Untitled

226
9. PHYLOGENETIC RELATIONSHIPS...
and cereals, not to study the evolution of the whole Pooideae sub-family. Thereby no wild
species were sampled.
Similarly, somehow surprising is the closer affinity of F. pratensis and P. pratensis to
Triticeae than to Lolium. F. pratensis and P. pratensis formed a small cluster that further
joined with cereals. On the other hand, the central placement of both species may reflect the
ancestral position to Lolium that is in agreement with previous data. The ancestral position
of F. pratensis is strongly supported by the Principal Component Analysis. In the PCA scatterplot
F. pratensis was placed somewhere in the middle between two Lolium clades and
P. pratensis (Figure 9.2). The last mentioned species was grouped with cereals. Noteworthy,
the placement of L. loliaceum between L. rigidum and L. perenne provides further credence
to earlier conclusion that L. loliaceum and L. perenne evolved from a common ancestor that
had the closest affinity to L. rigidum.
Not unexpectedly, the species delimitation is worse in the tree derived from cereal sequence
tagged sites (STS). The most surprising was the dispersion of two Lolium clades
into separate clusters (Figure 9.3). The Perenne clade formed a cluster with the Aveneae
and Triticeae tribes whereas the Temulentum clade was joined with F. pratensis and
P. pratensis. There may be plenty of biological and purely statistical explanations why the
Perenne clade clusters with A. sativa and A. strigosa. However, two possibilities are the