Untitled

6. DO ANY SPECIES BOUNDARIES...
149
breeding that can otherwise lead to expression of deleterious alleles. Indirectly it means that
strong incompatibility reaction within L. multiflorum should promote crossing with L. perenne.
Clearly, this interpretation stands out against the thesis that L. multiflorum and L. perenne
are different species.
The DNA data from the mapping population between L. multiflorum and L. perenne
(BR3 x NZ15) fully confirm the enzymatic outcomes. In total 26% of DNA markers are distorted
in the L. multiflorum x L. perenne mapping population in comparison with 25% of
enzymatic loci. The same proportion of disturbed segregation (20%) has been found in the
first interspecific mapping population constructed by Hayward et al. (1998). The level of DNA
marker distortions observed in the L. multiflorum x L. perenne population in the current studies
is typical of intraspecific crosses as indicated by 22% AFLP and RFLP markers distorted
in L. perenne mapping population, p150/112 (Jones et al. 2002b). From these comparisons
one notable conclusion is that very little genomic incompatibilities exist between L. multiflorum
and L. perenne. If this interpretation is correct it provides strong evidence of none
reduction in recombination in F 1
hybrids, thus confirming the earlier studies on chromosome
pairing (Naylor 1960).
Nevertheless, a case in point involves that some distortions at DNA loci are observed
in L. multiflorum and L. perenne. Any deviation from Mendelian segregation ratios is an
important feature that prevents the free exchange of genes and very often is attributable
to a birth of reproductive barriers. Unequal representation of parental alleles in a progeny
population appears to be related with unequal representation of alleles in gametes, differential
survival or fertilisation success of haploid gametes, and differential survival of the diploid
zygotes (Fishman and Willis 2005). Although the dominant mode of inheritance of the markers
used for the map construction does not allow discriminating among the potential causes
of distortions, it is unquestionable that they have resulted from various mechanisms at the
genome level. Structural rearrangements in chromosomes are between the most important
and frequent contributors to non-Mendelian inheritance in interspecific crosses (Rieseberg et
al. 1995). First, duplications, inversions and translocations in the heterozygous stage lead to
reduced pairing of chromosomes during the first meiotic division in F 1
hybrids. This inevitable
entails unbalanced gametes with additional copies of genes or/and without some genes. If
these aberrant gametes fuse the resultant zygotes are non-functional and hybrid sterility or
semi-sterility is observed. At the genetic map level it is observed as large distorted regions on
particular chromosomes. Results from L. multiflorum and L. perenne appear not to be consistent
with the above mechanism. Distorted regions are rather equally distributed between all
linkage groups. About 22-33% of markers are distorted on each linkage group with the only
exception of LG7, where 13% of markers deviate from the expected ratios. However, even
in this case all regions were scattered throughout the whole length of this group. A second
rationale for rejecting the chromosomal alteration hypothesis is the map length of 1 014 cM.
In general structural rearrangements have the effect of reducing map lengths. Meantime, the
present map is comparable or even longer than all other published maps of Lolium (Table
6.12). It is also comparable with the length of the genetic maps of barley (GrainGenes 2007),
whose genome size is very alike Lolium.
On the molecular level deviations from expected Mendelian segregation ratios are often
attributed to the effects of genic interaction. The widespread distributions of distorted regions