Table 1. Results of statistical tests of different evolutionary and ecological hypotheses that produce alternative tree topologies for the CTV p20 gene.
Hypotheses | Statistics* | ||
SH†Likelihood | K-treedistK-score‡ | TOPD/FMTSSplit distance§ | |
H1 (original tree) | −1859.71 | 0 | 0 |
H2 (hosts) | −2394.82* | 0.066 | 0.87 |
H3 (geographical locations) | −4155.41* | 0.067 | 0.97 |
H4 (collection years) | −3266.35* | 0.066 | 0.96 |
H5 (virulence) | −1859.34 | 0.005 | 0.70 |
These statistical tests compare maximum likelihood trees based on different hypothesis in which the phylogenetic relationships are correlated to the host species (H2), geographical location (H3), collection year (H4) and virulence (H5) with respect to the reference tree (H1).
Log likelihood for each tree based on the Shimodaira-Hasegawa test. Significant values (P<0.01) are indicated by asterisks.
The K-score, the minimum branch length distance from the original tree, was estimated with the program Ktreedist.
The split distance, the smallest number of transformations required to obtain one topology from the other, was estimated with the program TOPD/FMTS.