Greener
Journal of  Agricultural Sciences
Vol. 3 (2), pp.164-175,
February 2013

 ISSN: 2276-7770 

Research
Paper

Manuscript
Number:122112341

DOI: http://dx.doi.org/10.15580/GJAS.2013.2.122112341

Study of Some Fruit and Seed Traits Relationship and
Assessment of Multicollinearity in Date Palm (Phoenix Dactylifera L) Accessions of Nigeria by Correlation and
Principal Component

Analysis

Odewale J.O., Agho Collins, Ataga C.D., Odiowaya G.,

Hamza

A., Uwadiae E.O. and Ahanon M.J.

Plant Breeding Division, Nigerian Institute for Oil Palm
Research (NIFOR).P.M.B 1030, Benin City, 

Edo State,
Nigeria.

Corresponding Author’s
Email: collinsagho @ yahoo. com

Abstract:

Collinearity (or multicollinearity) is the undesirable
situation where the correlations among the independent
variables are strong. Multicollinearity misleadingly
inflates the standard errors and results in incorrect
conclusions about relationships between dependent and
predictor variables. Thus, it makes some variables
statistically insignificant while they should be otherwise
significant. It is like two or more people singing loudly at
the same time. One cannot discern which is which. They
offset each other. Most explanatory variables in the
biological sciences tend to correlate and this leads to
incorrect identification of the most important predictors.
Just like other crops, in date palm the seed characters most
times are used as predictors of the quality and quantity of
the fruits such as the fruit weight and size which are
important pricing index of date palm fruits and so, it
becomes important to develop a model to explain the
relationships between these two data sets without
collinearity. The changing environments accompany by floods
and high temperature coupled with the increase in population
in Nigeria and Africa as a whole is a concern to
agriculturist. Famine is fast approaching and crop modelling
is one of the solutions. The main purpose of this study is
to show how we can use multivariate analysis based on
principal component scores to establish a model that can
explain the relationship between the fruit and seed traits
of date palm and compare its ability to reduce
multicollinearity with the method of the ordinary least
square while also studying the variability that exist among
the germplasm collections for genetic improvement. The
result of the descriptive statistics which indicated the
values of the coefficient of variation for the different
characters of the fruit and seeds of date palm reveals the
possibility of genetic improvements of these characters.
Principal component analysis (PCA) was applied to predictor
variables to address the problem of multicollinearity. The
result of the principal component analysis indicated that
the contribution of the first two factors with Eigen value
greater than unity accounted for 86.5 % of the total
variation which was well above average and thus, explains
the use of PCA in data reduction. The results showed that
the principal component regression (PCR) was sufficient in
eliminating multicollinearity with a variance inflation
factor (VIF) and tolerance value (TOL) of unity and P < 0.05
and it was possible to explain a high percentage of the
total variance with a reduced number of principal components
as two principal components (PRIN 1 and PRIN 2) accounted
for most of the variability in seed characters observed
among the date palm germplasm collections from different
locations. There was a high level of variation in some of
the seed traits studied which could serve as the basis for
genetic improvement of date palm fruit. There was a direct
positive relationship between the groove width and the fruit
traits in the multiple regression analysis and the principle
component analysis, thus, indicating the importance of
groove width in the genetic improvement of date palm fruits.
 
Key words: date fruit, tolerance value, variance
inflation factor, multicollinearity, principal component
regression and principal component analysis.

Return to Content     View 
[Full Article – PDF]   

[Full Article – HTML]            [Full Article – EPUB]