Syst. Biol. 45(4):393--414, 1996

Data set incongruence and the phylogeny of crocodilians

Steven Poe

Department of Zoology and Texas Memorial Museum, University of Texas, Austin, Texas 78712-1064, USA;

Abstract.--- Published sets of systematic data on crocodilians (18S and 28S nuclear ribosomal DNA [rDNA] restriction fragments, mitochondrial rDNA restriction fragments, 12S mitochondrial rDNA sequences, osteology, external morphology, dentition, nest type, albumin, parasites) were used to test hypotheses of data set incongruence and phylogeny. Comparing incongruence indices between molecular versus molecular data set comparisons and molecular versus morphological data set comparisons showed the morphological;shmolecular comparisons to be significantly more incongruent, and experiments removing taxa suggested that morphological incongruence can be localized to the separation of Gavialis from Tomistoma. Significance tests of incongruence between the five larger crocodilian data sets relative to each other and to combined data sets supported these hypotheses and demonstrated that only 1 (12S sequences vs. morphology) of the 10 pairwise comparisons of data sets show significant incongruence. Three hypotheses of crocodilian phylogeny were evaluated using combined parsimony analysis, separate parsimony analyses, and evaluation of uncombinable data. The (alligatorids(crocodylids(Gavialis, Tomistoma))) hypothesis of crocodilian relationships was most well supported. Although this hypothesis is not supported by one of the molecular data sets and requires additional morphological homoplasy beyond that required in most-parsimonious trees based on morphology, other hypotheses require even more homoplasy, and any particular hypothesis of crocodilian evolution requires additional homoplasy in more than one data set. Alligatorid relationships were robustly supported in both combined and separate analyses. Crocodylus relationships were not well resolved in most-parsimonious trees from any individual data set but were completely resolved in the combined analysis.
[Combined analysis; crocodilians; incongruence; phylogeny; separate analysis.]

Syst. Biol. 45(4):415--448, 1996

A phylogenetic analysis of geese and swans (Anseriformes: Anserinae), Including Selected Fossil Species

Bradley C. Livezey

Section of Birds, Carnegie Museum of Natural History, 4400 Forbes Avenue, Pittsburgh, Pennsylvania 15213-4080, USA;

Abstract.--- A phylogenetic analysis of modern and selected fossil geese and swans was performed using 165 characters of the skeleton, trachea, and natal and definitive integument. Five shortest trees were found (length = 318; consistency index for informative characters = 0.634), which differed only in details of relationships among three species of Branta. The trees supported (1) +Cnemiornis as sister group to other taxa analyzed; (2) a sister group relationship between the moa-nalos of Hawaii and other geese and swans; (3) Cereopsis as sister group of Anser, Branta, Coscoroba, and Cygnus (contra Livezey, 1986, Auk 103:737--754); (4) monophyly of Anser, Branta, and +Geochen and confirmation of generic monophyly of each; and (5) Coscoroba as sister group to Cygnus. Selective exclusion of fossil taxa from the analysis variably affected inferred relationships and had substantial impacts on computational efficiency. Some nodes were not robust to bootstrapping: (1) nodes relating species groups within Anser, Cygnus, and +Thambetochen and (2) the node uniting Anser, Branta, Coscoroba, and Cygnus relative to Cereopsis. Bremer (decay) indices indicated similar differences in relative support for nodes. Skeletal characters were comparatively important in establishing higher order relationships, whereas integumentary characters were critical for lower order inferences. Constrained analyses revealed that other proposed phylogenetic hypotheses entailed variable penalities in parsimony. The shortest tree(s) was considered with respect to selected ecomorphological attributes (e.g., body mass, sexual size dimorphism, clutch size) and biogeography, and a revised phylogenetic classification of the geese and swans is proposed.
[Anseriformes; Anserinae; cladistics; comparative analysis; fossils; geese; morphology; swans; waterfowl.]

Syst. Biol. 45(4):449--470, 1996

Molecular phylogenetics of desmognathine salamanders (Caudata: Plethodontidae): a reevaluation of evolution in ecology, life history, and morphology

Tom A. Titus 1,2,3 and Allan Larson 2

1 Museum of Natural History, University of Kansas, Lawrence, Kansas 66045-2454, USA
2 Department of Biology, Washington University, St. Louis, Missouri 63130-4899, USA
3 Department of Biology, University of Oregon, Eugene, Oregon 97403-1210, USA;
E-mail: titus@darkwing.uoregon.edu4

Abstract.--- Phylogenetic relationships were reconstructed for salamanders of the plethodontid subfamily Desmognathinae to examine evolution of morphology, ecology, and life history. Mitochondrial DNA sequences encoding 12S and 16S ribosomal RNA and the valine transfer RNA provided 259 phylogenetically informative sites from approximately 1,200 nucleotide positions for 21 specimens representing 15 species and subspecies. These data were analyzed in conjunction with 13 morphological and reproductive characters to generate phylogenetic hypotheses. The directly developing terrestrial desmognathines Phaeognathus hubrichti and Desmognathus wrighti represent, respectively, the first and second phylogenetic branching events within the subfamily, and the remaining terrestrial lineage, D. aeneus, also branches near the base of the phylogenetic tree. These results challenge earlier hypotheses that within Desmognathus the small nonmetamorphosing species, D. aeneus and D. wrighti, represent the end of a graded phylogenetic decrease in size and decrease in use of aquatic habitats. In contrast to previous hypotheses, our results suggest that desmognathine evolution includes transformations in the direction of larger body sizes, lengthened larval periods, and greater use of aquatic habitats.
[Desmognathinae; ecology; life history; mitochondrial DNA; molecular phylogenetics; ribosomal DNA; salamanders.]

Syst. Biol. 45(4):471--493, 1996

A reconsideration of songbird phylogeny, with emphasis on the evolution of titmice and their sylvioid relatives

Frederick H. Sheldon 1 and Frank B. Gill 2,3

1 Museum of Natural Science, Louisiana State University, Baton Rouge, Louisiana 70803, USA;
2 The Academy of Natural Sciences, 1900 Benjamin Franklin Parkway, Philadelphia, Pennsylvania 19103, USA

Abstract.--- The phylogeny of oscine passerines was estimated by comparing 27 species using DNA--DNA hybridization. In the process, the finer structure of the "sylvioids" was examined (1) to assess the phylogenetic proposals of Sibley and Ahlquist (1990, Phylogeny and classification of birds, Yale Univ. Press, New Haven, Connecticut) and (2) to develop a framework for studies of sylvioid historical ecology. Many of Sibley and Ahlquist's phylogenetic proposals were supported, including their division of the oscines into two clades: corvids and passerids. However, their division of the passerids into three clades, Muscicapoidea, Sylvioidea, and Passeroidea, was not supported; neither their Sylvioidea nor their Passeroidea is monophyletic. The improved picture of oscine phylogeny presented here permits a more rigorous historical analysis of convergence, adaptation, phylogenetic constraint, and other evolutionary phenomena. For example, the sister group of the seed-caching Paridae is the Remizidae (including the verdin, Auriparus), not the nuthatches (Sittidae), which also cache seeds. Thus, seed caching arose separately in the Paridae and Sittidae and is likely to be a key innovation for these groups, i.e., an adaptation responsible for their diversification. Similar cases of convergence and thus potential opportunities for ecophylogenetic study are common throughout the passerines. Unfortunately, such study is hampered by the difficulty of resolving passerine phylogeny, which is characterized by many short internodes.
[Character mapping; DNA hybridization; historical ecology; Passeriformes; phylogeny; Sylvioidea.]

Syst. Biol. 45(4):494--513, 1996

Molecular and morphological data provide phylogenetic resolution at different hierarchical levels in Andira

R. Toby Pennington

Royal Botanic Garden Edinburgh, 20a Inverleith Row, Edinburgh EH3 5LR, Scotland;

Abstract.--- Cladistic analyses of molecular (chloroplast DNA restriction site) and morphological data matrices for species of Andira, a genus of Neotropical woody plants, were performed. The molecular data provided considerable phylogenetic resolution but failed to resolve relationships within species groups. The morphological data produced large numbers of equally most-parsimonious trees that yielded an entirely unresolved strict consensus tree. This lack of resolution reflects the small number of macromorphological characters in Andira that display discontinuous variation and thus are suitable for cladistic analysis. A directly combined cladistic analysis resulted in four far more highly resolved equally most-parsimonious trees. In contrast, combining the data sets using consensus techniques yielded poor phylogenetic resolution. The increased resolution of the directly combined analysis is explained by the molecular and morphological characters providing phylogenetic resolution at different hierarchical levels. The utility of the phylogenetic framework provided by this combined analysis is demonstrated by using the phylogeny to study the evolution of dispersal syndromes in Andira. Patterns of morphological variation in Andira are typical of many species-rich genera of woody tropical plants, which are an important component of diversity in tropical vegetation. This example demonstrates that cladistic analysis of combined molecular and morphological data may greatly improve phylogeny reconstruction of these organisms.
[Andira; bat dispersal; data combination; Leguminosae; rodent dispersal; tropical trees.]

Syst. Biol. 45(4):514--521, 1996

Accuracy of neighbor joining for n-taxon trees

Korbinian Strimmer and Arndt von Haeseler

Zoologisches Institut, Universität München, Postfach 202136, D-80021 München, Germany;
E-mail: (A. vH.)

Abstract.--- A Monte Carlo approach was used to estimate the accuracy of a given tree reconstruction method for any number of taxa. In this procedure, we sampled randomly over all possible bifurcating trees assigning substitution rates (branch lengths) to each edge from an exponential distribution to obtain a biologically sensible maximal observed distance. Three different sets of trees were studied: the unrestricted tree space, the biologically meaningful tree space as introduced by Nei et al. (1995, Science 267:253-254), and the population data tree space. We used this technique to elucidate the performance of neighbor joining as a function of the number of taxa, assuming that distances are uncorrected as sequences evolve according to the Jukes-Cantor model. The accuracy of neighbor joining decreases almost exponentially with the number of taxa. However, the rate of decrease depends on the tree space studied. Although the accuracy decreases towards zero, teh similarity, i.e., the number of partitions that are identical between model tree and reconstructed tree, is in all cases studied much higher than the value expected for two randomly chosen trees. Although the probability of recovering the true tree is dramatically influenced by sequence length, the average similarity does not decrease substantially if branch lenghts are not too short.
[Assigning edge lengths; Felsenstein zone; finite sequence length; Jukes-Cantor model; Monte Carlo sampling; neighbor joining.]

Syst. Biol. 45(4):522--543, 1996

Testing for phylogenetic conflict among molecular data sets in the tribe Triticeae (Gramineae)

Roberta J. Mason-Gamer and Elizabeth A. Kellogg

Harvard University Herbaria, 22 Divinity Avenue, Cambridge, Massachusetts 02138, USA;
E-mail: (R.M.-G.), (E.A.K.)

Abstract.--- Four molecular data sets are available for the diploid intersterile genera of the cereal grain tribe Triticeae, and there are numerous differences among the four published trees. All six pairwise combinations of data sets were examined using tree comparisons, the incongruence length difference test, the Wilcoxon signed ranks test, and a permutation test. We describe some advantages, disadvantages, and properties of the different comparison methods. Test results provide no evidence for significant differences in the phylogenetic signal among the three nuclear data sets, with the exception of the placement of a single taxon. The chloroplast DNA restriction site data, however, support a significantly different tree, and the differences probably reflect a separate evolutionary history of the chloroplast genome.
[Gramineae; incongruence length difference test; permutation test; phylogenetic congruence; Poaceae; T-PTP test; Triticeae; Wilcoxon signed ranks test.]

Syst. Biol. 45(4):544--556, 1996

A likelihood-ratio test of monophyly

John P. Huelsenbeck, 1 David M. Hillis,2 and Rasmus Nielsen 1

1 Department of Integrative Biology, University of California, Berkeley, California 94720-3140, USA;
E-mail: (J.P.H.), (R.N.)
2 Department of Zoology, University of Texas, Austin, Texas 78712, USA;
E-mail: hillis;

Abstract.--- Many phylogenetic analyses are inspired by or depend upon the monophyly of a group specified a priori. Also, many evolutionary problems for which phylogenies are useful do not require every detail of the phylogeny to be estimated correctly but depend upon the monophyly (or lack thereof) of a particular group. We propose a likelihood-ratio test that compares whether the best trees estimated with and without the constraint of monophyly are significantly different. Simulation suggests that the test is conservative when the null hypothesis (a particular specified constraint) is correct. We applied the likelihood-ratio test of monophyly to the question of the relationship of the presumed-extinct marsupial wolf (Thylacinus). Specifically, we examined the null hypotheses that (1) the marsupial wolf does not form a monophyletic group with dasyurids and (2) the marsupial wolf is a basal member of the Australian radiation of mammals. Both hypotheses were rejected using the likelihood-ratio test of monophyly.
[Likelihood-ratio test; maximum likelihood; monophyly; phylogenetic methods; Thylacinus.]