Untitled

84
4. GENETIC DIVERSITY...
Hamrick and Godt have suggested (1989) that the majority of variation in allogamous
species is distributed within populations. This appears as relatively high average gene diversity
(H S
) and mean G ST
value of 0.099 that can be roughly explained that only 9.9%
of the total variation is distributed between populations. A subset of enzymatic and DNA
data has confirmed high within-population diversity of L. multiflorum (Oliveira et al. 1997),
L. perenne (Fernando et al. 1997; Balfourier et al. 1998) and L. rigidum (Balfourier et al.
1998). The among-population diversity has been low and ranged from 8% to 14%. By contrary,
all marker types used in the present work, show quite different pattern of variation. The
levels of genetic diversities among populations are two-three times higher (50-60%) than
those previously reported. It can not be excluded that nonrandom distribution of dominant
DNA markers through the genome could distort the results considerable. In general, at least
four times more dominant markers are needed to obtain the same efficiency as with codominant
ones. With 15 enzymatic loci and the minimum number of 55 dominant ISJ loci this
condition has been fulfilled pretty well. It is striking however, that similar results have been
observed for codominant enzymatic loci thus, indicating another reason for the discrepancy
between the present and literature data. The population genetic parameters are associated
neither with the number of populations nor number of plants per population as long as the
population size is above approximately ten plants (Nybom 2004). Most of authors cited above
used a similar sampling scale. Therefore, either evolutionary or historical processes have
shaped the present differentiation of Italian and perennial ryegrass populations. Pros is the
high variation between L. perenne accessions revealed by RAPDs (Bolaric et al. 2005b),
AFLPs (Guthridge et al. 2001) and cpDNA (Balfourier et al. 2000).
The relatively high G ST
values reflect spatial genetic structure and suggest restricted
levels of gene flow. In many plant species, gene flow via pollen is sufficiently limited to less
than a few hundred individuals occupying areas less than 50 m 2 . In the grasses L. perenne
and F. pratensis, pollen dispersal decreases rapidly within 75 m from the donor field (Giddings
2000; Rognli et al. 2000). Although any immigrant pollen can be found a kilometer
away from the source it is not likely to have an effect on genetic variation. Like L. multiflorum
and L. perenne in the present studies, many wind-pollinated species have a relatively strong
population structure suggesting that wind may not be a particularly effective agent for pollen
dispersal (Avise 2004). However, given the lack of differences in genetic structure of
L. multiflorum and L. perenne, the hypothesis of limited gene flow does not seem to be the
most convincing. Furthermore when among-population diversity is calculated separately on
a cultivar and ecotype basis, the G ST
value drops down and it fits better with values typical of
allogamous species.
Another explanation involves the role of glacial and interglacial cycles. The last glacial
period reached a peak between 25 000 and 18 000 years ago. During this period, land temperatures
dropped as much as 20 o C. These climatic changes led to changes in the distribution
of many plant species. Afterwards, in the current interglacial period, species migrated
northwards (Roberts 1998). These historical associations among populations, rather than
patterns of ongoing gene flow, may play a predominant role in shaping patterns of genetic
structure. Molecular analysis of Abies species from southern Mexico and Guatemala indicates
that these populations may have passed through a number of genetic bottlenecks that
led to a loss of genetic diversity and interpopulational differentiation due to genetic drift (Agu-