Page 8 - Geraci_al_2004

144
Figure 4. Dendrogram of 16 wild Brassica populations from the distance matrix described in Table 1 and the UPGMA algorithm.
a population but does not describe the organization of
the variation in the population. For example two
populations of B. incana are very different for H
values (0.23 pop14, 0.09 pop15) but their allelic
composition is very similar; on the contrary B. incana
(pop14) B. villosa subsp. bivoniana (pop2), B. rupes-
trus subsp. hispida (pop13) have similar values (0.23,
0.24 and 0.21 respectively) but a very different allelic have an influence on the number of heterozygotes,
composition.
Several factors have been associed with differences
in values of polymorphism (P), allelic number (A),
gene diversity (H) such as life form, regional dis-
tribution, geographical range, breeding system (Ham-
rick et al. 1979; Hamrick 1990). According with these
parameters, high values are generally associed with
mixed system of reproduction between autogamy and
allogamy; low values are found in autogamous and in
endemic plants. In our case, we must keep in mind
that three species (B. macrocarpa, B. rupestris, B.
villosa) are endemic, some populations (as pop5) are
constitued by very few individuals and other popula-
tions as B. rupestris subsp. hispida (pop12) and B.
macrocarpa (pop16) that are larger in size, are iso-
lated from the geographic point of view.
Considering the fixation index, a great number of
populations, in particular large sized populations,
presented an excess of heterozygotes.This could be
explained by assuming the presence of subpopulations
or competition phenomena in a population as well as
extensive gene exchanges (Brown 1979) and also
hypothesizing the self-incompatibility systems, often
found in many Brassica populations (Watanabe and
Hinata 1999). The presence of rare alleles could also
because the probability to find them as heterozygotes
is higher than to find them as homozygotes. For two
populations, pop4 (B. villosa subsp. drepanensis) and
pop14 (B. incana), an excess of homozygous indi-
viduals was found. For these populations, the gene
exchange seems restricted either by geographical
isolation or by biological process.
Gst values could be related with breeding system
and seed dispersal mechanism (Loveless and Hamrick
1984). In Brassica populations investigated, mean
values of Gst were comprised between 0.30–0.37.
Therefore a breeding system mixed between au-
togamy and allogamy (probably prevailing because of
self-incompatibility) and a seed dispersal-mechanism
for gravity or provoked by animals could be hypoth-
esized.
The same structure of the diversity was observed
for each species with the intra-populational diversity