Phylogenetic relationships within the speciose family Characidae (Teleostei: Ostariophysi: Characiformes) based on multilocus analysis and extensive ingroup sampling

Table 4 Likelihood-based tests for alternative topologies

Topology^a	Ln Likelihood	Diff.^b	SH	AU
(1,(2,(3,((4,(5, 6)),((7, 8), R)))))	-159175.3	0.0	1.0000	0.9139
(7,(1,(2,(3,(4,((5, 6),(8, R)))))))	-159195.4	20.1	0.4269	0.2226
(1,(2,(7,(3,(4,((5, 6),(8, R)))))))	-159195.7	20.4	0.4109	0.0514
(1,(7,(2,(3,(4,((5, 6),(8, R)))))))	-159198.9	23.6	0.3037	0.0766
(7,(1,(2,(3,(6,(8,(5,(4, R))))))))	-159200.5	25.2	0.2912	0.1972
(7,(1,(3,(2,(4,((5, 6),(8, R)))))))	-159201.4	26.1	0.2531	0.0882
((1, 3),(7,(2,((5, 6), (8, R)))))^c	-159212.5	37.2	0.1071	0.0000*
(7,(1,(2,(3,((4, 6),(5,(8, R))))))	-159215.6	40.3	0.0446*	0.0000*
(1,(2,(7,(3,(4,(5,(6,(8, R))))))))	-159219.3	44.0	0.0371*	0.0000*
(1,(4,(7,(2,(3,(6,(8,(5, R))))))))	-159220.9	45.6	0.0485*	0.0295
Calcagnotto et al. [31]	-159685.2	509.9	0.0145*	0.0000*
Lucena and Menezes [12]	-159692.4	517.1	0.0128*	0.0000*
Mirande [22], node 176 only	-160345.9	1170.6	0.0000*	0.0000*
Mirande [22]	-166637.0	7461.7	0.0000*	0.0000*

SH and AU are probability values obtained for the Shimodaira-Hasegawa and the Approximately Unbiased tests (Shimodaira 2002). Asterisks denote significant values (P < 0.05 for SH and P < 0.01 for AU), that imply the topology is rejected.
^a Topologies are sorted by likelihood values (obtained with Treefinder); at the top is the unconstrained ML topology summarized in Figure 4. Alternative hypotheses tested are depicted by parenthetic notation, where numbers represent the following taxa: 1: Distichodontidae+Citharinidae, 2: Crenuchidae, 3: Alestidae+Hepsetidae, 4: Erythrinidae, 5: Parodontidae, 6: Cynodontidae+Anostomoidea+Serrasalmidae+Hemiodontidae, 7: Ctenoluciidae, 8: Lebiasinidae, R: the rest (node 31 in Figure 4). Other topologies tested were taken from Calcagnotto et al. (2005); Lucena and Menezes (1998): the monophyly of Cynodontidae according to these authors groups: Cynodon, Hydrolycus, Rhaphiodon, Gilbertolus, and Roestes; Mirande (2010): either his full cladogram or only a constraint to impose his node 176 (Salmininae, Agoniatinae, Acestrorhynchinae, Cynodontidae).
^b Difference in Ln likelihood score with the best tree and the alternative topologies tested.
^c This topology is consistent with the reciprocal monophyly of African and Neotropical species.

ISSN: 2730-7182