Fig 1.
Phylogenetic analysis of mosasauroid relationships using different methods.
(A) Co-UMP: strict consensus of 30 most parsimonious trees (450 steps each) (CI = 0.350; RI = 0.692). (B) Mu-UMP: strict consensus of 84 most parsimonious trees (445 steps each) (CI = 0.329; RI = 0.660; length). (C) IWMP (fit = 45.45942; CI = 0.360; RI = 0.706; length = 449 steps). (D) ML tree. For the ML tree, branches are proportional to their length, values above branches indicate bootstrap support and scale bar represents branch lengths. Abbreviations: Ai, Aigialosauridae; Do, Dolichosauridae; Ha, Halisaurinae; Mo, Mosasaurinae; Pl, Plioplatecarpinae; Te, Tethysaurinae; Ty, Tylosaurinae; Ya, Yaguarasaurinae.
Table 1.
Model likelihoods and bayes factors for the analyses performed.
Fig 2.
Bayesian majority rule consensus tree drawn from 15,002 posterior trees (lognormal prior on rate variation across characters).
(A) Using an exponential hyperprior on the shape of the lognormal distribution. (B) Using a uniform hyperprior on the shape of the lognormal distribution. Branches are proportional to their length. Values above branches indicate clade probabilities and scale bar represents branch lengths. Abbreviations: Ai, Aigialosauridae; Do, Dolichosauridae; Ha, Halisaurinae; Mo, Mosasaurinae; Pl, Plioplatecarpinae; Te, Tethysaurinae; Ty, Tylosaurinae; Ya, Yaguarasaurinae.
Table 2.
Summary of the Wilcoxon signed-ranks and Shimodaira-Hasegawa tests (SH) test results for topologies generated under the different search methods.
Fig 3.
Maximum posterior clade credibility tree.
Tree obtained with a lognormal prior on rate variation across characters and a uniform hyperprior (see text for details). Abbreviations: Ha, Halisaurinae; Mo, Mosasaurinae; Pl, Plioplatecarpinae; Ru, Russellosaurina; Te, Tethysaurinae; Ty, Tylosaurinae; Ya, Yaguarasaurinae.
Fig 4.
Maximum parsimony ancestral trait optimization on the IWMP best fit tree.
Characters maps indicate evolutionary change of character 117 on the pelvic condition (left) and 123 on the pedal condition (right). White branches indicate state “0”, black branches indicate state “1”, and branches in shades of gray indicate missing data or ancestral state ambiguity.
Table 3.
Ancestral state likelihood reconstruction (Mk model) for pelvic and pedal characters obtained from the ML tree.
Fig 5.
Likelihood ancestral trait reconstruction on the Bayesian MCC tree.
Characters maps indicate evolutionary change of character 117 on the pelvic condition (left) and 123 on the pedal condition (right). White nodes indicate 100% likelihood for state “0”, black nodes indicate 100% likelihood for state “1”, and nodes in shades of gray indicate missing data or ancestral state ambiguity. Nodes with both black and white colors indicate the proportional likelihoods of state “0”(white) and state “1” (black) in a pie chart format.
Table 4.
Ancestral state likelihood reconstruction (Mk model) for pelvic and pedal characters obtained from the Bayesian maximum clade credibility tree.