Screening of new strains of sugarcane using augmented block designs

Mathematical Theory and Modeling www.iiste.org
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.4, No.8, 2014
153
Screening of New strains of sugarcane using
Augmented Block Designs
Otulo Wandera Cyrilus 1
Ojung’a Okoth Samson2
and Otumba Edgar Ouko3
1
Rongo University College P.O Box 103-40405 Rongo,Kenya
2
Ministry of Devolution and Planning P.O Box 115, Pap Onditi, Kisumu, Kenya
3
School of Mathematics, Applied Statistics and Actuarial science Maseno University
P. O. Box 333-40105, Maseno, Kenya
mcojijosamsa@gmail.com
Abstract
Sugar production has over time experienced a number of challenges, that is, the choice of the variety to plant;
soil nutrients variation and market competition amongst others have greatly affected sugar production. This
study has effectively and efficiently employed the technique of experimental designs to ascertain family
selection by comparing Augmented Block designs and Randomized Complete block designs. The augmented
block design is widely used in breeding programs, particularly in screening and selection of large number of
germplasm lines with non-replicated test treatments and replicated control treatments to estimate the
experimental errors. The study establishes a relationship between augmented block designs in screening and
completely randomized block design in screening new strains of Sugarcane. In the two designs analyzed, we
consider 5 test treatments and 2 control treatments for augment design and the same number of treatments for
Randomized Complete Block Design. In the event of screening new sugarcane varieties, attempts have been
made to find the effectiveness of augmented block designs and completely randomized block designs in test
families and control checks where the results reveal that Augmented Block Design is 11.86 times more efficient
than a Randomised Block Design. In the conclusion of this study we have shown that Augmented Block Design
is better suited when the plots are limited and Randomized Complete Block Design is better suited when
treatments are many.
Keywords: Augmented Block Design, Balanced incomplete block design, Randomized complete Block Design,
Latin Square, Unreduced BIB Design.
1. Introduction
Statistics is a science of extracting information from a complex and noisy data with uncertainty while applied
statistical methods help in analyzing data to serve specific purpose in application. This study has effectively
analyzed new sugarcane varieties using Augmented Block Design and Randomized Complete Block Design in
test families and check controls by the strength of their effectiveness. Most of research investigation involving
design of experiments helps make meaningful comparisons [1]. In plant variety selection programmes where
comparisons is between newly introduced strains and the commercial controls, the vital designs used are
augmented block designs, completely randomized block designs, reinforced balanced incomplete designs, Latin
squares and Gaeco Latin squares.
In this study we have used augmented block designs and randomized complete block design to screen new
sugarcane strains by concentrating on their effectiveness and efficiencies in tests verses control experiments as
applied to plant breeding of the new cane varieties. Farmers have several alternatives of varieties for field trials.
The tested entries are related to their pedigrees of yield for performance since families selection approach is
preferred due to its objectivity and useful traits of low hereditary as opposed mass selection. The analysis of
different designs used in screening the newly introduced varieties of sugarcane thus suitable for family selection
with replications have their relative efficiencies discussed into details with particular emphasis laid to an
augmented block designs and randomized complete block design.
2 Augmented Block Designs and Randomized Complete Block Designs
This study analyzes different designs seeking to give a statistical approach to farmers choice of varieties by

Vol.4, No.8, 2014
154
screening the newly introduced varieties of sugarcane to eliminate poor selection of varieties through augmented
block design and randomized complete block design.
2.1 Basic Concepts and Notation
i. Design
In this study, we take the view common among statisticians that a design is an allocation Θ of a set v
treatments to a set of Ω plots. The design may be thought of as a function f from Ω to Θ: - plot ω receives
treatments, and if f(ω) =λ is partitioned into blocks then f is called a block design. A block is proper if all its
blocks have the same size k. If k = v then we have a complete block design, otherwise if k < v it’s an
incomplete block design.
ii. Balanced incomplete block design
This is an arrangement of v treatment in b blocks such that any block has k treatments, any treatment occurs
in r blocks and any pair of treatments occurring in blocks. The five parameters are not independent but
satisfy these two relations;
vr = bk and λ (v -1) = r(k-1).
A Balanced Incomplete Block Design is then written as (v, k, and b) with
b = v(v- 1) λ
k(k-1)
And
λ (v-1)
r =
k-1
With respect to this study, we define a design in general as an allocation of a set Θ of v treatments to a set Ω
of b blocks thus the design may be thought of as a function of f from Ω on to Θ.
Plot ω receives a treatment if f(ω) = λ and is partitioned into blocks, then f is called a block design with the
size k for blocks.
If k = v then we have a balanced incomplete block design.
iii. Randomization
This is a method of dealing with nuisance factor by balancing the nuisance factor across the experiment. It is
done using random device or pseudo-random device to choose which factor combination is allocated to each
experimental unit. Like blocking, randomization is another technique of dealing with nuisance factors only that
for randomization the factors are unknown and uncontrollable. Majumda [2] in his handbook of statistics stated
that randomization balances out the impact of nuisance factors across the experiment.
iv. Replication
This is a convenient way of increasing the size of the experiment and precision it implies making more than one
measurement at the same combination factor level. It is necessary to let all sources of random variables operate

Vol.4, No.8, 2014
155
independently on every occasion so that replication can count as proper [3]. According Rao [4] in his book stated
that replication can be exploited to produce a pure error estimate of σ which is employed in this study.
v. Latin squares
A Latin square of order s is an arrangement of s symbols in an s x s array such that each symbol occurs once in
each row and once in each column of the array i.e. for a 4x4 Latin square in symbols ABCD would be
ABCD ABCD ABCD
BADC CDAB DCBA
CDAB DCBA BADC
DCBA BADC CDAB
These 4x4 Latin squares are mutually orthogonal since they are pair-wise orthogonal. A complete set of S-1
mutually orthogonal Latin squares exists for any S = pn
where p is a prime number [5]. This study uses Latin
square method of constructing an augmented block design with given parameters.
vi. Unreduced B.I.B design
According to Mike Jacroux [6] these designs are obtained by taking all combinations of the v treatments k at a
time with parameters given by
(v; k; b) =v
ck (2.1)
v =v-1
ckr (2.2)
And
λ=v-2
cK 2 (2.3)
vii. Reinforced incomplete block design
According to Das [7], if a number of control treatment p is added in every block of existing incomplete block
design with the resulting design having p + v treatment distributed in b blocks each of size k + p, such that
newly introduced p treatments are replicated b times and the original v treatments r times, this design is now
called reinforced incomplete block design.
viii. Augmented block design
An augmented experimental design is any standard experimental design in standard treatments to which
additional (new) treatments have been added. The additional treatments requires enlargement of the complete
blocks and incomplete block row column designs. Augmented designs that eliminates heterogeneity in one
directions are called augmented block designs [8].
ix. Randomized complete block design
This is the simplest type of layout where in this study treatments are allotted to the experimental unit at random.
This design essentially removes variability between blocks from the experimental errors. This was shown by
Rajender and Gupta [9] in their abstract of augmented design and randomised complete block design with a two
way elimination of heterogeneity.
3 Model building
A statistical model is basically an assumption relating effects of different levels of factors involved in an
experiment alongside one or more terms representing the error effects. To provide more insight understanding of

Vol.4, No.8, 2014
156
the design, this study develops a two factor experiment in screening various sugarcane varieties; we therefore use
the analogy of any agricultural experiment and refer to two-way classification as treatments and blocks as
proposed by Das and Gupta in 1992 under model of communication in statistics [7].
In this study, we consider v treatments and b blocks and one experimental value which are the yields of
sugarcane per acre corresponding to each treatment (plots) and blocks (variety) for ith
treatment and jth
block
denoted by Xij as in the analysis of variance.
3.1 Analysis of Variance
The ANOVA is a powerful statistical tool for test of significance. Since t-distribution is not adequate, ANOVA
which is based on f-distribution is employed with a basic purpose to test homogeneity of several means. ANOVA
is mainly used to dealing with analysis of agronomical data where variation is inherent and may be caused by
either chance causes or assignable cause.
3.2 Two Way Classification
The values of response in this classification are affected by two factors. In this study, we have the yield of
sugarcane varying for different treatments coded from one up to five and two others are used as controls like
rations as well as the differences in variety. Let us now suppose that n breeds of sugarcane are divided into h
different plots with each plot containing k varieties, if we consider the effect of k treatments, that is, rations given
at random to cane variation in each plot on the yield of sugarcane.
Let Xij= (yield of sugarcane from plot with jth
variety on ration i) i=1, 2,
3.......k,j=1,2,3.........h
We let the yield of sugarcane to be expressed as variable value in k X h two way table as below.
Treatment Varieties of Sugarcane Raw Raw
(Rations) 1 2....j....h Total Means
1 X
11 X
12 ....X1j....X1h R1 X1.
2 X
21 X
22 ....X2j....X2h R2 X2.
. . . ...... . .
. . . ...... . .
. . . ......
.
. .
i Xi1 Xi2....Xij....Xih
R
i Xi.
. ........ . .
. ........ . .
K Xk1 Xk2....Xkj....Xkh Rk Xk.
column total C1 C2.....Cj....Ch
column means X:1 X:2.....X:j....X:h
Table 1: ANOVA Table for General model

Vol.4, No.8, 2014
157
From the table 1 above
as the sum of observations in the ith
row.
i=1,2….k, as the sum of observations in the jth
column.
j=1,2,….h . = mean yield of the ith
treatment
mean yield of the ith
sugarcane
Overall mean
3.3 The Mathematical Model and its Assumptions
Since we had only one observation per cell the mathematical model for the means is
Xij = µij + ԑij; i = 1; …..k; j = 1; 2……………h (3.1)
Where Xij is the yield from sugarcane of ith
variety and the random variables having normal distribution with
means µij and common but unknown variance σ2
i.e. N ( µij,σ 2
)
The various effects are assumed additive and becomes;
(a) The general mean effect (3.2)
(b) The effect i due to the ith
plot is given by i = i = assuming that
(3.3)
(c) The effect j due to jth
cane variety is given by j = j- , assuming
(3.4)
(d) Interaction effect ( )ij when the ith
plot and jth
variety of cane occurs simultaneously is given by:
)ij = ij - i - j + (3.5)
Taking into account one observation per cell the assumptions that
i=0 (3.6)
and

Vol.4, No.8, 2014
158
Therefore the mathematical model for the effects used in this study is
Xij = + i + j + ij
3.4 Randomized Complete Block Design
This is the simplest type of layout in which treatment are allotted to the experimental unit at random. The design
in this study has considered one observation per treatment in each block, the order in which treatments run in
each block is done randomly. The model we assumed in this study for RCBD is of the form;
Xij = + i + j + ij (3.8)
with estimates
,
These estimates of , and j help in minimizing error.
Since the study considers treatments and block effects as deviations from the overall mean so that
and
.
The means for the model then appears as in the following table; Treatment effects are relative
Blocks
Treatment 1 2 3 .... B Means
1 11 12 13 ... 1b X
1:
2 21 22 23 ... 2b X
2:
. . . . . . .
. . . . . . .
. . . . . . .
t t1 t2 t3 ... tb X
t:
Mean X
:1 X
:2 X
:3 ... X
:b X
::
Table 2: The means for the model

Vol.4, No.8, 2014
159
Such that the null hypothesis
Ho
T
: No treatment effects: t1 = t2 = …..tb =0 or otherwise
Ho
B
: No blocks effects: b1 = b2 = ……bb = 0 or otherwise
(usually not of interest but assessed to determine if blocking was successful in reducing the variability in the
experimental units)
variability in the experimental units)
Source of variation SS df MSS F-Ration
Treatment SST v-1 S1=
Due to blocks SSB (b-1) S2 1
Due to Error SSE (v-1)(b-1) S2 =
SSE = S3
(v 1)(b 1)
Total TSS (bv-1)
Table 3: The ANOVA Table for RCBD
3.5 Augmented Block Design (ABD)
ABD is any experimental design in standard treatment to which additional (new) treatment have been added.
Augmented block design were introduced by Federer in 1956 [11] as an alternative to the systematically
arranged checks and new treatments. ABDs provides several advantages in screening new treatments such as
Genotypes, Insecticides and Drugs.
We consider an experimental design where ω tests treatment are to be compared with control treatment using
n experimental units arranged in r blocks such that jth
block is of size kj > , for augmented block design each of
the control treatment is replicated b times and occurs once in every block, while test treatment occurs only once
in one block. The analysis of variance generated from ABD with v = + ω treatments comprising of ω tests
and controls arranged in b blocks having k1 plots in block 1, k2 plots in block 2,....up to kb plots in block b
such that k1 + k2 + ……+ kb = n will appear as in the table below;

Vol.4, No.8, 2014
160
Source of variation SS Df MSS F-Ratio
Blocks eliminating treatment b-1 ASSB M_SSB = S1
S1/S4
Treatment eliminating blocks v-1 ASST
Among tests w-1 SST M_SST = S2 S2/S4
Among control u-1 SSC M_SSC = S3
S3/S4
Error n-v-b+1 SSE MSE = S4
Total n-1 TSS
Table 4: The ANOVA Table for Augmented Block Design with v = + ω
4 Discussions of Results
4.1 ABD and RCBD
The two block designs are similar in the sense that in Randomized block design the control treatments are added
to the design containing the test treatments whereas in the Augmented block design the test treatments are added
to a design containing the control treatments. The outcomes produced by the two methods were insignificantly
different.
Augmented Block Design Scenario
Treatments RCBD ABD N(0,1) ABD N(0,5) ABD N(0,25)
Treatments1 86.22 80.32 80.71 80.45
Treatments2 57.34 62.77 59.56 62.46
Treatments3 49.32 51.82 49.74 51.46
Treatments4 98.03 99.77 97.63 100.81
Treatments5 31.17 30.37 28.96 25.63
N39 59.56 61.25 62.66 64.29
CO945 60.68 59.81 60.31 59.30
Table 5: Comparison of treatment means generated by both ABD and RCBD
RCBD design produced higher figures for treatment1 though it is still within the treatment standard deviation.

Vol.4, No.8, 2014
161
4.2 Conclusion and Recommendations
In agricultural experiments, setting a block will always have some degree of intra block variation. The major
factors affecting the variation are the number of plots per block. The more the number of plots or experimental
units in a block, the higher the degree of intra blocks variations and vice-versa. A randomized complete block
design is discouraged by this study when evaluating large number of treatments instead augmented block designs
are encouraged due to the capability of having few plots per block. The study has revealed that in any
randomized complete block designs the number of plots in every block is dictated by the number of treatments
being evaluated. In conclusion, the study has revealed that augmented block designs are more efficient than
randomized complete block design for a finite error variance. In most cases under agricultural experiments, error
variances are assumes to be finite as it has been shown in the study.
Finally, we say that augmented block designs are more efficient than randomized complete block design in test
treatments versus control treatment experiments hence more valuable in screening new stains of sugarcane. The
efficiency of augmented block design when it is tailored to fit strongly in large inter block variation and scarce
test treatment. This study recommends the use of augmented block design in screening new strains of sugarcane
due to it’s high efficiency as opposed to the randomized complete block designs which only appear to be simple.
References
[1] Gupta V.K, and Parsad R. and Singh, V. P (1996),”Trace of optimal block designs; unequal block sizes for
comparing two disjoint sets of treatments”,Sankhya, Series B (8), pp 414-417. 1
[2] Majumdar, D. (1996);” Optimal and efficient treatment control designs”,Handbook of statistics, 13, S. Gosh
and C. R. Rao, eds., Elsevier Science, Amsterdam, pp 1007-1053. 2
[3] Jacroux, M. (1984);” Optimality and usage of block designs in in comparing test treatments with standard
treatments”, Journal of Royal Statistics Society, Series B (Methodological),(10) pp 315-322. 3
[4] Rao, C. R (1961);”Study of designs with replications”, Sankhya, (10), pp 117-127. 4
[5] Fisher R.A (1940); ”An explanation of possible different solutions of a problem in incomplete blocks”,
Annals of Eugenics,(5), pp 353-360. 5
[6] Jacroux, M. (2000;” On the determination and Construction of A-optimal and MV- optimal block designs
for comparing a set of treatments to asset of standard treatments”, Sankhya, Series B (9), pp 276-289. 6
[7] Das A, Gupta V.k (1992); “Estimating optimal block designs under Heteroscedastic Models”.
Communications in statistics, Theory and methods,(7), pp 1651- 1666 7
[8] Parsad .R. and Gupta V.K (2000); ”A note on Augmented block designs”,Indian Journal of Plant Genetics
Resources,pp 53-58. 8
[9] Rajender, P. and Gupta (2004);” Computer Aided construction and analysis of Augmented designs”,Journal
of Indian Social Statistics (special volume ), (9),pp 4-25. 9

The IISTE is a pioneer in the Open-Access hosting service and academic event
management. The aim of the firm is Accelerating Global Knowledge Sharing.
More information about the firm can be found on the homepage:
http://www.iiste.org
CALL FOR JOURNAL PAPERS
There are more than 30 peer-reviewed academic journals hosted under the hosting
platform.
Prospective authors of journals can find the submission instruction on the
following page: http://www.iiste.org/journals/ All the journals articles are available
online to the readers all over the world without financial, legal, or technical barriers
other than those inseparable from gaining access to the internet itself. Paper version
of the journals is also available upon request of readers and authors.
MORE RESOURCES
Book publication information: http://www.iiste.org/book/
IISTE Knowledge Sharing Partners
EBSCO, Index Copernicus, Ulrich's Periodicals Directory, JournalTOCS, PKP Open
Archives Harvester, Bielefeld Academic Search Engine, Elektronische
Zeitschriftenbibliothek EZB, Open J-Gate, OCLC WorldCat, Universe Digtial
Library , NewJour, Google Scholar

Business, Economics, Finance and Management Journals PAPER SUBMISSION EMAIL
European Journal of Business and Management EJBM@iiste.org
Research Journal of Finance and Accounting RJFA@iiste.org
Journal of Economics and Sustainable Development JESD@iiste.org
Information and Knowledge Management IKM@iiste.org
Journal of Developing Country Studies DCS@iiste.org
Industrial Engineering Letters IEL@iiste.org
Physical Sciences, Mathematics and Chemistry Journals PAPER SUBMISSION EMAIL
Journal of Natural Sciences Research JNSR@iiste.org
Journal of Chemistry and Materials Research CMR@iiste.org
Journal of Mathematical Theory and Modeling MTM@iiste.org
Advances in Physics Theories and Applications APTA@iiste.org
Chemical and Process Engineering Research CPER@iiste.org
Engineering, Technology and Systems Journals PAPER SUBMISSION EMAIL
Computer Engineering and Intelligent Systems CEIS@iiste.org
Innovative Systems Design and Engineering ISDE@iiste.org
Journal of Energy Technologies and Policy JETP@iiste.org
Information and Knowledge Management IKM@iiste.org
Journal of Control Theory and Informatics CTI@iiste.org
Journal of Information Engineering and Applications JIEA@iiste.org
Industrial Engineering Letters IEL@iiste.org
Journal of Network and Complex Systems NCS@iiste.org
Environment, Civil, Materials Sciences Journals PAPER SUBMISSION EMAIL
Journal of Environment and Earth Science JEES@iiste.org
Journal of Civil and Environmental Research CER@iiste.org
Life Science, Food and Medical Sciences PAPER SUBMISSION EMAIL
Advances in Life Science and Technology ALST@iiste.org
Journal of Biology, Agriculture and Healthcare JBAH@iiste.org
Journal of Food Science and Quality Management FSQM@iiste.org
Journal of Chemistry and Materials Research CMR@iiste.org
Education, and other Social Sciences PAPER SUBMISSION EMAIL
Journal of Education and Practice JEP@iiste.org
Journal of Law, Policy and Globalization JLPG@iiste.org
Journal of New Media and Mass Communication NMMC@iiste.org
Journal of Energy Technologies and Policy JETP@iiste.org
Historical Research Letter HRL@iiste.org
Public Policy and Administration Research PPAR@iiste.org
International Affairs and Global Strategy IAGS@iiste.org
Research on Humanities and Social Sciences RHSS@iiste.org
Journal of Developing Country Studies DCS@iiste.org
Journal of Arts and Design Studies ADS@iiste.org

Screening of new strains of sugarcane using augmented block designs

More Related Content

What's hot (20)

Similar to Screening of new strains of sugarcane using augmented block designs (20)

More from Alexander Decker (20)

Screening of new strains of sugarcane using augmented block designs