APPLICATION OF COVARIANCE TECHNIQUE IN LONG TERM FERTILIZER EXPERIMENT
Loading...
Date
41835
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Data generated in AICRP on LTFE for the 2007-2012 has been considered in this
study. Primary nutrients such as Nitrogen, phosphorous, Potassium and Sulphur have
positive significant influence with yield in addition to Zinc, Manganese, Iron and Organic
Carbon. Rest have Positive non significant correlation with yield.
Path coefficient analysis is used to study direct /indirect effect of soil parameters
with the yield. The association is recorded after neutralizing the negative direct effect by
the positive indirect effect while causing interaction effect in response. Balanced (NPK)
nutrients are having better association (low residual effect) with yield. Higher residual
effect noticed with higher dose of balanced nutrition (150% NPK) and also with
imbalanced nutrition indicating poor association.
Canonical correlation of response group (grain and straw yield) with primary
nutrients group is more followed by secondary and tertiary nutrition group. This indicated
primary nutrients are more responsible for productivity. First canonical covariate itself is
explaining more variation in all the years and for the pooled data. Strong relation between
them is observed by the higher Eigen values compared to the second canonical variate.
ANCOVA revealed that majority of the significantly related variables (soil
parameters) as covariates are having less error mean sum of squares than ANOVA. It can
be inferred that, analysis of covariance will minimize the error component of analysis
when analyzed with significantly related components. ANCOVA is relatively more
efficient than ANOVA, since value of relative efficiency is more than 1 with the
significant variables as covariates.