Cryptic Diversification of Two Widespread Species in Madagascar

Home , Ambarijeby, Amboasary Sud, Ambodiangezoka, Andranomena, Anjozorobe, Antsahabe

Loyola University Chicago Loyola eCommons

Master's Theses Theses and Dissertations

2018

Cryptic Diversification of woT Widespread Species in Madagascar

Lynika Sharlice Strozier Loyola University Chicago

Follow this and additional works at: https://ecommons.luc.edu/luc_theses

Part of the Biology Commons

Recommended Citation Strozier, Lynika Sharlice, "Cryptic Diversification of woT Widespread Species in Madagascar" (2018). Master's Theses. 3708. https://ecommons.luc.edu/luc_theses/3708

This Thesis is brought to you for free and open access by the Theses and Dissertations at Loyola eCommons. It has been accepted for inclusion in Master's Theses by an authorized administrator of Loyola eCommons. For more information, please contact [email protected].

LOYOLA UNIVERSITY CHICAGO

CRYPTC DIVERSIFICATION OF TWO WIDESPREAD SPECIES IN MADAGASCAR

A DISSERTATION SUBMITTTED TO

THE FACULTY OF THE GRADUATE SCHOOL

IN CANDIDACY FOR THE DEGREE OF

MASTER OF SCIENCE

PROGRAM IN BIOLOGY

LYNIKA S. STROZIER

CHICAGO, IL

MAY 2018

ACKNOWLEDGMENTS

I would like to thank all those who made this thesis possible. First, I would like to express my sincere gratitude to my advisor Dr. Reddy for her continued support throughout my studies.

Without her support throughout this process I do not believe I would have made it this far and this defense would not have been possible. A very special thanks to Sarah Kurtis who assisted with the laboratory experiments and thesis edits. She was very supportive and encouraging throughout my studies. I would also like to thank Dr. Jane Younger for her support throughout my thesis; Chris Kyriazis who helped with the thesis edits; my collaborator Steve M. Goodman whom helped throughout my thesis studies and who gave great feedback and suggestions; Dr.

Matthew Von Konrat who was very encouraging and supportive.

I would also like to thank my thesis committee Dr. Reddy, Dr. Grande, Dr. Moreau and Dr.

Laten. I appreciate the advice and feedback they have provided me. I would also like to thank

Virginia Lorenzo and Frank Inglima for their support and being able to answer any of my questions and concerns. I would also like to give a special thank you to Audrey Berry for being so helpful and supportive throughout my studies at Loyola.

I would also like to thank my family who has been there for me. They were able to give love, support and encouragement. I am especially grateful to have Mario Trujillo by my side, encouraging me to never give up and to keep striving for success. I realize that my thesis work would not be completed if it had not been for assistance from

iii

Loyola University Chicago. Finally, I would like to thank my grandmother Sharon Wright for keeping me on track. I could not have made it this far without my grandmother’s encouragement, constant support and her always pushing me to never give up no matter how hard it gets.

To my grandmother, my aunt, and my Mom

TABLE OF CONTENTS

ACKNOWLEDGMENTS ...... iii

LIST OF FIGURES ...... vii

LIST OF TABLES ...... ix

ABSTRACT ...... x

CHAPTER ONE: THE IMPORTANCE OF PHYLOGEOGRAPHIC STUDIES AMONG BIRDS ON MADAGASCAR ...... 1

CHAPTER TWO: USING MOLECULAR TOOLS TO DETERMINE DIVERSIFICATION PATTERNS AMOUNG GENUS NEWTONIA ...... 18

CHAPTER THREE: DETERMINING THE GENETIC DIVERSITY AMONG GENUS NEWTONIA ...... 38

CHAPTER FOUR: UNCOVERING HIDDEN DIVERSITY AMONG TWO WIDESPREAD SPECIES OF BIRDS N. AMPHICHROA AND N. BRUNNEICAUDA ...... 64

CHAPTER FIVE: IDENTIFYING DISTINCT SPECIES PRESENT WITHIN N.AMPHICHROA ...... 72

REFERENCE LIST ...... 74

VITA ...... 78

LIST OF FIGURES

Figure 1. The phylogeographic study among Dicrurus forticatus...... 12

Figure 2. The phylogeographic study among Bernieria madagascariensis...... 13

Figure 3. The phylogeographic study among Xanthomixis zosterops...... 14

Figure 4. The phylogeographic study among Genus Nesillas...... 15

Figure 5. Images of the three Newtonia species were taken at the Field Museum of Natural History...... 16

Figure 6. Map of sampling localities of the two species a) N. amphichroa and b) N. brunneicauda on Madagascar...... 37

Figure 7. Phylogenetic relationships of Newtonia using maximum likelihood and Bayesian inference of the combined five genes...... 50

Figure 8. Phylogenetic relationships of Newtonia using maximum likelihood of ND3 only...... 51

Figure 9. Phylogenetic relationships of Newtonia using maximum likelihood of CYTB only...... 52

Figure 10. Phylogenetic relationships of Newtonia using maximum likelihood of FIB5 only...... 53

Figure 11. Phylogenetic relationships of Newtonia using maximum likelihood of GAPDH only...... 54

Figure 12. Phylogenetic relationships of Newtonia using maximum likelihood of MUSK only...... 55

Figure 13. GenGIS analysis for N. amphichroa. Distinct clades are associated with isolated montane regions found along the eastern humid forest...... 56

Figure 14. GenGis analysis for N. brunneicauda. There are no distinct clades are associated with geographic regions...... 57

vii

Figure 15. The Median-Joining Haplotype Network of CYTB haplotypes within N. amphichroa...... 59

Figure 16. The Median-Joining Haplotype Network of CYTB haplotypes within N. brunneicauda...... 60

Figure 17. Box-and-whiskers plot showing elevation distribution of N. amphichroa and N. brunneicauda...... 63

viii

LIST OF TABLES

Table 1. A physical description of the three species of Newtonia...... 17

Table 2. Sample accession number, locality, elevation, latitude, and longitude information for each specimen of Newtonia used in this study...... 25

Table 3. Primers used for this study...... 36

Table 4. DNA sequences obtained using five Markers...... 45

Table 5. Population genetics statistics for three species of Newtonia...... 58

Table 6. Species delimitation (BPP) analysis...... 61

Table 7. Elevation ranges for each species set comparing the elevation ranges of the different Newtonia species and the misidentified samples...... 62

ABSTRACT

Using phylogenetic and phylogeographic tools to uncover hidden diversity within the genus

Newtonia in Madagascar

Madagascar is known for its rich biodiversity and high level of endemic species that are found nowhere else. Cryptic diversification, defined as genetically and evolutionarily distinct species that are hard to detect because they are morphologically indistinguishable from their closest relatives, has been hypothesized to occur in many groups on Madagascar. Currently it is unclear to what extent this phenomenon occurs in birds because only a few studies have been conducted. My study examined the phylogenetic and phylogeographic patterns within a songbird genus, Newtonia, that is endemic to Madagascar. This genus includes four species: N. amphichroa, N. brunneicauda, N. archboldi, and N. fanovanae. The objective of my study was to determine the following: 1. What are the phylogenetic relationships among the species of

Newtonia? 2. Is there phylogeographic structure to indicate cryptic differentiation or recent speciation within Newtonia species? 3. If there is evidence of differentiation, are lineages diverging due to habitat type or geography? 4. Are there misidentified specimens or hybrids in certain geographic areas or elevations? I used DNA sequences to examine the evolutionary relationships and diversification patterns among and within the Newtonia species. I conducted a phylogenetic analysis using two-mitochondrial loci, ND3 and CYTB, and three nuclear loci,

GAPDH, FIB5, MUSK. The phylogenetic

x analysis showed strong support for monophyly among the three Newtonia species. I examined geographic structure within two widespread species to determine the potential for cryptic species. My study found that N. amphichroa is divided into two deeply divergent clades associated with distinct regions in the eastern humid forest. The two major clades of N. amphichroa split 1.25 MYA, indicating a deep split within this population showing that they have been separated for quite some time and on the order of other sister species pairs. Within

Newtonia, N. archboldi separated from N. amphichroa and N. brunneicauda roughly 7.36 MYA and N. amphichroa separated from N. brunneicauda 4.38 MYA. Both the species delimitation analysis and the molecular clock analyses showed that the two divergent clades of N. amphichroa are distinct species that have been separated long enough to be considered different species. However, N. brunneicauda did not show any geographic structure and no genetically distinct lineages to indicate segregation within distinct areas. I further examined the issue of misidentifications - some sequences from individuals identified in the field as N. brunneicauda consistently grouped within the N. amphichroa clade and vice versa in the phylogenetic analysis.

I was able to rule out lab error by repeating extractions and sequencing. I then examined museum specimens of these species in a comparative series, and found that although the individuals of the two species are similar to one another, there are some features that clearly separate these species.

Therefore, careful examination of plumage indicated that these individuals were misidentified in the field. My study is an important step in better understanding the phylogenetic relationships and the phylogeographic patterns of endemic birds of Madagascar in order to uncover hidden diversity.

CHAPTER ONE

INTRODUCTION

THE IMPORTANCE OF PHYLOGEOGRAPHIC STUDIES AMONG

BIRDS ON MADAGASCAR

Madagascar is the fourth largest island in the world and has long been recognized as one of the world’s biodiversity hotspots because of its high number of endemic species, which are found nowhere else (Myers et al. 2000; Goodman and Benstead 2003). Madagascar separated from Africa 183–158 Ma, from India 96–65 Ma, and from Antarctica 130 Ma during the breakup of Gondwana and has been isolated ever since (Vences et al. 2009). Most modern species colonized Madagascar well after this time.

Madagascar’s long-term isolation and few colonization events make it a model region for studying how species evolve (Vences et al. 2009). A key aspect to understanding evolutionary patterns is addressing fundamental questions of what constitutes distinct species, how many species are in a given area, and what are the phylogenetic relationships of these species (Cracraft

2002). Madagascar’s evolutionary and biogeographical isolation makes it model system for explaining stochastic and deterministic influences on diversity patterns (Vences et al. 2009).

Over the last two millennia, Madagascar forests have been reduced substantially, and it is estimated that only between 10% and 20% of primary forest cover is left on the island (Watson et al. 2004). The degradation of the remaining forests is occurring as a consequence of many factors: most notably logging, slash and burn agriculture (Tavy), charcoal

1 2

production, and fuel wood collection (Watson et al. 2004). Deforestation and habitat degradation have been shown to have adverse impacts on forest bird communities and many other species

(Watson et al. 2004). With continued anthropogenic impacts in Madagascar, scientists predict the further decline of many bird species (Watson et al. 2004). They found over time that the forest edge had significantly fewer trees, less canopy cover, less litter cover, and greater shrub cover than the core habitat (Watson et al. 2004). The species were affected by internal degradation of forest habitat due to the loss of the larger trees. This type of degradation loss can influence species richness and individual species abundance (Beier et al. 2002).

Bird Diversity on Madagascar

Because Madagascar has a rich endemic flora, it is one of the top hotspots in the world for biodiversity conservation (Vences et al. 2009). Madagascar has a rather species-poor bird community at 209 species. A high proportion of species at 51% are endemic to Madagascar.

There are seven families and 60 genera that are endemic to the region (Safford et al. 2013).

These include Mesitornithidae (two genera, three species), Brachypteraciidae (four genera, five species), Leptosomatidae (one species), Philepittidae (two genera, four species), Bernieridae

(seven genera, ten species), and Vangidae (fourteen genera, twenty one species) (Safford et al.

2013). The Vangidae family is important because genus Newtonia belongs to this family and this family is a great example of adaptive radiation that has yielded a wide range of foraging strategies as seen by the striking difference in bill morphology that allowed them to exploit diverse foraging niches (Reddy et al.2012; Jonssson et al. 2012). Bernieridae is a Malagasy endemic family that contains seven genera and ten species (Block 2012). The Bernieridae family

is intriguing because it is the first family that provides an example of the effects of cryptic diversity when using diversification analysis (Block 2012). Cryptic diversification is defined as genetically and evolutionarily distinct species that are hard to detect because they are morphologically indistinguishable from their closest relatives. As a result of this study, I became interested in exploring if other bird groups on Madagascar showed cryptic diversity by using multiple analyses.

Recent studies have shown that some species have been classified incorrectly, and this could be primarily due to two reasons. First, taxonomy based solely on phenotypic similarities can sometimes be misleading because the appearance of these species may be influenced by their ecological and environmental factors rather than their evolutionary histories. For example, the two largest avian groups on Madagascar, the families Vangidae and Bernieridae, include species that previously were placed in several different families but now with genetic data have been shown to comprise monophyletic radiations with ecological diversity (e.g., Cibois et al. 2001;

Reddy et al. 2012). In one study they find that within the vangas of Madagascar there was a pattern of diversification consistent with the ecological opportunity model (Reddy et al. 2012).

They also show foraging specializations in this group was due to a common ancestry and led to further speciation. Both lines of evidence point towards speciation in vangas being driven by adaptation into unoccupied and novel ecological niches (Reddy et al. 2012). In the next study,

Block found that there were three cryptic lineages within Bernieria. This genus represents the first documented instance of avian cryptic species maintained by selection against hybrids (Block

2012). Lastly, Bernieridae provides the first example of the effects of cryptic diversity (Block

2012). Second, current taxonomy might not accurately reflect true species diversity because some species might very well belong to multiple distinct evolutionary lineages. Previous studies of vertebrates show that current classifications are often misleading when organisms exhibit only slight morphological differences (e.g., Yoder et al. 2005); therefore the diversity of the extant avifauna may be obscured due to cryptic diversification. There was an assumption that the two populations belong to a single clade, but their findings reveal that the two populations are paraphyletic with respect to the western species. The results strongly suggest that to the contrary of the current recognition of a single species, Microebus rufus, there are at least two species of mouse lemurs in the eastern regions of Madagascar (Yoder et al. 2005).

Microcebus have changed dramatically in the past several years and there coloration patterns and other morphological features are not identifiable (Yoder et al. 2005). They show the importance of resolving species boundaries in nature and explain that identification of species requires a multidimensional approaches involving both morphological and genetic variation

(Yoder et al. 2005). The diversity of the extant avifauna may also be obscured due to cryptic diversification, when evolutionarily distinct species are hard to detect because they are morphologically similar to their closest relatives. Species are comprised of discrete groups of similar organisms and are the basic units of evolution and biodiversity. Species concepts are a way of defining species and are important because they allow us to propose hypotheses about the ontology of nature. Different species concepts generally imply a different ontology (Cracraft

2002). One needs a clear idea of the entities of nature so that one can count and describe patterns of diversity (Cracraft 2002). Understanding of species comes through training in ecology,

genetics, and evolution. What constitutes a species depends on the evolutionary history, morphology, and DNA analysis. I believe one of these species concepts relates to my research, and it is the phylogenetic species concepts. The phylogenetic species concept (PSC) applies to a group of organisms that all share one or more features pointing to a unique common ancestor, which in turn are not shared by members of other species (Hey 2009). Failure to resolve taxonomic relationships may result in the neglect of certain geographic regions and species for adequate conservation measures. For species that are experiencing exceptional loss of habitat, conservation planning depends on accurate species level taxonomy. This occurs when a single nominal species is being underestimated and conservation priorities are improperly assigned because endemic species may be overlooked (Lohman et al. 2010). Managing the genetic diversity is a key component to managing endangered species. Without genetic diversity a population could not evolve in response to changes in the environment. If there is a loss of this unique genetic diversity, we are actually losing global biodiversity.

Phylogeographic Studies on Madagascar

There have been many studies involving lemurs and lizards that show deep divergence across habitat types on Madagascar. In many cases, these divergences have dramatically increased the number of recognized species. For example, in lemurs, there was a sharp rise in species numbers that can be explained by the promotion of known populations or subspecies to full species status with analyses showing that these divergences are deeper than previously thought (Vences et al. 2009). There are species that might have been influenced by bioclimatic disparities. Madagascan populations adapt to dry versus humid conditions, and then they diverge

into separate species. Similarly, the total number of recognized species of amphibians in

Madagascar has increased from 133 species in 1991 to 244 species in 2009 (Vences et al. 2009).

Phylogeographic analyses also point to some strong biogeographic patterns across the landscapes of Madagascar that may have played a key role in driving the diversification of the species that inhabit them (Vences et al. 2009). In mouse lemurs (Microcebus), a primarily north- south split is found across multiple species (Vences et al. 2009). Researchers hypothesize that this separation is due to the isolation of the northern Malagasy mountains massifs from the central highlands by 100 km of largely low elevations (Vences et al. 2009). Researchers also find that several species of Malagasy reptiles displayed clear geographic clustering of distinct clades in the northern, central northern, eastern, and western regions of Madagascar (Boumans et al.

2007). These types of studies indicate that there are geographic patterns among some widespread species that might also be apparent in other groups. Unfortunately, although there has been much ongoing research on lemurs and lizards, there are only a few phylogeographic studies on birds in

Madagascar. Phylogeography is the study of spatial and temporal distribution in a gene sequence within a population. Phylogeography describes the geographic structure through the genetic signaling.

There have been only four recent studies of avian phylogeographic patterns among widespread species on Madagascar, and these uncover two very different results. The first study finds that there is no genetic structure among the populations of widespread species Dicrurus forficatus on Madagascar (Fuches et al. 2013). Dicrurus forticatus is found throughout the island and was likely a recent colonizer. Therefore, its lack of genetic differentiation can be explained

by the limited time for the accumulation of genetic differences between eastern and western populations within nuclear DNA, which has slower mutation rates and longer coalescent times in comparsion to mitochondrial DNA (Figure 1). Furthermore, ecological niche modeling using bioclimatic data shows that climatic oscillations over the past 0.120 million years lead to repeated expansion and contraction of populations but little differentiation (Fuchs et al. 2013).

The second study on a small forest bird, Bernieria madagascariensis, shows high genetic divergence across morphologically identical populations (Figure 2). This study shows that

Bernieria madagascariensis is comprised of three distinct parapatric clades, with genetic divergences among them older than those found in many avian species complexes (Block 2012).

Block also shows that the Bernieria clades are cryptic species that diverged allopatrically and then expanded their range into parapatry. This study indicates that some species or evolutionary units may have complex histories that can be important for understanding biogeographic patterns on Madagascar (Block 2012).

The third study found that mtDNA sequencing of Xanthomixis zosterops, a forest- dependent Malagasy passerine belonging to the endemic family Bernieridae, shows multiple sympatric, cryptic, and deeply divergent clades split geographically by elevation (Figure 3).

Xanthomixis zosterops provides the first example of this biogeographic pattern in birds (Block et al. 2015). The phylogeographic patterns within this species give us insight into the potential barriers to gene flow such as rivers and altitudinal segregation on Madagascar (Block et al.

2015).

The fourth study examines the colonization history and phylogeography of the brush- warblers (Nesillas) that are endemic to islands of Madagascar, Comoros, and Aldabra Atoll

(Fuchs et al. 2016). The diversification of Nesillas was shaped by long distance dispersal events and inter-island colonization (Fuchs et al. 2016). These species show little to no genetic differentiation within Madagascar (Figure 4). Given the variable results in these studies, predicting the geographic patterns among species on an isolated island like Madagascar is difficult and needs to be explored extensively with additional studies on other endemic species.

Newtonia

My study focuses on using molecular and biogeography to better understand patterns of diversity across the genus Newtonia, a group of small insect-eating songbirds that generally fly only short distances. There are four species in this endemic genus found in forest habitats across the island. N. amphichroa is strictly forest dwelling and occurs in the eastern humid forest

(widespread, eastern humid forest); N. archboldi occurs in the southern spiny forest on

Madagascar (restricted, southern spiny forest); N. brunneicauda is strictly forest dwelling and found in multiple habitat types including humid, dry, humid-subhumid, and dry spiny forest

(widespread, throughout Madagascar); and N. fanovanae is distributed in the dry humid forest

(restricted, southeastern Madagascar). Newtonia fanovanae is only known from one specimen.

Traditionally the genus Newtonia was thought to belong to the warbler family, Sylviidae; however, phylogenetic studies find that this genus belongs to the Vangidae family (Yamagishi et al. 1999; Reddy et al. 2012). One study reconstructed a complete species-level phylogeny of the

22 members of the Vangidae family on Madagascar (Jonsson et al. 2012). They included only a

few representatives of each species of Newtonia in their phylogenetic analysis: four individuals of N. amphichroa, three individuals of N. brunneicauda, one individual of N. archboldi, and one individual of N. fanovanae (Jonsson et al. 2012). Although their combined analysis shows that each species is monophyletic, a closer look at each gene tree showed contradictory results. The gene tree from the nuclear region of GAPDH indicated non-monophyly of N. brunneicauda and

N. amphichroa. In some genes, these species were paraphyletic – N. brunneicauda comprised the majority of a clade, but few individuals of N. amphichroa are also present. This indicates that the relationships between the species are not consistent. This is important because unclear species definitions may contribute to the common misidentification of these species.

A major goal of this study is to investigate if there is cryptic differentiation within two of the widespread Newtonia species, N. amphichroa and N. brunneicauda. If there is differentiation within these species, I will examine if distinct lineages are delimited by geographic region or habitat type. If there is no differentiation, this indicates that the species are moving around more than suspected and gene flow is occurring across all the populations. These species are typically diagnosed by plumage; for example the N. amphichroa has darker plumage color and differing habitat preferences (Safford et al. 2013). Newtonia brunneicauda is also known as Common

Newtonia, is greyish-brown in color, and is found in multiple habitats (Safford et al. 2013).

Newtonia archboldi has a rufous-brown markering on its forehead (Safford et al. 2013). The

Field Museum specimens were hard to identify, but genetic data can be useful in the identification process. Newtonia species are difficult to differentiate based on morphology because they look very similar across all four species (Goodman and Benstead 2003; Safford and

Hawkins 2013). This included observing their upper-parts and under-parts and checking for their coloration (Figure 5). Based on the descriptions from the guidebooks and my observations of the museum specimens, these species are similar in size, length, and plumage patterns (Table 1;

Figure 5). These features (plumage patterns, elevation, and song) are often used as a tool for identifying birds. However, field identification can sometimes be difficult if the species are similar in appearance and therefore can lead to misidentifications in the field (Steve Goodman, personal communication). Recent studies on Malagasy birds have uncovered cryptic diversification using genetic tools, and the extent cryptic differentiation within Newtonia species is currently unknown. In this study, I also use genetic tools to examine cryptic diversification within Newtonia.

Objectives

The objectives of this study are to determine the following:

1. What are the phylogenetic relationships among the species of Newtonia? I conducted a

phylogenetic analysis using multiple molecular markers in order to examine how the

Newtonia species evolved.

2. Is there phylogeographic structure to indicate cryptic differentiation or recent

speciation within Newtonia species?

3. If there is evidence of differentiation, are lineages diverging due to habitat type or

geography?

Here I used genetic sequences to determine if there was any geographic structure within

the two widespread species of Newtonia.

4. Are there issues of misidentified specimens that I can overcome with

additional data? I examined three alternative hypotheses of why some individuals were

misplaced in the wrong clade in the phylogenetic analysis. I tested if misidentification

was due to (a) errors in the lab, (b) hybridization, or (c) misidentifications in the field.

Figure 1.Figure 2b from Fuchs et al. (2013) showing no genetic structure among the populations of widespread species, Dicrurus forficatus. Recent colonizer therefore lack genetic differentiation.

Figure 2. Figure 3.1 from Block (2012) showing Bernieria madagascariensis has high genetic divergence across morphologically identical populations. This species is comprised of three distinct parapatric clades with deep divergences.

Figure 3. Figure 1a from Block et al. (2015) showing Xanthomixis zosterops comprises multiple sympatric, cryptic, and deeply divergent clades. This indicates that rivers and elevation are potential barriers to gene flow.

Figure 4. Figure 2 from Fuchs et al. (2016) shows that in the genus Nesillas diversification was shaped by long distance dispersal events and interisland colonization with little genetic differentiation within Madagascar.

1. 2. 3. 4. 5. 6.

A. B. C.

Figure 5. Images of the three Newtonia species. Images taken by Lynika Strozier at the Field Museum of Natural History. A) The under-parts of N. brunneicauda (1, 2), N. amphichroa (3, 4), and N. archboldi (5, 6). B) The side view of the three species. C) The view of the upper-parts among the three species. The numbers bolded in red were the individuals that were identified as the same species in the field but genetic data showed they belong to a different clade.

Description N. amphichroa N. brunneicauda N. archboldi Top of Head Dark olive brown Greyish brown Greyish brown Lore’s Dark olive brown Lighter greyish Dark grey brown Tail Olive brown Greyish brown Grey brown Eye color Golden Yellow Pale dull, pale cream Pale Whitish yellow or golden yellow Weight Unsexed 9-18g Unsexed 7-14.5g Unsexed 7.2,8.3g Table 1. A physical description of the three species of Newtonia: N. amphichroa, N. brunneicauda and N. archboldi. Data was obtained from Safford and Hawkins (2013).

CHAPTER TWO

METHODS

USING MOLECULAR TOOLS TO DETERMINE

DIVERSIFICATION PATTERNS AMOUNG GENUS NEWTONIA

I obtained tissue samples from 150 individuals of the three species of Newtonia from across their range, including 45 tissue samples of N. amphichroa, 96 tissue samples of N. brunneicauda, and 9 tissue samples of N. archboldi. Newtonia fanovanae tissue was not available for this study, and this species thus eliminated from study (Figure 6; Table 2). For out- groups, I included samples of Vanga curvirostris, Falculea palliata, and Leptopterus chabert from genbank. All tissue samples were obtained from the collections of the Field Museum of

Natural History and the University of Antananarivo, Antananarivo, Madagascar collections.

Sequencing

Genomic DNA was isolated from tissue samples by using the Qiagen Dneasy Blood &

Tissue Extraction Tissue Kit (Qiagen Valencia,CA). I amplified DNA using Polymerase Chain

Reaction (PCR) for two-mitochondrial loci and three nuclear introns. The sequences for this dataset were: 351 bp of NADH dehydrogenase subunit 3 (ND3; primers ND3-L10751 and ND3-

H11151; Pastorini et al. 2002), 1007 bp of Cytochrome b (CYTB; primers CB-L14851and CB-

H15563; Cabanne et al. 2008), 359 bp of glyceraldehyde -3-phosphate dehydrogenase intron 11

(GAPDH; primers GapdL and GapdH; Fuchs et al. 2013), 557 bp of B-fibrinogen intron 5 (FIB5; primers Fib5F and Fib6R; Kimball et al. 2009), and 480 bp of Muscle Specific Kinase intron 3

18 19

(MUSK; Musk-13F and Musk-13R; Kimball et al. 2009). All primer sequences are listed in

Table 3. I ran standard PCR reactions consisted of per reaction 2.5µL10X PCR buffer (15mM),

0.5µL MgCl2 (25mM), 0.125µL Taq DNA Polymerase (5 units/µl), 17.375µL H2O (Nuclease

Free), 0.5µL (10mM) dNTPs, 1µL of (10uM) Primer L (forward) and (10uM) of Primer H

(reverse), and 2µL of template DNA.

PCR cycling conditions started with an initial 3-minute denaturing step at 94oC followed by 35 amplification cycles of 30 seconds denaturing at 94o C, annealing at 52oC for 30 seconds, extension at 72oC for 45 seconds and the final extension step at 72oC for 10 minutes. PCR products were run on 1% high melting point agarose gels to verify whether amplification was successful and of sufficient quantity for sequencing. The amplification products were cleaned up using Agencourt AMPure XP–PCR purification, Beckman Coulter (Indianapolis, Indiana).

The cycle sequence step included per reaction 1µL(5uM) of primer L and 1µL(5uM) primer H, 3µL of (5X) sequence buffer, 1µL Big Dye terminator v.3.1, and 5µL DNA template.

After cycle sequencing, I performed a purification on the 10µL sequencing reaction using ethanol/EDTA solution and 70% ethanol. Samples were dried down for 10 minutes in the 65oC degree oven. Next, samples were re-suspended in 10µL of Hi-Di formamide for the 3730XL sequencer. The sequencing reactions were run on ABI 3730 48-capillary electrophoresis DNA analyzer sequencer (Applied

Biosystems, Foster City, CA, USA) in the Pritzker Laboratory at the Field Museum of Natural

History. The sequences were aligned and edited using Geneious v7.1.5 (Kearse et al. 2012). I was able sequence the vast majority of amplicons for most individuals but was unable to amplify all genes for all samples due to various reasons including quality of DNA or specificity of primers (Table 3).

Phylogenetic Analysis

To test for monophyly among N. amphichroa, N. brunneicauda, and N. archboldi, I conducted a phylogenetic analysis by using the standard methods of maximum likelihood (ML).

PartitionFinder v. 1.1.1 (Lanfear et al. 2012) was used to infer the best fitting partitioning scheme. In RaxML v.7.0.4 (Stamatakis 2014), I conducted a rapid bootstrap of 1000 replicates using GTRCAT model and subsequent search for the best ML tree. I used the best partition scheme to search for the best ML tree and to calculate the bootstrap in RAxML. I used FigTree v7.0.4 (Rambaut and Drummond 2009) to view RAxML trees and Abobe illustrator CC 2014 to prepare these for publication.

Biogeographic Analysis

GenGis (Donovan et al. 2013) makes use of a digital map data, sample site information, sequence data, and one or more phylogenetic trees to reconstruct the phylogeographical patterns and assess whether there was an influence of geography on the phylogenetic results. In this study, I was interested in determining whether lineages diverged caused by habitat type or geography. Files required for this analysis were a digital map, location file, sequence file, and the tree file. The locations files indicate the geographic coordinates of each of the sample site.

Additional information from sequences that were collected at each sample site can be specified

21 in the sequence files. The tree files are associated with the node names, which have a unique identifier such as sample sites or sequences, and the map are used for visualizing the 3D terrain.

With program GenGis, I was able to look at the phylogenetic data overlaid across a geographic map.

Population Genetics

In order to determine if there was any phylogeographic structure to indicate cryptic differentiation or recent expansion within the Newtonia species, I first measured the amount of

DNA sequence variation within the three Newtonia species by examining haplotype diversity, nucleotide diversity, number of polymorphic segregation sites (S), Fu’s F statistic, and Tajima’s

D. Haplotype diversity calculates how many different haplotypes and their corresponding frequencies in that population/species. Nucleotide diversity was used to measure the degree of polymorphism within a population. This means that the higher the Hd value, the higher the chance to find different haplotypes in that population when you sequence the gene region. To test for any recent demographic changes, I calculated the Fu’s Fs (Fu 1997). To test for neutrality and demographic changes within species by loci, I used DNASP to calculate Tajima D (Tajima

1989).I used DNASP 5.0 (Librado and Rozas 2009), a software package for analyzing the nucleotide polymorphism from aligned DNA, to determine the amount of variation across my data. This analysis was done by first separating the sequences by loci, next separating by species in Geneious I exported the sequence files as a fasta files in Geneious. I then imported the files into the DNASP program. I selected the DNA polymorphism data and selected run and after the analysis was performed the program gave an output file that contained the population genetics statistics for each species by loci.

Haplotype Network Analysis

I next created a haplotype network using median joining networks of my CYTB sequence data. The haplotype network was performed using software called PopART

(http://popart.otago.ac.nz), which is a method for visualizing relationships between individual genotypes at the population level by using sequencing data. The files were imported as a NEXUS format including the alignment of the haplotype sequence files and a Newick format tree of the sequences. The nodes were colored coded by geographic area. I used two color schemes: first by broad-areas such as north, central, and south regions; then by 9 different biogeographic zones across Madagascar that included 1. South sub-arid, 2. Western dry forest, 3.Northwest dry forest,

4.Northeast humid forest, 5.Eastern humid forest, 6.Southeast humid forest, 7.Southeast montane regions, 8.Central montane region and 9.Northern montane regions.

Species Delimitation Analysis

In order to determine the presence of distinct genetic groups, I conducted a Bayesian species delimitation analysis using BPP V 3.2 (Yang 2015). BPP (for Bayesian Phylogenetics and Phylogeography) is a Bayesian Markov Chain Monte Carlo program for analyzing DNA sequence alignments under the multispecies coalescent model (Yang 2015). BPP provides a statistical method for testing the alternative models that may describe the number of distinct groups in a clade. The goal of the BPP analysis was to determine if the two distinct clades of N. amphichroa are divergent enough to be called different species. I used four genes (excluding

GAPDH due to poor coverage) and set up four alternative hypotheses to test: 1- all Newtonia should be considered one species; 2- Newtonia should be considered as two distinct species (N. archboldi and N. brunneicauda [including N. amphichroa]); 3- Newtonia should be considered

23 as three species (same as current taxonomy: N. archboldi, N. brunneicauda, N. amphichroa); 4-

Newtonia should be considered as four species (splitting N. amphichroa as two species). For population size parameters, I assigned the gamma prior G (2,1000), with mean 2/2000 = 0.001. I ran each analysis at least twice to confirm consistency between runs. Each run was for 100,000 samples, sampling frequency was 5 and the burn in was set to 20,000.

Checking for Misidentified Specimens

To examine if the misidentified samples are from particular geographic regions or elevation, I took a closer look at the elevation ranges of the individuals that were misidentified.

In my initial genetic analysis there were individuals that were identified as N. amphichroa grouped with N. brunneicauda and individuals that were identified as in N. brunneicauda grouped with N. amphichroa. I extracted the elevation data for all N. amphichroa and N. brunneicauda samples and plotted them to see if there was a large zone of overlap in their range.

Furthermore, I visually surveyed specimens of N. amphichroa, N. brunneicauda, and N. archboldi at the Field Museum of Natural history in order to view their similarities and differences in their physical appearance to get a better understanding of just how difficult it is to identify these individuals in the field.

Molecular Clock

In order to estimate the time of divergence within the Newtonia species, I performed a calibrated Bayesian phylogenetic analysis using the mitochondrial sequences (ND3 and CYTB) for 24 individuals using BEAST 2.4.4 (Bouckaert et al. 2014). The individuals were selected from their major clades identified in my full Maximum Likelihood analysis. To calibrate the divergences, I use a divergence rate of CYTB for Passeriformes of 2.07% MYA (± 0.20) per

24 million years as a reference rate (Weir and Schulter 2008). I ran BEAST twice to ensure convergence of runs. Each analysis was run for 150 million generations and convergence across taxa was confirmed in Tracer v1.6 (Rambaut et al. 2014). I used Tree Annotator (Bouckaert et al.

2014) to generate the maximum clade credibility tree.

Table 2.Sample accession number, locality, elevation, latitude, and longitude information for each specimen of Newtonia used in this study. * UADAB = University of Antananarivo Department of Animal Biology; FMNH = Field Museum of Natural History.

Accesssion Elevation Species Institution* No. Locality (m) Latitude Longitude Province Mahajanga, Fivondronana Mandritsara, 12 Km SSE Marotandrano, Marotandrano, amphichroa UADAB MJR0264 Anjiambolo 950 16.28 48.80 Province Mahajanga, Fivondronana Mandritsara, 12 Km SSE Marotandrano, Marotandrano, amphichroa UADAB MJR0270 Anjiambolo 950 16.28 48.80 Province Antananarivo, Fivondronana Anjozorobe, 6,7 Km SE Anjozorobe, amphichroa UADAB MJR0358 Antsahabe 1250 18.42 47.94 Province Antananarivo, Fivondronana Anjozorobe, 6,7 Km SE Anjozorobe, amphichroa UADAB MJR0362 Antsahabe 1250 18.42 47.94 Province Antananarivo, Fivondronana Anjozorobe, 6,7 Km SE Anjozorobe, amphichroa UADAB MJR0366 Antsahabe 1250 18.42 47.94 Province Antananarivo, Fivondronana Anjozorobe, 6,7 Km SE Anjozorobe, amphichroa UADAB MJR0367 Antsahabe 1250 18.42 47.94 Province Antananarivo, Fivondronana Anjozorobe, 8 Km SE Anjozorobe, amphichroa UADAB MJR0375 Andasin'i Saotra 1300 18.43 47.95 Province Antananarivo, Fivondronana Anjozorobe, 12,5 Km SSE Anjozorobe, amphichroa UADAB MJR0394 Ambohimanga 1325 18.47 47.96

26 Province Antananarivo, Fivondronana Anjozorobe, 5,5 Km E Alakamisy, amphichroa UADAB MJR0411 Iaban'Ikoto 1280 18.52 47.97 Province Antananarivo, Fivondronana Anjozorobe, 5,5 Km E Alakamisy, amphichroa UADAB MJR0416 Iaban'Ikoto 1280 18.52 47.97 Province Antananarivo, Fivondronana Anjozorobe, 5,5 Km E Alakamisy, amphichroa UADAB MJR0418 Iaban'Ikoto 1280 18.52 47.97 Province Antananarivo, Fivondronana Anjozorobe, 17 Km NE Anjozorobe, amphichroa UADAB MJR0434 Anorana 1250 18.31 48.02 Province Antananarivo, Fivondronana Anjozorobe, 12 Km SE Mangamila, amphichroa UADAB MJR0455 Andasin'i Tovo 1270 18.64 47.94 Province Antananarivo, Fivondronana Ankazobe, Ambohitantely, Jardin amphichroa UADAB MJR0526 Botanique 1540 18.17 47.28 Province Toamasina, Région Alaotra-Mangoro, Commune rurale Lakato, Forêt de Sahambaky, le long de la Rivière Sahatandra, 31,5 km NNO du village de Lakato, amphichroa UADAB MJR0562B Sahambaky 980 19.07 48.34 Province Toamasina, Région Alaotra-Mangoro, Commune rurale Lakato, Forêt de Sahambaky, le long de la Rivière Sahatandra, 31,5 km NNO du village de Lakato, amphichroa UADAB MJR0563A Sahambaky 980 19.07 48.34 Province Fianarantsoa, Fivondronana Befotaka, 11,5 Km SW Befotaka village, Andranomigodo, amphichroa UADAB MJR0676 Midongy 1055 23.89 46.90 Province de Toamasina, Forêt de Maromiza, 8.3 km SE du village d'Andasibe (Périnet), amphichroa UADAB MJR0688 Maromiza 980 18.98 48.46

27 Province Toamasina, Région Alaotra-Mangoro, Commune rurale Lakato, Forêt de Sahambaky, le long de la Rivière Sahatandra, 31,5 km NNO du village de Lakato, amphichroa UADAB MJR0707 Sahambaky 980 19.07 48.34 Province de Toamasina, Forêt de Maromiza, 8.3 km SE du village d'Andasibe (Périnet), amphichroa UADAB MJR0716 Maromiza 980 18.98 48.46 Forêt Toamasina, Forêt d'Analamay, Ambatovy, 9 km NE du village d'Ambohimanarivo, amphichroa UADAB MJR0751 Analamay 1106 18.80 48.32 Forêt Toamasina, Forêt d'Analamay, Ambatovy, 9 km NE du village d'Ambohimanarivo, amphichroa UADAB MJR0752 Analamay 1106 18.80 48.32 Forêt Toamasina, Forêt d'Analamay, Ambatovy, 9 km NE du village d'Ambohimanarivo, amphichroa UADAB MJR0753 Analamay 1106 18.80 48.32 Province de Toamasina, Forêt d'Analamay, Ambatovy, 10.5 km NE du village d'Ambohimanarivo, amphichroa UADAB MJR0770 Analamay 1105 18.79 48.33 Province de Toamasina, Forêt d'Analamay, Ambatovy, 10.5 km NE du village d'Ambohimanarivo, amphichroa UADAB MJR0771 Analamay 1105 18.79 48.33 Alaotra Mangoro, Moramanga, 12.5 km NE du village d'Ambohimanarivo, amphichroa UADAB MJR0820 Analamay 1006 18.81 48.36 Alaotra Mangoro, Moramanga, 12.5 km NE du village d'Ambohimanarivo, amphichroa UADAB MJR0821 Analamay 1006 18.81 48.36

28 Province Toamasina, Forêt d'Anivonimaro Ambalafary, 14.5 km SW of Andasibe/Périnet, amphichroa UADAB MJR0894 Lakato 997 19.04 48.35 Province de Mahajanga, Forêt de Bemanevika, amphichroa UADAB MJR01061 Bemanevika 1586 14.35 48.58 Province de Mahajanga, Région de Sofia, Forêt d'Ambanin'Andrakanala, amphichroa UADAB MJR01079 Bemanevika 1517 14.38 48.59 Province de Mahajanga, Région de Sofia, Forêt d'Ambanin'Andrakanala, amphichroa UADAB MJR01080 Bemanevika 1517 14.38 48.59 Foret de Marovony, 19 km amphichroa FMNH 352930 NNE Manantenina 50 -24.00 47.37 Province Antananarivo, Fivondronana Anjozorobe, 12 Km SE Mangamila, amphichroa UADAB MJR0464 Andasin'i Tovo 1270 18.64 47.94 Ambalamanenjana- Ambatomboay Trail, 45 amphichroa FMNH 363856 km S Ambalavao 720 -22.00 47.02 Andringitra, 38 km S Ambalavao, E of amphichroa FMNH 363858 Volotsangana River 1625 -22.00 46.97 Andranomay, 2 km NNE, amphichroa FMNH 384721 13 km SE Anjozorobe 1300 -18.47 43.95 RNI d'Andohahela, 15.0 amphichroa FMNH 384791 km NW Eminiminy 1500 -24.57 46.72 Foret de Betaolana, 8.5 km NW Ambodiangezoka, amphichroa FMNH 393364 Ambolokopatrika Ri 875 -14.54 49.44 Foret de Betaolana, 8.5 km NW Ambodiangezoka, amphichroa FMNH 393365 Ambolokopatrika Ri 875 -14.54 49.44 Foret de Betaolana, 11.0 amphichroa FMNH 393366 km NW Ambodiangezoka 1200 -14.54 49.43 Foret de Betaolana, 11.0 amphichroa FMNH 393367 km NW Ambodiangezoka 1200 -14.54 49.43 Foret de Betaolana, 11.0 amphichroa FMNH 393368 km NW Ambodiangezoka 1200 -14.54 49.43 Foret de Betaolana, 11.0 amphichroa FMNH 393369 km NW Ambodiangezoka 1200 -14.54 49.43 Foret de Betaolana, 11.0 amphichroa FMNH 393370 km NW Ambodiangezoka 1200 -14.54 49.43 Foret de Betaolana, 11.0 amphichroa FMNH 393371 km NW Ambodiangezoka 1200 -14.54 49.43

29 Foret de Betaolana, 11.0 amphichroa FMNH 393372 km NW Ambodiangezoka 1200 -14.54 49.43 Anjanaharibe-Sud, W slope, 13.5 km SW amphichroa FMNH 393373 Befingotra 1200 -14.75 49.42 Anjanaharibe-Sud, W slope, 13.5 km SW amphichroa FMNH 393374 Befingotra 1200 -14.75 49.42 Anjanaharibe-Sud, W slope, 13.5 km SW amphichroa FMNH 393377 Befingotra 1200 -14.75 49.42 Anjanaharibe-Sud, W slope, 13.0 km SW amphichroa FMNH 393379 Befingotra 1600 -14.75 49.42 Anjanaharibe-Sud, W slope, 13.0 km SW amphichroa FMNH 393380 Befingotra 1600 -14.75 49.42 Manambolo Forest, 19.5 amphichroa FMNH 393382 km SE Sandrisoa 1600 -22.16 47.04 Manambolo Forest, 19.5 amphichroa FMNH 393383 km SE Sandrisoa 1600 -22.16 47.04 Manambolo Forest, 19.5 amphichroa FMNH 393384 km SE Sandrisoa 1600 -22.16 47.04 Manambolo Forest, 19.5 amphichroa FMNH 393387 km SE Sandrisoa 1600 -22.16 47.04 Manambolo Forest, along Andohabatotany River, amphichroa FMNH 393389 17.5 km SE Sandris 1300 -22.15 47.02 Anjanaharibe-Sud, W slope, 13.0 km SW amphichroa FMNH 395989 Befingotra 1600 -14.75 49.42 Anjanaharibe-Sud, W slope, 13.0 km SW amphichroa FMNH 395990 Befingotra 1600 -14.75 49.42 Foret de Mahatsinjo, Andasivodihazo, 10 km amphichroa FMNH 396200 SE Tsinjoarivo 1550 -19.67 47.77 Foret de Mahatsinjo, Andasivodihazo, 10 km amphichroa FMNH 396201 SE Tsinjoarivo 1550 -19.67 47.77 Foret de Mahatsinjo, Andasivodihazo, 10 km amphichroa FMNH 396203 SE Tsinjoarivo 1550 -19.67 47.77 RS de Manongarivo, 14.5 amphichroa FMNH 396207 km SW Antanambao 1240 -14.00 48.43 RS de Manongarivo, 14.5 amphichroa FMNH 396208 km SW Antanambao 1240 -14.00 48.43 RS de Manongarivo, 14.5 amphichroa FMNH 396209 km SW Antanambao 1240 -14.00 48.43 Foret d'Andranomay, amphichroa FMNH 429706 Anjozorobe - -18.48 47.96

30 Foret d'Andranomay, amphichroa FMNH 429707 Anjozorobe - -18.48 47.96 Parc National de Marojejy, 13.0 km SE amphichroa FMNH 431265 Doany 1175 -14.43 49.80 Parc National de Marojejy, 13.0 km SE amphichroa FMNH 431266 Doany 1175 -14.43 49.80 Province Antananarivo, Fivondronana Anjozorobe, amphichroa UADAB MJR0415 5,5 Km E Alakamisy 1280 18.52 47.97 Province Mahajanga, Fivondronana Analalava, 7,5 Km NE Ambarijeby brunneicauda UADAB MJR0312 village, Belambo 150 14.89 47.73 Province Toliara, Fivondronana Ampanihy, 7,5 Km NE Ambarijeby brunneicauda UADAB MJR0317 village, Tongaenoro 120 24.74 44.03 Province Toliara, Fivondronana Ampanihy, 7,5 Km NE Ambarijeby brunneicauda UADAB MJR0318 village, Tongaenoro 120 24.74 44.03 Province de Toliara, Forêt de Vohondava, 10 km SSO de Tranomaro, le long de la rivière brunneicauda UADAB MJR0332 Vohondava, Vohondava 225 24.69 46.45 Province de Toliara, Forêt de Mahavelo, le long de la Rivière Santoria, 4 km NNE du village d'Ifotaka, brunneicauda UADAB MJR0344 Camp 5, Mahavelo 100 24.77 46.15 Province Antananarivo, Fivondronana Anjozorobe, 6,7 Km SE Anjozorobe, brunneicauda UADAB MJR0370 Antsahabe 1250 18.42 47.94 Province Antananarivo, Fivondronana Anjozorobe, 12 Km SE Mangamila, brunneicauda UADAB MJR0469 Andasin'i Tovo 1270 18.64 47.94 Province Toliara, Fivondronana Belo sur Tsiribihina, 5,7 Km NNE Masoarivo village, brunneicauda UADAB MJR0487 Ambalimby, Masoarivo 1 110 19.62 44.77 Province Toliara, Fivondronana Belo sur Tsiribihina, 5,7 Km NNE Masoarivo village, brunneicauda UADAB MJR0497 Ambalimby, Masoarivo 1 110 19.62 44.77

31 Province Toliara, Fivondronana Belo sur Tsiribihina, 5,0 Km NNE Masoarivo village, brunneicauda UADAB MJR0499 Ambalimby, Masoarivo 2 120 19.61 44.74 Province Toliara, Fivondronana Belo sur Tsiribihina, 5,0 Km NNE Masoarivo village, brunneicauda UADAB MJR0501 Ambalimby, Masoarivo 2 120 19.61 44.74 Province Toliara, Fivondronana Belo sur Tsiribihina, 5,0 Km NNE Masoarivo village, brunneicauda UADAB MJR0502 Ambalimby, Masoarivo 2 120 19.61 44.74 Province Mahajanga, Parc National d'Ankarafantsika, 10 km NNE du village d'Andranofasika, PN brunneicauda UADAB MJR0524 Ankarafantsika 150 16.30 46.93 Province Antananarivo, Fivondronana Ankazobe, Ambohitantely, Jardin brunneicauda UADAB MJR0528 Botanique 1540 18.17 47.28 Province Toliara, Fivondronana Morondava, 11,8 Km SE Belo sur mer, Kirindy Mite, brunneicauda UADAB MJR0543 Amponiloaky 40 20.79 44.10 Province Toliara, Fivondronana Morondava, 11,8 Km SE Belo sur mer, Kirindy Mite, brunneicauda UADAB MJR0544 Amponiloaky 40 20.79 44.10 Province Toamasina, Région Alaotra-Mangoro, Commune rurale Lakato, Forêt de Sahambaky, le long de la Rivière Sahatandra, 31,5 km NNO du village de Lakato, brunneicauda UADAB MJR0561 Sahambaky 980 19.07 48.34 Province Toamasina, Région Alaotra-Mangoro, Commune rurale Lakato, Forêt de Sahambaky, le long de la Rivière Sahatandra, 31,5 km NNO du village de Lakato, brunneicauda UADAB MJR0580 Sahambaky 980 19.07 48.34

32 Province Antananarivo, Réserve Spéciale d'Ambohitantely, Forêt d'Ambohitantely-Jardin brunneicauda UADAB MJR0743 Botanique, Ambohitantely 1540 18.23 47.29 Province Toamasina, Forêt d'Analamay, Ambatovy, 10.5 km NE du village d'Ambohimanarivo, brunneicauda UADAB MJR0774 Analamay 1105 18.79 48.33 Province Toamasina, Région Alaotra Mangoro, Moramanga, Forêt d'Analamay, 10 km E du village d'Ambohimanarivo, brunneicauda UADAB MJR0798 Analamay 1060 18.81 48.34 Province Toliara, Réserve Spéciale d'Andranomena, Forêt de Tamaro, 5 km E du village Andranomena, brunneicauda UADAB MJR0872 Andranomena 18 20.18 44.54 Province Toliara, Parc National de Tsimanampetsotsa, 6.5 km au NE du village d'Efoetse, Analasoa, brunneicauda UADAB MJR0908 Tsimanampetsotsa 19 24.03 43.74 Province Toliara, Parc National de Tsimanampetsotsa, 6.5 km au NE du village d'Efoetse, Analasoa, brunneicauda UADAB MJR0909 Tsimanampetsotsa 19 24.03 43.74 Province Mahajanga. Forêt d'Ambinda Nord, District de Maintirano, brunneicauda UADAB MJR0936 Région Melaky, Beanka 168 17.94 44.47 Province Mahajanga. Forêt d'Ambinda Nord, District de Maintirano, brunneicauda UADAB MJR0938 Région Melaky, Beanka 168 17.94 44.47 Province Toliara, Forêt de Salary, 6 km NNE du village de Salary Nord, brunneicauda UADAB MJR0951 Salary-Bekodoy 20 22.51 43.30 Province Toliara, Forêt de Salary, 6 km NNE du village de Salary Nord, brunneicauda UADAB MJR0963 Salary-Bekodoy 20 22.51 43.30

33 Province Toliara, Parc National de Tsimanampetsotsa, 6.5 km au NE du village d'Efoetse, Tsimanampetsotsa, brunneicauda UADAB MJR01038 Analasoa 19 24.03 43.74 Province de Mahajanga, Forêt de Bemanevika, brunneicauda UADAB MJR01068 Bemanevika 1586 14.35 48.58 Foret d'Analalava, 7 km N brunneicauda FMNH 345885 Manantenina 40 -24.00 47.28 Foret d'Analalava, 7 km N brunneicauda FMNH 345886 Manantenina 40 -24.00 47.28 Marosohy Forest, along Enakara-Antseva Trail, brunneicauda FMNH 345887 15.5 km WNW Ranoma 750 -24.00 46.80 Foret Cascade, 12.2 km NE Manambaro, 8.3 km brunneicauda FMNH 352931 NW Fort Dauphin 200 -24.00 46.93 Foret Cascade, 12.2 km NE Manambaro, 8.3 km brunneicauda FMNH 352932 NW Fort Dauphin 200 -24.00 46.93 Foret d'Ankapoky, 13 km NE Amboasary Sud, 21 brunneicauda FMNH 352933 km NW Ranopiso 75 -24.00 46.52 Foret d'Ankapoky, 13 km NE Amboasary Sud, 21 brunneicauda FMNH 352938 km NW Ranopiso 75 -24.00 46.52 Foret d'Itapera, 19.5 km brunneicauda FMNH 352941 NE Fort Dauphin - -24.00 47.12 Foret de Marovony, 19 km brunneicauda FMNH 352943 NNE Manantenina 50 -24.00 47.37 Foret de Betaolana, 11.0 brunneicauda FMNH 393375 km NW Ambodiangezoka 1200 -14.54 49.43 Foret de Betaolana, 11.0 brunneicauda FMNH 393376 km NW Ambodiangezoka 1200 -14.54 49.43 Anjanaharibe-Sud, W slope, 13.0 km SW brunneicauda FMNH 393378 Befingotra 1600 -14.75 49.42 Anjanaharibe-Sud, W slope, 13.0 km SW brunneicauda FMNH 393381 Befingotra 1600 -14.75 49.42 Station Forestiere de brunneicauda FMNH 396064 Tampolo 10 -17.28 49.42 Foret de Mahatsinjo, Andasivodihazo, 10 km brunneicauda FMNH 396202 SE Tsinjoarivo 1550 -19.67 47.77 Foret de Mahatsinjo, Andasivodihazo, 10 km brunneicauda FMNH 396204 SE Tsinjoarivo - -19.67 47.77

34 RS de Manongarivo, 17.3 brunneicauda FMNH 396205 SW Antanambao 1600 -14.02 48.42 Parc National de l'Isalo, 3.8 km NW Ranohira, brunneicauda FMNH 396206 Namaza River 800 -22.54 45.38 Parc Nat d'Andringitra, Foret de Ravaro, 12.5 km brunneicauda FMNH 427345 SW Antanifotsy 1500 -22.20 46.83 Parc Nat d'Andringitra, Foret de Ravaro, 12.5 km brunneicauda FMNH 427346 SW Antanifotsy 1500 -22.20 46.83 Parc Nat d'Andringitra, Foret de Ravaro, 12.5 km brunneicauda FMNH 427347 SW Antanifotsy 1500 -22.20 46.83 PN Tsimanampetsotsa, 6.5 brunneicauda FMNH 434458 km NE Efoetsa 50 -24.05 43.75 PN Tsimanampetsotsa, 6.5 brunneicauda FMNH 434459 km NE Efoetsa 50 -24.05 43.75 PN Tsimanampetsotsa, brunneicauda FMNH 434460 21.5 km NE Efoetsa 100 -24.00 43.88 PN Tsimanampetsotsa, 6.5 brunneicauda FMNH 434461 km NE Efoetsa 50 -24.05 43.75 PN Tsimanampetsotsa, 6.5 brunneicauda FMNH 434462 km NE Efoetsa 50 -24.05 43.75 Foret des Mikea, 9.5 km brunneicauda FMNH 436435 W Ankiloaka 80 -22.77 43.52 Foret des Mikea, 9.5 km brunneicauda FMNH 436436 W Ankiloaka 80 -22.77 43.52 Foret des Mikea, 9.5 km brunneicauda FMNH 436437 W Ankiloaka 80 -22.77 43.52 Foret des Mikea, 9.5 km brunneicauda FMNH 436439 W Ankiloaka 80 -22.77 43.52 Foret des Mikea, 7.5 km brunneicauda FMNH 436440 NE Tsifota 60 -22.80 43.43 Foret des Mikea, 7.5 km brunneicauda FMNH 436441 NE Tsifota 60 -22.80 43.43 Foret des Mikea, 7.5 km brunneicauda FMNH 436442 NE Tsifota 60 -22.80 43.43 Foret des Mikea, 7.5 km brunneicauda FMNH 436443 NE Tsifota 60 -22.80 43.43 Foret des Mikea, 7.5 km brunneicauda FMNH 436444 NE Tsifota 60 -22.80 43.43 Foret des Mikea, 7.5 km brunneicauda FMNH 436445 NE Tsifota 60 -22.80 43.43 RNI de Namoroka, 22 km N Andranomavo, brunneicauda FMNH 436522 Ampandra River 120 -16.43 45.40 RNI de Namoroka, 22 km N Andranomavo, brunneicauda FMNH 436523 Ampandra River 120 -16.43 45.40

35 RNI de Namoroka, 22 km N Andranomavo, brunneicauda FMNH 436524 Ampandra River 120 -16.43 45.40 RNI de Namoroka, 22 km N Andranomavo, brunneicauda FMNH 436525 Ampandra River 120 -16.43 45.40 RNI de Namoroka, 22 km N Andranomavo, brunneicauda FMNH 436526 Ampandra River 120 -16.43 45.40 RNI de Namoroka, 22 km N Andranomavo, brunneicauda FMNH 436527 Ampandra River 120 -16.43 45.40 Province de Toliara, Forêt de Vohondava, 10 km SSO de Tranomaro, le long de la rivière archboldi UADAB MJR0330 Vohondava 225 24.69 46.45 Province Toliara, Forêt de Salary, 6 km NNE du archboldi UADAB MJR0952 village de Salary Nord 20 22.51 43.30 Province Toliara, Forêt de Salary, 6 km NNE du archboldi UADAB MJR0957 village de Salary Nord 20 22.51 43.30 Province Toliara, Forêt de Salary, 6 km NNE du archboldi UADAB MJR0958 village de Salary Nord 20 22.51 43.30 Province Toliara, Forêt de Salary, 6 km NNE du archboldi UADAB MJR0959 village de Salary Nord 20 22.51 43.30 Toliara, Foret d'Ankapoky, 13 km NE Amboasary Sud, 21 km archboldi FMNH 352944 NW Ranopiso 75 -24.00 46.52 Toliara,,PN Tsimanampetsotsa, 21.5 archboldi FMNH 434463 km NE Efoetsa 100 -24.00 43.88 Toliara, Foret des Mikea, archboldi FMNH 436438 9.5 km W Ankiloaka 80 -22.77 43.52 Toliara, Foret des Mikea, archboldi FMNH 436446 9.5 km W Ankiloaka 80 -22.77 43.52 Toliara, Foret des Mikea, archboldi FMNH 436447 7.5 km NE Tsifota 60 -22.80 43.43

Gene Region Primer L Primer H L Forward H Reserve Sequence Sequence ND3 ND3-L10755 ND3-H11151 5'-GAC TTC 5'-GAT TTG CAA TCT TTA TTG AGC AAA TCT GG- CGA AAT 3' CAA C-3'

CYTB CB-L14851 CB-H15563 5'-CCT ACT 5-'GCG TAT TAG GAT GCG AAT CAT TCG AGG AAA CCC T-3' TA-3'

GAPDH GapdL GapdH 5'-ACC TTT 5'-CAT CAA intron11 AAT GCG GTC CAC GGT GCT AAC ACG GGC ATT GC- GTT GCT GT- 3' 3'

FIB5 intron 5 Fib5F Fib6R 5'-CGC CAT 5-GCC ATC ACA GAG CTG GCG TAT ACT ATT CTG AA- GTG ACA T-3' 3'

MUSK intron 3 Musk-13F Musk-13R 5-'CTT CCA 5-GCC ATC TGC ACT CTG GCG ACA ATG ATT CTG AA- GGA AA-3' 3'

Table 3. Primers used for this study.

N. amphichroa N. brunneicauda 45 samples from 23 sites 96 samples from 32 sites

A B Figure 6. Map of sampling localities of the two species a) N. amphichroa and b) N. brunneicauda on Madagascar.

CHAPTER THREE RESULTS DETERMINING THE GENETIC DIVERSITY AMONG GENUS NEWTONIA

Sequence Data

The full alignment of the sequence data was 2,754 bp. The alignment length of each marker was 351 bp for ND3, 1007 bp for CYTB, 359 bp for GAPDH, 480 bp for MUSK, and

557 bp for FIB5. The total number of DNA sequences obtained was 720: for N. amphichroa we obtained 314 DNA sequences, for N. brunneicauda 356 DNA sequences, and N. archboldi 50

DNA sequences (Table 4).

Phylogenetic Analysis

The Maximum Likelihood and Bayesian Inference analyses showed that in the combined five-gene tree, there was one individual originally identified as N. brunneicauda but based on my genetic data highlighted in red was N. archboldi (Figure 7). There were twenty-five individuals that were identified as N. brunneicauda, but my genetic data highlighted in blue indicated that they were N. amphichroa (Figure 7). There were four individuals originally identified as N. amphichroa, but my genetic data highlighted in green showed that they were N. brunneicauda

(Figure 7). In the combined five-gene tree, all three species (N. amphichroa, N. brunneicauda, and N. archboldi) clades were supported by bootstrap values of 100 each (Figure 7). The misidentified

38 39 individuals showed the same results throughout all five individual gene trees; however, the tree topologies were different for the individual gene trees. Within the combined five-gene tree, N. amphichroa and N. brunneicauda are sister groups. Newtonia archboldi is the sister group to both N. amphichroa and N. brunneicauda. In the individual gene trees of ND3, CYTB,

FIB5, MUSK, and GAPDH, the sister relationships change but are not strongly supported in the gene trees. Within the ND3 gene tree, all three species (N. amphichroa, N. brunneicauda and N. archboldi) clades were supported by bootstrap values of 100 each (Figure 8). The CYTB gene tree showed bootstrap support of 100 for N. archboldi, 97 for N. brunneicauda, and 74 for N. amphichroa (Figure 9). The FIB5 gene tree showed bootstrap support of 100 for N. archboldi, 99 for N. brunneicauda, and 42 for N. amphichroa (Figure 10). The GAPDH gene tree showed a bootstrap support of 100 for N. archboldi, 86 for N. amphichroa, and 24 for N. brunneicauda

(Figure 11). The MUSK gene tree showed bootstrap support of 89 for N. archboldi, 50 for N. brunneicauda, and 12 for N. amphichroa (Figure 12). There was strong bootstrap support within the combined five-gene tree; however, the bootstrap support values varied across the individual gene trees. As expected, due to their slower rate of evolution, the signal from nuclear genes

(FIB5, GAPDH, MUSK) is lower in comparison to mitochondrial genes. After the identification of individuals of Newtonia was cleared up within the combined five-gene tree, there was strong bootstrap support on the nodes showing 100% for each species, and each of the three Newtonia species is monophyletic in all partitions of the datasets.

Phylogeography

Within N. amphichroa, two deeply divergent clades isolated in discrete montane regions were observed (Figure 13). Using GENGIS to place this phylogeny on a map, we showed that one clade of N. amphichroa is located in the southern region and the second clade is located in the northern and central regions of the eastern humid forest of Madagascar (Figure 13). There are several mountain ranges that run a north-south direction through the island. The northern mountains include Montagne d’Ambre, Manongarivo, and Tsaratanana; the center region consists of the Ankaratra Massif. The Andringitra lies at the southern end of the island

(Goodman et al. 2013; Figure 13).

The GenGIS analysis showed that within N. brunneicauda, there was not a clear geographic pattern (Figure 14). Newtonia brunneicauda is widely distributed along various habitats on

Madagascar, yet closely related individuals are not genetically clustered together. This species can be found within and in the vicinity of forest habitats including the southwestern spiny forest and patches on the central plateau and eastern low lands at elevations from sea level to at least

2,300 m (Safford et al. 2013). Since there are no geographic barriers, N. brunneicauda appears to be moving around more than initially suspected, and gene flow is abundant between populations.

Population genetics analysis

I wanted to know if there was population structure among the individuals of Newtonia by using a permutation test for significance. The haplotype diversity (Hd), nucleotide diversity (π),

Fu’s Fs test, and Tajima’s D were estimated with DnaSP v5.10 to determine demographic changes from the sequence data among N. amphichroa, N. brunneicauda, and N. archboldi. A

41 more detailed explanation of how and why I used these statistics to investigate the population structure is presented down below.

Haplotype diversity

My results showed that both species of N. amphichroa and N. brunneicauda showed higher haplotype diversity in comparison to N. archboldi (Table 5). The species of N. archboldi showed lower haplotype diversity; however, this species also has a smaller sample size.

Nucleotide Diversity

N. amphichroa and N. brunneicauda showed high nucleotide diversity in comparison to

N. archboldi, which showed low nucleotide diversity across the five markers: ND3, CYTB,

GAPDH, FIB5 and MUSK (Table 5). However, the N. archboldi sample size is nine individuals, which is significantly lower than that of the other two species; therefore, we need to be cautious about drawing any conclusions for this particular species.

Fu’s F Statistic and Tajima’s D Test

To test for any recent demographic change, I used Fu’s Fs test. To test for neutrality and demographic changes within species by loci I used Tajima D test. Within N. amphichroa, CYTB,

GAPDH, FIB5 and MUSK showed weakly significant negative values (Table 5). Newtonia brunneicauda showed significantly negative values across all five markers. This indicated that this species might have undergone recent population expansion. Within N. archboldi, four of the markers (ND3, CYTB, GAPDH and MUSK) showed negative values; therefore, I needed to be cautious about drawing any conclusions for this particular species (Table 5).

Haplotype Network

42 Haplotype networks were used to visualize the genetic structure and the geographic distributions between N. amphichroa and N. brunneicauda. A median-joining network of N. amphichroa CYTB haplotypes from 59 individuals showed that there were two distinct clades that clustered between the northern region and the southern region on Madagascar (Figure 15).

Median-joining networks of N. brunneicauda CYTB haplotypes from 72 individuals showed no clear genetic structure compared to geography (Figure 16). These individuals are widespread and can be found across multiple habitats on Madagascar. This indicates that N. brunneicauda are dispersing more frequently to new areas on Madagascar and might not have enough time to diverge genetically.

Species Delimitation

Bayesian species delimitation results using the program BPP indicated that the highest probability was for four distinct species of Newtonia (Table 6). These results rejected the current taxonomy of three species of Newtonia with a very low posterior probability of 0.00001. There was, however, very high support for the splitting of N. amphichroa into two species showing four species with a posterior probability of 0.97 (Table 6).

Misidentified samples

I first confirmed all misidentified individuals within the combined five-gene tree and the individual gene trees and observed the same results across all six trees. Each misidentified sequence was placed in the same wrong clade on each phylogenetic tree. Once the identification was corrected using the genetic results as most accurate, I next looked at the elevation ranges among the misidentified individuals. The elevation ranges of N. amphichroa and N. brunneicauda were compared to determine if there was any overlap in the ranges of these two

43 species, which might help explain when and where misidentifications occur in the field. The results showed that there was a wide range of overlap in the overall elevation ranges of these two species (Table 7), but there is clear separation when looking at the abundance of these species

(Figure 17). Newtonia amphichroa occurred in the higher elevations of 1250m -1625m and N. brunneicauda is found in the lower elevations of 10m -1600m. The box-and-whiskers plot

(Figure 17) demonstrates that though some individuals of N. brunneicauda are found in the higher elevations, the vast majority is found below 1000 m. The misidentified samples came from the zone of overlap in the higher elevations (Table 7).

I examined the specimens at the Field Museum of Natural History to see if the misidentified samples were difficult to distinguish morphologically. Previously misidentified samples of N. amphichroa were morphologically more similar to other correctly identified N. amphichroa samples, as the genetic results suggested. The same was true for N. brunneicauda and N. archboldi. N. amphichroa had darker under-parts and upper-parts in comparison to N. brunneicauda and N. archboldi that have lighter under-parts and upper-parts (see Table 1; Figure

5). I concluded that these samples were easy to misidentify in the field in the absence of comparative specimens. In order to accurately say that there is a new species of N. amphichroa, further attributes need to be assessed such as morphometrics, plumage differences, and vocalization to concretely say that there is a new species.

Molecular Clock Analysis

All dates from the molecular clock analyses are reported as median ages. Our analysis showed that the genus Newtonia diverged from other vangas 10.18 million years ago (MYA).

Within Newtonia, N. archboldi separated from N. amphichroa and N. brunneicauda roughly 7.36

44 MYA. I estimated that N. amphichroa separated from N. brunneicauda 4.38 MYA. Within N. amphichroa, the two major clades of southern and northern plus central diverged 1.25 MYA, and the northern and central clades separated 370,000 years ago. I believe these dates are reasonable; however, there are some controversial concerns with molecular clock analysis. To confirm if these dates are reasonable, I took a look at the molecular clocks percent errors. The 95% HPB was (4.8154, 7.1389) at 6.5 MYA for the separation between N. amphichroa, N. brunneicauda, and N. achboldi. The 95% HPB was (3.2759,5.0587) at 5 MYA for the separation of N. amphichroa and N. brunneicauda. The 95% HPB was (0.8487,1.6379) at 2.5 MYA for the separation between two major clades of N. amphichroa. The 95% HPB was (0.0541,0.2945) and

( 0.1825,0.5677) at 1 MYA for the two less distinct clades of N. amphichroa.

45 Table 4. DNA Sequences obtained using 5 Markers Accesssion# CLADE ND3 CTYB GAPDH Musk FIB5 352930 amphichroa x x x x x 363856 amphichroa x x x - - 363858 amphichroa x x x x x 384721 amphichroa x x - x x 384791 amphichroa x x x x x 393364 amphichroa x x x x - 393365 amphichroa x x x - x 393366 amphichroa x x x - x 393367 amphichroa x x x x x 393368 amphichroa x x x x x 393369 amphichroa x x x x - 393370 amphichroa - x x x - 393371 amphichroa x x x x - 393372 amphichroa x x - x - 393373 amphichroa x x x x x 393374 amphichroa x x x x x 393377 amphichroa x x x x - 393379 amphichroa x x x x x 393380 amphichroa x x x x x 393382 amphichroa x x x x - 393383 amphichroa x x x x x 393384 amphichroa x x x x - 393387 amphichroa x x x - x 393389 amphichroa x x x x x 395989 amphichroa x x - x x 395990 amphichroa x x - x x 396200 amphichroa x x - x - 396201 amphichroa x x x x x 396203 amphichroa x x - x - 396207 amphichroa x x - x - 396208 amphichroa x x x x x 396209 amphichroa x x x x - 429706 amphichroa x x x x x 429707 amphichroa x x x x x

46 431265 amphichroa x x x x x 431266 amphichroa x x - x x MJR01061 amphichroa x x x - x MJR01079 amphichroa x x x x - MJR01080 amphichroa x x - - x MJR0264 amphichroa x x x - x MJR0270 amphichroa x x - x x MJR0358 amphichroa x x x x x MJR0362 amphichroa x x x x x MJR0366 amphichroa x x x - x MJR0367 amphichroa x x - - x MJR0375 amphichroa x x - x - MJR0394 amphichroa x - x x - MJR0411 amphichroa x x - x - MJR0415 amphichroa x x - - - MJR0416 amphichroa x x x x x MJR0418 amphichroa x x x - x MJR0434 amphichroa x x x x x MJR0455 amphichroa x x x x x MJR0464 amphichroa x x x x x MJR0526 amphichroa x x - x x MJR0562B amphichroa x x x x x MJR0563A amphichroa x x x x x MJR0676 amphichroa x x x x x MJR0688 amphichroa x x x x x MJR0707 amphichroa x x x x x MJR0716 amphichroa x x x x x MJR0751 amphichroa - x x x x MJR0752 amphichroa x x x x x MJR0753 amphichroa x x x x x MJR0770 amphichroa x x x x x MJR0771 amphichroa x x x x x MJR0820 amphichroa x x x x x MJR0821 amphichroa x x x x x MJR0894 amphichroa x x - x x 345885 brunneicauda x x x x x 345886 brunneicauda x x - x x 345887 brunneicauda x x x x x

352931 brunneicauda x x - x x 352932 brunneicauda x x - x x 352933 brunneicauda x x - x x 352938 brunneicauda x x x x - 352941 brunneicauda x x - x x 352943 brunneicauda x x - x x 393375 brunneicauda x x x x x 393376 brunneicauda x x - x - 393378 brunneicauda x x x x - 393381 brunneicauda x x x x x 396064 brunneicauda x x x x x 396202 brunneicauda x x - x x 396204 brunneicauda x x x x x 396205 brunneicauda x x - x x 396206 brunneicauda x x - x x 427345 brunneicauda x x x x x 427346 brunneicauda x x x x x 427347 brunneicauda x x - x x 434458 brunneicauda x x x x x 434459 brunneicauda x x - - x 434460 brunneicauda x x - x x 434461 brunneicauda x x - x x 434462 brunneicauda x x x x x 436435 brunneicauda x x - x x 436436 brunneicauda x x - x x 436437 brunneicauda x x x x x 436439 brunneicauda x x - x x 436440 brunneicauda x x x - x 436441 brunneicauda x x x x x 436442 brunneicauda x x x x x 436443 brunneicauda x x x x x 436444 brunneicauda x x x x x 436445 brunneicauda x x - - x 436522 brunneicauda x x x - x

436523 brunneicauda - x x - x 436524 brunneicauda x x - x x 436525 brunneicauda x x x x x 436526 brunneicauda x x - x x 436527 brunneicauda x x - x x MJR01038 brunneicauda x x x x x MJR01068 brunneicauda x x x x x MJR0312 brunneicauda x x - x x MJR0317 brunneicauda x x - x - MJR0318 brunneicauda x x x x - MJR0332 brunneicauda x x x x x MJR0344 brunneicauda x x - x x MJR0370 brunneicauda x x - x x MJR0469 brunneicauda - x x x x MJR0487 brunneicauda x x x x - MJR0497 brunneicauda x x x x - MJR0499 brunneicauda x x - x - MJR0501 brunneicauda x x - x x MJR0502 brunneicauda x x x - x MJR0524 brunneicauda x x x x x MJR0528 brunneicauda x x - x x MJR0543 brunneicauda x x - x x MJR0544 brunneicauda x x x x x MJR0561 brunneicauda x x - x x MJR0580 brunneicauda x x x x x MJR0743 brunneicauda x x x - x MJR0774 brunneicauda x x - - x MJR0798 brunneicauda x x x x x MJR0872 brunneicauda x x x - x MJR0908 brunneicauda - x - x - MJR0909 brunneicauda x x x x x MJR0936 brunneicauda x x - x x MJR0938 brunneicauda x x - x - MJR0951 brunneicauda x x - x x MJR0963 brunneicauda x x x x - MJR0330 archboldi x x x x x MJR0952 archboldi x x x x x MJR0957 archboldi x x x x x MJR0958 archboldi x x x x x

MJR0959 archboldi x x x x x 352944 archboldi x x x x x 434463 archboldi x x x x x 436438 archboldi x x x x x 436446 archboldi x x x x x 436447 archboldi x x x - x Leptopterus JQ23927 JQ23921 JQ71341 JQ71339 Outgroup chabert 1 5 8 - 2 Falculea JQ23926 JQ23921 JQ71341 Outgroup palliata 6 0 6 - - Vanga JQ23928 JQ23922 JQ71343 JQ71340 Outgroup curvirostris 5 8 4 - 3

Figure 7. Phylogenetic relationships of Newtonia using maximum likelihood and Bayesian inference of the combined five genes. Colored labels are misidentified individuals.

Vanga curvirostris Falculea palliata Leptopterus chabert 352944 100 434463 0957 436447 N. archboldi 0330 0952 436446 0959 0958 436438 363856 393383 89 393382 384791 393389 393384 393387 0676 352930 363858 0270 393367 393368 393370 395990 393380 98 100 0264 393373 393369 N. amphichroa 395989 431266 393366 393364 01080 01079 393372 01061 393371 396209 393365 393377 393374 393379 431265 396207 396208 0821 0751 0455 0707 0366 0362 0394 99 0358 0416 0464 0526 429706 0771 0434 0411 396200 0752 429707 0688 0418 396203 0894 0753 0563A 384721 0375 0562B 0770 0820 0716 0367 396201 434458 345887 100 01038 0332 0908 434462 434461 0318 434460 434459 352933 0344 436437 0501 0499 427347 0938 N. brunneicauda 352938 0936 0502 0487 0497 0317 0909 396205 0872 396206 0963 436436 0951 436441 436445 436439 436435 436440 436444 436442 436443 0544 0543 436526 427346 396202 0580 0798 0561 0469 393381 436522 0524 0774 0312 0743 396204 436523 427345 0528 352941 345885 352931 393376 352943 0370 436527 345886 01068 436524 396064 436525 393378 393375 352932 0.03

Figure 8. Phylogenetic relationships of Newtonia using maximum likelihood of ND3 only. ML bootstrap support values are shown on the nodes. Falculea palliata Leptopterus chabert Vanga curvirostris 00487 00502 00497 4434459 00317 00499 00909 4436437 100 00501 4427347 3352938 00938 4434462 4434460 4434461 001038 00344 4434458 3352933 00318 58 00332 3345887 3396064 3393381 00580 00798 N. brunneicauda 00561 4427346 3396202 001068 3393376 3352932 00370 3352943 3345886 3393375 4427345 00528 3352941 3393378 3352931 56 00774 00743 00312 4436524 4436525 4436522 4436527 3345885 345885 3396204 00524 3396205 4436526 00543 00963 3396206 00872 4436441 4436435 00951 4436436 4436440 4436444 4436442 4436439 4436445 4436443 4436447 100 4436446 00957 00330 51 4434463 4436438 N. archboldi 00952 00958 00959 3352944 100 00563A 00434 00770 3396201 3396200 3384721 00455 00771 4429706 00707 00416 00688 00418 00820 00753 00716 00362 00411 N. amphichroa 00375 00526 00366 00821 00464 00367 00894 3396203 00415 00358 00752 4429707 3393383 99 00676 3393382 3363856 3352930 3393387 3393384 3363858 3384791 3393389 00562B 3393372 3393365 3393380 3395990 3393373 00270 00264 3393368 3393369 3395989 3393377 3393367 3396209 001080 3393374 3393371 001079 3393364 3396208 3393366 4431265 3393379 3396207 4431266

0.03

Figure 9. Phylogenetic relationships of Newtonia using maximum likelihood of CYTB only.

Vanga curvirostris Falculea palliata 436447 100 434463 352944 0330 0952 436446 N. archboldi 0957 0958 436438 Leptopterus chabert 363856 393387 352930 393382 0676 393383 384791 363858 384721 395989 393372 393366 74 393371 393379 396208 393365 431266 395990 81 393380 0264 393368 0270 393373 393367 431265 393377 01061 393364 396207 396209 N. amphichroa 01079 01080 0820 0751 0562B 78 0455 0716 _0770 0707 396201 0688 0752 429707 0771 0367 0563A 0894 0366 0464 0416 0415 0418 0434 0394 0375 0411 0358 0526 429706 0362 97 0317 436437 0909 0499 0501 0938 _427347 0497 0487 0502 0936 352938 345887 0332 01038 434459 434458 _0908 .434462 0344 0318 _434460 434461 396205 0543 0544 436442 436443 436444 436440 0951 396206 436435 N. brunneicauda 0963 436441 436439 0872 436526 427346 396202 0580 0798 0561 396204 01068 352943 352931 436527 0370 0528 436524 393375 393376 345886 352941 0774 393378 436523 0743 0524 0312 345885 436522 393381 0469 352932 427345 345885

0.03

Figure 10. Phylogenetic relationships of Newtonia using maximum likelihood of FIB5 only.

Leptopterus chabert Vanga curvirostris 100 3352944 00958 00952 00959 N. archboldi 4436446 4436447 00957 90 4434463 00330 4436438 00963 00951 4436436 99 4434458 3345887 3345885 001038 4434459 00909 4434460 4436522 3345886 4436440 4436524 00370 00312 4436439 4436527 N. brunneicauda 00798 4436441 3393376 001068 00524 4427346 3352943 00580 00561 4436526 00774 00344 4436444 4436435 3393381 3352941 4434461 00469 4436437 00936 00502 4436445 00501 4427345 00544 3396204 4436442 4436523 4436443 00543 3396205 00743 396064 4427347 3352932 3393375 00872 3393378 3396202 4434462 00332 345885 3352931 3393367 00676 3363858 00751 00707 42 00752 3395990 00418 3384721 3393383 3393382 396208 3393389 N. amphichroa 001061 3384791 4431265 3393379 3393380 00464 00526 00416 3393373 3395989 00434 00771 00264 00562B 00455 00688 114025 114022 00716 00270 00367 3393387 3352930 3393374 00358 3396201 3393366 114027 114026 3393368 00753 4431266 00820 00563A 3393370 00894 00770 3393372 4429707 00366 00362 0.03

Figure 11. Phylogenetic relationships of Newtonia using maximum likelihood of GAPDH only.

Falculea palliata Vanga curvirostris Leptopterus chabert 0957 352944 1974_460 0959 436438 92 436447 N. archboldi 0330 0952 434463 0958 100 436524 429707 0770 396201 0716 393367 01061 0264 0455 0751 86 0821 396209 393370 393371 114022 0707 436525 393377 393389 0497 396064 393378 396200 436436 393375 0487 352930 393366 393379 0563A 114026 0464 114027 114025 0562B 393369 N. amphichroa 0753 393372 396207 396208 0820 431265 0358 0752 0771 393365 393373 363858 363856 393387 429706 393383 0434 0394 0688 393380 0418 0362 393374 01068 0366 393364 0676 393384 01079 393382 384791 0416 0798 436527 0909 352938 436522 67967 0318 436444 436443 N. brunneicauda 24 345887 396204 434458 393381 436441 0332 345885 431266 114021 434459 345885 0872 0580 427346 0524 427345 436440 0502 436442 393376 436437 434462 01038 0963 0469 436523 114020 0544 0743 352941 434461

0.03

Figure 12. Phylogenetic relationships of Newtonia using maximum likelihood of MUSK only.

393367 0752 0688 0820 396203 393389 0418 393379 0716 0562B 0411 431265 0770 0563A 0753 393372 0367 0270 393377 0375 0751 396200 N. amphichroa 393373 396207 393380 396208 393374 393383 395990 395989 352930 396201 00771 00362 429706 12 0676 0821 429707 00366 363858 393368 393371 01079 0358 393364 384791 393370 0394 0464 431266 393384 393382 393369 0526 89 436438 0958 0959 0330 436446 26 0952 N. archboldi 0957 352944 434463 0707 0455 427346 345886 50 436444 434461 0501 0544 436443 0909 0774 393378 0963 0951 0370 0499 0502 436524 01068 0580 0743 436527 0938 436439 0317 434460 0312 345887 396204 393375 434458 434462 393381 396202 345885 0497 N. brunneicauda 436437 0524 0798 0908 352933 352943 436525 436441 436526 0487 0936 01038 427347 352932 396206 396205 352931 396064 436435 436436 0561 0528 427345

0.03

Figure 13. GenGIS analysis for N. amphichroa. Distinct clades are associated with isolated montane regions found along the eastern humid forest. Colors indicate three distinct clades - northern (orange), central (blue), and southern (purple). High mountain peaks are noted.

Montagne d’Ambre

Tsaratanana Marojejy

Ankaratra

Andringitra

57 Figure 14. GenGis analysis for N. brunneicauda. There are no distinct clades are associated with geographic regions. This species was found throughout Madagascar with no obvious geographic structure of populations.

TABLE 5. Population genetics statistics for 3 species of Newtonia.

Marker Sequences Haplotype Nucleotide Polymorphic Number Fu's F Tajima's D Gene Diversity Segregation of Statistic Diversity Sites Mutations

ND3 66 Hd: 0.675 Pi: 0.01396 S:6 Eta:7 0.765 0.25330 ND3 68 Hd: 0.355 Pi:0.01901 S:2 Eta:2 -0.12 -0.22419 ND3 10 Hd:0.200 Pi:0.00087 S:1 Eta:1 -0.339 -1.11173

CytB 67 Hd:0.733 Pi:0.00959 S:14 Eta:15 -0.196 -0.40825 CytB 72 Hd:0.968 Pi: 0.01116 S: 54 Eta: 56 -18.973 -0.86353 CytB 10 Hd:0.667 Pi: 0.00101 S: 4 Eta:4 -2.847 -1.66706

GapdH 57 Hd:0.663 Pi: 0.01199 S:8 Eta:8 -4.576 -0.41098 GapdH 44 Hd: 0.621 Pi: 0.00835 S:6 Eta: 6 -2.404 -0.45555 GapdH 10 Hd: 1.000 Pi: 0.00047 S:1 Eta:1 -0.879 -1.16439

Fib5 60 Hd:0.721 Pi: 0.00920 S:9 Eta:10 -1.075 -0.32152 Fib5 64 Hd:0.908 Pi: 0.01004 S:24 Eta:24 -12.392 -1.09729 Hd:0.200 Fib5 10 Pi: 0.00046 S:1 Eta:1 -0.339 -1.11173

Musk 60 Hd:0.551 Pi:0.00849 S:4 Eta:4 -0.78 -0.34680 Musk 64 Hd:0.092 Pi:0.00096 S:4 Eta:4 -6.633 -1.66226 Musk 5 Hd:0.600 Pi: 0.00165 S:1 Eta:1 0.626 1.22474

Figure 15. The Median-Joining Haplotype Network of CYTB haplotypes within N. amphichroa. Colors on map show different geographic areas of endemism in Madagascar (adapted from Vences et al. 2009). Matching colors were used to show geographic distribution of haplotypes in the network.

Figure 16. The Median-Joining Haplotype Network of CYTB haplotypes within N. brunneicauda. Colors on map show different geographic areas of endemism in Madagascar (adapted from Vences et al. 2009). Matching colors were used to show geographic distribution of haplotypes in the network.

Combination Model Posterior Table 6. BPP (Bayesian Probability Phylogenetics and Two species of 0.972 Phylogeography) N.amphichroa analysis Current taxonomy: 1 0.00001 species of N. amphichroa

62 Table 7. Elevation ranges for each species set comparing the elevation ranges of the different Newtonia species and the misidentified samples.

Species sets N Elevation (Minimum - Maximum) N. amphichroa 69 720m-1625m N. brunneicauda 72 10m-1600m N. archboldi 9 20m-225m Field ID: N. brunneicauda; Genetic ID: N. amphichroa 26 875m-1600m Field ID: N. brunneicauda; Genetic ID: N. archboldi 1 80m Field ID: N. amphichroa; Genetic ID: N. brunneicauda 4 800m-1600m

63 Figure 17. Box-and-whiskers plot showing elevation distribution of N. amphichroa and N. brunneicauda. Dark bar are the median values, boxes include first to third quartiles, dashed lines (whiskers) indicate minimum and maximum values. 1500 1000 Elevation 500 0

N. amphichroa N. brunneicauda Species

CHAPTER FOUR DISCUSSION UNCOVERING HIDDEN DIVERSITY AMONG TWO WIDESPREAD SPECIES OF BIRDS N. AMPHICHROA AND N. BRUNNEICAUDA My study provides insight into the evolutionary history within a passerine group on

Madagascar, the genus Newtonia. Other studies have found that the isolation of populations across barriers is often the main driving force of diversification among birds on Madagascar

(Vences et al. 2009). This study provides a unique opportunity to investigate evolutionary patterns between two widespread species and examine possible hidden diversity among these drab plumage birds.

The first objective was to determine the phylogenetic relationships among the species of

Newtonia. In my research, it was important to look at Newtonia using a larger sample size across their ranges and multiple markers in order to uncover their genetic relationships. My findings showed that N. amphichroa, N. brunneicauda, and N. archboldi are well-supported monophyletic groups. For the individual gene trees the topologies were different, and the bootstrap support values varied across the individual gene trees. Introgression is unlikely here because the clades are reciprocally monophyletic. However, my preliminary analyses showed some individuals of

N. amphichroa grouping with N. brunneicauda and vice versa. This led me to ask if this was due to errors in the lab, misidentifications in the field, or possible hybridization between these species. With the consistent placement of misidentified individuals within the combined five- gene tree and the individual gene trees, I concluded that these 64 65 samples were misidentified in the field. With the corrected identifications, these three species of

Newtonia were monophyletic and consistent across all trees. The second and third objectives were to determine if there is phylogeographic structure to indicate recent speciation within

Newtonia species and if so, whether there is evidence for differentiation due to habitat type or geography. A strength of my study is that both N. amphichroa and N. brunneicauda are well represented from across their ranges. N. amphichroa has 45 individuals from 23 sites (Figure 6A) and N. brunneicauda has 96 samples from 32 sites (Figure 6B). This thorough sampling has allowed me to get a better understanding about the geographic structure in order to uncover the true biotic diversity.

N. amphichroa

My results show that there were two distinct clades of N. amphichroa: one located in the northern/central region and the second clade located in the southern region. There are many mountains found along Madagascar, and these clades have differentiated due to isolation on these massifs. The northern massifs include Tsaratanana, Manongarivo, and Montagne d’ Ambre

(Goodman et al. 2013). The central plateau includes the forests of Manjakatompo on Ankaratra

Massif, Ambohitantely, and Kalambatritra. The southern region includes the Ivakoany massifs.

Other studies have similarly shown that these mountain ranges do have an impact on the separation among species such as Microcebus mouse lemurs (Vences et al. 2009). Furthermore, other studies of widespread reptile species also uncovered a primarily north-south split between lineages (Vences et al. 2009). Researchers have shown that species adapted to higher elevations are more likely to be endemic and remain isolated on mountains during climatic shifts (Vences et al. 2009). My study shows that montane regions have a significant impact on the separation

66 within N. amphichroa, which is found in high elevations between 1250 m and 1625 m (Figure

15). On Madagascar, there are three mountains reaching heights above 2000 m: one in the north

(Tsaratanana), one in the center (Ankaratra), and one in the south (Andringitra). Mountainous areas are ideal topographic locations for allopatric speciation of populations that remained isolated in such montane refugia. There are two possibilities for how this might occur. First, the species are adapted to the high altitude mountain peak habitats, become geographically isolated from each other on these mountain peaks, and the distance between may be too great for them to allow gene flow between the mountain peaks. Second, they are not adapted to the low altitude habitat between the mountains and cannot undergo gene flow between populations, leading to their isolation from each other.

A key aspect to understanding evolutionary patterns is addressing fundamental questions of what constitutes distinct species, how many species are in a given area, and what are the phylogenetic relationships of these species (Cracraft 2002). Cryptic species are genetically and evolutionarily distinct species that are hard to detect because they are morphologically indistinguishable from their closest relatives. I used genetic data in order to determine if the two clades of N. amphichroa are distinct species. BPP (for Bayesian Phylogenetics and

Phylogeography) is a Bayesian Markov chain Monte Carlo (MCMC) program for analyzing

DNA sequence alignment under the multispecies coalescent model (Yang 2015). The (MSC) multispecies coalescent model provides a natural framework for addressing a number of important problems in evolutionary biology such as species delimitation (Yang 2015). Species delimitation is an act of identifying species level biological diversity by using genetic sequence data. BPP is able to delimit distinct genetic groups. The BPP analysis showed very high support

67 for the splitting of N. amphichroa into two species. What these results indicate is that the two groups within N. amphichroa are not interbreeding, there is minimum gene flow, and they are not moving around. To conclude, the species delimitation model can delimit distinct genetic groupings, not species. To call it a new species, one will need to assess additional evidence such as morphometrics, plumage differences, and vocalization differences.

Molecular clocks are used to date taxonomic groups, determine the impact of climatic and geological events on diversification, estimate rates of speciation and extinction, determine the timing of dispersal events, and date the origin of gene families (Weir et al. 2008). The two major clades of N. amphichroa split 1.25 MYA, indicating a deep split within this population.

They have been separated for quite some time and on the order of other sister species pairs (Price

2008). The split of the two major clades of N. amphichroa was a reasonable date, and this was confirmed with percent errors (%HPB), which had showed a range of (0.8487, 1.6379) at 2.5

MYA. These findings indicate that both N. amphichroa and N. brunneicauda represent two divergent lineages. N. brunneicauda and N. amphichroa diverged from N. archboldi during the

Miocene and N. amphichroa from N. brunneicauda during the Pliocene. These events generated disjunct patterns of distribution in many bird groups and other vertebrate lineages (Barker et al.

2004). Both the BPP and the molecular clock analyses showed that the two divergent clades are distinct groups that have been separated long enough to be considered different species.

N. brunneicauda

The Fu’s F test and Tajima’s D test were used to infer whether the populations have intraspecific genetic structure and to detect signatures of demographic change. The results showed strongly significant negative values for the Fu’s F test and Tajima’s D test across the five

68 gene markers, indicating that the population has undergone recent population expansion within

N. brunneicauda in comparison to N. amphichroa. The haplotype network of N. brunneicauda did not show any clear geographic patterns to indicate segregation within distinct areas. There are two obvious explanations. The first is that there is a high degree of gene flow among these different regions, and the second is that there has been a recent expansion in numbers of N. brunneicauda individuals as they spread into new areas on Madagascar and have not had time to diverge genetically yet. In contrast, N. amphichroa showed a clear pattern in the haplotype network of a northern/central clade and a southern clade.

The fourth objective was to determine if some specimens were misidentified in the field.

I examined three alternative hypotheses of why some individuals were mixed up in the wrong clade in the phylogenetic analysis by testing if this was due to (a) errors in the lab, (b) hybridization, or (c) misidentifications in the field. To address the first possibility, I repeated some lab experiments by redoing some extractions, PCRs, and sequencing and always got the same consistent results. To address the second consideration, I compared the phylogenies using mtDNA genes and nuclear genes. If there are hybrids, I should expect the signal in maternally inherited mtDNA to be different from biparentally inherited nuclear genes. But these individuals were consistent in their placement in each tree. Finally, I explored the possibility of field misidentifications. My collaborator Steve Goodman, who helped collect many of the samples I used, indicated that this was a possibility since these birds are so similar looking. Researchers in the field use elevation, plumage patterns, and song to help identify species. We did not have access to song so we focused on the elevation and plumage patterns.

I examined specimens of N. amphichroa, N. brunneicauda, and N. archboldi at the Field

Museum of Natural History to compare their plumage patterns and to confirm species. In my initial diagnosis of the Newtonia species, it was hard to distinguish between N. amphichroa and

N. brunneicauda. My diagnosis included looking first at the upper-parts and then looking at the under-parts. Then I looked at the coloration, checking to see if there were any key differences in coloration among the individuals. However, when I separated out the genetically identified N. amphichroa from the N. brunneicauda and vice versa, it was clear the N. amphichroa were overall darker in appearance and were more similar to each other than to N. brunneicauda, which are lighter in appearance. So I concluded that these samples were misidentified in the field.

Conservation Concerns

Over the past few decades, Madagascar has been losing forest at alarming rates due to slash and burn agriculture, logging, charcoal production, and collection of wood for fuel. The diversity of bird life on Madagascar is still poorly understood; it is unclear how great this threat is. Natural processes can lead to isolation and adaptation on montane regions, like in the case of

N. amphichroa, and these restricted species are vulnerable to habitat destruction. Anthropogenic causes, such as forest fragmentation, can lead to smaller populations being isolated and declines in population size. Fragmented forests have already had an impact on the bird communities on

Madagascar (Goodman et al. 2003). More than 70% of the eastern evergreen forests have been cleared since the arrival of humans on Madagascar 2,000 years ago. As the forests of

Madagascar are degraded and reduced in coverage, the remaining natural habitats are being cut into smaller and smaller pieces (Goodman et al. 2003). The effects of habitat fragmentation have important ramifications for the forest-dwelling endemic birds on the island (Goodman et al.

2003). A recent study found that the number of forest-dependent species declines with decreasing fragment size (Andrianarimisa et al. 2000).

The diversity patterns of most animal and plant groups in Madagascar relate to the primary biomes: rainforest, deciduous forest, spiny forest, and grasslands (Vences et al. 2009).

The habitat conditions in all these areas are becoming increasingly degraded as humans are impacting these regions. Further studies are needed to understand the impact on bird populations.

After determining the phylogenetic relationships and geographic structure within N. amphichroa, I have observed two distinct clades separated by a montane regions. Due to this clear separation, it is important to have a good understanding about how many distinct species are present, and my study is a great opportunity for species discovery. If there is a different species present as shown within N. amphichroa, managing the genetic diversity is a key component to managing new species. Knowing what species are out there is the first step in saving what’s out there. Conservation measures should be taken into consideration because as these populations get smaller and more isolated due to habitat degradation, species do not have the necessary ability to be resilient in the face of change.

Knowing the specific geographic areas or habitat types being impacted by environmental factors like forest fragmentation will help in providing protection against degradation.

Conservation strategies require a multifaceted approach that involves a community-based technique. These include environmental education, reforestation, sustainable agriculture, and research. Cryptic species require special consideration in conservation planning because they are rare individuals. If there are multiple species present as shown with N. amphichroa, this means that different species might require different conservation strategies (Bickford et al. 2006). These

71 newly discovered cryptic species provide a great opportunity for conservation management

(Bickford et al. 2006). Failure to resolve taxonomic relationships may result in the neglect of certain regions and species if adequate conservation measures are not taken. These conservation measures include estimation of species richness and endemism. The discovery of geographical and habitat-related patterns in cryptic species can uncover diversity and sites that are in need of conservation.

CHAPTER FIVE CONCLUSION IDENTIFYING DISTINCT SPECIES PRESENT WITHIN N.AMPHICHROA

The objective of my study was to determine the phylogenetic relationships and the phylogeographic patterns. Determining if the Newtonia species were misidentified in the field, the number of distinct species present, and the time of divergence. These key objectives allowed for the uncovering of the hidden diversification among the three Newtonia species. Cryptic diversification is when genetically and evolutionarily distinct species that are hard to detect because they are morphologically indistinguishable from their closest relatives. This has been hypothesized to occur in many groups on Madagascar. Currently, it is unclear to what extent this phenomenon occurs in birds because only a few studies have been conducted. Previous studies of vertebrates have shown that the current classifications are often misleading when organisms exhibit only slight morphological differences (Yoder et al. 2005), therefore the diversity of the extant avifauna may be obscured due to cryptic diversification. The identification, and description of cryptic species have important implications for conservations and natural resource protection and management (Birkford et al.2006). Therefore, Madagascar has a high need for conservation concerns for these forest dwelling bird groups. As a result of this concern many of the unique birds groups on Madagascar are probably at risk of extinction, but how great this threat maybe it is still unclear, because the diversity of

72 73

Madagascar’s bird life is still poorly understood. Failure to resolve taxonomic relationships may result in the neglect of certain regions and species if adequate conservation measures are not taken.

The results of my study has met these objectives with the well-supported phylogenetic tree indicating the three species of Newtonia are monophyletic. The GenGis results from N. amphichroa showed two distinct clades separated by montane regions. The BPP analysis indicated very high support for the splitting of N. amphichroa into two species indicating the presence of speciation. As previously mentioned, N. amphichroa was considered be one species.

My analysis confirmed the presence of cryptic species within N. amphichroa. There were two major N. amphichroa clades diverged 1.25 MYA, indicating these two distinct clades have been separated long enough to call them different species.

Cryptic species requires special consideration in conservation on Madagascar. Previous researchers found dual problems: (I) species already considered endangered or threatened might be composed of multiple species that maybe rare and (II) if there are different species present they might require different conservation strategies (Brickford et al. 2007). Although there are multiple habitats on Madagascar some are fast disappearing. Knowing the number of distinct species will help in monitoring the habitats and it will provide additional justification and protection for the unique habitats on Madagascar. Studying phylogenetic relationships and phylogeographic patterns can help in uncovering the hidden diversity among bird groups on

Madagascar.

REFERENCE LIST

Alexandros S. 2014. RAxML Version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30 (9): 1312-1313.

Andrianarimisa A, Bachmann L, Ganzhorn JN, Goodman SM, Tomiuk J. 2000. Effects of forest fragmentation on genetic variation in endemic understory forest birds in central Madagascar. Journal für Ornithologie 141:152-159.

Beier P, Drielen MV, Kankam BO. 2002. Avifaunal Collapse in West African Forest Fragments. Conservation Biology 16: 1097-1111.

Barker FK, Cibois A, Schikler P, Feinstein J, and Cracraft J. 2004. Phylogeny and diversification of the largest avian radiation. Proceedings of the National Academy of Sciences. vol.101:11040- 11045.

Bickford D, Lohman DJ, Sodhi NS, Ng PKL, Meier R, Winker K, Ingram KK, Das I. 2007. Cryptic species as a window on diversity and conservation. Trends in Ecology and Evolution 22 (3): 148-155.

Block, NL. 2012. Cryptic Diversity and Phylogeography in the Bernieridae, an Endemic Malagasy Passerine Radiation. PhD Dissertation.University of Chicago.

Block NL, Goodman SM, Hackett S, Bates JM, Raherilalao MJ. 2015. Potential merger of ancient lineages in a passerine discovered based on evidence from host-specific ectoparasites. Ecology and Evolution 5(17): 3743-3755.

Boumans L, Vieites DR, Glaw F, Vences M. 2007. Geographical patterns of deep mitochondrial differentiation in widespread Malagasy reptiles. Molecular Phylogenetics and Evolution 45: 822- 839.

Bouckaert R, Kuhnert D, Vaughan T, Wu CH, Xie D, Suchard MA, Rambaut A, Drummond AJ. 2014. Beast 2:A Software Platform for Bayesian Evolutionary Analysis. PLOS Computational Biology.

Cabanne GS, d’Horta FM, Sari HER, Santos FR, and Miyak CY. 2008. Nuclear and mitochondrial phylogeography of the Atlantic forest endemic Xiphorhynchus fuscus (Aves: Dendrocolaptidae): Biogeography and systematics implications. Molecular Phylogenetics and Evolution 49:760-773.

Cibois A, Slikas B, Schulenberg TS, Pasquet E. 2001. An endemic radiation of

74 75 Malagasy songbirds is revealed by mitochondrial DNA sequence data. Evolution 55:1198-1206.

Cracraft J. 2002. The seven great questions of systematic biology: An essential foundation for conservation and the sustainable use of biodiversity. Annals of the Missouri Botanical Garden 89: 127-144.

Fuchs J, Parra JL, Goodman SM, Raherilalao MJ, Vanderwal J, Bowie RCK. 2013. Extending ecological niche models to the past 120,000 years corroborates the lack of strong phylogeographic structure in the Crested Drongo (Dicrurus forticatus forficatus) on Madagascar. Biological Journal of the Linnean Society 108: 658-676.

Fuchs, J. Lemoine D, Parra JL, Pons JM, Raherilalao MJ, Jones RP, Warren BH, Goodman SM. 2016. Long- distance dispersal and inter-island colonization across the western Malagasy Region explain diversification in brush-warblers (Passerformes: Nesillas). Biological Journal of the Linnean Society.

Goodman SM, Raherilalao MJ. 2013. Atlas of Selected Land Vertebrates of Madagascar. Association Vahatra, Antananarivo.

Goodman SM, Benstead JP. 2003. The Natural History of Madagascar. The University of Chicago Press. Chicago and London.

Hey J. 2009. Why Should We Care about Species? Nature Education 2:2.

Hoyo JD, Elliott A, Christie D. 2007. Handbook of the Birds of the World. Volume 12. Picathartes to Tits and Chickadees. Lynx Edicions.

Jønsson KA, Fabre P-H, Fritz SA, Etienne RS, Ricklefs RE, Jørgensen TB, Fjeldså Rahbek JC, Ericson PGP, Woog F, Pasquet E, Irestedt M. 2012. Ecological and evolutionary determinants for the adaptive radiation of the Madagascan vangas. Proceedings of the National Academy of Sciences 109: 6620-6625.

Kimball RT, Braun EL, Barker FK, Bowie RCK, Braun MJ, Chojnowski JL, Hackett SJ, Han KL, Harshman J, Torres VH, Holznagel W, Huddleston CJ, Marks BD, Miglia KJ, Moore WS, Reddy S, Sheldon FH, Smith JV, Witt CC, and Yuri T. 2008. A well-tested set of primers to amplify regions spread across the avian genome. Molecular Phylogenetics and Evolution 50: 654-660.

Lanfear R, Calcott B, Ho SYW, Guindon S. 2012. PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Molecular.Biology and Evolution 29 (6): 1695-1701.

Librado P, Rozas J. 2009. DnaSP v5: software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: 1451-1452.

Lohman DJ, Ingram KK, Prawiradilaga DM, Winker K, Sheldon FH, Moyle RG, Ng PKL, Ong

76 PS, Wang KL, Braile TM, and Astuti D. 2010. Cryyptic genetic diversity in “widespread” Southeast Asian bird species suggests that Philippine avian endemism is gravely underestimated. Biological Conservation 143: 1885-1890.

Myers N. Mittermeier RA, Mittermeier CG, Fonseca da GAB, Kent J. 2000. Biodiversity hotspots for conservation priorities. Nature. Volume 403.

Parks DH, Mankowski T, Zangooei S, Porter MS, Armanini DG, Baird DJ, Langille MGI, and Beiko RG.2013. GenGis 2: Geospatial analysis of traditional and genetic biodiversity with new gradient algorithms and an extensible plugin framework. PLOS One.

Pastorini J, Thalmann U, Martin RD. 2003. A molecular approach to comparative phylogeography of extant Malagasy lemurs. PNAS 100: 5879-5884.

PopArt Program. Website: http://popart.otago.ac.nz.

Price T. 2008. Speciation in Birds. Robert & Company Publishers. Greenwood Village Colorado.

Rambaut A, Suchard MA, Xie D, Drummond AJ. 2014. Tracer v1.6, Available from http://beast.bio.ed.ac.uk/Tracer.

Rambaut, A.2010. FigTree. 1.3.1. Scotland: University of Edinburgh.

Reddy S, Driskell A, Rabosky DL, Hackett SJ, Schulenberg TS. 2012. Diversification and the adaptive radiation of the vangas of Madagascar. Proceedings of the Royal Society 279: 2062- 2071.

Safford R, Hawkins F. 2013. The Birds of Africa. Volume VIII. The Malagasy Region. Christopher Helm, London.

Schulenberg TS. 2003. Vangidae, Vangas. In The Natural History of Madagascar (eds Goodman SM, Benstead JP), University of Chicago Press, Chicago, IL.1138-1143.

Vences M, Wollenberg KC, Vieites DR, Lees DC. 2009. Madagascar as a model region of species diversification. Trends in Ecology and Evolution 24: 456-465.

Watson EM, Whittaker RJ, Dawson TP. 2004. Habitat structure and proximity to forest edge affect the abundance and distribution of forest-dependent birds in tropical coastal forests of southeastern Madagascar. Biological Conservation 120: 311-327.

Weir JT, Schluter D. 2009. Calibrating the avian molecular clock. Molecular Ecology 17: 2321- 2328.

Wilme L, Goodman SM, and Ganzhorn JU. 2006. Biogeographic evolution of Madagascar’s microendemic biota. Science 312:1063-1065.

77 Yamagishi S, Nakamura M. 2009. Handbook of the Birds of the World, eds., del Hoyo J, Elliott A, Christie DA. Lynx Edicions.

Yang Z. 2015. The BPP program for species tree estimation and species delimitation. Current Zoology 61(5):854-865.

Yoder AD, Olson LE, Hanley C, Heckman KL, Rasoloarison R, Russell A L, Ranivo J, Soarimalala V, Karanth KP, Raselimanana AP, Goodman SM. 2005. A multidimensional approach for detecting species patterns in Malagasy vertebrates. Proceedings of National Academy of Sciences 102: 6587-6594.

VITA

Lynika Strozier earned a Bachelor of Science at Dominican University in the spring

2011. During that same year she had the opportunity to be teachers assistant for the microbiology lab sections at Dominican University. After graduating from Dominican University she was a part of a one year Post-Baccalaureate Research Education Program at the University of Chicago in the summer of 2011.

As an undergraduate she served as a research assistant at the Field Museum of Natural

History with Dr. Matthew Von Konrat. There she studied the evolutionary relationships among early land plants. Beginning in the summer of 2009, Lynika began an internship and later became staff at the Field Museum of Natural History for Dr. Matthew von Konrat in the Botany

Department. She contributed to the understanding of the evolutionary history of early land plants by using molecular tools. Her hard work allowed her to coauthor two peer reviewed journal articles one in 2011 and the second in 2015. Over the years she not only worked on early land plants she also performed DNA Sequencing on Florida Keys ants, and on lichens.

Lynika came to Dr. Sushma Reddy’s lab in 2013 when she began her Master of Science in biology at Loyola University Chicago. Here she had the opportunity to continue exploring the evolutionary relationships among birds in the genus

78 79

Newtonia endemic to Madagascar. This work from her thesis has been presented at the American

Ornithologists Union Conference in September of 2014. During her time here she has been on a panel for “This Is How We Science” at the Field Museum of Natural History where she had the opportunity to be one of the members to speak to high school students about being a graduate student in the sciences. During her summers, she trained undergraduate students at the Field

Museum of Natural History on DNA sequencing and assisted with lab projects.