Environmental Engineering Reference
In-Depth Information
•
Bunch planting:
According to Ladd et al. (1974), this is a recommended method when
there is a large number of individuals from a previously untested crossing, because the
simultaneous evaluation of many stems in the same area is possible. Normally, 5-10
seedling bundles are planted (Mangelsdorf 1953) and it is expected that natural selec-
tion eliminates the inferior ones, based on competition (Urata 1970), although this seems
questionable, because it is not known whether this may actually occur in heterogeneous
bunches. Matsuoka et al. (1999a) pointed out that the main disadvantages are the inability
to select more than one stem per bunch, as the mixture prevents individual identification
and makes it difficult to assess tillering, which is positively correlated with budding and
yield. Skinner et al. (1987) argue that there are examples of success of the bunch planting,
although there are no conclusive data about its superiority.
• Family selection:
It is based on the fact that for the important traits, the heritability
between families is higher than the heritability between individual plants. As a result,
several breeding programs prefer to carry out the selection between the families, choos-
ing those with higher phenotypic averages. Note that this procedure also serves to identify
superior crosses (Skinner et al. 1987). Matsuoka et al. (1999a) mentioned that this proce-
dure can lead to the disposal of superior individuals in families with a low average, but
with high variance.
• Selection of individual plants:
Many breeding programs carry out this type of selection,
using only the traits of high heritability as a criterion in this phase, such as Brix and
resistance to flowering and disease. Latter (1964) mentions that in some cases this selec-
tion strategy leads to better results than the selection between families and also reduce
the risk of discarding superior individuals in families with a low average. According with
Matsuoka et al. (1999a) the economic feasibility of assessing each plant individually, rather
than evaluating a large number of seedlings should be taken into consideration. In practice,
the individual evaluations are usually possible, because in the early phases of selection
many genotypes are discarded visually based only on their vigor (Skinner et al. 1987;
Matsuoka et al. 1999a).
21.6.3 E
valuation
of
c
lonES
in
f
iEld
t
rialS
In this phase, the genotypes selected are compared based on experiments using appropriate statis-
tical designs. Because of the high number of genotypes being evaluated and the small number of
stems to be used in experiments, the plots are usually small, e.g., a 5-m-long row per replication,
place and time of evaluation. It is very common in this step to use increased complete randomized
blocks (Federer 1956). Other traits that were not evaluated in the previous phase are now con-
sidered, and assessments of the regrowth capacity are also commonly made. Because of limited
experimental precision, the values of heritability are still low and therefore the selection should
not be too intense (Matusuoka et al. 1999a). Many sugarcane breeding programs in the world have
applied approximately 10-30% intensities (Skinner et al. 1987).
21.6.4 f
iEld
t
rialS
for
p
rEcommErcial
g
EnotypES
The clones selected as above are then evaluated in experiments with many replications, in several
environments, and throughout various rattons. Such experiments in Brazil are usually installed
in randomized block designs. After intensive evaluations through several cuts, environments and
years, the new cultivars can be recommended for commercial use. An interesting feature of this
phase is that sugarcane growers become greatly involved because most experiments are installed in
commercial farms and evaluated under conditions that are very similar to those in actual cultivation.
