FKB's Research

PROJECTS IN PROGRESS

Temporal and Biogeographic Patterns of Basal Passerine Diversification
I have studied the timing and spatial patterning of taxic diversification in passerines at several hierarchical levels. Most recently, I have used DNA sequence data from two nuclear genes to infer relationships among major lineages of passerine birds. The order Passeriformes contains some 60% of extant avian species diversity, and is distributed worldwide in nearly all terrestrial biomes. Any explanation of avian diversification as a whole is incomplete without an understanding of the diversification of passerines in particular. Historical analysis of passerine evolution has been hampered by the relative morphological uniformity of the group, which takes to an extreme the general trend in birds. My postdoctoral studies in collaboration with George Barrowclough and Jeff Groth at the American Museum have revealed strongly-supported phylogenetic structuring within passerines using nuclear DNA sequence characters. Our data suggest many novel hypotheses of passerine relationships. For example, we reconstruct the basal divergence within the order between the New Zealand wrens (Acanthisittidae) and all other passerines, yielding paraphyly of the suboscines (non-songbird passerines). Even more interestingly, our data strongly contradict the monophyly of the large suborder Corvida, previously proposed by Charles Sibley and Jon Ahlquist in the comprehensive summary of their DNA-DNA hybridization work, “Phylogeny and Classification of Birds” (1990, Yale University Press). This group of primarily Australasian taxa has been thought to represent a large monophyletic radiation of passerines which originated on the isolated Australian continental mass, with subsequent dispersal of many lineages to the north. Paraphyly of the Corvida, and distribution of the basal lineages of oscine passerines (songbirds) in Australasia, strongly indicates an Australian origin not only for Corvida, but for oscines as a whole. Thus, our data pinpoint the origin of a large (nearly 80% of passerines are oscines), widely-distributed group to a single Gondwanan landmass. As a whole, our results support the contention that passerines had their origin in a late-Gondwanan context, as basal divergences within passerines are among taxa distributed on several Gondwanan landmasses. In collaboration with Joel Cracraft (also at the American Museum) and others, I have also examinined the temporal signal in our molecular data in order to test for congruence between the timing and spatial patterning of passerine diversification, in the context of geological reconstructions of the end-stage Gondwanan breakup. In collaboration with Pam Beresford, Tim Crowe, and Peter Ryan, I have clarified the relationships of some surprisingly ancient elements of the African avifauna (e.g., the enigmatic sugarbirds, genus Promerops, the “babblers” Arcanator and Modulatrix, the “bulbul” Nicator, and various endemic “warblers”. The relationships of these taxa suggest an unexpectedly early role for Africa in diversification of the Passerida. Current research at this level is focused on resolving relationships within the Sylvioidea, a diverse group of primarily Old World birds that historically have proven difficult to classify.

**Sequence data strongly suggest the origin of all oscine passerines in Australasia, rather than just the parvorder Corvida as previously hypothesized based on DNA-DNA hybridization data.**

Phylogenetic Approaches to Studying Avian Participation in the Great American Interchange
The Great American Interchange (GAI) refers to the massive interchange of northern and southern biotas which occurred upon closure of the Panamanian Isthmus some 3.5-2.5 Ma. The temporal and to some degree ecological details of this interchange have been well documented by rich mammalian fossil faunas from North, Central, and South America. Unfortunately, a similarly detailed record does not exist for the avifauna of the region. Highly vagile organisms, birds may have commenced intercontinental exchange much earlier than mammal groups, potentially facilitated by Caribbean island-hopping. Some groups (e.g., hummingbirds, New World suboscines) are clearly southern, and have recently expanded northward. One focus of my research has been on the biogeography of a more enigmatic group, the wrens (Troglodytidae). Wrens are distributed throughout the New World (with one representative in the Old World), with a center of diversity in southern Central America and northern South America. Directionality of dispersal in this group is unclear, though Ernst Mayr speculated that the group was North American in origin. Using both nuclear and mitochondrial DNA sequence data, I have inferred relationships among nearly all wren genera (see summary). Unfortunately, basal relationships within the group remain poorly resolved due to problems in rooting the tree. However, this study suggests that basal divergences within the family could be among lineages distributed in North America, pointing to a northern origin for the group. I have gathered additional support for this hypothesis by examining the interchange at an even finer scale, focusing on relationships within a single wren genus, Campylorhynchus, distributed throughout Central and South America (see maps). This study indicated a northern origin for the genus, with several inferred dispersal events into South America. At this finer scale, I also analyzed the temporal component of the data, which indicated that early dispersals of these wrens occurred much earlier than completion of a strictly terrestrial corridor. Molecular phylogenetic approaches provide clues to the directionality and timing of dispersal in the wrens, and offer promise for other widely-distributed groups of New World birds. Future work will focus on resolving relationships within wrens at a finer scale (see work on Thryothorus, below).

Recently, I began a study on the widespread and ecologically diverse New World nine-primaried oscines (subfamily Emberizinae), in collaboration with Kevin Burns, John Klicka, Scott Lanyon, and Irby Lovette. This study will go far toward resolving phylogenetic relationships among these problematic birds, which compose one of the most species-rich lineages of oscines (with 824 species, nearly 15% of all passerines, and over 8% of all birds). This study will provide important insights into the relationships among dispersal, ecology, and diversification in New World passerines.

Behavioral Evolution in Wrens and Blackbirds
Within wrens, I am studying the relative contribution of environmental (namely, climate) and historical (phylogenetic) factors in determining social group size within the genus Campylorhynchus. Species within the genus vary from pair breeding (brunneicapillus) through highly social breeders with groups containing up to 14 individuals (nuchalis; see this page for a survey of social behavior in wrens). Broad scale comparative analyses have identified a variety of climatic factors which may be important determinants of sociality in avian species. However, no clear correlations of environmental factors with sociality have been identified using explicitly historical (i.e. phylogenetic) comparative methods. These methods must be used, as it is clear that sociality is significantly overrepresented in certain avian groups. Current analyses of sociality within Campylorhynchus strongly indicate phylogenetic structuring, and suggest that the severity of drought conditions may place significant constraints on the formation and maintenance of large social groups.

Another complex of behaviors that varies significantly among wren species is the vocal duet. In wrens, these duets involve male and female vocalizations that are presented with a highly variable degree of temporal coordination, from precisely timed antiphonal songs where it may be difficult to separate the contributions of two individuals, to broadly overlapping songs of similar character. Likewise, males to a degree, but especially females vary significantly in their repertoire size. In collaboration with Peter Slater's group at the University of St. Andrews, I am studying the evolution of duetting behavior in wrens (and particularly in the genus Thryothorus) using phylogenetic comparative methods. Our results to date indicate first that the genus Thryothorus is paraphyletic, falling into four more or less distantly related clades. Second, song coordination and female repertoire size seem to correlate quite strongly with clade membership, with clades being characterized by differing combinations of these two characteristics. Finally, within clades, some species appear to depart from the “norm.” These species offer excellent opportunities for mechanistic and demographic tests of the causes and consequences of duetting behavior in these birds.

In collaboration with Scott Lanyon, I am studying host specificity in species of the blackbird genus Molothrus. The five species in this genus (excluding the bay-winged cowbird M. badius, and including the giant cowbird M. oryzivorus) are brood parasites, laying their eggs exclusively in the nests of other species, to be raised along with their “legitimate” nestmates, without care from their own genetic parents. We are trying to quantify the diversity of hosts parasitized by different species of Molothrus, which vary widely in the number of host species used (from 3 to over 200). However, in addition to species diversity, we are interested in quantifying host nest type diversity, life history diversity, and phylogenetic divergence from their parasites. These measures will give us insight into the behavioral and physiological adaptations required by the parasitic lifestyle, and help to distinguish between alternative models of host-parasite coevolution.

Testing for Phylogenetic Incongruence
Prevous work by myself and others has shown that one popular procedure for testing phylogenetic congruence among data sets—the ILD or incongruence length difference test—is subject to an excessive Type I error rate. This effect is apparent when data sets contrast significantly in a number of characteristics, including overall evolutionary rates, patterns of among site rate heterogeneity, or lineage-specific evolutionary rate variation. In a wide range of situations, the ILD will strongly reject congruence (or “homogeneity”), even though data sets are known to share the same underlying history of branching patterns. I am developing a new parsimony-based alternative to the ILD, which appears to ameliorate this effect. However, a second problem with the test—a lack of statistical power—can probably only be overcome by use of explicitly model-based analyses.

Nominal Type I error rates for the ILD test under various evolutionary models, for a given underlying tree and contrasting rates (rate 1: rate 2) of sequence evolution.

Phylogeny and Molecular Evolution of Nuthatches (Sittidae)
One problem currently of interest combines empirical and theoretical approaches in the study of a particularly thorny problem of inference. To date, much research has focused on variation among genes in inferred trees and subsitution dynamics. However, there can be considerable variation in substitution dynamics among lineages, as well. New methods are beginning to be developed to deal with these issues (e.g., the LogDet transform and non-stationary maximum likelihood), but the underlying molecular mechanisms driving this variation remain largely obscure. I am studying among-taxon variation in nucleotide base composition at degenerate sites of nuclear protein-coding genes, with specific focus on the nuthatches (Sittidae) and their allies (wrens, gnatcatchers and creepers; see this page). I previously discovered a significant deviation in base composition at the Sitta RAG-1 (nuclear recombination-activating gene 1) locus relative to all other passerine birds. I am currently sequencing RAG-1, the closely-linked gene RAG-2, and two other nuclear genes for several representatives of Sitta, as well as for a variety of outgroup taxa. In collaboration with Eric Pasquet at the Museum National d'Histoire Naturelle (Paris), I am using data from these and other loci to reconstruct phylogenetic relationships among species of nuthatches. More broadly, I am attempting to isolate the evolutionary transition between the AT-rich state (found in most taxa) and the GC-rich state (found in Sitta), and also to determine the effect of this shift on phylogenetic inference within the clade of Sitta and its relatives. Preliminary analyses of other nuclear data sets for orders of birds (in collaboration with George Barrowclough and Jeff Groth) indicate that such shifts in base composition may be fairly common, and represent a formidable challenge to phylogenetic inference, even when using methods designed to ameliorate this effect. Ultimately, this work will lead to comparative population-level studies of individual loci in AT- and GC-rich taxa which could be used to help unravel the mechanistic and evolutionary forces determining base composition. At least one other shift in composition at a nuclear locus within passerines has been identified, and may be studied in a similar fashion.

Nuthatches (genus Sitta) show overall higher GC content at RAG1 third positions than is found in most passerines (represented here by Certhia), in particular at the 3' end of the gene.