Likelihood and Bayesian Analyses with the Full Data

Figure 1 shows a maximum likelihood cladogram (-ln likelihood = 211,110.54) for the 16.4 kb data set under the GTR + r + I model of sequence evolution.

Rabbit Pika Cladogram

Whale Dolphin Hippo Ruminant

Llama Rhino Tapir Horse

Caniform Pangolin Flying fox Rousette fruit bat Rhinolophoid bat Phyllostomid microbat Free-tailed bat Hedgehog Shrew Mole Mouse Rat

Hystricid Caviomorph Sciurid Rabbit Pika

Flying lemur Tree shrew Strepsirrhine Human Sloth Anteater Armadillo Tenrec Golden mole s. e. elephant shrew l. e. elephant shrew Aardvark Sirenian Hyrax Elephant Opossum Diprotodontian

Cetartiodactyla

Perissodactyla

Carnivora Pholidota

Chiroptera

Eulipotyphla

Rodentia

| Lagomorpha

Dermoptera Scandentia

Primates

| Pilosa Cingulata | Afrosoricida

Macroscelidea

Tubulidentata Sirenia Hyracoidea Proboscidea

Xenarthra

Afrotheria

|Marsupialia

Figure 1. Maximum likelihood cladogram, with bootstrap values shown below branches, for the 16.4 kb data set. Asterisks indicate clades that were constrained in maximum likelihood analyses. The range of Bayesian posterior probabilities from four independent MCMCMC runs with the 16.4 kb data set is given above each branch (also see Tables 1-4). In cases where only a single value is shown, this value was obtained in all four runs. Abbreviations: s.e.=short-eared; i. e=long-eared.

Bootstrap support percentages are shown below each branch. Four independent Bayesian (MCMCMC) analyses, two that were run for 600,000 generations and two for 300,000 generations, resulted in the same topology. The range of posterior probabilities (all four runs) is shown above each branch, with posterior probabilities expressed as percentages. Tables 1-4 show posterior probabilities for individual MCMCMC runs. Although the maximum likelihood tree and Bayesian trees (from independent runs) were topologically identical, bootstrap proportions were generally lower than posterior probabilities.

Consistent with the maximum likelihood analyses presented in Madsen et al. (2001) and Murphy et al. (2001a), our analyses recovered four major clades of placental mammals: Xenarthra (Cingulata, Pilosa); Afrotheria (Afrosoricida, Macroscelidea, Tubulidentata, Hyracoidea, Proboscidea, Sirenia); Euarchontoglires (Rodentia, Lagomorpha, Scandentia, Der-moptera, Primates); and Laurasiatheria (Eulipotyphla, Chiroptera, Carnivora, Pholidota, Perissodactyla, Cetartiodactyla). There is also moderate to strong support for relationships between these groups. Euarchontoglires and Laurasiatheria are sister-taxa that together constitute a clade named Boreoeutheria (Springer and de Jong, 2001). The basal split among living placental mammals is between Afrotheria and Boreoeutheria + Xenarthra (herein named Notolegia, see below for definition). Bootstrap support for Notolegia was 76% and posterior probabilities were 0.99-1.00 in independent MCMCMC analyses with the complete data set. Two competing hypotheses for the position of the root that had the next highest likelihood scores are: (i) at the base of Xenarthra (i.e., a basal split between Xenarthra and Epitheria; -ln likelihood = 211,119.66) and (ii) between Atlantogenata (i.e., Afrotheria + Xenarthra) and Boreoeutheria (-ln likelihood = 211,115.95). SOWH (Swofford-Olsen-Waddell-Hillis) tests (Swofford et al., 1996; Goldman et al., 2000) rejected these locations for the root (Murphy et al., 2001b). There were no other positions for the root that had nonzero bootstrap percentages or nonzero probabilities in likelihood, and Bayesian analyses, respectively.

Most relationships within the major clades were also resolved. In Xenarthra, pilosans (sloth and anteater) cluster to the exclusion of the armadillo (Cingulata). In Afrotheria, there is a basal separation of paenungulates (Hyracoidea, Proboscidea, Sirenia) and a clade containing Tubulidentata, Macroscelidea, and Afrosoricida, with the latter two orders as sister-taxa. The basal split in Laurasiatheria is between Eulipotyphla and other taxa (i.e., Scrotifera of Waddell et al., 1999; hereafter called Variamana for reasons that are discussed

Table 1. likelihood bootstrap support and Bayesian posterior probabilities for the major clades of placental mammals*

Data sets and analysis Clade

Table 1. likelihood bootstrap support and Bayesian posterior probabilities for the major clades of placental mammals*

Data sets and analysis Clade

Boreo-

eutheria

Epitheria

Atlantogeneta

Xenarthra

Boreoeutheria

Laurasia-theria

Euarchonto-glires

1.

Full data set

ML-Boot

100

76

4

18

NA

100

100

100

(16,397 bp)

600,000 - 1st

1.00

0.99

0.00

0.01

1.00

1.00

1.00

1.00

600,000 - 2nd

1.00

1.00

0.00

0.00

1.00

1.00

1.00

1.00

300,000 - 1st

1.00

1.00

0.00

0.00

1.00

1.00

1.00

1.00

300,000 - 2nd

1.00

1.00

0.00

0.00

1.00

1.00

1.00

1.00

2.

Outgroup

-Opossum

1.00

0.96

0.00

0.04

1.00

1.00

1.00

1.00

jackknifing

-Diprotodontian

1.00

0.99

0.00

0.01

1.00

1.00

1.00

1.00

(16,397 bp)

3.

Subsets of

Nuclear

1.00

1.00

0.00

0.00

1.00

1.00

1.00

1.00

genes

(14,750 bp)

Protein-coding

1.00

0.98

0.00

0.02

1.00

1.00

1.00

1.00

UTRs

1.00

0.75

0.20

0.05

1.00

1.00

1.00

1.00

Protein-coding

1.00

0.26

0.07

0.67

1.00

1.00

1.00

1.00

1+2

UTRs+protein-

1.00

0.98

0.02

0.00

1.00

1.00

1.00

1.00

coding 3

Mitochondrial

1.00

0.01

0.00

0.99

1.00

0.01

0.00

0.00

rRNA

*With the exception of the maximum likelihood bootstrap analysis that was run for the full data set, all other analyses reported in this table are MCMCMC runs with MrBayes 2.01 (Huelsenbeck and Ronquist, 2001). All Bayesian analyses with outgroup jackknifing and subsets of genes were run for 300,000 generations, with burnins set at 60,000 or 75,000 generations based on empirical evaluation. NA, not applicable because clade was constrained in ML bootstrap analysis.

Table 2. Likelihood bootstrap support and Bayesian posterior probabilities for relationships within Afrotheria*

Data sets and analysis Clade

Table 2. Likelihood bootstrap support and Bayesian posterior probabilities for relationships within Afrotheria*

Data sets and analysis Clade

Paen-ungulata

Tethytheria

Proboscidea + Hyracoidea

Sirenia + Hyracoidea

Tubulidentata + Macroscelidea + Afrosoricida

Macroscelidea + Afrosoricida

Afrosoricida

1.

Full data set

ML-Boot

NA

38

22

40

79

70

81

(16,397 bp)

600,000 - 1st

1.00

0.02

0.22

0.76

1.00

0.99

1.00

600,000 - 2nd

1.00

0.03

0.40

0.58

1.00

0.99

1.00

300,000 - 1st

1.00

0.04

0.23

0.73

1.00

0.98

1.00

300,000 - 2nd

1.00

0.01

0.27

0.72

1.00

0.99

1.00

2.

Outgroup

-Opossum

1.00

0.04

0.48

0.48

1.00

1.00

1.00

jackknifing

-Diprotodontian

1.00

0.02

0.21

0.77

1.00

1.00

1.00

(16,397 bp)

3.

Subsets

Nuclear (14,750 bp)

1.00

0.01

0.88

0.11

1.00

1.00

1.00

of genes

Protein-coding

1.00

0.00

0.83

0.18

1.00

0.98

1.00

UTRs

1.00

0.28

0.49

0.23

0.95

0.97

0.09

Protein-coding 1+2

1.00

0.01

0.99

0.00

1.00

0.83

1.00

UTRs+protein-

1.00

0.08

0.17

0.75

0.54

0.88

0.99

coding 3

Mitochondrial

1.00

0.95

0.01

0.04

0.98

0.02

1.00

rRNA

Table 3. Likelihood bootstrap support and Bayesian posterior probabilities for relationships within Euarchontoglires*

Data sets and analysis

Clade

Glires

Rodentia

Euarchonta

Primatomorpha

Scandentia+

Scandentia+

Primates

Dermoptera

1. Full data set

ML-Boot

100

100

77

15

0

85

(16,397 bp)

600,000 - 1st

1.00

1.00

1.00

0.00

0.00

1.00

600,000 - 2nd

1.00

1.00

1.00

0.00

0.00

1.00

300,000 - 1st

1.00

1.00

1.00

0.00

0.00

1.00

300,000 - 2nd

1.00

1.00

0.99

0.00

0.00

1.00

2. Outgroup

-Opossum

1.00

1.00

0.99

0.00

0.00

1.00

jackknifing

-Diprotodontian

1.00

1.00

0.99

0.00

0.00

1.00

(16,397 bp)

3. Subsets of genes

Nuclear (14,750 bp)

1.00

1.00

0.94

0.03

0.00

0.97

Protein-coding

1.00

1.00

0.56

0.58

0.01

0.41

UTRs

1.00

0.02

0.25

0.01

0.03

Protein-coding 1+2

1.00

1.00

0.61

0.77

0.04

0.13

UTRs+protein-coding 3

1.00

1.00

0.88

0.01

0.00

0.99

Mitochondrial rRNA

0.00

0.00

0.00

0.00

0.00

0.98

Table 4. Likelihood bootstrap support and Bayesian posterior probabilities for relationships within Laurasiatheria*

Data sets and analysis Clade

Table 4. Likelihood bootstrap support and Bayesian posterior probabilities for relationships within Laurasiatheria*

Data sets and analysis Clade

Vari-

Chiroptera+

Fereu-

Carnivora+

Hippo+

Eulipo

Erinaceidae+

Yinptero-

amana

Eulipotyphla

ungulata

Pholidota

Cetacea

typhla

Soricida

chiroptera

1.

Full data set

ML-Boot

94

<5

59

91

100

100

100

75

(16,397 bp)

600,000 - 1st

1.00

0.00

0.98

1.00

1.00

1.00

1.00

1.00

600,000 - 2nd

1.00

0.00

0.98

1.00

1.00

1.00

1.00

1.00

300,000 - 1st

1.00

0.00

0.99

1.00

1.00

1.00

1.00

1.00

300,000 - 2nd

1.00

0.00

0.99

1.00

1.00

1.00

1.00

1.00

2.

Outgroup

-Opossum

1.00

0.00

0.95

1.00

1.00

1.00

1.00

1.00

jackknifing

-Diprotodontian

1.00

0.00

0.92

1.00

1.00

1.00

1.00

1.00

(16,397 bp)

3.

Subsets

Nuclear (14,750 bp)

1.00

0.00

0.65

1.00

1.00

1.00

1.00

1.00

of genes

Protein-coding

1.00

0.00

0.98

1.00

1.00

1.00

1.00

1.00

UTRs

0.06

0.00

0.00

1.00

0.97

1.00

0.99

0.69

Protein-coding 1+2

1.00

0.00

0.06

0.91

0.53

1.00

0.84

0.69

UTRs+protein-

0.48

0.01

0.38

1.00

1.00

1.00

1.00

1.00

coding 3

Mitochondrial

0.95

0.00

0.00

0.02

0.11

0.03

0.03

0.92

rRNA

below). Within Variamana, the next split is between Chiroptera and Fereuungulata (carnivores, pholidotans, perissodactyls, cetartiodactyls). There is strong support for a carnivore-pholidotan clade within Fereuungulata. Euarchontoglires is divided into Glires (Lagomorpha + Rodentia) and Euarchonta (Dermoptera + Primates + Scandentia), both of which receive high bootstrap support percentages and posterior probabilities. Within Euarchonta, Scandentia and Dermoptera are sister-taxa (bootstrap support = 85%; posterior probabilities = 1.00 in four independent MCMCMC runs).

Was this article helpful?

0 0

Post a comment