/ACC//ANA
SWM4578-56M-3Y-3M-2Y-OM
Faisalabad 85
MAYA/MON//KVZ/TRM
CM44083-N-3Y-1M-1Y-1M-1Y-OB
Falke
BUC/4/TZPP//IRN46/CN067/3/PRT
CM56744-7Y-2Y-1M-1Y-OM
Flicker "S"A
WE//CNO/NO/3/ZBZ
CM8954-B-7M-1Y-1M-OY
Flicker "S"B
WE//CNO/NO/3/ZBZ
CM8954-B-7M-1Y-1M-1Y-OM
Flicker "S"B ReSel
WE//CNO/NO/3/ZBZ
CM8954-B-7M-1Y-1M-1Y-OM-(1-18B)
Flicker "S"D
WE//CNO/NO/3/ZBZ
CM8954-B-7M-1Y-1M-1Y-OM-2PTZ-OY
G 162
GUIZHOU LARGE HEAD NO.7/YANGMAI NO.3
Gabo
BOBIN W39//BOBIN W39/GAZA
Gaboto
BAGE 2018//F44/SVL/3/BAGE 1971.37
Galvez 87
BB/GLL//CARP/3/PVN
CM33483-C-7M-1Y-OM-5B-OY
Gara
AU//KAL/BB/3/WOP
CM33203-K-9M-19Y-3M-3Y-OM
Genaro 81
KVZ/BUHO//KAL/BB
CM33027-F-12M-1Y-6M-OY
Giza 139
HINDI 90/KENYA B 256
Giza 144
REGENT/2*GIZA 139
Giza 150
MIDA/CADET
REGENT/2*GIZA 139//MIDA/CADET/
Giza 155
2*HINDI 62
Glenlea
PEMBINA*2/BAGE//CB100
Glennson 81
KVZ/BUHO//KAL/BB
CM33027-F-8M-1Y-8M-1Y-2M-OY
Gloria 74
H/2*RAF2
H/2*RAF2
II41593-1R-3M-1S-2M-OS
HB 501
E5477/SN64
HD 1941
HD2160/5/TOB/CNO//BB/2/NAI*2//
HD 2172
TT/SN64/3/HD1954
HD1962//E4870/K65/3/HD1593
HD 2177
HD 2236
HD2119/HD1981
HD2119/HD1912/HD1592//3/
HD 2278
HD1962//E4870/K65
HD2160/HD1912/HD1592//3/
HD 2281
HD1962//E4870/K65
HD2252/UP262
HD 2329
HD2267/HD2236
HD 2402
GLL/AUST II 61.157//CNO/N066/3/
HI 977
Y50E/3*KAL
HP 1209
E4871/PJ62
HUW 206
KVZ/BUHO//KAL/BB
CM33027-F-15M-4Y-4M-2Y-2M

89

Uniq. Abbr.

Common Kame
Selection History

1108
1084
1020
1099

HUW37
HUW55
HW135
c19343

1192

HAHN

1246

HAVIK

82

HZ1l2

79

HZ2152

370
364
218

HZ806
HZ695
HORK

108

HORK A

1237

HORK B

KAL/5331//H01982
HUW 37
HUW 55
E4870/H01982//INIA66/3/H02189
HW 135
HY5LOP/PVN
HYSLOP/PVN
CI019343 no selection history
CBC148/KA/7/HK/MOA38/4/4777/3/
Hahn
REI//Y/KT/5/YR/6/TUC
CM33682-L-IY-IY-4M4Y-I00B-501Y-500B-OY
Havik
VEE/H499.71A//4*JUP
CMH80A.383-1B-IY-IB-IY-IB-IY-OB
Hazera 112
21931/CH53//ANOE5/3/GA56/4/AN64
1I2985-5H-2H-IIH
Hazera 2152
YT54A 3*/NI0B
II8474-A-8T-12B
Hazera 806/1976
Hazera 895
Hork
HOP/RON//KAL
CM8874-K-IM-IY-IM-2Y-OM
Hork "5"A
HOP/RON//KAL
CM8874-K-IM-1Y-OM
Hork 5 Re5el
HOP/RON//KAL
CM8874-K-IM-IY-OM-(1-356Y)-(l-200B)
Huacamayo
YR//5N64/NT5207.85/3/CNO/7C//GTO
CM8671-B-IM-IY-IM-IY-IM-OY
Huelquen
MO/MCM//EXCHANGE
IA 78112
KVZ//CNO/PJ
IA 7846
KVZ/3/CN067/CHR//ON
IA 7873
CNO/GLL
lAC 24
IA551/4/5N64/Y50E//GTO/3/2*CNO
IA5 54
IA516/4/Y53/NI0B//Y50/3/KT54B
1I30565
CC/INIA66/3/CNO//ELGA/5N64
II30565-22M-OY-OT
1I35953
INIA66/NAR59/3/CIN/PJ62
1I35953-2R-5M-4R-OM
INIA/BB
INIA/BB
1I26478-74M-OY-4T-IT-OT
15WYNll Ell
SN64/TZPP/3/NAI60//ST464/BZA/4/
not in PMS
21931/3/CHA53/AN//GB55/5/AN64
ISWYN12 E20
ISWYN17 E8
BUHO/5/SN64/TZPP//Y50/3/NAPO
ISWYN24 E38
/4/LAC617(67A)
E-II-75-1935-1E9E-IE-IE-IE-OE
TG//BJ66/INIA66/3/15WRN74-141
ISWYN25 E23
not in PMS
INIA66/5/ELGA/SN64/4/
ISWYN25 E41
TG/3/5N64//TZPP/NAI60
not in PMS
ISWYN5 E31
L 1418-3463L 1231 x 23L1274-111 (L)
ISWYN5 E44
36896/2*CJ54//YT54 A(H)
4265/HD3/4/MO//K/Y/3/WIS/SUP F6
ISWYN5 E49
MARA//(SUPREMO-MENTANA-MCM) ?
ISWYN8 E13
not in PMS
LEE/NI0B//GB55/3/GB56
ISWYN9 E37
JUP/BJY
Iapar 17 Caete
CM39992-8M-7Y-OM
Iapar 18 Marumbi
PF72640/PF7326//PF7065/ALONORA
Fl1933-0-500M-5n3Y-502M-500Y-lG-OG
Iapar 6 Tapejara
UNKNOWN CROSS
Iassul 20
COLONIAS//FN/K58
Icta Sara 82
TAST/4/TP//CNO/NO/3/CNO/7C/5//JUP
CM3089-G-IY-4M-IY-3M-IY-OY
Iliniza
Imbabura

366

HUAC

40
1112
1113
1077
1101
226
165

HQN
i 78112
IA7846
IA7837
IAC24
IA554
i30565

109

i35953

169

i26478

81

l1I511

34
1064
1249

121520
17I58
241538

1130

251523

1104

251541

187
188
86
41

51531
51544
5I549
81513

1115
1202

91537
CAETE

1252

IA18

1134
32
1168

IA6
IA520
5ARA

1021
1022

ILNZ
IBB

CroBB

90

Uniq. Abbr.

CORlOn Hame
Selection History

1169

IMU

1122

IND66

Imuris T79
BY/MAYA/4/BB//HD832.5.5/0N/3/CNO/PJ
CM31678
Indus 66
PJ/GB55
II8156
Inia 66
LR64/SN64
II19008-83M-I00Y-I00M-I00Y-I00C
Itapua 25
PI/3/LR64//TZPP/KNTT2
III8790-1R-IT-2Y-IC
ltapua 30
JUP/ALD
CM36867-18Y-21M-3Y-OM-OE
JP12867/PL
J9128.67/PALOMA
CM4900-5Y-IM-3Y-OM
Jahuara 77
TZPP/PL//7C
CM5287-J-IY-2M-IY-OM
Jaral "S"(A)
SN64//TZPP/NAl60
II18889
Jaral "S"(B)
SN64//TZPP/NAl60
IlI8889-6T-4T-2T-IT-2B
Jaral "S"(C)
SN64//TZPP/NAl60
II18889-101M-IR-3C-IT-2B-OY
Jaral 66
SN64//TZPP/NAI60
lII8889-101M-IR-3C-4Y
Jinmai 4058
OFN//HUADONG#5/ABD
Junco
BB/GLL//CARP/3/PVN
CM33483-C-7M-IY-OM-20B-OY-2PTZ-OY
Jupateco 73
III2300//LR64/8156/3/NOR
II30842-31R-2M-2Y-OM
Juriti
lAC5/ALDAN
CM46961-16M-113PR-IT-OT
Justin
THATCHER/KENYA
FARMER//LEE/MlDA/3/CONLEY
CI13462
KI34/VEE
KI34(60)/VEE
PBTI0-IA-3A-OA
KLPE/RAF
KLPE/RAF
ClD6552 no selection history
Kalyan Bluebird
KAL/BB
ll26992-30M-IY-IM-3Y-OM
Kalyan Bluebird ReSel
KAL/BB
II26992-30M-IY-IM-3Y-OM(I-208B)
Kanchan
UP301/C306
Kauz
JUP/BJY//URES
CM67458-4Y-IM-3Y-IM-3Y-OB
Kauz ItS"
JUP/BJY//URES
CM67458-4Y-IM-3Y-IM-2Y-OB-3Y-OY
Kavko "5"
KVZ/3/CNO/CHR//ON
SE375-12S-3S-0S
Kenya 6106.3
Kenya Kifaru
Kenya Leopard
LAGEADlNHO/3*K354P//CII2632/3/3*K354P
1346 A.2 A.l
Kenya Nungu
WlS245/lI.50.17//Cl8154/2*
FR/3/2*TOB66
Kenya Paa
Kenya Tumbili
KTB/GIZAI55//NAD63/T238.1.5.8.17.10//KL ATL/TOB66//CFN/BB
Khyber 79
WREN//CNO/GLL/3/C271
Kitt
TH/2*SUPREZA/3/FN//K58/N/7/
PEMBINA/FN/5/TH/6/MD//KI17A/2/ ... ?

110

INIA66

1255

ITP25

1196

ITP30

115

m4900

301

JAH

17

JAR SA

JAR SB

JAR SC

112
1121
1190

JAR
JM4058
JUN

113

JUP

1212

JUR

205

IN

1265

PBTI0

c6552

123

KALBB

116

KALBBR

1136
1224

BAW28
KAUZ

1223

KAUZ S

1272

KVK

1023
1024
85

K61063
KKI
KLD

1138

KNN

1025
1110

KPAA

1139
1131

KYB79
KITT

KTUM

Cross

91

Uniq. Abbr.

COJllmOn BaJIl8
Selection History

Cross

1260

KLCB

RELEN/BAGE//KLPE/5/BB/8156/4/
12300//MASS 5/GTO/3/JAR

KLRE

Klein Chamaeo
KL B103-71-20Y8M-llOOYK ?
Klein Rendidor

60
1198

WM1353
KHN

1111
170

Ij2484
6IS25

118

i19008

1026
1137
80
1027

LU265
LV
115682
LKH757

1028
117

LAP286
LR64A

373
181
1261

LIE

1209

LIRA

1029
999
1030
182

LIZ
LOCCK
LMAI10
LU

1189

m33254

1164

m29251

1116
225
1031
1080
1033
1102
1034
1035
1213

MLKS11
MN7083
MN7086
MN7357
MN7529
MN7663
m82128
MRSIM
MCU

233
1036
1037
36
37
705
10

LMP

LINEA

MG41
MGP
MHS18
MLB
MITU
MITUIS
tell
MJI

223

MAYA

102

MAYA S

24

MDS

KLEIN COMETA/KLEIN33 AG//SINVALOCHO

MA/KL H33 AG

Kloka WM 1353
OREF1158/FDL//MFO/2*TIBA63/3/COC
Kohinoor
CM37987-I-1Y-5M-OY
LAJ 2484
KVZ//CNO/PJ
LR/3*P4160
LR/3*P4160
not in PMS
LR64/SN64
LR64/SN64
II19008-52M-6Y-7M-101C
LU 265
Lagoa Vermelha
VERANAPOLIS*2//MARROQUIT/NEWTHATCH
Lakhish Line #1568/2
Lakhish Line #757
not in PMS
Lap 286
Lerma Rojo 64A
YSO/N10N//L52/3/2*LR
II8724-8Y-1C-1Y
Liesbeck
CNO/INIAen
Limpopo
SN64//TZPP/NAI60/3/TOKWE
Linea E 1979
MSN-F6-79-3E-OE
Lira
KVZ/TRM/ /PTM/ANA
CM43903-H-4Y-1M-1Y-3M-3Y-OB
Liz
Local Check
Long Mai #10
Lundi
MARA//LEE/SELKIRK
S595-A1-A6-B2
MAI/PJ62//EMU
MAIPO/PJ62//EMU
CM33254-T-1M-1Y-6M-3Y-OM
MAYA/MON
MAYA74/MONCHO
CM29251-3M-17Y-4M-OY
MULTILINE OF KALANSONA
MLKS-11
MN 7083
MN 7086
CRIM/2*ERA//MN69235
MN 7357
MN 7529
MN6988/MN69169
MN 7663
MN 82128
MR SIM
IAS64/ALDAN
Macuta
CM47207-6M-103PR-2T-OT
LINEA G/SINVALOCHO MA//SINVALOCHO
Magnif 41
MA/MAGNIF MG
JAR/NAPO/3/LR64//TZPP/3*ANE/4/
Magpie
BB/NOR//CNO/7C/3/CAL
CM20668-D-4Y-4M-1Y-OY
Mahissa 18
Malabadi
TH*7/FN//CANTHATCH/3/TH*6/PI170925
Manitou
Manitou Insensitive
Mapache (tel)
Marcos Juarez Inta
SN64/KLRE
II19975-68Y-1J-6Y-1J-4Y-1J-OB
Maya 74
CNO/GLL
1127829
Maya 74"S"
CNO/GLL
II27829-19Y-1M-500Y-501M-OY
Mendos
EUREKA/CI 12362//2*GABO/3/MENTANA/
6*GABO/4/SPICA/KODA//GABO

92

Uniq. Abbr.

25
421

MENG
durum1

179

MEX

26

MX

78

MX226

42

MFN

1197

MNV

1148

MN082

1100
120

MTC
MO

121

MON

1284

MON SA

1233

MON SB

67
43
1118
1038
1039
1040
68
69
70

MOTI
c6753

1220

N0610
n51817
NL410
NL459
NP824
NP832
NP852
NP880
NP881
NR861

1274

NR86II

1041
1105
150

n14.13
n51.28
n54.17
NAC

232

NAC SA

151

NAC SB

122

NAI60

71
72

1072

1043
1285
224

NJG484
NJ2049
NAFN

49

NAP0

50

NAR

171
1204

NAYAB
NKT A

COllllDOn Mame
Selection History

Cross

Mengavi
Mexicali 75 (Ourum)
GOOVZ469/3/TO//61.130/LOS
CM470-1M-3Y-OM
Mexicano 1481
YT54/NI0B 126.1C
1I7064
Mexico 120
YT54/NI0B
1I7064
Mexico 226
SN64/KLRE
1I19975-68Y-IJ-6Y-3J-2Y
Mexifen
SN64//6*SKE/3*ANE
1II8903-17M-2R-3C
Minivet"S"
BB/CNO//INIA66/S0TY/3/SPRW/4/PVN
CM37705
Minuano 82
TRINTANI*2/SELKIRK FL33//LINHAGEM
B12598-0A-10A-IA-OA
Mitacore
lAS 50/JARAL
Mochis 73
SON64/KLRE//BB
1I26502-8Y-3M-1T-1M-5S-0M
Mancha
WE/GTO//KAL/BB
CM8288-A-3M-7Y-OM
Mancha S A
WE/GTO//KAL/BB
CM8288-A-3M-6Y-5M-1Y-1M-OY
Mancha S B
WE/GTO//KAL/BB
CM8288
Mati
YT54/NI0B//NP852
NAR*2IPJ
NAR*2/PJ
CI06753
NO 610
N0526/KITT//BTT
NK 7751817
NL 410
NL 459
NP 824
W245:44-25-7-5/NP770//C518/NPI65
NP 832
E145 (KENYA)/PISSI LOCAL
NP 852
E1915/NP761//NP761
NP 880
C281/NP790
NP 881
NR86-1
TTR/JUN
CM59123-3M-IY-2M-IY-2M-2Y-OM
NR86-I1
GOK16/0GA//AU/JTS179/3/VEE
SWM12779-BN-BK-6N
NS 14.13
NS 51.28
AU/7C
NS 54.17
JARAL/BEZOSTAJA 1
Naeozari 76
TZPP/PL//7C
CM5287-J-IY-2M-2Y-3M-OY
Nacozari S A
TZPP/PL//7C
CM5287-J-1Y-2M-IY-IM-OY
Nacozari S B
TZPP/PL//7C
CM5287-J-1Y-2M-IY-4M-OY
Nainari 60
SPO/MTA//GB-KEN/3/TH/Q//KENYA/
P4160-6H-3Y-2C
MTA/4/GB-KEN
Nanjing 4840
Nanjing 82049
PCH/4/2*KT54A/N10B//KT54B/3/NAR59
Naofen
T2494-14T-4T-IV
FR//FN/Y48/3/NAR
Napa 63
1I9314-22T-4B-IT
FCR/MCM//KT48/Y48
Narino 59
1I4777-2B-1B-I0T-IT-?
Nayab 70
Neelkant "S"A
HOI220/3*KAL//NAC
CM40454-11M-4Y-IM-IY-44M-IY-OB

93

Uniq. Abbr.
t

Common Rame
Selection History

1167

NKT B

1044
1119
1141
1045
1262

NG8201
NG8319
NG8331
NG8401
NKCH

Neelkant "S"B
HD1220/3*KAL//NAC
CM40454-1M-4Y-2M-3Y-OM
Ning 8201
NINGMAI N04/0LSEN//ALD/YANGMAI N03
Ning 8319
NINGMAI NO.4/0LESON//ALD/YANGMAI NO.3
Ning 8331
YANGMAI NO.4/NING 7840
Ning 8401
Norkin Churrinche
ERA/POLK//TOB
N7568-9A-1G-1G-OA-OG
Noroeste 66
LR64/SN64
1I19008-52M-4Y-4M-2Y
Norteno 67
LR64/SN64
1I19008-52M-6Y-3M-2Y
Novafen "s"
RULOFEN*2//FKN/N10B
CH10593-2P-3P-6P-1P
Novojoa "S" (tel)
MAYAll/ARM
X2802-38N-3M-6N-5M-OY
Oasis 86
AGATHA/3*YR
CMH77A.485-8B-5Y-1B-1Y-OB
Ollanta
Olmill
NORIN 721NORIN 12
Opata 85
BJY/JUP
CM40038-6M-4Y-2M-1Y-2M-1Y-OB
Orso
FUNO/PRODUTTORE
Oxley
PJ62/GB56//TZPP/NAI60/3/2*WW15
PAl 4
INIA66/CAL//INIA66/CC
CM28647-67Y-1M-1M-OY
PAT 73121
PF 70100
PF 8237
PFAU
HORK/YMH//KAL/BB
CM38212
PFAU "s"
HORK/YMH//KAL/BB
CM38212-1-7Y-2M-1Y-3M-2Y-OM
PJ62/GB55//NAI60
PJ62/GB55//NAI60
PK594-80A-1A-OA
PR 3
Pak 20
21931/3/CH53/AN//GB56/4/AN64
II20985-511-2II-1I1
Pakistan 5725
Pakistan 5747
Palmira 1
TH/STC//FR
1I5962-4T-2B-1T-2B
JAR/CHR//CC/JAR
Pampa Inta
H2139-1P-4B-2P-1P-OP
Papago 86
BUC/PVN
CM52359-2M-3Y-1Y-2M-1Y-OM
Parana 2
2. 71-74T-1T-1T-2T-OY
Parana 68/1116
SK/J9281. 67
Pat 19
B530-0J-52C-1C-OC
TZPP/SON64A//NAR59
Pato (B)
II21974
TZPP/SON64A//NAR59
Pato Argentino
1I21974-4R-4M-2B-OY-OP-OY
VCM//CNO/7C/3/KAL/BB
Pavon 76
CM8399-D-4M-3Y-1M-1Y-1M-OY
VCM//CNO/7C/3/KAL/BB
Pavon ReSel
CM8399-D-4M-3Y-1M-OY-(1-26B)-OY
VCM//CNO/7C/3/KAL/BB
Pavon Sl
CM8399-D-4M-3Y-1M-1Y-OM
VCM//CNO/7C/3/KAL/BB
Pavon S2
CM8399-D-4M-3Y-1M-OY

124

NO

125

NOR

44

NFN

1280

tel2

1245

OASIS

178
1120
1203

OLN
OML
OPATA

1088
22
1163

ORSO
OXY
PAI4

1046
1047
1048
1199

p73121
f70100
PF8237
PFAU

1200

PFAUA

176

PK594

1049
174

PR3
PAK20

172
173
51

Pk5725
Pk5747
PAL

1254

PPAI

1216

PGO

1147

PARA2

11
216

PAR68
PAT19

12

PATOB

13

PATO

327
1236

PVN
PVNRS

127

PVN1

328

PVN2

Cross

94

Uniq. Abbr.

329

PVN3

330

PVN4

1235

PVN5

128

PJ

1214

PDZ

371

PHO

14

1160

PTS
m21692

130

PI

372
91

PLNK
PLO

133

PTM

206

PRT

1266

PJB81

1050
1051
1052
363

QT4081
QT4083
QFG2
QZ75

1257

QUM

1231

RABE

1253

RTNI

1140
134

ROMI
RQ

180

S1103

212

JIT43

1053
1128

SA75
m40392

1270

s83125

1054
16

SD2968
c6781

177

6IS24

138

i18892

207

I84690

1082
73

SADO
SL

COJlllllOn Name
Selection History

Cross

Pavon 53
VCM//CNO/7C/3/KAL/BB
CM8399-D-4M-2Y-2M-3Y-1M-OY
Pavon 54
VCM//CNO/7C/3/KAL/BB
CM8399-D-4M-3Y-OM-OBK
VCM//CNO/7C/3/KAL/BB
Pavon 55
CM8399
Penjamo 62
FKN/N10B
II7078-1R-6M-IR-IM
Perdiz
IAS58/BJY//BNQ
CM47971-A-4M-I05PR-1T-OT
Phoebe
CAL/KVZ//TRM
CM30831-D-4Y-2M-1Y-OM
Piamontes INTA
TH//LA ESTANZUELA/RAF 6MA/3/FN
Pionero Inta
PV18A/CNO//JAR
CM21692-1B-4B-1B-3B-OB-?
pitic 62
YT54/NIOB
II7064-1Y-1H-1R-2M
Planinka
Pollo
BB/GLL/3/CHO/SN64//BB
II35129-26Y-2M-IY-1M-IY-OM
Potam 70
INIA66/NAP063
II22402-6M-4Y-IM-1Y-OM
Protor
TOB/CNO
II24908-30M-3Y-3M-OY
punjab 81
INIA/3/SN64/P4160(E)//SN64
PK6841-2A-2A-IA-OA
QT 4081
QT 4083
Qian Feng #2
Quetzal 75
II786-1X-1X-3X-OX-?
CNO//SN64/KLRE/3/8156
Quimori
II23584
GOV/AZ//MUS/3/R37/GHL121//
Rabe "s"
KAL/BB/4/ANI
CM79516-025Y-4M03Y-01M-2Y-OB
Retacon Inta
JAR/CHR//CN067
H2138-12P-1B-2P-IP-OP
Y50E/8156//KAL/3/TOB/CNO
Romi
Roque 73
SN64//6*SKE/3*ANE
II18883-6M-6R-I1C-1Y
JARAL/3/MARA//LEE/SK
S1103
S1103
S331/NOR67
S331/NOR
JIT-43-2L
NAI/CB151//S984
SA75
SAP/MON
SAP/MON
CM40392
5C 83125
SC83125-1H-IH
SD 2968
SN64//SKE/LR64A
SN//SKE/LR
CID6781 no selection history
SN64/C271
5N64/C271
not in PMS
SN64/KNTT A
SN64/KNTT A
II18892-2M-3Y-5M-2Y
SOTY/3*JUSTIN
SN64//TZPP/Y54/3/3*JUSTIN
II84690-2Y-6C-6Y-4C-2Y-1C
Sado
DGA/SN64
Safed Lerma
Y50//N10B/L52/3/LR64
IIl5444-118Y-2C

95

Uniq. Abbr.

52

229

Cross

Selection History
SAK68
SAP

1187

SRHD

38
1180

SK
SERI

1142

SHA4

1146

SHA5

316
302
136

COlllDen Naae

STA
STM
7C

1217

SIND83

1055

SIPA

197

SOL

74

SKA

361
360
1258

SOND
SGL
SKI

137

SN64

140

SPRW

1277

STAR

1056
1057
1109
142

11B .15
SUN278
SNLG
SX

1058
1145

SU220
SUZ1

1143

SUZ8

189
1124

T80l7

1125

T8020

SRX

196
59

T64.2W
i22964

1226

m20106

1268

575110

29
30
149

TR236
TR380
i19025

20
190

TAC
TAlC

BZA/2*AFM
Samaea 68
1I13022-9B-lT-2B-1T
INIA66/S0TY//CZHO
Sapsucker
BR69-1Y-3M-OY
AU//KAL/BB/3/WOP
Sarhad 82
CM33203-K-9M-24Y-OM
Selkirk
MCMURACHY/EXCHANGE//3*REDMAN
Seri 82
KVZ/BUHO//KAL/BB
CM33027-F-15M-500Y-OM-87B-OY
Shanghai #4
-14B-OY
Shanghai #5
-7B-OY
Shasta
INIA66/ANZA
Shortim
siete Cerros
PJ/GB55
1I8156-1M-2R-4M
Sindi 83
TZPP/PL/l7c
CM5287
Sipa INIA
SIS/PVN
CM30697-2M-14Y-OM
Soltane
SN64/KLRE
1I19975-68Y-1J-6Y-1J-3Y
Sonalika
1I54-388/AN/3/YT54/N10B//LR64
IIl8427-4R-1M
Sonderend
Song len
SN64/KLRE//BB
Sonka Inia
1I26502-8Y-6M-2Y-OM
Sonora 64
YT54/N10B//2*Y54
1I8469-2Y-6C-4C-2Y-1C
Sparrow
FN/MD//Kl17A/3/2*COFN/4/SN64/
CM2182-5M-1Y-2M-3Y-OM
KLRE/3/CNO//2*LR64/SN64
Star
LFN/SDY//PVN
SWM7215-2Y-2Y-OY-2Y-OY-41M-OY
Sun 11B-15
Sun 278
Sunelg
KITE*4//CS3D/AG#14
PJ/GB55
Super X
II8156-lM-2R-4M
Suweon 220
Suzhoe #1
-5B-OY
Suzhoe #8
-31B-OY
YT54A/N10B//3*LR64
Syrimex
ROLLO/MAGNIF/4/S0N/TZPP//
T 8017
NAI/3/MOYSTAD
ROLLO/MAGNIF/4/S0N/TZPP
T 8020
//NAI/3/MOYSTAD
K338/ECDH//KOUDIAT 17 KT Y
T-64-2-W
TOB/8156
TOB/8156
1I22964-3Y-5M-OY
TOB/CC//PTO/3/BB/GLL
TOB/CC//PATO/3/BB/GLL
CM7l-20106-10E-15E-4E-1E
TOK*3/S111LAI
TOK*3/S111LA1
S75110-2-3
IRN 59.111//3*GABO/CHARTER
TR 236
TR 380
TZPP/AN64
TZPP/AN64
1I19025-l00M-101Y-100C-2Y
Tacuari
MASSAUX#5/GABOTO
Taichung 31
SHOAWASE/SAITAMA

96

uniq. Abbr.

1165
143
1060
925

TAN
TI
TEJO
barley

III

TES

1250

THB

1251

THB S

1166

TSH

1195

TSH S

53

TIBA63

27

TG

307

TTM

145

TOB

147

TOL73

1276

TONI

45

TQFN

148

TRM

368

TOW

1061
28
1162
105

TRIG01
TDK
TRI
TRI A

1228

TUI

1248

TNG

1240

TURA

1241

TURB

1242

TURC

184

TRP

1206

TRT

1135
1063
1126
75

UP1109
UP201
UP262
UP301

76
219

UP310
JIT35

COllllllOn Hame
Cross
Selection History
Tanager
SIS/PVN
CM30697-2M-8Y-1M-1Y-1B-OY
Tanori 71
SN64/CNO//INIA66
1I25717-11Y-3M-1Y-OM
Tejo
Tequila
MINN126/CM67
CMB-72-189-11Y-3B-1Y-OB
Tesopaco 76
INIA66/S0TY//CZHO
BR69-1Y-3M-3Y-OM
Thornbird
IAS63/ALD//GV/LV
F11915-A-502M-1Y-3F-701Y-15F-OY
Thornbird "s"
IAS63/ALD//GV/LV
Fl1915-A-502M-1Y-3F-701Y-SF.700Y
Thrush
BJY/GJO/4/MAI/3/BB//TOB/CNO
CM34742-E-2M-SY-3M-500Y-100B-501Y-OM
Thrush "S"
BJY/GJO/4/MAI/3/BB//TOB/CNO
CM34742
Tiba 63
FN/3/T48/MAY054//MENKEMEN
II10668-4T-2B-1T
Timgalen
AGUILERA/KENIA//MAROQUI/SUPREMO
/3/GB/4/WINGLEN
Titmouse
PL/3/INIA66/CNO//CAL/4/BJY
CM30136-3Y-1Y-OM
Tobari 66
TZPP/SON64A
II19021-4M-3Y-102M-100Y-101C
Toluca F73
INIA66/NAPO//CNO
II28036-111M-1R-2M-1T-OM
Toniehi 81
CAR422/ANA
SWM4610-2Y-20M-OM
Toquifen "S"
90S/FN*2//4160/3/YT54/N10B/4/2*C14
CH7817-3P-4P-1P-2P-1P
Torim 73
BB/INIA
1I26591-1T-7M-OY-5SY-OM
Towhee
CLI//RQ/SOTY/3/SIS/PVN
CM34709-G-15M-5Y-OM
Trigo 1
Triple Dirk
URUGUAY 1084/NX DIRK 48
Trisa INIA
CNO/INIA66//BB
CM28339-17Y-4M-2Y-OM
CNO/INIA66//BB
Trisa INIA A
CM28339-17Y-1M-1Y-OM
Tui
HER/ SAP / /VEE
CM74849-2M-2Y-3M-2Y-OB-48M-OY
Tungurahua
AMZ/5/FR*2/FN//3*Y/3/2*4777
E-II-685512-7E-OE-6E-OY
/4/FR//MYS4*2/4777
Turaeo "S"A
CN079*2/PRL
CM90312-A-2B-12Y-1B-OY
Turaco "S"B
CN079*2/PRL
CM90312-A-2B-3Y-3B-OY
Turaeo "S"C
CN079*2/'PRL
CM90312-D-3B-8Y-6B-OY
Turpin 7
MY54/N10B//P4160
IIS715-7Y-1C
Tyrant
DGA/BJY
CM40610-25Y-4M-1Y-1M-1Y-OB
UP 1109
UP262/UP368
UP 201
S308/BAJIO 66
UP 262
LR64/SN64
UP 301
1I19008
KLPE/RAF//LR64/2*SN64
UP 310
UP301//SN64/PI62
UP3011/SN/PI62
JIT-35-2L

97

Uniq. Abbr.

Coaaon Raae
Selection History

1243

m90351

1174

URES

106S
1066
1067
1068
1069
1070
1172

V1130
V1287?
V5648
V79143
V79353
V878
VEE#4

303

VEE#5

URES*2/PRL
URES*21PRL
CM90315-A-2B-2Y-1B-OY
Ures 81
KVZ/BUHO//KAL/BB
CM33027-F-12M-1Y-4M-2Y-2M-OY
V 1130
V 1287.GII
V 5648
V 79143
V 79353
V 878
Veery #4
KVZ/BUHO//KAL/BB
CM33027-F-12M-1Y-10M-1Y-3M-1Y-OM
Veery #S
KVZ/BUHO//KAL/BB
CM33027-F-1SM-SOOY-OM
Veery #S"S"
KVZ/BUHO//KAL/BB
CM33027-F-1SM-SOOY-OM-110B-OY
Veery #7
KVZ/BUHO//KAL/BB
CM33027-F-15M-4Y-4M-3Y-2M-1Y-OM
Veery #8
KVZ/BUHO//KAL/BB
CM33027-F-12M-1Y-1M-1Y-1M-OY
Veery 7"S"
KVZ/BUHO//KAL/BB
CM33027-F-15M-4Y-4M-2Y-1M-1Y-OM
Vicam 71
INIA66/NAP063
II22398-39M-1R-OY-101M-OY
Victor I
FN//K58/N/3/II-SO-3S/4/3*MARA
Victoria Inta
JAR//BB/CNO
Cl191-7B-3B-1T-3P-OP
Vireo
INIA66/0N//INIA66/BB/3/COC
CM28235-2Y-6Y-OM
Vulture
YR RESEL(B)/TRF//RSK/TRM
CM3604-A-1M-5Y-OM
W84/11
K20/MENG/5/SN64//TZPP/NAI60/3/DROM/4/SKE
W84/14
NAD/LR//BB/3/IBWSN264
BE2S93-BU1-BUl-SEl-TM
CARTHAGE//Kal/BB
WL 2265
CM7806
WL 410
SON63/S326//KAL
WL 711
S309/CHRIS//KAL
West bred 911
Sln male-sterile facil. Rec. Sln.
CRIM/2*ERA//BUI/GALLO
Wheaton
Xelaju
FKN/GB56
II8325-4M-2R-2M
Y50E/3*KAL
Y50E/3*KAL
II3518-5M(F1)-31Y-OM-8M-OY
YW00890
ELGAU/SN64//7C/NAPOEN
YW00890-2S-11-12-21-11-2B
Yang Mai 6
NEWTHATCH/MARROQUI 588
Yaqui 50
II120-3C-(9-11C)-24C
JO/AA//FG
Yavaros (durum)
CM9799
Yecorata "S"
INIA66/CNO//CAL/3/BB#2 RESEL
II40041-5M-2R-2M-4S-0M
INIA66/CNO//CAL/3/BB#2 RESEL
Yecorato 77
II40041-9M-2R-6M-4S-0M
ZA75/ZP
ZA75/ZP
SA79014-9-11
Zaafrane
SN64/KLRE
II1997S-68Y-IJ-1Y-1J-5Y-1T
Zambezi
8156//LEE/ND74
SS9S-A1-A6-B2

1179

VEE#5S

1178

VEE#7

1173

VEE#8

1176

VEE#7S

155

VCM

83
1152

VI
VTI

365

VIREO

312

VUL

1107

W84/11

1117

W84/14

1230

WL2265

1129
1127
1278

WL410
WL711
WBD911

1081
61

WHT
XJ

158

i3518

203

yw0890

1071
156

YMI6
Y50

1106

durum4

239

YRT S

339

YRT

1269

ZA/ZP

IS

ZAF

183

ZBZ

Cross

98

Uniq. Abbr.

220

ZA75

119

ZA75 A

376

ZA75 B

1247

ZAGU

54

ZIPA

COllllDOn Kame
Selection History

Cross

MENG/8156
Zaragoza 75
1I22364-1Y-6C-1Y-1C-4Y-3C-2R-100Y
MENG/8156
Zaragoza 75A
1I22364-1Y-6C-1Y-1C-4Y-3C-1Y-2B-300Y-OY
MENG/8156
Zaragoza 75B
1I22364-1Y-1C-1Y-IC-4Y-3C-IOOY
Zaraguro "S"
E-II-3958-2E-OE-OE-9E
Zipa 68
F/Y48//2*AFM
1I12924-20B-3T-2B-IT

99

CIMMYT Wheat Special Report Series

(As of April 1, 1992)
No.1. Russian Wheat Aphid Research at CIMMYT, 27 pages.
No.2. Wheat and Wheat Breeding in China, 14 pages.
No.3. Impact of Crop Management Research in Bangladesh, 15 pages.
No.4. Wheat Cultivar Abbreviations (In Press).
No.5. A Guide to the CIMMYT Bread Wheat Breeding Program, 43 pages.
No.6. Wheat Production and Grower Practices in the Yaqui Valley,
Sonora, Mexico, 75 pages.
No.7. Kamal Bunt Research at CIMMYT (In Press).
No.8. Management and Use of International Trial Data for Improving
Breeding Efficiency, 100 pages.
No.9. Durum Wheats: Challenges and Opportunities (In Press).

100

CIO
N l

Apart~do

L DE MEJOAAMIENTO DE MAIZ

TRIGO

Z
AND WHEAT IMPROV ME T CENTER
Postal 6 6 1 06600 Mil ico. D. F.
e Ie