A Case Study in Amazonian Biogeography: Vocal and DNA -Sequence Variation in Hemitriccus Flycatchers

Louisiana State University LSU Digital Commons

LSU Historical Dissertations and Theses Graduate School

2000 A Case Study in Amazonian Biogeography: Vocal and DNA -Sequence Variation in Hemitriccus Flycatchers. Mario Cohn-haft Louisiana State University and Agricultural & Mechanical College

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses

Recommended Citation Cohn-haft, Mario, "A Case Study in Amazonian Biogeography: Vocal and DNA -Sequence Variation in Hemitriccus Flycatchers." (2000). LSU Historical Dissertations and Theses. 7348. https://digitalcommons.lsu.edu/gradschool_disstheses/7348

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact [email protected]. INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy subm itted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

Bell & Howell Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 800-521-0600

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A CASE STUDY IN AMAZONIAN BIOGEOGRAPHY: VOCAL AND DNA-SEQUENCE VARIATION IN HEMITRICCUS FLYCATCHERS

A Dissertation

Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy

The Department of Biological Sciences

by Mario Cohn-Haft B.A., Dartmouth College, 1983 M.S., Tulane University, 1995 December 2000

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 9998667

Ail rights reserved.

___ ® UMI

UMI Microform 9998667 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DEDICATION

To the memory of Richard B. Warren (1916-1998), my loving stepfather,

Dick, who couldn't have cared less about titles and degrees. I think he would have been proud of me for finishing and would have enjoyed hearing the story

told here.

To the memory of Anthony "Papa Nino" Capraro (1891-1975), my

maternal grandfather. He was the first intellectual in my family that I'm aware

of, a self-taught scholar and courageous fighter for social justice. His belief in the socializing and liberating power of education, his fierce dedication to acting

on what he knew was right, and his loving faith in (and expectations for) his

grandchildren inspire me to this day, despite the fuzziness of my memories of

him. To the memory of my youth (1961-?), which sometime between the start

and finish of this degree quietly but definitively flew the coop.

iii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGMENTS

I would like to have dedicated this dissertation to the geniuses who

invented the polymerase chain reaction, tape recorders, and computers, and those who wrote all the relevant software. They made it possible for a regular

"schmoe" like me to chug through analyses that not long ago weren't even

ideas. Even if the substance of the ideas presented here did not build in any

profound way on work that was done before me, the mechanics that went into this dissertation depended completely on the successful and brilliant inventions

of others. I would like to have dedicated this document to them, but I recognize

that they might not be interested in the genetics or voices of Amazonian birds

or the story told by those traits. So, rather than pester them with my study, I

thank them for the immeasurable contribution they made to it. I hope they are

very happy, or at least rich. I thank the members of my dissertation committee: Van Remsen, Fred

Sheldon, Mark Hafner, Bruce Williamson, Kam-biu Liu, and A1 Afton. Van was

the ideal adviser: patient, tolerant, trusting, encouraging, and almost always

right. Fred was like a second adviser, especially generous with his time and use

of his lab facilities. Bruce took a special interest in this study and offered some

great ideas and stimulating conversations. The administrative staffs of the Museum of Natural Science and

Department of Biological Sciences were helpful throughout my studies. Thanks

go to Alice Fogg, Gail Kinney, Jana Kloss, Katina Johnson, Wanda Miles, Prissy

Milligan, Marilyn Young, and especially Peggy Sims, for taking care of me like

one of her own. Mark Hafner, as Museum Director, made the museum not only

conducive to getting good work done but also very hard to leave, and a model

that any future job will have a hard time matching.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. My studies here received generous financial support from the LSU Board

of Regents, Lex and Chrissie Sant, the American Ornithologists' Union, the Frank M. Chapman Fund (American Museum of Natural History), the George

H. Lowery Fund, the Tropical Bird Research Fund, the Marantz family, Gerry Kusiolek, the Charles M. Fugler Fellowship, Mr. Peterson, Sigma Xi (National),

the Biological Dynamics of Forest Fragments Project, the LSU Museum of

Natural Science, the McDaniels Travel Award, and Biograds. In addition to

these funding sources, travel to research destinations was underwritten by Field

Guides Inc., the New York Botanical Gardens, Fundaqao Vitoria Amazonica,

Elderhostel, Ecotours and the staff of the Tucano, Paul Allen and Bill Gates, and

Henry and Lene Hagedom.

In addition to the LSU Museum of Natural Science, other institutions in

the United States and abroad, especially in Brazil, provided critical assistance in

the form of specimen loans, access to collections, and logistical support: Cornell

Laboratory of Ornithology Library of Natural Sounds, Chicago Field Museum, Academy of Natural Sciences in Philadelphia, Vienna Natural History Museum,

Smithsonian U.S. National Museum, Museu Paraense Emilio Goeldi, and

Instituto National de Pesquisas da Amazonia in Manaus. The Brazilian IBAMA

in Brasilia and local chapters provided collecting permits and logistical support

for field work. Access to Indian lands was authorized by FUNAI and FOIRN. Friends and colleagues from these and other institutions provided important

help and discussions: Tom Schulenburg, Doug Stotz, Kevin Zimmer, Jeff

Podos, Phil Stouffer, Kelli Gilbert, Sergio Borges, Magalli, David Oren, Ze

Maria, Jurgen Haffer, John Fitzpatrick, Ned Johnson, Carla Cicero, Manuel Plenge, Mort and Phyllis Isler. Jim Giamanco of the LSU Physics department invented and built a sound-attenuating adapter cable for the computer vocal

analyses.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Field companions were crucial to the success of this work. I would especially like to thank Curtis Marantz and Bret Whitney for helping me to find,

record, and collect a lot of tody-tyrants and other great birds. Jose Eremildes

and Ken Rosenberg also participated in some field expeditions. It is hard to

know how to express my thanks to the people all over the Brazilian Amazon

who opened their homes and properties, and gave their time and hospitality to

this stranger with a strange job. I hope I can find a way to return the favor.

I'm also grateful to those collectors and naturalists who gathered

specimens and made recordings that proved useful in this study, whether they

originally intended them to be or not. The following people volunteered the

use of their field tape recordings: Haven Wiley, Paul Coopmans, Fernando

Pacheco, Andy Whittaker, Kevin Zimmer, Tom Schulenburg, Ted Parker, Alex

Aleixo, Mark Robbins, Bret Whitney, Pepe Alvarez, Sergio Borges, Sjoerd

Mayer, Dan Lane, Dan Christian, and Phil Stouffer. I am grateful for the

collecting efforts of Alex Aleixo, Jason Weckstein, Dan Lane, and Pepe Alvarez,

who took time out from their other work to collect important specimens of

Hemitriccus, and especially to Bret Whitney, who helped and participated at

every stage and probably could have done this study better than I did. My eleventh-hour lab experience was introduced by Jason Weckstein,

enriched by Alexandre Aleixo and Kwai Hin Han, constantly rescued by David

Reed, speeded along by Rebecca Miller, encouraged at a distance by John Bates

and Kevin McCracken, and led to successful and rapid completion by the

careful and dedicated work of Nannette Crochet on the automated sequencer.

Thanks also to Jim Demastes for teaching me allozymes and to Frank Burbrink for convincing me not to use them.

It is impossible to overestimate the importance to me over the years of

my fellow graduate students, friends, and colleagues at the Museum. I'm

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. proud to have gotten to know these people and thank them for their help,

friendship, and intellectual stimulation: Danny Lane, Mark Portagee McRae, Dave Reed, Han E. Bee, Wacker, Alexio, John and Shannon, Lonnie Beavers,

Laurie Binford, Christopherberger Witt, Satheroni, Kalaniccus, Saaara,

Catherine Cummins, Jenn Chiddlesworth, Butlotink, Krats, Terrence, Manolito,

Jane Read, Kenny, Ted, JPO and Letty, Ben!, Doug Pratt, Rob Spigott, Rob

Merle, Jen Duhickey, Hockwoman, Andre Bundao, Soninha, Dave Moyer,

Heatherface, Jessica Wuzzah, Gooey Cox, Steven W. and Donna Ditts, Kazuya Nooki, Josie and her pussycat, Matt and Alison Styring and the dogs, Chico and

the chicklets, Thomas Frescura Valqui, Shannon Allen, Jeremus Kirchman,

Boundy, Mike Frichaux Seymore, David Good, Becky Saunders, Dr. Jimmy, and

Dr. Fitz. This study felt the influence of Ted Parker from the start. He knew what

I was up to and was really enthusiastic about the project. I wish he had stuck

around to share his insights with me, participate in the discoveries, and revel a

little longer in the beauty and wonder of the Amazon. I thank everyone in my family for encouraging me up to this point. My

mother, Athena Warren, believes in me unconditionally but can still do a great

job editing my writing. My father, Louis Cohn-Haft, taught me without even

trying about inferring history from the skimpy evidence available in the

present. He embodies most of what I like and admire about scholarship and

academia.

My wife, Rita, is also my best friend and toughest critic. I can't wait to get back to her and Butch and our life.

vii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS

Dedication ...... iii

Acknowledgments ...... iv

List of T ables...... ix

List of Figures ...... x

Abstract ...... xi

Introduction ...... 1

M ethods ...... 14

R esults ...... 41

D iscussion ...... 61

Literature C ited ...... 91

Appendix 1: List of Tape Recordings Analyzed ...... 101

Appendix 2: Tissue Specimens Sequenced ...... 104

Appendix 3: DNA Sequences ...... 105

Appendix 4: Pairwise Distance M atrix ...... 132

Vita ...... 137

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES

1. Study taxa ...... 16

2. Primers used in this study ...... 35

3. Percent sequence divergence ...... 56

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES

1. Amazon basin ...... 2

2. Distribution of subspecies of Hemitriccus zosterops...... 18

3. Distribution of Hemitriccus minor complex...... 19

4. Distribution of Hemitriccus inomatus complex...... 21 5. Field sites ...... 24

6. Localities of 17 sequenced specimens of Hemitriccus zosterops complex...... 31

7. Localities of 17 sequenced specimens of Hemitriccus minor complex...... 32

8. Localities of 10 sequenced specimens of Hemitriccus inomatus complex...... 33

9. Position of the 1026-bp portion of the mitochondrial genome sequenced...34 10. Distribution of vocal types in Hemitriccus zosterops complex...... 45

11. Distribution of vocal types in Hemitriccus minor complex...... 47

12. Distribution of vocal types in Hemitriccus inomatus complex...... 49

13. Discriminant function analysis classification of vocal types ...... 51 14. Parsimony consensus tree ...... 52

15. Parsimony bootstrap tree ...... 53

16. Single best maximum-likelihood tree ...... 54

17. Maximum likelihood bootstrap tree ...... 55

18. Corrected pairwise genetic distances ...... 60

19. Area cladograms ...... 67

20. Geographic distribution of variable phenotypic traits ...... 81

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT

Ongoing debate about the evolutionary history of the Amazonian

avifauna suffers from a shortage of data. Especially needed are thorough

geographic sampling, sensitive markers of genetic diversity, and historical data

in the form of explicit phylogenies. I examined patterns of geographic variation

in three widespread species complexes of nearly identical Amazonian

flycatchers in the genus Hemitriccus, based on spectrographic analysis of

vocalizations and phylogenetic analysis of DNA sequences (1026 base pairs of

cytochrome b from 44 individuals) from throughout Amazonia. Although

members of the three complexes are extremely difficult to distinguish

morphologically and have long been confused by taxonomists, they represent

three well-defined clades separated by a remarkable 10% sequence divergence.

Within clades, vocally defined populations correspond closely to genetic units

and more accurately than does current taxonomy. Area cladograms of the three species complexes are not strictly concordant and show five-fold differences in inter-regional sequence divergences, suggesting either different rates of

evolution or different ages of vicariance events. Two taxa, the H. zosterops and

H. inomatus complexes, show similar, but not identical, patterns of a basal

north-south split along the current path of the Amazon River and a more recent

east-west split within the northern population, consistent with the possibility of

a single set of vicariance events affecting both taxa similarly and

simultaneously. Nevertheless, within-population genetic variability differs

dramatically between the two groups, suggesting unique histories or life-

history traits. The third complex, H. minor, shows a different geographic

pattern of distribution and genetic divergence. Presently, no single hypothesis proposed to explain patterns of Amazonian biogeography adequately or

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. uniquely accounts for the patterns found in this study. A null hypothesis is

proposed, in which current environmental conditions are sufficient to explain

existing patterns without invoking particular historical changes. Based on the

results of this study, the null hypothesis cannot be rejected. More such studies

on many more taxa will be necessary before evolutionary patterns are

sufficiently well described to differentially implicate specific historical

processes.

x ii

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION

As early as the mid-1800s, some of the first naturalists to explore the

Amazon began to notice in the distributions of its animals and plants a pattern not seen anywhere else on earth. On opposite banks of the great rivers (Fig. 1),

including the Amazon itself and some of its tributaries, species differed, despite

the overall similarity of the forest on both sides (Wallace 1853, Hellmayr 1910,

Snethlage 1913). Although some of these species occurred on only one side of a

river, more often it was noticed that what appeared to be the same kind of

organism, occurring on both banks, looked different on opposite sides of the

river. For example, howler monkeys (genus Alouatta) on the Amazon's south

bank are black, whereas those immediately opposite on the north bank and

throughout northern Amazonia are red (Emmons and Feer 1990).

In some cases these different forms have been described as distinct

species, and in others as subspecies. Regardless of the taxonomic level applied to opposite-bank variants, however, it has always been clear that they are more

closely related to each other than to anything else found co-occurring with them

on the same side of the river. Thus, the predominant pattern is one of

widespread organisms, often occurring throughout the Amazon basin,

composed of geographically distinct variants separated by large rivers. The

different geographic representatives often are referred to as allotaxa or replacement taxa, because they have non-overlapping or allopatric distributions

and appear to replace each other geographically. Together, allotaxa form

widespread species groups or complexes often referred to as superspecies,

which are believed to be monophyletic, that is, all descended from a common

ancestral population.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10° S 10° 10° N 10° equator — — — Atlantic Ocean ^/tmazo& W1 Pacific Ocean Pacific Figure 1. Amazonbasin, showing country boundaries and principal rivers and topographic features believed relevant to the distributionabove of itsthe avifauna. range of the study Shadedstretches taxa. areas of which correspond For are simplicity, also known theto highlands name as "Marahon,""Amazon" (> 700 m elevation),here "Solimoes," identifiesand mostly"Amazonas." the entire main river, ro

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This pattern has several remarkable aspects. One is that there should be any variation at all in Amazonian organisms. Geographic (or racial) variation is

known in most organisms throughout the world and is thought to be the usual

precursor of spedation. However, in most cases of geographic variation there

are either obvious barriers separating distinct forms or environmental

differences that might differentially favor their various phenotypes. The

Amazon basin, however, appears at least superficially to consist of more or less

uniform rainforest, equally suitable throughout to its inhabitants, and not to contain any important barriers to dispersal, such as deserts, mountain ranges,

or major ecotones of the sort typically assodated with distributional limits or

geographic variation. So why should organisms vary at all throughout the

region or be restricted in their distributions only to certain areas within it?

Another remarkable feature of the variation found in Amazonia is that

its limits are defined by rivers. It seems especially improbable that birds, which

are so mobile, have river-delimited distributions. After all, the ornithological

literature is littered with reports of interior continental spedes appearing

occasionally on remote oceanic islands. Cattle Egrets (Bubulcus ibis) are

believed to have colonized the New World within the last two centuries by flying across the Atlantic Ocean from Africa (Crosby 1972, Telfair 1994). Even

tiny songbirds migrate twice annually across the Gulf of Mexico (Lowery 1946).

The success of the popular sport of birdwatching is in large part due to these

kinds of phenomena, which engender in its partidpants the expectation that at

least a few individuals of any bird spedes could eventually appear virtually

anywhere on the planet. Surely a river then, even a huge Amazonian one,

should pose no obstade to movements of birds. Indeed, nowhere else in the world can you encounter a different avifauna simply by crossing a river. Why

should that be the case in the Amazon?

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Beginning in the late 1960s, Jurgen Haffer (1969,1970,1974,1978,1982, 1985,1987,1992a, 1992b, 1993,1997a, 1997b; Haffer and Fitzpatrick 1985) drew

the attention of the scientific world to the problem by both documenting

patterns of distribution of South American birds in more detail than had ever

been done before and by offering a hypothetical explanation for these patterns.

His explanation, which came to be known as the "Pleistocene Refuge Theory,"

had a tremendous impact on the field of Amazonian biogeography by stimulating numerous studies, not only of other animal and plant groups, but

also in such diverse areas as climatology, geology, and paleontology. The

eventual success or failure of this hypothesis is the subject of ongoing debate

and has been reviewed repeatedly (e.g., Colinvaux 1996, Haffer 1997a, and

references therein). I will outline the hypothesis itself and attempt to evaluate

its current status later (see Discussion). For now, the point I wish to emphasize

is that the hypothesis was formulated to explain the avian distribution patterns

that Haffer himself described, the accuracy of which is crucial to the relevance

of any hypothesis proposed to explain them.

Haffer's description of avian distribution patterns was based on his

mapping the localities of distinctly recognizable plumage variations of widespread South American bird species and species complexes. By

superimposing the range maps of numerous such groups, he was able to

discern certain repeated patterns; particularly, he noted concordance in the

regions containing distinct geographic forms (areas of endemism) and in the

boundaries delimiting these regions of relatively uniform species composition (zones of contact).

Although Haffer assembled an enormous amount of data for the first

time and occasionally engaged in modest taxonomic revision (e.g., Haffer 1974),

most of his taxonomic assumptions were not controversial, and the general

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. areas of endemism he proposed remain largely unchallenged. Furthermore, one can argue that distributions of birds throughout the world are better

documented than those of any other animal or plant group; thus, basing

hypotheses of evolutionary history on bird distributions, even in the relatively

poorly studied Amazon basin, is a reasonable starting point. But are

Amazonian bird distributions described adequately to generate or evaluate

competing historical scenarios? I will argue that lack of data is a principal

obstacle to understanding the biogeography of the Amazon, and that the

patterns are still too poorly known to identify the processes that may have

caused them. Since Haffer's first formulation of the Refuge Theory (Haffer 1969),

biological field work in South America has increased explosively. Several new journals and bulletins have been created (e.g., Omitologia Neotropical, Ararajuba,

Cotinga), two large volumes edited (Buckley et al. 1985, Remsen 1997), and a

large database assembled (Stotz et al. 1996), all dealing specifically with the

Neotropical avifauna. Meanwhile, the more general ornithological journals contain burgeoning numbers of papers on South American birds (see Paynter

[1991] and Pacheco et al. [1999] for examples of this increased productivity in Brazil alone). Although Haffer made a monumental summary of the data

available at the time by examining specimens from museums around the world

(see Haffer 1974 for sampling methods and coverage), even now, after decades

of renewed interest in documenting the diversity of South America, the

geographic coverage of existing collections represents a mere fraction of the

area involved (Oren and Albuquerque 1991).

With increased study, countless range extensions have been

documented, gradually but steadily redrawing the range maps. In fact, only a

handful of lowland South American sites have avifaunas that can be said to be

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. thoroughly characterized (see Cohn-Haft et al. 1997). Also, with increased

recognition of subtle habitat distinctions within tropical rainforest and the very

specific habitat selection of some Neotropical bird spedes, distributions often

turn out to be remarkably patchy within the general region in which a spedes

has been found (e.g., Isler et al. 1997). A single newly studied locality can, with

its inevitable collection of unexpected bird records, cause a small revolution in

our understanding of distributions.

For example, results of recent avifaunal surveys on the west side of the

lower Rio Negro in Amazonian Brazil (Borges et al., in press) suggest that the

so-called "Imeri" area of endemism in northwestern Amazonia (Haffer 1969,

Cracraft 1985) had been prematurely characterized and delimited. Imeri now appears to contain a combination of bird spedes, some of which are tied to a

particular habitat type, and others of which occur in a particular geographic

region. Many spedes originally thought to be typical of Imeri occur in

campinarana or white-sand forest, which is a common habitat in that area, but

which occurs elsewhere patchily throughout the Amazon (Anderson 1981, Pires

and Prance 1985) and contains most of the same bird spedes wherever it occurs (Alvarez 1994; pers. observ.). On the other hand, other spedes known from

Imeri and found in typical upland (terra firme) forest, have now been found

many hundreds of kilometers away on the west bank of the lower Rio Negro

(Borgeset al., in press), suggesting that they may be widely distributed between

the Rio Negro and the Amazon. Similarly, bird spedes of riverine habitats

(Remsen and Parker 1983) not surprisingly have different patterns of distribution than terra firme spedes. These patterns have barely begun to be

documented, and in much of Haffer's work it was not known which spedes

occurred in which habitats.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Not only do our maps need revision, but our concept of the true diversity of Amazonian birds is changing. By the middle of the last century, it

was generally believed that there were virtually no new birds left to be

discovered on Earth (Mayr 1946). Consequently, it seemed plausible that our notions about Amazonian bird distributions, although based on rather scanty

museum collections, were not likely to change much in light of future field

work. Yet shortly thereafter, an explosion of discovery began in South America

that has continued unabated up to the present (see Amadon and Short 1992,

Peterson 1998). New species and subspecies of birds are described from South

America, including the Amazon, every year, and the rate has been increasing. By my own cursory tally from several prominent journals, over twenty new

species and two new genera have been described from tropical South America

in the last five years (since 1995)! In addition to the conspicuously new forms discovered, the "cryptic"

biodiversity of the Neotropics is only beginning to be appreciated. Morphological characteristics, such as size, bill shape, and plumage pattern, used traditionally to identify distinct populations of birds, are proving more

conservative than other traits that can be observed in living birds or detected

using modem techniques. In particular, vocalizations may differ

unambiguously in populations that had not been distinguished previously by their morphology. Based primarily or even exclusively on vocal differences, numerous taxa have been described or elevated to species status in recent years

(e.g., Lanyon 1978; Vielliard 1989; Howell and Robbins 1995; Robbins and

Howell 1995; Whitney et al. 1995a,b; Bierregaard et al. 1997; Isler et al. 1997,1998,

1999; Zimmer 1997; W hitney and Alvarez 1998; Krabbe et al. 1999; Whitney et al.

2000), and others have been realigned phylogenetically based on their vocal

similarity to taxa previously thought to be only distantly related (e.g., Whitney

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1992, Cohn-Haft 1996, Kratter and Parker 1997, Whitney and Pacheco 1997). I

predict that, when thorough surveys of vocal variation are completed and

evaluated, the recognized diversity of Amazonian bird species will as much as

double! Although the genetic basis of vocal differences has yet to be

demonstrated in many cases, molecular genetic techniques have been

developed in the last few decades that allow more or less direct assessment of

genetic variation within and among populations (Hillis et al. 1996b). These

approaches have still only been applied to a handful of Amazonian bird taxa

(Capparella 1988, Hackett and Rosenberg 1990, Hackett 1993, Bates et al. 1999)

but promise to represent the future wave of phylogenetic studies. One alternative to Haffer's Refuge Hypothesis, the River-barrier Hypothesis (Sick

1967, Willis 1969, Capparella 1991; see Discussion), gains support by the

observation that some bird taxa not differing in morphology on opposite banks of the Napo River in Peru in fact show considerable genetic divergence

(Capparella 1988,1991). It remains unknown how m any more taxa show a

similar pattern.

In the most general sense, our conception of the regions of endemism

originally identified by Haffer has remained remarkably intact, suggesting that

the broad patterns are reasonably well described. However, it is not yet

possible to evaluate how much of this tadt support merely represents a lack of

critical reevaluation. Amid the current upheaval in understanding, no major

taxonomic revision or synthesis of geographic distribution for all South

American or Amazonian birds has been completed in decades. In fact, active

field workers rushing to reach increasingly remote regions before their habitats

are permanentiy altered or destroyed have a backlog of discoveries and undivulged knowledge that will take years to appear in the literature. So, it

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. should be dear that the rate of change in our knowledge of the avifauna of

South America is too great for us to place great stock in what we believed was true just a few years ago.

In addition to the major modifications and improvements that we should expect in the spatial aspect of the study of geographic variation, modem

approaches also allow an assessment of the timing of evolutionary events

(Hillis et al. 1996b). The actual age of major evolutionary events can only be

estimated by independent information (such as radiometric dating of fossils or

of geological formations known to cointide with the event) or by assumption of

a constant rate of molecular change over time, the so-called "molecular clock".

The notion of a universal molecular clock, applicable to all organisms, is

currently out of favor, and the calibration of genetic differentiation in terms of

time since vicariance is fraught with difficulties (Avise 1994, Hillis et al. 1996a).

Nevertheless, relative genetic distances are still used as a rough estimate of

relative taxon age.

Less controversial, however, is that phylogenetic tree reconstruction

includes explicit statements of the temporal sequence of evolutionary events:

deeper bifurcations in a tree represent earlier population splits. When

populations are associated with geographic regions, phylogenetic trees become

explicit hypotheses of the historical relationships among areas (Rosen 1978,

Nelson and Platnick 1981, Cracraft and Prum 1988, Hackett 1993, Bates et al. 1998). Use of such "area tiadograms" is a potentially important tool, allowing

comparison with similar hypotheses generated by other sets of organisms and

with independent hypotheses (e.g., based on the fossil pollen record) of the

history of the region.

Ironically, although the description of patterns of geographic variation in birds was central to the stimulation of the study of Amazonian biogeography,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. emphasis quickly shifted to other sources of evidence, and bird distribution patterns have been treated as if they were completely understood. Meanwhile,

with current studies emphasizing distinct historical scenarios for the Amazon basin, even ornithologists seem content to propose new hypotheses while

supporting them with the same old distributional data set (e.g., Nores 1999). In

fact, because any reasonable hypothesis proposed m ust at least allow for the

distributions of birds as presented in the literature, it is becoming evident that

most historical hypotheses predict those distributions equally well. The inability to distinguish the now numerous hypotheses proposed (see

Discussion) by their predictions about bird distributions may be interpreted as

suggesting that the study of bird distributions has contributed all it can to the

question of the history of the Amazon basin (e.g., Haffer 1997a). However,

independent studies of the geomorphology, climate, and palynology of the region are conflicting, controversial, and inconclusive in their results (see

Colinvaux 1996, Haffer 1997a). Based on those studies, no clear picture of the history of the Amazon emerges. Thus, it seems that, if we are not to let the field

stagnate, the time has come for renewed biogeographic studies employing

larger samples and new techniques to reevaluate completely our understanding

of Amazonian bird biogeography. Recent studies of the biogeography of frogs (Lougheed et al. 1999) and

rodents (Silva and Patton 1998) using molecular phylogenetic data suggest

distinctly different patterns from those long accepted for birds; yet these recent

studies are based on few taxa or on geographically restricted regions. We

cannot say whether the differences found are due to real differences between

birds and mammals and frogs, to the use of modem versus old-fashioned methods, or are mere artifacts of the relatively small sample sizes of these

recent studies relative to the prior bird work.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I hope to have presented ample evidence that the patterns of distribution and geographic variation in Amazonian birds are far from thoroughly

characterized. This study is an attempt to begin to redress these sins of

omission in Amazonian ornithological research.

I examined patterns of geographic variation, sampled intensively and

extensively across the Amazon basin, in three closely related and broadly co

occurring species complexes. These are tody-tyrant flycatchers (family Tyrannidae) in the genus Hemitriccus: H. zosterops, H. minor, and H.

inomatus/minimus (see Methods). Virtually any avian taxon could be selected

for a study like this, and many more taxa should be studied. I chose Hemitriccus

because, in the course of prior field studies, I had noticed distinctly different

vocalizations in populations then believed, on the basis of morphology, to represent the same taxon. Simultaneously, I was involved in the rediscovery of

some virtually unknown species in the genus, discoveries of major range

extensions, and reidentifications of previously misidentified species at certain

sites. Some of these findings have already made their way into print (Cohn-

Haft 1996; Cohn-Haft et al. 1997; Borges et al, in press). Apparently, the strong

morphological similarity among species of Hemitriccus had led to considerable

taxonomic confusion (see Methods: Study taxa). Thus, this genus offered the

opportunity to make considerable revisions in prior notions of taxonomic limits

and geographic distributions, to apply new approaches emphasizing voice and

molecular genetics, and to examine the effects of these new data on existing

ideas about biogeography. Also, by limiting the study to one genus and one basic guild, some potentially confounding factors (such as variability in body

size, dispersal capabilities, generation times, metabolic rates, diets, or habitat

preferences) might be controlled.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Geographic variation was assessed independently in terms of vocalizations and DNA sequence data. This study had two principal objectives. One was to characterize the nature of geographic variation in each complex as

thoroughly as possible. By searching for these species throughout the Amazon

basin, I was able to improve confidence in estimates of their true geographic

distributions. Analyses of patterns of variation in the context of molecular

genetic phylogenies for each study species complex allowed evaluation of

current taxonomy (which is based only on external morphology of relatively

small and geographically limited samples) and tests of the utility of

vocalizations in recognizing cryptic genetic diversity, as measured directly by

DNA sequences.

The second main objective of the study was to explore historical area

relationships by treating the three phylogenies generated as independent hypotheses for the history of the Amazon. If all three species complexes show

the same areas of endemism and temporal sequences of diversification, then a

single series of historical events may have affected them all simultaneously.

This kind of concordance should raise optimism for future evaluation of

competing hypotheses, to the extent that the hypotheses make distinct

predictions about phylogeographic patterns. On the other hand, differences in patterns among the three groups would suggest the need to invoke more

complex historical models of evolution in the Amazon.

As in other Amazonian species, the details of Hemitriccus distributions

were not well known prior to this study (see Results). Yet described taxa (Traylor 1979) conform roughly in their purported distributions to the general

areas of endemism proposed for most Amazonian forest birds (Haffer 1969,

Cracraft 1985). Thus, results of this study could have general application,

serving as a model for Amazonian bird spedation. Unlike some Amazonian

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. birds, however, Hemitriccus species are morphologically and vocally

conservative, making them a challenging test for detecting meaningful

phenotypic variation. Furthermore, they also appear to be extremely similar to

one another ecologically.

This overall similarity among species suggests two alternative

expectations for phylogeographic comparison. One might predict that such

similar organisms should respond similarly to the same series of historical

events, meaning that, if any species would show concordant phylogenies, these

three taxa should. On the other hand, for such similar and apparently closely

related taxa to co-occur throughout the Amazon, then it is likely that they each evolved unique niches, such as microhabitat. If true, then it is reasonable to

suppose that each microhabitat responded differently to the same historical

(e.g., climatic) events, causing, in turn, distinct patterns of geographic variation

in its inhabitants. In this case, each species complex could prove to be a model

for all the species inhabiting its particular microhabitat. The results of this

study, then, will suggest the relative importance of various future lines of research.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. METHODS

Study Taxa Under current taxonomy (Traylor 1979), the genus Hemitriccus contains

some 20 species, all found in tropical South America. They are tiny (5-12 g), forest-based, insectivores, sharing a very similar overall morphology, including

nondescript plumage patterns and relatively flat, wide-based bills. As far as is

known, they are all ecologically very similar, foraging on arthropods by

sallying upwards for short distances from a perch, most often to the underside

of live vegetation (Fitzpatrick 1980). They give high-pitched, insect- or frog-like

songs, often consisting of a short series of notes (trill). As many as three species

in the genus may co-occur syntopically, where they seem to show vertical

stratification in the forest or differ in microhabitat use (e.g., edge versus

interior) (pers. observ.).

Hemitriccus species belong to a clade including another ten species in the genera Lophotriccus, Atalotriccus, Oncostoma, and Myiomis, all of which share a unique syringeal morphology (Lanyon 1988). I refer to this group inclusively as

the "tody-tyrants." Elsewhere (Cohn-Haft 1996), I have pointed out that the

genus Hemitriccus, as currently delimited, may be paraphyletic, such that some

or all of the other tody-tyrant genera might best be included in it. This situation, which I will be studying in more detail in the future, complicates the

choice of outgroups for phylogenetic analyses (see below). The tody-tyrants, in turn, are believed to be most closely related to the genera Todirostrum and

Poecilotriccus (sensu Lanyon 1988), with which they share certain derived cranial

features and nest morphology (Lanyon 1988).

All Hemitriccus species that occur widely in Amazonian forest were

included in this study. Eight Hemitriccus species are either primarily or entirely

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Amazonian in distribution. Of these, four (H. flammulatiis, H. josephinae, H.

striaticollis, and H. iohartnis) have restricted distributions within Amazonia,

rendering them unsuitable for a study of patterns of geographic variation across the Amazon basin. The remaining four species, H. minor, H. zosterops, H.

inomatus, and H. minimus, were the subjects of this study (Table 1). Their close

relationship is suggested by the fact that, at one time or another, each of these

taxa has been classified to include or be included in at least one of the others

(see Lanyon 1988, Stotz 1992, Cohn-Haft 1996, Cohn-Haft et al. 1997). Together

they may represent a monophyletic group, but this is not critical to the study

and will not be tested directly here. These study species represent three hypothesized widespread species

complexes (superspecies) (Table 1). The validity (i.e., monophyly) of each of

these complexes is critical for examining geographic patterns of speciation and

area relationships, and so will be tested in this study. Two non-Amazonian

taxa (H. z. naumburgae and H. spodiops) are also believed to belong to these complexes. If they prove to be ancestral (phylogenetically basal) to their

respective superspecies, then they will have no relevance to patterns of

speciation within Amazonia. However, because they might be derived from

Amazonian forms, they were included in phylogenetic analyses (see below). The "zosterops complex" consists of all known subspecies of H. zosterops,

which differ primarily in the richness of yellow coloration versus white in the

ventral plumage. Several of these taxa were originally described as full species,

but were lumped by Zimmer (1940) without detailed explanation. Presumably,

this was based on uniformity in size and shape (see measurements in Hellmayr

1927) and complementarity in geographic ranges of component taxa. The

monophyly of this group has never subsequently been questioned. This group

occurs throughout the Amazon (although distribution may be patchy or local)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1. Study taxa. Nomenclature follows Traylor (1979); arrangement in species complexes is based partly on hypotheses tested in this study (see Methods). Non-Amazonian taxa are shown in parentheses. Habitat preference is based on my own unpublished data (see text) and follows vegetation classification of Pires and Prance (1985).

Species Subspecies Habitat

zosterops complex H. zosterops H. z. zosterops terra firme forest interior H. z. flaviviridis terra firme forest interior H. z. griseipectus terra firme forest interior (H. z. naumburgae) terra firmeforest interior

minor complex H. minor H. m. minor terra firme and secondary forest interior H. m. snethlageae terra firme and secondary forest interior H. m. pallens igapo forest interior (H. spodiops) (monotypic) Andean cloud forest

inomatus complex H. inomatns monotypic campinarana, ridgetop, and igapo canopy H. minimus monotypic campinarana, ridgetop, and igapo canopy

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and also includes a disjunct population ( naumburgae) in the northern extreme of Brazil's coastal Atlantic forest (Fig. 2). Populations from Ecuador and northern

Peru have not been identified to subspecies. All members of the zosterops group occur in lower and middle strata (shaded interior) of lowland, primary terra

firme rainforest.

The "minor complex" consists of all subspecies of H. minor and the

monotypic H. spodiops of the foothills of the Bolivian Andes. Placement of these

two species in a single superspecies is based on certain shared, unusual

characteristics (presumed synapomorphies) of bill and nostril morphology, as

well as a strong overall similarity in morphology and vocalizations (Cohn-Haft

1996). Hemitriccus minor has a wide distribution and in many places occurs in

sympatry with H. zosterops, although it appears to be somewhat more limited

geographically than H. zosterops (Fig. 3). As in H. zosterops, the subspecies of H.

minor differ primarily in the richness of yellow ventrally. This difference,

however, is extremely subtle, and I am unable to distinguish morphologically

between the subspecies snethlageae and minor, whose distributional limits and

diagnostic characters are, in fact, poorly defined (Zimmer 1940, Traylor 1979,

Stotz 1992).

The minor group exhibits distinct ecological variation, unlike members of

the other two species complexes, which appear to be extremely uniform in their

microhabitat selection within a complex (Table 1). Subspecies minor and

snethlageae occur in the interior of terra firme forest and adjacent tall, non

flooded second growth; pallens, on the other hand, is only known from

igapo—seasonally flooded forest along blackwater rivers. Hemitriccus spodiops

is found only in a narrow mid-elevational range in Andean cloud forest (Cohn-

Haft 1996).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. naumburgae . . 2

H. rothschildi) . . 2 (H. (H. type locality shown because probably H. H. griseipectus z. based on current taxonomy (Traylor 1979). Large H. H. z. rothschildi Hemitriccus zosterops flaviviridis . . 2

zosterops H. . . 2

H. "T"s indicate"T"s type localities. Currently defunct subspecies question mark indicates questionregion ofknown marks occurrence indicate region where of possiblepopulation absence. not identifiedrefers Dots to indicate distinct to subspecies. range vocal type extensions (see SmallResults). discovered during this study. Figure 2. Distribution ofsubspecies of

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3. Distribution of Hemitriccus minor complex. Subspecies indicated follow Traylor (1979). Question marks indicate region of apparent absence. Dots indicate range extensions discovered during this study. "T" indicates type locality. The third spedes complex is composed of Hemitriccus inomatiis and H.

minimus. Both are currently considered monotypic and are among the most poorly known members of the genus. Shortly before the beginning of this

study, H. inomatus was entirely unknown in nature, represented in sdentific

collections only by the type spedmen, collected in 1831 (Pelzeln 1868).

Similarly, H. minimus was known until recently from only a few scattered

localities in southeastern Amazonia (Fig. 4). This lack of information and apparent rarity or extremely restricted distribution made them a seemingly

poor choice for a study of Amazon-wide geographic variation. However, new

and unpublished information about these taxa recently gathered by my colleagues (A. W. Whittaker, K. J. Zimmer, B. M. Whitney, and J. Alvarez) and

me, induding the rediscovery of H. inomatus in 1993, has led to the recognition

that both spedes are considerably more widespread and common than presumed. Furthermore, based on morphology, distribution, habitat

preference, and voice, we now believe the two to form a superspedes. In fact,

the greatest morphological difference between the two is ventral coloration:

inomatus is white below, whereas minimus is yellow, much like the variation

found within the other study complexes. Thus, I treat the spedes H. minimus and H. inomatus as geographic representatives of what I call here the "inomatus

complex" (comparable to the subspedes of H. zosterops, for example). This is a

working hypothesis, which is tested explidtly in this study (see Results).

Members of the inomatus complex appear to have fairly spedfic habitat

requirements, which may have contributed to their being overlooked in nature for so long. They occur in low-stature rainforest, often called campinarana

(Anderson 1981, Pires and Prance 1985). This general vegetation physiognomy

is typically found growing on patches of pure, white-sand soils, but also occurs

on low ridgetops, and in certain poorly drained areas, induding along

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A H . inomatus O H. minimus

Figure 4. Distribution of Hemitriccus inomatus complex. Filled symbols indicate localities known prior to this study. Open symbols are localities discovered during this study. "T" indicates type locality. blackwater rivers. In all of these situations, where the vegetation is somewhat thinner and lower than typical rainforest, members of the inomatus complex

occur. This patchily distributed and somewhat difficult to define habitat may

grade into typical terra firme or flooded igapo, and so it is possible to find a

representative of this group in syntopy with members of one or both of the other species complexes. Their coexistence with other Hemitriccus species may

also be facilitated by a preference for foraging in higher strata, usually in the

crowns of the tallest trees (even though those may be of relatively low stature).

Despite the semblance of taxonomic order imposed on these species

(Table 1), it must be stressed that they are all nearly identical-looking and have been subject to considerable revision over the years. In fact, within-complex

geographic variation in ventral color is often more impressive than among-

group differences. The taxon minor was originally described as a subspecies of

zosterops (although it was subsequently placed for some time in its own genus!),

and pallens was described as a subspecies of minimus, which in turn was lumped into minor, until very recently, when Stotz (1992) noticed that the type

series of minimus contained specimens of both typical minor and of a taxon that

had subsequently been named aenigma (=minimus)\ Under ideal circumstances,

specimens can be placed to species complex by a combination of

morphometries and nostril shape, along with other more subtle characters.

However, these flycatchers are slightly sexually dimorphic in size, causing

overlap between sexes of the different groups (Cohn-Haft et al. 1997); the other

characters, along with correct sex identification, depend on careful specimen

preparation. In practice, many specimens, especially old ones and those with

little supplementary data, are extremely difficult to identify with certainty, and

it is common to find specimens misidentified in collections. Even live birds in

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the hand are often misidentified, as was the case for nearly fifteen years at a

heavily studied field site (Cohn-Haft et al. 1997). Thus, confusion has reigned for most of the history of these taxa, and the

need is clear for revision applying new and objective criteria for diagnosis,

namely vocal and molecular genetic analyses. Data Collection

Most field work for this study was conducted between 1996-1998 in a

series of collecting expeditions to various localities throughout the Brazilian

Amazon. Localities were chosen to sample both banks of all rivers suspected of

representing taxonomic limits, to sample at the lower, middle, and upper

reaches of each of these rivers (to detect clinal variation around river

headwaters), and to sample at least one locality not associated with a major

river in each major interfluvium. Although in practice it was impossible to

reach all intended destinations during the study period (let alone find all study

taxa at each site on every trip), expeditions conducted prior to the main study

period provided further specimens and recordings, and I made additional trips

to other localities where specimen collection was not permitted, but vocal data

were collected (Fig. 5).

Individuals of study taxa were located by their vocalizations and were

tape recorded under natural conditions, using a Sony TCM-5000 cassette

recorder with a Sennheiser ME-66 directional shotgun microphone on Maxell UDXL-H 60-min cassette tape. After recording spontaneous, unstimulated

vocalizations, I conducted a series of playback experiments (not presented

here), tape recorded post-playback vocalizations, and then collected the bird

using a shot gun. Specimens were skinned in the field, and tissues were frozen on liquid

nitrogen. During preparation, all relevant data were noted (including

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure Figure 5. Field sites. Dots represent collecting sites; musical notes mark sites where only vocal data, but no specimens, were collected.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. coloration of bare parts, body mass, gonadal and molt conditions, habitat, and behavior—all available from the author on request), stomach contents and

syringes were saved (for future analyses), and cleaned skins were salted and stored for subsequent completion of taxidermy in the laboratory.

All my collecting was authorized by permits issued in my name since

1991 by the Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais

Renovaveis (IBAMA), under project numbers 1010/91-40 AM (general

collecting in the Amazon) and 02005.001226/91-79 (collecting in national parks and other restricted regions). In accordance with permit stipulations and

Brazilian law, all specimens were deposited at the Museu Paraense Emilio

Goeldi (MPEG), in Belem, Para, Brazil. Some specimens were subsequently

exchanged with the Louisiana State University Museum of Natural Science

(LSUMZ), and all tissue and stomach samples are on permanent loan to

LSUMZ. Additionally, all study specimens were loaned to me temporarily for

study at LSUMZ.

I collected as many individuals as possible of study taxa up to a limit of

ten per site. Typical samples ranged from 1-5 individuals per locality.

Numbers per locality and additional localities, especially those outside of Brazil

where I was unable to visit, were augmented whenever possible by collections

or recordings made by colleagues and tissue samples provided by systematic collections (see appendices 1,2).

Specimens collected were compared to study skins housed in major

museum collections in the New World (especially LSUMZ, MPEG, Colecdon

Boliviana de Fauna [CBF] of the Museo Nadonal de Historia Natural in La Paz,

Bolivia, and the American Museum of Natural History [AMNH] in New York).

Relevant type specimens were also examined at LSUMZ, AMNH, and the Vienna Natural History Museum.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Vocal Analyses The purpose of examining vocalizations was to see if they permit more successful recognition of geographic variation than notoriously unstable existing taxonomy, and to see which classification, vocal or current taxonomic,

better corresponds to natural populations (monophyletic groups). Therefore,

vocal classification was established prior to and independently of phylogenetic

analyses, which were based entirely on DNA-sequence data and included no

vocal characters. During the course of field work, scores of hours of Hemitriccus tape

recordings were amassed, and, inevitably, I developed a subjective sense of

individual and geographic variation in vocalizations of the study taxa. Based

on these observations, I proposed a classification of song types, diagnosable by

ear, and assigned each of these song types to a geographically proscribed region (see Results). Characteristics used to diagnose song types include a

combination of quantitative and qualitative characteristics (e.g., number of

notes, duration of song, relative pitch, tonal quality, relative frequency of songs

versus calls in vocalization bouts).

To evaluate the success of this vocal classification, discriminant function

analysis (DFA) was performed on songs classed into song types solely on the basis of geography, according to the distributional ranges outlined by the

previous qualitative classification (see song types in Results).

Vocalizations were analyzed spectrographically, using the program

Canary version 1.2 (Bioacoustics Research Program, Cornell Laboratory of Ornithology, Ithaca, New York) on a Macintosh PowerBook 180 computer. To

avoid problems of pseudoreplication, only one song per individual was selected

for spectrographic analysis, resulting in analysis of 147 songs (see Appendix 1).

Only natural (unstimulated) vocalizations were selected, except when

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. recordings of post-playback vocalizations were of markedly better quality than

natural (facilitating spectrographic analysis) and post-playback could be compared to natural songs of the same individual and found not to differ audibly. Also, only songs, not calls, scolds, alarms, etc., were examined. Songs

are defined here, functionally and contextually, as multi-note vocalizations

given by perched males, presumed territorial, while not engaged in other

activities such as agonistic encounters, copulation, or any known unusual

contexts. Thus, vocalizations of known females (i.e., collected and sexed

individuals) were excluded from analyses. It was not possible to distinguish

males from females in the field, however, and vocalizations of uncollected

individuals were assumed to be given by males (and so were analyzed) if the

vocalizing bird conformed in its behavior and context to observations of known

males. Only two of the dozens of collected vocalizing individuals proved to be

females, and their vocalizations were noticeably different from those of known

males, so inadvertent inclusion of female vocalizations in this study is not

believed to have occurred to any appreciable extent or to have biased results. The other major selection criterion was recording quality. Recordings

meeting other selection criteria were scanned by ear for the best portion

(highest signal-to-noise ratio), and up to 20 seconds of this portion was copied

digitally onto my computer as a sound file in the program Canary. Tape speed

was calibrated during copying to a recorded reference pitch (A-440 Hz). For

recordings provided by colleagues, which invariably lacked a reference pitch,

tape speed was adjusted as necessary so that the recordist's voice, if familiar to me, sounded natural. The copied segment was scanned visually as a waveform

graph (sound energy versus time) and the individual song with the best signal- to-noise ratio, but also not over-recorded (distorted by copying at too high a

level), was selected for spectrographic analysis.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For every song, fifteen parameters were measured using waveform, audiospectrogram, and spectrum representations in Canary:

1. number of notes (note defined as one temporally continuous

sound trace)

2. duration of song (in milliseconds)

3. first intemote interval (measured in milliseconds from the start of

the first note to the start of the second note)

4. second intemote interval (equals 0 for 2-note songs)

5. last intemote interval (same as #3 for 2-note songs)

6. fundamental frequency of first note (in kHz)

7. duration of first note (in milliseconds)

8. relative harmonic intensity of first note (defined as the intensity [=loudness in decibels] of the second harmonic minus the intensity of the

fundamental frequency at the same instant, mid-note)

9. fundamental frequency of second note

10. duration of second note

11. relative harmonic intensity of second note 12. fundamental frequency of last note

13. rank of highest note (sequential position of the note in the song

with the highest fundamental frequency)

14. fundamental frequency of highest note

15. relative harmonic intensity of highest note

Most of the these variables are standard "morphometries" of audiospectrograms (Catchpole and Slater 1995) and are similar to those used in

recent studies of species limits in Neotropical antbirds (Isler et al. 1998,1999).

Several variables may be correlated to one another. However, statistical

independence was not important, because variables were used only for

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. classification purposes and because DFA takes correlation among variables into account.

Because songs could vary from only two notes to nearly one hundred,

not all notes were measured. Variables were chosen to minimize the number of

empty or zero cells and the likelihood of surpassing the measurement accuracy

of the software, while still characterizing the salient or distinctive features of

songs. For example, by measuring the first, second, and last intemote intervals,

most systematic changes in rhythm throughout songs should be detected. By

measuring the frequency of the first, second, and last notes, as well as the

position (rank) and frequency of the highest note in the song, most simple

forms of modulation should be detected.

On the other hand, harmonics are frequently ignored in bird song

analyses. Because Hemitriccus songs contain conspicuous harmonics (voiced,

simultaneous, whole-number multiples of the fundamental frequency), it was

important to distinguish them from the fundamental frequency. I suspect that

some of the variation in tonal quality that allows a listener to distinguish the

songs of different species complexes is due to the harmonic structure of songs. Thus, basic frequency was measured using the fundamental (and not simply

the loudest frequency), and relative emphasis on different frequencies was

measured independently using harmonic intensity.

Based on the results of DFA, this classification was compared to current

taxonomy as a predictor of vocal characteristics. Then both taxonomic categories and this a priori vocal classification could subsequently be compared visually to an independently derived molecular phylogeny.

DNA Sequencing

A total of 46 specimens was sequenced, including members of the H.

zosterops group (n=17), H. minor group (n=17), H. inomatus group (n=10), and

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. presumed outgroup taxa Todirostrum chrysocrotaphum (n=l) and Poecilotriccus

latirostris (n=l). For the study taxa, every named taxon and every vocal type

was represented. One specimen was used from each locality (>100 km apart) to

describe the overall pattern of geographic variation, and a second individual

from at least one locality per species group was also included to provide an

indication of within-population variability (figs. 6-8). (See Appendix 2 for list of

specimens used.)

Deoxyribonucleic acid (DNA) was extracted from frozen or buffer-

preserved tissues (breast muscle, liver, or heart). To increase the chances of detecting any possible contamination when later comparing sequences,

individuals of the same taxon were not extracted sequentially, and a given

extraction run (one day) included a mix of study taxa with few repeats.

Extraction was effected using DNEasy extraction kits (Qiagen brand, Santa

Clarita, California), following the enclosed protocol for animal tissues. Using the Polymerase Chain Reaction (PCR) in a GeneAmp PCR System

2400 oil-free thermal cycler (Perkin Elmer Applied Biosystems, Norwalk,

Connecticut) and standard protocols (available on request), I amplified a

roughly 1100-base pair (bp) region of mitochondrial DNA (mtDNA), including

most of cytochrome b (cyt b) and some of the threonine tRNA gene immediately adjacent to the 3' end of cyt b. This region was amplified in two, partially

overlapping fragments (Fig. 9), using two pairs of primers (Table 2). To reduce

the chances of undetected contamination, all PCR reactions included negative controls, and the same approach used in extraction of avoiding consecutive

amplifications of multiple individuals of the same taxon in a single day's run

was employed for PCR. Reaction re-annealing temperatures ranged from 50-

55°C (increasingly higher temperatures were used throughout the study), and

only samples producing single, bright, and unambiguous bands of appropriate

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. com plex zosterops Sequenced specimens of complex. Specimen numbers within Hemitriccus zosterops fromAtlantic forest. H. H. z. naumburgae Figure 6. Localities ofsequenced 17 specimensof complex are sequential, appliedSpecimen roughly is 1 of clockwise from southeast to northeastaround Amazon basin.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. co m p lex minor Sequenced specimens of complex. Specimen numbers within Hemitriccus minor J Figure 7. Localities ofsequenced 17 specimens of complex are sequential, applied roughly clockwise from southeast to northeast around Amazon basin. o j N>

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. co m p lex inomatus Sequenced specimens of complex. Specimen numbers within Hemitriccus inomatus Figure 8. Localities of sequenced 10 complex specimens aresequential, of appliedAand mark B roughly localities clockwise ofknown werefrom southeast nottissue available specimens to northeast forstudy. from near around Iquitos, AmazonPeru, and basin. Acre, Brazil, respectively, Letters that u> GJ

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3’ H16064 tRNA • • : ryv:-| Fragment II Fragment H15499 cytochrome cytochrome b L15383 Fragment I L14987 — ND5 Figure 9. Position of the 1026-bp portion of the mitochondrial genome sequencedused in thisto amplify study and (shaded sequence region). them are shown by the arrows (see Table 2). Numbers indicate first andpolarity last nucleotide indicated positions,is for the light based strand. on chicken mtDNA Relative (Desjardinspositions of and the twoMorais fragments 1990); amplified and the primers 5' £

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ofKocher et al. (1989) thisstudy this study "H4a" ofHarshman (1996) Reference modified from "L14841" 5'-GCTTCC ATC CAA CAT CTC AGC ATG ATG-3' 5'-CCA AAC ACT CGT AGA ATG-3' 5'-GGG TTG TTTGAN CCTGTT TC-3' 5'-AAG TGG TAA GTC TTC AGT CTT TGG TTT ACA AGA CC-3' L15383 H15499 L14987 Location Nucleotide sequence (or direction) copied by thatprimer: L—light (forward), H—heavy (reverse). "N" in sequence indicates a fully degenerate site, atwhich any oneof the four possiblebases may occur. Table 2. Primers used in this study. Location is indicated by number of the nucleotide position at the primer's3'-end, based on the chicken mitochondrial genome (Desjardins and Morais 1990), preceeded by a letter referring to DNA strand $ H16064

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. length (when visualized under ultraviolet light after electrophoresis on agarose

starch gels with ethidium bromide) were kept. Successful amplification products were cleaned using QIAquick PCR purification kits (Qiagen brand)

and prepared for automated sequencing using a BigDye Terminator Cycle

Sequencing Kit (Perkin Elmer Applied Biosystems), using standard protocols

adapted for 10-pi reaction volumes, with the same primers and in the same thermocycler used for PCR. Cycle-sequence reaction products were sequenced

using an ABI377 Automated Sequencer (Perkin Elmer Applied Biosystems).

Sequences were visualized and aligned on a Macintosh computer using

the program Sequencher 3.1 (Gene Codes Corporation, Ann Arbor, Michigan).

Before alignment, I trimmed the short unreadable portions at the start of sequences and the primer regions at the ends and corrected obvious gel

anomalies (e.g., dye blobs). After automatic alignment, I visually inspected all sequences for discrepancies between light and heavy strands and for any

reading frame errors and the presence of stop codons, and I corrected manually

any mistakes that could be interpreted confidently as typical automatic gel-

reading errors. Problematic sequences were excluded from the study (see

below). All manual adjustments to sequences can be identified in the alignment

files (available on request). Complete sequences for all individuals were then aligned to one another

and trimmed to the same length, such that final sequences contain 1026

contiguous base pairs, all from within cyt b and including the final stop codon

at the end of the gene (Fig. 9). All sequences presented here are fully double stranded with no ambiguous base calls, except for one individual H. zosterops

(no. 17; LSUMZ B-25545), for which only the light strand was successfully

sequenced and six bases (of 1026) were left partially or entirely unidentified

(Appendix 3).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. There was no evidence of contamination in any of the sequences or in negative controls, and all sequences included in phylogenetic analyses are

believed to represent true mitochondrial DNA, not nuclear copies (also called

"numts" or "pseudogenes"). Some samples (not presented) produced sequences of possible pseudogenes and were excluded from the study. Suspected sequences contained one or more nucleotide mismatches in the roughly 100-bp region of overlap of the two amplified DNA fragments (see Fig. 9), each unambiguously read on both heavy and light strands, suggesting that

the two fragments did not derive from the same contiguous region of DNA.

Furthermore, one fragment in the mismatched pairs contained numerous

ambiguous base reads (double peaks) in both directions, suggesting the presence in that fragment of two DNA templates—either the mitochondrial

sequence and its nuclear copy, or both nuclear copies with some heterozygous

lod. Inclusion of these problematic sequences in phylogenetic analyses led either to alignment of the sample with the "wrong" taxa or to a basal position in

trees on unusually short branches. These kinds of results have been reported previously with avian pseudogenes (Sorenson and Fleischer 1996). An

alternative interpretation of these cases is cross-contamination of samples;

however, the total lack of DNA in PCR negative control lanes suggests that

contamination would have to have occurred during or prior to DNA extraction (including field or loan preparation of samples). These cases will be

investigated further at a future date.

Although it may not always be possible to recognize pseudogenes or ever to guarantee their exclusion, their accidental inclusion in phylogenetic

analyses intended to be based exclusively on mtDNA sequences can lead to

incorrect phylogeny reconstruction and distorted estimates of genetic distance (Arctander 1995, Sorenson and Fleischer 1996, Quinn 1997). Therefore, it is

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. important to take rigorous steps to identify pseudogenes and to avoid their

inadvertent inclusion with mtDNA in analyses.

I used recommended precautions and believe that all included sequences

represent mitochondrial DNA for the following reasons: (1) DNA was extracted only from tissue samples, which have high ratios of mitochondria to

nuclei relative to blood or skin samples; (2) no stop codons occurred within the

cyt b portion of any of the sequences; (3) sequences contain no insertions or

deletions relative to one another or to other known avian cyt b sequences; (4) as

expected in protein-coding DNA, base substitutions were most frequent in third

positions within codons, considerably less frequent in first positions, and rare in

second positions (see Appendix 3), as opposed to non-coding nuclear copies, in which substitutions can occur in roughly equal proportions at all positions (e.g.,

Arctander 1995); (5) superimposed base peaks were rare and occurred on only

one strand (and so could be explained by typical sequencing or gel-reading

anomalies and could be interpreted unambiguously); (6) sequences in both

DNA fragments amplified from each individual were identical and unambiguous in their region of overlap; (7) additional DNA amplifications,

performed on a few specimens, of the entire 1100-bp fragment in one

continuous piece, using only the two external primers, gave identical sequences

to those of the same individuals when amplified in two overlapping fragments;

and (8) in phylogenetic analyses, no samples appeared in unexpectedly basal portions of the tree or had exceptionally short branch lengths (see Results), both

of which, if present, would indicate the inclusion of slower-evolving nuclear

copies. My confidence that the sequences presented in this study are truly mitochondrial in origin is enhanced somewhat by my having identified what I

believe to be pseudogenes with characteristics (described above) different from

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. those of the included sequences. Nevertheless, it is conceivable that some or all

sequences represent recently transposed nuclear copies, which would differ

only slightly from their mitochondrial originals and be extremely difficult to

detect. However, such copies would also have little effect on phylogenetic analyses or genetic distances, especially considering the degrees of sequence

divergence found in this study (see Results).

Exact, step-by-step protocols, reaction recipes, lab notes, gel

photographs, digital and printed electropherograms, and Sequencher alignment

files may be examined directly on request to the author. All unused tissue

samples, DNA extracts, and successful PCR products were saved. Sequences

will be deposited in GenBank upon publication.

Phylogenetic Analyses

Aligned and justified sequences (1026 base pairs of cyt b) were analyzed

in PAUP* 4.0 (Swofford 1998) using equally weighted parsimony and

maximum-likelihood models. The former assumes equal probability of substitution at all base positions. Departures from this assumption were

incorporated in the maximum-likelihood model. Maximum likelihood analysis

used the gene evolution model "HKY+G" (Hasegawa et al. 1985) and model

parameters determined by the computer program Modeltest (Posada and

Crandall 1998; kindly provided by D. Posada). Those parameters were: Ti/tv

ratio=10.6642; nucleotide frequencies A=0.2857, C=0.3540, G=0.1116, T=0.2487;

gamma shape parameter=0.1616. Based on pairwise likelihood-ratio tests (Swofford et al. 1996) of increasingly complex models using Modeltest, this

model was determined to be the least complex one to offer statistically

significant improvement over simpler models. Support for optimal branching

diagrams was examined by bootstrapping, based on 1000 bootstrap replicates for parsimony and 100 for the maximum-likelihood model. More replicates

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were not run in maximum-likelihood due to computational constraints: 100

replicates took over eight days of continuous computation on the fastest available computer (a 667 MHz alpha processor running Linux version 6.1,

using PAUP* 4.0 compiled for Linux), and 1000 replicates could be expected to take more than 80 days.

Trees were rooted using the two outgroup taxa, Todirostrum

chrysocrotaphum and Poecilotriccus latirostris; these are representatives of the two

genera most closely related to Hemitriccus and allied tody-tyrants, based on a

study of syringeal morphology (Lanyon 1988). Other tody-tyrant genera (e.g., Lophotriccus, Oncostoma, and Myiomis) may belong within Hemitriccus (see

Cohn-Haft 1996) and so were not used in this analysis. The genus-level

relationships of the roughly thirty tody-tyrant species is the subject of ongoing research.

Alternative tree topologies were constructed in MacClade (Maddison

and Maddison 1992) and compared statistically using the Kishino-Hasegawa test (Kishino and Hasegawa 1989) in PAUP*.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RESULTS

Distributions My fieldwork throughout Amazonian Brazil and the unpublished work

of colleagues (assembled here) resulted in considerably improved

understanding of the distributions of Amazonian Hemitriccus species over that

available in the literature. Major range extensions were discovered, including probable new taxa, and apparently real gaps in distribution were identified.

Most strikingly, members of the H. inomatus complex were found to be

widespread throughout the Amazon (Fig. 4). Prior to this study, inomatus was

known only from the type specimen taken in 1831 from the upper Rio Negro in

northernmost Brazil (see Methods: Study Taxa). This study produced the

world's only five specimens of inomatus in addition to the type. All modem specimens and tape recordings come from within 200 km of the east bank of the

Rio Negro, where the species is locally common in patchily distributed, low-

stature, scrubby campinarana forests. The type locality is west of the uppermost

Rio Negro along the Rio Iqana, an extremely remote area that was not visited

during the study. Assignment of the modem specimens to inomatus is based on my comparison of these specimens to the type. The eastern distributional limits

of this species remain unknown, although inomatus was not found at study sites further east in the Brazilian Amazon, nor has it been found in Guyana (M. J.

Braun and M. B. Robbins, pers. comm.). However, it almost surely will be

found to occur in appropriate habitat in parts of Venezuela and Colombia

adjacent to the upper Rio Negro.

Prior to this study, H. minimus, proposed here as a member of the

inomatus complex (see below, Phylogenetic Hypotheses), was known from only

a few localities along the Tapajos River and in eastern Bolivia (Fig. 4). It has

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. now been collected and tape recorded throughout southeastern Amazonia east

of the Rio Madeira, and populations have been located along the west bank of the Rio Negro, near the Peruvian border in the Brazilian state of Acre (B. M.

Whitney, unpubl. data), and near Iquitos, Peru (J. Alvarez and B. M. Whitney,

unpubl. data). These last two localities are represented by specimens and tape

recordings, which I examined and which confirm their identification as

minimus; unfortunately, the tissues also collected from these specimens are housed in foreign collections and were not made available for this study (see

below, Phylogenetic Hypotheses).

Range extensions in the H. minor complex were found to the west, where

a population occurs just south of the Amazon River and east of the Peruvian

border (Fig. 3). This population is predicted to occur in Peru and also just north of the Amazon River into Colombia. It was found only in seasonally flooded

forest in blackwater tributaries of the Amazon ( igapo) and, based on this habitat

preference, voice (see below), and apparent contiguity of geographic

distribution, was assigned to H. m. pallens. The race pallens is known from comparable habitat throughout central

Amazonia west of the Madeira and Negro rivers, so my discovery of a population of H. minor west of the middle Madeira in non-flooded, upland

(terra firme) forest, and with a voice more like populations of snethlageae or minor

was unexpected. A museum specimen of H. minor that I found in a Bolivian

museum, misidentified as H. zosterops, from northern Bolivia, west of all the

major tributaries of the Madeira, and labeled as collected in terra firme forest, is

likely to refer to this same form. The late T. A. Parker's unpublished field notes also include frequent reference to H. minor in the same general area (see Fig. 3)

in terra firme forest. According to those notes, he tape recorded the birds, although no specimens were collected and the tapes have not been located.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In southeastern Amazonia, where H. minor occurs only in terra firme

forest and has a distinctive voice (see below), the species proved to be consistently absent from the region between the Tapajos and Xingu rivers, and

no museum specimens or literature references to the taxon from that region were located. A population of Lophotriccus galeatns occurs commonly in that

area, however, where it appears to occupy the same ecological niche as H. minor

and may simply replace it geographically (Zimmer et al. 1997). Also, no

member of the minor group has yet been confirmed from anywhere in

northeastern Amazonia east of the Rio Negro and north of the Amazon (Cohn-

Haft et al. 1997), where L galeatns also occurs commonly. These apparent lacunae in the distribution of H. minor appear to be real and not the result of

undersampling. No major range extensions were found for H. zosterops (Fig. 2).

Nevertheless, a previously unrecognized lack of specimens from the region

between the Madeira and Tapajos rivers may prove to represent a real

distributional gap. Because the species seems to be patchily distributed

throughout the Amazon and could not be found singing at certain times of year

at localities where recorded on other occasions, I believe that its distribution in

this part of the Amazon needs further sampling effort before the species is

considered genuinely absent. Vocal Variation

Tape recordings of vocalizations from all known populations of the

study taxa were examined. Based on the characteristics described below, I was

able to identify songs from all species complexes and classify them to

geographic regions. Because populations categorized by vocalizations did not necessarily correspond to named taxa, I refer to vocal-defined populations as

"song types."

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Songs could be assigned to categories based on a few simple guidelines described in the following classification:

Song Type 1— zosterops group This is a roughly 1-sec trill (series of similar notes), rising and then

falling slightly in pitch throughout, slowing toward the end, and preceded

immediately by a single note audibly distinct from the trill; individual notes can

be perceived, although they are delivered much too fast to count; "dry" tonal

quality. This song type occurs throughout the region east of the Branco and Negro rivers and north of the Amazon River, corresponding to the eastern

populations of nominate zosterops (Fig. 10). The type specimen of a taxon

named rothschildi (currently considered synonymous with zosterops; Traylor

1979) was described from this region, so the name rothschildi is tentatively

applied to members of this vocal type.

Song Type 2— zosterops group This is a short trill lasting roughly 0.5 sec and typically comprising 5-11 notes all on approximately the same pitch, although sometimes introduced or

ended with a noticeably higher note; slightly "metallic" tonal quality; notes are

delivered nearly slowly enough to count. This song type occurs throughout the

entire region north of the Amazon and west of the Branco and Negro rivers,

and south of the Amazon west of the Ucayali River; thus, it corresponds to the race flaviviridis, the western part of nominate zosterops, and all the forms found

in the region in between, which have not been assigned to any race (Fig. 10). Song Type 3— zosterops group

This is a very rapid, high-pitched, 2- or 3-note song, the first note lower-

pitched than the others; song differs from all other study taxa in that it is a

simple pair or triplet, not a trill. This song type occurs throughout the region

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Type 1 Type complex. Dots represent sample points. fMMMflMdAd A ft m n p f Type 3 Type Hemitriccus zosterops Type 2 Type Figure 10. Distribution ofAudiospectrogramsvocal types in all drawn to samescale.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. south of the Amazon and east of the Ucayali; thus, it corresponds to the subspecies griseipectus and rtanmburgae (Fig. 10).

Song Type 4— minor group This is usually an extremely rapid and typically long trill containing 35-

75 or more notes delivered in less than 1 sec and maintaining more or less

constant pitch; intemote intervals begin short and often diminish throughout

the song, causing an increase in delivery speed throughout the trill; notes

delivered so fast that they usually cannot be detected individually, but rather give the impression of a "buzz." Individuals just west of the Rio Madeira in

terra firme were recorded giving a slightly slower version of this song, and

localities just east of that river had a range of variation between the two, with

no obvious geographic pattern. Also, like individuals throughout southeast

Amazonia, they tend to call far more often than they sing. Thus, the entire range was classed as one type, although it may be the most variable song

category recognized. This song type occurs throughout southeastern Amazonia

to just west of the Rio Madeira; thus, it corresponds to the subspecies minor and

snethlageae and to the new population found at Humaita (Fig. 11).

Song Type 5— minor group This is a slower trill than the most examples of type 4, containing 14-30

notes in roughly the same time interval; this song type lacks as strong a

tendency to speed up throughout as similar-speed examples of the previous

type, and is given far more often than are calls; individual notes can be detected by ear, although delivered much too fast to count; this song resembles zosterops

song type 1 in speed, but usually lacks the discrete first note, is higher-pitched and more uniform in pitch and speed throughout, and typically is delivered in bouts of 3 or 4 songs in rapid succession (see also Cohn-Haft 1996). This song

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Type 4 'irxxvrvimcrimrr rm*r * * complex. Dots represent sample points. , Hemitriccus minor mu Type 6 MWMWMMMMMMMM a Type 5 Figure 11. Distribution ofvocalAudiospectrograms types in all drawn to same scale.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. type occurs west of the Negro and Madeira rivers in igapo and, as such, corresponds to the subspecies pallens (Fig. 11).

Song Type 6— minor group This is a long, fast trill, like type 4, but dropping in pitch throughout.

This song type corresponds to H. spodiops of the Andean foothills in Bolivia and

was described by Cohn-Haft (1996) (Fig. 11).

Song Type 7— inomatus group This is a distinctive and unique 2-part song, consisting of 3 introductory

notes given at a uniform and rather slow interval (roughly 0.1-0.3 sec) or with a

slightly longer interval between the second pair of notes, followed by a quick,

slightly rising trill of 3 or 4 notes; effect of this unusual rhythm is of a stuttered

trill; pure "tinkling" or "crystalline" tonal quality, unique to this species and

next, leading to my grouping them in the same species complex. This song type

occurs east of the Rio Negro and north of the Amazon, corresponding to the

species H. inomatus (Fig. 12).

Song Type 8 — inomatus group

This is a short trill of 6-12 notes of rather uniform pitch and speed; same "tinkling" tonal quality of the preceding, but lacking the two-part or stuttering

rhythm; sometimes rising or dropping slightly in pitch or slowing in speed

throughout, but with no detectable geographic pattern. This song type occurs

west of the Rio Negro and south of the Amazon; thus, it corresponds to H. minimus, including several newly discovered populations (Fig. 11).

Many characteristics used here to describe song types are not strictly quantitative or may show considerable variation. For example, I have been

unable to find any quantitative characteristic of audiospectrograms that

corresponds to tonal quality (e.g., "metallic"). What we perceive as tonal

quality may be a result of harmonic structure, the shape of the note, or a

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Type 8 complex. Dots represent sample points. Type 7 Hemitriccus inomatus Audiospectrograms all drawn to samescale. Figure 12. Distribution ofvocal types in 'sO

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. combination of the two. To further complicate analyses, these songs are so fast and composed of such short notes that extremely high-quality recordings are

necessary to yield analyzable spectrograms. Furthermore, differences in frequency and duration may be too subtle to be detected at computationally

efficient resolutions, and these variables can be affected by slight variation in

playback speed and other "noise" introduced by tape copying.

Despite the difficulties in quantitative analyses of these song

characteristics, discriminant function analysis (DFA) of 15 vocal parameters taken on 147 individual songs successfully placed >95% of songs to song type

(Fig. 13). Overall success was lower using DFA to class songs to named subspecies, reflecting the fact that vocal categories and current taxonomy differ

and that some taxa cannot be correctly diagnosed vocally.

These vocal types were determined a priori and then compared to

phylogenetic hypotheses to determine whether they contained phylogenetic

signal (see Discussion).

Phylogenetic Analyses Results of equally weighted parsimony and maximum-likelihood

analyses were generally concordant (figs. 14-17). In both analyses, the three

predefined species groups were reciprocally monophyletic. They differed from

one another by roughly 10% (raw sequence divergence; Table 3). The two

analyses also agreed in most of the clades identified within each species complex. The relationship between the two outgroup taxa could not be determined with certainty and had no effect on the relationships among the

ingroup taxa.

Under the HKY+G evolutionary model chosen, the single best maximum-likelihood tree (Fig. 16) is significantly more likely than any of the

others, including the maximum-likelihood bootstrap tree (P < 0.02, Kishino-

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. zosterops (97%) + Type1 (n=16) CSI

FACTOR 1

5 + + minor (98%) o Type 4 (n=21) CM *K x Type 5 (n=19) cc O C feO O o H + Type 6 (n=4)

-5L -5 5 FACTOR 1 inomatus (100%) X X X XX XXX Type 7 (n=9) XXX XXX XXX XXX Type 8 (n=29) X XXX 1 i§8_ IX XXX -20 10 SCORE 1

Figure 13. Discriminant function analysis classification of vocal types from 147 songs. (Raw data in Appendix 1.) Ellipses represent 95% confidence interval. Percent correct classification shown in parenthesis for each species complex, and sample sizes shown for each vocal type. For inomatus complex, discriminant function scores are plotted as frequency distribution along single axis, which accounted for 100% of data dispersion.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. zosterops complex zrottfiS zrothM minor complex zfla vtt inomatus complex ZZ08f\2 outgroups zzo stt3 zflavB zflav 10 zgris 8 zgris 7 zgris 6 zgris 5 zgris 2 zgris 3 zgris A z naum 1 m sneth S m sneth 6 m sneth 7 m sneth 8 m sneth 4 m minor 3 m minor 2 m minor 1 m sneth 9 m sneth 10 mSW 11 spodiops 12 m pallens 13 m pallens 15 m pallens 14 m pallens 17 m pallens 16 minimus 3 minimus 2 minimus 6 minimus 5 minimus 4 minimus 1 minimus NW 7 minimus NW 8 Inomatus 9 inomatus 10 Todiroatrum chrysocrotaphum Poocilotriccus latirostris

Figure 14. Parsimony consensus tree; equal character weighting. Strict consensus of 32 equally most parsimonious trees. Individual specimens labelled using abbreviated scientific name and specimen number (see figs. 6-8). Note, name rothschildi used here to distinguish H. zosterops specimens of vocal type 1 (see text). Regional descriptors (SW, NW) used to distinguish representatives of populations of unknown subspecific designation.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. zro th 14 zosterops complex zro th 15 minor complex zroth 16 zroth 17 inomatus complex zflav9 outgroups zflav 10 ztla v 11 z z o s t 12 z z o s t 13 100 zg ris 8 zg ris 7 zg ris 6 zg ris 5 100 zg ris 3 zg ris 2 zg ris A z naum 1 m sneth 5 m sneth 6 m sneth 7 m sneth 8 m sneth 4 m minor 3 m minor 2 m minor 1 100 m sneth 9 m sneth 10 mSW 11 m pallens 15 m pallens 13 100 m pallens 17 m pallens 14 m pallens 16 spodlops 12 minimus 3 minimus 2 minimus 6 minimus 5 minimus 4 100 minimus 1 100 minimus NW 7 minimus NW 8 Inomatus 9 Inomatus 10 Todlrostrum chrysocrotaphum Poecllotriccus latlrostrls

Figure 15. Parsimony bootstrap tree. Equally weighted characters, 50% majority rule, numbers indicate percent bootstrap support (of 1000 replicates) for each node.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. zro th 16 J"L zroth 14 — P zroth 15 *■ zroth 17 zosterops complex p zflav 9 minor complex • zflav 10 inomatus complex r zflav 11 "L zzo st 12 outgroups - z z o s t13 znaum l r zg ris8 r—P- zgrisl J L zgris6 — 5 changes 1— zgris 5 zgris A zgris 3 —I zgris 2 i m sneth 5 m sneth 6 m sneth 8 m sneth 7 L m sneth 4 m minor 3 - m minor 1 f m minor 2 m sneth 9 ML'm sneth 10 — mSW 11 m pallens 16 m pallens 17 m pallens 15 m pallens 13 m pallens 41 — spodlops 12 minimus 3 minimus 2 minimus 6 minimus 4 L' minimus 1 minimus 5 minimus NW 7 minimus NW 8 inomatus 9 H inomatus 10 Todlrostrum chrysocrotaphum Poecilotriccus latirostris

Figure 16. Single best maximum-likelihood tree. HKY+G model (see text). Branch lengths proportional to sequence divergence; scale bar represents 5 nucleotide substitutions.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. zroth 14 zroth 16 zroth 15 1 zroth 17 zosterops complex zflav 9 minor complex zflav 10 inomatus complex z flav 11 outgroups z z o s t 12 z z o s t 13 z gris 8 z gris 7 z gris 6 zg ris 5 z gris 3 zg ris 2 z gris 4 z naum 1 m sneth 5 62 m snet/i 6 70 m snef/j 7 m snet/i 8 72 m minor3 62 m minor 2 73 m minor 1 100 m sneth 4 73 m sneth 9 m sneth 10 m SW11 55 m pallens 15 62 72 m pallens 13 97 67 m pallens 17 m pallens 16 m pallens 14 spodiops 12 minimus 3 minimus 2 minimus 6 minimus 1 minimus 4 8 minimus 5 minimus NW 7 minimus NW8 g inomatus 9 inomatus 10 Todirostrum chrysocrotaphum Poecilotriccus latirostris

Figure 17. Maximum likelihood bootstrap tree. 50% majority rule, HKY+G model (see text), numbers at nodes indicate percent bootstrap support (of 100 replicates) for each node. Numbered bars right of specimen names indicate vocal types.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3. Percent sequence divergence (uncorrected p x 100) for selected pairs of operational taxonomic units (OTUs), representing major nodes in phylogeny. Ranges shown are of all possible individual comparisons at each node (number of comparisons in parentheses). All 1035 possible pairwise comparisons incorporated (from Appendix 4). For divergences along specific branches, see also Fig. 16, in which branch lengths are proportional to sequence divergence.

Pair of OTUs______Percent sequence divergence

outgroup: Todirostrum latirostre— Todirostnim chrysocrotaphum (1) 11.9 Todirostrum latirostre—ingroups (44) 10.7-13.2 Todirostrum chrysocrotaphum —ingroups (44) 11.5-12.9

among major clades (ingroup species complexes): inomatus complex— minor complex (170) 9.0-11.2 inomatus complex—zosterops complex (170) 10.6-11.6 minor complex—zosterops complex (289) 8.7-11.1

within zosterops dade: z. griseipectus/naumburgae— 2 . zosterops/flaviviridis/rothschildi (72) 2.4-3.8 2 . zosterops/flaviviridis—z. rothschildi (20) 1.0-1.7 2 . naumburgae—z. griseipectus (7) 1.3-1.7 2 . griseipectus—z. griseipectus (21) 0.0-2.3 2 . zosterops/flaviviridis—2 . zosterops/flaviviridis (10) 0.2-1.1 2 . rothschildi—z. rothschildi (6) 0.2-0.5

within minor clade: spodiops—m. pallens (5) 5.8-6.0 spodiops—m. snethlageae/m. minor/m. SW (11) 6.1-6.7 m. pallens—m. snethlageae/m. minor/m. SW (55) 5.8-6.4 m. SW—m. snethlageae/m. minor (10) 1.6-2.0 m. snethlageae/m. minor—m. snethlageae/m. minor (45) 0.0-1.5 m. pallens—m. pallens (10) 0.1-0.4

within inomatus dade: inomatus/minimus NW—minimus (24) 2.9-3.8 inomatus— minimus NW (4) 1.8 inomatus— inomatus (1) 0.0 minimus NW —minimus NW (1) 0.0 minimus—minimus (15) 0.0-0.8

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hasegawa test). Even under a parsimony model, it is not significantly less likely than the parsimony-derived phylogenies (P > 0.05, Kishino-Hasegawa

test). Nevertheless, some nodes in this tree received weak bootstrap support, so

the bootstrap likelihood tree (Fig. 17) is taken as the best, conservative working

hypothesis for phylogenetic relationships in the study organisms.

The most important difference between the parsimony and maximum-

likelihood trees is the paraphyly of the southern forms of H. zosterops

(griseipectus and naumburgae; vocal type 3) found in the maximum-likelihood

analysis (Fig. 16). Rather, various southern forms of H. zosterops were found to be the successive outgroups of the entire northern Amazonian clade. This

arrangement, however, received weak bootstrap support, collapsing into an

unresolved polytomy among southern forms and the northern clade (Fig. 17).

Parsimony analysis found strong bootstrap support for the same internal

relationships among individuals of griseipectus and naumburgae, but also

supported the monophyly of southern H. zosterops in 76% of bootstrap

replicates (Fig. 15). To test the likelihood of the monophyly of southern H.

zosterops, the same best maximum-likelihood topology (Fig. 16) was modified

only so that southern H. zosterops (specimens 1-8) were constrained to be

monophyletic and sister to the northern clade as in the arrangement found in

Figure 15. This topology did not differ significantly from the best tree under the maximum-likelihood criterion (P = 0.59, Kishino-Hasegawa test). Thus,

southern H. zosterops is just as likely to be monophyletic as it is to have the

relationship shown in Figure 16. Although I suspect that monophyly in this

group is correct (especially considering the vocal evidence; see Fig. 17 and

Discussion), I will consider both topologies in subsequent analyses.

In the zosterops group, northern forms are delimited by the Amazon

River for most of its length, but occur south of it west of the Ucayali River in

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Peru. Northern and southern forms differ by roughly 3% sequence divergence

(Table 3). The northern Amazonian clade consists of a monophyletic group in

the east, corresponding to a unique vocal type, but not recognized in current taxonomy (and tentatively referred to as H. z. rothschildi, see Vocal Analyses).

The remaining northern specimens form a monophyletic group in the

maximum-likelihood tree, corresponding to the northwestern vocal type. The

parsimony tree (Fig. 15) cannot resolve the relationship between the northeastern clade and the remaining northwestern specimens, expressing that

node as an unresolved polytomy. This is probably due to within-group

sequence divergences as high as those between this group and the northeastern

vocal type (> 1%; Table 3). In any case, neither method supports the distinction

between the Peruvian subspecies flaviviridis and nominate zosterops. Rather, the

topology suggests gene flow among distant regions of the northwestern group,

concordant with the vocal classification. The H. minor complex consists of three clear clades differing from each

other by roughly 6% sequence divergence: spodiops, pallens, and a group

containing all the remaining southern Amazonian terra firme populations. The

relationship among the three differed in the different analyses (figs. 14-17), but

in no case was H. spodiops found to be basal within the complex, suggesting that

H. minor as currently defined is not monophyletic. Within the southeastern

clade, the subspecies snethlageae was also found not to be monophyletic,

concordant with the lack of vocal distinction between it and nominate minor.

The specimen collected west of the middle Rio Madeira came out as the sister to

all other southeastern minor. This result is consistent with the predictions based

on habitat and voice, but that individual's considerable genetic divergence from specimens just across the Madeira suggests a fairly old split. More samples will

be necessary to confirm this form's status as representative of a unique lineage.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In both phylogenetic analyses, the inomatus group consisted of three tight dades. Northern and southern Amazonian populations are split by roughly 3% sequence divergence. The northern dade contains two groups on

opposite sides of the Rio Negro, differing by 1.8% sequence divergence.

Within-group variation is extremely low, with less than 0.8% sequence

divergence within all of southeastern Amazonia, and identical sequences

between pairs of individuals within the same interfluvium (i.e., specimens 2-3,

7-8,9-10; compare figs. 8 and 16). The vocal distinctiveness of inomatus parallels the evolutionary hypothesis based on DNA sequences, but neither the

genetic distinctiveness of northwestern Amazonian minimus nor its sister

relationship to inomatus were predicted by vocal analysis (Fig. 17). Based on

plumage coloration and voice, I applied the name minimus to a paraphyletic

group. This suggests that the shared characteristics I used to join the two

genetically distinct forms are ancestral in this complex (see Discussion).

Note that the uncorrected sequence divergences here are conservative

estimates of genetic difference (Fig. 18). Models, such as the maximum-

likelihood model used, that take into account the increased likelihood of

multiple changes (saturation) at the same nucleotide site over time yield much higher genetic differences among more highly diverged taxa.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. without prohibited reproduction Further owner. copyright the of permission with Reproduced

corrected equality between corrected and uncorrected values. uncorrected and corrected between equality evolutionary model used in this study (see text) text) (see study this in used model evolutionary raw sequence divergences (uncorrected p). Line represents represents Line p). (uncorrected divergences sequence HKY+G raw under distances genetic pairwise Corrected 18. Figure 0.0 0 0.3* 0.4- 0.5-1 . 2 0.00 - Pairwise genetic distances genetic Pairwise 0.05 uncorrected p uncorrected 60 HKY+G 0.10 vs. corresponding corresponding 0.15 DISCUSSION

Distribution and Taxonomic Limits

Recognition of the three major clades hypothesized in this study

(zosterops, minor, and inomatns complexes; see Methods) was strongly

supported in all molecular phylogenetic analyses. Furthermore, preliminary

analyses of mtDNA sequences from other Hemitriccus species (unpubl. data)

reinforce these groupings and suggest that other species once taxonomically confused with the study taxa (e.g., H. striaticollis , H. flammulatus, and H.

margaritaceiventer) are indeed phylogenetically distinct and differ from the

study taxa and each other by roughly the same amount of sequence divergence.

The concordance found at this level of classification between current

taxonomy and molecular phylogenies encourages confidence in both data sets. This result in itself is rather remarkable, considering that much of our current

classification of these taxa is based on subtle differences in external morphology, almost entirely determined from small samples of relatively old

and data-poor specimens analyzed by taxonomists who never set foot in the

natural habitats of these creatures or observed the birds alive. The general

accuracy of their classifications is a tribute to their skill. At the same time, it is clear that, with such poorly known and difficult to distinguish organisms,

addition of any new information regarding distribution, voice, and other

natural history traits derived from modem field work promises to improve steadily on earlier hypotheses. The addition of spodiops to the minor group

(Cohn-Haft 1996) and my current proposal of an inomatus-minimns dade, both

based on morphology and voice, are examples. What could not have been predicted from current taxonomy or modem

field observation is the extraordinarily high genetic divergence among such

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. phenotypicaUy conservative taxa. On the contrary, the expectation of this study was that, if any sympatric Amazonian taxa should be genetically close or of

relatively recent origin, then these congeneric flycatchers with their nearly

identical morphologies, history of taxonomic confusion, frequent

misdassification in collections, and apparent ecological similarity, would be

strong candidates. Remarkably, the roughly 10% raw sequence divergence

among them is equivalent to intergeneric divergences in most birds studied

previously (Johns and Avise 1998), induding Neotropical blackbirds (Lanyon

and Omland 1999). Within a Neotropical genus, the sympatric Amazonian

tanagers Ramphocelus carbo and R. nigrogularis, which are easily distinguished

morphologically, vocally, and ecologically, differ by only 3.4% sequence

divergence (Hackett 1996). Only potoos (Nyctibiidae) have been found to have

higher intrageneric sequence divergences (11.1-16.2%; Mariaux and Braun

1996). Although a strictly distance-based notion of genus might lead to the

recognition of each major dade here as a separate genus, I do not recommend

such a revision. The ecological and phenotypic similarity of all the study taxa

argues for their maintenance in Hemitriccus (see Sheldon and Winkler 1993).

The implications of these unusually great sequence divergences for the historical biogeography of the Amazon will be discussed further below.

Much as the interspecific divergences found here correspond to typical

measures of intergeneric divergence, sequence divergences among "subspedes"

or geographical forms (1.0-6.4%) within each major dade are comparable to

those found among congeneric spedes in the temperate zone (Johns and Avise 1998). Elevation of the major geographic subunits to full spedes would be

consistent with both a phylogenetic spedes concept (Cracraft 1983, McKitrick

and Zink 1988), given redprocal monophyly of vocally or morphologically

diagnosable groups, and with a biological spedes concept (Mayr 1969), given

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. no evidence of recent gene flow, the presumed importance of vocalizations in mate selection, and negative response to playback of songs among vocal groups

(unpublished data). Relatively high genetic divergence coupled with subtle or cryptic

morphological distinctiveness is proving to be the rule in the few studies of South American organisms to date (e.g. Bates et al. 1999). It is probable that

further genetic studies will lead to a major reclassification and a stunning

increase in the recognized species diversity of South American birds. It was not

the purpose of this study to revise the taxonomy of Amazonian Hemitriccus

flycatchers, and such a revision must await further data, including results of

ongoing sequencing of more individuals and of vocal and molecular analyses of

populations from type localities. However, likely results of such study include recognizing the taxa rothschildi, griseipectiis, and pallens as full species,

synonymizing flaviviridis with zosterops and snethlageae with minor, and naming

the northwestern Amazonian population of H. minimus as a new species. This

would result in no net change in the number of recognized study taxa (see

Table 1), but would increase the number of species-level taxa from five to ten.

In sum, although the monophyly of taxa recognized as species under

current taxonomy was supported, detailed field work and vocal and genetic

analyses led to substantial revisions of within-spedes patterns, major extensions of known distributional ranges, discovery of probable new taxa, and

recognition of higher than expected levels of genetic differentiation. There is no

reason to expect a qualitatively different situation from other Amazonian bird

taxa to be studied (see Bates et al. 1999). Vocalizations versus DNA sequences

For the most part, vocally defined groupings were concordant with

dades identified by molecular phylogenetic analyses (Fig. 17). In several cases,

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vocalizations either enabled recognition of genetically distinct groups not

previously recognized under morphology-based taxonomy (H. z. rothschildi), or

accurately reflected the lack of genetic distinctiveness of taxa that should be considered invalid (H. z. flaviviridis and H. m. snethlageae). Phylogenetic

structure within vocal groups discernible using molecular markers also was

unrecognized under current classification. Thus, overall lack of phenotypic

variation (in both voice and morphology) can mask genetic differentiation.

However, the evolutionary significance of such entirely cryptic genetic

differentiation remains to be explored.

In at least one case, that of the newly discovered populations of H.

minimus in northwestern Amazonia, both vocalizations and morphology

suggest lumping non-sister taxa on the basis of what are most likely shared

ancestral traits (Fig. 17). This suggests that vocal traits suffer from the same

difficulty in determining polarity as do other phenotypic characters, and argues

in favor of basing phylogenies on molecular data. The molecular phylogenies

can then be used to examine the evolution of vocal characters. This is likely to

be a fruitful line of research that I intend to pursue using this same data set.

Although vocal analyses missed some genetic diversity and suggested

one paraphyletic taxon, in no cases did vocalizations lead to the recognition of

polyphyletic taxa. This conservative characteristic of the vocal classification

suggests that vocal differences are reliable indicators of genetic distinctiveness,

whereas lack of vocal distinction may not necessarily indicate lack of genetic

distinctiveness. Thus, vocalizations in Hemitriccus flycatchers appear to be powerful genetic markers. The same has been found to be true in other tyrant

flycatchers (Johnson 1980) and is probably the case for most subosdnes and

birds in general (e.g., McCracken and Sheldon 1997).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The strong correlation between vocal type and haplotype (based on the

molecular phylogeny) is consistent with the proposition that flycatcher songs

are genetically hard-wired. Correlation does not, of course, demonstrate

causality and no one would argue that cyt b codes for song characteristics. More direct tests of song heritability in suboscines should be performed before

the generality of results from a few flycatchers (Kroodsma 1984) is accepted as

dogma. However, there is as yet no reason to doubt that all flycatchers have

inherited songs, and as long as patterns of vocal geographic variation reflect

molecular phylogenies, it is probably reasonable to assume that other

suboscines do as well. Area Relationships and Historical Biogeography

By substituting the taxa on the molecular phylogenetic trees with their

regions of geographic distribution (area cladograms), it is possible to hypothesize the historical relationships among geographic subregions of the

Amazon. If all of the three widespread Amazonian clades in this study show

the same or highly concordant area cladograms, then this supports a single

shared evolutionary history among them. In the case of these particular study

taxa, it is reasonable to expect concordance assuming that the widespread

ancestors of each major clade were all subject to the same historical events and

that their morphological and ecological similarity would lead them to respond

similarly to those events. On the other hand, seemingly subtle ecological

differences among the three groups, perhaps the same differences that allow

them to coexist sympatrically in the present (e.g., in microhabitat preferences),

could cause them to respond differently, tracking the different historical

trajectories of their respective preferred niches. Each of the three species complexes has a unique area dadogram, even

taking into account possible ambiguities in the original molecular phylogenetic

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hypotheses (Fig. 19a-d). The two most similar patterns of geographic

divergence are shared by the zosterops and inomatiis groups, if the former's griseipectus/naumburgae forms are assumed to be monophyletic (Fig. 19a,c). In that case, both groups show a basal split between northern and southern forms,

divided by the current course of the Amazon River. Northern forms

subsequently split on opposite sides of the lower Rio Negro, although the

division higher upstream differs between the two clades (figs. 10,12). In

southern Amazonia, the zosterops group appears to differentiate on opposite

banks of the Rio Madeira, although no vocal or morphological distinctions have been described. The westernmost populations of the inornatus group, from

Iquitos, Peru, and Acre, Brazil, are of unknown phylogenetic affinities. For

them to follow the pattern found in zosterops, the Acre population should belong to the southern Amazonian group, as sister to the southeastern clade,

and the Iquitos population (from north of the Amazon) should pertain to the

unnamed northwestern clade. Unfortunately, the vocal evidence is unclear, and

no tissues from these specimens are yet available for molecular analysis.

If the topology taken from the best maximum-likelihood tree (Fig. 16) is

assumed to be correct, then the zosterops group shows a complex geographic

pattern of differentiation (Fig. 19b). In this case, the two northern Amazonian clades retain their sister relationship, but the sister to them is the Atlantic forest

form, mumburgae. This entire group, in turn, shares a common ancestor with

southwestern populations of griseipechis, which all share common ancestors

with a succession of more easterly populations of griseipectiis. Not only does

this arrangement have weak bootstrap support (Fig. 17), but it suggests a

geographically complex series of evolutionary events, possibly involving long

distance dispersal and colonization, extinction of intermediate forms, or

"budding" from ancestral populations. Furthermore, the phenotypic

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NW (Imerf)

NW (Napo) NE

SE (Pa 2) SW

SE (Pa 1)

zosterops SE (Pa2) S + Atlantic A. SE (Pa1) SW NW or Atlantic NW NE NE

inomatus minor SE SE C. D. NW ■ Andes

NE ■ SW + NW

E. Passeriformes F. Tyrannidae (Bates e t al. 1 9 9 8 ) (Bates e t al.1998)

Atlantic Atlantic SE NW (Imerf) NE NW (Napo) NW (Imerf) SW NW (Napo) SE

Figure 19. Area cladograms based on phylogenetic analyses of this study (a- d; see text) and distributional data of Bates et al. (1998) (e,f). Map inset depicts major divisions of Amazon and adjacent regions as referred to in this study, and subdivisions of Bates et al. (1998) (marked with dashed lines, labeled in parentheses).

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. intermediacy of northwestern zosterops populations, both in terms of number of song notes and ventral coloration, between southern and northeastern forms

argues in favor of a clockwise direction of differentiation events from south to

northwest to northeast, directly conflicting with this topology.

For these reasons, I suspect that the vocally and morphologically

uniform southern Amazonian and Atlantic forest populations (griseipectns and

nawnburgae) form a monophyletic group, sister to the northern populations, as described by the parsimony trees (figs. 14,15). Opinions aside, however, if the

best maximum-likelihood tree does describe accurately the relationships among

southern populations of zosterops, then the area relationships are even more

distinctly different from those found in the other species complexes than described in the first scenario.

Regardless of the preferred topology chosen for the other complexes, the

minor group shows a unique phylogeographic pattern (Fig. 19d). In particular,

the presence of a western Amazonian clade including both sides of the upper

Amazon River with no genetic structure on opposite banks is novel among the

study taxa. The sister relationship between this western group ( pallens) and southeastern Amazonian conspedfics is not supported by any analysis. Thus,

Andean H. spodiops cannot be considered basal, and the possibility that it arose

from an Amazonian ancestor must be considered. This is another unique

characteristic of this group.

Not only do these three dades of Amazonian flycatchers show distinctly

different area relationships, but the relationships found here do not resemble

those described in other studies. Bates et al. (1998) estimated the historical

relationships of Neotropical areas of endemism based on the presence of shared

taxa. The phylogenies of none of my study taxa resemble either their consensus

tree for all Passeriformes nor their Tyrannidae tree (figs. 19e,f). This supports

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. their contention that there may be no single set of historical events that affected all Amazonian birds equally. However, it is not dear, even using their

methodology, whether they would have found the same area relationships they

did for tyrannids, had they used my new distributional information. As Bates

et al. (1998) themselves point out, use of phylogenies is the preferred means of

analyzing area relationships. Only now are phylogenies beginning to be

produced. Age of Taxa

Assumptions of a universal molecular dock are not recommended, based

on empirical variability in evolutionary rates, and are usually not very

informative when taking into account the large confidence intervals

surrounding them (Hillis et al. 1996a). Nevertheless, assuming relatively constant rates of evolution within Hemitriccus (as suggested by branch lengths

in Fig. 16), then the oldest major within-clade speciation events, assodated with

distinct vocal types or named taxa, are five times older than the most recent

ones. The time scale widens further when taking into account the conservative

nature of uncorrected sequence divergences (Fig. 18). Thus, it is probable that major spedation events did not all occur at the same time; rather, they occurred hierarchically, even within superspedes, with

long time lags between major events. Furthermore, if evolutionary rates prove

to be similar between Amazonian and north-temperate avifaunas, then it would

appear that these Amazonian birds are of relatively andent origin and that the

same historical events did not affect both faunas. Chesser (2000) assumed a 2% per million year divergence rate in mtDNA of flycatchers in the South American genus Muscisaxicola. This led him to the

condusion that most members of that genus, many of which are sympatric,

spedated during the mid- to late-Pleistocene. If this same rate is applied to

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hemitriccus flycatchers, then sympatric congeners (the "species" of current

taxonomy) would have spedated at least five million years ago, and geographic

replacement forms split during the mid- to early-Pleistocene at the latest, and

members of the minor group split some three million years ago. In any case,

even the most recent splits do not cointide with late Pleistocene.

If these geographic forms are indeed as old as temperate or Andean

species, then the interesting questions are, "why have they diverged so little

phenotypically?" and "what is keeping them from shifting niches and occurring

in sympatry with sister taxa after all this time?" One possible explanation is

that strong stabilizing selection (due to niche-packing in this very species-rich

tropical community) maintains phenotype and niche virtually unchanged over

time. Geographic replacement forms are unable to shift into new niches and are

unable to coexist due to competitive exclusion. Thus, they are true biological

species but maintain allopatric distributions. This possibility will be explored

further below. Support for Historical Spetiation Hypotheses

A number of hypotheses have been put forward to explain patterns of differentiation in the Amazon basin, although none has been tested rigorously.

Below I will consider each of the major hypotheses in light of the patterns of

spatial and temporal variation found in this study. Throughout this discussion, I will distinguish between the generation and maintenance of differentiation,

although the various hypotheses may not have explicitly recognized this

distinction in their original formulations.

River Barriers The oldest and perhaps most straightforward explanation for

differentiated taxa occurring on opposite banks of rivers is that the original

formation of those rivers was the vicariance event that split widespread

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ancestral populations into isolated interfluvial populations (Wallace 1853; Sick

1967; Willis 1969; Capparella 1988,1991). This mechanism requires that the

rivers become complete barriers to gene flow at some time in the past and

continue to be so up to the present; thus, the same process that generated differentiation is what maintains it.

Although intuitively it may seem unlikely that rivers ever completely

divide avian populations (see Introduction), lowland tropical rainforest birds

may be sedentary or have unusually poor dispersal abilities. In particular, it

may be postulated that terra firme forest understory species should be more

susceptible to river barriers than canopy- or edge-inhabiting species, which are

adapted to more open environments and so less averse to crossing open spaces,

and that species of riverine habitats should be the least influenced by river

barriers (Capparella 1988,1991). The river barriers hypothesis predicts that all rivers that completely

dissect a taxon's distribution and are above a certain (taxon-spedfic) threshold

width for their entire length should separate differentiated forms on opposite

banks. Thus, although some smaller rivers might be barriers to certain taxa and

not to others, any taxon divided by a small river must also be divided by the

larger ones. It follows from this, then, that there should be a hierarchy of river

importance as barriers. It is not necessary to measure the width of all rivers to

establish this hierarchy. Rather, when comparing any two taxa, the rivers

subdividing one into distinct forms must indude all those dividing the other or

be a subset of them; there can be no non-overlapping sets of river barriers.

Capparella (1988,1991) found cases in which apparent phenotypic identity on opposite river banks masked hidden genotypic differentiation, and he

suggested that this might be a common phenomenon, explaining the numerous

apparent exceptions to this corollary.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Another prediction of the river hypothesis is that splits among allopatric

forms should date to the formation of the rivers, presumably concurrent with

the uplift of the Andes mountains and considerably predating the late Pleistocene—thus, broadly speaking, somewhere between 1-10 million years

before present (Capparella 1988,1991). It follows from this, then, that

numerous taxa exposed to the same series of vicariant river-formation events

should show the same phylogenetic branching pattern, unless many important

rivers formed at roughly the same time, in which case the intemodes between

successive forks on the phylogeny should be relatively short. The three taxa studied here appear to violate these predictions in several

instances. Although the zosterops and inomatus groups show reasonably similar

geographic distribution patterns and remarkably similar patterns of branching

and genetic divergence, they differ in the exact location of the split between

northern clades. Assuming vocal differences correlate with genetic differences

as they appear to, then the zosterops group does not split across the upper Rio Negro, as does the inomatiis group, but does split somewhere in the vicinity of

the Rio Branco, where the inomatus group apparently does not.

The minor group, in turn, shows a rather different pattern from both the

other groups in several aspects. Beside the absence of any form in the northeast

and the unique presence of an Andean form (both of which could simply be explained by regional extinctions in this and the other species groups,

respectively), the most striking distributional difference is in the western form,

pallens, that extends to both sides of the upper Amazon (Solimoes) River

without any indication of genetic or vocal differentiation on opposite banks.

This immense river is a clear boundary in the zosterops group and is of

undetermined importance for the inomatus group. Also, it is known to separate clearly differentiated phenotypes in numerous (perhaps most) Amazonian taxa

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (see, for example, Haffer 1978, Cracraft 1985, Bates et al. 1998). Nevertheless, the fact that this form inhabits flooded riverine forest (Table 1) may explain its exceptional lack of differentiation across the upper Amazon, as allowed by the

river barrier hypothesis. The presence of an Andean foothill form, spodiops, in the minor group

implicates spedation agents other than river barriers because no major river

separates the Bolivian yungas, home to spodiops, from the ranges of pallens and

snethlageae. Although this kind of differentiation may represent an entirely

different phenomenon from that of intra-Amazonian differentiation, the

comparable sequence divergences between the two Amazonian forms and

between each of them and spodiops suggests that all three split roughly concurrently. In that case, there may have been a complex interplay between

Amazonia and the adjacent Andean foothills not induded in the river barriers

hypothesis. Furthermore, this geographic pattern is shown by several bird taxa

(Cohn-Haft 1996), so it may represent a general phenomenon.

Finally, the six-percent sequence divergences in the minor group are

nearly twice those of the deepest splits in the other two groups and five times

greater than their most recent splits. This suggests that minor not only did not

suffer the two distinct phases of vicariance seen in the others, but that differentiation in the minor group happened at a different time, perhaps much

earlier than in either zosterops or inomatus.

Among the three spedes complexes in this study, there is no strong

evidence of concordance in either the distributional relationships to rivers or

the historical differentiation events, both of which would be expected under the river barriers hypothesis. Thus, strictly speaking the river barriers hypothesis

must be seen as failing in this case. Modifications of the strict hypothesis can be

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. envisioned, but those involve different historical assumptions and are much

more difficult to test (see below). Pleistocene Refugia Haffer's (1969) explanation was considerably more complex than the

river barriers hypothesis. He proposed that, during the glacial maxima of the

Pleistocene, global cooling led to drier climates in the Amazon, causing humid

forest throughout the region to be replaced by savanna vegetation. Forest

survived only in isolated pockets (called "refugia") in the wettest areas, where populations of forest organisms survived in isolation. When refugial forests

expanded to recover the Amazon during the wet intergladals, their resident

organisms spread along with them. However, expanding populations

encountering rivers were prevented or deterred from crossing at the widest

stretches, but eventually crossed at narrower places when they arrived there. In

those cases in which refugial populations had differentiated sufficiently in isolation not to interbreed upon secondary contact, but without any ecological

niche shift, they would exclude each other geographically by competition. The

observed pattern of opposite bank replacement forms could exist due to the

inability of newly differentiated and expanded populations to cross rivers, or

due to their inability to establish themselves on the other bank, owing to the

presence of a competitor there. This explanation allowed for Haffer's

observation of numerous taxa whose distributions were partially delimited by

big rivers, but which appeared to "leak across" the narrower upper reaches or

headwaters regions. This kind of imperfect association of distributional boundaries with

rivers is one prediction of the Pleistocene refuge hypothesis. Haffer also made

specific predictions about the locations of presumptive refugia; however,

because bird distributions are expected to have expanded from them at

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. different rates to fill newly available post-refugial forest habitat, the accuracy of

those predictions cannot be tested directly by the modem bird distributions and is not critical to the success of the model in general. Nevertheless, a certain

amount of geographic concordance should occur across taxa in their centers of

endemism, although their exact distributional limits might vary. Another

prediction is that differentiation should date to the glacial maxima of the

Pleistocene (all roughly within the last two million years), especially the latest

ice age (approximately 18,000 years ago), which would have left the most recent

and so most recognizable mark on the genetic structure of populations. Thus,

as in the previous model, branching patterns should be concordant due to the

widespread and simultaneous nature of the forest change believed to have

caused differentiation. In contrast to the river barriers model, however, genetic

divergences should be relatively low, due to the very recent origin of most

differentiation. The results of this study uphold those predictions with mixed success.

The distribution of nominate zosterops on both sides of the upper Rio Negro,

which was problematic under the river barriers hypothesis, could be

interpreted as the kind of "leakage" or secondary dispersal across headwaters expected in this model. In general, the presence of roughly concordant

distributions of members of the zosterops group compared with the inomatus

group fits the model, and the concordance of branching patterns and genetic

divergences between those two groups is suggestive of a shared history of

vicariance (consistent with both models). It is unclear whether the rather different geographical pattern of differentiation found in the minor group is

acceptable under this model, but, in practice, Haffer (1974,1978, and thereafter)

was willing to accept distributional patterns far more divergent from this one as

reasonable results of his proposed mechanisms. The different branching

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pattern of the minor group phylogeny relative to the other two groups suggests a different evolutionary history, but again it is unclear whether that is acceptable within the expectations of the hypothesis. The most contradictory

result of the current study with respect to the Pleistocene refuge hypothesis is

that the degree of genetic divergence found appears to be far greater than one

would expect from late Pleistocene differentiation (see above, Age of Taxa).

Rivers and Refugia Without any clear, direct evidence for the existence of forest refugia amid expanses of savanna in the Pleistocene Amazon, another interpretation of

paleoecological data from the Amazon's perimeter is that Pleistocene climate

fluctuations caused the ecotone between humid Amazonian forest and adjacent

plant communities to move back and forth over a limited region. This scenario

proposes a relatively constant presence of rainforest throughout most of the Amazon basin, but with a retreat away from the perimeter during glacial

maxima when adjacent montane and savanna plant communities invaded the

region from the sides (Ayres and Clutton-Brock 1992). An effective shrinkage

of the lowland rainforest toward the center of the region and away from river

headwaters would strengthen the river-barrier effect by withdrawing bird

populations from areas of narrower rivers. In effect, then, during these periods

rivers would become more complete barriers, turning interfluvial regions into

refugia. Haffer (1997a) dubbed this the "river-refuge hypothesis." Although this hypothesis paints quite a different scenario for Pleistocene

Amazonian land cover, the evolutionary results might be virtually

indistinguishable from the previously outlined refugia hypothesis. Thus, my

results support this hypothesis equally well (or poorly). Future analyses of

much larger population-level samples, using coalescence theory (Kingman

1982, H udson 1990) might determine whether Pleistocene populations were

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. restricted into small refugia from which they later expanded outward over much of their present range, or whether they were simply more strictly isolated

temporarily in areas roughly the size of their current distributions (river

refugia).

Islands

Nores (1999) proposed that sea-level rises of 100 m or more during the

late Tertiary and Quaternary would have flooded much of the Amazon basin,

leading to the formation of islands on higher ground. Just as proposed for

refugial islands in a "sea" of savanna, these true oceanic islands would have

facilitated vicariant speciation of forms which would subsequently recolonize

reforested areas after sea level dropped. Once again, although the historical

scenario is radically different from others proposed, the likely outcome in terms

of distributional patterns is indistinguishable. Depending on the size and

distribution of these islands, the specifics of phylogenetic branching patterns

might differ from other hypotheses (although no hypothesis has made explicit

predictions regarding the topology of area cladograms), but the expectation of

concordance in phylogenetic patterns would be equivalent to that of the others.

The only critical difference is that these island-differentiated forms could be

older than the Pleistocene, which seems to be consistent with present data.

Tectonic Arches Contact zones between sister taxa of some small Amazonian mammals

have been described that coincide with tectonic arches formed during the uplift

of the Andes (Silva and Patton 1993,1998). These arches are presently

recognizable only as subterranean features completely covered by more or less uniform vegetation types. The authors of this observation make no specific

proposal of mechanism, but suggest that the geographic coincidence may be

due to some as yet undetermined mechanism related to the formation or

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. subsidence of these arches. The arches cut across the Amazon in roughly parallel diagonals, intersecting some rivers, such as the Jurua and Jau at right

angles.

Although the possibility was not addressed directly in my sampling

design, distributional patterns of the bird taxa studied here do not coincide at any point with the mammal patterns or the arches. On the other hand, because

birds are probably better dispersers than many small mammals, and better able

to track suitable habitat, it is unlikely that their current distributions would

coincide with these arches in the absence of existing habitat discontinuities

there. Thus, the possibility that birds and mammals originally differentiated

similarly under the influence of the same undetermined process no longer

detectable in the bird populations cannot be eliminated.

Climatic Cooling and Novel Plant Communities

Colinvaux (1987,1993,1996) argued that the Pleistocene Amazon may

not have dried significantly and was probably continuously covered by humid

forest. However, during glacial maxima the climate cooled sufficiently to affect

plant communities in peripheral and elevated parts of the basin. It is likely, he argues, that plant species assemblages under such modified climatic conditions

have no modem analogues and that vicariant spedation may have occurred in

ephemeral islands of cold-adapted forest amid the sea of basically unchanged

lowland humid forest (Colinvaux 1998). One prediction from this scenario is that most Amazonian fauna should

not have differentiated during the Pleistocene, due to the constant cover of

typical lowland forest. This is consistent with the apparent pre-Pleistocene

divergences found in my study. Another prediction is that peripheral endemic

taxa might be of Pleistocene origin and be cold-adapted. The existence of the

Andean foothill form of the minor group, H. spodiops, and of other

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. biogeographic links between the lowlands and the foothills (Cohn-Haft 1996) may be due to such a mechanism. However, exactly how this would work is

unspecified, and a Pleistocene age of divergence seems to be counterindicated in the case of the Hemitriccus studied here. On the other hand, as pointed out by K.-b. Liu (pers. comm.), long

periods throughout history dominated by novel and changing plant communities associated with fluctuating warm and cold climates might also be

expected to alter the demography of specific bird populations in unique and

unrepeatable ways. This could enhance the isolating effects of natural

demographic fluctuations as described below (see below, Demographic

Stochasticity Hypothesis).

Ecological Gradients

Concerned about the lack of alternatives to the Pleistocene refuge

hypothesis, Endler (1982) proposed a sort of null hypothesis. He suggested that

current ecological conditions might be sufficient to explain observed patterns of differentiation. The mechanism he proposed to generate such diversification

was natural selection along ecological gradients. Regional environmental

differences select for different traits in widespread inhabitants, leading

eventually to speciation among populations without major vicariance events.

Endler explained this parapatric speciation model in considerably more detail

earlier (Endler 1977), but the process is still not credited with playing an

important role in terrestrial vertebrate evolution, for which geographic isolation

is believed to be the sine qua non of genetic differentiation. Even if such a mechanism could work for birds (as proposed for some

African rainforest taxa by Smith et al. [1997]), there is no evidence that it has in

my study taxa. First, no consistent environmental differences have ever been

proposed among the traditionally accepted areas of Amazonian endemism. In

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fact, their apparent uniformity of habitat and dimate is remarkable, belying the need for any geographic variation at all (see Introduction). Furthermore, in the

Hemitriccus studied, two of the three spedes complexes (zosterops and inomatus)

show no evidence of niche shift or habitat differences among any of the

populations within a complex (Table 1).

Even allowing for more subtle environmental differences than I would

have detected in this study, could these differences have led to the

differentiation observed? Under this model, differences among populations

should have environmentally adaptive significance. However, it is not at all

clear that the phenotypic traits known to vary geographically in these birds are under any selective pressure whatsoever. Yellow vs. white ventral coloration,

short vs. long songs, or fast vs. slow trills appear unlikely to confer any

particular survival advantage in subtly distinct environments. More likely they

are traits that can change easily under stochastic population-genetic processes

without carrying any selective burden at all. Nevertheless, they may be subject to sexual selection (see below).

Finally, phenotypic variation might be expected to show similar regional

characteristics across taxa. For example, if slightly drier climate in one part of a

species' range selects for paler plumage coloration, then that region's

representative of other species might also be expected to be paler than their

conspedfics elsewhere. So, in this study, if we allow for the possibility of subtle undetected environmental differences among regions and real (but

unrecognized) selective pressure on these traits, then we might expect the traits

to show consistent geographic patterns across study taxa. However, we don't

see any such pattern (Fig. 20). There appears to be no geographic pattern to ventral coloration or song characteristics, suggesting that there is no regional

selection on these traits.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Trill speed Song duration Ventral color inomatus minor zosterops Figure 20. Geographic distribution ofvariable phenotypic traits in Amazoniantrill speed representatives is slow (-) to fast of each (+); in all traits, is intermediate. ”0" Shaded quadrant indicates region where study species complex. quadrants. Ovals are schematic Ventral maps ofcolorAmazon ranges basin from dividedwhite (-) to yellowspeciesinto NE, group song (+); NW, SW, and duration does SE not is occur.from short (-) to long (+); oo

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Alternative and Null Hypotheses None of the hypotheses suggested to date appears to be a perfect match

for the results of this study. Yet it is also difficult to reject most of them. In some cases this is because the historical scenarios hypothesized were too vague

to permit specific predictions. On the other hand, other distinct historical

scenarios lead to identical predictions for current patterns. This is a more

serious problem and calls into question the value of further study in this area.

In fact, I will propose here that virtually any speciation model proposed could

lead to the same distributional patterns found today.

The brilliance of Haffer's Pleistocene refuge hypothesis was in dissociating the mechanisms of generation and maintenance of diversity. Just

because enormous rivers help maintain the distributional patterns of differentiated forms does not mean those rivers caused the differentiation in the

first place. In fact, as shown in this study, it appears unlikely that a strict river barriers hypothesis is correct. But nothing in Haffer's vision of how currently

observed distribution patterns are maintained requires that they be generated

originally by Pleistocene refugia per se. (See Burbrink et al. [2000] for a case in

which Pleistocene refugia are believed likely to have generated modem river-

delimited species boundaries in North American rat snakes.) Haffer's model for the maintenance of diversity and its geographic patterns relies on secondary contact among variously differentiated forms, lack of ecological differentiation

(in some cases) leading to competitive exclusion, and large rivers as

inhibitors—but not necessarily complete barriers—to dispersal: all

characteristics that exist today. If Haffer was right about how these patterns are maintained, and I suspect he was, then I predict that any mechanism of genetic differentiation proposed and assumed to run in today's Amazon (under

contemporary environmental conditions) will produce the same geographic

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. patterns observed. This prediction could be tested with a spatially explicit model including different dispersal rates, competition coefficients, habitat

suitabilities, and "traversabilities" of river barriers, all of which are constant in

time and run under various spatial and temporal initial conditions of speciation.

One can easily invent an infinity of speciation "just-so stories" to propose as hypotheses to explain Amazonian diversification. For example, perhaps a

now-extinct herbivorous mammal colonized the Amazon basin consuming vast

amounts of vegetation. Before eating itself to extinction, it radically changed

the distribution of plant communities and caused vicariance in the other animal

inhabitants of those communities. After the mammal disappeared, forest

recovered, and newly spedated forms recolonized, and so on. Without

independent evidence for the existence of these mammals, this hypothesis is no more or less likely than any other.

Haffer's proposal of Pleistocene drying in the Amazon and the existence

of forest refugia was as hypothetical as his distinct proposal that this mechanism

caused differentiation in Amazonian organisms. The former is reasonably

testable, but not, of course, by the existence of the patterns of diversification he set out to explain, which would be circular reasoning. Unfortunately,

independent paleoecological evidence (summarized by Colinvaux 1996) does

not support the existence of the proposed Pleistocene refugia, and attempts to

reconstruct Amazonian paleo-communities are extremely controversial (see

Colinvaux 1996, Haffer 1997a). In other words, we simply don't know what

andent Amazonia looked like and have no strong evidence to suggest that,

during the Pleistocene, it was radically different from the present. So why did Haffer suggest Pleistocene refugia as opposed to any other

spedation mechanism? The only other existing hypothesis when he first

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. elaborated his refuge hypothesis (Haffer 1969,1974) was river barriers. Not only did that hypothesis rely on unlikely assumptions of total impermeability

of rivers and lead to inaccurate distributional predictions (by Haffer's analysis), but it invoked events, such as tectonism, Andean uplift, and river formation,

that Haffer believed were too ancient to correspond with the low levels of phenotypic divergence he observed in Amazonian birds. In particular, the

typically subtle morphological differentiation among related forms and the prevalence of allotaxon complexes suggested a recent history of vicariance, with

insufficient time to develop ecological differentiation and sympatry—i.e., to

"complete" the speciation process (sensu Mayr 1969). But genetic evidence

gathered since Haffer's initial proposal (including the present study)

consistently points to pre-Pleistocene divergence times. Thus, Haffer himself has come to propose earlier vicariance phenomena, invoking the same sorts of

climatic cycles in prior epochs (Haffer 1993,1997a).

Is it worth proposing more historical scenarios? Of course it is. One of

them may eventually even turn out to be right. As we have seen, however,

timing is one of the key features distinguishing various historical hypotheses, and we are still unsure of how to assess timing precisely in phylogenetic

studies. Nevertheless, this is likely to improve with more taxa, more

independent studies of molecular evolution, and the gradual accumulation of

independent paleoecological data. Meanwhile, however, our strongest

historical evidence comes from phylogenetic branching patterns, which, as we

have seen in the above discussion, are not predicted to differ among most of the major hypotheses proposed.

One of the objectives of this study (see Introduction) was to determine whether any single historical event or series of events is likely to have caused

the observed patterns of differentiation in all the study taxa. The answer

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. appears to be "no." Lack of strict concordance in distributions, phylogenetic branching patterns, and sequence divergences all point to distinct evolutionary

histories for each of the flycatcher species groups examined. Does this imply

chaos and an unknowable history? One testable possibility, as suggested

earlier, is that each species complex represents a syndrome for its particular

microhabitat. Perhaps, though, too many processes have been involved ever to

uncover the history of the Amazonian avifauna; nevertheless, obviously there

must have been some processes, and it would be premature to give up trying to

identify which one or ones played an important role just because they are difficult to distinguish (see Bush 1994).

What is lacking is a null hypothesis, which must be rejected before we

engage in seemingly endless speculation about equally probable and

unsubstantiated alternative historical scenarios. The river barriers hypothesis

postulated both generation and maintenance of biogeographical patterns by the

same mechanism (rivers barriers) and so could be seen as a sort of null, relative

to those hypotheses suggesting separate mechanisms. However, the river

barriers hypothesis appears to have serious flaws, as originally noted by Haffer

and corroborated by this study. But also, the river hypothesis still requires a specific and distinct historical event (the formation of the major Amazonian

rivers) to work. In that sense, then, it is a distinct hypothesis and not truly a

null. What is the null hypothesis in this case?

It would seem that the appropriate null hypothesis for Amazonian

biogeography is that current environmental conditions are sufficient to explain

the biogeographical patterns observed in the present. That is, that the current

patterns are an inevitable result of current conditions, sui generis, without

invoking other hypothetical (past) events or mechanisms. But how would that

work? If, as I proposed above and as can be tested by modeling, any speciation

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. mechanism allowed to run for enough generations will create the observed patterns, then the problem is to identify a speciation mechanism not associated

with a specific event, but rather that can occur more or less constantly without

invoking major historical changes. Endler's (1982) ecological gradients hypothesis provided such a mechanism in the form of parapatric speciation, but

that mechanism appears not to be supported (see above). Several years ago, G. B. Williamson (pers. comm.) pointed out to me a

mechanism whereby vicariant speciation might occur more or less continuously

in the Amazon. It requires three basic conditions that appear to apply to many

Amazonian bird populations: distribution over a large geographic area, low

population densities, and limited dispersal. These conditions alone might drive speciation under temporary isolation of some portion of a population. Factors

causing isolation might include localized extinctions due to disease, fires,

floods, major forest blowdowns, and other forms of occasional, but unpredictable, disturbance known to occur in the region. These conditions could be exacerbated by any sort of habitat patchiness, including that predicted

by alternating and prolonged periods of cooling and warming as described by

Colinvaux (1996). This sort of mechanism is a standard source of population

genetic structure in metapopulation models (e.g., McCauley 1993), but has

never before been applied to Amazonian birds or shown to be a possible source

of speciation without other vicariance events. Below, I will elaborate how what

I call the "demographic stochasticity hypothesis" might work in Amazonian

birds and might predict the biogeographic patterns observed in this study and

elsewhere.

Demographic Stochasticity Hypothesis Under this model, a hypothetical widespread Amazonian bird species is

mostly sedentary, with small permanent territories and year-round

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. monogamous pair bonds. It occurs at low densities in rainforest, having a

somewhat patchy distribution, probably related, at least in part, to some very specific microhabitat specialization or interspecific interaction (e.g.,

competition, mutualism, predation). It is physically capable of crossing any size river, but is averse to entering open spaces, and its probability of crossing a

river is inversely proportional to the river's width. Due to some stochastic

demographic factor (see above), a local subpopulation becomes genetically

isolated temporarily. If that isolation occurs for a long enough period or happens to isolate a few unusual individuals, then the isolated subpopulation

could take on unique genetic characteristics. In most cases, upon secondary

contact with adjacent populations, those characteristics will become swamped

by the others, or, if they confer some selective advantage, they will sweep

through the population at large. This sort of genetic population dynamics should be going on constantly.

However, without permanent reproductive isolation or differential selective

pressures across the range of the species (which we assume not to exist), it will

not lead to speciation. Only if the trait that becomes modified in isolation is

under sexual selection pressure might it lead to speciation. For example, if

female preference for male song differs in a temporary isolate, that difference

might quickly drive to fixation and, upon secondary contact with other

populations, effect speciation on the basis of prezygotic reproductive isolation. Having spedated without ecological differentiation, the rest of the

scenario is similar to that under any other proposed spedation mechanism.

These now distinct spedes will compete and one may drive the other to local

extinction by competitive exdusion. The distributional limit between the two

forms should converge quickly on some sort of dispersal "trough" where

reduced dispersal enables one form to dominate each side. This trough could

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. be any sort of habitat discontinuity, the most conspicuous of which in

Amazonia are rivers. As long as the two species remain ecologically identical, they will maintain allopatry, not because of an insurmountable physical barrier, but because of interspecific competition.

Because the relative effectiveness of a dispersal barrier should affect the

likelihood of finding a distributional limit there, spatial patterns across taxa

should be generally similar, especially with respect to the biggest rivers.

However, because this process can happen at any place in the basin and does not require that other spedes be affected simultaneously, the details of

distributional patterns should differ across species, espedally with respect to

lesser rivers and other less conspicuous impediments to dispersal. Also,

because this process can occur continuously in time and independently in

different taxa, phylogenetic branching patterns and genetic divergences should vary randomly across taxa.

Thus, this null hypothesis predicts only roughly concordant

distributional patterns and randomly variable phylogenetic branching patterns

and genetic divergences. Based on my results, the null hypothesis cannot be

rejected. Furthermore, the geographic variants in Hemitriccus taxa studied here appear to have spedes-level sequence divergences, to differ phenotypically

only in traits related to reproductive isolation or that carry no selective burden,

and to differ ecologically only among the most genetically divergent forms.

This is consistent with a view of potentially competing species selected for

similarly constrained (specialized) phenotypes over long periods of time in a constant environment and stable (and rich) interspecific milieu.

Clearly, this new hypothesis is a "just-so story" too, and a rather

elaborate one at that. However, if its assumptions prove to be valid, then it is a true null hypothesis in that it is more parsimonious than other hypotheses in

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. not invoking additional processes or events other than those that can be

observed in the present. Also, it differs from the others in that its assumptions are almost entirely ornithological in nature. Although almost none of the

assumed avian traits (e.g., lack of ecological differentiation, variable dispersal

rates, sexual selection on modified traits, etc.) has been studied in Amazonian

birds, they all can (and must) be tested by ornithologists in the field and lab.

This hypothesis need not wait for independent evidence from other disciplines.

Also, it may be applied to other groups of organisms, which could be expected to differ in their patterns, based on their dispersal abilities, habitat selectivity,

social systems, and so on, and, consequently, in their biogeographical patterns

as well.

Conclusions Active field work added considerably to our understanding of the

distributions of Amazonian Hemitriccus flycatchers and the nature of their geographic variation. In particular, vocal analyses permitted the recognition of

monophyletic units (as diagnosed using molecular phylogenetic techniques),

many of them morphologically cryptic, better than did existing taxonomy.

Voice should prove to be a powerful tool for recognizing geographic variation

in many more bird taxa in the Neotropics. Molecular analyses provided the finest degree of sensitivity to geographic variation and allowed phylogeny

reconstruction, which in turn enabled tests of historical biogeographic

hypotheses.

None of the previously proposed hypotheses of Amazonian

biogeography accounted for all the patterns found even in this small group of

three closely related flycatcher dades. Furthermore, many of these hypotheses were found not to differ in their predictions, given the kind of data gathered in

this study. I recommend testing results of biogeographic analysis against a null

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hypothesis that current environmental conditions are sufficient to explain

phylogeographic patterns. In this study, it was impossible to reject this null hypothesis. Results of many more such studies will help to uncover patterns of

biogeographical concordance, if such patterns really exist. Clearly, this is the beginning of the use of bird (and other organismal)

data, not a last ditch effort to make it fit existing hypotheses. This kind of work

needs to be done on a massive scale to be reevaluated in light of independent

paleo-historical data sets. Although never too soon to begin synthesizing and attempting to suggest processes to explain described patterns, it is clearly far

too early to stop describing and refining the patterns we wish to explain. A

hypothetical process that suitably explains a non-existent pattern is a weak

contribution at best; but the accurate description of more existing patterns than

are explained by a given process is the start of an understanding of the true

complexity of the problem.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LITERATURE CITED

Alvarez A., J. 1994. Abundanda y diversidad de espedes de aves en los nos Tigre y Corrientes. Unpublished thesis, Universidad Nadonal de la Amazonia Peruana, Iquitos.

Amadon, D., and L. L. Short. 1992. Taxonomy of lower categories—Suggested guidelines. Bulletin of the British Ornithologists' Club (Supplement) 112A: 11-38.

Anderson, A. B. 1981. White-sand vegetation of Brazilian Amazonia. Biotropica 13:199-210.

Arctander, P. 1995. Comparison of a mitochondrial gene and a corresponding nudear pseudogene. Proceedings of the Royal Sodety of London, series B 262:13-19.

Avise, J. C. 1994. Molecular markers, natural history and evolution. Chapman and Hall, New York.

Ayres, J. M., and T. H. Clutton-Brock. 1992. River boundaries and spedes range size in Amazonian primates. American Naturalist 140: 531-537.

Bates, J.M., S.J. Hackett, and J. Cracraft. 1998. Area-relationships in the Neotropical lowlands: an hypothesis based on raw distributions of Passerine birds. Journal of Biogeography 25: 783-793.

Bates, J.M., S.J. Hackett, and J.M. Goerck. 1999. High levels of mitochondrial DNA differentiation in two lineages of antbirds (Drymophila and Hypocnemis). Auk 116:1093-1106.

Bierregaard, R. O., Jr., M. Cohn-Haft, and D. F. Stotz. 1997. Cryptic biodiversity: An overlooked spedes and new subspedes of antbird (Aves:Formicariidae) with a revision of Cercomacra tyrannina in northeastern South America. Ornithological Monographs 48: 111-128.

Borges, S. H., M. Cohn-Haft, A. M. P. Carvalhaes, L. M. Henriques, J. F. Pacheco, and A. Whittaker. In press. Birds of the Jau National Park, Brazilian Amazon: Spedes checklist, biogeography and conservation. Omitologia Neotropical.

Buckley, P. A., M. S. Foster, E. S. Morton, R. S. Ridgely, and F. G. Buckley. 1985. Neotropical ornithology. Ornithological Monographs No. 36.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Burbrink, F. T., R. Lawson, and J. B. Slowinski 2000. Mitochondrial DNA phylogeography of the North American rat snakes (Elaphe obsoleta): A critique of the subspecies concept. Evolution (in press).

Bush, M. B. 1994. Amazonian speciation: A necessarily complex model. Journal of Biogeography 21: 5-17.

Capparella, A. P. 1988. Genetic variation in Neotropical birds: Implications for the speciation process. Acta Congr. Int. Omith. 19:1658-1664.

Capparella, A. P. 1991. Neotropical avian diversity and riverine barriers. Acta Congr. Int. Omith. 20:307-316.

Catchpole, C. K., and P. J. B. Slater. 1995. Bird song: Biological themes and variations. Cambridge University Press, Cambridge.

Chesser, R. T. 2000. Evolution in the high Andes: The phylogenetics of Muscisaxicola ground-tyrants. Molecular Phylogenetics and Evolution 15: 369-380.

Cohn-Haft, M. 1996. Why the Yungas Tody-Tyrant (Hemitriccus spodiops) is a Snethlagea, and why it matters. Auk 113:709-714.

Cohn-Haft, M., A. Whittaker, and P.C. Stouffer. 1997. A new look at the "spedes-poor" central Amazon: The avifauna north of Manaus, Brazil. Ornithological Monographs 48: 205-235.

Colinvaux, P.A. 1987. Amazon diversity in light of the paleoecological record. Quartemary Sdence Review 6:93-114.

Colinvaux, P.A. 1993. Pleistocene biogeography and diversity in tropical forests of South America. Pages 473-499 in Biological relationships between Africa and South America (P. Goldblatt, Ed.). Yale University Press, New Haven, Connecticut.

Colinvaux, P.A. 1996. Quaternary environmental history and forest diversity in the Neotropics. Pages 359-405 in Evolution and environment in tropical America (J. B. C. Jackson, A. F. Budd, and A. G. Coates, Eds.). University of Chicago Press, Chicago.

Colinvaux, P.A. 1998. A new vicariance model for Amazonian endemics. Global Ecology and Biogeography Letters 7: 95-96.

Cracraft,J. 1983. Spedes concepts and spedation analysis. Pages 159-187 in Current ornithology, vol. 1 (R. F. Johnston, Ed.). Plenum Press, New York.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cracraft, J. 1985. Historical biogeography and patterns of differentiation within the South American avifauna: Areas of endemism. Ornithological Monographs 36: 49-84.

Cracraft, J., and R.O. Prum. 1988. Patterns and processes of diversification: Speciation and historical congruence in some Neotropical birds. Evolution 42: 603-620.

Crosby, G. T. 1972. Spread of the Cattle Egret in the Western Hemisphere. Bird-Banding 43: 205-212.

Desjardins, P., and R. Morais. 1990. Sequence and gene organization of the chicken mitochondrial genome. Journal of Molecular Biology 212: 599- 634.

Emmons, L. H., and F. Feer. 1990. Neotropical rainforest mammals. University of Chicago Press, Chicago.

Endler, J. 1977. Geographic variation, speciation, and dines. Princeton University Press, Princeton, New Jersey.

Endler, J. 1982. Pleistocene forest refuges: Fact or fancy? Pages 179-200 in Biological diversification in the tropics (G. T. Prance, Ed.). Columbia University Press, New York.

Fitzpatrick, J. W. 1980. Foraging behavior of Neotropical tyrant flycatchers. Condor 82:43-57.

Hackett, S.J. 1993. Phylogenetic and biogeographic relationships in the Neotropical genus Gymnopithys (Formicariidae). Wilson Bulletin 105: 301-315.

Hackett, S.J. 1996. Molecular phylogenetics and biogeography of tanagers in the genus Ramphocelus (Aves). Molecular Phylogenetics and Evolution 5: 368-382.

Hackett, S.J., and K.V. Rosenberg. 1990. Comparison of phenotypic and genetic differentiation in South American antwrens (Formicariidae). Auk 107: 473-489.

Haffer, J. 1969. Spedation in Amazonian forest birds. Sdence 165:131-137.

Haffer,J. 1970. Art-Entstehung bei einigen Waldvogeln Amazoniens. J. Omith. I ll: 285-331.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Haffer, J. 1974. Avian spedation in tropical South America. Publications of the Nuttall Ornithological Club, No. 14.

Haffer, J. 1978. Distribution of Amazon forest birds. Bonn. Zool. Beitr. 29: 38- 78.

Haffer, J. 1982. General aspects of the refuge theory. Pages 6-24 in Biological diversification in the tropics (G. T. Prance, Ed.). Columbia University Press, New York.

Haffer, J. 1985. Avian zoogeography of the Neotropical lowlands. Ornithological Monographs 36:113-146.

Haffer, J. 1987. Biogeography of Neotropical birds. Pages 105-150 in Biogeography and Quaternary history in tropical America (T. C. Whitmore and G T. Prance, Eds.). Clarendon and Oxford University Press, Oxford.

Haffer, J. 1992a. Parapatric spedes of birds. Bulletin of the British Ornithologists' Club 112: 250-264.

Haffer, J. 1992b. On the "river effect" in some forest birds of southern Amazonia. Boletim do Museu Paraense Emilio Goeldi, ser. Zool. 8: 217- 245.

Haffer, J. 1993. Time's cyde and time's arrow in the history of Amazonia. Biogeographica 69:15-45.

Haffer, J. 1997a. Alternative models of vertebrate spedation in Amazonia: An overview. Biodiversity and Conservation 6:451-476.

Haffer, J. 1997b. Contact zones between birds of southern Amazonia. Ornithological Monographs 48:281-305.

Haffer, J., and J. W. Fitzpatrick. 1985. Geographic variation in some Amazonian forest birds. Ornithological Monographs 36:147-168.

Harshman, J. 1996. Phylogeny, evolutionary rates, and ducks. Ph.D. dissertation, University of Chicago, Chicago.

Hasegawa, M., H. Kishino, and T. Yano. 1985. Dating of the human-ape splitting by a molecular dock of mitochondrial DNA. Journal of Molecular Evolution 21:161-164.

Hellmayr, C. E. 1910. The birds of the Rio Madeira. Novitates Zoologicae 17: 257-428.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hellmayr, C. E. 1927. Catalogue of birds of the Americas and the adjacent islands. Part 5, Tyrannidae. Field Museum of Natural History Publication 242, Zoological Series, vol. 13.

Hillis, D. M., B. K. Mable, and C. Moritz. 1996a. Applications of molecular systematics: The state of the field and a look to the future. Pages 515-543 in Molecular sysematics, 2nd edition (D. M.Hillis, C. Moritz, and B. K. Mable, Eds.). Sinauer Associates, Sunderland, Massachusetts.

Hillis, D. M., C. Moritz, and B. K. Mable. 1996b. Molecular systematics. Sinauer Associates, Sunderland, Massachusetts.

Howell, S. N. G., and M. B. Robbins. 1995. Species limits in the Least Pygmy- Owl (Glaucidium minutissimum) complex. Wilson Bulletin 107: 7-25.

Hudson, R. R. 1990. Gene genealogies and the coalescent process. Oxford Surveys in Evolutionary Biology 7:1-44.

Isler, M.L., P.R. Isler, and B.M. Whitney. 1997. Biogeography and systematics of the Thamnophilus punctatus (Thamnophilidae) complex. Ornithological Monographs 48:355-381.

Isler, M.L., P.R. Isler, and B.M. Whitney. 1998. Use of vocalizations to establish species limits in antbirds (Passeriformes: Thamnophilidae). Auk 115: 577-590.

Isler, M.L., P.R. Isler, and B.M. Whitney. 1999. Species limits in antbirds (Passeriformes: Thamnophilidae): The Myrmothenila surinamensis complex. Auk 116:83-96.

Johns, G. C., and J. C. Avise. 1998. A comparative summary of genetic distances in the vertebrates from the mitchondrial cytochrome b gene. Mol. Biol. Evol. 15:1481-1490.

Johnson, N.K. 1980. Character variation and evolution of sibling species in the Empidonax difficilis-flavescens complex (Aves: Tyrannidae). Univ. Calif. Publ. Zool. 112:1-151.

Kingman, J. F. C. 1982. On the genealogy of large populations. J. Applied Probability 19A: 27-43.

Kishino, H., and M. Hasegawa. 1989. Evaluation of the maximum likelihood estimate of the evolutionary tree topolgies from DNA sequence data, and the branching order in Hominoidea. J. Molecular Evolution 31:170-179.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kocher, T. D., W. K. Thomas, A. Meyer, S. V. Edwards, S. Paabo, F. X. Villablanca, and A. C. Wilson. 1989. Dynamics of mitochondiral DNA evolution in animals: Amplicafidation and sequencing with conserved primers. Proceedings of the National Academy of Sciences USA 86: 6196-6200.

Krabbe, N., M. L. Isler, P. R. Isler, B. M. Whitney, J. Alvarez A., and P. J. Greenfield. 1999. A new species in the Myrmothemla haematonota superspedes (Aves; Thamnophilidae) from the western Amazonian lowlands of Ecuador and Peru. Wilson Bulletin 111: 157-302.

Kratter, A. W., and T. A. Parker IE. 1997. Relationship of two bamboo- spedalized foiiage-gleaners: Automolus dorsalis and Anabazenops fuscus (Fumariidae). Ornithological Monographs 48:383-397.

Kroodsma, D.E. 1984. Songs of the Alder Flycatcher ( Empidonax alnorum) and Willow Flycatcher ( Empidonax traillii) are innate. Auk 101:13-24.

Lanyon, S. M., and K. E. Omland. 1999. A molecular phylogeny of the blackbirds (Icteridae): Five lineages revealed by cytochrome-i? sequence data. Auk 116: 629-639.

Lanyon, W.E. 1978. Revision of the Myiarchus flycatchers of South America. Bull. Amer. Mus. Nat. Hist. 161:429-627.

Lanyon, W. E. 1988. A phylogeny of the flatbill and tody-tyrant assemblage of tyrant flycatchers. American Museum Novitates 2923:1-41.

Lougheed, S. C., C. Gascon, D. A. Jones, J. P. Bogart, and P. T. Boag. 1999. Ridges and rivers: A test of competing hypotheses of Amazonian diversification using a dart-poison frog (Epipedobatesfemoralis). Proceedings of the Royal Sodety of London B 266:1829-1835.

Lowery, G. H., Jr. 1946. Evidence of trans-gulf migration. Auk 63:175-211.

Maddison, W. P., and D. R. Maddison. 1992. MacClade: Analysis of phylogeny and character evolution. Verson 3.0. Sinauer, Sunderland, Massachusetts.

Mariaux, J., and M. J. Braun. 1966. A molecular phylogenetic survey of the nightjars and allies (Camprimulgiformes) with spedal emphasis on the potoos (Nyctibiidae). Molecular Phylogenetics and Evolution 6:228-244.

Mayr, E. 1946. The number of spedes of birds. Auk 63:64-69.

Mayr, E. 1969. Prindples of systematic zoology. McGraw-Hill, New York.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. McCauley, D. E. 1993. Evolution in metapopulations with frequent local extinction and recolonization. Oxford Surveys in Evolutionary Biology 9:109-134.

McCracken, K. G., and F. H. Sheldon. 1997. Avian vocalizations and phylogenetic signal. Proceedings of the National Academy of Sciences of the USA 94:3833-3836.

McKitrick, M. C., and R. M. Zink. 1988. Species concepts in ornithology. Auk 90:1-14.

Nelson, G. J., and N. I. Platnick. 1981. Systematics and biogeography: dadistics and vicariance. Columbia Univerisyt Press, New York.

Nores, M. 1999. An alternative hypothesis of the origin of Amazonian bird diversity. Journal of Biogeography 26:275-485.

Oren, D. C., and H. G. Albuquerque. 1991. Priority areas for new avian collections in Brazilian Amazonia. Goeldiana Zool. 6.

Pacheco, J. F., L. F. Silveira, F. C. Straube. 1999. Bibliografia recente da omitologia brasileira. Ararajuba7:149-152.

Paynter, R. A., Jr. 1991. The maturation of Brazilian ornithology. Ararajuba 2: 105-106.

Pelzeln, A. 1868. Zur Omithologie Brasiliens. Resultate von Johann Natterers Reisen in den Jahren 1817 bis 1835. Druck und Verlag von A. Pichler's Witwe & Sohn, Vienna, Austria.

Peterson, A. T. 1998. New spedes and new spedes limits in birds. Auk 115: 555-558.

Pires, J. M., and G. T. Prance. 1985. The vegetation types of the Brazilian Amazon. Pages 109-145 in Amazonia (G. T. Prance and T. E. Lovejoy, Eds.). Pergamon Press, Oxford.

Posada, D., and K. A. Crandall. 1998. Modeltest: Testing the model of DNA substitution. Bioinformatics 14: 817-818.

Quinn, T. W. 1997. Molecular evolution of the mitochondrial genome. Pages 3- 28 in Avian molecular evolution and systematics (D. P. Mindell, Ed.). Academic Press, San Diego, California.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Remsen, J.V., Jr. 1997. Studies in Neotropical ornithology honoring Ted Parker. Ornithological Monographs No. 48.

Remsen, J.V., Jr., and T. A. Parker m . 1983. Contribution of river-created habitats to bird species richness in Amazonia. Biotropica 15: 223-231.

Robbins, M. B., and S. N. G. Howell. 1995. A new species of pygmy-owl (Strigidae: Glaudicium) from the eastern Andes. Wilson Bulletin 107:1-6.

Rosen, D. E. 1978. Vicariant patterns and historical explanation in biogeography. Systematic Zoology 27:159-188.

Sheldon, F. H., and D. W. Winkler. 1993. Intergeneric phylogenetic relationships of swallows estimated by DNA-DNA hybridization. Auk 110:798-824.

Sick, H. 1967. Rios e enchentes na Amazonia como obstaculo para a avifauna. Atas do simposio sobre a biota amazonica 5:495-520.

Silva, M. N. F., and J. L. Patton. 1993. Amazonian phylogeography: mtDNA sequence variation in arboreal echimyid rodents (Caviomorpha). Molecular Phylogenetics and Evolution 2:243-255.

Silva, M. N. F., and J. L. Patton. 1998. Molecular phylogeography and the evolution and conservation of Amazonian mammals. Molecular Ecology 7:475-486.

Smith, T. B., R. K. Wayne, D J. Girman, and M. W. Bruford. 1997. A role for ecotones in generating rainforest biodiversity. Science 276:1855-1857.

Snethlage, E. 1913. Uber die Verbreitung der Vogelarten in Unteramazonien. J. Omith. 61:469-539.

Sorenson, M. D., and R. C. Fleischer. 1996. Multiple independent transpositions of mitochondrial DNA control region sequences to the nucleus. Proceedings of the National Academy of Science USA 93: 15239-15243.

Stotz, D.F. 1992. Specific status and nomenclature of Hemitriccus minimus and Hemitriccus aenigma. Auk 109:916-917.

Stotz, D. F., J. W. Fitzpatrick, T. A. Parker HI, and D. K. Moskovits. 1996. Neotropical birds; Ecology and conservation. University of Chicago Press, Chicago.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Swofford, D. L. 1998. PAUP*. Phylogenetic analysis using parsimony (* and other methods). Version 4. Sinauer Associates, Sunderland, Massachusetts.

Swofford, D. L., G. J. Olsen, P. J. Waddell, and D. M. Hillis. 1996. Phylogenetic inference. Pages 407-514 in Molecular sysematics, 2nd edition (D. M.Hillis, C. Moritz, and B. K. Mable, Eds.). Sinauer Associates, Sunderland, Massachusetts.

Telfair, R. C. II. 1994. Cattle Egret (Bubulcus ibis). In The Birds of North America, No. 113 (A. Poole and F. Gill, Eds.). Academy of Natural Sciences, Philadelphia; American Ornithologists' Union, Washington, D.C.

Traylor, M. A., Jr. 1979. Tyrannidae. Pages 1-240 in Check-list of birds of the world, vol. 8 (M. A. Traylor, Jr., Ed.). Harvard Univ. Press, Cambridge, Massachusetts.

Vielliard, J. 1989. Uma nova especie de Glacidium (Aves, Strigidae) da Amazonia. Revista Brasileira de Zoologia 6: 685-693.

Wallace, A.R. 1853. A narrative of travels on the Amazon and Rio Negro. Reeve, London.

Whitney, B.M. 1992. Observations on the systematics, behavior, and vocalizations of "Thamnomanes" occidentalis (Formicariidae). Auk 109: 302-308.

Whitney, B. M., and J. Alvarez A. 1998. A new Herpsilochmns antwren (Aves: Thamnophilidae) from northern Amazonian Peru and adjacent Ecuador: The role of edaphic heterogeneity of terra firme forest. Auk 115:559-576.

Whitney, B. M., and J. F. Pacheco. 1997. Behavior, vocalizations, and relationships of some Myrmotherula antwrens (Thamnophilidae) in eastern Brazil, with comments on the "plain-winged" group. Ornithological Monographs 48:809-819.

Whitney, B. M., J. F. Pacheco, D. R. C. Buzzetti, and R. Parrini. 2000. Systematic revision and biogeography of the Herpsilochmns pileatas complex, with description of a new species from northeastern Brazil. Auk 117: 869-891.

Whitney, B. M., J. F. Pacheco, P. R. Isler, and M. L. Isler. 1995a. Hylopezus nattereri (Pinto, 1937) is a valid species (Passeriformes: Formicariidae). Ararajuba 3:37-42.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Whitney, B. M., J. F. Pacheco, and R. Parrini. 1995b. Two species of Neopelma in southeastern Brazil and diversification within the Neopelma/Tyranneutes complex: implications of the subspecies concept for conservation (Passerformes: Tyrannidae). Ararajuba 3:43-53.

Willis, E. O. 1969. On the behavior of five species of Rhegmatorhina, ant- following antbirds of the Amazon basin. Wilson Bulletin 81: 363-395.

Zimmer, J. T. 1940. Studies of Peruvian birds. No. 34. The genera Todirostnim, Euscarthmomis, Snethlagea, Poecilotriccus, Lophotriccus, Myiomis, Pseiidotriccus, and Hemitriccus. American Museum Novitates 1066:1-23.

Zimmer, K. J. 1997. Species limits in Cranioletica vulpina. Ornithological Monographs 48:849-864.

Zimmer, K. J., T. A. Parker IE, M. L. Isler, and P. R. Isler. 1997. Survey of a southern Amazonian avifauna: The Alta Floresta region, Mato Grosso, Brazil. Ornithological Monographs 48: 887-918.

100

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. minor BMW AF-L BMW Carajas BMW Tap E BMW Acre 1 AW misc 00-2a-15 98-33a 96-12a borba 3 98-29b2 98-29b 1 98-30a 99-6a-14 98-31a SaoGab 97-206 JFP H. 9.14 -247 -1.1 -36.9 -22.4 95-6b-19 -15.7 -19.2 -18.6 -18.9 -26.9 Pepe4 -13.5 AW Urubu -15.2 -33.6 95-13a-15 -26.8 -33.4 2.78 696 97-26b-9 2.96 2.52261 -29.2 -20.9 CAM 96-22 AW 2.26 -20.1 BMWAcre 2 2.61252 •14.6 -185 SHBaJau SllBcjau 297 2.61269 -18.2 -25.7 TAP AltFlo 00-lb-2 1 2.61 -31.5 BMW AF-R1 2.35 2.43 2.26 2.52 2.35 2.52 -30.1 2.352.43 -18.8 Pepe 10 2.262.43 -16.5 -21.9 2.43 97-6b -27.5 CAM 97-20-2 2.35 2.96 2.782.87 -15.42.35 -13.1 S.G.2 2.87 2.172.43 -22.92.53 -21.1 AWamaz np -30.2 97-29a(l) CAM98-rl6-2 1 1 2.78 -17.4 Mao 3 8 9 4 2.69 -34.7 98-13a-0 5 2.61 4 14 2.96 3.4 11 2.43 11 2.09 -21.6 00-4b-25 12 13 14 15 Catalog Number 12 2.35 5 2.61 8 2.17 2 2.17 -24.72.78 97-9b 2.78 7 2.61 252 2.61 2.61 6 2.61 235 2.52 11 3.35 -7.22 2.09 -1.64 2.43 3 -15.6 2.35 9 -28.2 -14.9 2.61 -24.4 -28.1 -24.7 -14.4 8 9 18 -10.8 2.43 7 19 17 -12.2 2.43 6 16 -18.7 2.43 16 5.9 2.61 9 19 12 -27.3 2.17 6 29 22 -19.7 2.26 3 42 -16.4 26 -15.2 2.87 7 1.65 1.48 17 18.2 2.35 2.43 2.61 12 -26.9 2.61 7 2.17 2.35 22 -14.7 2.52 2.43 2.69 27 226 17 -14.3 2.43 6 2.87 32 -23.2 2.96 6 2.69 34 -20.6 261 1 2.69 -161 93<0cl)-l-19 2.09 2262.17 21 23 -25.2 -22.5 2.35 2.26 9 5 2.17 11 -23.1 2.35 7 2.26 21 -18.8 2.43 7 2.87 163 243 12 -131 2.66 2.35 5.52 694 2 16 -579 2.26 21 -IB 2.35 12 -3.01 1.64 -9.2 156 16 -714 217 8 2.26 -13.6 00-2a-15 -7.2 -7.37 2.26 20 -24.5 2.43 -265 252 32 -13.7 13.9 11.8 2 17 -102 -3.13 -324 2.35 24 -166 2 17 20 -17.8 2.17 -27.5 1.91 17 -4.9 2.17 1 -18.3 -27.8 252 31 -37.8 -19.5 2.61 28 -218 2.61 3 2.69 -30.8 -17.4 2.78 33 -22.4 2.78 -17.3 -17.3 -32.8 2.35 -29.7 2.78 35 -31.5 2.87 5 2.96 -33.8 -12.91 2.26 1 09 -1 2.52 12 -2.91 2.78 28 9 -149 19 11 -398 2.26 11 343 11 1.12 2 13 -17.7 2.17 8 17 14 10 13 15 10 16 12 20 32 -37.4 35 -33.6 287 42 -21.5 2.78 1 13 13 1.48 19 1.91 1.39 1.39 1.3 1.56 1.3 2.09 20 2.52 2.87 2.092.52 13 28 2.97 31 243 21 -326 261 37 -18.2 252 2 2.17 13 2.09 2.87 21 -6.98 13 16 2.35 12 0 39 87 1.56 54 2.35 26 71 2.26 65 96 41 1.22 15 10.7 159 15 10.5 2.26 4 62 50 2.43 22 57 269 29 -16.1 57 43 48 2.61 15 100 19 22 77 48 2.35 20 39 5362 43 1.3 12 69 45 39 89 68 87 2.17 19 48 5059 60 47 59 71 50 77 4 5 6 7 8 9 10 107 59 257 58 2.87 299 328 283 2 3 671 671 64 718 63 78 515 340 34 54 803 71 814 648 72 67 64 43 83 1 897 144 590 47 462 50 41 544 60 686 46 556 27 663 693 48513 46 53 53 70 2.17 17 866 779 197 167 347 264 51 513 46 495 59 941 941 104 328 58 2.78 25 844 121 419 73 80 74 518 450 62 32 98 41 2.17 15 -19 487 34 43 72 2.17 19 -17 947 9% 178 200 282 1139 1139 320 300 52 2.78 32 1030 240 8 6 6 9 1111 808 110 70 88 10 8 10 7 8 77 6 7 8 8 8 888 11 9 6 88 9 11 7 8 8 7 532 57 61 105 8 9 751 81 75 116 1.48 8 12 4 44 88 11 7 77 7 7 4 85 1157 67 VT 1 APPENDIX 1: OF LIST TAPE RECORDINGS ANALYZED Caxiuana 8 8 Locality Pari; BR: Amazonas;BR: Manaus BR: Amazonas;BR: Jau NP Amazonas;BR: jau NP 8 10 BR: Acre;BR: Sena do Moa MatoBR: Grosso; Alta Floresta BR: Amazonas;BR: jau NP 8 U BR: Amazonas;BR: Manaus Amazonas;BR: Manaus Amazonas;BR: BariuauR. 7 7 7 6 889 295 BR: MatoBR: Grosso; opp. Alta Floresta 8 8 489 49 57 63 2.35 12 -14.1 BR: MatoBR: Grosso; Alta Floresta MatoBR: Grosso; opp. Alla Floresta 8 8 12 12 BR: Amazonas;BR: Cueiras R. Amazonas;BR: GabrielS. Amazonas;BR: GabrielS. Acre;BR: Sena do Moa 8 11 495 43 46 39 BR; Amazonas;BR; Manaus Amazonas;BR: UrubuR. BR: Amazonas;BR: jau NP BR: Amazonas;BR: Borba Amazonas;BR: Ig. Tumbira Amazonas;BR: jau NP MatoBR: Grosso; Alta Floresta 8 13 BR: Parti;BR: Amazonia NP 8 BR: BR: BR: Pari;BR: Carajas 8 BR: Parti;BR: Caxiuana BR: Parti;BR: Caxiuana 8 7 520 59 BR: Pari;BR: opp. ManicoriPE: Iquitos 8 11 478 42 44 34 1.48 16 BR: Pari;BR: Caxiuana 8 Pari;BR: Breves PE: PE: Iquitos Pari;BR: Carajas 4 45 678 18 13 16 2.43 15 BR: Pari;BR: Itabaiuna Pari;BR: Maripa Pari;BR: opp. Humaita MatoBR: Grosso; 8 Alta Floresta 8 12 4 8 55 6 685 531 11 93 11 87 12 147 2.35 243 12 8 BR: Pari;BR: Genipapo BR: Pari;BR: Maripa BR: Parti;BR: Caxiuana BR: Pari;BR: Carajas 4 54 817 91 36 13 209 BR: Pari;BR: Carajas Taxon inomalus inornatus minimus minimus minimus inomalus minimus minimus minimus minimus minor minor minimus minor minor inomatusinomatus minimus inomalus minimus minimus inomatus minimus inomalus inomatus minimus inomatusinomalus minimus inomalus minimus inomalus inomalusinomalus inomatus inomalus inomatus minimus inomatus minimus inomalus inomatusinomatusinomalus inomatusinomalus inomalus minimus in the text (Results), and column numbers refer to vocal parameters described in Methods. inomatus inomalus inomalus inomatus inomalus Localities are abbreviated using the first two letters of the country name. Vocal types (VT) follow the numberingscheme inomalus inomatus minimus inomalus minimusinomalus inomalus inomalus minimus inomatus inomatus minimus inomalus minimus minor inomatus inomalus minimusinomalusinomalus minimus minimus minor minor Complex inomalus minimus minor inomatus minimus minor inomalus minimus minor

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DFLCush natl(2) DFLCush nat?3(l) DFLCush 2(1) JFPTefeO AA OP-005-8 11.m. Borbashort 11.m. 97-9a 2nd use 99-6b-29 CAM Ins 88300 TSS PNNKM later Catalog Number 98-29a:l 97-18b CAM 97-34-3 99-5a-14 95-10b-16 97-21b 1.26 BMW WAltflo2 5.59 7.9 99-5b-30 2.79 KJZ RGrande 7.19 99-3a-5 -1.09 TSS LNS42009 14.6 JEP AmLod 2 (1) 17.4 CAM 88409 use -7.9 99-la-14 -5.4 -5.2-5 ]MBSta. Cruz -2.71 -9.1 97-8b -152 99-la-4 -1.87 15 -18.8 HW 96(1) -34.6 -10.8 BMW -14.1 99-2a-23 -19.2 95-9a-22 -17.9 -17.5 95-10a-20 1.74 2.352.09 18.8 AW Acrenat -5.451 97-21a 252 -175 2.87 -37.72.26 Ins33708-1 -29.72.26 BMW M-S 1(1) -3.42 PcpePucacuro 3.04 -6.44 00-5a-6:2 plhk 3.22 2.782.96 -0.7 BMW AF Lbank (1) 261 -1.31 SM (SH) 57-2 natl 2.87 1528-1086 TSS 359 -17.435 99-3b-7 33 3.48 -10.8 95-9b-20 3.222.87 -8.49 -27 3.48 -27.4 95-8a 35 -4.05 96-5a-0 359 -26 1 3.13 2 2 7 3 4 2.78 -14.8 Ins65392-1 5 3 5 9 3.48 -41.1 CAM 98-rl4-6 5 4 2.96 -2.7 PCSJanauaca 3 304 10.25 Dieta 10-17 3.13 18 3.04 -19.3 SM 55-3 15 2.87 15 2.26 2.26 3 2.35 -25.4 2.17 5 2.17 -10.8 269 2.61 2 17 2.962.35 6 2.78 3.04 2.43 5 2.43 2.962.43 5 3 3.22 8.56 297 8 3.04 4.88 2.78 -2.67 2.17 4-02 2.17 226 3 2.26 -5.45148 2.09 2 2 2 4.68 1.83 5 2.78 -7.37 252 27 2.96 11.3 -8.36 3.2 24 -0.2 -16.3 -174 2.09 4 2.26 -19.8 217 -204 8 -7.76 2.87 3 3.04 -9.41 78 -7.21 -5.34 2.35 7 252 11 19 -9.31 3.22 4 3.48 17 -13.3 2.61 5 29 17.4 174 33 19 -2418 2.78 0 4 2.43 6 3.22 31 1619 5.25 -0.2 2.17 1.74 2 2.96 -0.2 H.m.p. Anav (publ.) 26 5.63 25 -24.5 2.43 3 20 -172 2.52 3 3.13 2.29 9 10 11 12 13 14 1.83 217 37 -324 2.17 3 2.17 -4.5 BMW Divisor PB 1 2.26 32 -3.42 2.17 2 2.09 2.69 2.26 2.78 13 -335 1.91 3 2 17 32 -36.4 2.17 3 3.32.26296 16 183.3 -11.7 16 3 3.22 -5.7 4 2 261 4 255 243 22 -152 2.17 2 2.43 2.35 17 2.43 72.35 0.6 2.78 7 2.87 14 -205 2.96 17 -38.7 243 5 184 247 2 8 18.1 2.52 30 3.04 156 1.74 -7.57 -5.84 304 -7.05 4 6 252 14 -0.28 165 -8.23 2.78 11 -6.12 3.13 49-4 16 2.96 5.13 20 -6.85 -0.3 -7.88 2.78 -0.8 296 -1.67 3.04 4 0 9 -271 2.78 18 -23.1 3.04 78 -199 19 10 -7.2 1510 553 -5.3 2.3513 22 2 13.1 10 -7.03 17 -2.25 2.69 17 -23.8 2.87 10 296 15 10 11 31 15 19 -02 2.78 16 -18 243 19 36 1.13 14 13 16 1.33 304 20 -02 3.13 40 -127 2.35 44 188 2.35 2 35 -11.4 1.83 19 2 2 32 -28.2 2 37 -11.1 243 13 -174 2.17 26 -20.3 2.17 27 2.872.17 16 21 -0.5 2.09 2.26 2.17 2.78 2.87 2.43 18 4.75 2.78 2.87 21 6 2.87 23 2.12 2.78 23 5 6 7 8 17 2.87 11 -14.2 2.87 1419 2.431114 10 2.17 424 14 19 15 209 1514 2.35 2.35 10 -2.81 252 13 14 2.26 7 3.75 252 8 11 48 1.56 66 7668 280 191 74 1 21 30 111 124 2 37 26 49 71 2.09 22 28 235 23 -16 2.43 21 3833 2.1741 2.5229 16 16 -09 -18.3 31 2.35 21 2.43 25 2.87 17 23 2.5225 26 2.26 -3.57 2.61 33 26137 16 243 16 374 28 2.26 14 -15.7 40 3029 2.7827 28 25 2.7837 -31.2 13 2.8726 2.78 23 20 -713 2.17 30 2.61 -144 304 2.43 14 2.87 11 -14 -26.81 25 2.9632 15 304 -346 3.3 14 179138 3.13 24 -34.6 2.17 51 117 209 51 387 217 45 100 191 138 2 17 32 132 2.17 36 -18.8 2.35 23 -31.5 0 0 0 0 11 15 17 16 17 10 11 11 13 13 74 76 37 33 39 39 53 42 37 41 99 91 97 30 33 20 30 37 32 28 26 29 45 51 34 30 35 67 65 77 54 39 42 91 4 121 117 117 108 9 18 16 19 15 68 80 29 81 77 45 33 39 86 36 41 153 143 100 79 48 854 639 115 894558 42 599 579 658 24 600 28 385 34 791 30 1115 76 1 2 3 5 319 88 62 6322 108 219 66 3 98 282 122 5 518 162 18 599 14 509 19 63915 32 452 32 16 702 133 19 572 24 20 650 114 38 48078 824 13 32 744 35 21 618 31 38 6049 766 682 12 13 61 739 15 20 627 36 22 3 10 3 3 154 41 3 2 115 2 8 543 2 4 474 2 5 537 147 666 53 30 48 945 588 858 55 13 534 17 42 514 4 444 754 954 974 36 1424 6 23 32 37 780 62 4 5 20 742 555 27 16 661 20 461 30 5 38 750 24 5 24 4 4 18 BR: Ama.;BR: W. Negro,opp. Manaus PE: PE: Pucacuro 2 PE: CushabalayPE: S. MartinPE: AcreBR: Acre;SerraBR: do Moa Amazonas;BR: Autazes MatoBR: Grosso; opp. Alta Floresta 3 3 2 BR: Am.; BR: Amazon, S. opp. Manaus Amazonas;BR: Tele 3 BR: Amazonas;BR: Jau NP ABR: m ;S. Amazon, opp. Manaus BO: PNNKMBO: CochabambaBO: BR: RondoniaBR: RondoniaBR: Rondonia;BR: Guajara-mirim E. CochabambaBO: BR: Pari;BR: opp. Humaita BR: Amazonas;BR: Janauaci L. 5 BR: Amazonas;BR: Janauari L. Amazonas;BR: Negro, R. Jacare Am.;BR: Negro, R. Sto. Antonio 5 5 ABR: m ; W. Negro, 5 opp. Manaus 5 BR: Amazonas;BR: Anavilhanas Amazonas;BR: Jau NP Amazonas;BR: opp. Manicori Amazonas;BR: Branco R. mouth 5 5 4 24 11 BO: PNNKM 4 BO: CochabambaBO: BR: Amazonas;BR: Jau NP Amazonas;BR: jau NP BO: Sta.BO: Cruz Amazonas;BR: Borba MatoBR: Grosso; Alta Floresta MatoBR: Grosso; Alta Floresta Pari;BR: Maripa Pari; 4 BR: Maripa Pari; 4 BR: opp. Humaita 4 86 CochabambaBO: 1023 4 11 BR: Amazonas;BR: Barcelos 5 14 495 41 BR: Amazonas;BR: Jau NP BR: Pari;BR: Caxiuana Amazonas;BR: Anavilhanas 5 26 BR: Amazonas;BR: Anavilhanas 5 BR: Amazonas;BR: Humaila 4 BR: Amazonas;BR: Autazes Amazonas;BR: B.Constant 5 Cushabalay PE: Cushabalay PE: ? Yasuni EC: NP,S Napo ? 7 EC: Morona-Santiago 2 spodiops spodiops minor sneth sneth spodiops spodiops pollens pollens pollens pollens pollens flaviviridis flaviviridis minor snetli minor minor pollens minor pollens minor pollens minor minor minor pollens minorminor pollens zosterops zosterops zosterops floviviridis zosterops zosterops gris zosteropszosterops floviviridis zosterops gris gris zosterops gris minor pollens minorminor pollens minorminor pollens minor pollens pollens minor pollens zasterops minor sneth minor minor zosterops minorminor minor snethminorminor snethminor sneth minor sneth sneth minor snethminorminorminor minor sneth sneth minor sneth zosteropszosterops gris gris minor pollens minor pollens minor pollens minorminor minor pollens pollens Complex Taxon Locality VT

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. D1ETA D1ETA 16(5) MBR 2 coll. 2 MBR D 9 Mao 1(3) 9844b 96-lb-0 9840b 1 9840b 3 9840a 2:1 DGC Camis nat? (2) 98-22a CAM 98-rl4-26 CAM98-rl5-U 98-23b TAP CAM Altflo88477(1) CAM 98-rl5-43 98-281)4 5.02 -4.59 -609 -0.19 PC Pto Inirida -9.43 9841a -5.09 16.1 BMW Alagoas nat 1 -2.76 TSS MD 2(1) 13.5 CAM 76086(1) -7.46 TSS Heath4 -34.8 AW Hemi copies-19.6 1 -14.9 -13 disjunct MBR (4) -15.6-19.3 D (1) 15 -21 D9 Mao8 min (3) -27 -14.9-22.6 98-24b:2 96-21)-0 1.91 1.74 -1.71 CAM LNS 88319(2) 1.91 1.83 1.91 2.09 -20.2 2.B7 2 -3.22 96-2b-13 1.91 -3.67 2.09 209 -15.5 2.35 -26.7 98-24b:l 3.39 -31.9 96-2b-5 2.17 2.612.17 13.3 22.2 22.17 22.3 -6.68 CAM Altflo88409b(4) 97-29b 2.09 2.09 -8.89 CAM Altflo6(l)2.26 -4.09 1 1 2.09 -13 9840a 1 1 2.17 1 7 2.96 -10.5 95-2b-25 7 2.17 5 3 2 1.91 8.21 CAM 75220(3) 2 3 2.09 -16.8 AA Xingu nat 2 2 12 13 14 15 Catalog Number 165 7 1.83 3.46 D Mao14 nat (1) 1.91 6 2.09 1.98 4 198 1.831.91 1 1.75 7 252 1.91 174 1 2 1.83 2 2.17 2 2.17 2 2.09 2 2.09 -3.02 TSS MD3(2) 2.09 2.09 2.26 2 2.26 -26 T SSM D natl(l) 11 147 0.8 1.48 10 2 -337 -6.24 1.74 -9.37 -4.59 2.09 2 2.09 -7.07 2.17 -9.29 -6.68 -3.02 -7.46 -2.76 1.91 2 -20.4 -10.2 165 9 2 -13.5 -11.4 -30.8 1.91 1 3.48 -27.1 11 11 18 15 13 15 1516 -089 16 -20.818 1.83 -15.2 1.74 -21 1 1 2.17 2.17 -21.5 98-24a 21 -10.6 2 31 21 -9.67 183 1 29 -4.12 2 27 0 2 5 2 -5.6 95-7a-23 46 3.58 2 2 2 3.58 CAM 75770 42 -18.5 33 49 40 9 10 1.91 166 20 1.3 1.39 1.91 25 1.74 1.91 1.91 1.74 30 19.1 2 1.91 31 8.21 1.91 30 -3.67 1.91 1.91 2.09 2.09 2.26 38 -34.9 2.35 3 2.35 -33.4 CAM Ziggy 13(3) 2.26 24 -13.9 1.83 1 2.26 2 2.26 2 2.17 252 34 -9.83 2.52 2 2.52 -9.83 904b 2.17 37 16.1 2.26 32 -4.09 2.26 2 2.61 35 13.3 2.61 2 2262.09 39 -22 2.17 2 2.26 -22 TSS MD4(1) 209 38 5.02 0.3 1.56 2.8 0 1.83 9.1 02 2.41 -2.35 2 31 2.76 2.17 6 217 -14.1 DIETA Ziggy(l)11 -5.51 -5.09 -5.24 11.6 209 29 -8.89 2.09 -14.6 209 27 -166 2.09 1 2.17 -14 6 95-14b-5(l) -29.7 2.09 44 0.4 2.09 1 3.04 -29.7 95-8b-20(2) -13.7 183 -19.3 -155 -21.5 -267 14 13 15 16 18 -14.9 20 -19.8 183 28 -34.8 2.26 20 -22.4 2.26 1 37 39 25 -155 32 26 16.532 217 11.6 20 2.35 22.2 30 2.17 13.5 2 235 2 2.35 39 38 1.91 1.91 23 1.91 39 1.56 29 0 1.83 1.91 29 -24 9 2.26 35 -26 2.87 2.17 31 304 21 3.39 26 -319 2.17 31 3.08 2 1 2.26 18 2.17 182.17 -15.12.35 17 1.91 18 -8.91 1.65 1 2.17 -15.1 2.09 5 6 7 8 50 2 18 -15.1 59 74 1 15 58 2.35 17 -196 6986 1.9147 21 1.74 24 -9.43 -1.71 60 2.17 23 -13.1 1.9145 21 1.56 87 209 54 2.17 15 -3.05 2.17 11 57 2.17 83 1.83 6777 2 27 -3.22 2 51 97 66 1.83 19 -17.1 1.74 20 -14.1 1.65 8 45 2.09 69 63 97 5452 1.65 1.74 30 1.63 82 2 42 -7.71 43 2.17 16 44 46 209 % 48 3.48 18 -27 1 64 98 2.17 58 2 8079 65 1 1.9170 21 26 1.83 -8.31 5.76 2.1760 36 49 174 25 11.1 2 52 1.65 25 2.47 65 92 107 2.09 25 -13 209 25 128 2.26 44 -25.7 154 2 34 -11.7 137189 2.09 28 -20.2 209 42 -178 1.91 1 190 2.17 20 -6 09 2 25 -15.4 2 1 2.17 108 1.89 55 -6 26 1.81 55 118 2 50 -136 209 41 -22.6 2 209 121 2.09 40 109 1.56 21 13.4 100 2.09 129 2 44 -138 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 84 85 30 34 29 45 22 53 90 87 96 81 41 32 35 28 30 26 28 61 39 82 78 40 4 138 115 112 107 110 130 189 150 122 185 109 129 67 86 98 64 49 79 45 116 100 111 153 101 148 357 208 2 3 159 113 139 100 109 80 339599 141 117 711 748 63 562 164 401 115 482 128 2 96 96 7 888 151 2 2 114 70 222 135 70 98 832 54 52 160 121 10 414 18 689 61 14 506 63 19 20 542 38 2023 789 810 172 76 1 1 1 18 11 8 201 476 677 22 177 163 683 1 13 804 1 17 851 1 24 753 25 1 20 822 301 22 8 3 965 406 190 2 IBS 2 4 8 532 663 170 141 222 7 6 5 594 370 183 86 2 14 710 33 3 2 11433 60 2 90 58 2 8 1337 3 2 33 3 2 314 104 131 65 VT 1 Carajas 3 Pari; BR: Amazonas;BR: Manaus BR: AmapaBR: 1 Amazonas;BR: ApuauR. 1 BR: AmapaBR: AmapaBR: AmapaBR: AmapaBR: 1 Amazonas;BR: Manaus 1 Amazonas;BR: opp.S. Gabriel 1 Amazonas;BR: Gabriel S. Amazonas;BR: S. Gabriel Amazonas;BR: Gabriel S. Am.;BR: W. Negro, opp. Manaus 2 2 5 6 438 PE: TambopataPE: TambopataPE: AlagoasBR: Alagoas;BR: Murici 3 3 Amazonas;BR: Barcelos Amazonas;BR: Barcelos Amazonas;BR: Barcelos Amazonas;BR: jad NP 2 2 4 4 412 Amazonas;BR: R. Preto Amazonas;BR: Gabriel S. Amazonas;BR: S. Gabriel Am.;BR: W. Negro, opp. Manaus 2 2 11 893 BR: Amazonas;BR: Manacapuru Amazonas;BR: Manaus Amazonas;BR: S. Gabriel 2 5 494 BR: BR: Caxiuana ParA; BR: Caxiuana ParA; BR: 3 BR: ParA; R. R. ParA; XinguBR: BR: Amazonas;BR: Manaus Amazonas;BR: Neblina NP BR: AmapaBR: PE: Madred eDios PE: CamiseaR. PE: TambopataPE: Tambopata AmapaBR: 3 3 3 2 95 Amazonas;BR: Manaus 1 14 Am.;BR: W. Negro,opp. Manaus 2 5 BR: MatoBR: Grosso; opp. Alta Floresta 3 2 PE: Madred eDios PE: Madred eDios 3 2 121 92 PE: PE: Heath CO: I’to Inirida GU: south GU: south BR: MatoBR: Grosso; opp. Alta Floresta MaloBR: Grosso; 3 opp. Alta Floresta 3 2 3 94 192 64 PE: Madred e Dios 3 2 82 52 zosterops zosterops zosterops zosterops gris gris gris gris gris gris zosterops zosterops zosterops zosterops zosterops zosterops zosterops zosterops zosteropszosterops zosterops zosteropszosterops zosterops zosterops zosterops zosterops zosteropszosterops gris zosterops gris zosterops gris zosterops gris zosterops gris zosterops naum naum zosteropszosterops zosterops zosterops zosterops zosterops zosterops zosterops zosterops zosterops zosterops zosterops zosteropszosterops zosterops zosteropszosterops zosterops zosterops zosteropszosterops zosterops zosterops zosterops zosterops zosterops zosterops zosterops zosterops zosteropszosterops zosterops zosterops zosteropszosterops gris zosterops zosteropszosterops zosterops zosterops zosterops zosterops zosteropszosterops zosterops zosterops zosterops zosterops zosterops gris zosterops zosterops zosterops gris zosterops zosteropszosterops gris gris Complex Taxonzosterops Locality zosterops gris

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX 2: TISSUE SPECIMENS SEQUENCED

Tissue numbers beginning with "B-" are from the Louisiana State Museum of Natural Science, UFPE=Universidade Federal de Pernambuco in Brazil, USNM=Smithsonian Intitution in Washington, D.C.

Specimen Tissue Number Locality Prepara tor H. z. naumburgae 1 AGA-052 (UFPE) BR: Pernambuco; Timbauba S. A. Roda H. z. gnseipectus 2 B-25556 BR: Para; Caxiuana Mario Cohn-Haft H. z. gnseipectus 3 B-25523 BR: Para; Carajas Mario Cohn-Haft H. z. griseipectus 4 B-35512 BR: Mato Grosso; opp. Alta Floresta Jason D. Weckstein H. z. griseipectus 5 B-25519 BR: Amazonas; Autazes Mario Cohn-Haft H. z. griseipectus 6 B-993 BO: La Paz; Puerto Linares Thomas S. Schulenberg H. z. griseipectus 7 B-9214 BO: Pando; Cobija Kenneth V. Rosenberg H. z. gnseipectus 8 B-11148 PE: Ucayali; Cerro Tahuayo Donna C. Schmitt H. z. flaviviridis 9 B-27808 PE: Loreto; Contamana Andrew W. Kratter H. z. flaviviridis 10 B-27692 PE: Loreto; Contamana Angelo P. Capparella H. z. flaviviridis 11 B-5603 PE: San Martin; Jirillo Tristan J. Davis H.;. zosterops 12 B-25576 BR: Amazonas; opp. S. Gabriel Mario Cohn-Haft H. z. zosterops 13 B-25461 BR: Amazonas; Jau NP Mario Cohn-Haft H. z. rothschildi 14 B-20303 BR: Amazonas; N Manaus Mario Cohn-Haft H. z. rothschildi 15 USNM B10934 GU: N. side Acari Mts. M. B. Robbins H. ;. rothschildi 16 USNMB11801 GU: W. bank upper Essequibo R. M. J. Braun H. z. rothschildi 17 B-25545 BR: Amapa; Porto Grande Mario Cohn-Haft H. m. minor I B-25557 BR: Para; Ilha de Mara jo, Breves Mario Cohn-Haft H. m. minor 2 B-25555 BR: Para; Caxiuana Mario Cohn-Haft H. m. minor 3 B-25522 BR: Para; Carajas Mario Cohn-Haft H. m. snethlageae 4 B-35492 BR: Mato Grosso; Alta Floresta Jason D. Weckstein H. m. snethlageae 5 B-14588 BO: Sta. Cruz; Catarata Arco Iris Curtis A. Marantz H. m. snethlageae 6 B-15068 BO: Sta. Cruz; Piso Firme John M. Bates H. m. snethlageae 7 B-25494 BR: Para; W. bank R. Tapajos, Maripa Mario Cohn-Haft H. m. snethlageae 8 USNM B03856 BR: Amazonas; Borba Mario Cohn-Haft H. m. snethlageae 9 B-25582 BR: Amazonas; opp. Humaita Mario Cohn-Haft H. m. snethlageae 10 B-36780 BR: Rondonia; E Guajara-mirim A. Aleixo H. m. SW 11 B-25580 BR: Amazonas; W Humaita Mario Cohn-Haft H. spodiops 12 B-22787 BO: La Paz; Cerro Asunta Pata Mario Cohn-Haft H. m. pallens 13 B-25520 BR: Amazonas; Autazes Mario Cohn-Haft H. m. pallens 14 USNM B03844 BR: Amazonas; L. Janauaca Mario Cohn-Haft H. m. pallens 15 B-25507 BR: Amazonas; Benjamin Constant Mario Cohn-Haft H. m. pallens 16 B-25403 BR: Amazonas; Jau NP Mario Cohn-Haft H. m. pallens 17 B-20248 BR: Amazonas; Anavilhanas Mario Cohn-Haft H. minimus 1 B-25528 BR: Parti; Carajas Mario Cohn-Haft H. minimus 2 B-25498 BR: Pari; W. bank R. Tapajds, Maripa Mario Cohn-Haft H. minimus 3 B-25418 BR: Amazonas; Borba Mario Cohn-Haft H. minimus 4 B-35326 BR: Mato Grosso; opp. Alta Floresta A. Aleixo H. minimus 5 B-35349 BR: Mato Grosso; Alta Floresta Jason D. Weckstein H. minimus 6 B-15325 BO: Sta. Cruz; Aserradero Moira Tristan J. Davis H. minimus NW 7 B-25575 BR: Amazonas; S. bank R. Uaupes Mario Cohn-Haft H. minimus NW 8 B-25471 BR: Amazonas; Jau NP Mario Cohn-Haft H. inomatus 9 B-25481 BR: Amazonas; R. Bariuau Mario Cohn-Haft H. inomatus 10 USNM B03847 BR: Amazonas; N Manaus Mario Cohn-Haft Todirostrum chrysocrotaphumB-25401 BR: Amazonas; Jau NP Mario Cohn-Haft Poecilotriccus latiroslris B-10907 PE: Ucayali; R. Abujao Donna C. Schmitt

104

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX 3: DNA SEQUENCES

Aligned 1026-base pair DNA sequences of all study specimens used in phylogenetic analyses. Letters A,T,C, and G refer to bases in nucleotides, read from the 5'- to 3'-end of the light strand, and grouped by codon. Where sequences do not differ from that of first specimen (H. z. naumburgae 1), shown in full, bases are indicated by a dot. All sequences fully double-stranded with no ambiguous base calls, except for H. z. rothschildi 17, which is based on light strand only, with six ambiguous calls; these are labeled "Y" (= T or C; n=5) and "N" (= uncertain base; n=l).

H. z . naumburgae 1 ATC TGC TTA GTC ACC CAG ATC CTC ACC GGC ATT CTA TTA H. z . g r is e ip e c tu s 2 A. . H. z. griseipectus 3 A. . H. z . c p r ise ip e c tu s 4 A. . H. z . g r is e ip e c tu s 5 A. ■ C. H. z . g r is e ip e c tu s 6 A. . . , T .C. H. z . g r is e ip e c tu s 7 A. . . T .r . H. z . g r is e ip e c tu s 8 .r . H. z . f l a v i v i r i d is 9 A. . . .T . . . C .. H. z . f l a v i v i r i d is 10 A. . . . . C .. H. z . f l a v i v i r i d is 11 A. . . . . c . . H. z . z o s te r o p s 12 A. . ... c.. H. z . z o s te r o p s 13 A. . . . . c . . H. z . r o th s c h il d i 14 ...... C. . A. . . . . c . . H. z . r o th s c h il d i 15 ...... c . . A. . . . . c . . H. z . r o th s c h il d i 16 ...... c . . A. . . . . c . . H. z . r o th s c h il d i 17 N...... C .. A. . . . . c . . H. m. m in o r 1 A. . . .A . . . G.T . . . c . . H. m. m in o r 2 A. . . .A . . . G.T . . . c . . H. m. m in o r 3 A. . . .A . . . G.T . . . c . . H. m. snethlageae 4 A. . . .A . . . G,T . . . c . . H. in. snethlageae 5 A. . . .A . . . G,T . . . c . . H. m. snethlageae 6 A. . . .A . . . G.T . . . c . . H. m. sn e th la g e a e 7 . .A . . . G.T . . . c . . H. iTi. snethlageae 8 A. . ..A ...G.T ... c.. H. m. s n e th la g e a e 9 A. , . .A . . . G.T . . . c . . H. m. s n e th la g e a e 10 A. . . .A . . . G.T . . . c . . H. 171. SW 11 A.T . .A . . . G.T . . . c . . H. s p o d iop 12 s . .T ..A ... G.T ...... T . . . c . . H. m. p a lle n s 13 A. . G, , . . . c . . H. iTi. p a lle n s 14 A. . G. . . . . c . . H. m. p a lle n s 15 A. . G. . . . . c . . H. in. p a lle n s 16 A. . G. . . . . c . . H. in. p a lle n s 17 A. . G. . . . . c . . H. m inim us 1 ACG ..A ... A. . ..T ... . ,c ... c.. H. m inim us 2 ACG . .A . . . A. . ..T ... , ,c ... c.. H. m inim us 3 ACG ..A ... A . .T . . . . ,r ... c.. H. m inim us 4 ACG ..A ... A. . ..T ... . ,r ... c.. H. m inim us 5 ACG . -A . . . A . .T . . . . .c ... c.. H. m inim us 6 ACG . .A ... A. . ..T ... . ,r . . . c . . H. m inim us NW 7 ACA . .A ... A. . ..T ... . .c ... c.. H. m inim us NW 8 ACA ..A ... A. . ..T ... . .c ... c.. H. in o m a tu s 9 ACA ..A . . . G . . ..T ... . ,c ... c.. H. in o m a tu s 10 ACA ..A . . . G . . . .T ... . .c ... c.. Todirostrum chrysocro taphum A. . ..A ... G . . . ,c ... c.. Poecilotriccus latirostris A.A ..A ... A. . . .T ... . ,c ... c..

105

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 GCC ATG CAT TAC ACA GCA GAT ACC ACC CTA GCA TTT ACA H. z . griseipectus 2 ...... H. z . griseipectus 3 ...... H. z . g r is e ip e c tu s 4 C ...... C... H. z . griseipectus 5 ...... H. z . griseipectus 6 ...... H. z . griseipectus 7 C ...... H. z . griseipectus 8 ...... H. z . flaviviridis 9 A ...... H. z . flaviviridis 10 A ...... H. z . f l a v i v i r i d is 11 A ...... H. z . z o s te r o p s 12 A ...... H. z . z o s te r o p s 13 A ...... H. z . r o th s c h il d i 14 A ...... C ...... H. z . r o th s c h ild i 15 A ...... C ...... H. z . r o th s c h il d i 16 A ...... C ...... H. z . rothschildi 17 A ...... C ...... H. m. m in or 1 A ...... C ...... G.. H. m. min or 2 A ...... G.. H. m. min or 3 A ...... G.. H. m. sn e th la g e a e 4 A ...... G.. H. m. sn e th la g e a e 5 A ...... G.. H. m. sn e th la g e a e 6 A ...... G.. H. m. sn e th la g e a e 1 A ...... G.. H. m. sn e th la g e a e 8 A ...... G.. H. m. sn e th la g e a e 9 A ...... G.. H. m. sn e th la g e a e 10 A ...... G.. H. m. SW 11 A ...... G.. H. sp o d iop 12 s A C .. T ...... H. m. p a lle n s 13 A ...... T ...... C ...... H. m. p a lle n s 14 A ...... T ...... C ...... H. /n. p a lle n s 15 A ...... T ...... C ...... H. m. p a lle n s 16 A ...... T ...... C ...... H. m. p a lle n s 17 A ...... T ...... C ...... H. minimus 1 A ..C ..T ...... C ...... H. minimus 2 A ..C ..T ...... C ...... H. minimus 3 A . .C . . T ...... C ...... H. minimus 4 A ..C ..T ...... C ...... H. minimus 5 A ..C ..T ...... C ...... H. minimus 6 A ..C ..T C ... G...... H. minimus NW 7 A ..C . . T ...... C ...... H. minimus NW 8 A ..C . . T ...... C ...... H. in o m a tu s9 A ..C ..T ...... C ...... H. in o m a tu s10 A ..C ..T ...... C ...... Todirostrum chrysocrotaphum . .T . .A . .C C . .A ...... Poecilotriccus latirostris . .T ...... C ...... C C . . .

106

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 TCA GTC CCC CAC ACA TGC CGT AAC GTC CAA TTC GGC TGA H. z . g r is e ip e c tu s 2 H. z . g r is e ip e c tu s 3 H. z . g r is e ip e c tu s 4 H. z . g r is e ip e c tu s 5 H. z . g r is e ip e c tu s 6 H. z . g r is e ip e c tu s 7 H. z . g r is e ip e c tu s 8 H. z . f l a v i v i r i d is 9 G H. z . f l a v i v i r i d is 10 H. z . f l a v i v i r i d is 11 T H. z . z o s te r o p s 12 T H. z . z o s te r o p s 13 H. z . r o th s c h il d i 14 H. z . r o th s c h ild i 15 H. z . r o th s c h ild i 16 H. z . r o th s c h ild i 17 H. m. m in or 1 H. m. m in or 2 H. m. m in or 3 H. m. sn e th la g e a e 4 H. m. sn e th la g e a e 5 H. m. s n e th la g e a e 6 H. m. sn e th la g e a e 7 H. m. sn e th la g e a e 8 H. m. sn e th la g e a e 9 H. m. sn e th la g e a e 10 H. m. SW 11 H. sp o d iop 12 s H. m. p a lle n s 13 H. m. p a lle n s 14 H. m. p a lle n s 15 H. m. p a lle n s 16 H. m. p a lle n s 17 H. minimus 1 H. minimus 2 H. minimus 3 H. minimus 4 H. minimus 5 H. minimus 6 H. minimus NW 7 H. minimus NW 8 H. in o m a tu s 9 H. in o m a tu s 10 Todirostrum chrysocrotaphum Poecilotriccus latirostris

107

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z. naumburgae 1 CTC ATC CGC AAC CTT CAT GCA AAC GGA GCC TCA TTC TTC H. z. griseipectus 2 ...... T ...... C ...... H. z. griseipectus 3 ...... T ...... C...... H. z. griseipectus 4 ...... T ...... H. z. griseipectus 5 H. z. griseipectus 6 H. z . griseipectus 7 H. z . griseipectus 8 H. z. flaviviridis 9 H. z . flaviviridis 10 H. z. flaviviridis 11 H. z. zosterops 12 H. z. zosterops 13 H. z . rothschildi 14 H. z. rothschildi 15 H. z . rothschildi 16 H. z . rothschildi 17 H. m. otin or 1 H. m. minor 2 H. m. o tin or 3 H. m. snethlageae 4 H. m. snethlageae 5 H. at. snethlageae 6 H. at. snethlageae 1 H. at. snethlageae 8 H. at. snethlageae 9 H. at. snethlageae 10 H. at. SW 11 H. spodiops 12 H. at. p a lle n s 13 H. at. p a lle n s 14 H. at. p a lle n s 15 H. at. p a lle n s 16 H. at. p a lle n s 17 H. atiniatus 1 H. atinimus 2 H. atiniatus 3 H. atiniatus 4 H. atiniatus 5 H. atiniatus 6 H. atiniatus NW 7 H. atiniatus NW 8 H. inomacus 9 H. in o m a tu s 10 Todirostrum chrysocrotaphum Poeci1otriecus latirostris

108

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z. naumburgae 1 TTT ATC TGT ATC TAC CTG CAC ATC GGA CGC GGA CTT TAT H. z . griseipectus 2 ...... T ...... H. z . griseipectus 3 ...... T ...... H. z. griseipectus 4 H. z. griseipectus 5 H. z. griseipectus 6 A H. z. griseipectus 7 A H. z . griseipectus 8 A H. z. flaviviridis 9 H. z . flaviviridis 10 H. z. flaviviridis 11 H. z . z o s te r o p s 12 H. z . z o s te r o p s 13 C H. z. rothschildi 14 H. z . rothschildi 15 H. z. rothschildi 16 H. z. rothschildi 17 H. m. m in or 1 H. m. m in or 2 . .C . . .c . . . T r H. m. m in or 3 . .C . . .c . .. T r H. m. snethlageae 4 . .C . . .c . . . , T c H. m. snethlageae 5 . .C . ..c ...... A . . . . . T. r H. m. sn e th la g e a e 6 . .C . ..c ...... A . . , T r H. m. snethlageae 7 . .C . ..c...... A . . . , , T c H. m. snethlageae 8 . .C . ..c ...... A . . . , T 0 H. m. snethlageae 9 . .C . .-C ... . T,c , H. m. snethlageae 10 . .C . . .c . . . , , T c H. m. SW 11 . .c . ..c ... T r H. sp o d iop 12 s . .c . ..c ... . . -T T c .c H. m. pallens 13 . .c . . .c T.c. . H. m. p a lle n s 14 . .c . . .c ... , T c H. m. p a lle n s 15 . .c . . .c T r , H. m. pallens 16 . .c . . .c ... T, c , H. m. p a lle n s 17 . .c . . .c T r H. m inim us 1 . .c . ..c ...... A ...... T . . .A . ., T, r ,.c H. minimus 2 . .c . . .c ...... A ...... T . .,. . .A . .. T, c ,.c H. m inimus 3 . .c . ..c ...... A ...... T . . .A . ., T r ,,c H. minimus 4 . .c . . .c ...... A ...... T . . -A . .. T,,0 . .c H. minimus 5 . .c . ..c ...... A ...... T . . .A . . T,c . .c H. minimus 6 . .c . ..c ...... A ...... T . . .A . . T r ,r H. minimus NW 7 . .c . ..c ...... A ...... T . ..A .. , T H. m inim us NW 8 . .c . . .c ...... A ...... T ...... A . ., T H. in o m a tu s 9 ..c ...... c ...... T . ..A .. . T, H. in o m a tu s 10 . .c . ..c ...... T . . .A . . . T, Todirostrum chrysocrotaphum . .c . . .c ...... A.A ...... A . .. T. .c Poecilotriccus latirostris ..c ...... A.A ...... A . . T, c ,

109

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 TAG GGA TCC TAC CTC TAG AAA GAA ACC TGA AAC ACC GGA H. z . g r is e ip e c tu s 2 H. z . g r is e ip e c tu s 3 H. z . g r is e ip e c tu s 4 H. z . g r is e ip e c tu s 5 H. z . g r is e ip e c tu s 6 H. z . g r is e ip e c tu s 7 H. z . g r is e ip e c tu s 8 H. z . f l a v i v i r i d is 9 H. z . f l a v i v i r i d is 10 H. z . f l a v i v i r i d is 11 H. z . z o s te r o p s 12 H. z . z o s te r o p s 13 H. z . r o th s c h ild i 14 H. z . r o th s c h ild i 15 H. z . r o th s c h ild i 16 H. z . r o th s c h ild i 17 H. m. m in or 1 H. m. m in or 2 H. m. m in or 3 H. m. sn e th la g e a e 4 H. m. snethlageae 5 H. m. snethlageae 6 H. m. sn e th la g e a e 7 H. m. sn e th la g e a e 8 H. m. sn e th la g e a e 9 H. m. sn e th la g e a e 10 H. m. SW 11 H. sp o d iop 12 s H. m. p a lle n s 13 H. m. p a lle n s 14 H. m. p a lle n s 15 H. m. p a lle n s 16 H. m. p a lle n s 17 H. minimus 1 H. minimus 2 H. minimus 3 H. minimus 4 H. minimus 5 H. minimus 6 H. minimus NW 7 H. minimus NW 8 H. in o m a tu s 9 H. in o m a tu s 10 Todirostrum chrysocrotaphum Poecilotriccus latirostris

110

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z. naumburgae 1 GTT ATC CTC TTA CTA ACC CTA ATA GCA ACA GCT TTC GTC H. z . g r is e ip e c tu s 2 H. 2 . griseipectus 3 H. 2 . griseipectus 4 H. 2 . griseipectus 5 H. 2 . griseipectus 6 H. 2 . griseipectus 7 H. 2 . griseipectus 8 H. 2 . flaviviridis 9 A.C H. 2 . flaviviridis 10 A.C H. 2 . flaviviridis 11 A.C H. 2 . z o s te r o p s 12 A.C H. 2 . z o s te r o p s 13 A.C H. 2 . rothschildi 14 A.C H. z . r o th s c h il d i 15 A.C H. z . r o th s c h il d i 16 A.C H. 2 . rothschildi 17 A.C H. m. m in or 1 H. m. m in or 2 . .. C.G ... . G. . T H. m. m in or 3 ... C.G . .. . G. . . .T . H. m. sn e th la g e a e 4 ... C.G .. , ,T H. m. sn e th la g e a e 5 . .. C.G .. . . T . H. m. sn e th la g e a e 6 ... C.G .. T H. m. sn e th la g e a e 7 ... C.G .. , T H. m. sn e th la g e a e 8 ... C.G .. , ,T , H. m. snethlageae 9 ... C.G .... G. . , ,T , H. m. sn e th la g e a e 10 . .. C.G .... G. . , ,T , H. m. SW 11 ... C.G .... G. . , . T , H. s p o d iop 12 s ... C. . . . , ,T , H. m. p a lle n s 13 ... G.. ... C.. .. , ,C H. m. p a lle n s 14 ... G. . ... c . . . . r . , H. m. pallens 15 ... G.. ... c . . , ,r H. m. p a lle n s 16 ... G. . ... c . . , .C H. m. p a lle n s 17 ... G.. ... c . . , r H. m inim us 1 . .A ...... c .. , ,T . . c ...... H. m inim us 2 ..AG.. ... c ...... G -, ,T ,. . c ...... H. m inim us 3 ..AG.. ... c ...... G ,. ,T ,. . c ...... H. m inim us 4 ..AG.. ... c ...... G ., ,T . . c ...... H. m inim us 5 ..AG.. ... c ...... G .. ,T ,. . c ...... H. m inim us 6 ..AG.. ... c ...... G ,, ,T . . c ...... H. minimus NW 7 ..AG.. ... c . . . , T . . c ...... H. m inim us NW 8 ..AG.. ... c.. . . ,T . . c ...... H. in o m a tu s 9 . .A ...... c .. , ,T . . c ...... H. in o m a tu s 10 . .A ...... c.. ., . ,T . . c ...... Todirostrum chrysocrotaphum A.C ...... c...... G. . . .G .. .G ... . ,r . . c ...... Poecilotriccus latirostris A.C .... .T ...... C . , T . . c ......

111

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 GGC TAC GTC CTC CCA TGA GGC CAA ATA TCT TTT TGA GGG H. z . g r is e ip e c tu s 2 H. z . g r is e ip e c tu s 3 H. z . g r is e ip e c tu s 4 <2

H. z . g r is e ip e c tu s 5 ^ H. z . g r is e ip e c tu s 6 H. z . g r is e ip e c tu s 7 H. z . g r is e ip e c tu s 8

H. z . f l a v i v i r i d is 9 ...... H. z . f l a v i v i r i d is 10 H. z . f l a v i v i r i d is 11 H. z . z o s te r o p s 12 H. z . z o s te r o p s 13 H, z , r o th s c h ild i 14 H. z . r o th s c h ild i 15 H. z . r o th s c h ild i 16 H. z . r o th s c h ild i 17 H. m. m in or 1 H. m. m in or 2 H. m. m in or 3 H. m. s n e th la g e a e 4 H. m. sn e th la g e a e 5 H. m. s n e th la g e a e 6 H. m. s n e th la g e a e 1 H. m. s n e th la g e a e 8 H. m. s n e th la g e a e 9 H. m. s n e th la g e a e 10 H. m. SW 11 H. s p o d iop 12 s H. m. p a lle n s 13 H. m. p a lle n s 14 H. m. p a lle n s 15 H. m. p a lle n s 16 H. m. p a lle n s 17 H. m inimus 1 H. minimus 2 H. m inimus 3 H. minimus 4 H. m inimus 5 H. m inimus 6 H. m inimus NW 7 H. m inimus NW 8 H. in o m a tu s 9 H. in o m a tu s 10 Todirostrum chrysocrotaphum Poecilotriccus latirostris

112

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 GCT ACA GTA ATC ACC AAC CTA TTC TCT GCC ATC CCC TAT H. z . g r is e ip e c tu s 2 T ...... H. z . g r is e ip e c tu s 3 T ...... H. z . g r is e ip e c tu s 4...... H. z . g r is e ip e c tu s 5 ...... H. z . g r is e ip e c tu s 6 ...... H. z . g r is e ip e c tu s 7 ...... H. z . g r is e ip e c tu s 8 ...... H. z . f l a v i v i r i d is 9 G. H. z . f l a v i v i r i d is 10 G . H. z . f l a v i v i r i d is 11 ...... H. z . z o s te r o p s 12 ...... H. z . z o s te r o p s 13 ...... H. z . r o th s c h il d i 14 C . . A ...... H. z. rothschildi 15 A ...... H. z . r o th s c h il d i 16 C . . A ...... H. z. rothschildi 17 A ...... H. m. m in or 1 T . . T ...... C ...... C H. m. m inor 2 T . . T ...... C H. m. m in or 3 T . . T ...... C H. m. s n e th la g e a e 4 T . . T ...... C H. m. s n e th la g e a e 5 T . . T ...... C H. m. s n e th la g e a e 6 T . . T ...... C H. m. s n e th la g e a e 7 T . . T ...... c H. m. s n e th la g e a e 8 T . . T ...... C H. m. sn e th la g e a e 9 T . . T ...... C H. m. sn e th la g e a e 10 T . . T ...... C H. m. SW 11...... T . . T ...... T ...... C H. sp o d iop 12 s ...... T ...... T . . T C H. m. p a lle n s13 T ...... T ...... T ...... C H. m. p a lle n s14 T ...... T ...... T ...... C H. m. p a lle n s15 T ...... T ...... T ...... C H. m. p a lle n s16 T ...... T ...... T ...... C H. m. p a lle n s17 T ...... T ...... T ...... C H. minimus 1 . .C ...... T . . T ...... T ...... C H. m inimus 2 . .C ...... T . . T ...... T ...... C H. m inim us 3 . .C ...... T . . T ...... T ...... C H. m inimus 4 . .C ...... T . . T ...... T ...... C H. m inim us 5 . .C ...... T . . T ...... T ...... C H. m inim us 6 . .C ...... T . . T ...... T ...... C H. m inimus NW 7 . .C ...... T ...... T . .C H. m inimus NW 8 . .C ...... T ...... T . .C H. in o m a tu s 9 . .C T . . T ...... T ...... T . .C H. in o m a tu s 10 . .C T . . T ...... T ...... T . .C Todirostrum chrysocrotaphum ...... T C .. T ...... T — Poecilotriccus latirostris T ...... T ..C

113

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 ATC GGC CAA ACA CTC GTA GAA TGA GCC TGA GGA GGA TTT H. z . g r is e ip e c tu s 2 H. z . g r is e ip e c tu s 3 H. z . g r is e ip e c tu s 4 H. z . g r is e ip e c tu s 5 H. z . g r is e ip e c tu s 6 H. z . g r is e ip e c tu s 7 H. z . c p r ise ip e c tu s 8 H. z . f l a v i v i r i d is 9 H. z . f l a v i v i r i d is 10 H. z . f l a v i v i r i d is 11 H. z . z o s te r o p s 12 H. z . z o s te r o p s 13 H. z . r o th s c h il d i 14 H. z . r o th s c h il d i 15 H. z . r o th s c h ild i 16 H. z . r o th s c h il d i 17 H. m. m in or 1 H. m. m inor 2 H. m. m in or 3 H. m. s n e th la g e a e 4 H. m. s n e th la g e a e 5 H. m. s n e th la g e a e 6 H. m. s n e th la g e a e 7 H. m. s n e th la g e a e 8 H. m. s n e th la g e a e 9 H. m. sn e th la g e a e 10 H. m. SW 11 H. sp o d iop 12 s H. m. p a lle n s 13 H. m. p a lle n s 14 H. m. p a lle n s 15 H. m. p a lle n s 16 H. m. p a lle n s 17 H. minimus 1 H. minimus 2 H. minimus 3 H. minimus 4 H. minimus 5 H. minimus 6 H. minimus NW 7 H. minimus NW 8 H. in o m a tu s 9 H. in o m a tu s 10 Todirostrum chrysocrotaphum Poecilotriccus latirostris

114

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z. naumburgae 1 TCA GTC GAC AAC CCC ACC CTC ACC CGA TTC TTC GCC ATC H. z . griseipectus 2 ...... ff. z . griseipectus 3 ...... H. z . griseipectus 4 ...... H. z . griseipectus 5...... A.. H. z. griseipectus 6 ...... ff. z . griseipectus 7 ...... ff. z . griseipectus 8 ...... ff. z . flaviviridis 9 ...... H. z . flaviviridis 10 ...... H. z. flaviviridis 11 ...... H. z. zosterops 12 ...... H. z . z o s te r o p s 13 ...... H. z . rothschildi 14 ...... H. z . rothschildi 15 ...... H. z . rothschildi 16 ...... H. z . rothschildi 17 ...... H. m. m in or 1 T . . T ...... T . . . H. m. min or 2 T . . T ...... T . . . H. m. min or 3 T . . T ...... T . . . H. m. sn e th la g e a e A T . . T ...... T . . . H. m. s n e th la g e a e 5 T . . T ...... T . . . H. m. sn e th la g e a e 6 T . . T ...... T . . . H. m. sn e th la g e a e 7 T . . T ...... T . . . H. m. sn e th la g e a e 8 T . . T ...... T . . . H. m. sn e th la g e a e 9 T . . T ...... T . . . H. m. s n e th la g e a e 10 T . . T ...... T . . . H. m. SW 11 T . . T ...... T . . . H. spodiops 12 T ...... T T . .T . .T H. m. p a lle n s13 T . . T ...... H. m. p a lle n s14 T . .T ...... H. m. p a lle n s15 T . . T ...... H. m. p a lle n s 16 T ...... H. m. p a lle n s17 T . . T ...... H. m inimus 1 T ...... A T . .T ...... H. m inimus 2 T ...... A T . .T ...... H. m inimus 3 T ...... A T . .T ...... H. minimus 4 ... ..T A T . .T ...... H. m inim us 5 T ...... A T . .T ...... H. minimus 6 T ...... A T . .T ...... H. minimus NW 7 T ...... A ...... T ...... H. m inimus NW 8 T ...... A ...... T ...... H. in o m a tu s 9 T ...... A T . .T ...... H. in o m a tu s 10 T ...... A T . .T ...... Todirostrum chrysocrotaphum ..C ..T T T ...... T ...... T Poecilotriccus latirostris T ...... T ...... T . . .

1 1 5

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z. naumburgae 1 CAC TTT CTC CTC CCA TTC ATC ATT GCA GGT CTC ACA TTA H. z . griseipectus 2 H. z . griseipectus 3 H. z . griseipectus 4 C ...... H. z . griseipectus 5 ...... G ...... A H. z . griseipectus 6 ...... G ...... H. z . griseipectus 7 ...... G C . . . H. z . griseipectus 8 ...... G C ... H. z. flaviviridis 9 ...... C ...... H. z . flaviviridis 10 ...... c ...... H. z . flaviviridis 11 ...... c ...... H. z. zosterops 12 ...... c ...... H. z. zosterops 13 ...... c ...... H. z . rothschildi 14 ...... c ...... H. z . rothschildi 15 ...... c ...... H. z. rothschildi 16 ...... c...... H. z. rothschildi 17 ...... C Y . . . H. m. m in or 1 C ...... T ...... C H. m. m in or 2 C ...... T ...... C H. m. m in or 3 C ...... T ...... C H. m. s n e th la g e a e 4 C ...... T ...... C H. m. snethlageae 5 C ...... T ...... C H. m. snethlageae 6 C ...... T ...... C H. m. snethlageae 7 C ...... T ...... C H. m. snethlageae 8 C ...... T ...... C H. m. snethlageae 9 C ...... T ...... C H. m. snethlageae 10 C ...... T ...... C H. m. SW 11 C ...... T ...... C H. sp o d iop 12 s C ...... T ...... T . .C ...... C ...... C H. m. pallens 13 C ...... T . . C ...... C ...... T H. m. p a lle n s 14 C ...... T . . C ...... C ...... T H. m. pallens 15 C ...... T . . C ...... C ...... T H. m. p a lle n s 16 C ...... T . .C ...... C ...... T H. m. pallens 17 C ...... T . . C ...... C ...... T H. minimus 1 C T ...... T ...... C ...... C.T H. minimus 2 C ...... T ...... T . . C ...... C ...... C.T H. m inimus 3 C ...... T ...... T . .C ...... C ...... C.T H. minimus 4 C ...... T ...... T . . C ...... C ...... C.T H. m inimus 5 C ...... T ...... T . . C ...... C ...... C.T H. m inimus 6 C ...... T ...... T . . C ...... C ...... C.T H. m inimus NW 7 C ...... T . . C ...... C ...... C.T H. m inimus NW 8 C ...... T . . C ...... C ...... C.T H. in o m a c u s 9 C ...... T ...... T . . C ...... C ...... C.T H. in o m a tu s 10 C ...... T ...... T . . C ...... C ...... C.T Todirostrum chrysocrotaphum C T C . .C ...... C Poecilotriccus latirostris C T T . . T C ..A ...... T

116

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 ATC CAT CTA ACC TTC CTC CAC GAA ACA GGT TCA AAC AAC H. z . g r is e ip e c tu s 2 H. z . g r is e ip e c tu s 3 H. z . g r is e ip e c tu s 4 H. z . g r is e ip e c tu s 5 H. z . g r is e ip e c tu s 6 T H. z . g r is e ip e c tu s 7 T H. z . g r is e ip e c tu s 8 T H. z . f l a v i v i r i d is 9 H. z . f l a v i v i r i d is 10 H. z . f l a v i v i r i d is 11 C H. z . z o s te r o p s 12 C H. z . z o s te r o p s 13 H. z . r o th s c h il d i 14 H. z . r o th s c h ild i 15 H. z . r o th s c h ild i 16 H. z . r o th s c h ild i 17 H. in. m in or 1 H. m. m in or 2 H. m. m in or 3 H. m. sn e th la g e a e 4 H. m. sn e th la g e a e 5 H. m. sn e th la g e a e 6 H. m. sn e th la g e a e 7 H. m. sn e th la g e a e 8 H. m. snethlageae 9 H. m. s n e th la g e a e 10 H. 771. SW 11 H. sp o d iop 12 s H. m. p a lle n s 13 H. 77i. p a lle n s 14 H. m. p a lle n s 15 H. in. p a lle n s 16 H. m. p a lle n s 17 H. m inim us 1 H. m inim us 2 H. m inim us 3 H. m inim us 4 H. m inim us 5 H. m inim us 6 H. m inim us NW 7 H. m inim us NW 8 H. in o m a tu s 9 H. in o m a tu s 10 Todirostrum chrysocrotaphum Poecilotriccus latirostris

117

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 CCC CTA GGC ATC CTC TCA AAC TGC GAT AAA ATT CCA TTC H. z . g r is e ip e c tu s 2 H. z . g r is e ip e c tu s 3 H. z . g r is e ip e c tu s 4 H. z . g r is e ip e c tu s 5 H. z . g r is e ip e c tu s 6 H. z . g r is e ip e c tu s 7 H. z . g r is e ip e c tu s 8 H. z , f l a v i v i r i d i s 9 H. z . f l a v i v i r i d is 10 H. z . f l a v i v i r i d is 11 T H. z . z o s te r o p s 12 H. z . z o s te r o p s 13 H. z . r o th s c h ild i 14 H. z. rothschildi 15 C H. z . r o th s c h ild i 16 H. z. rothschildi 17 c H. m. m in or 1 . .. TC. . r H. m. m in or 2 . .. TC. . . .T . H. m. m in or 3 ... TC. . . .T . H. m. sn e th la g e a e 4 .. . TC. . H. m. sn e th la g e a e 5 ... TC. . . ,C . H. m. sn e th la g e a e 6 ... TC. . . ,c . H. m. sn e th la g e a e7 . .. TC. . H. m. sn e th la g e a e 8 ... TC. . . .T . H. m. sn e th la g e a e 9 . .. TC. . . .T . .C . .0 . ,r , H. m. sn e th la g e a e 10 . .. TC. . . .T ., .C. . .0 . ,c , H. m. SW 11 . . . TC. . . .T . , .0 H. sp o d iop 12 s . .T .. .. TC. . . . G. . . .C . H. m. p a lle n s 13 .. . TC. . H. m. p a lle n s 14 ... TC. . . .T . H. m. p a lle n s 15 . .. TC. . . .T . H. m. p a lle n s 16 ... TC. . H. m. p a lle n s 17 . .. TC. . . .T . H. minimus 1 ... T...... C. . .. G.. ,c . ,c ,. .G H. minimus 2 ... .C. . . . G.. . .T . . ,c . H. minimus 3 ... T...... C. . . ,c . H. minimus 4 ... T...... C. . .. G.. . .T . . ,c .. .G . .. H. minimus 5 . .. T...... C. ... G.. . .c , H. m inim us 6 ... T...... C. . . . G. . . .T . . .c , H. m inim us NW 7 ... .C. . . . G. . . .T . . .c . H. minimus NW 8 ... .c. . .. G.. . .c . H. in o m a tu s 9 ... .c. . . . G. . . .T . . .c . H. in o m a tu s 10 Todirostrum chrysocrotaphum . . . T . . . .T ...T TC. . . . G. . . .T . Poecilotriccus latirostris ... TC. .

118

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 CAC CCC TAC TTC TCC ATA AAA GAC ATC CTA GGA TTC ATA H. z . g r is e ip e c tu s 2 H. z . g r is e ip e c tu s 3 H. z . g r is e ip e c tu s 4 C H. z . g r is e ip e c tu s 5 H. z . g r is e ip e c tu s 6 H. z . g r is e ip e c tu s 7 T H. z . g r is e ip e c tu s 8 T H. z . f l a v i v i r i d is 9 H. z . f l a v i v i r i d is 10 H. z . f l a v i v i r i d is 11 H. z . z o s te r o p s 12 H. z . z o s te r o p s 13 H. z . r o th s c h il d i 14 H. z. rothschildi 15 H. z . r o th s c h il d i 16 H. z . r o th s c h il d i 17 H. m. m in or 1 . . . T ...... T ...... T ...... C ...... H. m. m in or 2 . . . T ...... T ...... T ...... C ...... H. m. m in or 3 . . .T ...... T ...... T ...... C ...... H. m. s n e th la g e a e 4 . . . T ...... T ...... T ...... C ...... H. m. s n e th la g e a e 5 . . . T ...... T ...... T ...... C ...... H. m. snethlageae 6 . . . T ...... T ...... T ...... C ...... H. m. s n e th la g e a e 7 . . .T ...... T ...... T ...... C ...... H. m. s n e th la g e a e 8 . . .T ...... T ...... T ...... C ...... H. m. s n e th la g e a e 9 . . . T ...... T ...... T ...... C ...... H. m. s n e th la g e a e 10 . . . T ...... T ...... T ...... C ...... H. m. SW 11 . . .T ...... T ...... T ...... C ...... H. sp o d iop 12 s ...... T ...... C . .T . . . H. m. p a lle n s 13 ...... T ...... C ...... H. m. p a lle n s 14 ...... T ...... C ...... H. m. p a lle n s 15 ...... T ...... C ...... H. m. p a lle n s 16 ...... T ...... C ...... H. m. p a lle n s 17 ...... T ...... C ...... H. minimus 1 . ..T . . T ...... C ...... G H. miziimus 2 T . .T . .T ...... C ...... G H. minimus 3 T . .T . .T ...... C ...... G H. minimus 4 . ..T . . T ...... C ...... G H. minimus 5 . ..T . . T ...... C ...... G H. minimus 6 . ..T . .T ...... C ...... G H. minimus NW 7 . . .T . .T C . .T . .G H. minimus NW 8 . . .T . .T C . .T . .G H. in o m a tu s 9 . . .T . .T C . .T . .G H. in o m a tu s 10 . . .T ..T C . .T . .G Todirostrum chrysocrotaphum T ...... C...... T ...... C ...... Poecilotriccus latirostris . . . T ...... C T. . . .T . .T . .T

119

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 CTC CTC CTT CTC CCA CTA ATA ACC CTA GCT ATA TTC TCA H. z. griseipectus 2 ...... G...... H. z . g r is e ip e c tu s 3 ...... G...... H. z. griseipectus 4 H. z. griseipectus 5 H. z. griseipectus 6 H. z. griseipectus 7 H. z. griseipectus 8 H. z. flaviviridis 9 G H. z. flaviviridis 10 G H. z. flaviviridis 11 H. z. z o s te r o p s 12 C H. z . z o s te r o p s 13 H. z . r o th s c h ild i 14 H. z . r o th s c h ild i 15 H. z . r o th s c h ild i 16 H. z . r o th s c h ild i 17 H. m. m in or 1 H. m. m inor 2 H. m. m in or 3 H. m. sn e th la g e a e 4 H. m. sn e th la g e a e 5 H. m. sn e th la g e a e 6 H. m. sn e th la g e a e 7 H. m. sn e th la g e a e 8 H. m. sn e th la g e a e 9 H. m. s n e th la g e a e 10 H. m. SW 11 H. sp o d iop 12 s H. m. p a lle n s 13 H. m. p a lle n s 14 H. m. p a lle n s 15 H. m. p a lle n s 16 H. m. p a lle n s 17 H. minimus 1 H. minimus 2 H. minimus 3 H. m inimus 4 H. minimus 5 H. minimus 6 H. minimus NW 7 H. minimus NW 8 H. in o m a tu s 9 H. in o m a tu s 10 Todirostrum chrysocrotaphum Poecilotriccus latirostris

120

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z. naumburgae 1 CCA AAC CTT CTA GGC GAC CCA GAA AAT TTT ACA CCC GCC H. z. griseipectus T ..C ...... H. z . griseipectus ...... T . . C ...... H. z. griseipectus T ...... G ...... H. z . griseipectus ...... T ...... H. z. griseipectus T ...... H. z. griseipectus ...... T ...... H. z. griseipectus 8 T ...... H. z. flaviviridis 9 ...... C ...... G ...... H. z. flaviviridis 10 ...... C ...... G ...... H. z. flaviviridis 11 T . . C ...... G ...... H. z . z o s te r o p s 12 T . . C ...... G ...... H. z. z o s te r o p s 13 T . . C ...... G ...... H. z. rothschildi 14 T . . C ...... G ...... H. z. rothschildi 15 T . . C ...... G ...... H. z. rothschildi 16 T . .C ...... G ...... H. z. rothschildi 17 T . . Y ...... G ...... H. m. m in or 1 C . . C ...... H. m. m in or 2 ...... C . . C ...... T H. m. m inor 3 C . . C ...... H. m. sn e th la g e a e 4 C ..C ...... H. m. snethlageae 5 C ..C ...... H. m. snethlageae 6 C . . C ...... H. m. sn e th la g e a e 7 C . . C ...... H. m. snethlageae 8 C ..C ...... H. m. snethlageae 9 C ..C ...... H. in. snethlageae 10 C . . C ...... H. m. Sri 11 T C . . C ...... H. sp o d iop 12 s G . .T ..T C . .C ...... H. m. p a lle n s 13 c . . c ...... H. m. p a lle n s 14 ...... T C . . C ...... H. m. p a lle n s 15 C . . C ...... H. m. p a lle n s 16 c . . c ...... H. m. p a lle n s 17 c ..c ...... H. minimus 1 ...... T . .C T...... C . .C ...... H. minimus 2 T . .C T...... C . .C ...... H. minimus 3 T . .C T...... C . .C ...... H. minimus 4 T ,.C T...... C . .C ...... H. minimus 5 T . .C T...... C . .C ...... H. minimus 6 T ,.C T C . .C ...... H. minimus NW 7 ...... C ...... T C .. C ...... H. minimus tM 8 ...... C ...... T C .. C ...... H. in o m a tu s 9 T . .C ...... C ...... H. in o m a tu s 10 T . . C ...... C ...... Todirostrum chrysocrotaphum . .C C .. C ...... T Poecilotriccus latirostris . .T C . .C ......

121

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z. naumburgae 1 AAC CCA CTA GTA ACC CCT CCC CAT ATT AAG CCA GAATGG H. z. griseipectus 2 . .<7 . , A H. z. griseipectus 3 , .0 . .A H. z. griseipectus 4 . .A H. z. griseipectus 5 . .A H. z. griseipectus 6 H. z. griseipectus 7 . .A H. z. griseipectus 8 . .A H. z. flaviviridis 9 ...... T . , A . .A H. z. flaviviridis 10 ...... T H. z. flaviviridis 11 ...... T . .A H. z. z o s te r o p s 12 ...... T . .A H. z. z o s te r o p s 13 ...... T . .A H. z. rothschildi 14 ...... T . , A H. z. rothschildi 15 ...... T A H. z. rothschildi 16 ...... T . , A H. z. rothschildi 17 ...... T . .A H. m. m in or 1 . .0 A H. m. minor 2 ,r . .A M m mi nmr 3 . .c H. m. snethlageae 4 ,r . .A H. m. snethlageae 5 , ,c . .A . .A H. m. snethlageae 6 . .c . .A A H. m. snethlageae 7 . ,c . .A A H. m. snethlageae 8 , ,r . . A . .A H. m. snethlageae 9 . ,r . , A . .A H. m. snethlageae 10 . .c . .A . .A H. m. SW 11 . ,c . , A . , A H. sp o d iop 12 s . .c H. m. p a lle n s 13 . .T , T . .c H. 771. p a lle n s 14 . .T . ,c . .A H. m. p a lle n s 15 . .T . .c H. 771. p a lle n s 16 . .T T . .c H. m. p a lle n s 17 . .T , , T . .c . .A H. minimus 1 . .C . , T . .C . .A . .C H. minimus 2 . ,r . , T . .C . .A . .c H. m inim us 3 . .c . . T . .C . . A .c H. m inim us 4 . .c . T . .c . .A . .c H. minimus 5 . .r . , T . .c . .A . .c H. m inim us 6 . ,c . , T . .c . .A . .c H. minimus NW 7 . .c . T , ,r . .A . .0 H. minimus NW 8 . .C . . T , ,c .r , .0 H. inomatus 9 . .0 . .T ,c r ..(? H. in o m a tu s 10 . .0 . . T , , r , r . .0 Todirostrum chrysocrotaphum ...... T ...... T ...... C . . A ...... A Poecilotriccus latirostris C ..T .. C ...... A ...... A

122

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 TAC TTC CTA TTT GCC TAC GCC ATT CTC CGA TCA ATC CCC H. z . g r is e ip e c tu s 2 ...... H. z . g r is e ip e c tu s 3 ...... H. z. griseipectus 4 ...... H. z . g r is e ip e c tu s 5 ...... H. z . g r is e ip e c tu s 6 ...... H. z . g r is e ip e c tu s7 ...... H. z . g r is e ip e c tu s 8 ...... H. z . f l a v i v i r i d is 9 ...... H. z . f l a v i v i r i d is 10 ...... H. z . f l a v i v i r i d is 11 ...... H. z . z o s te r o p s 12 ...... H. z . z o s te r o p s 13 ...... H. z . r o th s c h il d i 14 ...... H. z . r o th s c h il d i 15 ...... H. z . r o th s c h il d i 16 ...... H. z . r o th s c h il d i 17 ...... H. m. m inor 1 C ...... C ...... C H. m. m inor 2 C ...... C ...... C H. m. m inor 3 C ...... C ...... C H. m. sn e th la g e a e 4 C ...... C ...... C H. m. sn e th la g e a e 5 C ...... C ...... C H. m. sn e th la g e a e 6 C ...... C ...... C H. m. sn e th la g e a e 1 C ...... C ...... C H. m. sn e th la g e a e 8 C ...... C ...... C H. m. snethlageae 9 C ...... C ...... C H. m. sn e th la g e a e 10 C ...... C ...... C H. m. SW 11 C ...... C ...... C H. sp o d iop 12 s C ...... C ...... C H. m. p a lle n s 13 C ...... c ...... C H. m. p a lle n s 14 C ...... C ...... C H. m. p a lle n s 15 C ...... C ...... C H. m. p a lle n s 16 C ...... C ...... C H. m. p a lle n s 17 C ...... C ...... C H. minimus 1 T H. minimus 2 ...... H. minimus 3 ...... H. minimus 4 ...... H. minimus 5 ...... H. minimus 6 ...... H. minimus NW 7 ...... H. minimus NW 8 ...... H. in o m a tu s 9 ...... H. in o m a tu s 10 ...... Todirostrum chrysocrotaphum T C . .T A ...... T ...... T Poecilotriccus latirostris C ...... T . .T . .C ...... G ...... T

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 AAC AAA CTG GGA GGA GTC CTG GCC CTT GCT GCC TCT GTC H. z . g r is e ip e c tu s 2 H. z . g r is e ip e c tu s 3 H. z . g r is e ip e c tu s 4 ...... C H. z . g r is e ip e c tu s 5 ...... C ...... H. z . g r is e ip e c tu s 6 A ...... H. z . g r is e ip e c tu s 7 H. z. griseipectus 8 H. z . f l a v i v i r i d is 9 T H. z . f l a v i v i r i d is 10 H. z . f l a v i v i r i d is 11 . . T .. , H. z . z o s te r o p s 12 H. z . z o s te r o p s 13 . . T .. . H. z . r o th s c h ild i 14 H. z . r o th s c h ild i 15 H. z . r o th s c h ild i 16 . . T .. , H. z . r o th s c h ild i 17 H. m. m in or 1 H. m. m in or 2 . .0 ...... A . . . . ,C . H. m. m in or 3 . .r, ...... A .. . . , ,r . H. m. sn e th la g e a e 4 . .T . . , .0 ...... A .. . . .c . H. m. sn e th la g e a e 5 H. m. sn e th la g e a e 6 . .G ...... A .. . . H. m. sn e th la g e a e7 , ,0 ...... A . . . . , ,r . H. m. sn e th la g e a e 8 . .0 ,r H. m. sn e th la g e a e 9 ,0 ...... A . . . . H. m. sn e th la g e a e 10 . . . .A ...... A . . . . H. m. SW 11 . . . .A , , . A , H. s p o d iop 12 s H. m. p a lle n s 13 . . . .A , . .0 , , , A . H. m. p a lle n s 14 H. m. p a lle n s 15 H. m. p a lle n s 16 H. m. p a lle n s 17 H. m inim us 1 H. m inim us 2 . .0 ., .. . .A .. . . . ,T . .A . H. m inim us 3 H. m inim us 4 H. m inim us 5 . .T . . . .0 . H. m inim us 6 . .T . . , .0 ...... A ...... T , .A . H. m inim us NW 7 . . . .A . . .c . H. m inim us NW 8 . . . .A , . .c . H. in o m a tu s 9 H. in o m a tu s 10 . .T . . . .0 , , . A . Todirostrum chrysocrotaphum . . T.A .. . . .A .. . . ,C , A . .T Poecilotriccus latirostris . . T.A . ..A ..A ..T . , A . T

124

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z . naumburgae 1 CTA ATT CTA TTT CTA GTC CCA TTC CTC CAC ATA TCA AAA H. z. griseipectus 2 C ...... H. z . g r is e ip e c tu s 3 C ...... H. z . g r is e ip e c tu s 4 H. z . g r is e ip e c tu s 5 H. z . g r is e ip e c tu s 6 H. z . g r is e ip e c tu s 7 H. z . g r is e ip e c tu s 8 H. z . f l a v i v i r i d is 9 H. z . f l a v i v i r i d is 10 H. z . f l a v i v i r i d is 11 ff. z. z o s te r o p s 12 H. z . z o s te r o p s 13 H. z . r o th s c h il d i 14 H. z . r o th s c h il d i 15 H. z. rothschildi 16 H. z. rothschildi 17 H. m. m in or 1 ...... C...... A ...... G H. m. m in or 2 ...... C...... A ...... G H. m. m in or 3 ...... C...... A ...... G H. m. s n e th la g e a e 4 ...... C...... A ...... G H. m. s n e th la g e a e 5 ...... C...... A ...... G H. m. s n e th la g e a e 6 ...... C...... A ...... G H. m. s n e th la g e a e 7 ...... C...... A ...... G H. m. s n e th la g e a e 8 ...... C...... A ...... G H. m. s n e th la g e a e 9 ...... C...... A ...... G H. m. s n e th la g e a e 10 ...... C A . . T ...... G H. m. SW 11 C ... .C...... A ...... G ff. sp o d iop 12 s ...... C...... ff. m. p a lle n s 13 C . . G ...... C...... ff. m. p a lle n s 14 C ...... C...... ff. m. p a lle n s 15 C . . G ...... C...... ff. in. p a lle n s 16 C ...... C...... H. m. p a lle n s 17 C ...... C...... H. minimus 1 ...... T H. m inimus 2 ...... T H. minimus 3 ...... T H. m inim us 4 ...... T H. minimus 5 ...... T H. minimus 6 ...... T H. minimus NW 7 ...... C...... T ...... H. minimus NW 8 ...... C...... T ...... H. inomaeus 9 ...... C...... T ...... H. in o m a tu s 10 ...... C...... T ...... Todirostrum chrysocrotaphum T.. . . C C ... .C...... T Poecilotriccus latirostris ...... C T. . .C...... T ......

125

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. H. z. naumburgae 1 CAA CGT ACA ATA ACC TTC CGC CCC CTT TCT CAA CTC CTA H. z . griseipectus 2 ...... C ...... H. z . griseipectus 3 ...... c ...... H. z . griseipectus 4 ...... c ...... H. z . griseipectus 5 H. z . griseipectus 6 T H. z . griseipectus 7 H. z . griseipectus 8 H. z . flaviviridis 9 C H. z . flaviviridis 10 C H. z . flaviviridis 11 H. z . z o s te r o p s12 H. z . z o s te r o p s13 H. z . rothschildi 14 H. z. rothschildi 15 H. z . rothschildi 16 H. z . rothschildi 17 H. m. m in or 1 H. m. m in or 2 ..C ...... A . . .c . . H. m. m in or 3 . .C ...... A . . .c . . H. m. snethlageae 4 . .C ...... A . . .c . . H. m. snethlageae 5 ..c ...... A . . .c . . H. m. snethlageae 6 . .c ...... A . . .c . . H. m. snethlageae 7 ..c ...... A . . .c . . H. m. snethlageae 8 ..c ...... A . . .c . . H. m. snethlageae 9 ..c ...... A . . .c . . H. m. snethlageae 10 ..c ...... A . . .c . . H. m. SW 11 ..c ...... A . . .c . . H. s p o d iop 12 s ..c ...... A . . .c . . H. m. pallens 13 , .0 .. .c ...... A . r . .c . . H. m. p a lle n s 14 .0 .,c ...... A . .c ,. .c . . H. m. p a lle n s 15 . .0 .,c ...... A . .c H. m. p a lle n s 16 . .G ,..c ...... A . .c . .c . . H. m. p a lle n s 17 , .0 . .c ... . ,c H. minimus 1 ...... A ...... c . . H. minimus 2 ...... A ...... c . . H. minimus 3 ...... A ...... c . . H. minimus 4 ...... A ...... H. minimus 5 . .c . . H. minimus 6 K. minimus NW 7 H. minimus NW 8 H. in o m a tu s 9 ...... A ...... H. in o m a tu s 10 ...... A ...... -C . . Todirostrum chrysocrotaphum , .G ..c ...... T . . . .c . ... AC. T.. Poecilotriccus latirostris ..c ...... T ...... A . c . .c ...... T ..

126

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ff. z. naumburgae 1 TTC TGA CTA CTA GTT TCA AAC CTC CTC GTC CTC ACA TGA ff. z. cpriseipectus 2 G ...... A...... ff. z. gpriseipectus 3 G ...... A...... ff. z. griseipectus 4 ...... A...... ff. z. griseipectus 5 ...... A...... ff. z. cpriseipectus 6 G ...... A...... ff. z. griseipectus 7 G ...... A...... ff. z. griseipectus 8 G ...... A...... ff. z. flaviviridis 9 ...... T ...... ff. z. flaviviridis 10 ...... T ...... ff. z. flaviviridis 11...... G ...... T ...... ff. z. zosterops 12...... G ...... T ...... ff. z. zosterops 13 ...... T ...... ff. z. rothschildi 14 ...... T ...... ff. z. rothschildi 15 ...... T ...... ff. z. rothschildi 16 ...... T ...... ff. z. rothschildi 17 ...... T ...... ff. m. m in or 1 T ...... A.T ...... ff. m. m in or 2 T ...... T A .T ...... ff. m. m in or 3 T ...... A.T ...... ff. m. s n e th la g e a 4 e T ...... T A .T ...... ff. m. sn e th la g e a 5 e T ...... T A .T ...... ff. m. s n e th la g e a 6 e T ...... T A .T ...... ff. m. sn e th la g e a 7 e T ...... T A .T ...... ff. m. snethlageae 8 T ...... T A .T ...... ff. m. sn e th la g e a e9 T ...... A.T ...... ff. m. s n e th la g e a e 10 T ...... A.T ...... ff. m. SW 11 T ...... A.T ...... ff. sp o d io p 12 s C ...... A...... G ff. m. p a lle n s 13 ..T ...... T .. . . C ...... A...... ff. m. p a lle n s14 ..T ...... T .. . . C ...... A...... ff. m. p a lle n s15 ..T ...... T .. . . C ...... A...... ff. m. p a lle n s16 ..T ...... T.. . . C ...... A...... ff. m. p a lle n s17 ..T ...... T. . . . C ...... A...... ff. m inimus 1 C ...... T A...... ff. m inim us 2 C ...... T A...... ff. m inim us 3 C ...... TA ...... ff. m inim us 4 C ...... T A...... ff. minimus 5 C ...... T...A...... ff. m inim us 6 C ...... T...A...... ff. m inim us NW 7 C ...... T A...... ff. minimus NW 8 C ...... T A...... ff. in o m a tu s9 C ...... T A...... ff. in o m a tu s10 C ...... T A...... Todirostrum chrysocrotaphum ...... T...... A...... Poecilotriccus latirostris T...... C ...... T ... A......

127

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ff. z. naumburgae 1 ATT GGT AGOCAA CCT GTA GAG CAA CCA TTT ATC ATT ATT ff. z. griseipectus 2 ff. z. griseipectus 3 ff. z. griseipectus 4 r ff. z. griseipectus 5 . .G ff. z. griseipectus 6 . .G . .G ff. z. griseipectus 7 . .0 , .G ff. z. griseipectus 8 . .A . .G ff. z. flaviviridis 9 ...... C ... ,A , , G ff. z. flaviviridis 10 ...... C ... . .A ,G ff. z. flaviviridis 11 ...... C ... A , G ff. z. z o s te r o p s 12 ...... c ... , , A . ,G ff. z. z o s te r o p s 13 ...... c ... , ,G ff. z. rothschildi 14 . .C ... ff. z. rothschildi 15 . ,G . .c ... ff. z. rothschildi 16 , G . .c ... ff. z. rothschildi 17 ...... Y ... . .G ff. m. m inor 1 . .G H. m. m inor 2 , ,G , .G ff. m. m inor 3 ff. m. snethlageae 4 . .A ... , ,G ff. m. snethlageae 5 ff. m. snethlageae 6 . .A ... . ,G ff. m. snethlageae 7 G . .A ... . .G ff. m. sn e th la g e a e 8 , ,G . .A G ,T ff. m. sn e th la g e a e 9 ...... C ... ff. m. snethlageae 10 ...... C ... , ,G , , A , T H. m. SW 11 ...... C ... ff. sp o d io p 12 s ff. m. p a lle n s 13 ...... C ... . .G . .A ... , ,G ...... c ff. m. p a lle n s 14 ...... C ... ff. m. p a lle n s 15 ...... c ... ff. m. p a lle n s 16 ...... c ... , , G . .A ... . .0 ...... c ff. m. p a lle n s 17 ...... c ... ff. minimus 1 ...... c ... . .C . .c ff. minimus 2 ...... c ... . .A r . ,c ff. minimus 3 ...... c ... ff. minimus 4 ...... c ... ff. minimus 5 ...... c ... ff. minimus 6 ...... c ... . .A ... . .A . .c . .c ff. minimus NW 7 ...... c ... . .A ... . .A . .c , r ff. minimus NW 8 ...... c ... . .A r . ,c ff. in o m a tu s 9 ...... c ... . .A ... . .A . .c . ,c ff. in o m a tu s 10 ...... c ... . .A r . .c Todirostrum chrysocrotaphum ...... c ... Poecilotriccus latirostris . .A . .c . .c

128

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ff. z. naumburgae 1 GGT CAA CTA GCC TCC ATC TCC TAC TTC TCC ATC CTC CTC ff. z. griseipectus 2 G ...... ff. z. griseipectus 3 G ...... ff. z. griseipectus 4 . . G ...... ff. z. griseipectus 5 ff. z. griseipectus 6 ff. z. griseipectus 7 ff. z. griseipectus 8 ff. z. flaviviridis 9 ...... T ...... ff. z. flaviviridis 10 ...... T ...... ff. z. flaviviridis 11 ...... T ...... ff. z. z o s te r o p s 12 ...... T ...... ff. z. zoscerops 13 ...... T ...... ff. z. rothschildi 14 ...... T ...... ff. z. rothschildi 15 ...... T ...... ff. z. rothschildi 16 ...... T ...... ff. z. rothschildi 17 ...... T ...... ff. m. m in or 1 ...... T ...... ff. m. m in or 2 ...... T ...... ff. m. m in or 3 ...... T ...... T ff. m. snethlageae 4 ...... T ...... ff. m. snethlageae 5 ...... T ...... ff. m. snethlageae 6 ...... T ...... ff. m. snethlageae 7 ...... T ...... ff. m. snethlageae 8 ...... T ...... ff. m. snethlageae 9 ...... T ...... ff. m. snethlageae 10 ...... T ...... ff. m. SW 11 C ...... T ...... ff. sp o d iop 12 s ...... T ...... T ...... T ...... ff. m. p a lle n s 13 ...... T . .T .. T ...... ff. m. p a lle n s 14 T . .T .. T ...... ff. m. p a lle n s 15 C ...... T . .T . .T ...... ff. m. p a lle n s 16 ...... T . .T .. T ...... ff. m. p a lle n s 17 ...... T . .T .. T ...... ff. minimus 1 ...... T ...... T ...... A . .T . ff. minimus 2 ...... T ...... T ...... A . .T . ff. minimus 3 ...... T ...... T ...... A . .T . ff. minimus 4 ...... T ...... T ...... A ... . ff. minimus 5 ...... T ...... T ...... A . .T . ff. m inim us 6 ...... T ...... T ...... A . .T . ff. minimus NW 7 ...... T ...... T ...... A . . . . ff. m inim us NW 8 ...... T ...... T ...... A . .. . ff. inomatus 9 . . .A . . T ...... T ...... A ... . ff. in o m a tu s 10 . . .A . . T ...... T ...... A ... . Todirostrum chrysocrotaphum G T ...... Poecilotriccus latirostris T . . . G.T . . T ...... A

129

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ff. z. naumburgae 1 ATT TTC TTC CCC CTC GTT GGA ACG CTA GAG AAC AAG C ff. z. griseipectus 2 ...... ff. z. griseipectus 3 ...... ff. z. griseipectus 4 ...... C . . . . .C ...... ff. z. griseipectus 5 ...... ff. z. griseipectus 6 ...... T ...... ff. z. griseipectus 7 ...... ff. z. griseipectus 8 ...... ff. z. flaviviridis 9 . . C ...... G ... . .A ... ff. z. flaviviridis 10 . . c ...... G ... . .A ... ff. z. flaviviridis 11 . . c ...... G . . . . .A ... ff. z. zosterops 12 . . c ...... G . . . . .A ... ff. z. zosterops 13 . . c ...... G . .A ... ff. z. rothschildi 14 . . c ...... A. . . .G . . . . .A ... H. z . r o th s c h ild i 15 . . c ...... G . .A . .A ... ff. z. rothschildi 16 . . c ...... G . .A . .A ...

ff. z. rothschildi 17 . . Y ...... A.. . .G . .A ..A ... o o o o ff. m. m in or 1 ...... A . .. A.T A.C ... G. A T.. . .A ... . .A T ff. m. m in or 2 ...... A ... A.T A.C ... G.A T.. ..A ... . .A T ff. m. m in or 3 ...... A . . . A.T A.C ... G. A T.. . .A ... . .A T ff. m. snethlageae 4 ...... A . . . A.T A.C . . . G.A T.. . .A ... . .A T ff. m. snethlageae 5 ...... A . . . A.T A.C . . . G.A T. . . .A ... . .A T ff. m. snethlageae 6 ...... A . .. A.T A.C . . . G.A T. . . .A ... . .A T ff. m. snethlageae 7 ...... A . . . A.T A.C . . . G.A T.. . .A ... . .A T ff. m. sn e th la g e a e 8 ...... A . . . A.T A.C . . . G.A T.. . .A ... . .A T ff. m. s n e th la g e a e 9 ...... A . . . A.T A.C .A. G.A T.. . .A ... . .A T ff. m. snethlageae 10 ...... A . .. A.T A.C • A. G.A T.. . .A ... . .A T ff. m. SW 11 ...... A . . . A.T A.C . . . G.A T.. . .A ... . .A ff. s p o d io p 12 s ...... A A.T A. . . .G G.A T. . . .A ... ff. jn. p a lle n s 13 . . C ...... A.T A.. . .G G.A T. . . .A ... ff. m. p a lle n s 14 . . C ...... A.T A.. . .G G.A T.. . .A ... ff. jn. p a lle n s 15 . . c ...... A.T A. . . .G G.A T. . . .A ... ff. m. p a lle n s 16 . . c ...... A.T A.. . .G G.A T.. . .A ... ff. m. p a lle n s 17 . . c ...... A.T A. . . .G G.A T.. . .A ... ff. m inim us 1 . . c ...... A.T A.C . .. G.A T. . . .A ... . .A ff. m inim us 2 . . c ...... A.T A.C . . . G.A T. . . .A ... . .A ff. m inim us 3 . - C ...... A.T A.C . . . G.A T. . . .A ... . .A ff. m inim us 4 . . c ...... A.T A.C . . . G.A T.. . .A ... . .A ff. m inim us 5 . . c ...... A.T A.C . . . G.A T.. . .A ... . .A ff. m inim us 6 . . c ...... A.T A.C ... G.A T.. . .A ... . .A ff. m inim us NW 7 . . c ...... A.T A.C ... G.A . .A ... ff. m inim us NW 8 . . c ...... A.T A.C . . . G.A . .A ... ff. incsmaCus 9 . . c ...... A.T A.C ... G.A . .A ... ff. in o m a tu s 10 . . c ...... A.T A.C ... G.A . .A ... Todirostrum chrysocrotaphum ...... T A. . A.C . .T . .A ... , A Poecilotriccus latirostris . . C ...... T , , A.TA.C.C. G.A » . . .A ... . .A

130

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ff. z. naumburgae 1 CTA AAC TTC TAA ff. z. griseipectus 2 ...... ff. z. griseipectus 3 ...... ff. z. griseipectus 4 ...... ff. z. griseipectus 5 ...... ff. z. griseipectus 6 ...... ff. z. griseipectus 7 ...... ff. z. griseipectus 8 ...... ff. z. flaviviridis 9 ...... ff. z. flaviviridis 10 ...... ff. z. flaviviridis 11 ...... ff. z. zosterops 12 ...... ff. z. zosterops 13 ...... ff. z. rothschildi 14 ...... H. z . r o th s c h ild i 15 ...... ff. z. rothschildi 16 ...... ff. z. rothschildi 17 ...... ff. m. m in or 1 ...... ff. m. m in or 2 ...... ff. m. m in or 3 ...... ff. m. s n e th la g e a e 4 ...... ff. m. snethlageae 5 ...... ff. m. snethlageae 6 ...... ff. m. s n e th la g e a e 7 ...... ff. m. snethlageae 8 ...... ff. m. snethlageae 9 ...... ff. m. s n e th la g e a e 10 ...... ff. m. SW 11 ...... ff. sp o d iop 12 s ...... ff. m. p a lle n s 13 ...... ff. m. p a lle n s 14 ...... ff. m. p a lle n s 15 ...... ff. m. p a lle n s 16 ...... ff. m. p a lle n s 17 ...... ff. m inimus 1 ...... ff. m inimus 2 ...... ff. m inimus 3 ...... ff. m inimus 4 ...... ff. m inim us 5 ...... ff. m inim us 6 ...... ff. m inim us NW 7 ...... ff. m inimus NW 8 ...... ff. in o m a tu s 9 ...... ff. inom atus 10 ...... Todirostrum chrysocrotaphum T C...... Poecilotriccus latirostris......

131

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX 4: PAIRWISE DISTANCE MATRIX

Distance matrix showing pairwise sequence divergences (uncorrected p) for all specimens sequenced.

Specim en 1 2 3 4 5 6 1 H. z. naumburgae 1 2 H. z . griseipectus 2 0.015 3 H. z. griseipectus 3 0.015 0.000 4 H. z . griseipectus 4 0.017 0.020 0.020 5 H. z . griseipectus 5 0.013 0.020 0.020 0.020 6 H. z . griseipectus 6 0.017 0.023 0.023 0.023 0.012 7 H. z. griseipectus 7 0.017 0.023 0.023 0.021 0 .012 0.006 8 H. z . griseipectus 8 0.017 0.023 0.023 0.023 0.012 0.006 9 H. z. flaviviridis 9 0.027 0.034 0.034 0.035 0 .035 0.037 10 H. z . flaviviridis 10 0.025 0.032 0.032 0.033 0.033 0.037 11 H. z . flaviviridis 11 0.026 0.033 0.033 0.034 0.032 0.036 12 H. z. zosterops 12 0.026 0.033 0.033 0.034 0.032 0.036 13 H. z . zosterops 13 0.024 0.031 0.031 0.032 0.030 0.034 14 H. z . rothschildi 14 0.029 0.035 0.035 0.034 0.034 0.038 15 H. z . rothschildi 15 0.029 0.035 0.035 0.033 0.034 0.038 16 H. z . rothschildi 16 0.029 0.035 0.035 0.033 0.034 0.038 17 H. z . rothschildi 17 0.027 0.033 0.033 0.030 0.031 0.035 18 H. m. minor 1 0.090 0.097 0.097 0.091 0 .097 0.094 19 H. m. minor 2 0.090 0.097 0.097 0.093 0 .097 0.094 20 H. m. minor 3 0.088 0.097 0.097 0.093 0 .097 0.094 21 H. m. snethlageae4 0.090 0.097 0.097 0.093 0 .097 0.094 22 H. m. snethlageae5 0.090 0.097 0.097 0.093 0.097 0.092 23 H. m. snethlageae6 0.090 0.097 0.097 0.093 0.097 0.092 24 H. m. snethlageae7 0.089 0.096 0.096 0.092 0.096 0.091 25 H. m. snethlageae8 0.089 0.096 0.096 0.092 0 .096 0.091 26 H. m. snethlageae9 0.089 0.098 0.098 0.094 0.096 0.093 27 H. m. snethlageae10 0.089 0.096 0.096 0.092 0.096 0.093 28 H. m. SW 11 0.090 0.093 0.093 0.090 0.095 0.092 29 H. spodiops 12 0.102 0.105 0.105 0.105 0.109 0.107 30 H. m. p a llen s 13 0.091 0.090 0.090 0.092 0.098 0.096 31 ff. m. p a llen s 14 0.091 0.090 0.090 0.092 0.098 0.096 32 ff. m. p a llen s 15 0.092 0.091 0.091 0.092 0.098 0.097 33 ff. m. p a llen s 16 0.088 0.087 0.087 0.089 0.095 0.093 34 ff. m. p a llen s 17 0.089 0.088 0.088 0.090 0.096 0.094 35 ff. minimus 1 0.108 0.109 0.109 0.109 0.115 0.113 36 ff. minimus 2 0.110 0.111 0.111 0.111 0.115 0.113 37 ff. minimus 3 0.110 0.111 0.111 0.111 0.115 0.113 38 ff. minimus 4 0.108 0.109 0.109 0.109 0.115 0.113 39 ff. minimus 5 0.108 0.109 0.109 0.109 0.115 0.113 40 ff. minimus 6 0.110 0.111 0.111 0.111 0.115 0.113 41 ff. minimus NW 7 0.107 0.108 0.108 0.109 0.113 0.111 42 ff. minimus NW 8 0.107 0.108 0.108 0.109 0.113 0.111 43 ff. in om atu s 9 0.112 0.109 0.109 0.110 0.116 0.114 44 ff. in om atu s 10 0.112 0.109 0.109 0.110 0.116 0.114 45 Todirostrum chrysocrotaphum0.125 0.126 0.126 0.128 0.129 0.129 46 Poecilotriccus latirostris 0.127 0.132 0.132 0.124 0.132 0.132 1 2 3 4 5 6

132

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 8 9 10 11 12 13 14 15 16 1 2 3 A 5 6 7 8 0.002 9 0.037 0.035 10 0.037 0.035 0.002 11 0.036 0.034 0.011 0.009 12 0.036 0.034 0.011 0.009 0.002 13 0.034 0.032 0.009 0.007 0.006 0.006 14 0.036 0.036 0.017 0.015 0.012 0.012 0.012 15 0.036 0.036 0.017 0.015 0.012 0.012 0.012 0.004 16 0.036 0.036 0.017 0.015 0.012 0.012 0.012 0.002 0.002 17 0.032 0.032 0.015 0.013 0.012 0.012 0.010 0.005 0.003 0.005 18 0.096 0.096 0.105 0.103 0.102 0.102 0.100 0.101 0.101 0.101 19 0.098 0.096 0.101 0.099 0.098 0.098 0.097 0.101 0.101 0.101 20 0.098 0.096 0.103 0 .101 0.100 0.100 0.098 0.103 0.103 0.103 21 0.098 0.096 0.099 0.098 0.098 0.0980.097 0.101 0.101 0.101 22 0.096 0.094 0.099 0.098 0.098 0.098 0.097 0.101 0.101 0.101 23 0.096 0.094 0.099 0.0980.098 0.098 0.097 0.101 0.101 0.101 24 0.095 0.093 0.098 0.097 0.098 0.0980.096 0.100 0.100 0.100 25 0.095 0.093 0.098 0.097 0.098 0.098 0.0960.100 0.100 0.100 26 0.097 0.095 0.098 0.097 0.098 0.098 0.096 0.102 0.100 0.102 27 0.097 0.095 0.100 0.098 0.099 0.099 0.0980.104 0.102 0.104 28 0.096 0.094 0.099 0.098 0.097 0.097 0.095 0.100 0.100 0.100 29 0.105 0.105 0.111 0.109 0.110 0.110 0.108 0.109 0.109 0.109 30 0.096 0.096 0.093 0.091 0.094 0.094 0.092 0.096 0.096 0.096 31 0.096 0.096 0.093 0.091 0.094 0.094 0.092 0.096 0.096 0.096 32 0.097 0.097 0.094 0.092 0.095 0.0950.093 0.097 0.097 0.097 33 0.093 0.093 0.090 0.088 0.091 0.091 0.089 0.093 0.093 0.093 34 0.094 0.094 0.091 0.089 0.092 0.092 0.090 0.094 0.094 0.094 35 0.113 0.113 0.114 0.112 0.113 0.113 0.107 0.113 0.111 0.113 36 0.113 0.113 0.116 0.114 0.115 0.115 0.109 0.115 0.115 0.115 37 0.113 0.113 0.116 0.114 0.115 0.115 0.109 0.115 0.115 0.115 38 0.113 0.113 0.114 0.112 0.113 0.113 0.107 0.113 0.113 0.113 39 0.113 0.113 0.114 0.112 0.113 0.113 0.107 0.113 0.113 0.113 40 0.113 0.113 0.116 0.114 0.115 0.115 0.109 0.115 0.115 0.115 41 0.109 0.109 0.111 0.109 0.112 0.112 0.106 0.113 0.113 0.113 42 0.109 0.109 0.111 0.109 0.112 0.112 0.106 0.113 0.113 0.113 43 0.112 0.112 0.114 0 .112 0.113 0.113 0.107 0.114 0.114 0.114 44 0.112 0.112 0.114 0.112 0.113 0.113 0.107 0.114 0.114 0.114 45 0.127 0.127 0.125 0.123 0.124 0.124 0.120 0.124 0.127 0.125 46 0.130 0.130 0.128 0.126 0.129 0.129 0.127 0.129 0.129 0.129 7 8 9 10 11 12 13 14 15 16

133

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17 18 19 20 21 22 23 24 25 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 0.096 19 0.096 0.008 20 0.098 0.008 0.004 21 0.096 0.012 0.006 0.008 22 0.096 0.012 0.006 0.008 0.004 23 0.096 0.012 0.006 0.008 0.004 0.000 24 0.095 0.011 0.005 0.007 0.003 0.001 0.001 25 0.095 0.011 0.005 0.007 0.003 0.001 0.001 0.000 26 0.094 0.015 0.013 0.013 0.013 0.011 0.011 0.012 0 .012 27 0.096 0.015 0.015 0.015 0.015 0.013 0.013 0.014 0 .014 0.006 28 0.094 0.021 0.017 0.019 0.019 0.019 0.019 0.018 0.018 0.016 29 0.104 0.067 0.063 0.065 0.061 0.061 0.061 0.062 0 .062 0.064 30 0.090 0.063 0.063 0.063 0.061 0.061 0.061 0.060 0.060 0.062 31 0.090 0.063 0.063 0.063 0.061 0.061 0.061 0.060 0.060 0.062 32 0.091 0.064 0.064 0.064 0.062 0.062 0.062 0.061 0.061 0.063 33 0.087 0.064 0.064 0.064 0.0620.062 0.062 0.061 0.061 0.063 34 0.088 0.063 0.063 0.063 0.061 0.061 0.061 0.060 0 .060 0.062 35 0.107 0.104 0.103 0.102 0.101 0.099 0.099 0.100 0.100 0.098 36 0.111 0.109 0.106 0.105 0.104 0.102 0.102 0.103 0 .103 0.103 37 0.111 0.109 0.106 0.105 0.104 0.102 0.102 0.103 0.103 0.103 38 0.109 0.107 0.104 0.103 0.102 0.100 0.100 0.101 0.101 0.101 39 0.109 0.107 0.104 0.103 0.102 0.100 0.100 0.101 0.101 0.101 40 0.111 0.109 0.106 0.105 0.104 0.102 0.102 0.103 0 .103 0.103 41 0.109 0.110 0.107 0.106 0.107 0.103 0.103 0.104 0 .104 0.106 42 0.109 0.110 0.107 0.106 0.107 0.103 0.103 0.104 0 .104 0.106 43 0.110 0.112 0.109 0.110 0.107 0.105 0.105 0.106 0.106 0.108 44 0.110 0.112 0.109 0.110 0.107 0.105 0.105 0.106 0.106 0.108 45 0.121 0.119 0.120 0.122 0.122 0.120 0.120 0.119 0.119 0.119 46 0.125 0.119 0.120 0.120 0.120 0.118 0.118 0.117 0.117 0.120 17 18 19 20 21 22 23 24 25 26

134

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27 28 29 30 31 32 33 34 35 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 0.016 29 0.067 0.061 30 0.062 0.059 0.060 31 0.062 0.059 0.058 0.002 32 0.063 0.058 0.060 0.001 0.003 33 0.063 0.060 0.059 0.003 0.003 0.004 34 0.062 0.059 0.060 0.002 0.002 0.003 0.001 35 0.098 0.101 0.100 0.096 0.0960.097 0.095 0.096 36 0.102 0.104 0.101 0.095 0.095 0.096 0.094 0.095 0.008 37 0.102 0.104 0.101 0.095 0.095 0.096 0.094 0.095 0.008 0.000 38 0.100 0.102 0.099 0.093 0.093 0.094 0.092 0.093 0.006 0.004 39 0.100 0.102 0.099 0.093 0.093 0.094 0.092 0.093 0.007 0.003 40 0.102 0.104 0.101 0.095 0.095 0.096 0.094 0.095 0.008 0.002 41 0.105 0.109 0.099 0.092 0.090 0.093 0.091 0.092 0.038 0.034 42 0.105 0.109 0.099 0.092 0.090 0.093 0.091 0.092 0.038 0.034 43 0.107 0.109 0.101 0.097 0.097 0.098 0.096 0.097 0.032 0.032 44 0.107 0.109 0.101 0.097 0.097 0.098 0.096 0.097 0.032 0.032 45 0.120 0.115 0.119 0.120 0.120 0.121 0.117 0 .118 0.120 0.122 46 0.119 0.117 0.122 0.112 0.112 0.113 0.110 0.111 0.115 0.116 27 28 29 30 31 32 33 34 35 36

135

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37 38 39 40 41 42 43 44 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 0.004 39 0.003 0.003 40 0.002 0.004 0.003 41 0.034 0.034 0.033 0.034 42 0.034 0.034 0.033 0.034 0.000 43 0.032 0.030 0.029 0.032 0.018 0. 018 44 0.032 0.030 0.029 0.032 0.018 0. 018 0.000 45 0.122 0.122 0.122 0.124 0.123 0. 123 0.123 0.123 46 0.116 0.114 0.114 0.116 0.107 0. 107 0.111 0.111 0. 119 37 38 39 40 41 42 43 44 45

136

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. VITA

Mario Cohn-Haft was bom in Northampton, Massachusetts, on

December 12,1961. At an early age, he developed a passionate interest in

nature that was encouraged and fostered by his family and by the teachers in the Northampton public schools and the Massachusetts Audubon Society's

Arcadia Day Camp. He began to focus seriously on birds at age fourteen. His

first exposure to the Neotropics was in 1982, in a field course on tropical

ecology in Costa Rica and Panama, during his undergraduate studies at

Dartmouth College, where he majored in biology. After several post-graduate years taking short-term jobs alternating between biological field research and

singing, he went to the Brazilian Amazon in 1987. Since then, he has essentially never left, returning to the United States only long enough to complete a master

of science degree at Tulane University and a doctorate at Louisiana State

University, both degrees focused on Amazonian birds. Also in Brazil, he met

his wife, Rita Mesquita, whom he married in 1991. They live, with their dog Butch, in Manaus, Brazil, where they plan to dedicate the rest of their careers to

research, teaching, and conservation efforts in the Amazon.

137

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DOCTORAL EXAMINATION AND DISSERTATION REPORT

Candidate: Mario Cohn-Haft

Major Field: Zoology

Title of Dissertation: a Case Study in Amazonian Biogeography: Vocal and DNA-sequence Variation in Hemitriccus Flycatchers

Approved:

jor Professor/and Chairman

Dei le L Graduate School EXAMINING/ COMMITTEE; yUjLUcJ/

Date of v-ravnijia.tion:

October 25, 2000

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.