Figure 4

Comparison of the tree topologies inferred from the various single and concatenated mitochondrial gene datasets. A. Maximum likelihood phylogenies for Fusarium based on mt protein coding nucleotide sequences. The corresponding sequences for three species Trichoderma reesi, Metarhizium anisopliae and Lecanicillium muscarium in the Hypocreales were used for outgroup purposes. Bootstrap values (>85%), based on 1000 replications are indicated at the internodes. 1: Phylogeny based on the concatenated data set for known mitochondrial genes. A similar topology (referred to as topology 1) was inferred from the single gene datasets for cox 2, cox 3, nad 2 and nad 5. 2: Phylogeny based on the atp 6 dataset with similar topologies (topology 2) inferred from the cob, nad 4 and nad 1single gene datasets. 3: Phylogeny based on the atp 8 dataset, with similar topologies (topology 3) inferred from the atp 9 and nad 4L single gene datasets. 4: Phylogeny (topology 4) inferred from the cox 1 dataset. 5: Phylogeny (topology 5) inferred from the nad 3 dataset. 6: Phylogeny (topology 6) inferred from the nad 6 dataset. B. Summarized results of Shimodaira-Hasegawa (SH) tests where support for the six tree topologies was tested against the various datasets. For each dataset, the tree receiving the best likelihood score are indicated with “Best”; those topologies that are significantly worse (P < 0.05) than the best tree are indicated with “Yes” and those that are not (P > 0.05) with “No”. See Additional file 1: Table S3 for the full results.