=Paper=
{{Paper
|id=Vol-1152/paper2
|storemode=property
|title=A Survey of the Applications of Main Biometrical Methodologies and Relative Informatics Tools Applied In Agricultural Research
|pdfUrl=https://ceur-ws.org/Vol-1152/paper2.pdf
|volume=Vol-1152
|dblpUrl=https://dblp.org/rec/conf/haicta/GousiosT11
}}
==A Survey of the Applications of Main Biometrical Methodologies and Relative Informatics Tools Applied In Agricultural Research==
https://ceur-ws.org/Vol-1152/paper2.pdf
A Survey of the Applications of main Biometrical
methodologies and relative Informatics tools
applied in agricultural research.
Vissarion Gousios1, StergiosTzortzios2
1
University of Thessaly, School of Agricultural Sciences, Laboratory of Biometry, Fytokou
Street - Nea Ionia, Volos, Greece, email: vgousios@uth.gr
2
University of Thessaly, School of Agricultural Sciences, Laboratory of Biometry, Fytokou
Street - Nea Ionia, Volos, Greece, email: stzortz@uth.gr
Abstract: Today, that the collection of biological data has been increasing at
explosive rate, the right processing of these data is something more than a
necessity. Fisher’s work (1925) in conjunction with the vast increases in
computer power that were implemented after the 1960s, have made possible
much more efficient and exciting methods of data analysis (Curnow, 1984),
thus opening the “bag of Aeolus” for the application of modern statistical
techniques in agricultural experimentation and research. This study mainly
seeks to survey biometrical methodologies that have been applied in the
agricultural research, emphasizing in the segregation of the agriculture in six
separate scientific fields.
Keywords: Biometrical methodologies, Agriculture, Research, Applications
1. Introduction
Many scientists have indicated the importance of statistical research in
agriculture (Fisher 1925; Box 1976; Finney 1978; Riley 1998; Wilson 1999).
Biometry moves towards this direction as it is implied from its various definitions: a
branch of biology that studies biological phenomena and observations by means of
statistical analysis (http://www.wordnetweb.princeton.edu/perl/ webwn), the
measurement of quantities in the living world (Source: Oxford Dictionary of
Statistics), the application of statistical methods to the analysis of continuous
variation in biological systems (http://www.expertglossary.com/ science) and others.
Some interesting concepts for the work that has been done in the field of
Biometry come from several authors. According to Box (1976) “Fisher's work
gradually made clear that the statistician's job did not begin when all the work was
over -it began long before it was started”. Preece (1984), Sadasivan (1982) and
Fielding (1990) summarized the poor quality of data collection and trial design
together with the ritualistic approach to data analysis, as main disadvantages in
biological research. Camacho and Carbonell (1993) reviewed the quality of statistics
in agricultural research in Costa Rica and concluded that there were weak links
____________________________________________
Copyright ©by the paper’s authors. Copying permitted only for private and academic purposes.
In: M. Salampasis, A. Matopoulos (eds.): Proceedings of the International Conference on Information
and Communication Technologies
for Sustainable Agri-production and Environment (HAICTA 2011), Skiathos, 8-11 September, 2011.
13
between agriculturalists and researchers. Riley and Onwueme (1996) concluded in
their work that level of biometric knowledge of Agricultural researchers in ACP
countries is poor. Onwueme et al. (1996) suggested that lack of access to computer
and statistical software have seriously hampered the teaching, learning and using of
biometry in developing countries. Akoroda (1996) believes that the poor and
inefficient use of biometrical methods in applied agricultural research can be
attributed to cultural background, school schedules, teachers orientation limited
resources a. o.
Towards this direction, a brief survey in the papers referring to agriculture in
the past conferences of the International Biometric Society (I.B.S.) has shown that
the rate of these papers is about 8% of the total papers submitted (Gousios and
Tzortzios, 2011).
The aim of this paper is to make a survey of the biometrical methodologies
and relative informatics tools applied in agricultural research as well as the
classification of these methodologies in different agricultural fields and the detection
of possible trends in their application. An extended survey of the biometrical
methodologies could lead to the best exploitation of the biological material.
2. Methodology
The general methodology followed for the implementation of this study is
consisted of four parts.
Initially, a personal survey was made in well known electronic scientific
databases (Elsevier, Wiley and Science Direct) for papers referred to applications of
biometrical methodologies in agriculture. To this purpose, we search for terms as
“statistical/biometric methodologies”, “statistical/biometric applications” and
“biometrics” which were included in the abstract, title or keyword of the papers. We
also restricted the research in the subject “agricultural and biological sciences” for
all the years that it was possible.
After this, we limited the results in journals relative with agricultural fields of
interest (Biotechnology, Crop Protection, Crop Production, Animal Production,
Food Science and Technology, Natural Resources Management). Totally, a number
of about 700 papers including the above criteria were reviewed to comprise the
backbone of this study. Furthermore, a classification of the surveying biometrical
methodologies has been made according to their applications in the above six
different agricultural fields.
Meanwhile, a questionnaire was constructed to help the survey of biometrical
methodologies and relative informatics tools applied in agriculture. The
questionnaire was addressed to researchers and scientists of Greek Agricultural
Institutes and Research Centers. Each one of the interviewed scientists belongs to
one of the six different agricultural fields, thus the results of the questionnaires were
classified per field.
Finally, an evaluation of the biometrical methodologies has been made,
according to the total number of their applications in the different fields of
agriculture and a comparison of the results between the bibliographical research and
14
the answers in the questionnaires.
3. Results and Discussion
In the Biotechnology field the survey showed that twelve methodologies
have been applied in the past (fig. 1). The most commonly used ones are:
Cluster Analysis with applications such as in the determination of groupings
in the genotypes or in the environments (Mungomery et al. 1974; Byth et al. 1976;
Lin and Thompson 1975; Lin 1982; Fox and Rosielle 1982; Ramey and Rosielle
1983 and Ghaderi et al. 1982) and the classification of locations on the basis of a
single trait for screening superior breeding lines (Campbell and Lafever 1980;
Ghaderi et al. 1980; Fox and Rosielle 1982; Baenziger et al. 1985; Yau et al. 1991;
Collaku 1991; Van Oosterom et al. 1993; Abdalla et al. 1996; Trethowan et al.
2001). Regression analysis with applications such as in the explanation of
Genotype x Environment interactions (Cruz et al. 1989; Eberhart and Russel 1966;
Finlay and Wilkinson 1963; Freeman and Perkins 1971; Perkins and Jinks 1968;
Shukla 1972; Silva 1995, 1998; Silva and Barreto 1985; Toler 1990; Toler and
Burrows 1998; Verman et al. 1978) and the examination of yield stability of various
genotypes (Finlay and Wilkinson 1963; Eberhart and Russell 1966). Discriminant
Analysis also shows a wide range of applications in this field such as in the
identification of cultivars with specific characteristics (Ebdon et al. 1998), in the
determination of genetic variation in populations (Nelson 2001; McElroy et al. 2002;
Ilarslan et al. 2002), in the assessment of the impact of a transgenic crop on soil
environment (Park et al. 2006) and the identification of appropriate parents and top
crosses in corn for future breeding and genetics program (Aydin et al. 2007).
Number of different
15
applications
10
5
0
Biometrical Methodologies
Fig 1. Number of different applications of Biometrical methodologies in Biotechnology.
In the Crop Protection field nineteen methodologies have been applied (Fig.
2), while in this field Analysis of Variance methodology appears many applications,
15
which is rather a consequence of the nature of the specific field. More specifically,
the most commonly applied methodologies in the Crop Protection field are:
Analysis of Variance with applications in the evaluation of different
insecticides/fungicides against pests/fungi in field crops (Bi et al. 2002; Obanor et
al. 2008), field evaluation of the resistance of different genotypes to parasitic weeds
(Abbes et al. 2007) and study of the influence of water and fertilization on biology
of insects / fungi in field crops and greenhouses (Archer et al. 1995; Jauset et al.
2000; Simoglou et al. 2006). Regression Analysis with applications such as in the
study of the effect of pesticides application in crop sequences on associated weed
composition, richness and diversity over five years (Puricelli, 2005), study of the
correlation of insect infestation in cultivars with abiotic and biotic factors
(Vayssieres et al. 2009), determination of the critical dates of cumulative catches of
Andean potato weevils and carabids in two potato rotation systems (Kroschel et al.
2009). General Linear Models with applications in the estimation of the dynamic
future populations of Diaprepes Abbreviatus in terms of time and space (Li et al.
2007), study of the performance of wheat cultivars and cultivar mixtures in the
presence of Cephalosporium stripe (Mundt, 2002) and study of the influence of the
different herbicide strategies on barley yields (Barroso et al. 2009).
15
Number of different applications
10
5
0
Biometrical Methodologies
Fig 2. Number of different Applications of Biometrical methodologies in Plant Protection.
In the Food Science and Technology field the survey showed fifteen
biometrical methodologies to be applied (fig.3), most of them with many
applications. Some of the common methodologies in the Food Science and
Technology field are:
Principal Components Analysis which has been applied for determination
of the combined effects of pressure, temperature, and co-solutes on Lactococcus
lactis (Killiman et al. 2006), classification of agricultural products according their
genotype/ physical properties/ origin (Campbell et al. 2000; Goodner et al. 2001;
Maeztu et al. 2001; Kallithraka et al. 2001; Bertelli et al. 2007), quality assessment
in grain/vegetables (Autran et al. 1986; Porretta, 1994). Discriminant Analysis with
applications in the determination of anthocyanins, flavonoids and colour parameters
in wines (Gomes – Cordoves et al. 1995), determination of mineral nutrients and
16
toxic elements in coffee (Éder José dos Santos et al. 2001) and detection of seed oil
adulteration in virgin olive oils in combination with spectroscopy (Vigli et al. 2003).
Survival Analysis with applications in studying sensory shelf life of foods (Hough
et al. 2003; Calle et al. 2003), determination of the affection of environmental
conditions to growth/no growth of typical spoilage yeasts (Evans et al. 2004),
estimation of optimum concentrations of a food ingredient (Gartita et al. 2006),
studying of optimum ripening time of fruits according to consumer data (Garitta et
al. 2008).
Number of different applications
15
10
5
0
Biometrical Methodologies
Fig 3. Number of different Applications of Biometrical methodologies in Food Science and
Technology.
In the Natural Resources Management and Agricultural Engineering
Field sixteen methodologies have been applied (fig. 4) while applications of most
common methodologies in the specific field are:
Number of different applications
12
10
8
6
4
2
0
Biometrical methodologies
Fig 4. Number of different Applications of Biometrical methodologies in Natural Resources
Management.
17
Principal Components Analysis with applications in: the examination of
crops’ yields according to fertilizer applications (Willson and Freeman 1970), the
evaluation of variation in biological, chemical and physical soil properties (Stenberg
et al. 1998), the determination of management effects on soil parameters (Sena et al.
2002) and the evaluation of microbial indices of soil fertility (Suzuki et al. 2005).
Discriminant Analysis with applications in: the determination of soil chemical
properties according to the presence of Azotobacter in soil (Cox and Martin 1937),
hard classifications in soil science (Hughes and Lindley 1955; Caroll et al. 2005)
and the determination of relationships between identified and measured soil nutrient
properties on fields (Splechtna and Klinka 2001). Factor Analysis with applications
in: Soil Quality Indexes identification and/ or interpretation. (Bachmann and Kinzel
1992; Wander and Bollero 1999; Brejda et al. 2000a,b; Andews et al. 2002; Shukla
et al. 2004a;2006), investigation of groundwater contamination (Grande et al. 1996;
Subbarao et al. 1995; Abu-Jaber et al. 1997; Jeong 2001), determination of
management discriminant properties in soils (Quiroga et al. 1998) and identification
of sources of soil pollutants (Carlosena et al. 1998).
In the Animal Production field the survey showed that eighteen
methodologies have been applied in the past and (fig. 5). Some of the most
commonly used methodologies in the Animal Production field are:
Number of different applications
15
10
5
0
Biometrical methodologies
Fig 5. Number of different Applications of Biometrical methodologies in Animal Production
Field.
The REML method with applications in: the examination of the effects of
level of fish oil inclusion in the diet on rumen digestion and fermentation parameters
in cattle offered grass silage based diets (Keady and Mayne 1999), estimation of
genotypic and phenotypic correlations in conjunction with specialized software
packages (Berry et al. 2002; Gilmour et al. 1999; Persson and Andersson 2003;
Conington et al. 2001; Legarra and Ugarte 2001; Neumaier and Groeneveld 1998;
Boldman et al. 1993; Bureau et al. 2001; Zhu and Weir 1996) and study of the
18
gestation length in Danish Holsteins (Hansen et al. 2004). The Survival Analysis
with applications in: Genetic Evaluation (Cox 1972; Cox and Oaks 1984; Ducrocq et
al. 1998), comparison of different dairy cow breeds on a seasonal grass-based
system of milk production (Dilloe et al. 2003), determination of factors affecting
culling early in the productive life of Holstein-Friesian cattle (Ojango et al. 2005)
and estimation of effects on longevity of beef cows (Szabó et al. 2009). The
General Linear Models with applications in: examination of relationship between
level of milk production and estrous behavior of lactating dairy cows (Lopez et al.
2003), effect of dietary phosphorus concentration on estrous behavior of lactating
dairy cows (Lopez et al. 2003) and investigation of the effects of maceration of rice
straw on voluntary intake and performance of growing beef cattle fed rice straw-
based rations (Nader et al. 2008).
In the Plant Production field the survey showed that fifteen methodologies
have been applied in the past and some of the most commonly used ones are:
Cart Analysis with applications such as in the detection of temporal and
spatial variability in crop yields (Perez-Quezada et al. 2003), the determination of
Number of different applications
15
10
5
0
Biometrical methodologies
Fig 6. Number of different Applications of Biometrical methodologies in Crop Production
Field.
relationships and interactions between soybean yields and a suite of soil and
agronomic variables (Zheng et al. 2009) and the examination of the primary
associations between environmental, agronomic and weed management variables to
crop and weed dependent variables (Williams et al. 2009). Cluster Analysis with
applications in: classification of varieties according to their genetic similarity
(Murphy et al. 1986), study the relationship between wheat grain yield and its
components under drought conditions (Leilah et al. 2005), determination of metals
in plants cultivated during the process of conversion from conventional to organic
agriculture (dos Santos et al. 2009) and study of the adjustment of modern rice
varieties in high-altitude regions (Steele et al. 2009). Multiple Regression with
application such as in: development of empirical models from large data sets, as has
been done for a number of canopy-level crop condition parameters (Shibayama and
Akiyama 1991; Osborne et al. 2002), identification of the relationship between rice
19
grain yields and soil mineralizable nitrogen (Tsujimoto et al. 2009) and study of the
cover cropping effects on yield and weed control of potato in transitional systems
(Campiglia et al. 2009).
In addition to the bibliographic survey, a questionnaire was constructed,
addressed to researchers and scientists in Greek Agricultural Institutes and Research
Centers. Its role was to survey the utilization of the biometrical methodologies and
relative informatics tools by scientists in the different fields of agriculture.
The whole procedure was implemented via personal interviews and each
scientist could answer the questions based on a five grade scale. This five grade
scale contains symbols *, 1, 2, 3, 4. The “*” means that somebody is not familiar
with the method, “1” means that the asking person simply doesn’t need that
methodology, “2” means that the asking person use this methodology in a rate 30%
of his work, “3” means that the asking person use this methodology in a rate 30-70%
of his work and “4” means that the asking person use the methodology in a rate
greater than 70% of his work. The total number of the questionnaires completed was
102 and the survey which began in 2009 is still in progress. (The results probably
must be indexed as preliminary). Finally, data from the questionnaires were
compared with data from the survey in the bibliography.
The results of the questionnaires so far, are presented below:
CROP PROTECTION
Agricultural Fields
NATURAL RESOURCES
4
FOOD SCIENCE 3
2
CROP PRODUCTION
1
ANIMAL PRODUCTION *
0% 20% 40% 60% 80% 100%
Utilization of ANOVA
Fig 7. Utilization of ANOVA in agricultural fields (Questionnaires)
Undoubtedly, ANOVA is a methodology very familiar to users from every
agricultural field. This condition is being confirmed from the fact that the rates of
scientists which apply ANOVA in their work range from 75-100% (fig.8), while
only a 9% in the Natural Resources field is not familiar with the method. Having in
mind the survey in the databases for ANOVA, this rate is justified because we have
found many applications of this methodology in each agricultural field (fig 1-6).
20
CROP PROTECTION
Agricultural Fields
NATURAL RESOURCES 4
3
FOOD SCIENCE
2
1
CROP PRODUCTION
*
ANIMAL PRODUCTION
0% 20% 40% 60% 80% 100%
Utilization of Regression Analysis
Fig 8. Utilization of Regression Analysis in agricultural fields (Questionnaires)
Regression Analysis is also a method with many applications in Agriculture,
as we have already seen from the survey. Figure 8 confirms this condition since its
use is very common within scientists in all agricultural fields with percentages
ranging from 75 to 85%. The utilization of this methodology is especially high in the
Natural Resources field, where the 85% of the researchers apply Regression
Analysis in a percentage larger than 75% of their work. The percentage of
researchers in all agricultural fields which are not familiar with that method ranges
from 17 to 25%.
Agricultural Fields
CROP PROTECTION
NATURAL RESOURCES 4
3
FOOD SCIENCE
2
CROP PRODUCTION 1
*
ANIMAL PRODUCTION
0% 10% 20% 30% 40% 50% 60% 70%
Utilization of Cluster Analysis
Fig 9. Utilization of Cluster Analysis in agricultural fields (Questionnaires)
The results of the survey in literature have showed that Cluster Analysis is a
method with many applications in all the agricultural fields except possibly from the
Crop Protection field. This fact is also being confirmed in fig. 9. Furthermore,
scientists in the Crop Production field, have to give more attention in this
methodology because of the great number of recorded applications (fig. 6). In
21
general, Cluster Analysis doesn’t seem to be well adopted from scientists in Crop
Protection (100%), Crop Production (51%) and Animal Production (55%).
Agricultural Fields
CROP PROTECTION
NATURAL RESOURCES
4
FOOD SCIENCE 3
2
CROP PRODUCTION
1
ANIMAL PRODUCTION
*
0% 20% 40% 60% 80% 100%
Utilization of PCA
Fig 10. Utilization of Principal Components Analysis in agricultural fields (Questionnaires).
For Principal Components Analysis the survey has shown that there are many
applications in Natural Resources Field and Food Science and Technology field. The
questionnaires (fig. 10) seem to confirm this view (Natural Resources: 45%, Food
Science: 75%), but still there are capabilities for a more extended application of this
methodology in these two fields. In Crop Production, Crop Protection and Animal
Production Fields there is a definite hysteresis in the use of this methodology, since
75%, 100% and 78% of the researchers are not familiar with Principal Components
Analysis.
CROP PROTECTION
Agricultural Fields
NATURAL RESOURCES
4
3
FOOD SCIENCE
2
CROP PRODUCTION
1
*
ANIMAL PRODUCTION
0% 10% 20% 30% 40% 50% 60%
Utilization of Factor Analysis
Fig 11. Utilization of Factor Analysis in agricultural fields
The survey showed that Factor Analysis has many applications especially in
Crop Production and Natural Resources fields. Results from the questionnaires show
22
that this methodology is being used in a medium percentage (value from 33% to 58
%) in all the scientific fields from Greek scientists. It is also obvious that scientists
from Animal Production (65%) and Food Science (75%) fields don’t prefer or don’t
have the knowledge to use this methodology in their work.
ANIMAL PRODUCTION CROP PRODUCTION FOOD SCIENCE NATURAL RESOURCES CROP PROTECTION
* 1 2 3 4 * 1 2 3 4 * 1 2 3 4 * 1 2 3 4
TABLES IN PAPER FORM COMPUTERS WORKSHEEP TABLES DATA FILES SPECIALLY DESIGNED DATABASES
Methods of Data Collection
Fig 12. Methods of Data Collection in agricultural fields (Questionnaires)
Regarding the data collection methods, in Animal Production field the 92%
of the researchers use worksheet tables, 33% use data files and only an 11% use
specially designed databases to store their data. In Crop Production field 100% use
worksheet tables, 50% use data files and 60% used specially designed database. In
Food Science field 100% use worksheet tables while just 25% use data files and
specially designed databases. All scientists from Natural Resources field use
worksheet tables, 55% of them use data files and 37% prefer specially designed
databases. Finally, half of the scientists in Crop Protection field use worksheet tables
whereas only 15% make use of data files (fig. 12).
As far as the informatics tools are concerned, in Animal Production field
56% of the researchers use ACCESS, 90% use SPSS and only 33% use a
programming language. In Crop Production field, 30% use Access and in general
there is an adequate knowledge in statistical packages. In addition, 15% of scientists
in Crop Protection use Visual basic and 8% use Pascal while 50 % of the scientists
in food science use Access, 75% use SPSS, 50% use STATISTICA and only 25%
use Visual basic. Scientists from Natural Resources field use ACCESS to a
significant extent (85%) and have a good knowledge of statistical packages. It is
impressive that 70% of them use FORTRAN. Finally, 35% of scientists in Crop
Protection field use ACCESS while 60% of them use SPSS and are familiar with
programming languages in rates from 15 to 30% (fig. 13).
23
ANIMAL PRODUCTION CROP PRODUCTION FOOD SCIENCE NATURAL RESOURCES CROP PROTECTION
EXCEL
DBASE
C,C++
S-PLUS, MINITAB
FOX PRO
SAS,JMP
ACCESS
ORACLE
AMIPRO
OTHER
SPSS,SYSTAT
STATGRAPHICS
PASCAL
FORTRAN
VISUAL BASIC
STATISTICA
DATABASES SPREADSHEET STATISTICAL SOFTWARE PROGRAMMING
Informatics tools
Fig 13. Informatics tools used in agricultural fields (Questionnaires).
4. Conclusions
The biometrical methodologies which are mostly applied, according to the
survey in databases/journals are:
- In the Biotechnology field: Cluster Analysis > Regression Analysis >
Discriminant Analysis > Biplot analysis >AMMI model.
- In the Crop Protection field: ANOVA > General Linear Models >
Regression Analysis > Discriminant Analysis.
- In the Food Science and Technology field: PCA > Cluster Analysis >
Discriminant Analysis > Regression Analysis > Survival Analysis >
Canonical Correlation Analysis.
- In the Natural Resources field: Regression Analysis > PCA >
Discriminant Analysis > Factor Analysis > ANOVA > Redundancy
Analysis.
- In the Animal Production field: Regression Analysis > Survival Analysis
> General Linear Models > ANOVA > REML.
- In the Plant Production field: ANOVA > Regression Analysis > Cluster
Analysis > Correlation Analysis > PATH analysis > CART Analysis.
According to the Questionnaires completed the most common methodologies are:
- In the Plant Protection field ANOVA and Regression.
- In the Natural Resources field ANOVA, Regression and Factor Analysis.
- In the Food Science and Technology field ANOVA, Regression and
Principal Component Analysis.
- In the Crop Production field ANOVA, Regression and Factor Analysis.
- In the Animal Production field Regression Analysis and ANOVA.
In the present study there is a definite hysteresis in the application of most
methodologies from Greek scientists, except for ANOVA, Regression and PCA.
24
Agricultural researchers could use the material described above in order to have
better results in their work.
Regarding the use of informatics tools it is a general fact that at least 30% of
scientists use Access whereas the vast majority of them (90%) use Excel. The
highest rates regarding the use of the statistical packages come from Animal
Production (90% use SPSS), Food Science (75% use SPSS), Crop Protection (60%
use SPSS), and Natural Resources (55% use Statgraphics). As for the Programming
languages, it is indicated by the questionnaires that Fortran is the most widely used
one, since 70% of Natural Resources Field Scientists opt for it.
REFERENCES
Akoroda O.M. (1996) Anatomy of problems facing users of statistical tools in agricultural
research: a Nigerian perspective. Proceedings of a CTA/University of Hohenheim Workshop,
Hohenheim, Germany 7-9 October 1996.
Box G. E. P. (1976) Science and statistics. Journal of the American Statistical Association,
71, p. 791-799.
Curnow, N.R. (1984) Present Position and Potentional Developments: Some Personal Views,
Statistics in Biometry and Agriculture. Journal of Royal Statistical Society, 147 (2), p 349-
358.
Ebdon JS, Petrovic AM, Schwager SJ. (1998) Evaluation of discriminant analysis in
identification of low-and highwater use Kentucky Bluegrass cultivars. Crop Science 38: 152-
157.
Evans, G., Everis, K. and Betts, D., (2004) Use of survival analysis and Classification and
Regression Trees to model the growth/no growth boundary of spoilage yeasts as affected by
alcohol, pH, sucrose, sorbate and temperature. International Journal of Food Microbiology,
92, p. 55– 67.
Finney J. (1978) Statistics and Statisticians in Agricultural Research. Journal of Agricultural
Science, 91, p. 653-659.
Garitta, L., Serrat, C., Hough, G. and Curia, A., (2006) Determination of optimum
concentrations of a food ingredient using survival analysis statistics. Journal of Food Science
71, p. 526–532.
Goodner, K., Rouseff, R. and Hofsommer, H., (2001) Orange, mandarin, and hybrid
classification using multivariate statistics based on carotenoid profiles. Journal of Agricultural
and Food Chemistry, 49, 1146–1150.
Jauset, A., Sarasu, M., Avilla, H and Albajes, R. (2000) Effect of nitrogen fertilization level
applied to tomato on the greenhouse whitefly. Crop Protection, 19, p. 255-261.
Keady, T. and Mayne, C., (1999) The effects of level of fish oil inclusion in the diet on rumen
digestion and fermentation parameters in cattle offered grass silage based diets. Animal Feed
Science and Technology, 81, p. 57±68.
25
McElroy, S., Walker, R.H. and Santen, E.V. (2002) Patterns of Variation in Poa annua
populations as Revealed by Canonical Discriminant analysis of Life History Traits. Crop
Science, 42, p. 513-517.
Obanor, F., Jaspers, M., Jones E. and Walter, M. (2008) Greenhouse and field evaluation of
fungicides for control of olive leaf spot in New Zealand. Crop Protection, 27, p. 1335– 1342.
Ojango, J., Ducrocq, V. and Pollot G. (2005) Survival Analysis to factors affecting culling
early in the productive life of Holstein-Friesian cattle in Kenya. Livestock Production
Science, 92, p. 317-322.
Perkins, J. and Jinks, J. (1968) Environmental and genotype-environmental components of
variability. IV. Non-linear interactions for multiple inbred lines. Heredity, 23, 525-535.
Preece, A.D. (1984). Biometry in the third world: Science not ritual. Biometrics, 40, p. 519-
523.
Puricelli, E. and Tuesca, D. (2005) Weed density and diversity under glyphosate-resistant
crop sequences. Crop Protection, 24, p.533–542.
Riley, J. (1998). Strengthening biometry and statistics in agricultural research. Study Report.
(CTA) Technical Centre for Agricultural and Rural Cooperation. Wageningen, The
Netherlands. 27 pp.
Sena, M., Frighetto, R., Valarini, P., Tokeshi, H. and Poppi, R. (2002) Discrimination of
management effects on soil parameters by using principal component analysis: a multivariate
analysis case study. Soil and Tillage Research, 67, p. 171–181.
Stenberg, B., Pell, M. and Torstensson L. (1998) Integrated evaluation of variation in
biological, chemical and physical soil properties, Ambio,27, p. 9–15.
Szabó, F. and Dákay, I. (2009) Estimation of some productive and reproductive effects on
longevity of beef cows using survival analysis. Livestock Science, 122 (2-3), p. 271-275.
Toler, J. E. and. Burrows, P.M. (1998). Genotype performance over environmental arrays: a
non-linear grouping protocol. Journal of Applied Statistics, [S.l.], 25, p. 131-143.
Vigli, G., Philippidis, A., Spyros, Apostolos, and Dais, P. (2003) Classification of edible oils
by employing 31P and 1H NMR spectroscopy in combination with multivariate statistical
analysis. A proposal for the detection of seed oil adulteration in virgin olive oils. Journal of
Agricultural and Food Chemistry, 51, p. 5715–5722.
Wander, M.M. and Bollero, G.A. (1999). Soil quality assessment of tillage impacts in Illinois.
Soil Science Society of America Journal, 63, p. 961–971.
Wilson, R.S. (1999). Evolution and Biometry. Biometrics, 55, p 333-337.
Zhu J. and Weir B.S. (1996) Mixed model approaches for diallel analysis based on a bio-
model. Genetic Resolution (Cambridge), 68:233–240
26