Classic Genetics and Breeding - Genetics, Genomics and Breeding of Crop Plants - page 76

work was done on dwarf varieties, whose reduced height was determined
by major genes. However, although many different sources were studied, in
all cases, yield was reduced, with problems of grain filling in the center of
capitula. In addition, reduced internode length may have led to problems in
light absorption by leaves. In recent years, breeding has concerned selection
of “short” varieties, but the tendency is still a positive correlation between
height and seed yield, possibly related to greater leaf area.
2.5 Classical Genetic Mapping and Cytogenetics
Since research on sunflower genetics is relatively recent (1960 onwards), no
chromosome map based on morphological characters was produced. With
17 pairs of chromosomes, the probability of finding linkages between the
few major genes known in sunflower was quite low. The main exception
was the recessive gene controlling male sterility that was found to be linked
with only 1% recombination to a gene controlling anthocyanin production
(Leclercq 1966). Most genetic studies, concerning male fertility restoration,
branching, downy mildew, broomrape or rust resistance were limited to
showing independence of the small number of loci controlling these
characters. Studies of isoenzymes (Quillet et al. 1991) started only a few
years before the first molecular markers were obtained, so that, although
this system has been used to check hybrid parentage, isoenzyme markers
were mapped only at the same time as restriction fragment length
polymorphism (RFLP) markers. The existence of 17 different chromosomes
has also meant that there have been a few cytogenetic studies and no real
definition of karyotypes.
2.6 Interest of Molecular Maps and Knowledge of Genomics
for Conventional Genetics and Breeding
For the few characters showing Mendelian inheritance, phenotypic studies
allowed definition of dominance/recessivity and independence/linkage.
However, as for most crops, other than these limited studies, sunflower
breeding up to 1990 was based on field trials for characters which appeared
quantitative, either because they were under polygenic control or because
they were affected by the environment. It was necessary to make expensive
factorial or diallel crossing plans to determine heritability and additivity
(both parental lines contributing to the value of a hybrid) or dominance
interactions (character not forecasted from parental values). With only
phenotypic studies, it was difficult to say whether two parents had the
same genotype for a complex character or produced the same effect from
different genes.