Biomedical Engineering Reference
In-Depth Information
As yield and fiber quality traits are quantitatively inherited, breeding strategies
will require approaches involving careful selection of parents, effective plant
selection, and large breeding populations. Parental selection can be done at random
or based on performance data from a previous season, its pedigree, or more
quantitative data such as breeding value or combining ability, gained from prior
mating studies. Once selection has been made, careful evaluation is necessary
before advancing. Experimental designs for field trials are based on reducing the
amount of variation within the trial; the more replications either through years,
locations, or within locations, strengthen a dataset. Larger breeding populations,
whether through the number of crosses or selections, provide a better chance of
identifying and confirming the desired trait.
In the USA, there are both public and private breeders with defined roles. Public
breeders are primarily focused on germplasm enhancement. Improved breeding
lines are then released and available for private breeders to use in generating
commercial cultivars. Bowman [
136
] in a survey of breeding practices in the
USA showed the pedigree method was used in up to 82 % of private programs,
and backcross methods were used in up to 28 % of public programs. The average
cotton breeding program had about 100 crosses annually, creating up to 3,700
nursery plots and 7,500 plant selections. Yield testing began at F
4
stage, but this
stage had considerable range across programs. Detailed evaluation of performance
of elite lines was done at about six sites in private companies. Other than having
many more plant selections, similar practices are used in Australia.
Breeding for improved fiber quality is only beneficial if yield is maintained or
simultaneously improved. Long-standing negative associations between yield and
fiber traits, due to linkage, pleiotropy, or physiological factors [
31
,
137
,
138
], slow
the development of cultivars with premium fiber quality which are competitive in
yield with existing cultivars. In a detailed analysis of 6-11 years of cotton breeding
data from Australia and the USA, Clement et al. [
138
] reported the magnitude of
these associations was greatest for fiber strength, and despite differences in yield
and fiber strength between countries and seasons, there was a consistent and strong
negative association (Fig.
10.5
). For Australian data, a strength improvement from
32 to 34 g/tex was associated with a mean yield reduction of 1,000 kg lint/ha.
Fortunately, there were outliers in that association, albeit at low frequency
(1.4 % in Australian data), that have allowed some progress to be made in
improving fiber quality without loss of yield. It must be noted that in the last
30 years, the definition of high or premium fiber quality has shifted to much greater
values of fiber length and strength. For example, Culp and Harrell [
139
] and Scholl
and Miller [
140
] reported strength values of approximately 22.4 and 21.3 g/tex,
respectively, in their strongest lines which are well below current base grade values
(Table
10.4
). The triple hybrid line, TH 131-5, was 26 g/tex with 601 kg/ha lint
yield [
137
]; this shows the strongest genotype at the time was barely at the current
base grade, while the low yield displays the negative association. Figure
10.3
also
shows that fiber properties considered above average 30 years ago would now be
considered below average. Thus the task of achieving high yield and premium fiber
quality has also become more challenging. Breeding practices for improving fiber
Search WWH ::
Custom Search