An understanding of the relationship among sugar yield
components could increase the efficiency of selection for
sugar yield, once in recent decades, the increase in sugar
yield has been achieved primarily by increasing cane yield
rather than sugar content (Jackson 2005). Milligan et al.
(1990) pointed that selection for increased sucrose content
should emphasize cane yield with concentration on stalk
number. Thus, the evaluation of the yield components as
also as correlations between their components is an
important step in the characterization of a sugarcane family.
The quality parameters Fiber%, Brix and Pol%Cane are
the most relevant agronomic traits used as an indicator of
sugarcane quality (Aitken et al. 2006). The evaluation of the
genetic variability distribution, heritability and correlation
between these traits has practical implications in a sugarcane
breeding program, allowing to breeders a more effi-
cient selection of parents and prediction of genetic gains.
The high polyploid nature of sugarcane associated to the
heterozygosity condition of the majority of the loci allows to
obtains considerable levels of genetic variability on the first
generation of a cross involving two (bi-parental cross) or
more (polycross) parents (Chaudhary 2001; Hoarau et al.
2007). Progenies derived from bi-parental crosses have been
used in the construction of convectional maps in sugarcane.
Most of the genetic maps published were constructed based
on progenies derived from interespecific crosses (Ming et al.
2001, 2002). However, progenies clones derived from
crosses between varieties or elite clones have usually high
concentration of favorable alleles of breeder interests. In
addition, the segregation behavior of important traits in a
cross with a large number of progenies involving such type
of parents is relatively scarce for sugarcane.
The present work aimed to evaluate the genetic variability
of quality parameters (Fiber%, Brix and Pol%Cane)
and yield components at two crop cycles (plant cane and
ratoon cane) of a large bi-parental genetic mapping population
and also to estimate the heritabilities and pair-wise
correlation of these traits. The phenotypic characterization
of the mentioned traits as well as the knowledge of the
correlation among them is the first step to quantify the
potential of the cross for further QTL (quantitative trait
loci) detection. Moreover, the analysis of trait correlation
will allow the evaluation of the potential of a cross
involving modern varieties to identify clones that combine
desirable traits of high Pol%Cane and Fiber%.