The Evolutionary Diversification and Biogeography of Parrots (Aves: Psittaciformes): an Integrative Approach

Home , Alisterus, Aprosmictus, Arini (tribe), Guaiabero, Psilopsiagon, Psittacinae

Eawag_06930

The evolutionary diversification and biogeography of parrots (Aves: Psittaciformes): an integrative approach

Inauguraldissertation der Philosophisch-naturwissenschaftlichen Fakultät der Universität Bern

vorgelegt von

Manuel Schweizer

von Hasle bei Burgdorf BE

Leiter der Arbeit:

Prof. Marcel Güntert Naturhistorisches Museum der Burgergemeinde Bern

Prof. Ole Seehausen Institut für Ökologie und Evolution Universität Bern

The evolutionary diversification and biogeography of parrots (Aves: Psittaciformes): an integrative approach

Inauguraldissertation der Philosophisch-naturwissenschaftlichen Fakultät der Universität Bern

vorgelegt von

Manuel Schweizer

von Hasle bei Burgdorf BE

Leiter der Arbeit:

Prof. Marcel Güntert Naturhistorisches Museum der Burgergemeinde Bern

Prof. Ole Seehausen Institut für Ökologie und Evolution Universität Bern

Von der Philosophisch-naturwissenschaftlichen Fakultät angenommen.

Der Dekan / Die Dekanin: Bern, 1.11.2011 Prof. Dr. Silvio Decurtins

Table of contents

TABLE OF CONTENTS

Chapter 1 5 Introduction

Chapter 2 25 Summary of chapters and synthesis

Chapter 3 53 The evolutionary diversification of parrots supports a taxon pulse model with multiple trans-oceanic dispersal events and local radiations

Chapter 4 67 Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations

Chapter 5 95 Nectarivory in parrots is a key innovation that triggered parallel adaptations and species proliferation through a nonadaptive radiation

Chapter 6 133 Out of the Bassian province: historical biogeography of the Australasian platycercine parrots

Chapter 7 159 Phylogeny and biogeography of the parrot genus Prioniturus (Aves: Psittaciformes)

Chapter 8 181 Disparity versus diversity and the role of ecological opportunity in the evolution of Neotropical parrots

Acknowledgements 207

Erklärung 209

Curriculum vitae 211

Chapter 1

INTRODUCTION

In writing this PhD thesis I have attempted to gain an understanding of biogeographic and macroevolutionary patterns in the diversification of a particular species-rich avian group, namely parrots (Aves: Psittaciformes), by integrating mainly molecular genetic, but also morphological data in an ecological and historical context. A well resolved phylogenetic hypothesis is the prerequisite for predictions about evolutionary diversification processes and mechanism in a particular group. Therefore, the reconstruction of the phylogeny of the major groups of parrots provided the basis for this work. Biodiversity in the sense of species richness is neither randomly nor uniformly distributed on earth with some regions being much more diverse than others and different regions being inhabited by distinct species assemblages (Myers et al. 2000; Mittelbach et al. 2007; Gotelli et al. 2009; Lomolino et al. 2010). Moreover, species richness is highly unevenly distributed among taxonomic groups (Newton 2003; Ricklefs et al. 2007; Soltis 2007; Alfaro et al. 2009). Since the beginning of modern science, biologists have searched for explanations for these discernible spatial and taxonomic patterns of contemporary global diversity. The study of processes that generate and maintain biological diversity is still a major research issue at the interface not only between micro- and macroevolution, but also between different scientific disciplines such as evolutionary biology, ecology, paleontology, geology or climatology. One important goal is to infer the influence of ecological and evolutionary mechanisms on the relationships between time, dispersal and diversification (Wiens 2011). Understanding this is not only important in a historical context, but it is also prerequisite to predicting ecological changes and ecosystem function in the context of ongoing human-induced mass extinction (Ricklefs 1987). However, the timing and rate of diversification and their causes are still particularly poorly understood or controversially discussed for different organismic groups (e.g. Penny and Phillips 2004; Weir 2006; Bininda-Emonds et al. 2007; Brown et al. 2008; Nishihara et al. 2009). Species richness is basically the results of speciation, extinction and dispersal (Wiens 2011). These processes are dependent on time (e.g. clade age) or geographical area, but also influenced by historical events like continental drift, climate change or mountain orogenesis, by biotic factors like competition or ecological limits and intrinsic factors such as the potential to colonize new areas or to adapt to new ecological conditions (Hunter 1998; Price 2008; Thomas et al. 2008; Rabosky 2009; Lomolino et al. 2010; Yoder et al. 2010; Wiens 2011). Hence, diversity is the result of interactions between properties of the organisms themselves and their environment (Newton 2003). Diversification is promoted among other things by vicariance events, range expansions and ecological opportunities subsequent to habitat change, the colonization of new areas or the evolution of key innovations. In particular, evolutionary key innovations have often been considered as essential for promoting diversification (Heard and Hauser 1995; Hunter 1998). All these processes play an important role in different parts of this thesis and will be briefly introduced here.

7 Evolutionary diversification and biogeography of parrots

Speciation and Extinction

Several species concepts have been proposed to categorize the discernible discontinuities among groups of organisms observed in nature (e.g. Helbig et al. 2002). A discussion of these is clearly beyond the scope of this introduction. The so- called biological species concept puts the focus on reproductive isolation among groups of populations (Mayr 1942) and research on speciation has mainly focused on mechanisms leading to such reproductive isolation (Coyne and Orr 2004; Price 2008). It is now widely accepted that species arise in many cases by means of natural selection (Price 2008; Schluter 2009). The process of speciation can be viewed under different categorizations. One considers speciation in a geographical context, and a common classification scheme ranges from allopatric over parapatric to sympatric speciation (Price 2008). The formation of two species from geographically separated diverging populations is called allopatric speciation. It can proceed via vicariance when the range of an ancestral species is geographically separated by a barrier, e.g. through climatic or geographical events (Coyne and Orr 2004, see also below). Another mode of allopatric speciation is called peripatric speciation. In this case, the separation of an ancestral range of a species is reached through a founder event by the colonization of a new area (e.g. islands) usually by a small subset of the ancestral species population (Coyne and Orr 2004; Lomolino et al. 2010). Speciation can also take place in spatially overlapping populations (Coyne and Orr 2004). It is called sympatric speciation, when overlap between ancestral populations is extensive and interbreeding is possible throughout the process. In the case of parapatric speciation, however, areas of ancestral populations only partially overlap, making initial interbreeding possible (Coyne and Orr 2004; Price 2008; Lomolino et al. 2010). Another classification of the speciation process is concerned with the mechanisms and processes responsible for the evolution of reproductive isolation (Rundle and Nosil 2005; Price 2008). In this context, a distinction is made between ecologically-based and non-ecological or mutation-order processes that generate reproductive isolation (Rundle and Nosil 2005; Schluter 2009). In non-ecological processes, chance is thought to play an important role in the achievement of reproductive isolation through processes of genetic drift, founder events and population bottlenecks, hybridization or polyploidization (Coyne and Orr 2004; Rundle and Nosil 2005). Nevertheless, selection may be involved such as in speciation through some modes of sexual selection or the fixation of incompatible alleles (Schluter 2001; Rundle and Nosil 2005). However, this selection is not directly related to the environment, and fixed distinct mutations can be advantageous in both environments of the initial populations (Coyne and Orr 2004; Schluter 2009). In contrast, reproductive isolation can evolve directly or indirectly through divergent ecologically-based selection on traits between populations in contrasting environments (Schluter 2001; Rundle and Nosil 2005). This process is referred to as ecological speciation. It is a common mechanism by which new species arise, and strong connections between reproductive isolation and selection on phenotypic traits are often found (Schluter 2009). Ecological speciation can proceed in any geographical context (e.g. in sympatry or allopatry) and divergent ecologically-based selection can be triggered by environmental differences between populations, by differences in mate preferences between environments as well as by ecological interactions between populations that are basically sympatric (Rundle and Nosil 2005). Extinction can be considered to be the antagonistic process to speciation. The balance between speciation and extinction is expressed in diversification rates (e.g. Rabosky

8 Chapter 1

2009) and the net diversification rate is the difference between the rates of speciation and extinction. In the context of the current human-induced biodiversity crisis, extinction is usually perceived negatively. However, extinction is the final evolutionary process and the inevitable fate of every species, and it has been a crucial driver of evolution in the past (Purvis et al. 2000; Erwin 2008). Most species that once occurred on earth have become extinct and the current multicellular diversity only accounts for approximately 1-2% of all species that have existed during the past 600 million years (Erwin 2008). Extinction rates not only vary geographically (e.g. Mittelbach et al. 2007), but have also not been constant through time and the background extinction rate has been exceeded several times in the history of life on earth (Benton 1995). Five major mass extinction events have been identified during the Phanerozoic, when a significant fraction of the organismic diversity was extirpated (e.g. Jablonski 2005; Erwin 2008). These biotic crises created new ecological opportunities and changed the course of evolution by mediating faunal and floral replacements (Jablonski 1991; Erwin 2001). Mass extinctions usually led to ‘surviving intervals’ with low diversity followed by variable recovery phases with a rapid increase in morphological disparity and taxic diversity (Erwin 2001, 2008). Ecological and evolutionary dominance patterns consequently often changed during mass extinctions and previously unimportant taxa may have become prominent biotic elements in the aftermath (Jablonski 1991; Erwin 2008). In the context of this thesis, the end Cretaceous mass extinction 65 million years ago (Ma), with the supersession of pterosaurs and non-avian dinosaurs by ‘modern birds’ and placental mammals (Benton 1995; Penny and Phillips 2004) is particularly important (see below).

Dispersal vs. vicariance

For a long time, vicariance approaches have dominated historical biogeography. Vicariance is generally considered to be the split of a geographical range of a species into one or more parts through a barrier caused by historical events like mountain orogenesis or ocean formation through tectonic rifting (de Queiroz 2005; Lomolino et al. 2010). Such barrier formation was considered to lead to episodes of allopatric speciation in multiple clades, generating congruent biogeographic patterns among them (cf. Halas et al. 2005; Upchurch 2008). Vicariance was opposed to dispersal, the extension of a species’ geographical range through active movement (de Queiroz 2005; Lomolino et al. 2010). It was argued by vicariance or cladistic biogeographers that dispersal could potentially explain any distribution pattern, and dispersal hypotheses were thus claimed to be “not capable of falsification” and therefore unscientific (cf. de Queiroz 2005; McGlone 2005). The increased availability of phylogenetic trees during the last decades and their usage in a biogeographic context has provided a means of testing whether the hypothesized vicariance history is congruent with the branching patterns among taxa that have been analyzed (de Queiroz 2005). Moreover, the development of sophisticated methods to estimate divergence times from molecular genetic data in recent years (e.g. Drummond et al. 2006; Ho and Phillips 2009) has allowed researchers to explicitly compare temporal associations of branching patterns and potential drivers of vicariance. Vicariance is then usually considered the null hypothesis that is rejected if phylogenetic divergence dates are incongruent with potential vicariance events, leaving dispersal as the most likely alternative (e.g. Trenel et al. 2007; Pramuk et al. 2008). Reliable calibrations are the prerequisite for dating

9 Evolutionary diversification and biogeography of parrots divergence times from molecular data (Ho and Phillips 2009). However, rather than using independent dating points, e.g. from well accepted fossils, hypothetical relationships between biological and geological events have often been used to calibrate molecular phylogenies in different organismic groups, including parrots (e.g. Tavares et al. 2006; Ribas et al. 2007; Wright et al. 2008). Such calibrations have been called into question as cases of circular reasoning because assessments of evidence in favor of the temporal congruence of the biological and geological phenomena suffers from non-independence (Waters and Craw 2006; Upchurch 2008). Nevertheless, several recently published studies have provided firm evidence for discordance between sequences of geological events and phylogenetic patterns, often in combination with divergence time estimates (e.g. Raxworthy et al. 2002; Vences et al. 2003; Yoder et al. 2003; Pereira et al. 2007; Harshman et al. 2008; Voelker et al. 2009). Hence, these studies invoked dispersal as an explanation for the current distribution patterns of different animal groups. This has led to the conclusion that dispersal is an important process in speciation and the build-up of regional faunas and that the significance of trans-oceanic and long-distance dispersal has been severely underestimated in historical biogeography (de Queiroz 2005; McGlone 2005; Cowie and Holland 2006; Yoder and Nowak 2006; Heaney 2007). This supported the suggestion of Halas et al. (2005) that the null hypothesis for historical biogeography may have to be changed from vicariance to taxon pulses. A taxon pulse model of diversification as introduced by Erwin (1981) incorporates both dispersal and vicariance and differs from a vicariance driven model of biotic diversification in three ways (Halas et al. 2005). First, diversification is driven by biotic expansion, and current distribution patterns of groups of organisms are the result of vicariance and dispersal. Second, areas may show a biogeographically reticulated history as they may have been colonized by multiple lineages at different times and often from different sources. Third, the absence of particular clades in a particular area is often the result of a lack of participation in biotic expansion rather than the result of extinction.

Colonization, key innovations, ecological opportunity and adaptive radiations

Colonization is regarded as the dispersal of a species to an area where it has previously not occurred, followed by the establishment of a viable population (Bellemain and Ricklefs 2008). The probability of successful colonization depends on many factors, including dispersal ability, habitat selection, reproductive strategy or capability to cope with new ecological challenges like interspecific interactions (Lomolino et al. 2010). The study of colonization has been broadly influenced by theories on island biogeography pioneered by MacArthur & Wilson (1967), which were later expanded to metapopulation concepts (e.g. Hanski 1999). In this context, colonization has been considered to proceed mainly unidirectionally, flowing from larger to smaller sites, e.g. from continents to islands. However, several recent studies have found evidence for the occurrence of ‘reverse colonization’, i.e. the recolonization of continents from islands, suggesting that small sites like oceanic islands must not be considered to be merely sinks for biodiversity, but can contribute to the build-up of biota on larger landmasses (Raxworthy et al. 2002; Filardi and Moyle 2005; Heaney 2007; Bellemain and Ricklefs 2008; Jonsson et al. 2010). Colonization of new areas (e.g. continents, island groups or newly generated habitats by mountain orogenesis or climate change) potentially results in subsequent radiations (e.g. Losos and DeQueiroz 1997; Lovette et al. 2002; Burbrink and Pyron 2009;

10 Chapter 1

Losos and Ricklefs 2009; Voelker et al. 2009). Newly colonized areas may offer under-utilized adaptive zones facilitating diversification. Colonization can thus lead to ecological opportunities, i.e. the presence of various underexploited or inefficiently exploited resources (Price 2008). In general, ecological opportunities lead to relief from natural selection in ecologically important traits followed by ecological release (Nosil and Reimchen 2005; Yoder et al. 2010). The latter term describes the process of an expansion of resource use, an increase in population density or in morphological and behavioral variation of organisms following the relaxation of selection constraints (Losos and DeQueiroz 1997; Yoder et al. 2010). As a consequence, ecological opportunities may promote lineage and morphological diversification (e.g. Schluter 2000; Kassen 2009; Losos 2010). Apart from the colonization of new environments, ecological opportunities may also be provided by the extinction of antagonists or by evolutionary key innovations (Simpson 1949). Key innovations can be considered attributes of an organism that are associated with taxonomic or evolutionary success, e.g. with taxonomic diversity, such as the number of species in a group (Heard and Hauser 1995; Hunter 1998; Lynch 2009). Key innovations include functionally important morphological changes which allow individuals to enter a new adaptive zone and to expand their use of energy, to increase their individual fitness or their potential for reproductive or ecological specialization (Heard and Hauser 1995; Vermeij 1995; Hunter 1998; Price et al. 2010). Ecological opportunities may lead to strong directional selection and fast adaptations and may be followed by adaptive radiations in some cases (Hunter 1998; Kassen 2009; Losos 2010; Yoder et al. 2010). The concept of adaptive radiations links ecological differentiation with ecological speciation (Price 2008). According to Schluter (2000), adaptive radiations concern monophyletic clades that have diversified under increased speciation rates into a group of ecologically separated species, which differ in traits adapted to exploit their different ecological niches. Adaptive radiations are thus characterized by rapid multiplication of lineages early in their history in combination with an increase of ecological and morphological disparity among these lineages (Schluter 2000; Harmon et al. 2003; Harmon et al. 2010; Slater et al. 2010). Adaptive radiations have been postulated to be responsible for much of the biodiversity found on earth (Simpson 1953), however, as pointed out by Harmon et al. (2010), “identifying adaptive radiations has proven difficult due to a lack of broad-scale comparative datasets”. The few well-documented spectacular examples of adaptive radiations stem basically from isolated lakes or islands (Harmon et al. 2010) and include, e.g., African cichlid fishes, Darwin finches on Galapagos, Hawaiian honeycreepers or Anolis lizards in the Caribbean (Pratt 2005; Seehausen 2006; Grant and Grant 2008; Price 2008; Losos 2009). Theories on adaptive radiations predict that the early burst of lineage diversification and morphological evolution should be followed by a slowdown as a consequence of density dependence when ecological space becomes filled and saturated, leaving fewer opportunities for speciation (Harmon et al. 2003; Seehausen 2006; Rabosky and Lovette 2008a; Harmon et al. 2010). The decrease in speciation rate should thus be inversely correlated with species richness through the progress of the radiation (Rabosky and Lovette 2008b; Burbrink and Pyron 2009). A decline in diversification rates after an early burst has been detected in various groups of organisms (Harmon et al. 2003; Kozak et al. 2005; Ruber and Zardoya 2005; Weir 2006; Phillimore and Price 2008; Rabosky and Lovette 2008a; Burbrink and Pyron 2009; Brock et al. 2011). In contrast, a pattern of initially rapid morphological evolution followed by a slowdown through time seems to be rare in comparative data (Harmon et al. 2010).

11 Evolutionary diversification and biogeography of parrots

Adaptive radiations need to be distinguished from nonadaptive radiations. Multiple speciation events early in a clade’s history may not only result from rapid ecological speciation. Such a pattern can also be caused by a more or less simultaneous fragmentation of populations into multiple geographically isolated lineages with subsequent speciation primarily by non-ecological modes (Price 2008; Rundell and Price 2009). The resulting collection of related ecologically similar species which replace each other geographically is called a nonadaptive radiation (Rundell and Price 2009). After (nonecological) speciation in terms of reproductive isolation has been completed in allopatry, secondary contact between such species might trigger ecological differentiation through character displacement e.g. in the course of ecological opportunity (Rundell and Price 2009).

Evolution of modern birds

The phylogeny and temporal diversification patterns of parrots have to be viewed in the context of the evolution of birds in general. The monophyly of modern birds, Neornithes, is now well established (Cracraft 2001). Within Neornithes, a split between Palaeognathae (ratites and tinamous) and Neognathae is robustly supported by both molecular and morphological data (e.g. Groth and Barrowclough 1999; Hugall et al. 2007; Livezey and Zusi 2007; Hackett et al. 2008; Mayr 2011). The other strongly supported node at the base of the tree of modern birds concerns the split between Neoaves and Galloanseres (land- and waterfowl) within Neognathae (e.g. Cracraft 2001; Livezey and Zusi 2007; Brown et al. 2008; Hackett et al. 2008; Mayr 2011). However, disentangling the relationships among higher taxa, usually ranked as orders within Neoaves, has been difficult and controversial (Sibley and Ahlquist 1990; Brown et al. 2008; Hackett et al. 2008; Pratt et al. 2009; Mayr 2011; Pacheco et al. 2011). The parrots comprise one of several other lineages within Neoaves whose relationships could not be unambiguously resolved so far (e.g. reviewed in Mayr 2011). It has even been argued that the evolutionary relationships between the different groups within Neoaves can probably never be fully clarified due to simultaneous radiation of multiple lineages (Poe and Chubb 2004, see below). Similarly, the time scale of the radiation of modern birds is still a matter of controversy. Based on a strict interpretation of the fossil record, Feduccia (1995, 2003) proposed that birds underwent a mass extinction after the global perturbations at the Cretaceous–Paleogene (K–Pg) boundary 65 Ma. Modern birds should then have evolved from the few surviving lineages in an explosive radiation paralleling that of mammals. Under this scenario, birds and mammals should have inherited practically the entirety of the terrestrial vertebrate adaptive landscape from the now extinct other dinosaur groups and the pterosaurs, and should have rapidly filled the many newly vacant ecological niches (Feduccia 1995, 2003). However, this ‘Tertiary or Paleogene radiation hypothesis’ is in disagreement with the results of several recent molecular phylogenetic studies which have dated the origin of several lineages of modern birds before the K–Pg boundary (Hedges et al. 1996; Cooper and Penny 1997; Pereira and Baker 2006; Slack et al. 2006; Brown et al. 2007; Brown et al. 2008; Pratt et al. 2009; Pacheco et al. 2011). Moreover, recent fossil findings indicate that at least five basal avian splits must have occurred as early as in the Cretaceous (Clarke et al. 2005; Brown et al. 2008). Hence, it is still unclear what was the influence of the mass extinction event at the K-Pg boundary for the evolution of modern birds and which

12 Chapter 1 evolutionary model explains best the supersession by birds (and mammals) of other dinosaurs and pterosaurs (cf. Penny and Phillips 2004). The integration of molecular phylogenetic and distributional data with the geological context led to the hypothesis that the early diversification of birds took place in Gondwana (Cracraft 2001). Gondwana was an ancient supercontinent in the southern hemisphere, which included today’s landmasses of Africa and the Arabian Peninsula, Antarctica, Australia, the Indian subcontinent, Madagascar and South America (Upchurch 2008). The break-up of Gondwana from about 160-30 Ma was supposed to have produced ‘vicariance patterns’ that explain the geographic distribution and phylogenetic relationships of different groups of organisms, including birds (Hedges et al. 1996; Cracraft 2001; Upchurch 2008). Within birds, Palaeognathae have played a crucial role in the arguments surrounding the biogeography of Gondwana, and vicariance in the late Cretaceous caused by continental drift was thought to have primarily shaped their diversification (Cooper et al. 2001; Cracraft 2001; Haddrath and Baker 2001). Parrots are another group of birds whose diversification was thought to have been shaped by the break-up of Gondwana (Cracraft 2001; de Kloet and de Kloet 2005; Wright et al. 2008). However, the causal relationships between these geological and biological events have not been assessed independently and dispersal has been neglected as a potential mechanism to explain current distribution patterns of parrots and other bird groups.

Parrots

With 353 species currently recognized, Psittaciformes represent one of the most species-rich avian groups (next to Passeriformes) traditionally taxonomically ranked as ‘orders’ (Collar 1998; Rowley 1998). They radiated extensively in South America and Australia and also underwent a minor radiation in Africa and tropical Asia (Smith 1975; Forshaw 1989; Collar 1998; Juniper and Parr 1998; Rowley 1998; Cracraft 2001). Parrots have fascinated man for a long time because of their colorful plumages, their companionship and ability to imitate human language. As a consequence, they are among the most familiar and popular pets. On the other hand, parrots have also been persecuted for their negative impact on human food production (Collar 1998; Juniper and Parr 1998; Rowley 1998). However, habitat loss and fragmentation together with collection of live specimens for illegal bird trade are the main source of threats. Moreover, several parrot species consist of small and thus susceptible island populations (Collar 1998). As a consequence, 93 parrot species are currently recognized as being at risk of global extinction (BirdLifeInternational 2000). Parrots therefore include a comparatively large number of threatened species, and 18 species even became extinct between 1600 and 1980, as listed in 1981 (Collar 1998; Juniper and Parr 1998).

Phylogenetic relationships Parrots form a very distinct group of birds characterized by several synapomorphies (Smith 1975), but identifying their ancestral links and the relationships of the whole order has proven difficult (Brown et al. 2008; Hackett et al. 2008; Pratt et al. 2009; Mayr 2011, see also above). The phylogenetic relationships within parrots have also been controversial and the varying interpretations of the relationships and taxonomical arrangements of the different forms and groups have depended on which array of morphological characters was used (Salvadori 1891; Peters 1937; Verheyen

13 Evolutionary diversification and biogeography of parrots

1956; Smith 1975; Homberger 1980; Collar 1998; Juniper and Parr 1998; Mayr 2008, 2010). Several molecular phylogenetic studies published prior to this thesis have shed some light on the evolutionary history of different groups (Leeton et al. 1994; Miyaki et al. 1998; Brown and Toft 1999; Groombridge et al. 2004; Russello and Amato 2004; de Kloet and de Kloet 2005; Tavares et al. 2006). Nevertheless many controversies remained, and a robust and widely accepted phylogenetic hypothesis was lacking. Results stemming from his thesis and papers which appeared in the course of this work (e.g Wright et al. 2008; Mayr 2010; Joseph et al. 2011; White et al. 2011) have, however, considerably improved our understanding of parrot phylogenetics. A synthesis of these is presented in the next chapter.

Temporal and spatial diversification patterns and drivers of diversification Several stem group parrots from the Paleogene have been described from European fossil deposits, but the earliest fossils of crown group members stem from the Miocene (Mayr 2009; Worthy et al. 2011). The allocation of an avian dentary symphysis from the Cretaceous Lance Formation to the parrots by Stidham (1998) has been convincingly called into question (Dyke and Mayr 1999). Molecular phylogenetic studies revealed that the New Zealand parrot taxa Nestor and Strigops form the monophyletic sister group of the remaining parrots (de Kloet and de Kloet 2005). This in turn led to the assumption that the separation of New Zealand from Gondwana 82–85 Ma in the Cretaceous coincided with this early split within modern parrots (de Kloet and de Kloet 2005), thus predating the oldest fossil parrots for millions of years. This bio- and palaeogeographic evidence was then used to calibrate divergence dates between and within several groups of Neotropical parrots (Tavares et al. 2006; Ribas et al. 2007; Ribas et al. 2009), despite the criticism that such calibrations are cases in which geological and biological evidence lack independence (Waters and Craw 2006). Wright et al. (2008) compared divergence estimates, again using such a calibration with the assumption of a minimum age of 50 Ma for the initial split within parrots. This latter estimate corresponded to a hypothesized divergence between modern parrots and fossil forms in the Lower Eocene, accounting for the hypothesis of a radiation of crown group parrots during the Paleogene. Based on this comparison, Wright et al. (2008) considered a Paleogene origin of modern parrots to be less likely than a Cretaceous origin, because a Paleogene scenario would require several trans-oceanic dispersal events to New Zealand, Madagascar and South America from Australia in order to explain current distribution patterns. As parrots are mostly non-migratory birds today, it was argued that their ancient patterns of diversification were not much influenced by dispersal and primarily shaped by vicariance. However, as the relationships, especially of the African taxa, remained mostly unresolved, it has been problematic to assess the likelihood of possible colonization routes and vicariance events in order to draw firm conclusions about the roles of dispersal and vicariance in the evolutionary history of parrots. Vicariance was not only thought to have been an important driver of parrot diversification through continental drift, but was also proposed to have played a role within continents through passive transportation of taxa to high elevations by mountain building, leading to highland-lowland disjunctions (Ribas et al. 2007). However, given the many insular parrot taxa, especially in South East Asia and the Australasian region (Collar 1998), it is hard to imagine that dispersal and colonization of new areas may not have played an important role in the evolutionary history of parrots.

14 Chapter 1

The partly striking ecological differences found among recent parrot species moreover imply that ecological speciation through niche expansion after the colonization of new areas could have been an important driver of parrot diversification. Accordingly, an important role for ecological speciation following ecological opportunities provided by new open areas during phases of climate change, especially in the Miocene, was proposed for Australian and South American parrots (Christidis et al. 1991; Tavares et al. 2006). Particularly the ecologically and morphologically disparate clade of the Neotropical Arini may comprise an (old) adaptive radiation. However, these hypotheses have never been tested based on an independently evaluated temporal and spatial framework. Additionally, the specialization to different food sources may have been an important force in the evolutionary history of some parrot groups. While parrots feed mainly on seeds and fruits, the chiefly Australasian Loriinae (lories) are specialized on a nectarivorous diet. In addition, the swift parrot Lathamus discolor of Australia, the genus Loriculus of Australasia and Indo-Malaysia as well as the genus Brotogeris of the Neotropics are also supposed to depend on nectar as food (Homberger 1980; Güntert 1981; Forshaw 1989; Collar 1998; Rowley 1998). The diet shift to nectarivory may have been a driver of diversification, especially in the species-rich lories, which consist of 53 species today (Collar 1998). Several morphological specializations of the feeding tract to a nectarivorous diet have been described for parrots, which may have been essential for them to effectively feed on nectar and pollen (Güntert and Ziswiler 1972; Güntert 1981; Richardson and Wooller 1986; Gartrell 2000; Gartrell et al. 2000). However, whether these reported differences indeed represent a phenotype-environment correlation has never been tested using a phylogenetic comparative approach. Hence, the influence of diet shifts on the diversification of parrots has remained unclear.

Literature cited

15 Evolutionary diversification and biogeography of parrots

Brown, J. W., R. B. Payne, and D. P. Mindell. 2007. Nuclear DNA does not reconcile 'rocks' and 'clocks' in Neoaves: a comment on Ericson et al. Biology Letters 3:257-259. Brown, J. W., J. S. Rest, J. Garcia-Moreno, M. D. Sorenson, and D. P. Mindell. 2008. Strong mitochondrial DNA support for a Cretaceous origin of modern avian lineages. BMC Biology 6. Burbrink, F. T., and R. A. Pyron. 2009. How does ecological opportunity influence rates of speciation, extinction, and morphological diversification in New World ratsnakes (tribe Lampropeltini)? Evolution 64:934-943. Christidis, L., R. Schodde, D. D. Shaw, and S. F. Maynes. 1991. Relationships among the Australo-Papuan parrots, lorikeets, and cockatoos (Aves, Psittaciformes) - protein evidence. Condor 93:302-317. Clarke, J. A., C. P. Tambussi, J. I. Noriega, G. M. Erickson, and R. A. Ketcham. 2005. Definitive fossil evidence for the extant avian radiation in the Cretaceous. Nature 433:305-308. Collar, N. J. 1998. Family Psittacidae (parrots). Pp. 280-477 in J. del Hoyo, A. Elliott, and J. Sargatal, eds. Handbook of the birds of the world, vol. 4. Lynx Edicions, Barcelona. Cooper, A., C. Lalueza-Fox, S. Anderson, A. Rambaut, J. Austin, and R. Ward. 2001. Complete mitochondrial genome sequences of two extinct moas clarify ratite evolution. Nature 409:704-707. Cooper, A., and D. Penny. 1997. Mass survival of birds across the Cretaceous- Tertiary boundary: molecular evidence. Science 275:1109-1113. Cowie, R. H., and B. S. Holland. 2006. Dispersal is fundamental to biogeography and the evolution of biodiversity on oceanic islands. Journal of Biogeography 33:193-198. Coyne, J. A., and H. A. Orr. 2004. Speciation. Sinauer Associates, Sunderland. Cracraft, J. 2001. Avian evolution, Gondwana biogeography and the Cretaceous- Tertiary mass extinction event. Proceedings of the Royal Society B-Biological Sciences 268:459-469. de Kloet, R. S., and S. R. de Kloet. 2005. The evolution of the spindlin gene in birds: sequence analysis of an intron of the spindlin W and Z gene reveals four major divisions of the Psittaciformes. Molecular Phylogenetics and Evolution 36:706-721. de Queiroz, A. 2005. The resurrection of oceanic dispersal in historical biogeography. Trends in Ecology & Evolution 20:68-73. Drummond, A. J., S. Y. W. Ho, M. J. Phillips, and A. Rambaut. 2006. Relaxed phylogenetics and dating with confidence. PLoS. Biol. 4:699-710. Dyke, G. J., and G. Mayr. 1999. Did parrots exist in the Cretaceous period? Nature 399:317-318. Erwin, D. H. 2001. Lessons from the past: biotic recoveries from mass extinctions. Proceedings of the National Academy of Sciences of the United States of America 98:5399-5403. Erwin, D. H. 2008. Extinction as the loss of evolutionary history. Proceedings of the National Academy of Sciences of the United States of America 105:11520- 11527. Erwin, T. L. 1981. Taxon pulses, vicariance, and dispersal: an evolutionary synthesis illustrated by carabid beetles. Pp. 159-196 in G. Nelson, and D. E. Rosen, eds. Vicariance biogeography: a critique. Columbia University Press, New York.

16 Chapter 1

Feduccia, A. 1995. Explosive evolution in Tertiary birds and mammals. Science 267:637-638. Feduccia, A. 2003. 'Big bang' for Tertiary birds? Trends in Ecology & Evolution 18:172-176. Filardi, C. E., and R. G. Moyle. 2005. Single origin of a pan-Pacific bird group and upstream colonization of Australasia. Nature 438:216-219. Forshaw, J. M. 1989. Parrots of the world. Landsdowne Press, Melbourne. Gartrell, B. D. 2000. The nutritional, morphologic, and physiologic bases of nectarivory in Australian birds. Journal of Avian Medicine and Surgery 14:85- 94. Gartrell, B. D., S. M. Jones, R. N. Brereton, and L. B. Astheimer. 2000. Morphological adaptations to nectarivory of the alimentary tract of the swift parrot Lathamus discolor. Emu 100:274-279. Gotelli, N. J., M. J. Anderson, H. T. Arita, A. Chao, R. K. Colwell, S. R. Connolly, D. J. Currie, R. R. Dunn, G. R. Graves, J. L. Green, J.-A. Grytnes, Y.-H. Jiang, W. Jetz, S. Kathleen Lyons, C. M. McCain, A. E. Magurran, C. Rahbek, T. F. L. V. B. Rangel, J. Soberón, C. O. Webb, and M. R. Willig. 2009. Patterns and causes of species richness: a general simulation model for macroecology. Ecol. Lett. 12:873-886. Grant, P. R., and B. R. Grant. 2008. How and why species multiply: the radiation of Darwin’s finches. Princeton University Press, Princeton. Groombridge, J. J., C. G. Jones, R. A. Nichols, M. Carlton, and M. W. Bruford. 2004. Molecular phylogeny and morphological change in the Psittacula parakeets. Molecular Phylogenetics and Evolution 31:96-108. Groth, J. G., and G. F. Barrowclough. 1999. Basal divergences in birds and the phylogenetic utility of the nuclear RAG-1 gene. Molecular Phylogenetics and Evolution 12:115-123. Güntert, M. 1981. Morphologische Untersuchungen zur adaptiven Radiation des Verdauungstraktes bei Papageien (Psittaci). Zool. Jahrb. Abt. Anat. Ontog. Tiere 106:471-526. Güntert, M., and V. Ziswiler. 1972. Konvergenzen in der Struktur von Zunge und Verdauungstrakt nektarfressender Papageien. Rev. Suisse Zool. 79:1017-1026. Hackett, S. J., R. T. Kimball, S. Reddy, R. C. K. Bowie, E. L. Braun, M. J. Braun, J. L. Chojnowski, W. A. Cox, K. L. Han, J. Harshman, C. J. Huddleston, B. D. Marks, K. J. Miglia, W. S. Moore, F. H. Sheldon, D. W. Steadman, C. C. Witt, and T. Yuri. 2008. A phylogenomic study of birds reveals their evolutionary history. Science 320:1763-1768. Haddrath, O., and A. J. Baker. 2001. Complete mitochondrial DNA genome sequences of extinct birds: ratite phylogenetics and the vicariance biogeography hypothesis. Proceedings of the Royal Society B-Biological Sciences 268:939-945. Halas, D., D. Zamparo, and D. R. Brooks. 2005. A historical biogeographical protocol for studying biotic diversification by taxon pulses. Journal of Biogeography 32:249-260. Hanski, I. 1999. Metapopulation ecology. Oxford University Press, Oxford. Harmon, L. J., J. B. Losos, T. J. Davies, R. G. Gillespie, J. L. Gittleman, W. B. Jennings, K. H. Kozak, M. A. McPeek, F. Moreno-Roark, T. J. Near, A. Purvis, R. E. Ricklefs, D. Schluter, J. A. Schulte, O. Seehausen, B. L. Sidlauskas, O. Torres-Carvajal, J. T. Weir, and A. O. Mooers. 2010. Early

17 Evolutionary diversification and biogeography of parrots

bursts of body size and shape evolution are rare in comparative data. Evolution 64:2385-2396. Harmon, L. J., J. A. Schulte, A. Larson, and J. B. Losos. 2003. Tempo and mode of evolutionary radiation in iguanian lizards. Science 301:961-964. Harshman, J., E. L. Braun, M. J. Braun, C. J. Huddleston, R. C. K. Bowie, J. L. Chojnowski, S. J. Hackett, K. L. Han, R. T. Kimball, B. D. Marks, K. J. Miglia, W. S. Moore, S. Reddy, F. H. Sheldon, D. W. Steadman, S. J. Steppan, C. C. Witt, and T. Yuri. 2008. Phylogenomic evidence for multiple losses of flight in ratite birds. Proceedings of the National Academy of Sciences of the United States of America 105:13462-13467. Heaney, L. R. 2007. Is a new paradigm emerging for oceanic island biogeography? Journal of Biogeography 34:753-757. Heard, S. B., and D. L. Hauser. 1995. Key evolutionary innovations and their ecological mechanisms. Historical Biol 10:151-173. Hedges, S. B., P. H. Parker, C. G. Sibley, and S. Kumar. 1996. Continental breakup and the ordinal diversification of birds and mammals. Nature 381:226-229. Helbig, A. J., A. G. Knox, D. T. Parkin, G. Sangster, and M. Collinson. 2002. Guidelines for assigning species rank. Ibis 144:518-525. Ho, S. Y. W., and M. J. Phillips. 2009. Accounting for calibration uncertainty in phylogenetic estimation of evolutionary divergence times. Systematic Biology 58:367-380. Homberger, D. G. 1980. Funktionell- morphologische Untersuchungen zur Radiation der Ernährungs- und Trinkmethoden der Papageien (Psittaci). Bonn. Zool. Monogr. 13:1-92. Hugall, A. F., R. Foster, and M. S. Y. Lee. 2007. Calibration choice, rate smoothing, and the pattern of tetrapod diversification according to the long nuclear gene RAG-1. Systematic Biology 56:543-563. Hunter, J. P. 1998. Key innovations and the ecology of macroevolution. Trends in Ecology & Evolution 13:31-36. Jablonski, D. 1991. Extinctions - a paleontological perspecitive. Science 253:754-757. Jablonski, D. 2005. Mass extinctions and macroevolution. Paleobiology 31:192-210. Jonsson, K. A., R. C. K. Bowie, R. G. Moyle, L. Christidis, J. A. Norman, B. W. Benz, and J. Fjeldsa. 2010. Historical biogeography of an Indo-Pacific passerine bird family (Pachycephalidae): different colonization patterns in the Indonesian and Melanesian archipelagos. Journal of Biogeography 37:245- 257. Joseph, L., A. Toon, E. E. Schirtzinger, and T. F. Wright. 2011. Molecular systematics of two enigmatic genera Psittacella and Pezoporus illuminate the ecological radiation of Australo-Papuan parrots (Aves: Psittaciformes). Molecular Phylogenetics and Evolution 59:675-684. Juniper, T., and M. Parr. 1998. Parrots: a guide to the parrots of the world. Pica Press, Sussex. Kassen, R. 2009. Toward a general theory of adaptive radiation insights from microbial experimental evolution. Pp. 3-22. Year in Evolutionary Biology 2009. Blackwell Publishing, Oxford. Kozak, K. H., A. A. Larson, R. M. Bonett, and L. J. Harmon. 2005. Phylogenetic analysis of ecomorphological divergence, community structure, and diversification rates in dusky salamanders (Plethodontidae: Desmognathus). Evolution 59:2000-2016.

18 Chapter 1

Leeton, P. R. J., L. Christidis, M. Westerman, and W. E. Boles. 1994. Molecular phylogenetic affinities of the night-parrot (Geopsittacus occidentalis) and the ground-parrot (Pezoporus wallicus). Auk 111:833-843. Livezey, B. C., and R. L. Zusi. 2007. Higher-order phylogeny of modern birds (Theropoda, Aves: Neornithes) based on comparative anatomy. II. Analysis and discussion. Zoological Journal of the Linnean Society 149:1-95. Lomolino, M. V., B. R. Riddle, R. J. Whittaker, and J. H. Brown. 2010. Biogeography, fourth edition. Sinauer Associates, Inc., Sunderland. Losos, J. B. 2009. Lizards in an evolutionary tree: ecology and adaptive radiation of anoles. University of California Press, Berkeley. Losos, J. B. 2010. Adaptive radiation, ecological opportunity, and evolutionary determinism. American Naturalist 175:623-639. Losos, J. B., and K. DeQueiroz. 1997. Evolutionary consequences of ecological release in Caribbean Anolis lizards. Biological Journal of the Linnean Society 61:459-483. Losos, J. B., and R. E. Ricklefs. 2009. Adaptation and diversification on islands. Nature 457:830-836. Lovette, I. J., E. Bermingham, and R. E. Ricklefs. 2002. Clade-specific morphological diversification and adaptive radiation in Hawaiian songbirds. Proceedings of the Royal Society of London Series B-Biological Sciences 269:37-42. Lynch, V. J. 2009. Live-birth in vipers (Viperidae) is a key innovation and adaptation to global cooling during the Cenozoic. Evolution 63:2457-2465. MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton. Mayr, E. 1942. Systematics and the origin of species. Columbia University Press, New York. Mayr, G. 2008. The phylogenetic affinities of the parrot taxa Agapornis, Loriculus and Melopsittacus (Aves: Psittaciformes): hypotarsal morphology supports the results of molecular analyses. Emu 108:23-27. Mayr, G. 2009. Paleogene fossil birds. Springer-Verlag, Berlin Heidelberg. Mayr, G. 2010. Parrot interrelationships - morphology and the new molecular phylogenies. Emu 110:348-357. Mayr, G. 2011. Metaves, Mirandornithes, Strisores and other novelties - a critical review of the higher-level phylogeny of neornithine birds. Journal of Zoological Systematics and Evolutionary Research 49:58-76. McGlone, M. S. 2005. Goodbye Gondwana. Journal of Biogeography 32:739-740. Mittelbach, G. G., D. W. Schemske, H. V. Cornell, A. P. Allen, J. M. Brown, M. B. Bush, S. P. Harrison, A. H. Hurlbert, N. Knowlton, H. A. Lessios, C. M. McCain, A. R. McCune, L. A. McDade, M. A. McPeek, T. J. Near, T. D. Price, R. E. Ricklefs, K. Roy, D. F. Sax, D. Schluter, J. M. Sobel, and M. Turelli. 2007. Evolution and the latitudinal diversity gradient: speciation, extinction and biogeography. Ecol. Lett. 10:315-331. Miyaki, C. Y., S. R. Matioli, T. Burke, and A. Wajntal. 1998. Parrot evolution and paleogeographical events: mitochondrial DNA evidence. Molecular Biology and Evolution 15:544-551. Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca, and J. Kent. 2000. Biodiversity hotspots for conservation priorities. Nature 403:853-858. Newton, I. 2003. The speciation and biogeography of birds. Academic Press, London. Nishihara, H., S. Maruyama, and N. Okada. 2009. Retroposon analysis and recent geological data suggest near-simultaneous divergence of the three superorders

19 Evolutionary diversification and biogeography of parrots

of mammals. Proceedings of the National Academy of Sciences of the United States of America 106:5235-5240. Nosil, P., and T. E. Reimchen. 2005. Ecological opportunity and levels of morphological variance within freshwater stickleback populations. Biological Journal of the Linnean Society 86:297-308. Pacheco, M. A., F. U. Battistuzzi, M. Lentino, R. Aguilar, S. Kumar, and A. A. Escalante. 2011. Evolution of modern birds revealed by mitogenomics: timing the radiation and origin of major orders. Molecular Biology and Evolution 28:1927-1942. Penny, D., and M. J. Phillips. 2004. The rise of birds and mammals: are microevolutionary processes sufficient for macroevolution. Trends in Ecology & Evolution 19:516-522. Pereira, S. L., and A. J. Baker. 2006. A mitogenomic timescale for birds detects variable phylogenetic rates of molecular evolution and refutes the standard molecular clock. Molecular Biology and Evolution 23:1731-1740. Pereira, S. L., K. P. Johnson, D. H. Clayton, and A. J. Baker. 2007. Mitochondrial and nuclear DNA sequences support a Cretaceous origin of Columbiformes and a dispersal-driven radiation in the Paleogene. Systematic Biology 56:656-672. Peters, J. L. 1937. Checklist of the birds of the world. 3. Harvard University Press, Cambridge. Phillimore, A. B., and T. D. Price. 2008. Density-dependent cladogenesis in birds. PLoS. Biol. 6:483-489. Poe, S., and A. L. Chubb. 2004. Birds in a bush: five genes indicate explosive evolution of avian orders. Evolution 58:404-415. Pramuk, J. B., T. Robertson, J. W. Sites, and B. P. Noonan. 2008. Around the world in 10 million years: biogeography of the nearly cosmopolitan true toads (Anura: Bufonidae). Global Ecology and Biogeography 17:72-83. Pratt, H. D. 2005. The Hawaiian honecreepers. Oxford University Press, Oxford. Pratt, R. C., G. C. Gibb, M. Morgan-Richards, M. J. Phillips, M. D. Hendy, and D. Penny. 2009. Toward resolving deep Neoaves phylogeny: data, signal enhancement, and priors. Molecular Biology and Evolution 26:313-326. Price, S. A., P. C. Wainwright, D. R. Bellwood, E. Kazancioglu, D. C. Collar, and T. J. Near. 2010. Functional innovations and morphological diversification in parrotfish. Evolution 64:3057-3068. Price, T. 2008. Speciation in birds. Roberts and Company, Greenwood Village, CO. Purvis, A., K. E. Jones, and G. M. Mace. 2000. Extinction. Bioessays 22:1123-1133. Rabosky, D. L. 2009. Ecological limits and diversification rate: alternative paradigms to explain the variation in species richness among clades and regions. Ecol. Lett. 12:735-743. Rabosky, D. L., and I. J. Lovette. 2008a. Density-dependent diversification in North American wood warblers. Proceedings of the Royal Society B-Biological Sciences 275:2363-2371. Rabosky, D. L., and I. J. Lovette. 2008b. Explosive evolutionary radiations: decreasing speciation or increasing extinction through time? Evolution 62:1866-1875. Raxworthy, C. J., M. R. J. Forstner, and R. A. Nussbaum. 2002. Chameleon radiation by oceanic dispersal. Nature 415:784-787. Ribas, C. C., C. Y. Miyaki, and J. Cracraft. 2009. Phylogenetic relationships, diversification and biogeography in Neotropical Brotogeris parakeets. Journal of Biogeography 36:1712-1729.

20 Chapter 1

Ribas, C. C., R. G. Moyle, C. Y. Miyaki, and J. Cracraft. 2007. The assembly of montane biotas: linking Andean tectonics and climatic oscillations to independent regimes of diversification in Pionus parrots. Proceedings of the Royal Society B-Biological Sciences 274:2399-2408. Richardson, K. C., and R. D. Wooller. 1986. The structures of the gastrointestinal tracts of honeyeaters and other small birds in relation to their diets. Aust. J. Zool. 34:119-124. Ricklefs, R. E. 1987. Community diversity: relative roles of local and regional processes. Science 235:167-171. Ricklefs, R. E., J. B. Losos, and T. M. Townsend. 2007. Evolutionary diversification of clades of squamate reptiles. J. Evol. Biol. 20:1751-1762. Rowley, I. 1998. Family Cacatuidae (cockatoos). Pp. 246-279 in J. Del Hoyo, A. Elliott, and J. Sargatal, eds. Handbook of the birds of the world. Lynx Edicions, Barcelona. Ruber, L., and R. Zardoya. 2005. Rapid cladogenesis in marine fishes revisited. Evolution 59:1119-1127. Rundell, R. J., and T. D. Price. 2009. Adaptive radiation, nonadaptive radiation, ecological speciation and nonecological speciation. Trends in Ecology & Evolution 24:394-399. Rundle, H. D., and P. Nosil. 2005. Ecological speciation. Ecol. Lett. 8:336-352. Russello, M. A., and G. Amato. 2004. A molecular phylogeny of Amazona: implications for Neotropical parrot biogeography, taxonomy, and conservation. Molecular Phylogenetics and Evolution 30:421-437. Salvadori, T. 1891. Catalogue of the Psittaci, or parrots, in the collection of the British Museum, Vol. 20. Br. Mus. (Nat. Hist.), London. Schluter, D. 2000. The ecology of adaptive radiation. Oxford University Press, Oxford. Schluter, D. 2001. Ecology and the origin of species. Trends in Ecology & Evolution 16:372-380. Schluter, D. 2009. Evidence for ecological speciation and its alternative. Science 323:737-741. Seehausen, O. 2006. African cichlid fish: a model system in adaptive radiation research. Proceedings of the Royal Society B-Biological Sciences 273:1987- 1998. Sibley, C. G., and J. E. Ahlquist. 1990. Phylogeny and classification of birds: a study in molecular evolution. CT: Yale University Press, New Haven. Simpson, G. G. 1949. Tempo and mode in evolution. Columbia University Press, New York. Simpson, G. G. 1953. The major features of evolution. Columbia University Press, New York. Slack, K. E., C. M. Jones, T. Ando, G. L. Harrison, R. E. Fordyce, U. Arnason, and D. Penny. 2006. Early penguin fossils, plus mitochondrial genomes, calibrate avian evolution. Molecular Biology and Evolution 23:1144-1155. Slater, G. J., S. A. Price, F. Santini, and M. E. Alfaro. 2010. Diversity versus disparity and the radiation of modern cetaceans. Proceedings of the Royal Society B- Biological Sciences 277:3097-3104. Smith, G. A. 1975. Systematics of parrots. Ibis 117:18-68. Soltis, D. E. 2007. Reconstructing the three of life: taxonomy and systematics of species rich taxa. Evolution 61:3007-3011. Stidham, T. A. 1998. A lower jaw from a Cretaceous parrot. Nature 396:29-30.

21 Evolutionary diversification and biogeography of parrots

Tavares, E. S., A. J. Baker, S. L. Pereira, and C. Y. Miyaki. 2006. Phylogenetic relationships and historical biogeography of Neotropical parrots (Psittaciformes: Psittacidae: Arini) inferred from mitochondrial and nuclear DNA sequences. Systematic Biology 55:454-470. Thomas, G. H., C. D. L. Orme, R. G. Davies, V. A. Olson, P. M. Bennett, K. J. Gaston, I. P. F. Owens, and T. M. Blackburn. 2008. Regional variation in the historical components of global avian species richness. Global Ecology and Biogeography 17:340-351. Trenel, P., M. H. G. Gustafsson, W. J. Baker, C. B. Asmussen-Lange, J. Dransfield, and F. Borchsenius. 2007. Mid-Tertiary dispersal, not Gondwanan vicariance explains distribution patterns in the wax palm subfamily (Ceroxyloidea: Arecaceae). Molecular Phylogenetics and Evolution 45:272-288. Upchurch, P. 2008. Gondwanan break-up: legacies of a lost world? Trends in Ecology & Evolution 23:229-236. Vences, M., D. R. Vieites, F. Glaw, H. Brinkmann, J. Kosuch, M. Veith, and A. Meyer. 2003. Multiple overseas dispersal in amphibians. Proceedings of the Royal Society of London Series B-Biological Sciences 270:2435-2442. Verheyen, R. 1956. Analyse du potentiel morphologique et projet d'une nouvelle classification des Psittaciformes. Bull. Inst. R. Sci. nat. Belg. 32:1-54. Vermeij, G. J. 1995. Economics, volcanos, and phanerozoic revolutions. Paleobiology 21:125-152. Voelker, G., S. Rohwer, D. C. Outlaw, and R. C. K. Bowie. 2009. Repeated trans- Atlantic dispersal catalysed a global songbird radiation. Global Ecology and Biogeography 18:41-49. Waters, J. M., and D. Craw. 2006. Goodbye Gondwana? New Zealand biogeography, geology, and the problem of circularity. Systematic Biology 55:351-356. Weir, J. T. 2006. Divergent timing and patterns of species accumulation in lowland and highland Neotropical birds. Evolution 60:842-855. White, N. E., M. J. Phillips, M. T. P. Gilbert, A. Alfaro-Nunez, E. Willerslev, P. R. Mawson, P. B. S. Spencer, and M. Bunce. 2011. The evolutionary history of cockatoos (Aves: Psittaciformes: Cacatuidae). Molecular Phylogenetics and Evolution 59:615-622. Wiens, J. J. 2011. The causes of species richness patterns across space, time, and clades and the role of “ecological limits". Q. Rev. Biol. 86:75-96. Worthy, T. H., A. J. D. Tennyson, and R. P. Scofield. 2011. An early Miocene diversity of parrots (Aves, Strigopidae, Nestorinae) from New Zealand. Journal of Vertebrate Paleontology 31:1-15. Wright, T. F., E. E. Schirtzinger, T. Matsumoto, J. R. Eberhard, G. R. Graves, J. J. Sanchez, S. Capelli, H. Mueller, J. Scharpegge, G. K. Chambers, and R. C. Fleischer. 2008. A multilocus molecular phylogeny of the parrots (Psittaciformes): support for a Gondwanan origin during the Cretaceous. Molecular Biology and Evolution 25:2141-2156. Yoder, A. D., M. M. Burns, S. Zehr, T. Delefosse, G. Veron, S. M. Goodman, and J. J. Flynn. 2003. Single origin of Malagasy Carnivora from an African ancestor. Nature 421:734-737. Yoder, A. D., and M. D. Nowak. 2006. Has vicariance or dispersal been the predominant biogeographic force in Madagascar? Only time will tell. Annual Review of Ecology Evolution and Systematics 37:405-431. Yoder, J. B., E. Clancey, S. Des Roches, J. M. Eastman, L. Gentry, W. Godsoe, T. J. Hagey, D. Jochimsen, B. P. Oswald, J. Robertson, B. A. J. Sarver, J. J.

22 Chapter 1

Schenk, S. F. Spear, and L. J. Harmon. 2010. Ecological opportunity and the origin of adaptive radiations. J. Evol. Biol. 23:1581-1596.

Chapter 2

SUMMARY OF CHAPTERS

Chapter 3

The diversification of parrots (Psittaciformes) has been assumed to be mainly influenced by vicariance, the splitting of evolutionary lineages following a separation of their range by geological events (de Kloet and de Kloet 2005; Wright et al. 2008). However, as the relationships especially of the African parrot taxa have remained mostly unresolved so far, it has been difficult to draw firm conclusions about the roles of dispersal and vicariance in the evolutionary history of parrots. This chapter reports on our analysis of the broadest taxon sampling of old world parrots ever using nucleotide sequence data. The sample consisted of 60 old world species and one additional Neotropical taxon and was based on 3219 bp (base pairs) stemming from the three nuclear genes c-mos, RAG-1 and Zenk. All three markers are single copy nuclear exons and have been widely used in phylogenetic inferences (e.g. Groth and Barrowclough 1999; Lovette and Bermingham 2000; Chubb 2004b, a; Overton and Rhoads 2004; Tavares et al. 2006; Pereira et al. 2007; Treplin et al. 2008). Our phylogenetic analyses based on model-based approaches (Bayesian inference, maximum likelihood) and maximum parsimony provided a well-resolved and congruent phylogenetic hypothesis, which was in some cases in conflict with traditional views of the evolution of Psittaciformes. Agapornis of Africa and Madagascar was found to be the sister group to Loriculus of Australasia and Indo- Malaysia, and together they clustered with the Australasian Loriinae, Melopsittacus and Cyclopsittacini. Poicephalus and Psittacus from mainland Africa formed the sister group of the Neotropical Arini, whereas Coracopsis from Madagascar and adjacent islands may be the closest relative of Psittrichas from New Guinea. Our ancestral area reconstruction based on a maximum likelihood approach using our phylogenetic hypothesis revealed that several major trans-oceanic dispersal events must be considered to explain these biogeographic relationships. The current distribution patterns of parrots seem to be best explained by independent colonizations of the African continent via trans-oceanic dispersal from Australasia and Antarctica in the Paleogene, following what may have been vicariance events in the late Cretaceous and/or early Paleogene. Trans-oceanic dispersal thus played a more important role than previously thought and was the prerequisite for range expansion into new continents, as has been shown in other bird groups (e.g. Shapiro et al. 2002; Fuchs et al. 2006; Fuchs et al. 2007; Pereira et al. 2007; Jonsson et al. 2008; Voelker et al. 2009). A taxon pulse model of diversification (Erwin 1981; Halas et al. 2005) can best explain the global spread and radiation of parrots, because initial vicariance events within the center of parrot diversity in Australasia were followed by episodes of dispersal to and in-situ radiation on all other tropical continents. Moreover, both Africa and Indo-Malaysia were colonized by multiple lineages at different times, resulting in a biogeographically reticulated history

Chapter 4

The initial split within crown group parrots separated the monophyletic clade of the New Zealand taxa Nestor and Strigops from the remaining extant species (de Kloet

27 Evolutionary diversification and biogeography of parrots and de Kloet 2005; Wright et al. 2008; Pratt et al. 2009; Schweizer et al. 2010). This divergence event was considered to coincide with the separation of New Zealand from Gondwana 82–85 million years ago (Ma) assuming that the diversification of parrots was mainly shaped by vicariance (de Kloet and de Kloet 2005; Tavares et al. 2006; Wright et al. 2008). However, as demonstrated in chapter 3 (Schweizer et al. 2010), the distribution patterns of several extant parrot groups cannot be explained without invoking trans-oceanic dispersal, thus challenging this assumption. The aim of the present chapter was therefore to generate a temporal and spatial framework for the diversification of parrots using external avian fossils as calibration points in order to evaluate the relative importance of the influences of past climate change, plate tectonics and ecological opportunity. Phylogenetic relationships were investigated using partial sequences of the nuclear genes c-mos, RAG-1 and Zenk as in chapter 3 (Schweizer et al. 2010). Compared to the latter study, we added important new taxa of the genera Poicephalus, Prioniturus and Psephotus, the Neotropical taxon Arini, the Philippine endemic Bolbopsittacus lunulatus, the Australasian Northiella haematogaster and the genus Psittacella. To be able to date the phylogeny of parrots with external fossils as calibration points, we added sequences of 21 avian species outside the parrots. Divergence dates and confidence intervals were estimated using a Bayesian relaxed molecular clock approach (Drummond et al. 2006). Biogeographic patterns were evaluated using a model-based approach taking temporal connectivity between areas into account (Ree et al. 2005; Ree and Smith 2008). We moreover tested if net diversification of parrots remained constant over time by a comparison of lineage-through-time plots of our data with different null models of constant diversification rates through time. The comparative method MEDUSA (Alfaro et al. 2009) was used to infer whether some parrot groups were more species-rich than expected given their age. We found that crown group diversification of parrots started only about 58 Ma in the Paleogene, significantly later than previously thought. The Australasian lories and possibly also the Neotropical Arini were found to be unexpectedly species-rich. Diversification rates were probably increased around the Eocene/Oligocene boundary and in the middle Miocene, during two periods of major global climatic aberrations characterized by global cooling. Our results demonstrated that the diversification of parrots was shaped by climatic and geological events as well as by key innovations. Initial vicariance events caused by continental break-up were followed by trans-oceanic dispersal and local radiations as has been hypothesized in chapter 3 (Schweizer et al. 2010). Habitat shifts caused by climate change and mountain orogenesis may have catalyzed the diversification of parrots by providing new ecological opportunities and challenges as well as by leading to isolation caused by habitat fragmentation. Especially the spread of open dry habitats on all southern continents may have been an important driver of diversification. The lories comprise the only highly nectarivorous parrot clade and their diet shift, associated with morphological innovations, may have acted as an evolutionary key innovation (Heard and Hauser 1995; Hunter 1998), allowing them to explore underutilized niches and promoting their diversification.

Chapter 5

Specialization to a nectarivorous diet resulted in several radiations of birds including different groups of parrots (Collar 1998; Fleming and Muchhala 2008). One of them,

28 Chapter 2 the Australasian lories, was shown in chapter 4 (Schweizer et al. in press) to be unexpectedly species-rich given their age compared to the remaining parrot lineages. Their shift to nectarivory may thus have created an ecological opportunity and promoted species proliferation. Besides the lories, the swift parrot Lathamus discolor of Australia, the genus Loriculus of Australasia and Indo-Malaysia as well as the genus Brotogeris of the Neotropics are all supposed to depend on nectar as food (Homberger 1980; Güntert 1981; Forshaw 1989; Collar 1998). Several morphological specializations of the feeding tract to a nectarivorous diet have been described in these parrots (Churchill and Christensen 1970; Güntert and Ziswiler 1972; Güntert 1981; Richardson and Wooller 1990; Gartrell 2000; Gartrell et al. 2000; Gartrell and Jones 2001). However, these reported adaptations were mostly based on analyses of a small taxon sampling and have never been statistically assessed implementing a correction for phylogenetic non-independence. Consequently, they may reflect phylogenetic contingency rather than a high level of adaptation (e.g. Felsenstein 1985; Freckleton 2009). In the research reported in this chapter, we thus applied a phylogenetic comparative approach to test if morphological variation in nectarivorous parrots indeed reflects dietary adaptations. Our analyses were based on a broad taxon sampling of 78 parrot species and 15 continuous characters of the digestive tract measured from altogether 354 individuals. To control for phylogenetic non-independence among species, we used a phylogenetic reconstruction based on partial sequences of the three nuclear genes c-mos, RAG-1 and Zenk, as described in chapters 3 and 4 (Schweizer et al. 2010; Schweizer et al. in press), and additionally of the mitochondrial gene ND2 (NADH2). The final alignment was 4254 bp in length. We investigated potential dietary adaptations using generalized least squares (PGLS) ANCOVA (Grafen 1989; Garland and Ives 2000), implementing diet as a covariate in the model. Moreover, we tested if the diet shift to nectarivory in parrots influenced the rate of morphological evolution based on a Brownian motion model correcting for phylogenetic and time related effects (O'Meara et al. 2006; Price et al. 2011). Additionally, we developed a new statistical approach to infer if a subset of species in a phylogenetic tree shows a change in the rate of trait evolution at the base of their radiation. Our analyses showed that the nectarivorous parrot species differed from the remaining parrots in several traits of their digestive tract. These trait-changes may have allowed them to effectively feed on nectar and may indicate a phenotype-environment correlation and parallel or convergent evolution by natural selection under the same or similar ecological conditions (cf. Losos et al. 1998; Schluter et al. 2004; Colosimo et al. 2005). The ecological expansion to nectarivory did not lead to a change in the rate of morphological evolution in all nectarivorous species. But it was associated with significantly hastened morphological evolution at the base of the radiation of the lories. Hence, the diet shift and the corresponding morphological adaptations can be considered an evolutionary key innovation in this group (cf. e.g. Heard and Hauser 1995; Vermeij 1995; Hunter 1998; Yoder et al. 2010). It promoted significant nonadaptive lineage diversification through allopatric partitioning of the same new niche. However, it did not result in further ecological specializations. Interestingly, the diet shift to nectarivory in parrots only promoted diversification in the lories.

29 Evolutionary diversification and biogeography of parrots

Chapter 6

The Australian continent was predominantly covered by rainforest in the past, but aridification, especially from the mid-Miocene onwards, led to a fragmentation of mesic biomes and expansion of arid habitats (Jacobs et al. 1999; Martin 2006; Byrne et al. 2008; Byrne et al. in press). This had a major influence on the diversification of Australian terrestrial organisms and the general direction of their radiation is supposed to have been from rainforests into drier habitats (Schodde 2006; Byrne 2008). This has also been hypothesized for the platycercine parrots (Platycercini) (Christidis et al. 1991). Today, they occur in a wide variety of habitats in Australia and also occupy islands in Melanesia and in New Zealand (Collar 1998). The beginning of the radiation of the platycercine genera which comprise species adapted to arid habitats was shown in chapter 4 to coincide with the onset of the aridification of Australia (Schweizer et al. in press). In this chapter’s study, we aimed at providing a picture of the radiation of Platycercini based on a spatial and temporal framework. We tested the hypothesis that mesic biomes should map as ancestral in their phylogeny and examined colonization patterns in Melanesia and New Zealand. Phylogenetic relationships were investigated using partial sequences of the nuclear genes c-mos, RAG-1 and Zenk, as in chapters 3 and 4 (Schweizer et al. 2010; Schweizer et al. in press) and additionally of the two mitochondrial markers ND2 (NADH2) and Cytb (Cytochrome b) with the concatenated alignment consisting of 5121 bp. Temporal patterns of the diversification of the platycercine parrots were inferred using a Bayesian relaxed molecular clock approach (Drummond et al. 2006) based on secondary calibration points adopted from the results of chapter 4 (Schweizer et al. in press). Spatial diversification patterns were evaluated on the basis of model-based biogeographic reconstructions (Ree et al. 2005; Ree and Smith 2008). The phylogenetic relationships that were revealed partially contradicted traditional taxonomic treatment which was mainly based on non-molecular traits. The mesic biome was confirmed as the ancestral area of Platycercini and arid environments were colonized from there. Diversification within two of the three robustly supported platycercine clades started from the middle Miocene onward, coinciding with the beginning of severe aridification in Australia. The associated habitat shifts may thus have catalyzed the diversification of Platycercini through colonization of and adaptation to the new open habitats. Moreover, fragmentation of forest habitats led to vicariance speciation within non-arid Australian platycercine parrots as predicted. Colonization of Pacific islands proceeded from Australia over Melanesia to New Zealand. Small oceanic islands thus contributed as stepping zones to the build-up of the biota of the larger landmass of New Zealand. Such a pattern has been recently revealed in several organismic groups, contradicting traditional biogeographic paradigms which considered oceanic islands as evolutionary dead-ends (Filardi and Moyle 2005; Bellemain and Ricklefs 2008; Jonsson et al. 2010a).

Chapter 7

Parrots have colonized the Indo-Malayan region several times independently from Australasia, mainly after these two regions came into close contact around 25 Ma, as shown in chapters 3 and 4 (Schweizer et al. 2010; Schweizer et al. in press). However, the exact pattern of colonization of Sundaland, the Philippines and Wallacea, a geologically dynamic region with a complex history of land and sea, could not be

30 Chapter 2 inferred unambiguously. The aim of the study reported in this chapter was thus to reconstruct dispersal and colonization patterns of the parrot genus Prioniturus across and within Wallacea and the Philippines based on a molecular phylogeny. The genus occurs in the oceanic Philippines, Palawan and Wallacea (Collar 1998), and the described taxa have been variously treated in different assemblages, and various numbers of species have been recognized (Salomonsen 1953; White and Bruce 1986; Collar 1998; Dickinson 2003; Forshaw 2006). We consequently further wanted to test whether traditional groupings would be confirmed by our molecular phylogenetic framework. We included all but one recognized Prioniturus taxa in our analyses, with the majority of samples stemming from toe pads of museum skins. We generated partial sequences of the two mitochondrial genes ND2 (NADH2) and Cytb (Cytochrome b) with a final alignment consisting of 1287 bp and investigated phylogenetic relationships using Bayesian inference and maximum likelihood analyses. We revealed three well supported clades within Prioniturus which were partially congruent with proposed groupings based on morphological similarities. The different subspecies of the traditionally recognized species P. discurus were found not to be monophyletic and we thus proposed to treat P. mindorensis comb. nov. as a separate species. Further studies using additional data and specimens are necessary to clarify whether the genetically and morphologically differentiated taxa P. d. discurus and P. discurus whiteheadi also merit species status. Moreover, we corroborated that P. montanus and P. waterstradti are specifically distinct. Hence, species diversity was found to be underestimated and, depending on the taxonomic source used, it was found that one to two or (probably) even three additional species should be recognized in the Philippines. Our biogeographic reconstruction based on a maximum likelihood approach revealed that five major dispersal events between the different island groups need to be invoked to explain the current distribution patterns of Prioniturus. While the species assemblage of Prioniturus on Sulawesi/Wallacea was found not to be monophyletic, the Philippine taxa were the result of a single colonization event followed by in-situ speciation and subsequent dispersal to Sulawesi, Palawan and the Sulu archipelago.

Chapter 8

Avian diversity is higher in South America than in all other tropical regions combined (e.g. Weir 2006; Price 2008). Accordingly, the Neotropical parrots (Arini) were found to be likely more diverse compared to the remaining parrot groups, given their age in the analyses of chapter 4 (Schweizer et al. in press). Arini colonized South America most probably from Antarctica after they split in the late Eocene/early Oligocene from a clade consisting of African parrots today (Psittacini). The newly colonized Neotropics may have presented under-utilized adaptive zones that provided several ecological opportunities facilitating diversification, as hypothesized in chapters 3 and 4 (Schweizer et al. 2010; Schweizer et al. in press). However, although probably influenced by sampling artefacts, there seems to have been a time lag between the arrival of parrots in South America at around 35 Ma and the beginning of their diversification in the late Oligocene, as shown in chapter 4 (Schweizer et al. in press). Andean uplift and climate change in combination with habitat alterations may have later driven diversification in parrots from the Miocene onwards, as in other organismic groups (e.g. Weir 2006; Thomas et al. 2008; Hoorn et al. 2010). Using

31 Evolutionary diversification and biogeography of parrots molecular and morphological data, we thus tested whether the Neotropical parrots diversified in an adaptive radiation after their colonization of South America or if their diversification occurred later and was influenced by historical processes from the Miocene onwards. We generated a time calibrated phylogeny of more than 66% of all Arini species, implementing a Bayesian relaxed molecular clock approach (Drummond et al. 2006) based on a secondary calibration point adopted from the results of chapter 4 (Schweizer et al. in press). We therefore used partial sequences of the nuclear exon RAG-1 and the mitochondrial genes COI (cytochrome oxidase subunit I), Cytb and ND2 extracted from GenBank with a final alignment of 5506 bp. We tested if diversification remained constant over time and inferred whether some groups within Arini were more species-rich than expected given their age through the use of different methods (Pybus and Harvey 2000; Rabosky 2006b; Alfaro et al. 2009; Brock et al. 2011). We used overall size (total and tarsus length) and relative bill length as traits which are likely to represent ecological differences among different parrot species (Collar 1998). Morphological evolution was investigated by calculating mean subclade disparity through time (Harmon et al. 2003) and by fitting different likelihood models of continuous character evolution to our data (cf. Burbrink and Pyron 2009; Harmon et al. 2010). It was revealed that the unusual diversity of Arini was not due to particular clades within them being more diverse than others. The signature of an adaptive radiation was found in the early history of Neotropical parrots with increased diversification rates, a concentration of morphological size evolution and a partitioning of size niches probably representing different ecological adaptations. Relative bill size evolution was not different from a Brownian motion process. We found no sign of a decrease in diversification rates through time as is predicted by theories of adaptive radiations and often corroborated empirically (Schluter 2000; Harmon et al. 2003; Kozak et al. 2005; Ruber and Zardoya 2005; Weir 2006; Phillimore and Price 2008; Rabosky and Lovette 2008; Burbrink and Pyron 2009; Brock et al. 2011). Increased Andean uplift, climate change and associated habitat shifts from the Miocene onwards (Farias et al. 2008; Garzione et al. 2008; Thomas et al. 2008) may have provided new underutilized or vacant niche space with several ecological opportunities, not limiting the number of parrot species in South America.

GENERAL SUMMARY AND SYNTHESIS

Our understanding of the global spatial and temporal patterns of biodiversity has increased fundamentally since early explorers and naturalists discovered the variation in species richness across geographical regions and between different groups of organisms. Yet the historical and ecological processes that drive diversification in terms of speciation and morphological evolution are still controversially discussed (e.g. Schluter 2000; Penny and Phillips 2004; Weir 2006; Mittelbach et al. 2007; Brown et al. 2008; Price 2008; Rabosky 2009; Harmon et al. 2010; Hoorn et al. 2010; Wiens 2011). The relative influence of different drivers of diversification processes can only be assessed using a temporal and spatial framework based on a robustly

32 Chapter 2 resolved phylogeny. The increasing availability in the last few decades of phylogenetic hypotheses mainly based on molecular data in combination with new analytical tools such as the confident estimation of divergence times or evolutionary rates provides a firm basis for the study of diversity patterns (e.g. Pybus and Harvey 2000; Harmon et al. 2003; Drummond et al. 2006; O'Meara et al. 2006; Rabosky 2006b, a; Alfaro et al. 2009; Ho and Phillips 2009; Harmon et al. 2010; Brock et al. 2011). Moreover, the field has profited from an integration of ecology with evolutionary biology (Ricklefs 2004; Rabosky 2009). The study of particularly species-rich and widely distributed groups like the parrots can provide important insights into the mechanisms and processes driving the accumulation of species richness. However, such diverse clades often pose major systematic problems and challenges (Soltis 2007). This is also the case in the parrots, and before I could infer macroevolutionay patterns in their diversification by integrating molecular and morphological data, I had to tackle some important phylogenetic questions.

Phylogenetic relationships

Only a handful molecular phylogenies of parrots had been published at the beginning of the work on this thesis (Christidis et al. 1991; Miyaki et al. 1998; Brown and Toft 1999; Eberhard and Bermingham 2004; Ribas and Miyaki 2004; de Kloet and de Kloet 2005; Ribas et al. 2005; Astuti et al. 2006; Tavares et al. 2006). Most classifications and taxonomic frameworks were then still based on qualitatively assessed morphological features which had not been integrated into a cladistic framework (Smith 1975; Homberger 1980; Forshaw 1989; Collar 1998; Rowley 1998). Several papers published during work on this thesis, in combination with our results, have considerably improved our knowledge of parrot phylogeny. In particular, the work presented in chapters 3 and 4 (Schweizer et al. 2010; Schweizer et al. in press) aimed at elucidating the deep evolutionary relationships in parrots. We especially succeeded in resolving the phylogenetic position of several controversial basal taxa with high nodal support and confirmed the results of other recent molecular phylogenies which were in conflict with traditional views of the systematics of parrots. Several morphological and behavioral characters used in previous classifications were thus revealed to be phylogenetically uninformative and to show a high degree of homoplasy, while others were found to correlate well with molecular phylogenies (cf. Tavares et al. 2006; Mayr 2010). Consequently, a concise revision of the taxonomy of the groups within parrots usually ranked as ‘families’, ‘subfamilies’ or ‘tribes’ is badly needed. However, as there are still some controversies and gaps in our understanding of the phylogenetic history of parrots, this would be premature at the moment. It is now widely accepted that the New Zealand taxa Nestor and Strigops are the sister group of all other parrots and that the mainly Australasian Cacatuidae is the taxon that branches off next (cf. de Kloet and de Kloet 2005; Tavares et al. 2006; Tokita et al. 2007; Brown et al. 2008; Wright et al. 2008; Pratt et al. 2009; Mayr 2010; Pacheco et al. 2011). We identified the Neotropical Arini as the sister group of the African taxa Poicephalus and Psittacus. The latter two had been grouped together with Coracopsis from Madagascar and adjacent islands as Psittacini, in part because their relationships to other groups were not clear (Collar 1998). In contrast to this traditional treatment, we revealed a close relationship of Coracopsis with Psittrichas fulgidus from New

33 Evolutionary diversification and biogeography of parrots

Guinea. The clade consisting of (Arini + (Psittacus + Poicepalus)) most probably forms the sister group to all remaining parrot taxa except the basal clades of Nestor, Strigops and Cacatuidae. However, the position of the lineage leading to the cluster of Coracopsis and Psittrichas could not (yet) be unambiguously placed in the parrot phylogeny. A clade of Micropsittini (Micropsitta) and Psittaculini (sensu Collar (1998), but excluding Agapornis, Loriculus and Psittacella) was found to be the sister group of a mixed cluster of African, Australasian and Indo-Malayan taxa. This contained Platycercini, Loriinae, Cyclopsittacini as well as the genera Agapornis, Loriculus and Psittacella. Cyclopsittacini (sensu Collar (1998), but excluding Bolbopsittacus) formed the sister group to a clade consisting of Melopsittacus and the Loriinae, while Bolbopsittacus was found to be the sister group of Agapornis and Loriculus. Thus we revealed that Cyclopsittacini and Psittaculini are polyphyletic in their traditional treatment (cf. Smith 1975; Collar 1998). Melopsittacus has for a long time been included in Platycercini (Collar 1998), but its close relationship to Loriinae also renders the Platycercini polyphyletic. Whether the remaining Platycercini form a monophyletic group is still a matter of debate. Although not robustly supported throughout, we reported in chapter 6 a clade consisting of Pezoporus, Neopsephotus and Neophema as the sister group of all other Platycercini. We thus found no indication for a close relationship between the former clade and Agapornis-Loriculus in contrast to Wright et al. (2008, though Pezoporus was not included in their study). Similarly, Joseph et al. (2011) concluded that the cluster consisting of Pezoporus, Neopsephotus and Neophema was likely the sister group of the remaining Platycercini, although this relationship was not robustly supported in their data either. The phylogenetic position of Pezoporus-Neopsephotus- Neophema also remains unclear from morphological characters (Mayr 2010). Thus, more data need to be analyzed to resolve their relationship robustly and to confirm the monophyly of Platycercini. The same is true for the genus Psittacella from New Guinea. Like Joseph et al. (2011), we clearly showed that it is neither a member of Psittaculini, where it has traditionally been placed (Collar 1998), nor that it has close affinities to Platycercini (cf. Christidis et al. 1991). However, its phylogenetic position could not yet be robustly resolved, and it may well take a basal position within the cluster of old world parrots. Several papers published during the last few years have shed light on the phylogenetic relationships among and within genera of Neotropical parrots (Arini) (Eberhard and Bermingham 2004; Ribas and Miyaki 2004; Russello and Amato 2004; Ribas et al. 2005; Ribas et al. 2006; Tavares et al. 2006; Ribas et al. 2007a; Ribas et al. 2007b; Wright et al. 2008; Ribas et al. 2009). The phylogenetic hypothesis for Arini presented in chapter 8 is highly congruent with the results of these studies. The discrepancies revealed with the traditional systematic treatment of Arini again call for a concise systematic revision. Chapter 6 and 7 dealt with phylogenetic relationships of old world parrots on the genus and/or species level. The results of these chapters demonstrated that polyphyly not only occurs in taxonomically higher ranked taxa of parrots, but also within genera and species. The phylogenetic relationships within the ‘core Platycercini’ (chapter 6) and Prioniturus (chapter 7) were found in some cases to contradict traditional thinking and taxonomic treatment, which was mainly based on non-molecular traits. We consequently showed that shared plumage features may not always be a sign of a close phylogenetic relationship, but may instead indicate a plesiomorphic pattern or convergent evolution. Similar results were found for the Neotropical parrot genera

34 Chapter 2

Amazona, Aratinga and Pionopsitta (Ribas and Miyaki 2004; Russello and Amato 2004; Ribas et al. 2005; Ribas et al. 2007b). As shown in chapter 7, a thorough phylogenetic analysis on the species level can reveal an underestimation of the number of recognized species or even cryptic diversity within species. Depending on the taxonomic source used, one to two or probably even three additional species in the genus Prioniturus should be recognized in the Philippines. This is in line with other studies, which demonstrated that the alpha diversity of the Philippine avifauna is strongly underestimated (Peterson 2006; Lohman et al. 2010; Oliveros and Moyle 2010). As Asian birds were generally found to be over-lumped in recent studies compared to e.g. the Neotropical avifauna (e.g. Collar 2003), several cryptic parrot species probably await formal recognition especially in polytypic species or species assemblages distributed over different island groups in Southeast Asia and the Australasian region. Species usually form the basis for conservation assessment and strategies (Mace 2004) and an incorrect interpretation of alpha diversity can result in dire consequences through a lack of conservation measures. These may be particularly fatal for the often small and susceptible populations of island taxa (e.g. Hazevoet 1996; Lovette et al. 1999; Sangster 2000). Species-level phylogenies and systematic revisions for insular old world parrot taxa like Charmosyna, Geoffroyus, Loriculus, Micropsitta, Tanygnathus and Trichoglossus are thus urgently needed. It is still unclear which group of Neoaves forms the sister taxon of Psittaciformes. Although this issue was not the focus of this thesis, several groups of Neoaves which have previously been allied to the parrots were included in the analyses presented in chapter 4 (Schweizer et al. in press). An affiliation with Falconidae or Coraciiformes was indicated there but gained no robust support. Other taxa combinations were proposed as close relatives in recently published molecular phylogenies. In the phylogenomic study of Hackett et al. (2008), the parrots formed a clade together with Falconidae and Passeriformes, with the latter revealed to be the parrots’ sister taxon. In contrast, Psittaciformes clustered with Passeriformes, Falconidae and Cariamidae in a study of nuclear exons and introns (Ericson et al. 2006). Analyses of complete mtDNA genomes also led to controversial results. While the parrots were either placed as the most basal lineage within Neoaves or as the basal group of the latter together with the ‘WoodKing’ (Coliiformes, Coraciiformes, Trogoniformes, Piciformes) in the study of Pratt et al. (2009), Pacheco et al (2011) found evidence indicating that Strigiformes might be the sister taxon of Psittaciformes. In another study using mtDNA, the parrots formed a cluster with Strigiformes, Coraciiformes and Piciformes (Brown et al. 2008). It has been argued that the evolutionary relationships between the different groups within Neoaves can probably never be fully clarified due to simultaneous radiation of multiple lineages (Poe and Chubb 2004). Although we are still far from a robustly supported phylogenetic hypothesis (e.g. Mayr 2011), recent studies have demonstrated that the phylogenetic positions of the deep lineages of Neoaves could indeed be resolvable by using a great deal of data, reducing noise and incorporating prior knowledge (Pratt et al. 2009).

Macroevolution and biogeography of parrots: temporal and spatial diversification patterns

Early diversification of parrots It has for a long time been hypothesized that Gondwana played an important role in the early evolution of modern birds (Cracraft 2001). Our data presented in chapter 3

35 Evolutionary diversification and biogeography of parrots and 4 (Schweizer et al. 2010; Schweizer et al. in press) indeed suggest that a common ancestor of all extant parrots lived in Australasia. However, the break-up of Gondwana starting in the late Cretaceous was most probably not a driving force for the early diversification of parrots as has been proposed by others (e.g. Cracraft 2001; de Kloet and de Kloet 2005; Wright et al. 2008). This is in congruence with other studies showing for different taxonomic groups that the distribution patterns among organisms occurring on the different continents of the southern hemisphere may not be sufficiently explained by vicariance promoted by the fragmentation of Gondwana (Sanmartin and Ronquist 2004; McGlone 2005; Waters and Craw 2006; Harshman et al. 2008; Upchurch 2008; Phillips et al. 2010). In agreement with other studies (Brown et al. 2008; Pratt et al. 2009; Pacheco et al. 2011), we found in chapter 4 (Schweizer et al. in press) that the stem lineage of parrots was probably in existence in the Cretaceous. However, we have strong evidence that the diversification of the crown group parrots started most probably after the K–Pg boundary in the late Palaeocene/early Eocene. This is clearly later than the assumption of an initial split at 80–85 Ma, coinciding with the separation of New Zealand from Gondwana (Tavares et al. 2006; Ribas et al. 2007a; Wright et al. 2008; Ribas et al. 2009). In congruence with our results, Pacheco et al. (2011) dated the same split between 54.13 and 61.44 Ma, while Pratt et al. (2009) placed it even later, at 50.38 Ma (mean age). Recent geological evidence indicates that a land bridge between New Zealand and Australia may have existed until the Early Eocene, up to 52 Ma (Gaina et al. 1998; Tennyson 2010). Thus, even if the initial split within the crown group of parrots occurred later than hitherto assumed, then it could still have been caused by vicariance after the final complete separation of New Zealand from Australia. The time scale of the radiation of modern birds has been a matter of controversy for years and the influence of the mass extinction event at the K-Pg boundary on their evolution is still disputed (cf. Penny and Phillips 2004). Our results support a growing body of evidence that many extant avian lineages including parrots were in existence in the Cretaceous, well before the extinction of other dinosaurs and pterosaurs (Hedges et al. 1996; Cooper and Penny 1997; Clarke et al. 2005; Pereira and Baker 2006; Slack et al. 2006; Brown et al. 2007; Brown et al. 2008; Pratt et al. 2009; Pacheco et al. 2011). These results refute the hypothesis of an explosive adaptive radiation of modern birds after the K-Pg boundary (Feduccia 1995, 2003). Accordingly, we could find no evidence for a burst of diversification in parrots at the beginning of the Paleogene in response to the extinction of other taxa.

Dispersal and Vicariance Several long-distance dispersal and colonization events have to be invoked to explain the current distribution patterns of the different parrot groups as shown in chapters 3, 4 (Schweizer et al. 2010; Schweizer et al. in press), 6 and 7. This result is in line with a recent change in biogeographic thinking away from a vicariance driven model of diversification toward a resurrection of long-distance and oceanic dispersal as an important process in the build-up of regional fauna (de Queiroz 2005; McGlone 2005; Cowie and Holland 2006; Yoder and Nowak 2006). Accordingly, several recently published studies convincingly proposed long-distance dispersal as an explanation for current distribution patterns of different animal groups (e.g. Raxworthy et al. 2002; Vences et al. 2003; Yoder et al. 2003; Pereira et al. 2007; Harshman et al. 2008; Voelker et al. 2009; Jonsson et al. 2010a; Jonsson et al. 2010b).

36 Chapter 2

Major dispersal events during the evolutionary history of parrots included the colonization of Africa and South America from Antarctica as a consequence of the latter continent becoming increasingly covered with ice during the late Oligocene/early Eocene (Zachos et al. 2001). Moreover, a common ancestor of each of the genera Coracopsis and Agapornis has independently colonized Africa and or Madagascar via trans-oceanic dispersal across the Indian Ocean from Australasia in the Oligocene/early Miocene. This biogeographic association between Australasia and Africa/Madagascar is surprising, and, within birds, it has so far only been convincingly proposed for the cuckoo-shrikes (Campephagidae) (Jonsson et al. 2010b). Dispersal from Australasia to the Indo-Malayan region also occurred several times independently. While the Philippine endemic Bolbopsittacus probably at first colonized the Indo-Malayan region in the Oligocene, all other colonization events occurred after Australasia came into contact with and reached its present position relative to Indo-Malaysia around 20–25 Ma (Li and Powell 2001; Hall 2002). The new proximity of these two landmasses in combination with emergent land and island arcs squeezed in between serving as stepping stones facilitated biotic exchange between Australasia and the Indo-Malayan region via Wallacea (Hall 1998; Moss and Wilson 1998; Hall 2002; Jonsson et al. 2008). The exact pattern of colonization of the Indo-Malayan region by ancestors of Loriculus, Prioniturus, Psittinus, Psittacula and Tanygnathus could not be inferred unambiguously based on the available data and the biogeographic reconstruction presented in chapters 3 and 4 (Schweizer et al. 2010; Schweizer et al. in press). In chapter 7, we provided a picture of dispersal and colonization patterns of Prioniturus across and within Wallacea and the Philippines. Prioniturus has diversified by a complex combination of colonization of island groups and subsequent divergence in allopatry among and within island groups. Dispersal between Sulawesi/Wallacea and the Philippines occurred twice and documents a rare case of faunal exchange between these two regions (cf. Woodruff et al. 1999; Smith et al. 2000; Evans et al. 2003; Jones and Kennedy 2008; Esselstyn et al. 2009). As detailed in chapter 6, long-distance over-water dispersal also played an important role in one lineage of the Australasian platycercine parrots (Platycercini). Again contrary to traditional biogeographic paradigms, Platycercini provide an example that islands may not be considered only as sinks for biodiversity, but can contribute to the build-up of the biota of larger regions, which is congruent with other recent studies (e.g. Filardi and Moyle 2005; Heaney 2007; Bellemain and Ricklefs 2008; Jonsson et al. 2010a). The agreement between geological and biological evidence for some cladogenetic events indicates that vicariance through continental rifting may have nevertheless played a role in the early evolutionary history of parrots. A convincing example is presented in chapter 4 (Schweizer et al. in press): a common ancestor of the Neotropical Arini and the African Psittacini most probably lived on Antarctica, and the timing of its split from the remaining Australasian lineages corresponds to the final physical separation of Australia and Antarctica around 40 Ma (Li and Powell 2001). Whether the initial split within crown group parrots was caused by dispersal or vicariance is still controversially discussed (see above). The ability to disperse over long distances may be considered as a prerequisite for parrots becoming a widely distributed and species-rich clade. Dispersal capabilities might be associated with traits favored by special types of ecological distributions (Jonsson et al. 2011). One group of oscine passerines (core Corvoidea) has colonized different continents from the islands of the proto-Papuan archipelago (Jonsson et al. 2011) and the island setting in this region might have provided opportunities for

37 Evolutionary diversification and biogeography of parrots vagile species and favored the evolution of abilities needed to disperse over long distances. Interestingly, two of the lineages of parrots that colonized Africa via trans- oceanic dispersal also seem to have had their origin in the Australo-Papuan region, as demonstrated in chapters 3 and 4 (Schweizer et al. 2010; Schweizer et al. in press). Similar as in core Corvoidea (Jonsson et al. 2011), the likelihood of successful long- distance dispersal might have been increased by several morphological characteristics of parrots, which could have enhanced flexibility and facilitated adaptations to new life-history strategies: high cognitive abilities in combination with large brain size and high levels of inquisitiveness and exploratory behavior (Collar 1998; Mettke- Hofmann et al. 2002; Pepperberg 2002; Iwaniuk et al. 2005), complex social relationships and flocking behavior (Collar 1998; Pepperberg 2002; Wright and Dahlin 2007; Dahlin and Wright 2009), as well as high manipulative ability of feet, bill and tongue (Homberger 1980; Collar 1998). In future studies it would be interesting to test if the lineages of parrots that include successful dispersers differ in these aspects from the remaining, more endemic groups.

Diversification rates and drivers of diversification It has often been proposed that species richness is positively correlated with clade age (e.g. Wiens 2011). However, no relationship between the age of a clade and its number of species has been found in many groups (e.g. Magallon and Sanderson 2001; Seehausen 2006; Ricklefs et al. 2007). Accordingly, the results of different chapters of this thesis demonstrated that age is certainly not a good predictor of diversity in parrots. Diversification rates of parrots may not have been constant over time, and the chiefly Australasian Loriinae (lories) and probably also the Neotropical Arini were found to be exceptionally species-rich given their age compared to the remaining parrot clades (Chapter 4, Schweizer et al in press.). Intrinsic and extrinsic factors can regulate clade diversity and ecological limits in particular might imply a decoupling of clade ages and species richness (Hunter 1998; Ricklefs 2003; Rabosky 2009). A complex array of such factors also influenced the diversification rates of parrots. Ecological opportunities, i.e. the presence of various underexploited or inefficiently exploited resources leading to a release from natural selection, can promote diversification (e.g. Nosil and Reimchen 2005; Price 2008; Losos 2010; Yoder et al. 2010). Ecological opportunities can inter alia be provided by evolutionary key innovations (e.g. Hunter 1998; Yoder et al. 2010). An evolutionary key innovation is often considered to be important for promoting diversification (Heard and Hauser 1995; Hunter 1998) and indeed played a role in parrots as demonstrated in chapters 5 and 4 (Schweizer et al. in press) for the nectarivorous lories. In addition, traits facilitating long-distance dispersal might also be considered as key innovations in different groups of parrots (see above). A key innovation hypothesis can be convincingly supported by taking two components into account, a comparative test and an ecological or functional argument (Heard and Hauser 1995). As demonstrated in chapter 4 (Schweizer et al. in press), the nectarivorous lories are unexpectedly species-rich with regard to their age. Additionally, we showed in chapter 5 that their diet shift to nectarivory was associated with rapidly evolved morphological adaptations in the digestive tract which probably allowed them to effectively feed on nectar and pollen. Nectar may have provided a spatially widespread underutilized niche, which would have allowed lories to expand their geographical ranges. This might have often been followed by allopatric speciation, driving the diversification of the lories into an exceptionally species-rich clade. The lories may thus be considered

38 Chapter 2 an example of a nonadaptive radiation (Rundell and Price 2009). Nectar seems to be an unpredictable resource in the Australasian region (Fleming and Muchhala 2008) and this may have prevented further ecological specialization within the lories. Hence, we provide an example that an ecological opportunity provided by a key innovation can promote significant lineage diversification in the absence of an adaptive radiation through allopatric partitioning of the same new niche. The parallel evolution of similar adaptations in the digestive tract in the other nectarivorous parrot lineages beside the lories indicates a phenotype-environment correlation. However, it was not associated with an increased species proliferation in these groups. Although the diet shift to nectarivory has the potential of being an evolutionary key innovation as demonstrated by the lories, various factors including ecological circumstances like interspecific competition or the lack of opportunities for allopatric speciation can inhibit diversification. Hence, not every evolutionary innovation necessarily promotes diversification (cf. Heard and Hauser 1995; Vermeij 2001; Price et al. 2010). High diversity within particular clades does not always have to be related to intrinsic key innovations of these organisms, instead extrinsic circumstances like special ecological conditions or episodes of expansion of geographic distributions may also promote diversification (e.g. Ricklefs 2003). Such crucial evolutionary events have also been important in the evolutionary diversification of several parrot groups. As detailed in chapter 4 (Schweizer et al. in press), two potential phases of accelerated net diversification rates for the parrots as a whole coincided with periods of global climate change characterized by cooler conditions. The older increase in diversification rate for the parrots as a whole took place around the Eocene/Oligocene boundary when Antarctica became increasingly ice-encrusted (Zachos et al. 2001). In parrots, major lineages emerged then, as in other birds groups (Moyle 2005; Baker et al. 2006; Pereira and Baker 2008), and Africa and South America were colonized. The late Eocene/Oligocene also marked a significant period of dispersal out of the Papuan area for many lineages within one group of oscine passerines (core Corvoidea) (Jonsson et al. 2011). Newly colonized areas may provide several ecological opportunities that trigger fast adaptations and initiate adaptive radiations in some cases (Burbrink and Pyron 2009; Kassen 2009; Losos 2010; Yoder et al. 2010). The Neotropical parrots (Arini) indeed show characteristics consistent with an adaptive radiation early in their evolutionary history, as demonstrated in chapter 8. However, it is not clear if their initial diversification occurred directly after their colonization of South America. Size niches probably representing different ecological adaptations seem to have been partitioned early in the diversification of Arini. An adaptive basis of trait differentiation is essential for an adaptive radiation (e.g. Losos 2010) and additional studies should thus investigate the adaptive values of size variation in Neotropical parrots. Neotropical parrots may have not yet reached their equilibrium diversity, because no decrease in diversification rates through time could be detected as would be expected in the course of adaptive radiations. Historical processes from the Miocene onwards may have continuously provided new niche spaces that did not limit their number of species (see below). Weir (2006) hypothesized that lowland diversity may approach its capacity in Neotropical birds, whereas the diversification of highland taxa seems currently not to be limited. In parrots however, next to evidence for Pleistocene speciation in the Andes for one parrot group (Ribas et al. 2007a), recent speciation events seemingly also happened in the lowlands. These may have been influenced by Pleistocene climate changes and associated habitat shifts through dynamic interactions of wet and dry forests, isolation

39 Evolutionary diversification and biogeography of parrots of open dry areas and speciation along forest/open-area ecotones (Ribas and Miyaki 2004; Ribas et al. 2007b; Ribas et al. 2009). Future studies with a more complete taxon sampling should attempt to disentangle these various influences on the diversification rates of Neotropical parrots. Further research should also try to identify potential ecological factors which may have limited the diversification of the African Psittacini. The common ancestor of this sister group of the Neotropical parrots colonized Africa probably at the time when South America was invaded by a common ancestor of Arini, however, the Psittacini remained comparatively species-poor. The younger acceleration of the net diversification rate in parrots as a whole probably occurred from the middle Miocene onwards. Global climate change then influenced continental vegetations and was associated with shrinking rain forests and an increase in areas of open vegetation in Australia, Africa and South America (Jacobs et al. 1999; Martin 2006). Additionally, uplift of the Andes in South America and of the Tibetan plateau in Asia during roughly the same time period led to further vegetation shifts and habitat fragmentation. These environmental changes may have catalyzed the diversification of parrots by promoting vicariance events and by providing ecological opportunities leading to lineage diversification through ecological speciation, e.g. across forest-savanna ecotones (cf. Smith et al. 2001; Smith et al. 2005). Accordingly, several dry-adapted lineages of the Australasian, African and probably Neotropical region (Tavares et al. 2006) parrots emerged at this time. Aridification of Australia from the mid-Miocene onwards may have been especially important for the diversification of the Australasian cockatoos (White et al. 2011) and platycercine parrots, as presented in chapter 6. In general, ancestors of recent arid taxa in the Australian fauna may have either diverged from their closest relatives before the onset of aridification and adapted to arid environment in situ or originated from mesic ancestors through ecological speciation by multiple colonization events of the precursors of modern arid environments from mesic biomes (reviewed in Byrne et al. 2008). As we show in chapter 6, both processes played a role in the platycercine parrots. On the other hand, the divergence events within Australian mesic and monsoon biome platycercine parrots seem to have been primarily caused by vicariance across proposed faunal barriers (e.g. Schodde 2006) due to habitat fragmentation following the spread of arid environments. This seems to mirror a general pattern found in Australian mesic and monsoon biome taxa (Bowman et al. 2009; Byrne et al. in press).

Final conclusions

The causes of diversification in the large and widely distributed clade of the parrots appear to be complex, as expected in most systems. “Nature was and is complex; an explanation of nature, if satisfactory, will also be complex”, as pointed out by Erwin (1981) clearly applies also to the parrots. However, the integrative approach applied, using both molecular and morphological data on the basis of a robust phylogenetic hypothesis and a spatial and temporal framework, allowed me to pinpoint some of the more important events in the evolutionary history of parrots and to identify mechanisms and processes driving their diversification. The diversity of parrots was influenced by geological and climatological alterations, and triggered by a unique mixture of biotic expansions and vicariance events followed by within-area radiations or extinctions. Exceptionally high diversity within clades was reached via a

40 Chapter 2 nonadaptive radiation initiated by an evolutionary key innovation and an adaptive radiation after the colonization of a new continent. Hence, the evolutionary diversification of parrots is best circumscribed with a taxon-pulse model (Erwin 1981; Halas et al. 2005), as already hypothesized in chapter 3 (Schweizer et al. 2010). Although the phylogeny of the parrots is still not completely resolved and the driving or limiting factors of diversification remain to be identified in several groups, this work provides a basis for further studies on the evolutionary history of parrots. The results presented herein are not only interesting in a historical context, but can also help us to understand and predict how current and future climate and environmental change can impact biodiversity on the global and local scale. They will also hopefully stimulate other researchers to study diversification patterns using an integrative approach in other bird and organismic groups.

Literature cited

Alfaro, M. E., F. Santini, C. Brock, H. Alamillo, A. Dornburg, D. L. Rabosky, G. Carnevale, and L. J. Harmon. 2009. Nine exceptional radiations plus high turnover explain species diversity in jawed vertebrates. Proceedings of the National Academy of Sciences of the United States of America 106:13410- 13414. Astuti, D., N. Azuma, H. Suzuki, and S. Higashi. 2006. Phylogenetic relationships within parrots (Psittacidae) inferred from mitochondrial cytochrome-b gene sequences. Zoological Science 23:191-198. Baker, A. J., S. L. Pereira, O. P. Haddrath, and K. A. Edge. 2006. Multiple gene evidence for expansion of extant penguins out of Antarctica due to global cooling. Proceedings of the Royal Society B-Biological Sciences 273:11-17. Bellemain, E., and R. E. Ricklefs. 2008. Are islands the end of the colonization road? Trends in Ecology & Evolution 23:461-468. Bowman, D. M. J. S., G. K. Brown, M. F. Braby, J. R. Brown, L. G. Cook, M. D. Crisp, F. Ford, S. Haberle, J. Hughes, Y. Isagi, L. Joseph, J. McBride, G. Nelson, and P. Y. Ladiges. 2009. Biogeography of the Australian monsoon tropics. Journal of Biogeography 37:201-216. Brock, C. D., L. J. Harmon, and M. E. Alfaro. 2011. Testing for temporal variation in diversification rates when sampling is incomplete and nonrandom. Systematic Biology 60:410-419. Brown, D. M., and C. A. Toft. 1999. Molecular systematics and biogeography of the cockatoos (Psittaciformes: Cacatuidae). Auk 116:141-157. Brown, J. W., R. B. Payne, and D. P. Mindell. 2007. Nuclear DNA does not reconcile 'rocks' and 'clocks' in Neoaves: a comment on Ericson et al. Biology Letters 3:257-259. Brown, J. W., J. S. Rest, J. Garcia-Moreno, M. D. Sorenson, and D. P. Mindell. 2008. Strong mitochondrial DNA support for a Cretaceous origin of modern avian lineages. BMC Biology 6. Burbrink, F. T., and R. A. Pyron. 2009. How does ecological opportunity influence rates of speciation, extinction, and morphological diversification in New World ratsnakes (tribe Lampropeltini)? Evolution 64:934-943. Byrne, M. 2008. Evidence for multiple refugia at different time scales during Pleistocene climatic oscillations in southern Australia inferred from phylogeography. Quat. Sci. Rev. 27:2576-2585.

41 Evolutionary diversification and biogeography of parrots

Byrne, M., D. A. Steane, L. Joseph, D. K. Yeates, G. J. Jordan, D. Crayn, K. Aplin, D. J. Cantrill, L. G. Cook, M. D. Crisp, J. S. Keogh, J. Melville, C. Moritz, N. Porch, J. M. K. Sniderman, P. Sunnucks, and P. H. Weston. in press. Decline of a biome: evolution, contraction, fragmentation, extinction and invasion of the Australian mesic zone biota. Journal of Biogeography. Byrne, M., D. K. Yeates, L. Joseph, M. Kearney, J. Bowler, M. A. J. Williams, S. Cooper, S. C. Donnellan, J. S. Keogh, R. Leys, J. Melville, D. J. Murphy, N. Porch, and K. H. Wyrwoll. 2008. Birth of a biome: insights into the assembly and maintenance of the Australian arid zone biota. Molecular Ecology 17:4398-4417. Christidis, L., R. Schodde, D. D. Shaw, and S. F. Maynes. 1991. Relationships among the Australo-Papuan parrots, lorikeets, and cockatoos (Aves, Psittaciformes) - protein evidence. Condor 93:302-317. Chubb, A. L. 2004a. New nuclear evidence for the oldest divergence among neognath birds: the phylogenetic utility of ZENK (i). Molecular Phylogenetics and Evolution 30:140-151. Chubb, A. L. 2004b. Nuclear corroboration of DNA-DNA hybridization in deep phylogenies of hummingbirds, swifts, and passerines: the phylogenetic utility of ZENK (ii). Molecular Phylogenetics and Evolution 30:128-139. Churchill, D. M., and P. Christensen. 1970. Observations on pollen harvesting by brush-tongued lorikeets. Aust J Zool 18:427-437. Clarke, J. A., C. P. Tambussi, J. I. Noriega, G. M. Erickson, and R. A. Ketcham. 2005. Definitive fossil evidence for the extant avian radiation in the Cretaceous. Nature 433:305-308. Collar, N. J. 1998. Family Psittacidae (parrots). Pp. 280-477 in J. del Hoyo, A. Elliott, and J. Sargatal, eds. Handbook of the birds of the world, vol. 4. Lynx Edicions, Barcelona. Collar, N. J. 2003. How many bird species are there in Asia? Oriental Bird Club Bull 38:20-30. Colosimo, P. F., K. E. Hosemann, S. Balabhadra, G. Villarreal, M. Dickson, J. Grimwood, J. Schmutz, R. M. Myers, D. Schluter, and D. M. Kingsley. 2005. Widespread parallel evolution in sticklebacks by repeated fixation of ectodysplasin alleles. Science 307:1928-1933. Cooper, A., and D. Penny. 1997. Mass survival of birds across the Cretaceous- Tertiary boundary: molecular evidence. Science 275:1109-1113. Cowie, R. H., and B. S. Holland. 2006. Dispersal is fundamental to biogeography and the evolution of biodiversity on oceanic islands. Journal of Biogeography 33:193-198. Cracraft, J. 2001. Avian evolution, Gondwana biogeography and the Cretaceous- Tertiary mass extinction event. Proceedings of the Royal Society B-Biological Sciences 268:459-469. Dahlin, C. R., and T. F. Wright. 2009. Duets in yellow-naped amazons: variation in syntax, note composition and phonology at different levels of social organization. Ethology 115:857-871. de Kloet, R. S., and S. R. de Kloet. 2005. The evolution of the spindlin gene in birds: sequence analysis of an intron of the spindlin W and Z gene reveals four major divisions of the Psittaciformes. Molecular Phylogenetics and Evolution 36:706-721. de Queiroz, A. 2005. The resurrection of oceanic dispersal in historical biogeography. Trends in Ecology & Evolution 20:68-73.

42 Chapter 2

Dickinson, E. D. E. 2003. The Howard and Moore complete checklist of the birds of the world. Princeton University Press, Princeton, New Jersey. Drummond, A. J., S. Y. W. Ho, M. J. Phillips, and A. Rambaut. 2006. Relaxed phylogenetics and dating with confidence. PLoS. Biol. 4:699-710. Eberhard, J. R., and E. Bermingham. 2004. Phylogeny and biogeography of the Amazona ochrocephala (Aves: Psittacidae) complex. Auk 121:318-332. Ericson, P. G. P., C. L. Anderson, T. Britton, A. Elzanowski, U. S. Johansson, M. Kallersjo, J. I. Ohlson, T. J. Parsons, D. Zuccon, and G. Mayr. 2006. Diversification of Neoaves: integration of molecular sequence data and fossils. Biology Letters 2:543-U541. Erwin, T. L. 1981. Taxon pulses, vicariance, and dispersal: an evolutionary synthesis illustrated by carabid beetles. Pp. 159-196 in G. Nelson, and D. E. Rosen, eds. Vicariance biogeography: a critique. Columbia University Press, New York. Esselstyn, J. A., R. M. Timm, and R. M. Brown. 2009. Do geological or climatic processes drive speciation in dynamic archipelagos? The tempo and mode of diversification in Southeast Asian shrews. Evolution 63:2595-2610. Evans, B. J., R. M. Brown, J. A. McGuire, J. Supriatna, N. Andayani, A. Diesmos, D. Iskandar, D. J. Melnick, and D. C. Cannatella. 2003. Phylogenetics of fanged frogs: testing biogeographical hypotheses at the interface of the Asian and Australian faunal zones. Systematic Biology 52:794-819. Farias, M., R. Charrier, S. Carretier, J. Martinod, A. Fock, D. Campbell, J. Caceres, and D. Comte. 2008. Late Miocene high and rapid surface uplift and its erosional response in the Andes of central Chile (33 degrees-35 degrees S). Tectonics 27:22. Feduccia, A. 1995. Explosive evolution in Tertiary birds and mammals. Science 267:637-638. Feduccia, A. 2003. 'Big bang' for Tertiary birds? Trends in Ecology & Evolution 18:172-176. Felsenstein, J. 1985. Phylogenies and the comparative method. American Naturalist 125:1-15. Filardi, C. E., and R. G. Moyle. 2005. Single origin of a pan-Pacific bird group and upstream colonization of Australasia. Nature 438:216-219. Fleming, T. H., and N. Muchhala. 2008. Nectar-feeding bird and bat niches in two worlds: pantropical comparisons of vertebrate pollination systems. Journal of Biogeography 35:764-780. Forshaw, J. M. 1989. Parrots of the world. Landsdowne Press, Melbourne. Forshaw, J. M. 2006. Parrots of the world: an identification guide. Princeton University Press, Princeton. Freckleton, R. P. 2009. The seven deadly sins of comparative analysis. J. Evol. Biol. 22:1367-1375. Fuchs, J., C. Cruaud, A. Couloux, and E. Pasquet. 2007. Complex biogeographic history of the cuckoo-shrikes and allies (Passeriformes: Campephagidae) revealed by mitochondrial and nuclear sequence data. Molecular Phylogenetics and Evolution 44:138-153. Fuchs, J., J. Fjeldsa, R. C. K. Bowie, G. Voelker, and E. Pasquet. 2006. The African warbler genus Hyliota as a lost lineage in the oscine songbird tree: molecular support for an African origin of the Passerida. Molecular Phylogenetics and Evolution 39:186-197.

43 Evolutionary diversification and biogeography of parrots

Gaina, V., W. R. Roest, R. D. Müller, and P. Symonds. 1998. The opening of the Tasman Sea: a gravity anomaly animation. American Meteorological Society, AMS JournalsOnLine 2:1-23. Garland, T., and A. R. Ives. 2000. Using the past to predict the present: confidence intervals for regression equations in phylogenetic comparative methods. American Naturalist 155:346-364. Gartrell, B. D. 2000. The nutritional, morphologic, and physiologic bases of nectarivory in Australian birds. Journal of Avian Medicine and Surgery 14:85- 94. Gartrell, B. D., and S. M. Jones. 2001. Eucalyptus pollen grain emptying by two Australian nectarivorous psittacines. Journal of Avian Biology 32:224-230. Gartrell, B. D., S. M. Jones, R. N. Brereton, and L. B. Astheimer. 2000. Morphological adaptations to nectarivory of the alimentary tract of the swift parrot Lathamus discolor. Emu 100:274-279. Garzione, C. N., G. D. Hoke, J. C. Libarkin, S. Withers, B. MacFadden, J. Eiler, P. Ghosh, and A. Mulch. 2008. Rise of the Andes. Science 320:1304-1307. Grafen, A. 1989. The phylogenetic regression. Philos. Trans. R. Soc. Lond. Ser. B- Biol. Sci. 326:119-157. Groth, J. G., and G. F. Barrowclough. 1999. Basal divergences in birds and the phylogenetic utility of the nuclear RAG-1 gene. Molecular Phylogenetics and Evolution 12:115-123. Güntert, M. 1981. Morphologische Untersuchungen zur adaptiven Radiation des Verdauungstraktes bei Papageien (Psittaci). Zool. Jahrb. Abt. Anat. Ontog. Tiere 106:471-526. Güntert, M., and V. Ziswiler. 1972. Konvergenzen in der Struktur von Zunge und Verdauungstrakt nektarfressender Papageien. Rev. Suisse Zool. 79:1017-1026. Hackett, S. J., R. T. Kimball, S. Reddy, R. C. K. Bowie, E. L. Braun, M. J. Braun, J. L. Chojnowski, W. A. Cox, K. L. Han, J. Harshman, C. J. Huddleston, B. D. Marks, K. J. Miglia, W. S. Moore, F. H. Sheldon, D. W. Steadman, C. C. Witt, and T. Yuri. 2008. A phylogenomic study of birds reveals their evolutionary history. Science 320:1763-1768. Halas, D., D. Zamparo, and D. R. Brooks. 2005. A historical biogeographical protocol for studying biotic diversification by taxon pulses. Journal of Biogeography 32:249-260. Hall, R. 1998. The plate tectonics of Cenozoic SE Asia and the distribution of land and sea. Pp. 99-131 in R. Hall, and J. D. Holloway, eds. Biogeography and geological evolution of SE Asia. Backhuys Publishers, Leiden. Hall, R. 2002. Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based reconstructions, model and animations. J. Asian Earth Sci. 20:353-431. Harmon, L. J., J. B. Losos, T. J. Davies, R. G. Gillespie, J. L. Gittleman, W. B. Jennings, K. H. Kozak, M. A. McPeek, F. Moreno-Roark, T. J. Near, A. Purvis, R. E. Ricklefs, D. Schluter, J. A. Schulte, O. Seehausen, B. L. Sidlauskas, O. Torres-Carvajal, J. T. Weir, and A. O. Mooers. 2010. Early bursts of body size and shape evolution are rare in comparative data. Evolution 64:2385-2396. Harmon, L. J., J. A. Schulte, A. Larson, and J. B. Losos. 2003. Tempo and mode of evolutionary radiation in iguanian lizards. Science 301:961-964. Harshman, J., E. L. Braun, M. J. Braun, C. J. Huddleston, R. C. K. Bowie, J. L. Chojnowski, S. J. Hackett, K. L. Han, R. T. Kimball, B. D. Marks, K. J.

44 Chapter 2

Miglia, W. S. Moore, S. Reddy, F. H. Sheldon, D. W. Steadman, S. J. Steppan, C. C. Witt, and T. Yuri. 2008. Phylogenomic evidence for multiple losses of flight in ratite birds. Proceedings of the National Academy of Sciences of the United States of America 105:13462-13467. Hazevoet, C. J. 1996. Conservation and species lists: taxonomic neglect promotes the extinction of endemic birds, as examplified by taxa from eastern Atlantic islands. Bird Conservation International 6:181-196. Heaney, L. R. 2007. Is a new paradigm emerging for oceanic island biogeography? Journal of Biogeography 34:753-757. Heard, S. B., and D. L. Hauser. 1995. Key evolutionary innovations and their ecological mechanisms. Historical Biol 10:151-173. Hedges, S. B., P. H. Parker, C. G. Sibley, and S. Kumar. 1996. Continental breakup and the ordinal diversification of birds and mammals. Nature 381:226-229. Ho, S. Y. W., and M. J. Phillips. 2009. Accounting for calibration uncertainty in phylogenetic estimation of evolutionary divergence times. Systematic Biology 58:367-380. Homberger, D. G. 1980. Funktionell- morphologische Untersuchungen zur Radiation der Ernährungs- und Trinkmethoden der Papageien (Psittaci). Bonn. Zool. Monogr. 13:1-92. Hoorn, C., F. P. Wesselingh, H. ter Steege, M. A. Bermudez, A. Mora, J. Sevink, I. Sanmartin, A. Sanchez-Meseguer, C. L. Anderson, J. P. Figueiredo, C. Jaramillo, D. Riff, F. R. Negri, H. Hooghiemstra, J. Lundberg, T. Stadler, T. Sarkinen, and A. Antonelli. 2010. Amazonia through time: Andean uplift, climate change, landscape evolution, and biodiversity. Science 330:927-931. Hunter, J. P. 1998. Key innovations and the ecology of macroevolution. Trends in Ecology & Evolution 13:31-36. Iwaniuk, A. N., K. M. Dean, and J. E. Nelson. 2005. Interspecific allometry of the brain and brain regions in parrots (Psittaciformes): Comparisons with other birds and primates. Brain Behav. Evol. 65:40-59. Jacobs, B. F., J. D. Kingston, and L. L. Jacobs. 1999. The origin of grass-dominated ecosystems. Ann. Mo. Bot. Gard. 86:590-643. Jones, A. W., and R. S. Kennedy. 2008. Evolution in a tropical archipelago: comparative phylogeography of Philippine fauna and flora reveals complex patterns of colonization and diversification. Biological Journal of the Linnean Society 95:620-639. Jonsson, K. A., R. C. K. Bowie, R. G. Moyle, L. Christidis, J. A. Norman, B. W. Benz, and J. Fjeldsa. 2010a. Historical biogeography of an Indo-Pacific passerine bird family (Pachycephalidae): different colonization patterns in the Indonesian and Melanesian archipelagos. Journal of Biogeography 37:245- 257. Jonsson, K. A., R. C. K. Bowie, J. A. A. Nylander, L. Christidis, J. A. Norman, and J. Fjeldsa. 2010b. Biogeographical history of cuckoo-shrikes (Aves: Passeriformes): transoceanic colonization of Africa from Australo-Papua. Journal of Biogeography 37:1767-1781. Jonsson, K. A., P. H. Fabre, R. E. Ricklefs, and J. Fjeldsa. 2011. Major global radiation of corvoid birds originated in the proto-Papuan archipelago. Proceedings of the National Academy of Sciences of the United States of America 108:2328-2333. Jonsson, K. A., M. Irestedt, J. Fuchs, P. G. P. Ericson, L. Christidis, R. C. K. Bowie, J. A. Norman, E. Pasquet, and J. Fjeldsa. 2008. Explosive avian radiations and

45 Evolutionary diversification and biogeography of parrots

multi-directional dispersal across Wallacea: evidence from the Campephagidae and other crown Corvida (Aves). Molecular Phylogenetics and Evolution 47:221-236. Joseph, L., A. Toon, E. E. Schirtzinger, and T. F. Wright. 2011. Molecular systematics of two enigmatic genera Psittacella and Pezoporus illuminate the ecological radiation of Australo-Papuan parrots (Aves: Psittaciformes). Molecular Phylogenetics and Evolution 59:675-684. Kassen, R. 2009. Toward a general theory of adaptive radiation: insights from microbial experimental evolution. Pp. 3-22. Year in Evolutionary Biology 2009. Blackwell Publishing, Oxford. Kozak, K. H., A. A. Larson, R. M. Bonett, and L. J. Harmon. 2005. Phylogenetic analysis of ecomorphological divergence, community structure, and diversification rates in dusky salamanders (Plethodontidae: Desmognathus). Evolution 59:2000-2016. Li, Z. X., and C. M. Powell. 2001. An outline of the palaeogeographic evolution of the Australasian region since the beginning of the Neoproterozoic. Earth- Science Reviews 53:237-277. Lohman, D. J., K. K. Ingram, D. M. Prawiradilaga, K. Winker, F. H. Sheldon, R. G. Moyle, P. K. L. Ng, P. S. Ong, L. K. Wang, T. M. Braile, D. Astuti, and R. Meier. 2010. Cryptic genetic diversity in "widespread" Southeast Asian bird species suggests that Philippine avian endemism is gravely underestimated. Biological Conservation 143:1885-1890. Losos, J. B. 2010. Adaptive radiation, ecological opportunity, and evolutionary determinism. American Naturalist 175:623-639. Losos, J. B., T. R. Jackman, A. Larson, K. de Queiroz, and L. Rodriguez-Schettino. 1998. Contingency and determinism in replicated adaptive radiations of island lizards. Science 279:2115-2118. Lovette, I. J., and E. Bermingham. 2000. c-mos variation in songbirds: Molecular evolution, phylogenetic implications, and comparisons with mitochondrial differentiation. Molecular Biology and Evolution 17:1569-1577. Lovette, I. J., E. Bermingham, and R. E. Ricklefs. 1999. Mitochondrial DNA phylogeography and the conservation of endangered Lesser Antillean Icterus orioles. Conserv. Biol. 13:1088-1096. Mace, G. M. 2004. The role of taxonomy in species conservation. Philos. Trans. R. Soc. Lond. Ser. B-Biol. Sci. 359:711-719. Magallon, S., and M. J. Sanderson. 2001. Absolute diversification rates in angiosperm clades. Evolution 55:1762-1780. Martin, H. A. 2006. Cenozoic climatic change and the development of the arid vegetation in Australia. Journal of Arid Environments 66:533-563. Mayr, G. 2010. Parrot interrelationships - morphology and the new molecular phylogenies. Emu 110:348-357. Mayr, G. 2011. Metaves, Mirandornithes, Strisores and other novelties - a critical review of the higher-level phylogeny of neornithine birds. Journal of Zoological Systematics and Evolutionary Research 49:58-76. McGlone, M. S. 2005. Goodbye Gondwana. Journal of Biogeography 32:739-740. Mettke-Hofmann, C., H. Winkler, and B. Leisler. 2002. The significance of ecological factors for exploration and neophobia in parrots. Ethology 108:249-272. Mittelbach, G. G., D. W. Schemske, H. V. Cornell, A. P. Allen, J. M. Brown, M. B. Bush, S. P. Harrison, A. H. Hurlbert, N. Knowlton, H. A. Lessios, C. M. McCain, A. R. McCune, L. A. McDade, M. A. McPeek, T. J. Near, T. D.

46 Chapter 2

Price, R. E. Ricklefs, K. Roy, D. F. Sax, D. Schluter, J. M. Sobel, and M. Turelli. 2007. Evolution and the latitudinal diversity gradient: speciation, extinction and biogeography. Ecol. Lett. 10:315-331. Miyaki, C. Y., S. R. Matioli, T. Burke, and A. Wajntal. 1998. Parrot evolution and paleogeographical events: mitochondrial DNA evidence. Molecular Biology and Evolution 15:544-551. Moss, S. J., and M. E. J. Wilson. 1998. Biogeographic implications of the Tertiary palaeogeographic evolution of Sulawesi and Borneo. Pp. 133-163 in R. Hall, and J. D. Holloway, eds. Biogeographic and geological evolution of SE Asia. Backhuys Publishers, Leiden. Moyle, R. G. 2005. Phylogeny and biogeographical history of Trogoniformes, a pantropical bird order. Biological Journal of the Linnean Society 84:725-738. Nosil, P., and T. E. Reimchen. 2005. Ecological opportunity and levels of morphological variance within freshwater stickleback populations. Biological Journal of the Linnean Society 86:297-308. O'Meara, B. C., C. Ane, M. J. Sanderson, and P. C. Wainwright. 2006. Testing for different rates of continuous trait evolution using likelihood. Evolution 60:922-933. Oliveros, C. H., and R. G. Moyle. 2010. Origin and diversification of Philippine bulbuls. Molecular Phylogenetics and Evolution 54:822-832. Overton, L. C., and D. D. Rhoads. 2004. Molecular phylogenetic relationships based on mitochondrial and nuclear gene sequences for the todies (Todus, Todidae) of the Caribbean. Molecular Phylogenetics and Evolution 32:524-538. Pacheco, M. A., F. U. Battistuzzi, M. Lentino, R. Aguilar, S. Kumar, and A. A. Escalante. 2011. Evolution of modern birds revealed by mitogenomics: timing the radiation and origin of major orders. Molecular Biology and Evolution 28:1927-1942. Penny, D., and M. J. Phillips. 2004. The rise of birds and mammals: are microevolutionary processes sufficient for macroevolution. Trends in Ecology & Evolution 19:516-522. Pepperberg, I. M. 2002. Cognitive and communicative abilities of grey parrots. Curr. Dir. Psychol. 11:83-87. Pereira, S. L., and A. J. Baker. 2006. A mitogenomic timescale for birds detects variable phylogenetic rates of molecular evolution and refutes the standard molecular clock. Molecular Biology and Evolution 23:1731-1740. Pereira, S. L., and A. J. Baker. 2008. DNA evidence for a Paleocene origin of the Alcidae (Aves: Charadriiformes) in the Pacific and multiple dispersals across northern oceans. Molecular Phylogenetics and Evolution 46:430-445. Pereira, S. L., K. P. Johnson, D. H. Clayton, and A. J. Baker. 2007. Mitochondrial and nuclear DNA sequences support a Cretaceous origin of Columbiformes and a dispersal-driven radiation in the Paleogene. Systematic Biology 56:656-672. Peterson, A. T. 2006. Taxonomy is important in conservation: a preliminary reassessment of Philippine species-level bird taxonomy. Bird Conservation International 16:155-173. Phillimore, A. B., and T. D. Price. 2008. Density-dependent cladogenesis in birds. PLoS. Biol. 6:483-489. Phillips, M. J., G. C. Gibb, E. A. Crimp, and D. Penny. 2010. Tinamous and moa flock together: mitochondrial genome sequence analysis reveals independent losses of flight among ratites. Systematic Biology 59:90-107.

47 Evolutionary diversification and biogeography of parrots

Poe, S., and A. L. Chubb. 2004. Birds in a bush: five genes indicate explosive evolution of avian orders. Evolution 58:404-415. Pratt, R. C., G. C. Gibb, M. Morgan-Richards, M. J. Phillips, M. D. Hendy, and D. Penny. 2009. Toward resolving deep Neoaves phylogeny: data, signal enhancement, and priors. Molecular Biology and Evolution 26:313-326. Price, S. A., R. Holzman, T. J. Near, and P. C. Wainwright. 2011. Coral reefs promote the evolution of morphological diversity and ecological novelty in labrid fishes. Ecol. Lett. 14:462-469. Price, S. A., P. C. Wainwright, D. R. Bellwood, E. Kazancioglu, D. C. Collar, and T. J. Near. 2010. Functional innovations and morphological diversification in parrotfish. Evolution 64:3057-3068. Price, T. 2008. Speciation in birds. Roberts and Company, Greenwood Village, CO. Pybus, O. G., and P. H. Harvey. 2000. Testing macro-evolutionary models using incomplete molecular phylogenies. Proceedings of the Royal Society B- Biological Sciences 267:2267-2272. Rabosky, D. L. 2006a. LASER: a maximum likelihood toolkit for detecting temporal shifts in diversification rates from molecular phylogenies. Evol Bioinform Online 2:257-260. Rabosky, D. L. 2006b. Likelihood methods for detecting temporal shifts in diversification rates. Evolution 60:1152-1164. Rabosky, D. L. 2009. Ecological limits and diversification rate: alternative paradigms to explain the variation in species richness among clades and regions. Ecol. Lett. 12:735-743. Rabosky, D. L., and I. J. Lovette. 2008. Density-dependent diversification in North American wood warblers. Proceedings of the Royal Society B-Biological Sciences 275:2363-2371. Raxworthy, C. J., M. R. J. Forstner, and R. A. Nussbaum. 2002. Chameleon radiation by oceanic dispersal. Nature 415:784-787. Ree, R. H., B. R. Moore, C. O. Webb, and M. J. Donoghue. 2005. A likelihood framework for inferring the evolution of geographic range on phylogenetic trees. Evolution 59:2299-2311. Ree, R. H., and S. A. Smith. 2008. Maximum likelihood inference of geographic range evolution by dispersal, local extinction, and cladogenesis. Systematic Biology 57:4-14. Ribas, C. C., R. Gaban-Lima, C. Y. Miyaki, and J. Cracraft. 2005. Historical biogeography and diversification within the Neotropical parrot genus Pionopsitta (Aves: Psittacidae). Journal of Biogeography 32:1409-1427. Ribas, C. C., L. Joseph, and C. Y. Miyaki. 2006. Molecular systematics and patterns of diversification in Pyrrhura (Psittacidae), with special reference to the picta- leucotis complex. Auk 123:660-680. Ribas, C. C., and C. Y. Miyaki. 2004. Molecular systematics in Aratinga parakeets: species limits and historical biogeography in the 'solstitialis' group, and the systematic position of Nandayus nenday. Molecular Phylogenetics and Evolution 30:663-675. Ribas, C. C., C. Y. Miyaki, and J. Cracraft. 2009. Phylogenetic relationships, diversification and biogeography in Neotropical Brotogeris parakeets. Journal of Biogeography 36:1712-1729. Ribas, C. C., R. G. Moyle, C. Y. Miyaki, and J. Cracraft. 2007a. The assembly of montane biotas: linking Andean tectonics and climatic oscillations to

48 Chapter 2

independent regimes of diversification in Pionus parrots. Proceedings of the Royal Society B-Biological Sciences 274:2399-2408. Ribas, C. C., E. S. Tavares, C. Yoshihara, and C. Y. Miyaki. 2007b. Phylogeny and biogeography of yellow-headed and blue-fronted parrots (Amazona ochrocephala and Amazona aestiva) with special reference to the South American taxa. Ibis 149:564-574. Richardson, K. C., and R. D. Wooller. 1990. Adaptation of the alimentary tract of some Australian lorikeets to a diet of pollen and nectar. Aust. J. Zool. 38:581- 586. Ricklefs, R. E. 2003. Global diversification rates of passerine birds. Proceedings of the Royal Society of London Series B-Biological Sciences 270:2285-2291. Ricklefs, R. E. 2004. A comprehensive framework for global patterns in biodiversity. Ecol. Lett. 7:1-15. Ricklefs, R. E., J. B. Losos, and T. M. Townsend. 2007. Evolutionary diversification of clades of squamate reptiles. J. Evol. Biol. 20:1751-1762. Rowley, I. 1998. Family Cacatuidae (cockatoos). Pp. 246-279 in J. Del Hoyo, A. Elliott, and J. Sargatal, eds. Handbook of the birds of the world. Lynx Edicions, Barcelona. Ruber, L., and R. Zardoya. 2005. Rapid cladogenesis in marine fishes revisited. Evolution 59:1119-1127. Rundell, R. J., and T. D. Price. 2009. Adaptive radiation, nonadaptive radiation, ecological speciation and nonecological speciation. Trends in Ecology & Evolution 24:394-399. Russello, M. A., and G. Amato. 2004. A molecular phylogeny of Amazona: implications for Neotropical parrot biogeography, taxonomy, and conservation. Molecular Phylogenetics and Evolution 30:421-437. Salomonsen, F. 1953. Taxonomic notes on the genus Prioniturus. Videnskabelige Meddelelser fra Dansk Naturhistorisk Forening 115:222-235. Sangster, G. 2000. Taxonomic stability and avian extinctions. Conserv. Biol. 14:579- 581. Sanmartin, I., and F. Ronquist. 2004. Southern hemisphere biogeography inferred by event-based models: plant versus animal patterns. Systematic Biology 53:216- 243. Schluter, D. 2000. The ecology of adaptive radiation. Oxford University Press, Oxford. Schluter, D., E. A. Clifford, M. Nemethy, and J. S. McKinnon. 2004. Parallel evolution and inheritance of quantitative traits. American Naturalist 163:809- 822. Schodde, R. 2006. Australia's bird fauna today - origins and development. in J. R. Merrick, M. Archer, G. M. Hickey, and M. S. Y. Lee, eds. Evolution and biogeography of Australasian vertebrates. Auscipub, Sidney. Schweizer, M., O. Seehausen, M. Güntert, and S. T. Hertwig. 2010. The evolutionary diversification of parrots supports a taxon pulse model with multiple trans- oceanic dispersal events and local radiations. Molecular Phylogenetics and Evolution 54:984-994. Schweizer, M., O. Seehausen, and S. T. Hertwig. in press. Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations. Journal of Biogeography.

49 Evolutionary diversification and biogeography of parrots

Seehausen, O. 2006. African cichlid fish: a model system in adaptive radiation research. Proceedings of the Royal Society B-Biological Sciences 273:1987- 1998. Shapiro, B., D. Sibthorpe, A. Rambaut, J. Austin, G. M. Wragg, O. R. P. Bininda- Emonds, P. L. M. Lee, and A. Cooper. 2002. Flight of the dodo. Science 295:1683-1683. Slack, K. E., C. M. Jones, T. Ando, G. L. Harrison, R. E. Fordyce, U. Arnason, and D. Penny. 2006. Early penguin fossils, plus mitochondrial genomes, calibrate avian evolution. Molecular Biology and Evolution 23:1144-1155. Smith, D. R., L. Villafuerte, G. Otis, and M. R. Palmer. 2000. Biogeography of Apis cerana F. and A. nigrocincta Smith: insights from mtDNA studies. Apidologie 31:265-279. Smith, G. A. 1975. Systematics of parrots. Ibis 117:18-68. Smith, T. B., R. Calsbeek, R. K. Wayne, K. H. Holder, D. Pires, and C. Bardeleben. 2005. Testing alternative mechanisms of evolutionary divergence in an African rain forest passerine bird. J. Evol. Biol. 18:257-268. Smith, T. B., C. J. Schneider, and K. Holder. 2001. Refugial isolation versus ecological gradients. Genetica 112:383-398. Soltis, D. E. 2007. Reconstructing the three of life: taxonomy and systematics of species rich taxa. Evolution 61:3007-3011. Tavares, E. S., A. J. Baker, S. L. Pereira, and C. Y. Miyaki. 2006. Phylogenetic relationships and historical biogeography of Neotropical parrots (Psittaciformes: Psittacidae: Arini) inferred from mitochondrial and nuclear DNA sequences. Systematic Biology 55:454-470. Tennyson, A. J. D. 2010. The origin and history of New Zealand's terrestrial vertebrates. N. Z. J. Ecol. 34:6-27. Thomas, G. H., C. D. L. Orme, R. G. Davies, V. A. Olson, P. M. Bennett, K. J. Gaston, I. P. F. Owens, and T. M. Blackburn. 2008. Regional variation in the historical components of global avian species richness. Global Ecology and Biogeography 17:340-351. Tokita, M., T. Kiyoshi, and K. N. Armstrong. 2007. Evolution of craniofacial novelty in parrots through developmental modularity and heterochrony. Evolution & Development 9:590-601. Treplin, S., R. Siegert, C. Bleidorn, H. S. Thompson, R. Fotso, and R. Tiedemann. 2008. Molecular phylogeny of songbirds (Aves: Passeriformes) and the relative utility of common nuclear marker loci. Cladistics 24:328-349. Upchurch, P. 2008. Gondwanan break-up: legacies of a lost world? Trends in Ecology & Evolution 23:229-236. Vences, M., D. R. Vieites, F. Glaw, H. Brinkmann, J. Kosuch, M. Veith, and A. Meyer. 2003. Multiple overseas dispersal in amphibians. Proceedings of the Royal Society of London Series B-Biological Sciences 270:2435-2442. Vermeij, G. J. 1995. Economics, volcanos, and phanerozoic revolutions. Paleobiology 21:125-152. Vermeij, G. J. 2001. Innovation and evolution at the edge: origins and fates of gastropods with a labral tooth. Biological Journal of the Linnean Society 72:461-508. Voelker, G., S. Rohwer, D. C. Outlaw, and R. C. K. Bowie. 2009. Repeated trans- Atlantic dispersal catalysed a global songbird radiation. Global Ecology and Biogeography 18:41-49.

50 Chapter 2

Waters, J. M., and D. Craw. 2006. Goodbye Gondwana? New Zealand biogeography, geology, and the problem of circularity. Systematic Biology 55:351-356. Weir, J. T. 2006. Divergent timing and patterns of species accumulation in lowland and highland Neotropical birds. Evolution 60:842-855. White, C. M. N., and M. D. Bruce. 1986. The birds of Wallacea: (Sulawesi, the Moluccas and Lesser Sunda Islands, Indonesia). An annotated checklist. British Ornithologists' Union. White, N. E., M. J. Phillips, M. T. P. Gilbert, A. Alfaro-Nunez, E. Willerslev, P. R. Mawson, P. B. S. Spencer, and M. Bunce. 2011. The evolutionary history of cockatoos (Aves: Psittaciformes: Cacatuidae). Molecular Phylogenetics and Evolution 59:615-622. Wiens, J. J. 2011. The causes of species richness patterns across space, time, and clades and the role of “ecological limits". Q. Rev. Biol. 86:75-96. Woodruff, D. S., M. P. Carpenter, E. S. Upatham, and V. Viyanant. 1999. Molecular phylogeography of Oncomelania lindoensis (Gastropoda: Pomatiopsidae), the intermediate host of Schistosoma japonicum in Sulawesi. J. Molluscan Stud. 65:21-31. Wright, T. F., and C. R. Dahlin. 2007. Pair duets in the yellow-naped amazon (Amazona auropalliata): phonology and syntax. Behaviour 144:207-228. Wright, T. F., E. E. Schirtzinger, T. Matsumoto, J. R. Eberhard, G. R. Graves, J. J. Sanchez, S. Capelli, H. Mueller, J. Scharpegge, G. K. Chambers, and R. C. Fleischer. 2008. A multilocus molecular phylogeny of the parrots (Psittaciformes): support for a Gondwanan origin during the Cretaceous. Molecular Biology and Evolution 25:2141-2156. Yoder, A. D., M. M. Burns, S. Zehr, T. Delefosse, G. Veron, S. M. Goodman, and J. J. Flynn. 2003. Single origin of Malagasy Carnivora from an African ancestor. Nature 421:734-737. Yoder, A. D., and M. D. Nowak. 2006. Has vicariance or dispersal been the predominant biogeographic force in Madagascar? Only time will tell. Annual Review of Ecology Evolution and Systematics 37:405-431. Yoder, J. B., E. Clancey, S. Des Roches, J. M. Eastman, L. Gentry, W. Godsoe, T. J. Hagey, D. Jochimsen, B. P. Oswald, J. Robertson, B. A. J. Sarver, J. J. Schenk, S. F. Spear, and L. J. Harmon. 2010. Ecological opportunity and the origin of adaptive radiations. J. Evol. Biol. 23:1581-1596. Zachos, J., M. Pagani, L. Sloan, E. Thomas, and K. Billups. 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science 292:686-693.

Chapter 3

THE EVOLUTIONARY DIVERSIFICATION OF PARROTS SUPPORTS A TAXON PULSE MODEL WITH MULTIPLE TRANS-OCEANIC DISPERSAL EVENTS AND LOCAL RADIATIONS

Manuel Schweizera, Marcel Günterta, Ole Seehausenb,c, Stefan T. Hertwiga

aNaturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, CH 3005 Bern, Switzerland bAquatic Ecology and Macroevolution, Institute of Ecology & Evolution, University of Bern, Baltzerstrasse 6, CH 3012 Bern cFish Ecology and Evolution, EAWAG, Seestrasse 79, CH 6047 Kastanienbaum, Switzerland

Corresponding author: Manuel Schweizer, Naturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, CH 3005 Bern, Switzerland, e-mail [email protected], tel +41 31 350 71 11, fax +41 31 350 74 99

Published as: Schweizer, M., O. Seehausen, M. Güntert, and S. T. Hertwig. 2010. The evolutionary diversification of parrots supports a taxon pulse model with multiple trans-oceanic dispersal events and local radiations. Molecular Phylogenetics and Evolution 54:984- 994. (http://www.sciencedirect.com/science/article/pii/S1055790309003467)

55 Molecular Phylogenetics and Evolution 54 (2010) 984-994

Contents lists available at ScienceDirect

Molecular Phylogenetics and Evolution

ELSEVIER journa l homepage: www.elsevier.com/locate/ympev

The evolutionary diversification of parrots supports a taxon pulse model with multiple trans-oceanic dispersal events and local radiations

Manuel Schweizera··. Ole Seehausen b.c. Marcel Giinterta. Stefan T. Hertwiga

• Naturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, CH 3005 Bern, Switzerland b Aquatic Ecology and Macroevolution, Institute of Ecology & Evolution, Uni11ersity of Bern, Baltzerstrasse 6, CH 3012 Bern, Switzerland

ARTICLE INFO ABSTRACT

Article history: Vicariance is thought to have played a major role in the evolution of modern parrots. However, as the Received 12 June 2009 relationships especially of the African taxa remained mostly unresolved, it has been difficult to draw firm Accepted 17 August 2009 conclusions about the roles of dispersal and vicariance. Our analyses using the broadest taxon sampling Available online 21 August 2009 of old world parrots ever based on 3219 bp of three nuclear genes revealed well resolved and congruent phylogenetic hypotheses. Agapomis ofAfrica and Madagascar was found to be the sister group to Loriru/us Keywords: of Australasia and Indo Malayasia and together they clustered with the Australasian Loriinae, Cyclo Parrots psittacini and Melopsittacus. Poicepha/us and Psittacus from mainland Africa formed the sister group of Biogeography the Neotropical Arini and Coracopsis from Madagascar and adjacent islands may be the closest relative Dispersal ViG1riance of Psittrichas from New Guinea. These biogeographic relationships are best explained by independent col Nuclear genes onization of the African continent via trans oceanic dispersal from Australasia and Antarctica in the Taxon pulse model Paleogene following what may have been vicariance events in the late Cretaceous and{or early Paleogene. Our data support a taxon pulse model for the diversification of parrots whereby trans oceanic dispersal played a more important role than previously thought and was the prerequisite for range expansion into new continents. © 2009 Elsevier Inc. All rights reserved

1. Introduction 1975; Wright et al., 2008). Recent molecular phylogenetic studies agree that the New Zealand taxa Kea (Nestor notabilis) and Kakapo Species richness is highly unevenly distributed among taxo (Strigops habroptilus) are the sister group of all other parrots (de nomic groups and studying the most diverse clades is often associ Kloet and de Kloet, 2005; Tavares et al., 2006; Tokita et al., ated with major systematic challenges (Soltis, 2007). As species 2007; Wright et al., 2008). The mainly Australasian Cacatuidae is richness is influenced by both the biological traits of the organisms the taxon that branches off next (Tokita et al., 2007; Wright themselves and their environment (Newton, 2003), the study of et al., 2008) and the monophyly of the new world parrots is well species rich groups can provide insights into the mechanisms of supported (de Kloet and de Kloet, 2005; Tavares et al., 2006; Tokita speciation and accumulation of species diversity, as well as into et al., 2007; Wright et al., 2008 ~ These phylogenetic and biogeo historical biogeography. For a long time, vicariance approaches graphic relationships have been interpreted as supporting a have dominated historical biogeography; recent works suggest vicariance speciation model and as modern parrots are mostly however, that dispersal is an important process in speciation and non migratory (Collar, 1998), it has been suggested that their the build up of regional fauna and that the importance of oceanic diversification pattern and evolutionary history may be less influ dispersal has been strongly underestimated (de Queiroz, 2005; enced by dispersal than in other avian groups. This has been taken Cowie and Holland, 2006; McGlone, 2005; Yoder and Nowak, in turn to suggest that vicariance was a major force in parrot diver 2006). sification following the break up of Gondwana (cf. Wright et al., With 353 species currently recognized, parrots represent one of 2008 ). There is still some controversy about the timescale of the the most species rich groups of birds (Collar, 1998; Rowley, 1998). evolution of parrots similar to other groups of modem birds with They have radiated extensively in the Neotropical and Australasian the fossil record suggesting a Cenozoic origin for most lineages region and to a smaller extent in the Afrotropical and Indo Mala whereas molecular genetic approaches date the origin of the same yan region (Collar, 1998; Cracraft, 2001; Rowley, 1998; Smith, lineages in the Cretaceous before the K Pg boundary (e.g. Brown et al., 2007, 2008; Cooper and Penny, 1997; Cracraft, 2001; Ericson • Corresponding author. Fax: +41 31 350 74 99. et al., 2006; Pratt et al., 2009). Although no representative of crown E-mail address: [email protected] (M. Schweizer). group Psittaciformes is known from Paleogene fossil deposits

(Mayr, 2009), molecular dating analyses calibrated with external et al. (2008) revealed a sister group relationship between the Psit dating points and/or biogeographic approaches suggest a Gondwa taciformes and the Passeriformes with both forming a clade to nan origin of the parrots during the late Cretaceous (de Kloet and gether with the Falconidae. We selected therefore two suboscine de Kloet, 2005; Tavares et al., 2006; Wright et al., 2008). However, Passeriformes (Pipra and Pitta), two oscine Passeriformes (Corvus a recent molecular dating of the origin of the major Neoaves lin and Picathartes) as well as one Falco as outgroups and rooted trees eages based on complete mitochondrial genomes and using two with the latter taxon (Table 1). fossil calibration points suggested that the Kakapo split from the remaining parrots only after the K Pg boundary (Pratt et al., 2.2. Laboratory methods 2009). Wright et al. (2008) compared two internal calibration points to estimate the divergence times within the parrots. First, Frozen tissues of birds collected in the wild or from captivity as the separation of New Zealand from Gondwana 82 85 MYA assum well as feathers sampled from captive birds were used for the iso ing that this date coincides with the separation of the New Zealand lation of genomic DNA (Table 1). Total genomic DNA was isolated Kea and Kakapo from all other Parrots. This calibration point was using peqGOLD tissue DNA Mini Kit following manufacturer’s rule also used by other authors (Ribas et al., 2007; Tavares et al., (Peqlab). Partial sequences of the three nuclear genes c mos, RAG 1 2006). Second, a minimum age of 50 MYA for the same initial split and Zenk (second exon) were amplified with Polymerase chain based on fossil records of stem parrots in Europe accounting for the reaction (PCR) using different sets of published primer sequences hypothesis of a radiation of modern bird orders during the Paleo (Table 2). PCR reaction volumes were 20 ll containing 10 ll PCR gene. Wright et al. (2008) considered the latter scenario less likely Master Mix S (c mos) or PCR Master Mix Y (RAG1, ZENK) (Peqlab), as it would have required oversea dispersal to New Zealand, Mad 2 3 ll genomic DNA, 2 ll of each primer with a concentration of agascar and South America from Australia. 10 lM and 3 4 ll ddH O. PCR was performed on a Techne TC However, to assess the likelihood of possible dispersal or colo 2 512 thermo cycler. Amplifications of c mos was performed with nization routes and vicariance events, we need to answer a few the following parameters: initial denaturation of 94 °C for 2 min critical phylogenetic questions. Especially the phylogenetic and followed by 33 cycles of denaturation at 90 °C for 30 s, annealing biogeographic history of the recent African parrots remains contro at 55 °C for 30 s, and extension at 72 °C for 1 min, with a final versial. African parrots include four endemic genera, Coracopsis (2 extension at 72 °C for 5 min. The PCR reaction profile published species) from Madagascar, the Comoro Islands and the Seychelles, by Groth and Barrowclough (1999) was used for RAG1 with the ini the African and Malagasy Agapornis (9 species) as well as the Afri tial denaturation step reduced to 2 min. For Zenk, the PCR profile of can genera Poicephalus (9 species) and Psittacus (1 species). Further, Chubb (2004a) was used with the annealing temperature set to Psittacula echo occurs on Mauritius and Africa has an endemic sub 53.5 °C. PCR products were examined by gel electrophoresis to species of the otherwise Asian Psittacula krameri (Collar, 1998). confirm the amplification of the target fragment. PCR products Various phylogenetic relationships have been proposed for African were either excised from gels and cleaned using the WizardÒ SV parrots and it was unclear how the different groups reached this Gel and PCR Clean UP System (Promega) or directly purified with continent (de Kloet and de Kloet, 2005). the above mentioned kit or with the peqGOLD MicroSpin Cycle The aim of the present study is therefore to investigate the phy Pure Kit (Peqlab). To increase the quantity of DNA for problematic logenetic and biogeographic history of parrots to infer the sister samples, the products of two independent PCR runs were put to group relationship especially of African parrots in order to test gether before the cleaning or a second PCR was performed after the role of dispersal vs. vicariance in their diversification. Com the cleaning. Sequencing was carried out with Microsynth AG pared to previous studies, we increased the taxon sampling and (Balgach, Switzerland) using the same primers as for amplification. chose a different set of markers consisting of the three nuclear All three genes were sequenced from both sides leading to com exons c mos, Rag 1 and Zenk. Rag 1 is a single copy nuclear exon plete overlapping fragments for Rag 1 and c mos and an overlap involved in recombination and has been widely applied to phylo ping fragment of about 600 bp for Zenk. Sequencing files were genetic studies at genus level within avian orders and families checked with Chromas (Technelysium Pty., Ltd.) and ambiguities (e.g. Groth and Barrowclough, 1999; Paton et al., 2003; Barker were assigned standard IUB codes. The Alignment of the sequences et al., 2004; Griffiths et al., 2004; Tavares et al., 2006; Pereira was done manually with BioEdit 7.0.5.2 (Hall, 1999). We checked et al., 2007; Treplin et al., 2008). Zenk is a single copy nuclear tran individual Sequences and the whole alignment further for quality scription factor that has utility to resolve divergences among by searching for apparent stop codons after the translation of se (Chubb, 2004a; Long and Salbaum, 1998) as well as within avian quences into amino acids and for indels that were not a multiple orders (Chubb, 2004b; Treplin et al., 2008). Finally, the single copy of three bases. nuclear proto oncogene c mos has been used successfully for resolving phylogenetic relationships among intermediate and more distantly related species of birds (Lovette and Bermingham, 2.3. Sequence and phylogenetic analyses 2000; Overton and Rhoads, 2004). A Chi square test of homogeneity of base frequencies across taxa as implemented in PAUP* (Swafford, 2001) was used to test 2. Materials and methods the variation of base frequencies between taxa for each gene. Data sets of the different genes were tested pairwise for heterogeneity 2.1. Taxon sampling and character sampling using the incongruent length difference test (ILD) (Farris et al., 1995) implemented in PAUP* to asses combinability of the differ We concentrated on sampling parrots from the old world as the ent data sets with taxa missing from either data set excluded (heu monophyly of the new world taxa is well supported (see above). ristic search, 1000 replicates, number of max trees limited to We sampled 60 old world species plus one Neotropical species. 1000). Additionally, we tested the potential loss of information in We analyzed between two and five species from those genera that the third codon position of all three genes due to substitution were of particular interest because their sister group relationships saturation estimating the index of substitution saturation (Iss) by could not be resolved in previous studies (five species of Agapornis, Xia et al. (2003) with DAMBE (Xia and Xie, 2001). The phylogenetic four of Neophema, three of Loriculus and two each of Micropsitta, informativeness per site was calculated as the quotient of the num Coracopsis and Poicephalus). A phylogenomic study by Hackett ber of parsimony informative sites and the sequence length to 986 M Schweizer et al./ Moleculor Phylogenetics and Evolution 54 (2010) 984- 994

Table 1 Species sampled. Museum and collection number, Gi!nBank Accession numbers for the three genes analyzed and sample type.

Species Museum/collection Collection nb. c-mos R.3g-1 Zenk type Agapomis canus NMBE 1056201 GQ505083 GQ505191 GQ505138 lissue Agapomis fischeri NMBE 1056202 GQ505084 GQ505192 GQ505139 lissue Agapomis lilianae NMBE 1056205 GQ505087 lissue Agapomis nigrigenis NMBE 1056203 GQ505085 GQ505193 GQ505140 lissue Agapomis roseicoUis NMBE 1056204 GQ505086 GQ505194 GQ505141 lissue Alisterus chloropterus NMBE 1056207 GQ505091 GQ505199 GQ505145 lissue Alister us scapu loris NMBE 1056206 GQ505090 GQ505198 GQ505144 lissue AprosmiausjonquiUaceus NMBE 1056208 GQ505092 GQ505200 GQ505146 lissue Barnardius zonarius NMBE 1056210 GQ505095 GQ505203 GQ505149 lissue Cacatua galerita fitzroyi NMBE 1056235 GQ505120 GQ505231 GQ505173 lissue Cacatua moluccensis NMBE 1056236 GQ505121 GQ505232 GQ505174 lissue Calyptorhynchus funereus NMBE 1056233 GQ505118 GQ505229 lissue Calyptorhynchus lotirostris NMBE 1056234 GQ505119 GQ505230 GQ505172 lissue cnarmosyna pulcheUa NMBE 1056241 GQ505126 GQ505237 GQ505179 lissue Coraropsis nigra GQ505114 GQ505224 Feathers Coraropsis vasa UWBM 85986/2004-001 GQ505113 GQ505223 GQ505167 lissue Cyanoramphus auriceps NMBE 1056221 GQ505104 GQ505213 GQ505158 lissue Cyanoramphus novaezelandiae NMBE 1056220 GQ505103 GQ505212 GQ505157 lissue Cyclopsitta diophthalma GQ505130 Feathers Eclecrus rorarus NMBE 1056248 GQ505135 GQ505244 GQ505187 lissue Eos cyanogenia NMBE 1056237 GQ505122 GQ505233 GQ505175 lissue Eunymphicus comutus comutus NMBE 1056223 GQ505106 GQ505215 GQ505159 lissue Eunymphicus comutus uvaeensis NMBE 1056224 GQ505107 GQ505216 GQ505160 lissue Lathamus disrolor NMBE 1056219 GQ505102 GQ505211 GQ505156 lissue Loriculus catamene ZMUC 115584 GQ505088 GQ505195 GQ505142 lissue Loriculus galgulus UWBM 73841 /2002-006 GQ505089 GQ505196 lissue Loriculus philippensis ZMUC 130608 GQ505197 GQ505143 lissue Lorius garrulus NMBE 1056240 GQ505125 GQ505236 GQ505178 lissue Melopsittacus undularus UWBM 60748/1998-068 GQ505222 GQ505166 lissue Micropsitta finschii tristrarni UWBM 66040/2000--022 GQ505128 GQ505240 GQ505182 lissue Micropsitta pusio UWBM 67905/2001-054 GQ505129 GQ505241 GQ505183 lissue Neophema chrysogaster NMBE 1056227 GQ505110 GQ505219 GQ505163 lissue Neophema chrysostoma NMBE 1056228 GQ505111 GQ505220 GQ505164 lissue Neophema pulchello NMBE 1056226 GQ505109 GQ505218 GQ505162 lissue Neophema splendida NMBE 1056225 GQ505108 GQ505217 GQ505161 lissue Neopsephotos bourkii UWBM 57542/1996-109 GQ505112 GQ505221 GQ505165 lissue Nestor notabilis NMBE 1056242 GQ505238 GQ505180 lissue Platycercus caledonicus NMBE 1056212 GQ505097 GQ505205 GQ505151 lissue Platycercus eximius NMBE 1056213 GQ505206 GQ505152 lissue Platycercus fiaveolus NMBE 1056215 GQ505099 GQ505208 lissue Platycercus venustus NMBE 1056214 GQ505098 GQ505207 GQ505153 lissue Poicephalus rufiventris NMBE 1056230 GQ505116 GQ505226 GQ505169 lissue Poicephalus senegalus NMBE 1056231 GQ505227 GQ505170 lissue Polytelis alexandrae NMBE 1056209 GQ505093 GQ505201 GQ505147 lissue Polytelis anthopeplus NMBE 1056657 GQ505094 GQ505202 GQ505148 lissue Prionirurus discurus NMBE 1056246 GQ505133 GQ505186 lissue Prionirurus luconensis NMBE 1056247 GQ505134 lissue Prosopeia tabuensis NMBE 1056252 GQ505105 GQ505214 lissue Psephotus dissimilis NMBE 1056218 GQ505101 GQ505210 GQ505155 lissue Psephotus varius NMBE 1056217 GQ505100 GQ505209 GQ505154 lissue Psittacula eupatria NMBE 1056250 GQ505137 GQ505246 GQ505189 lissue Psittaculirostris desmarestii NMBE 1056244 GQ505131 GQ505242 GQ505184 lissue Psittaculirostris edwardsii NMBE 1056245 GQ505132 GQ505243 GQ505185 lissue Psittacus erithacus erithacus NMBE 1056229 GQ505115 GQ505225 GQ505168 lissue Psitteuteles goldiei NMBE 1056239 GQ505124 GQ505235 GQ505177 lissue Psittinus q~murus NMBE 1056251 GQ505247 GQ505190 lissue Psittrichas fulgidus NMBE 1056243 GQ505127 GQ505239 GQ505181 lissue Purpureirephalus spurius NMBE 1056211 GQ505096 GQ505204 GQ505150 lissue Tanygnathus megalorhynchus NMBE 1056249 GQ505136 GQ505245 GQ505188 lissue Trichoglossus johnstoniae NMBE 1056238 GQ505123 GQ505234 GQ505176 lissue Tridaria malachitacea NMBE 1056232 GQ505117 GQ505228 GQ505171 lissue Pipra coronata AY056951 AY057020 AF492518 Pitta AY056952 AY057021 EF568299 Corvus rorone AY056918 AY056989 EF568306 Picathartes gymnocephalus AY056950 AY057019 EF568314 Falco AY447974 AY461399 AF490155 account for the different number of characters sampled from each mixed model approach and evaluated the significance of six differ gene (Treplin et al., 2008). ent biologically relevant ways to partition our data into gene and/ Phylogenetic analyses were conducted using model based ap or codon positions. Separate models for the sequences of the differ proaches (Bayesian inference BI and maximum likelihood ML) ent data partitions were evaluated with Mr. Modeltest 2.3 (Nyland and maximum parsimony analyses (MP). Bl of the phylogenetic er, 2004) using the Akaike information criteria (Akaike, 1974). The relationships was done with MrBayes 3.1 (Huelsenbeck and relevance of the different partitions was evaluated with the Bayes Ronquist, 2001; Ronquist and Huelsenbeck, 2003 ). We chose a Factor (BF) (Brown and Lemmon, 2007; Kass and Raftery, 1995) and M. Schweizer et al / Molecular Phylogenetics and Evolution 54 (2010) 984- 994 987

Table 2 search, 10 random taxon addition replicates, TBR branch swap Primers used for the amplification of the three genes analyzed in this study. ping, number of max trees limited to 100). MP search was con Gene Primer name References ducted for the combined data set and each gene separately. c-mos Clades were considered as supported by our analyses when boot F944 Cooper and Penny (1997) strap values were ;;,: 70% (Hillis and Bull, 1993) and clade credibility R1550or05 Overton and Rhoads (2004) values for the Bl ;;,: 0.95 (Huelsenbeck and Ronquist, 2001 ). R1550hb99 Hughes and Baker (1999) Rag-1 2.4. Reconstruction of ancestral area states R8 Groth and Barrowclough (1999) R11B Groth and Barrowclough (1999) R17 Groth and Barrowclough (1999) We choose a maximum likelihood approach to infer the ancestral R18 Groth and Barrowclough (1999) areas where the different groups of parrots have originated based on Zenk the tree topology obtained with the BI and ML analysis. We inferred ZlF Chubb (2004a) for each node the area assignment that maximizes the probability of Z9R Chubb (2004a) arriving at the observed areas in the terminal taxa given a stochastic Markov model of evolution (Lewis, 2001) as implemented in Mes quite 2.5 (build j77) (Maddison and Maddison, 2006, 2008). The cur the harmonic mean calculated by MrBayes was used as an estima rent distribution areas of the different species were categorized in tion of the marginal likelihood of the data (Kass and Raftery, five discrete character states based on the following biogeographic 1995 ). A more complicated model (i.e. more partitions) was favored realms (Newton, 2003 ): Afro tropical, Australasian, lndo Malayan compared to a simpler model if 21nBF was greater than 10 (Brown and Neotropical. We further included Madagascar and adjacent is and Lemmon, 2007 ). Base frequencies, rate matrix, shape parameter lands as a discrete character state. The distribution areas assigned and proportion ofinvariable sites were allowed to vary between par to the different species follow Collar (1998) and Rowley (1998 ). Lori titians. For each data partition, we ran two independent runs of culus catamene occurs on Sangihe in Wallacea, the transition zone Metropolis coupled Markov chain Monte Carlo analyses each con between the lndo Malayan and the Australasian realms. However, sisting of one cold chain and three heated chains with default tern it is part of a superspecies complex which contains taxa of New Gui perature of 0.2. The chains were run for 5 million generations with nea, the Bismarck Archipelago, Halmahera, Sulawesi and adjacent sampling every 100 generations and the first 25% of samples were island and was therefore assigned to the Australasian region. Our discarded as burn in (12,500trees) and we checked that the average species sampling covered the whole distribution areas of the differ standard deviation of split frequencies converged towards zero. ent groups with the exception that several species of the Loriinae oc For the maximum likelihood search, we used RAxML 7.0.4 (Sta cur additionally in the Oceanian region, one Cockatoo species enters matakis, 2006) running the program on the Web server with 100 the lndo Malayan region and that within the Psittaculini, Psittacula rapid bootstrap inferences with all free model parameters esti krameri has a subspecies occurring in Africa and Psittacula echo oc mated by the software (Stamatakis et al., 2008). We partitioned curs on Mauritius. our alignment into three genes with splitting each gene into three codon positions (9 partitions) as this partition was chosen as the 3. Results best model in our Bl (see below) and allowed RAxML to estimate an individual model of nucleotide substitution for each partition. 3.1. Sequence characteristics Analyses using the maximum parsimony criterion were con ducted in PAUP* (heuristic search, 1000 random taxon addition The final alignment was 3219 bp in length, consisting of 603 bp replicates, TBR branch swapping). Gaps were treated as fifth char for c mos, 1461 bp for Rag 1 and 1155 bp for Zenk (Table 3). It con acter state. Nodal support was estimated with a maximum parsi tained one indel of four amino acids for c mos, one indel of three mony bootstrap analysis (1000 pseudo replicates, heuristic amino acids and one indel of one amino acid for Rag 1 as well as

Table3 Sequence characteristics and model parameters for the three genes and their different partitions analyzed. Base substitution rates are the mean values obtained with MrBayes.

c-<:mos Rag-1 Zenk comb. data set 1&2 2 3 1&2 2 3 1&2 2 3 1&2 2 3 pos pos pos pos pos pos pos pos pos pos pos pos pos pos pos pos Size (hp) 603 1461 1155 3219 %A 024 026 028 0.17 0.32 0.33 0.34 0.35 029 024 0.30 027 025 0.17 0.28 0.31 0.31 0.30 024 %C 024 022 023 0.29 020 021 020 0.19 021 0.35 0.27 0.35 0.43 0.38 0.27 024 026 028 027 %G 0.30 0.37 0.30 0.29 022 027 024 020 020 020 0.20 0.18 0.15 023 0.22 026 023 0.19 023 n 022 0.16 020 0.25 025 0.19 023 027 0.30 020 0.23 021 0.17 023 0.23 0.19 021 023 026 Model GTR HKY HKY K80 HKY GTR GTR GTR GTR HKY HKY HKY GTR GTR GTR HKY HKY GTR GTR HKY 1i/Tv ratio 3.06 4.09 5.85 3.74 4.16 3.32 2.14 3.57 2.32 3.91 A- C O.D7 O.D7 0.13 0.09 0.04 O.D7 O.D7 O.D7 0.08 0.04 A- G 0.31 0.34 029 0.33 0.36 0.48 0.43 0.52 0.39 0.41 A-T 0.04 0.04 0.06 0.04 0.04 0.04 0.06 0.05 0.04 0.04 C-C 0.05 O.D7 0.08 0.10 0.13 0.09 0.18 0.04 0.08 0.10 C-T 0.49 0.43 0.40 0.40 0.39 021 021 028 0.36 0.37 G-T 0.04 0.04 0.05 0.04 0.04 0.10 0.05 0.04 0.05 0.03 Prop Inv 0.42 0.72 0 0 0 0.00 0.00 0.00 0.48 0.00 0.00 0.00 0.76 0.86 0.00 0.38 0.45 0.58 0.66 0.00 Gamma 1.03 - 0.10 0.11 126 0.44 0.45 027 0.62 123 0.31 0.23 - 1.62 0.97 0.87 0.79 0.64 1.03 Parslnf 128 304 174 606 inormativeness 021 021 0.15 0.19 per site Resolved nodes 102 108 93 123 in MP search ( nb) 988 M. Schweizer et al. / Molecular Phylogenetics and Evolution 54 (2010) 984–994 four indels of one amino acid and each one indel of three and two taxa could be resolved with large confidence. Different models of amino acids for Zenk. Due to missing data at the end of some se nucleotide evolution were selected for the different partitions of quences, their length ranged from 502 to 603 bp (88%>580 bp) our data set (Table 3). The six partitions tested with BI yielded al for c mos, 701 1461 bp (79%>1300 bp) for Rag 1 and 763 most identical trees with only small changes in the value of the 1155 bp (95%>1075 bp) for Zenk. The translation into amino acids clade credibility intervals. The only topological difference con did not reveal any unexpected stop codons and the indels were a cerned the placement of the clade of Coracopsis and Psittrichas in multiple of three bases. Both the Chi square test of homogeneity the 50% majority rule consensus tree of the concatenated data of base frequencies and the pairwise incongruence length differ set compared to the other partitions. After calculation of the Bayes ence test revealed no significant heterogeneity. Substantial satura Factors, the partitioning into three genes with splitting each gene tion at the third codon position of each gene did not influence our into three codon positions (nine partitions) was chosen as the phylogenetic inference as tested by Iss statistics (all P < 0.05). best fit model. The 50% majority rule consensus tree of this analy sis was identical in topology to the best scoring maximum likeli 3.2. Phylogenetic analysis hood tree (Fig. 1). As no significant heterogeneity between the three genes was re The model based approaches and the MP analysis yielded highly vealed with the pairwise incongruence length difference test, MP congruent trees and the sister group relationships of most African analysis was performed with the concatenated data set where

Fig. 1. 50% Majority-rule consenus tree of the Bayesian inference with clade credibility values and bootstrap values of the ML likelihood tree indicated at nodes. *Indicate clade credibility values of 1 or bootstrap values of 100. The notations for the different clades used in the text based on traditional taxonomy are indicated to the right of the tree. M. Schweizer et al. / Molecular Phylogenetics and Evolution 54 (2010) 984–994 989

120 trees were equally parsimonious with 123 resolved nodes in of ((Loriinae + Melopsittacus) + Cyclopsittacini). Coracopsis from the strict consensus tree (Fig. 2). Apart from different numbers of Madagascar and adjacent islands clustered with Psittrichas from resolved nodes, there were no conﬂicts among the topologies of New Guinea in the BI and ML analysis, but this clade was not ro the strict consensus trees for the single markers and of the com bustly supported. The sister group position of the Micropsittini to bined data set (Table 3) and the latter did not show any conﬂict the Psittaculini group A was well supported in the model based to the BI and ML tree. approaches. As expected, the sister group position of Nestor to the remaining parrots was highly supported overall (we did not include Strigops 3.3. Reconstruction of ancestral area states in our sampling). The next group branching off were the Cacatui dae. A clade consisting of ((Poicephalus + Psittacus)+Triclaria) re Our ancestral area reconstruction revealed that several major ceived robust support throughout, thus revealing a sister group trans oceanic dispersal events must be considered to explain the relationship between two of the endemic African genera and the current distribution patterns of parrots (see below). The Markov Neotropical parrots. A clade containing all Platycercini, the Lorii ML reconstruction of ancestral areas indicated that the common nae, the Cyclopsittacini as well as Loriculus and Agapornis was sup ancestor of all extant parrots occurred in Australasia (Fig. 3). The ported by the BI and ML analysis. Within this clade, the monophyly common ancestor of Psittrichas and Malagassy Coracopsis as well of the analyzed Platycercini group A (cf. Fig. 1) was supported over as of the ‘‘Loricoloriinae” lived in Australasia, and so did the com all, as was a clade containing Agapornis, Loriculus, the Loriinae, the mon ancestor of Loriculus and the African and Malagassy Agapornis. Cyclopsittacini and Melopsittacus. Within the latter clade, the clus However, the ancestral areas of the common ancestor of Agapornis ter of the African Agapornis and the Indo Malayan/Australasian and of that of the clade containing the Arini, Poicephalus and Psitta Loriculus was the sister group of an Australasian clade consisting cus could not be unambiguously resolved.

Fig. 2. Strict consensus tree of the maximum parsimony analysis of the concatenated data set with bootstrap values indicated at each node. 990 M. Schweizer et al. / Molecular Phylogenetics and Evolution 54 (2010) 984–994

Fig. 3. Cladogram with Markov-ML reconstruction of ancestral areas. Pie charts at each node represent the proportion of the total likelihood received by each biogeographic region as the ancestral area of a given clade. The present distribution of the taxa is indicated by the color of the circles at the tips of the tree. Inferred dispersal events are indicated at the clades concerned. The colonization of Africa from Indo-Malaysia by Psittacula krameri is also indicated, although this species is not included in our sampling.

4. Discussion to reconstruct phylogenies at least for regions in the tree with clo sely spaced branching events like at the base of Neoaves (Choj 4.1. Utility of marker system nowski et al., 2008). However, in our analyses using exons, tree resolution and nodal support were higher for some clades within Our analyses based on 3219 bp of three nuclear exons with the and among the different old world parrot taxa compared to a re broadest taxon sampling of old world parrots analyzed so far cent phylogenetic study of parrots based on nuclear introns and recovered with high support the phylogenetic relationship of some mitochondrial genes (Wright et al., 2008). This difference could controversial taxa. We revealed exactly the same topology that be due to our broader taxon sampling (e.g. Poe and Swofford, Hackett et al. (2008) recovered with about 10 times more bp when 1999; Tavares et al., 2006; Zwickl and Hillis, 2002). We sampled comparing the taxa sampled in both studies (Hackett et al. (2008) more old world taxa than Wright et al. (2008) and included two included only 7 parrot species in their study). Hence, we can have to more representatives of controversial taxa. Our results support some conﬁdence that the relationships among all other taxa sam the utility of nuclear genes for the reconstruction of phylogenetic pled are reasonably reliably resolved. A recent simulation study relationships within avian orders consistent with recent data for provided evidence that introns have a greater potential than exons the Passeriformes (Treplin et al., 2008). M. Schweizer et al. / Molecular Phylogenetics and Evolution 54 (2010) 984–994 991

In the MP analyses, Zenk showed the lowest informativeness sister group of the Psittaculini group A. The Australasian Platycer per site and the lowest amount of resolved nodes in the MP strict cini group A were revealed to be monophyletic in contrast to consensus tree compared to Rag 1 and c mos. This is in contrast to Wright et al. (2008), where the position of Neopsephotus and Neop the study of Treplin et al. (2008) where the informativeness per hema was not consistently resolved. The latter were found to be site was similar between Zenk and Rag 1 and the number of re the sister group to the remaining Platycercini group A. The mono solved nodes in the strict consensus tree of the MP analyses was typic genus Lathamus was for the first time analyzed with molecu higher for Zenk than for Rag 1. However, we sequenced only the lar data here. It has been variously treated as a member of the second exon of Zenk, whereas Treplin et al. (2008) included 5020 Loriinae, but our data confirm its placement within the Platycercini bp of the 30 UTR of Zenk additional to 1149 bp of the second exon. (cf. Collar, 1998). The second exon of Zenk together with the 30 UTR has the greatest utility in resolving splits among deep lineages within major avian groups which occurred roughly between 60 and 10 MYA (Chubb, 4.3. Multiple events of long distance trans oceanic dispersal 2004b), but was also able to resolve even older splits within the Passeriformes based on a broader taxon sampling (Treplin et al., Our data suggest Australasia is where a common ancestor of all 2008). In our data set, the second exon of Zenk failed to resolve extant parrots lived. This was also suggested recently by Wright deep splits in the Psittaciformes similar to c mos, while Rag 1 per et al. (2008). To explain the current distribution patterns of the dif formed best throughout. Since at least the deep splits within the ferent groups based on our robustly supported phylogenetic Psittaciformes were thought of as having similar ages as within hypothesis, several long distance dispersal and colonization events the Passeriformes (Barker et al., 2004; Wright et al., 2008) a similar have to be invoked regardless if modern parrots initially split be tree resolution of Zenk compared to Rag 1 would have been ex fore or after the K Pg boundary, and especially the African conti pected for these two groups. The 30 UTR of Zenk could therefore nent must have been colonized independently more than once be more useful to resolve such deep splits than its second exon. (Fig. 3). However, alternatively, highly variable substitution rates among The origin of the Psittacini of Africa (Poicephalus, Psittacus, avian lineages may account for these differences (Hackett et al., excluding Coracopsis) was dated around the K Pg boundary by 2008). Wright et al. (2008) when they used a Gondwana calibration point. African was already separated from the southern continents in the 4.2. Phylogenetic relationships within the parrots late Cretaceous no matter which of several different models for the break up of Gondwana is assumed (Upchurch, 2008). The origin of We succeeded to resolve the phylogenetic relationships of sev the Psittacini can therefore not be explained by vicariant evolution eral controversial taxa with high nodal support and confirmed re as a consequence of the separation of Africa from Gondwana. We cent results of molecular phylogenies which were in conflict with found the Psittacini to be the sister group of the Neotropical Arini, traditional views of the evolutionary history of the parrots (Collar, but the ancestral distribution area of their common ancestor could 1998). Previous molecular studies could not resolve the position of not be unambiguously resolved. We hypothesize that a common the African taxa Poicephalus and Psittacus, however, we identified ancestor of the Arini and the Psittacini became separated from the Neotropical Arini as their sister group. Psittacus and Poicephalus the Australasian lineages on Antarctica through vicariant evolution have been grouped together with Coracopsis from Madagascar and coinciding with the beginning of seafloor spreading between Aus adjacent islands as Psittacini in part because their relationships to tralia and Antarctica in the late Cretaceous and their final separa other groups were not clear (Collar, 1998). However, de Kloet and tion about 40 MYA (Li and Powell, 2001). At the beginning of the de Kloet (2005) hypothesized a close relationship of Coracopsis Paleogene, Antarctica was ice free, warmer and wetter than today with Psittrichas fulgidus from New Guinea. Our study recovered this and separated into West and East Antarctica by a seaway (Lawver clade but without robust support, which leaves the affinities of and Gahagan, 2003). We hypothesize that the common ancestor of these taxa still not unambiguously resolved. the Psittacini and that of the Arini diverged between East and West In congruence with Wright et al. (2008), we found a clade of Antarctica and that the major lineages of the Arini identified by African, Australasian and Indo Malayan taxa containing the Platy Tavares et al. (2006) evolved subsequently on West Antarctica, cercini group A and the Loricoloriinae. Melopsittacus was tradition the southern cone of South America and the rest of this continent. ally included within the Platycercini (Collar, 1998), however, we When climate change began during the Eocene with a trend to found it to belong to a clade consisting of the Loriinae and the wards cooler conditions, continental ice sheets expanded rapidly Cyclopsittacini. This is congruent with other molecular studies on Antarctica in the earliest Oligocene (Zachos et al., 2001). Parrots (de Kloet and de Kloet, 2005; Wright et al., 2008). The sister group probably dispersed out of Antarctica as was proposed for the Arini of this clade could not be unambiguously resolved by Wright et al. by Tavares et al. (2006). The ancestors of the Psittacini may have (2008). We revealed it to be a clade consisting of Agapornis from reached Africa by trans oceanic dispersal from East Antarctica via Africa and Madagascar and the Indo Malayan/Australasian Loricu the Kerguelen Plateau (de Kloet and de Kloet, 2005; Fig. 4, see lus. A close relationship between the latter two genera had already below). been suggested earlier (cf. Smith, 1975), and was found recently The genus Agapornis is restricted to Africa and Madagascar, also by Wright et al. (2008). Within the Agapornis species analyzed, whereas its sister genus Loriculus is distributed in the Indo Mala A. canus was the sister taxon to the remaining species. A. canus is yan and Australasian region with eight of 13 species occurring the only species of this genus occurring exclusively on Madagascar, exclusively in Australasia when the Wallace’s line is taken as a while all others are confined to Africa. Within Loriculus, we found boundary between the Australasian and Indo Malayan region (Col that the Australasian L. catamene is the sister species of the Indo lar, 1998). We identified Australasia as the ancestral area for a Malayan species pair L. philippensis and L. galgulus. Agapornis and common ancestor of Agapornis and Loriculus. The ancestor of Aga Loriculus were traditionally treated as Psittaculini, but our findings pornis, hence, must have reached Africa from Australasia. The split corroborate the proposition of Mayr (2008) based on hypotarsal between these genera is clearly younger than the separation of morphology that these two genera are closely related to the Lorii Gondwana, and thus requires another event of trans oceanic dis nae, the Cyclopsittacini and Melopsittacus. Mayr (2008) proposed persal. We hypothesize that the common ancestor of Agapornis dis the name Loricoloriinae for this clade. However, he included Micro persed from Australasia to Madagascar and finally, Agapornis psitta (Micropsittini) in the Loricoloriinae which we found to be the colonized the African continent from Madagascar (Fig. 4). 992 M. Schweizer et al. / Molecular Phylogenetics and Evolution 54 (2010) 984–994

thought (de Queiroz, 2005; Cowie and Holland, 2006; McGlone, 2005; Yoder and Nowak, 2006). Vicariance is usually considered the null hypothesis that is rejected if phylogenetic divergence dates are incongruent with potential vicariance events, leaving dis persal as the most likely alternative (e.g. Pramuk et al., 2008; Trenel et al., 2007). Trans oceanic dispersal has recently been in voked as a plausible mechanism to explain distribution patterns of several bird taxa, e.g. Turdus thrushes (Voelker et al., 2009), cuckoo shrikes and allies (Campephagidae) (Fuchs et al., 2007; Jonsson et al., 2008), Columbiformes (Pereira et al., 2007; Shapiro et al., 2002), vangas, bushshrikes and allies (Fuchs et al., 2006b). Trans oceanic dispersal is thought to have played a role in the diversification of other terrestrial vertebrates too, including cha meleons (Raxworthy et al., 2002), lizards (Vicario et al., 2003), frogs (Heinicke et al., 2007; Vences et al., 2003), Carnivora and le murs (Yoder et al., 2003) as well as monkeys (Schrago and Russo, 2003). Our data suggest that trans oceanic dispersal may have also played a major role in the spread and diversification of parrots. Their ability to disperse long distances over water is shown by their colonization of distant islands in the Pacific (Collar, 1998). We suggest multiple dispersal events between the Afrotropical, Indo Malayan, Neotropical and Australasian regions as well as Ant arctica during the Paleogene, at similar rate and timescale as has Fig. 4. Inferred trans-oceanic dispersal routes of parrots. Emergent continents been recently proposed for the Columbiformes (Pereira et al., above sea level today are shaded in grey and continental shelves are indicated with 2007). Halas et al. (2005) proposed the null hypothesis for histor black lines (redrawn from Li and Powell (2001)), corresponding to the middle ical biogeography may have to be changed from vicariance to taxon Eocene, when the dispersal of Agapornis, the Psittacini and Coracopsis could have occurred. Following Fjeldså and Bowie (2008), the Broken Ridge and the Kerguelen pulses. A taxon pulse model of diversification may indeed best ex archipelago are indicated where volcanic islands could have served as stepping plain the global spread and radiation of parrots because initial stones. The dispersal of Loriculus, Psittacus and Psittinus probably occurred later vicariance events within the center of parrot diversity were fol around 20–25 MYA. lowed by episodes of dispersal to and in situ radiation on all other tropical continents. In fact both Africa and the Indo Malaysia were colonized by multiple lineages at different times, and more often A third case of trans oceanic dispersal to the African region, and than not these lineages subsequently radiated, resulting in a bioge specifically again to Madagascar, is indicated by the relationship ographically reticulated history (cf. Erwin, 1981; Halas et al., 2005). between Psittrichas fulgidus and Coracopsis (see above). The Mala Tavares et al. (2006) proposed an important role for ecological spe gasy avifauna is supposed to be strictly comprised of Cenozoic ciation via niche diversification in combination with the coloniza trans oceanic dispersers and most groups are thought to have orig tion of open dry habitats for a clade of Neotropical parrots. inated from African ancestors (Yoder and Nowak, 2006). However, Contrasting with conclusions of a recent study on the radiation trans oceanic dispersal from Australasia to Madagascar or Mauri of plant lineages in the Southern Hemisphere, which found biome tius was also suggested for Alectroenas pigeons (Shapiro et al., conservatism prevalent (Crisp et al., 2009), relatively frequent 2002), Anas dabbling ducks (Johnson and Sorenson, 1999) and biome shifts through dispersal followed by local adaptive radiation cuckoo shrikes, whose genus Coracina further colonized Africa into new habitats seem responsible for much of the diversification via Madagascar (Fuchs et al., 2007; Jonsson et al., 2008). Around of parrots. The newly colonized, previously parrot free areas may 40 MYA, dispersal from Australasia to Madagascar and Africa could have presented under utilized adaptive zones with several ecolog have been facilitated by volcanic islands in the southern Indian ical opportunities facilitating diversification in comparable ways as Ocean possibly serving as stepping stones when the Broken Ridge when newly emerged or remote and ecologically under utilized is of Western Australia was connected with the Kerguelen archipel lands are colonized (cf. Losos and Ricklefs, 2009). ago (Fjeldså and Bowie, 2008; Fuchs et al., 2006a; Fig. 4). Indo Malayasia was probably colonized by Psittacula, Psittinus Acknowledgments and Loriculus in three independent colonization events when Aus tralasia approached Southeast Asia and reached its present posi We especially thank the Silva Casa Foundation for financial tion around 20 25 MYA (Li and Powell, 2001). Parrots of the support of this project. We are grateful to S. Birks (University of genus Psittacula then further colonized Africa. This fourth case of Washington, Burke Museum), Dr. R. Burkhard, Dr. A. Fer successfully colonizing Africa may have happened as recently as genbauer Kimmel, J. Fjeldså and J. B. Kristensen (Zoological Mu the Plio Pleistocene boundary (Groombridge et al., 2004). Finally, seum, University of Copenhagen), H. Gygax, Dr. P. Sandmeier, T. Prioniturus and Tanygnathus may have dispersed to Indo Malaysia and P. Walser and Dr. G. Weiss for kindly providing us with tissue from Australasia as they include species from both regions. or feather samples. We further thank the following people for valu Wright et al. (2008) considered a Paleogene origin of the Psit able support: B. Blöchlinger, R. Burri, H. Frick, M. Hohn, S. Klopf taciformes less likely than a Cretaceous origin because the former stein, S. Lauper, L. Lepperhof, M. Maan, M. Rieger, T. Roth and C. requires several oversea dispersal events. Importantly, our data Sherry. suggests that trans oceanic dispersal has to be invoked no matter if parrots originated in the Paleogene or in the Cretaceous. Vicari References ance had long been considered more plausible than dispersal to ex plain major distribution patterns, but recent work suggests that Akaike, H., 1974. New look at statistical-model identification. IEEE Transactions on trans oceanic dispersal may be more important than previously Automatic Control AC 19, 716–723. M. Schweizer et al. / Molecular Phylogenetics and Evolution 54 (2010) 984–994 993

Barker, F.K., Cibois, A., Schikler, P., Feinstein, J., Cracraft, J., 2004. Phylogeny and Hillis, D.M., Bull, J.J., 1993. An empirical-test of bootstrapping as a method for diversification of the largest avian radiation. Proceedings of the National assessing confidence in phylogenetic analysis. Systematic Biology 42, 182–192. Academy of Sciences of the United States of America 101, 11040–11045. Huelsenbeck, J.P., Ronquist, F., 2001. MrBayes: Bayesian inference of phylogenetic Brown, J.M., Lemmon, A.R., 2007. The importance of data partitioning and the utility trees. Bioinformatics 17, 754–755. of bayes factors in Bayesian phylogenetics. Systematic Biology 56, 643–655. Hughes, J.M., Baker, A.J., 1999. Phylogenetic relationships of the enigmatic hoatzin Brown, J.W., Payne, R.B., Mindell, D.P., 2007. Nuclear DNA does not reconcile ‘rocks’ (Opisthocomus hoazin) resolved using mitochondrial and nuclear gene and ‘clocks’ in Neoaves: a comment on Ericson et al. Biology Letters 3, 257– sequences. Molecular Biology and Evolution 16, 1300–1307. 259. Johnson, K.P., Sorenson, M.D., 1999. Phylogeny and biogeography of dabbling ducks Brown, J.W., Rest, J.S., Garcia-Moreno, J., Sorenson, M.D., Mindell, D.P., 2008. Strong (genus: Anas): a comparison of molecular and morphological evidence. Auk 116, mitochondrial DNA support for a Cretaceous origin of modern avian lineages. 792–805. BMC Biology, 6. Jonsson, K.A., Irestedt, M., Fuchs, J., Ericson, P.G.P., Christidis, L., Bowie, R.C.K., Chojnowski, J.L., Kimball, R.T., Braun, E.L., 2008. Introns outperform exons in Norman, J.A., Pasquet, E., Fjeldså, J., 2008. Explosive avian radiations and multi- analyses of basal avian phylogeny using clathrin heavy chain genes. Gene 410, directional dispersal across Wallacea: evidence from the Campephagidae and 89–96. other Crown Corvida (Aves). Molecular Phylogenetics and Evolution 47, 221– Chubb, A.L., 2004a. New nuclear evidence for the oldest divergence among 236. neognath birds: the phylogenetic utility of ZENK (i). Molecular Phylogenetics Kass, R.E., Raftery, A.E., 1995. Bayes Factors. Journal of the American Statistical and Evolution 30, 140–151. Association 90, 773–795. Chubb, A.L., 2004b. Nuclear corroboration of DNA–DNA hybridization in deep Lawver, L.A., Gahagan, L.M., 2003. Evolution of Cenozoic seaways in the circum- phylogenies of hummingbirds, swifts, and passerines: the phylogenetic utility Antarctic region. Palaeogeography Palaeoclimatology Palaeoecology 198, 11– of ZENK (ii). Molecular Phylogenetics and Evolution 30, 128–139. 37. Collar, N.J., 1998. Family Psittacidae (Parrots). In: del Hoyo, J., Elliott, A., Sargatal, J. Lewis, P.O., 2001. A likelihood approach to estimating phylogeny from discrete (Eds.), Handbook of the Birds of the World, vol. 4. Lynx Edicions, Barcelona, pp. morphological character data. Systematic Biology 50, 913–925. 280–477. Li, Z.X., Powell, C.M., 2001. An outline of the palaeogeographic evolution of the Cooper, A., Penny, D., 1997. Mass survival of birds across the Cretaceous-tertiary Australasian region since the beginning of the Neoproterozoic. Earth-Science boundary: molecular evidence. Science 275, 1109–1113. Reviews 53, 237–277. Cowie, R.H., Holland, B.S., 2006. Dispersal is fundamental to biogeography and the Long, K.D., Salbaum, J.M., 1998. Evolutionary conservation of the immediate-early evolution of biodiversity on oceanic islands. Journal of Biogeography 33, 193– gene ZENK. Molecular Biology and Evolution 15, 284–292. 198. Losos, J.B., Ricklefs, R.E., 2009. Adaptation and diversification on islands. Nature 457, Cracraft, J., 2001. Avian evolution, Gondwana biogeography and the Cretaceous- 830–836. tertiary mass extinction event. Proceedings of the Royal Society B-Biological Lovette, I.J., Bermingham, E., 2000. c-mos variation in songbirds: molecular Sciences 268, 459–469. evolution, phylogenetic implications, and comparisons with mitochondrial Crisp, M.D., Arroyo, M.T.K., Cook, L.G., Gandolfo, M.A., Jordan, G.J., McGlone, M.S., differentiation. Molecular Biology and Evolution 17, 1569–1577. Weston, P.H., Westoby, M., Wilf, P., Linder, H.P., 2009. Phylogenetic biome Maddison, W.P., Maddison, D.R., 2006. StochChar: a package of Mesquite modules conservatism on a global scale. Nature 458, 754–756. for stochastic models of character evolution. Version 1, 1. de Kloet, R.S., de Kloet, S.R., 2005. The evolution of the spindlin gene in birds: Maddison, W.P., Maddison, D.R., 2008. Mesquite: a modular system for evolutionary sequence analysis of an intron of the spindlin W and Z gene reveals four major analysis: Available at http://mesquiteproject.org. divisions of the Psittaciformes. Molecular Phylogenetics and Evolution 36, 706– Mayr, G., 2008. The phylogenetic affinities of the parrot taxa Agapornis, Loriculus and 721. Melopsittacus (Aves: Psittaciformes): hypotarsal morphology supports the de Queiroz, A., 2005. The resurrection of oceanic dispersal in historical results of molecular analyses. Emu 108, 23–27. biogeography. Trends in Ecology & Evolution 20, 68–73. Mayr, G., 2009. Paleogene Fossil Birds. Springer-Verlag, Berlin Heidelberg. Ericson, P.G.P., Anderson, C.L., Britton, T., Elzanowski, A., Johansson, U.S., Kallersjo, McGlone, M.S., 2005. Goodbye Gondwana. Journal of Biogeography 32, 739–740. M., Ohlson, J.I., Parsons, T.J., Zuccon, D., Mayr, G., 2006. Diversification of Newton, I., 2003. The Speciation and Biogeography of Birds. Academic Press, Neoaves: integration of molecular sequence data and fossils. Biology Letters 2, London. 543. U1. Nylander, J.A.A., 2004. MrModeltest v2. Evolutionary Biology Centre, Uppsala Erwin, T.L., 1981. Taxon pulses, vicariance, and dispersal: an evolutionary synthesis University Uppsala, Sweden. illustrated by carabid beetles. In: Nelson, G., Rosen, D.E. (Eds.), Vicariance Overton, L.C., Rhoads, D.D., 2004. Molecular phylogenetic relationships based on Biogeography: A Critique. Columbia University Press, New York, pp. 159–196. mitochondrial and nuclear gene sequences for the Todies (Todus, Todidae) of the Farris, J.S., Kallersjo, M., Kluge, A.G., Bult, C., 1995. Constructing a significance test Caribbean. Molecular Phylogenetics and Evolution 32, 524–538. for incongruence. Systematic Biology 44, 570–572. Paton, T.A., Baker, A.J., Groth, J.G., Barrowclough, G.F., 2003. RAG-1 sequences Fjeldså, J., Bowie, R.C.K., 2008. New perspectives on the origin and diversification of resolve phylogenetic relationships within Charadriiform birds. Molecular Africa’s forest avifauna. African Journal of Ecology 46, 235–247. Phylogenetics and Evolution 29, 268–278. Fuchs, J., Cruaud, C., Couloux, A., Pasquet, E., 2007. Complex biogeographic history of Pereira, S.L., Johnson, K.P., Clayton, D.H., Baker, A.J., 2007. Mitochondrial and nuclear the cuckoo-shrikes and allies (Passeriformes: Campephagidae) revealed by DNA sequences support a Cretaceous origin of Columbiformes and a dispersal- mitochondrial and nuclear sequence data. Molecular Phylogenetics and driven radiation in the Paleogene. Systematic Biology 56, 656–672. Evolution 44, 138–153. Poe, S., Swofford, D.L., 1999. Taxon sampling revisited. Nature 398, 299–300. Fuchs, J., Fjeldså, J., Bowie, R.C.K., Voelker, G., Pasquet, E., 2006a. The African warbler Pramuk, J.B., Robertson, T., Sites, J.W., Noonan, B.P., 2008. Around the world in 10 genus Hyliota as a lost lineage in the oscine songbird tree: Molecular support for an million years: biogeography of the nearly cosmopolitan true toads (Anura: African origin of the Passerida. Molecular Phylogenetics and Evolution 39, 186–197. Bufonidae). Global Ecology and Biogeography 17, 72–83. Fuchs, J., Fjeldså, J., Pasquet, E., 2006b. An ancient African radiation of corvoid birds Pratt, R.C., Gibb, G.C., Morgan-Richards, M., Phillips, M.J., Hendy, M.D., Penny, D., (Aves: Passeriformes) detected by mitochondrial and nuclear sequence data. 2009. Toward resolving deep Neoaves phylogeny: Data, signal enhancement, Zoologica Scripta 35, 375–385. and priors. Molecular Biology and Evolution 26, 313–326. Griffiths, C.S., Barrowclough, G.F., Groth, J.G., Mertz, L., 2004. Phylogeny of the Raxworthy, C.J., Forstner, M.R.J., Nussbaum, R.A., 2002. Chameleon radiation by Falconidae (Aves): a comparison of the efficacy of morphological, oceanic dispersal. Nature 415, 784–787. mitochondrial, and nuclear data. Molecular Phylogenetics and Evolution 32, Ribas, C.C., Moyle, R.G., Miyaki, C.Y., Cracraft, J., 2007. The assembly of montane 101–109. biotas: linking Andean tectonics and climatic oscillations to independent Groombridge, J.J., Jones, C.G., Nichols, R.A., Carlton, M., Bruford, M.W., 2004. regimes of diversification in Pionus parrots. Proceedings of the Royal Society B- Molecular phylogeny and morphological change in the Psittacula parakeets. Biological Sciences 274, 2399–2408. Molecular Phylogenetics and Evolution 31, 96–108. Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference Groth, J.G., Barrowclough, G.F., 1999. Basal divergences in birds and the under mixed models. Bioinformatics 19, 1572–1574. phylogenetic utility of the nuclear RAG-1 gene. Molecular Phylogenetics and Rowley, I., 1998. Family Cacatuidae (Cockatoos). In: Del Hoyo, J., Elliott, A., Sargatal, Evolution 12, 115–123. J. (Eds.), Handbook of the Birds of the World, vol. 4. Lynx Edicions, Barcelona, pp. Hackett, S.J., Kimball, R.T., Reddy, S., Bowie, R.C.K., Braun, E.L., Braun, M.J., 246–279. Chojnowski, J.L., Cox, W.A., Han, K.L., Harshman, J., Huddleston, C.J., Marks, Schrago, C.G., Russo, C.A.M., 2003. Timing the origin of New World monkeys. B.D., Miglia, K.J., Moore, W.S., Sheldon, F.H., Steadman, D.W., Witt, C.C., Yuri, T., Molecular Biology and Evolution 20, 1620–1625. 2008. A phylogenomic study of birds reveals their evolutionary history. Science Shapiro, B., Sibthorpe, D., Rambaut, A., Austin, J., Wragg, G.M., Bininda-Emonds, 320, 1763–1768. O.R.P., Lee, P.L.M., Cooper, A., 2002. Flight of the dodo. Science 295, 1683. Halas, D., Zamparo, D., Brooks, D.R., 2005. A historical biogeographical protocol for Smith, G.A., 1975. Systematics of Parrots. Ibis 117, 18–68. studying biotic diversification by taxon pulses. Journal of Biogeography 32, Soltis, D.E., 2007. Reconstruction the three of life: Taxonomy and systematics of 249–260. species rich taxa. Evolution 61, 3007–3011. Hall, T.A., 1999. BioEdit, a user-friendly biological sequence alignment editor and Stamatakis, A., 2006. RAxML-VI-HPC: maximum likelihood-based phylogenetic analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41, analyses with thousands of taxa and mixed models. Bioinformatics 22, 2688– 95–98. 2690. Heinicke, M.P., Duellman, W.E., Hedges, S.B., 2007. Major Caribbean and Central Stamatakis, A., Hoover, P., Rougemont, J., 2008. A rapid bootstrap algorithm for the American frog faunas originated by ancient oceanic dispersal. Proceedings of RAxML web servers. Systematic Biology 57, 758–771. the National Academy of Sciences of the United States of America 104, 10092– Swafford, D.L., 2001. PAUP*: Phylogenetic Analysis Using Parsimony (*and other 10097. methods), Version 4.0. Sinauer Associates, Sunderland, MA. 994 M. Schweizer et al. / Molecular Phylogenetics and Evolution 54 (2010) 984–994

Tavares, E.S., Baker, A.J., Pereira, S.L., Miyaki, C.Y., 2006. Phylogenetic relationships Voelker, G., Rohwer, S., Outlaw, D.C., Bowie, R.C.K., 2009. Repeated trans-Atlantic and historical biogeography of Neotropical parrots (Psittaciformes: Psittacidae: dispersal catalysed a global songbird radiation. Global Ecology and Arini) inferred from mitochondrial and nuclear DNA sequences. Systematic Biogeography 18, 41–49. Biology 55, 454–470. Wright, T.F., Schirtzinger, E.E., Matsumoto, T., Eberhard, J.R., Graves, G.R., Sanchez, Tokita, M., Kiyoshi, T., Armstrong, K.N., 2007. Evolution of craniofacial novelty in J.J., Capelli, S., Mueller, H., Scharpegge, J., Chambers, G.K., Fleischer, R.C., 2008. A parrots through developmental modularity and heterochrony. Evolution & multilocus molecular phylogeny of the parrots (Psittaciformes): support for a Development 9, 590–601. Gondwanan origin during the Cretaceous. Molecular Biology and Evolution 25, Trenel, P., Gustafsson, M.H.G., Baker, W.J., Asmussen-Lange, C.B., Dransﬁeld, J., 2141–2156. Borchsenius, F., 2007. Mid-tertiary dispersal, not Gondwanan vicariance Xia, X., Xie, Z., 2001. DAMBE: software package for data analysis in molecular explains distribution patterns in the wax palm subfamily (Ceroxyloideae: biology and evolution. Journal of Heredity 92, 371–373. Arecaceae). Molecular Phylogenetics and Evolution 45, 272–288. Xia, X.H., Xie, Z., Salemi, M., Chen, L., Wang, Y., 2003. An index of substitution Treplin, S., Siegert, R., Bleidorn, C., Thompson, H.S., Fotso, R., Tiedemann, R., 2008. saturation and its application. Molecular Phylogenetics and Evolution 26, 1–7. Molecular phylogeny of songbirds (Aves: Passeriformes) and the relative utility Yoder, A.D., Burns, M.M., Zehr, S., Delefosse, T., Veron, G., Goodman, S.M., Flynn, J.J., of common nuclear marker loci. Cladistics 24, 328–349. 2003. Single origin of Malagasy Carnivora from an African ancestor. Nature 421, Upchurch, P., 2008. Gondwanan break-up: legacies of a lost world? Trends in 734–737. Ecology & Evolution 23, 229–236. Yoder, A.D., Nowak, M.D., 2006. Has vicariance or dispersal been the predominant Vences, M., Vieites, D.R., Glaw, F., Brinkmann, H., Kosuch, J., Veith, M., Meyer, A., biogeographic force in Madagascar? Only time will tell. Annual Review of 2003. Multiple overseas dispersal in amphibians. Proceedings of the Royal Ecology Evolution and Systematics 37, 405–431. Society B-Biological Sciences 270, 2435–2442. Zachos,J.,Pagani,M.,Sloan,L.,Thomas,E.,Billups,K.,2001.Trends,rhythms, Vicario, S., Caccone, A., Gauthier, J., 2003. Xantusiid ‘‘night” lizards: a puzzling and aberrations in global climate 65 Ma to present. Science 292, 686–693. phylogenetic problem revisited using likelihood-based Bayesian methods on Zwickl, D.J., Hillis, D.M., 2002. Increased taxon sampling greatly reduces mtDNA sequences. Molecular Phylogenetics and Evolution 26, 243–261. phylogenetic error. Systematic Biology 51, 588–598.

Chapter 4

MACROEVOLUTIONARY PATTERNS IN THE DIVERSIFICATION OF PARROTS: EFFECTS OF CLIMATE CHANGE, GEOLOGICAL EVENTS AND KEY INNOVATIONS

Manuel Schweizer1*, Ole Seehausen2,3, Stefan T. Hertwig1

1Naturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, CH 3005 Bern, Switzerland

2Aquatic Ecology and Macroevolution, Institute of Ecology & Evolution, University of Bern, Baltzerstrasse 6, CH 3012 Bern, Switzerland

3Fish Ecology and Evolution, EAWAG, Seestrasse 79, CH 6047 Kastanienbaum, Switzerland

*Correspondence: Manuel Schweizer, Naturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, CH 3005 Bern, Switzerland. E-mail [email protected].

Published as: Schweizer, M., O. Seehausen, and S. T. Hertwig. in press. Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations. Journal of Biogeography. (http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2699.2011.02555.x/abstract)

69 Journal of Biogeography (1. Biogeogr.) (2011)

Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations

1 2 1 Manuel Schweizer *, Ole Seehausen ,3 and Stefan T. Hertwig

1Naturhistorisches Museum der ABSTRACT Burgergemeinde Bern, Bernastrasse 15, CH 2 Aim Parrots are thought to have originated on Gondwana during the Cretaceous. 3005 Bern, Switzerland, Aqua tic Ecology and Macroevolution, Institute of Ecology and The initial split within crown group parrots separated the New Zealand taxa from Evolution, University of Bern, Baltzerstrasse 6, the remaining extant species and was considered to coincide with the separation CH 3012 Bern, Switzerland, 3Fish Ecology and of New Zealand from Gondwana 82 85 Ma, assuming that the diversification of Evolution, EA WAG, Seestrasse 79, CH 6047 parrots was mainly shaped by vicariance. However, the distribution patterns of Kastanienbaum, Switzerland several extant parrot groups cannot be explained without invoking transoceanic dispersal, challenging this assumption. Here, we present a temporal and spatial framework for the diversification of parrots using external avian fossils as calibration points in order to evaluate the relative importance of the influences of past climate change, plate tectonics and ecological opportunity.

Location Australasian, African, Indo Malayan and Neotropical regions.

Methods Phylogenetic relationships were investigated using partial sequences of the nuclear genes c mos, RAG 1 and Zenk of 75 parrot and 21 other avian taxa. Divergence dates and confidence intervals were estimated using a Bayesian relaxed molecular clock approach. Biogeographic patterns were evaluated taking temporal connectivity between areas into account. We tested whether diversification remained constant over time and if some parrot groups were more species rich than expected given their age.

Results Crown group diversification of parrots started only about 58 Ma, in the >a Palaeogene, significantly later than previously thought. The Australasian lories .c and possibly also the Neotropical Arini were found to be unexpectedly species a. rich. Diversification rates probably increased around the Eocene/Oligocene m boundary and in the middle Miocene, during two periods of major global .. climatic aberrations characterized by global cooling. c:n M ain conclusions The diversification of parrots was shaped by climatic and 0 geological events as well as by key innovations. Initial vicariance events caused by GI continental break up were followed by transoceanic dispersal and local radiations. c:n Habitat shifts caused by climate change and mountain orogenesis may have acted 0 as a catalyst to the diversification by providing new ecological opportunities and challenges as well as by causing isolation as a result of habitat fragmentation. The ·-m lories constitute the only highly nectarivorous parrot clade, and their diet shift, 'l- associated with morphological innovation, may have acted as an evolutionary key innovation, allowing them to explore underutilized niches and promoting their o diversification. m *Correspondence: Manuel Schweizer, Keywords - Naturhistorisches Museum der Burgergerneinde Oimate change, dispersal, diversification, Gondwana, historical biogeography, c Bern, Bemastrasse 15, CH 3005 Bern, Switzerland. key innovation, molecular clock, molecular phylogeny, Psittaciformes, vicari ..:s E mail: [email protected] ance. ...0 © 2011 Blackwell Publishing Ltd http://wileyonlinelibrary.com/journaVjbi doi:l0.1111/j.1365 2699.2011 .02555.x M. Schweizer et al.

et al., 2010), the temporal patterns of their diversification INTRODUCTION remain controversial. The finding that the New Zealand taxa A robust temporal and spatial framework for the speciation Nestor and Strigops formed the monophyletic sister group of events in a group of organisms is a prerequisite for understand the remaining taxa led to the assumption that the separation of ing the evolutionary dynamics responsible for its current New Zealand from Gondwana 82 85 Ma coincided with this diversity. In this context, an assessment of the relative influences early split within modern parrots (de Kloet & de Kloet, 2005). of plate tectonics, past climate change and ecological opportu This bio and palaeogeographic evidence was used to calibrate nity on the diversification process is especially important. the diversification of several groups of Neotropical parrots Despite the eminent efforts that have been made to reconstruct (Ribas et al., 2005, 2009; Tavares et al., 2006). However, such the phylogenies of major vertebrate groups such as birds and calibrations based on New Zealand biogeography have been mammals, the time scale of their radiations is still a matter of criticized as being a case in which geological and biological controversy. In the past, the strict interpretation of the fossil evidence lacked independence and always rely on implicit record led to the hypothesis that modern birds evolved in an assumptions about vicariance and dispersal (Waters & Craw, explosive radiation paralleling that of mammals after the global 2006; Ho & Phillips, 2009; Trewick & Gibb, 2010). It was perturbations that caused mass extinctions at the Cretaceous argued in the case of parrots, however, that the diversification Palaeogene (K Pg) boundary 65 Ma. In this scenario, birds and of these today mostly non migratory birds was shaped mammals inherited practically the entirety of the terrestrial primarily by vicariance and in fact not much influenced by vertebrate adaptive landscape from the other dinosaur groups dispersal. Following the same reasoning, Wright et al. (2008) and the pterosaurs, and rapidly filled the many recently vacant considered a Palaeogene origin of modern parrots to be less ecological niches (Feduccia, 1995, 2003). However, several likely than a Cretaceous origin, because a Palaeogene scenario recent molecular phylogenetic studies have dated the origin of would require several transoceanic dispersal events to explain modern birds before the K Pg boundary (Hedges et al., 1996; current distribution patterns. In contrast, Schweizer et al. Cooper & Penny, 1997; Pereira & Baker, 2006; Slack et al., 2006; (2010) demonstrated that transoceanic dispersal between the Brown et al., 2007, 2008; Pratt et al., 2009). Furthermore, the Afrotropical, Indo Malayan, Neotropical and Australasian description of a well preserved fossil anseriform (Vegavis) from regions as well as Antarctica has to be invoked to explain the the very late Cretaceous pushed at least five basal avian splits back distribution patterns of parrots, no matter if they originated in into the Cretaceous (Clarke et al., 2005; Brown et al., 2008). the Palaeogene or in the Cretaceous. This in turn challenged The integration of molecular phylogenetic data with the the value of taking the separation of New Zealand from geological context resulted in the conclusion that the conti Gondwana as a calibration point for the initial split within nental break up of Gondwana during the Cretaceous shaped parrots. Indeed, molecular dating based on complete mito the diversification not only of the deep lineages of birds, but chondrial genomes, involving fossil calibrations outside the also those of mammals (Hedges et al., 1996; Cracraft, 2001; parrots, dated the split of Strigops from two (Agapornis, Nishihara et al., 2009). Within birds, the ratites (Palaeognat Melopsittacus) (Pratt et al., 2009) to six (Agapornis, Aratinga, hae) have played a crucial role in the arguments surrounding Brotogeris, Forpus, Melopsittacus, Nymphicus) (Pacheco et al., the biogeography of Gondwana. They were mainly thought to 2011) other genera of parrots after the K Pg boundary. have diverged as a result of vicariance in the late Cretaceous In the present work, we generated a comprehensive tempo caused by continental drift with the exception of the kiwi ral framework for the diversification of parrots based on a (Apteryx) and the ostrich (Struthio), which dispersed later robust phylogenetic hypothesis, independent calibration points (Cooper et al., 2001; Cracraft, 2001; Haddrath & Baker, 2001). and a relaxed molecular clock approach (Drummond et al., However, the causal relationship between these geological and 2006) to test the hypothesis that the initial split within crown biological events has been called into question because group parrots coincided with the separation of New Zealand assessment of evidence in favour of the temporal congruence from Australia. In addition, we aimed at establishing a detailed of the two phenomena has often suffered from non indepen hypothesis of biogeographic and dispersal patterns and tested dence, and dispersal has been neglected as a potential whether the rates of diversification remained constant over mechanism to explain current distribution patterns (Waters time or if there was indeed the hypothesized early burst after & Craw, 2006; Upchurch, 2008). Indeed, discordance between the K Pg boundary in response to the extinction of other taxa. molecular phylogenies, in combination with divergence time Finally, we asked if some groups within the parrots were more estimates and patterns of continental break up, has recently species rich than expected given their age. been shown for the palaeognath birds, and vicariance alone is no longer considered the best explanation for ratite distribu MATERIALS AND METHODS tion patterns (Harshman et al., 2008; Phillips et al., 2010). Another group of birds that is thought to have originated in Sampling Gondwana is the parrots (Psittaciformes). While several recently published molecular studies shed light on the Our sample comprised a total of 75 out of 353 extant parrot phylogenetic relationships within parrots (de Kloet & de species and included representatives of all the major groups of Kloet, 2005; Tavares et al., 2006; Wright et al., 2008; Schweizer this taxon (Table 1, Appendix S1 in Supporting Information)

2 Journal of Biogeography ª 2011 Blackwell Publishing Ltd Macroevolutionary patterns in the diversiﬁcation of parrots

Table 1 Tribal membership and distribution of the parrot taxa Table 1 Continued sampled for this study following Collar (1998) and Rowley (1998). Note that recent phylogenetic studies have revealed Melopsittacus Tribe Species included Range to cluster together with the Loriini and the Cyclopsittacini, away Eunymphicus cornutus cornutus Australasian from the remaining Platycercini. Agapornis and Loriculus form the Eunymphicus cornutus uvaeensis Australasian sister group to those taxa, away from the remaining Psittaculini Lathamus discolor Australasian (Schweizer et al., 2010; Wright et al., 2008). The name Loricol Melopsittacus undulatus Australasian oriinae has been proposed for this clade (Mayr, 2008). Compare Neophema chrysogaster Australasian this traditional taxonomic treatment with the relationship of Neophema chrysostoma Australasian Bolbopsittacus, Coracopsis, Psittacella and Psittrichas as revealed in Neophema pulchella Australasian this study. Neophema splendida Australasian Tribe Species included Range Neopsephotos bourkii Australasian Northiella haematogaster Australasian Nestorini Nestor notabilis Australasian Platycercus caledonicus Australasian Cacatuini Cacatua galerita fitzroyi Australasian Platycercus eximius Australasian Cacatua moluccensis Australasian Platycercus flaveolus Australasian Calyptorhynchus funereus Australasian Platycercus venustus Australasian Calyptorhynchus latirostris Australasian Prosopeia tabuensis Australasian Psittrichadini Psittrichas fulgidus Australasian Psephotus chrysopterygius Australasian ‘Psittacini’ Coracopsis nigra Malagasy Psephotus dissimilis Australasian Coracopsis vasa Malagasy Psephotus varius Australasian Poicephalus gulielmi Afrotropical Purpureicephalus spurius Australasian Poicephalus meyeri Afrotropical ‘Cyclopsittini’ Bolbopsittacus lunulatus Indo Malayan Poicephalus rufiventris Afrotropical Cyclopsitta diophthalma Australasian Poicephalus senegalus Afrotropical Psittaculirostris desmarestii Australasian Psittacus erithacus erithacus Afrotropical Psittaculirostris edwardsii Australasian Arini Ara macao Neotropical Lorini Trichoglossus johnstoniae Indo Malayan Amazona aestiva Neotropical Lorius garrulus Australasian Amzona dufresniana Neotropical Psitteuteles goldiei Australasian Amazona pretrei Neotropical Eos cyanogenia Australasian Deroptyus accipitrinus Neotropical Charmosyna pulchella Australasian Guarouba guarouba Neotropical Pionus menstruus Neotropical Triclaria malachitacea Neotropical [nomenclature follows Collar (1998) and Rowley (1998)]. ‘Psittaculini’ Agapornis canus Malagasy Compared to a previous paper (Schweizer et al., 2010), we Agapornis fischeri Afrotropical added species of the genera Poicephalus, Prioniturus and Agapornis lilianae Afrotropical Psephotus, the Neotropical taxon Arini, the Philippine endemic Agapornis nigrigenis Afrotropical Agapornis roseicollis Afrotropical Bolbopsittacus lunulatus, the Australasian Northiella haematog Alisterus chloropterus Australasian aster and the genus Psittacella, which is analysed here for the Alisterus scapularis Australasian first time in a molecular genetic study. To be able to date the Aprosmictus jonquillaceus Australasian phylogeny of parrots with external fossils as calibration points, Eclectus roratus Australasian we added sequences taken from GenBank of 20 avian species Loriculus catamene Australasian belonging to the Neognathae (Appendix S1). We further used Loriculus galgulus Indo Malayan Struthio as the outgroup for all analyses to account for the Loriculus philippensis Indo Malayan well accepted split between Palaeognathae and Neognathae Polytelis alexandrae Australasian (e.g. Livezey & Zusi, 2007; Hackett et al., 2008). Partial Polytelis anthopeplus Australasian sequences of the three nuclear genes c mos, RAG 1 and Zenk Prioniturus discurus Indo Malayan (second exon) were generated following the laboratory proto Prioniturus luconensis Indo Malayan Prioniturus montanus Indo Malayan col described in Schweizer et al. (2010) (Appendix S2). The Psittacella brehmii Australasian alignment of the sequences was done manually after translation Psittacula eupatria Indo Malayan into amino acids with BioEdit 7.0.5.2 (Hall, 1999). Psittinus cyanurus Indo Malayan Tanygnathus megalorhynchus Australasian Phylogenetic analyses Micropsittini Micropsitta finschii tristrami Australasian Micropsitta pusio Australasian Phylogenetic hypotheses were established using Bayesian infer ‘Platycercini’ Barnardius zonarius Australasian ence (BI), maximum likelihood (ML) and maximum parsimony Cyanoramphus auriceps Australasian (MP). BI was conducted with MrBayes 3.1 (Huelsenbeck & Cyanoramphus novaezelandiae Australasian Ronquist, 2001; Ronquist & Huelsenbeck, 2003) using a mixed

Journal of Biogeography 3 ª 2011 Blackwell Publishing Ltd M. Schweizer et al. model approach. We evaluated alternative biologically relevant the best parameter settings as found by the MrBayes analyses parameter settings for our concatenated data corresponding to (different substitution models for the genes and their codon separate models with varying base frequencies, rate matrix, positions, see below). As no representative of crown group shape parameters and proportion of invariable sites for the Psittaciformes is known from Palaeogene fossil deposits (Mayr, various genes and/or their codon positions. Models were 2009), we used well accepted fossils outside the parrots as selected based on Akaike information criterion (AIC) values calibration points. We incorporated the earliest known pen using MrModeltest 2.3 (Nylander, 2004). We performed two guin fossil, Waimanu, into our dating analyses and used a independent runs of Metropolis coupled Markov chain Monte mean estimate of 66 Ma with a normal distribution Carlo (MCMC) analyses, each consisting of one cold chain and (95% ± 6 Ma or SD = 3.06) for the split between Sphenisc three heated chains with a default temperature of 0.2. The chains iformes (penguins) and other seabird lineages (Slack et al., were run for 10 million generations with sampling every 100 2006; Pratt et al., 2009). Waimanu has been considered as generations. We checked that the average standard deviation of particularly useful for providing prior information for the split frequencies converged towards zero, and the length of the calibration of molecular phylogenies (Pratt et al., 2009). First, ‘burn in’ period was calculated by visually inspecting trace files aquatic birds such as penguins can be expected to have a better with Tracer 1.4.1 (Rambaut & Drummond, 2007) and by fossil record than land animals. Furthermore, penguins are monitoring the change in cumulative split frequencies using rather large compared with other birds and have more solid awty (Wilgenbusch et al., 2004; Nylander et al., 2008). The first and not hollow bones. Moreover, penguins are distinct and 25% of samples were then discarded as burn in (25,000 trees) thus easier to identify than other bird groups. All these points well after the chains reached stationarity. We further compared minimize the potential problem of the oldest fossil potentially likelihoods and posterior probabilities of all splits to assess underestimating the correct age of a group. The lower bound convergence among the two independent runs using Tracer of the prior chosen accounts for potential dating errors of the and awty. When the chains did not mix appropriately, the fossil, and the upper bound takes into account that two temperature was set to 0.1. The relevance of the different putative members of the Gaviiformes have been described parameter settings was evaluated using the Bayes factor (BF) from the late Cretaceous (cf. Mayr, 2009). Nevertheless, we (Kass & Raftery, 1995; Brown & Lemmon, 2007). The harmonic tested the influence of using more conservative priors for this mean calculated by MrBayes was used as an estimation of the split between penguins and other seabird lineages. We marginal likelihood of the data (Kass & Raftery, 1995). A more additionally used a uniform prior distribution with a lower complicated model was favoured over a simpler model if 2lnBF bound of 60 Ma and an upper bound of 124 Ma (see below for was greater than 10 (Brown & Lemmon, 2007). Indels were justification of this upper bound). We also tested a lognormal treated as missing data; however, based on the best fitting model prior distribution for the same split with a zero offset at 60 Ma (see Results, Phylogeny) MrBayes was rerun with alignment (mean = 1, SD = 1.5). The sister group of Sphenisciformes is gaps coded as binary characters and appended to the matrix not yet unambiguously resolved, so two alternative hypotheses using simple gap coding (Simmons & Ochoterena, 2000) as on their phylogenetic relationships were analysed separately implemented in SeqState 1.4.1 (Muller, 2005, 2006). The ML with beast. First, we treated the penguins as a monophyletic search was performed using RAxML 7.0.4 (Stamatakis, 2006) on clade with Procellariiformes (tubenoses), as suggested by the web server with 100 rapid bootstrap inferences, with all free Hackett et al. (2008) and Pratt et al. (2009). Second, we model parameters estimated by the software (substitution rates, defined the penguins as the sister group of a clade consisting of gamma shape parameter, base frequencies) based on the best Gaviiformes (loons), Procellariiformes, Pelecaniformes and parameter settings found by MrBayes (Stamatakis et al., 2008). Suliformes, based on our results (see below). As a second fossil The MP analyses for the concatenated data were conducted using calibration point we used a minimum age of 30 Ma for the paup* (Swafford, 2001) (heuristic search, 1000 random taxon stem Phoenicopteriformes (flamingos) (Ericson et al., 2006; addition replicates, tree bisection reconnection (TBR) branch Brown et al., 2008). Podicipediformes (grebes) are well swapping, gaps as fifth character state or missing data). Nodal supported as the sister group of Phoenicopteriformes, both support was estimated with a MP bootstrap analysis (1000 by molecular and morphological data (Mayr, 2005; Brown pseudo replicates, heuristic search, 10 random taxon addition et al., 2008; Hackett et al., 2008; Pratt et al., 2009) and by our replicates, TBR branch swapping, number of max trees limited to data (see below). In our beast analyses we thus defined these 100). Clades were considered as supported when clade credibility two groups as monophyletic. We used a uniform prior for values of the BI were ‡ 0.95 (Huelsenbeck & Ronquist, 2001) and their split, with a lower bound at 30 Ma accounting for stem when bootstrap values were ‡ 70 (Hillis & Bull, 1993). group Phoenicopteriformes from the late Eocene/early Oligo cene (Ericson et al., 2006; Brown et al., 2008; Mayr, 2009). Putative stem group Phoenicopteriformes have also been Molecular dating described from the middle Eocene (cf. Mayr, 2009) and even We used beast 1.4.8 (Drummond & Rambaut, 2007) to some, albeit doubtful, remains from the late Cretaceous (Olson estimate divergence times, applying a relaxed molecular clock & Feduccia, 1980). Given this uncertainty, we used a conser with an uncorrelated lognormal distribution of branch lengths vative upper bound of 124 Ma, which considers that well and a Yule tree prior with linked trees and clock models using sampled fossil sites stemming from up to 124 Ma are not

4 Journal of Biogeography ª 2011 Blackwell Publishing Ltd Macroevolutionary patterns in the diversification of parrots described to contain any fossils of modern birds (cf. Pratt 40 Ma, and that between Australasia and the Indo Malayan et al., 2009). In addition, we used a uniform prior distribution, region was set to 1 after 20 Ma. All other dispersal rates were again with an upper bound of 124 Ma and a lower bound of set to 0.1. 66 Ma for the split between Galloanserae and Neoaves. The lower bound was chosen considering that the fossil Vegavis Rates of diversification belonged to Galloanserae at 66 Ma (Clarke et al., 2005; Pratt et al., 2009). Default prior distributions were chosen for all Rates of diversification were analysed using the R packages other parameters, and the MCMC was run for 25 million ‘Laser’ 2.3 (Rabosky, 2006a) and ‘Geiger’ 1.3.1 (Harmon et al., generations with sampling every 1000 generations. Tracer was 2008). Temporal variation in diversification rates was visual used to confirm appropriate burn in and the adequate effective ized using semi logarithmic lineage through time (LTT) plots. sample sizes of the posterior distribution. Three independent The 1000 last trees from the posterior of the best fitting beast chains were run for each of the topological constraints and model were used with the root node set to 58.587 Myr (mean each prior setting, and we compared likelihoods and posterior value recovered with the best fitting beast model), and all probabilities of all splits to assess convergence among the runs non parrot taxa and one subspecies of Eunymphicus cornutus using Tracer and awty. The resulting maximum clade were pruned. We plotted the mean LTT from these 1000 trees credibility tree and the 95% highest posterior density (HPD) along with the 95% confidence intervals. To compute these, we distributions of each estimated node were analysed with used the intervals of node ages over the 1000 trees at every FigTree 1.2.1 (Rambaut, 2008). The relevance of the two lineage added to the tree, starting from the root node. We topological constraints was evaluated with the Bayes factor compared the results with two null models of constant rate (BF) in Tracer, with the marginal likelihood of the data diversification under two extreme relative extinction rates, estimated using the approach proposed by Suchard et al. with speciation (k) set to 0.2 and extinction (l) set either to 0 (2001) (smoothed estimate method, 1000 bootstrap replicates). (relative extinction rate a = l/k = 0, pure birth model) or to 0.18 (a = l/k = 0.9) (Couvreur et al., 2010). One thousand phylogenetic trees with each diversification rate were simulated Biogeographic reconstruction in Mesquite 2.72 (Maddison & Maddison, 2009) to generate We used the software Lagrange to reconstruct the biogeo the null distributions. To account for incomplete taxon graphic history of the parrots (Snapshot version for web sampling, the simulated phylogenies contained 353 tips configuration tool at http://www reelab.net/lagrange) (Ree representing the current species diversity of parrots (Collar, et al., 2005; Ree & Smith, 2008) based on the maximum clade 1998; Rowley, 1998) and were then pruned to 74 taxa in credibility tree from the best fitting model in beast.Lag reflection of our taxon sampling. As our taxon sampling is not range treats dispersal and local extinctions as stochastic random, but biased towards the inclusion of more deeply processes, incorporating a continuous time model for geo diverging lineages (phylogenetically over dispersed sampling), graphic range evolution through dispersal, extinction and the pruning of missing taxa was done non randomly using the cladogenesis (the DEC model), and can take connectivity method of Brock et al. (in press). This method incorporates a between areas into account (Ree et al., 2005; Ree & Smith, scaling parameter a to control the degree to which the 2008; Ree & Sanmartin, 2009). According to their current sampling is phylogenetically over dispersed. When a =0, distribution (Collar, 1998; Rowley, 1998), the terminal taxa pruning of taxa from a phylogenetic tree is completely were assigned to the Afrotropical, Australasian, Indo Malayan, random. When a = 1, a higher proportion of taxa with shorter Neotropical or Malagasy region (Table 1; Newton, 2003; tip branches are pruned, resulting in a higher number of tip Schweizer et al., 2010). As the current distributions of parrot branches that attach to the tree at deeper nodes. Higher values genera only exceptionally span over more than a single of a lead to an increase of the sampling bias towards the root, biogeographic area, the maximum range size was restricted and only the oldest nodes are retained. We used various values to two areas, and all combinations of areas were allowed in the between 0.1 and 10 for a to non randomly prune taxa from adjacency matrix. Baseline rates of dispersal and local extinc our simulated trees. The root node of the resulting trees was set tion were estimated by the software. We considered two to 58.587 Ma, and mean LTT curves were computed. The models of dispersal opportunities in our analyses. The first did mean LTT curve from the posterior distribution of the beast not constrain dispersal between areas over time (dispersal rate analyses was compared with the mean curves of the two null between all areas set to 1). The second incorporated the models for various values of a by carrying out a Kolmogorov following geographical information that may have facilitated Smirnov goodness of fit test. dispersal among areas: the connection between Australasia and To test for temporal variation in diversification rates, the the Neotropical region via Antarctica until about 40 Ma, and birth death likelihood (BDL) approach of Rabosky (2006b) the connection between Australasia and the Indo Malayan and the (c) statistic of Pybus & Harvey (2000) are often region from about 20 Ma (Li & Powell, 2001; Hall, 2002). All applied. However, it has recently been shown that rate other break up events of Gondwana took place before the downturns should not be inferred with these methods unless initial split within the parrots. The dispersal rate between > 80% of species in a particular clade have been sampled Australasia and the Neotropical region was set to 1 before (Cusimano & Renner, 2010). We therefore applied the recently

Journal of Biogeography 5 ª 2011 Blackwell Publishing Ltd M. Schweizer et al. developed comparative method MEDUSA (Alfaro et al., 2009). strict consensus trees when gaps were treated either as a fifth This uses a diversity tree as its basis, which corresponds to a character state or as missing data. time calibrated phylogeny with a species richness value Galliformes were found to be the sister group to Neoaves. assigned at each tip of the tree. MEDUSA first fits a single However, the phylogenetic relationships within Neoaves could birth death model to the entire diversity tree based on the not be robustly resolved. Phoenicopteriformes and Podiciped likelihood function of Rabosky et al. (2007). Then a model iformes always clustered together. Sphenisciformes were found with two birth rates and two death rates, including a shift to be the sister taxon of the seabird lineages (Gaviiformes, location parameter, is fitted to the diversity tree and its AIC Procellariiformes, Pelecaniformes, Suliformes) with no robust score is compared with that of the first model. This process is support. The monophyly of Passeriformes was well supported, continued until the AIC score of a more complex model with as was the sister group position of Acanthisitti (New Zealand additional rate shifts and rate parameters is not less than a wrens) to a clade containing the suboscines and the oscines. defined threshold number. The maximum clade credibility tree Psittaciformes formed an unresolved, not robustly supported from the best fitting model in beast was used for the diversity clade with the Coraciiformes + Falco in the Bayesian and ML tree and pruned down to 31 tips to account for missing species. analyses. The beast and MP analyses revealed a clade These tips represented clades that were well supported in our consisting of the parrots and Coraciiformes, while Falco was phylogenetic inference and to which missing species could be the sister group to all the remaining Neoaves. assigned based on the results of other molecular phylogenetic We confirmed the sister group relationship between the studies (Tavares et al., 2006; Wright et al., 2008). The genera African Psittacini and the Neotropical Arini as proposed by Callocephalon, Ognorhynchus, Oreopsittacus and Psilopsiagon, Schweizer et al. (2010). The division of the Arini into two well which were not included in these earlier studies, were assigned supported clades is in congruence with other molecular to clades based on taxonomic information (Forshaw, 1973). analyses (Tavares et al., 2006; Wright et al., 2008). Psittacella Pezoporus and Geopsittacus were included, together with clustered together with Platycercini without robust support in Psittacella, as sister taxa of Platycercini (cf. Leeton et al., 1994). the ML and Bayesian analyses, while its position in a clade together with Platycercini and Loricoloriinae was not resolved in the MP analyses. The monophyly of the Platycercini was not RESULTS robustly supported either. Bolbopsittacus was found to be the sister taxon of Agapornis and Loriculus. The sister group Sequence characteristics relationship between Psittrichas from Australasia and Corac The final alignment consisted of 3222 bp (c mos, 603 bp; opsis from the Malagasy region was robustly supported. This RAG 1, 1461 bp; Zenk 1158 bp; Appendix S3). The alignment relationship was first proposed by de Kloet & de Kloet (2005), of RAG 1 contained one indel of three amino acids, two indels but either not confirmed (Wright et al., 2008) or not robustly of one amino acid, and one indel of two amino acids. For supported (Schweizer et al., 2010). The cluster of Coracopsis c mos, the alignment contained one indel of four amino acids, and Psittrichas was found to be the sister group of the Arini while for Zenk there were six indels of one amino acid, one and the Psittacini in the MP analyses. In contrast, MrBayes indel of three amino acids and one indel of two amino acids. and RAxML revealed it as the sister group of a cluster There were no ambiguously aligned amino acids. containing all Old World parrots except the Psittacini, Arini, Cacatuidae and Nestor, but with no robust support. beast found it to be the sister group of the cluster of Psittacella + Phylogeny Platycercini (Fig. 2). No conflict was detected between the topologies of the trees resulting from the different parameter settings with Divergence time estimate MrBayes (Appendix S3). After calculation of the Bayes factors, the parameter setting using different models of The comparison of the three independent runs for all beast sequence evolution for the three genes and three codon analyses revealed high convergence among the various positions with temperature set to 0.1 was chosen as the best parameters, and effective samples sizes were > 293 for all fitting model (Fig. 1). Based on this parameter setting, we parameters. The analyses based on two different topological reran the MrBayes analyses using 20 million generations constraints resulted in very similar divergence time estimates with sampling every 100 generations; however, this had no (Table 2). The mean values indicated that the initial split influence on the results. The coding of gaps as binary within the extant parrots occurred after the K Pg boundary characters had no impact on the tree topology and did not (Fig. 3). The Bayes factors revealed the topological constraint significantly influence node support values. The topologies based on the relationships of Sphenisciformes found with of the best scoring trees obtained with beast,MrBayes and MrBayes as the best fitting model. The mean value for the RAxML were highly congruent with the strict consensus MP separation of Nestor from the remaining parrots was dated at tree (Figs 1 & 2, Appendix S4). The ML and BI trees were 58.6 Ma (95% HPD: 44.9 72), and the Cacatuidae split at even identical for the parrots (Fig. 1, Appendix S3). In the 47.4 Ma (36.3 59.4). The subsequent splitting events between MP analyses, there was no difference in the topology of the the remaining major recent groups of parrots occurred

6 Journal of Biogeography ª 2011 Blackwell Publishing Ltd Macroevolutionary patterns in the diversification of parrots

LORICOLORllNAE

......

PLATYC£ACINI

~mphu#llOWl~~H "'~ "~5~:~-~~~ f Vll"/l'lll)llltll• ()Orlfl(lf\4 Vl'eHM.1$ ~"'~* HOQflhomo .,,rendl'dlt f.;=:=$~( PSITTACINAE .,, flltc/)11WMC/Jr)'1»SICrt"'9 '-----~photos boUlkH

PSITTAC\ILINI

MlCROPSITTN 1/1()() l.h:ro/Ultl.ltp..#(J I r:0-:-,..,:-:8::-1 --{)Z•lf~oo;;:c::;::,,:.s• I P$1TTRICKIWINI ==---- ~$Mfl/i:llA p,,...... ,,.,.,. t.119 Potf»pMrv& tutivetitrit I ~-~ PStTTACINI f"O

t------~ "100 0.!!1fw::=:l1tOUN t/100 c.Jwrom.)'l'\Ghw;~' I '------lmoo «>~=:,.~ C..C.m.moluoc«lsl.c '------c-.,-"'-...--HHtwnotobllnl NESTORINI L_---{·~..~ .. ~· !::::::::::::::::::::::::::::::::::::::~~~-----_:______,...,. r."'=.::----C•~lf~ooC::::::::::::::::::::::..'.'...,_~----•~· 1------lt/t()l) L_------{!m~~~::::::::::::~°"""'::'.':______P'fce1Jt$1tn '------Annfhisit!.a 1------~L~ ~------~

L-----~·~,.,,!1:::::::::::::::~""°'~""""~~~::______"°""""' '------

Figure 1 A 50% majority rule consensus tree of the Bayesian inference of parrots and other avian taxa. Clade credibility values and bootstrap values of the maximum likelihood inference above 50 are indicated at each node. The notation used in the text for the various parrot clades is indicated to the right of the tree. between 32 and 40 Ma (Fig. 2, Table 2). The use of two Reconstruction of historical biogeography different more conservative priors for the split between the penguins and the other seabird lineages (based on the The global ML at the root node for the unconstrained topological constraint found to be the best fitting model) model was -64.87 (dispersal rate = 0.002237, extinction resulted in highly similar time estimates for the different rate = 0.00 1918), while the constrained model had a global splits, and our results can thus be considered as robust ML at the root node of -63.94 (dispersal rate= 0.008502, (Table 2). extinction rate = 0.001623). Of the 149 splits at nodes, 136

I I

26 27

; I I I I I 221 I , I 21 ---- I .____ l I I - ~ I I ' 1•1 - - 7 15 I I I 11 1 -~ I I 181 I - I I - I I 1: ----= 191: .~ I 20 I ,_ I 8 9 ' _r:::-E I ' - 3 I ' I 13 ~ - 12 1 i I ,. I rt;:::: 25 ~

~ 10 I 11 2 '--- ' ' •I I - ' ·-- I ---r- 5 23 I I ~ ' -== 28 ' I ~ I I I -

50 125 100 75 50 0 Palaeooone I Eocene "'-'J5 M~ I CRETACEOUS PALAEOGENE NEOGENE

Figure 2 Maximum clade credibility tree of the dating analysis of parrots and other avian taxa using BEAST based on the best fitting model. Mean node ages and the 95% highest posterior density distributions are shown. Posterior summaries were only calculated for the nodes in the tree that had a posterior probability greater than 0.5. The nodes used for calibration are indicated by circles.

were unambiguously resolved between the two models. With Rates of diversification regard to the 13 remaining nodes, however, the models either yielded different splits between areas (nodes l, 2 and When an improvement of the AIC score of~ 4 was considered 7, Table 3, Fig. 4) or considered the same splits as most as a significant increase in model fit (Burnham & Anderson, likely but revealed more than one split within two log 2002; Alfaro et al., 2009), MEDUSA found one period in which likelihood units of the maximum for the respective node the tempo of diversification changed and led to the excep (Table 3, Fig. 4). At nodes l and 2, the two models differed tionally species rich clade of the lories (Loriinae) (Fig. 5). in that the constrained model included, in addition to When models with an improvement of the AIC score of ~ 2 Australasia, the Neotropical region in the range of a were considered as supported (Burnham & Anderson, 2002), common ancestor of these taxa. At node 7, the uncon we had an indication for a second event of increased strained model favoured a split between the Indo Malayan diversification rate for the clade leading to the Arini and Australasian + Indo Malayan realms, while the con (MIC= 3.4). Both these increased diversification rates were strained model found a split between Australasia and Indo associated with rather high species turnover, with death rates Malaysia as more likely. being 95.4% and 89.4%, respectively, of the birth rate. The

Table 2 Divergence dates of parrots and other avian taxa for various nodes estimated with a Bayesian relaxed clock approach. The mean values and the 95% highest posterior density (HPD) distributions are given for two topological constraints and for different prior distributions for the calibration of the split between penguins and other seabird lineages. The ﬁrst topological constraint deﬁned Sphe nisciformes (penguins) as the sister group of a clade consisting of Gaviiformes (loons), Procellariiformes (tubenoses), Pelecaniformes and Suliformes, based on our results, whereas the second one treated Sphenisciformes as a monophyletic clade with Procellariiformes, as suggested by Hackett et al. (2008) and Pratt et al. (2009). Node numbers refer to those in Fig. 2.

Topological constraints and prior for the calibration of the split between penguins and the other seabird lineages:

Hackett et al. (2008), Pratt This study et al. (2009) This study This study Normal prior Normal prior Uniform prior Lognormal prior

Node number Mean 95% HPD (Ma) Mean 95% HPD (Ma) Mean 95% HPD (Ma) Mean 95% HPD (Ma)

1 58.587 44.869 71.961 61.734 49.414 75.350 58.769 (45.375 72.343) 56.991 (44.104 70.470) 2 47.381 36.264 59.395 49.853 39.283 61.394 47.612 (36.779 58.826) 46.043 (35.349 57.082) 3 40.760 31.763 51.280 42.920 33.954 53.016 40.851 (31.625 49.997) 39.631 (30.608 49.006) 4 35.135 26.004 45.069 36.944 27.537 46.270 35.193 (26.631 44.636) 34.306 (25.863 43.498) 5 25.260 17.256 34.267 26.471 18.235 35.431 25.316 (17.421 34.087) 24.638 (17.266 33.053) 6 12.922 7.028 19.634 13.491 7.082 20.609 13.243 (6.821 20.253) 12.671 (6.924 19.249) 7 38.817 31.763 51.2799 40.851 32.157 50.418 38.903 (30.319 47.923) 37.728 (29.019 46.597) 8 36.137 27.429 45.009 38.036 29.672 47.233 36.210 (27.937 44.594) 35.004 (26.753 43.539) 9 27.609 17.792 37.948 29.342 18.122 40.053 27.945 (17.185 37.986) 26.851 (16.559 36.870) 10 31.130 22.278 40.249 33.023 24.264 41.708 31.253 (23.143 39.007) 30.223 (21.958 38.610) 11 18.465 10.765 26.496 19.574 11.258 28.475 18.665 (11.003 26.657) 18.005 (10.484 25.861) 12 19.850 12.722 27.802 20.888 13.106 28.743 19.817 (13.042 27.340) 19.220 (12.676 26.596) 13 14.016 7.730 20.959 14.605 8.094 21.421 13.889 (7.791 20.368) 13.451 (7.671 20.134) 14 9.802 5.816 14.327 10.331 5.996 14.876 9.856 (5.919 14.405) 9.471 (5.540 13.600) 15 32.650 24.179 41.281 34.176 25.606 42.752 32.600 (24.457 40.860) 31.449 (23.241 39.676) 16 18.589 11.715 26.352 19.587 12.482 27.706 18.662 (11.842 26.103) 18.034 (11.309 25.101) 17 14.165 8.213 20.884 14.899 8.559 21.829 14.168 (8.406 20.709) 13.720 (8.036 19.924) 18 28.526 20.680 36.727 29.928 21.862 38.675 28.513 (20.742 38.883) 27.531 (19.553 35.697) 19 23.672 16.328 31.230 24.701 17.241 32.363 23.614 (16.393 31.059) 22.808 (15.706 30.348) 20 13.832 8.072 20.270 14.352 8.356 20.705 13.776 (7.830 20.271) 13.409 (7.677 20.005) 21 33.193 23.547 42.683 35.008 25.442 44.404 33.335 (24.598 42.941) 32.279 (23.065 41.142) 22 25.242 17.249 33.480 26.751 18.466 35.562 25.470 (17.055 33.768) 24.552 (17.053 32.882) 23 112.902 97.157 124.000 115.412 101.975 123.999 112.749 (96.133 123.999) 110.942 (93.829 124.00) 24 80.650 58.979 101.899 81.717 61.759 100.406 78.833 (56.939 97.689) 78.867 (59.013 98.364) 25 95.638 79.197 111.523 98.647 84.511 112.247 95.753 (79.004 111.979) 93.390 (76.338 110.109) 26 70.362 52.600 88.130 73.081 57.364 90.358 71.165 (53.926 89.449) 69.299 (52.555 87.724) 27 62.406 45.517 79.743 65.177 48.929 81.072 63.357 (45.441 80.813) 61.963 (44.955 78.725) 28 49.148 31.177 66.126 48.901 30.817 65.317 49.250 (31.496 66.187) 47.044 (30.423 63.090) 29 64.527 58.715 70.421 64.083 58.229 69.790 65. 387 (60.000 74.930) 62.302 (60.013 67.589) background diversification rate was characterized instead by a (Kolmogorov Smirnov test, P = 0.0002). However, it was only low turnover, with the death rate only accounting for 1.9% of significantly different from a null model based on a constant the birth rate (Fig. 5). diversification rate and a rather high extinction rate for The LTT plot, too, indicated that the diversification rate of a > 1.08, and the simulated mean LTT curves then showed an the parrots was not constant over time (Fig. 5). The first excess of younger branching events. However, the effects of change occurred in the upper Eocene (around 40 Ma), with an extinction can be difficult to separate from those of increased acceleration in the diversification rate that lasted until the early speciation towards the present, a phenomenon termed the Oligocene (around 30 Ma). A second increase in the diversi ‘pull of the present’ (Nee et al., 1994). Extinction may not have fication rate started in the middle Miocene (after 15 Ma). The had any dominant effect in our case, as the diversification of decrease towards the present is probably influenced by the species rich lories started at the same time as the second incomplete lineage sampling, with some genera and many indicated increase in diversification rate and probably influ species missing (cf. Ricklefs et al., 2007) and is not discussed enced overall net diversification rates. Furthermore, MEDUSA further. The mean LTT curve differed significantly from the identified a rather low turnover for the background diversi expectations under the null model based on a constant fication rate. As it is problematic to infer extinction rates from diversification rate and no extinction for any value of a tested molecular phylogenies in the absence of fossil data (Rabosky,

Journal of Biogeography 9 ª 2011 Blackwell Publishing Ltd M. Schweizer et al.

K-Pg boundary part of the parrot phylogeny could probably be enhanced by I the inclusion of Pezoporus and Geopsittacus in further molec I Topological constrainls ular studies, as it was shown from cytochrome b data that these \ I - this study enigmatic genera may be linked with the platycercine parrots \.I --- Hackett el al. (2008). I \ Pratt et al. (2009) (Leeton et al., 1994). The Philippine endemic Bolbopsittacus 0.8 has traditionally been considered a member of either the Psittaculini (Smith, 1975) or the Cyclopsittacini (Smith, 1981; I\ I\ Collar, 1998), but we revealed it to be the sister taxon of c'(ii I• c 11 Agapornis and Loriculus, in congruence with Wright et al.

Table 3 Nodes for which the biogeographic reconstruction of parrots with Lagrange differed between the two models applied. The numbering of the nodes corresponds to that in Fig. 4. The split format reads as follows: [left|right], where ‘left’ and ‘right’ are the ranges inherited by each descendant branch, with ‘left’ corresponding to the upper branch, and ‘right’ to the lower branch in the phylogenetic tree in Fig. 4. All reconstructions within two log likelihood units (lnL) of the maximum for each node are shown, with the relative probability (Rel. Prob.) being the relative probability (fraction of the global likelihood) of a split. A, Australasian; AT, Afrotropical; IM, Indo Malayan; M, Madagascar; NT, Neotropical.

Unconstrained model Constrained model

Node Split lnL Rel. Prob. Split lnL Rel. Prob.

1 [A|A] 65.78 0.40 [A|A+NT] 64.79 0.43 [A|A+NT] 66.06 0.31 [A|A+AT] 65.05 0.33 [A|A+AT] 66.36 0.2262 [A|A] 65.52 0.21 2 [A|A] 65.88 0.36 [A+NT|A] 64.73 0.46 [A+NT|A] 65.90 0.36 [A+AT|A] 64.99 0.35 [A+AT|A] 66.2 0.26 [A|A] 65.65 0.18 3 [A|NT] 65.60 0.48 [A|NT] 64.62 0.51 [A|AT] 65.91 0.35 [A|AT] 64.89 0.39 [A|A] 67.33 0.09 4 [AT|NT] 65.44 0.56 [AT|NT] 64.39 0.64 [NT|NT] 66.52 0.19 [NT|NT] 65.67 0.18 [A|NT] 67.37 0.08 5 [M|A] 65.41 0.59 [M|A] 64.38 0.65 [A|A] 65.93 0.35 [A|A] 65.19 0.29 6 [A|A] 65.24 0.69 [A|A] 64.39 0.64 [A|IM] 66.64 0.17 [A|IM] 65.3 0.26 7 [IM|A+IM] 65.96 0.34 [IM|A] 65.24 0.27 [A|A] 66.57 0.18 [IM|IM] 65.35 0.25 [IM|A] 66.59 0.18 [A|A] 65.38 0.24 [IM|IM] 66.65 0.17 [IM|A+IM] 66.02 0.13 8 [A|AT] 66.59 0.18 [A|AT] 65.34 0.25 [A+IM|IM] 66.66 0.17 [A|M] 65.42 0.23 [A+IM|A] 66.66 0.17 [IM|AT] 65.98 0.13 [A|M] 66.66 0.17 [IM|M] 66.06 0.12 [IM|AT] 67.15 0.10 [A|A] 66.56 0.07 [IM|M] 67.31 0.09 [A+IM|A] 66.73 0.06 [A|A] 67.62 0.06 [A+IM|IM] 66.73 0.06 [IM|IM] 68.35 0.03 [IM|IM] 67.25 0.04 9 [M|AT] 65.47 0.55 [M|AT] 64.29 0.71 [A|AT] 67.05 0.11 [IM|AT] 67.21 0.10 10 [A|A] 65.15 0.76 [A|A] 64.23 0.75 [A|A+IM] 66.28 0.2439 [A|A+IM] 65.36 0.24 11 [A|A] 65.14 0.76 [A|A] 64.27 0.72 [A|A+IM] 66.42 0.21 [A|A+IM] 65.55 0.20 12 [IM|A+IM] 64.98 0.89 [IM|A+IM] 64.29 0.71 [IM|A] 65.64 0.18 13 [A|A+IM] 64.97 0.90 [A|A+IM] 64.25 0.73 [A|A] 65.74 0.17

et al., 2006; Brown et al., 2007, 2008; Pratt et al., 2009; exclude the possibility that earlier diversity within the group Pacheco et al., 2011). A similar pattern has been found in had gone extinct by the K Pg boundary. mammals, and the mass extinction event at the K Pg boundary does not seem to have had a major influence on the Biogeography, species richness and temporal diversification of today’s mammals (Bininda Emonds et al., diversification patterns within parrots 2007; Nishihara et al., 2009). The diversification of the crown group of the parrots, however, did not start until the The causes of diversification in the large and widely distributed Palaeocene, and well after the K Pg boundary, but we cannot clade of the parrots appear to be complex. However, our

Journal of Biogeography 11 ª 2011 Blackwell Publishing Ltd M. Schweizer et al.

Millions of years ago 75 50 25 0 Palaeocene I Eocene I Oligocene Miocene

CRETACEOUS PALAEOGENE NEOGENE

Figure 4 Biogeographic reconstruction obtained using LAGRANGE of the area splits at the various nodes of parrots. The red circles indicate nodes either at which the reconstruction of the models yielded different splits between areas or where both models considered the same splits to be most likely but more than one split was revealed within two log likelihood units of the maximum for the respective node. The alternative reconstructions for these nodes are given in Table 3, along with the likelihood values.

analysis of temporal and spatial patterns in diversity allowed us force in this cladogenetic event. The constrained and uncon to pinpoint some of the more important events in their strained biogeographic models agreed in their reconstruction evolutionary history. of the ranges of these two taxa, although different scenarios It has been hypothesized that a common ancestor of the were revealed to be similarly likely (node 3, Table 3 ). Arini and the Psittacini Jived in Antarctica and became Biogeographic reconstruction is hampered here by the fact separated from the Australasian lineages when Antarctica that Antarctica cannot be implemented in the model as no began to split from Australia (Tavares et al., 2006; Schweizer recent taxa occur on this continent As a result of climate et al., 2010). The two continents finally separated at about change towards cooler conditions during the F.ocene, conti 40 Ma (Li & Powell, 2001), which corresponds well with our nental ice sheets expanded rapidly on Antarctica in the earliest estimate for the split between those two groups (node 3, Figs 2 Oligocene (Zachos et al., 2001). It was hypothesized that & 4). This agreement between geological and biological parrots then colonized the Neotropics and Africa from evidence indicates that vicariance may have been the major Antarctica, giving rise to the Arini and Psittacini (Tavares

(a) Alinl Clade 1 (84 species) M n; dade 2 (64 opecie•) ---~==~~====~~g~~~~~~~~~~~~~ Poice(JhlJM~uidM(2 (91 specicspedes3)) r P&msws ( 1 species) ,------!------...--- P•NtoO001" ) r_j ,---~~---{======j=~~ ~_t.,.(1opecies){6$PCOie$}

r--- PratyeM:US (8 species) '-I--- 8Mwdius (2 ,pee;.. ) ....------+---- Lllth0mrus (1 $p41CliS) '----1 Ptoscpeia (3 species) .+--- C~us (5 &pecies) '-1---- Ec.u'l~ ( 1 spe.cies) L ----c=:jt:======::::=== CMIC<¥>Sls(2"'8desP$ittrlCM$ (1 pecies) ) rf------..-- A9-,,;.{9$j)e(;&$) LoricvM(13Specieo) ~---+------+---'+------..-- ~( 1 $poci0$) ~-----+---- Mel0p"$itracus(1 species)

Lorini ci.o. 1 (33 $p.eio) lu-. ---c=:i:=== l.Orlni Cl$de 2 (20 $J)f)Cies) Clade a M IC L P.$1itlaculir0$IM (3specie&) ------{=!=== Cydopsitt• (2 $1)0cit$} I 1 o.091 o.954 7.79 ,------f------!;...--l-Wa(6 s-) '----~-! Alisterus, Aprosmiclus. Pol'f'elis (8 species) 2 0.006 0.894 3.40 ,------....--- Ptfonoturw (9specie$) Bg. O.D78 0.019 '--'---ll-----{===:t~~~ PsJttac.ull•. G«lff10yus, P:ltrlt'lus, T"'1ygttMhus (22 $J)ecles) EC/e'Ctv$ ( 1 ~C$) (b) '---;r------;.------..;...------.P..-- Neslor. S'1ipop& (3spec!es)

--- MeanBEASTLTI -- Ci 95% • • • • • • • • Cl So/• ------• = 0(11'°.2.ir-O). a = 0.1 ------a = O(A=0.2, µ=0),a = 10 •••·••• ••• 8 • 0.9 (A •0.2,µ•0.18), Cl • 0.1 -········"a = 0.9 (A =0.2,µ:0.18), o =10

Millions of years ago 50 25 0 Eocene Oligocene Miocene

PALAEOGENE NEOGENE

Figure 5 Diversity tree and semi logarithmic lineage through time (LTI) plot fo r parrots. (a) Diversity tree of parrots used for the MEDUSA analyses. Clades for which we found an indication for an unusual diversification rate are assigned numbers that represent the order in which rate shifts were added by the stepwise Akaike information criterion (AIC) procedure. The estimated net diversification rates (r = A.- µ) and relative extinction rates (a = µ/A.) for the clades denoted by numbers and the background rates (Bg.) are indicated. We revealed the lories (Loriinae) to be exceptionally species rich and have an indication that the clade leading to the Arini showed an increased diversification rate. (b) Semi logarithmic LTI plot of the 1000 last trees from the posterior of the best fitting model obtained in BEAST showing the mean (Mean BE.AST LTI) and the 5% and 95% confidence interval (CI) curves. In addition, the mean curves of the 1000 simulated trees for the two null models with constant diversification rates are given for two extreme values of the scaling parameter ct (see text for further details). Arini clade 1 includes Anadorhynchus, Ara, Aratinga, Cyanoliseus, Cyanopsitta, Deroptyus, Diopsittaca, Enicognathus, Guarouba, Leptosittaca, Nandayus, Pionites, Propyrrhura, Pyrrhura, Ognorhynchus, Orthopsittaca, Rhynchopsitta. Arini clade 2 includes Amazona, Bolborhynchus, Brotogeris, Forpus, Graydidascalus, Hapalopsittaca, Myopsitta, Nannopsittaca, Pionopsitta, Pionus, Psilopsiagon, Touit, Triclaria. Loriinae clade 1 includes Chalcopsitta, Eos, Glossopsitta, Lorius, Neopsittacus, Oreopsittacus, Pseudeos, Psitteuteles, Trichoglossus. Loriinae clade 2 includes Channosyna, Phygis, Vini. et al., 2006; Schweizer et al., 2010). We indeed found this Schweizer et al. (2010) proposed two further major dispersal divergence event to have occurred in the late Eocene or early events during the evolutionary diversification of parrots, Oligocene (at around 35 Ma) (node 4, Figs 2 & 4). namely the colonization of Madagascar from Australasia by a

Journal of Biogeography 13 © 2011 Blackwell Publishing Ltd M. Schweizer et al. common ancestor of Coracopsis, and the colonization of They differ from the remaining parrots in being highly Madagascar and later Africa, again from Australasia, by a specialized to a nectarivorous diet and are characterized by common ancestor of Agapornis. We found that a common several morphological specializations of their feeding tract to ancestor of Coracopsis dispersed to Madagascar and split from nectarivory, including a modified gizzard musculature and a Psittrichas at around 28 Ma, in the middle Oligocene (node 5, brush tip tongue allowing them to harvest nectar rapidly Fig. 4, Table 3; node 9, Fig. 2, Table 3). The split between (Guntert, 1981; Richardson & Wooller, 1990; Collar, 1998). Agapornis and Loriculus was found to have taken place later, at Their diet shift and the associated anatomy may represent a around 24 Ma (node 19, Fig. 2, Table 2). The biogeographic key innovation that promoted their radiation and account for reconstruction revealed several colonization scenarios of their large species richness. The other highly nectarivorous bird Madagascar and Africa by ancestors of Agapornis as almost group of Australasia comprises the honeyeaters (Meliphagi equally likely. We hypothesize that a colonization of Mada dae), which are among the most species rich and most diverse gascar from Australasia followed by dispersal from Madagascar group of birds in this region (Newton, 2003). The ecological to Africa (Schweizer et al., 2010) is most probable, as the relationships of the lories and the honeyeaters with plants are Madagascan endemic Agapornis canus is the sister group to the not as species specific as they are in some other nectarivorous African mainland Agapornis species (nodes 8 and 9, Fig. 4, birds such as the Neotropical hummingbirds or the sunbirds in Table 3). Thus, Coracopsis and Agapornis have independently Africa (Fleming & Muchhala, 2008). Adaptation to and colonized Africa/Madagascar through long distance dispersal coevolution with individual plant species may thus not have across the Indian Ocean from Australasia. Within birds, such a been a primary driving force of the diversification of the lories. biogeographic pattern has so far only been convincingly However, nectar may have provided a spatially widespread proposed for the cuckoo shrikes (Campephagidae) (Jonsson underutilized niche, which would have allowed lories to et al., 2010). expand their geographical ranges and successfully establish Colonization of Indo Malaysia from Australasia occurred populations on oceanic islands, which was often followed by independently several times. Bolbopsittacus, which is endemic allopatric speciation, and eventually followed by secondary to the Philippines, split from its closest relatives in Australasia sympatry through range expansion. Even today, however, about 28 Ma, which probably represents the first colonization congeneric species of the lories occur generally in allopatry of the Indo Malayan region (node 18, Fig. 2, Table 2). This (Collar, 1998). was perhaps facilitated by the East Philippines Halmahera The radiation of the lories started at a time when the LTT South Caroline Arc, which was approaching the Australian plots indicate an accelerated diversification rate for the parrots plate at this time (Hall, 2002). Australasia reached its present as a whole, from the middle Miocene onwards and contem position relative to Indo Malaysia around 20 25 Ma (Li & poraneous with a period of gradual cooling following the Powell, 2001; Hall, 2002), and all other splits between middle Miocene climate transition (MMCT) (Zachos et al., Australasian and Indo Malayan taxa seem to have occurred 2001; Shevenell et al., 2004). During this period, the majority after the two realms came into close contact. The proximity of of the extant genera of parrots arose, similar to the case of these two regions in combination with a complex pattern of other bird groups such as toucans (Ramphastidae) (Patane tectonic movement and the development of archipelagos in et al., 2009), buteonines (Accipitridae) (do Amaral et al., this area could have provided new dispersal opportunities and 2009) and some Passeriformes such as tapaculos, cuckoo facilitated faunal exchange between them (Hall, 1998, 2002). shrikes and other crown Corvida (Jonsson et al., 2008; Mata However, the exact pattern of colonization of the Indo et al., 2009). A major change in vegetation began to take place Malayan region by ancestors of Loriculus, Prioniturus, Psittinus, contemporaneously in the middle/late Miocene, associated Psittacula and Tanygnathus could not be inferred unambigu with the shrinking of rain forests and an increase in areas of ously with our biogeographic reconstruction. It possibly open vegetation in Australia, Africa and South America (Jacobs included dispersal back to Wallacea by Prioniturus and et al., 1999; Martin, 2006). While this led to a fragmentation of Tanygnathus in a way similar to that shown for the cuckoo forest habitat into refugia, the newly opened savanna areas shrikes (Campephagidae) (Jonsson et al., 2008). Groombridge provided new ecological opportunities and may have been et al. (2004) dated the split of Psittacula from other parrots colonized by parrots from forests through niche expansion, between 3.4 and 9.7 Ma using cytochrome b rates estimated for prompting ecological speciation (Tavares et al., 2006; Fjeldsa˚ & other avian species. This is consistent with our estimate Bowie, 2008). The mechanisms of such ecological speciation regarding the split between Psittacula and Psittinus/Tanygna across forest savanna ecotones have been demonstrated in thus at a mean value of 6.25 Ma. other bird species (Smith et al., 2001, 2005). We indeed found The analyses using MEDUSA found the lories to be that several dry adapted Australasian lineages emerged at this unexpectedly species rich given their age (Fig. 5). The lories time (Melopsittacus, Polytelis alexandrae, Northiella/Psephotus). split from their closest relatives in the middle Miocene and are In Africa, Poicephalus, a genus that includes several dry thought to have radiated through the islands off northern landscape taxa, split from Psittacus, a bird of lowland forests, Australia and colonized areas west to Sulawesi and Bali, north then, and this was followed by diversification of Poicephalus to the Philippines, east to several Pacific islands and south to (Collar, 1998). In South America, the uplift of southern Australia (Christidis et al., 1991; Collar, 1998; Schodde, 2006). portions of the Andes (Farıás et al., 2008) followed by the

14 Journal of Biogeography ª 2011 Blackwell Publishing Ltd Macroevolutionary patterns in the diversification of parrots uplift of the Central Andean plateau, beginning in the late influence of climate driven habitat change and geological Miocene (Garzione et al., 2008), led to further habitat change events, especially from the middle Miocene onwards, on the and fragmentation at around the same time, and could have speciation processes of parrots, future studies should attempt promoted the radiation of Neotropical parrots (see below). to integrate knowledge of phylogenetic relationships of closely Finally, environmental changes in Asia following the increased related species into a spatial and temporal framework. uplift of the Tibetan plateau and the onset of the Indian and east Asian monsoons during the late Miocene (Ruddiman & ACKNOWLEDGEMENTS Kutzbach, 1990; Zhisheng et al., 2001) probably also prompted the diversification of parrots there, as was proposed for the We especially thank the Silva Casa Foundation for financial early diversification of the genus Psittacula in south and support of this project, and Marcel Guntert for valuable Southeast Asia by Groombridge et al. (2004). discussions and assistance. We are grateful to S. Birks A much earlier possible increase of net diversification rates (University of Washington, Burke Museum), R. Burkhard, for the parrots as a whole took place around the Eocene/ A. Fergenbauer Kimmel, J. Fjeldsa˚ and J. B. Kristensen (Zoo Oligocene boundary and again coincided with a major global logical Museum, University of Copenhagen), H. Gygax, M.B. climatic aberration. It was characterized by cooler conditions Robbins and A. Nyari (The University of Kansas, Natural History just above the Eocene/Oligocene boundary and Antarctica Museum and Biodiversity Research Center), P. Sandmeier, becoming increasingly ice encrusted (Zachos et al., 2001). T. and P. Walser, D. Willard (Field Museum of Natural History) Diversification around the Eocene Oligocene boundary has andG.Weisforkindlyprovidinguswithtissueorfeather also been found in other bird groups, including auks (Pereira samples. & Baker, 2008), penguins (Baker et al., 2006), trogons (Moyle, Chad D. Brock and Luke J. Harmon kindly provided the code 2005), pigeons and doves (Pereira et al., 2007). During this for non randomly pruning taxa from phylogenetic trees. We period, the major lineages of parrots emerged, and Africa and further thank the following people for valuable support: South America were colonized (see above). These newly S. Bachofner, B. Blochlinger, L. Cathrow, V. de Pietri, M. colonized areas may have provided several ecological oppor Hohn, L. Lepperhof, R. Morales Hojas, M. Rieger and C. Sherry. tunities for the facilitation of diversification. We indeed have an indication based on the analyses with MEDUSA that the REFERENCES clade leading to the Neotropical Arini showed an increased diversification rate and that the Arini may be exceptionally Alfaro, M.E., Santini, F., Brock, C., Alamillo, H., Dornburg, A., species rich given their age. However, there seems to have been Rabosky, D.L., Carnevale, G. & Harmon, L.J. (2009) Nine a time lag between the arrival in South America and the exceptional radiations plus high turnover explain species beginning of the diversification, although this needs to be diversity in jawed vertebrates. Proceedings of the National corroborated with a broader taxon sampling. More generally, Academy of Sciences USA, 106, 13410 13414. avian species richness is higher in South America than in any do Amaral, F.R., Sheldon, F.H., Gamauf, A., Haring, E., Rie other continent, and almost one third of the world’s bird sing, M., Silveira, L.F. & Wajntal, A. (2009) Patterns and species occur there (Price, 2008). High speciation rates in the processes of diversification in a widespread and ecologically Neotropics are probably caused by a spatially and temporally diverse avian group, the buteonine hawks (Aves, Accipitri heterogeneous combination of mechanisms, with a strong dae). Molecular Phylogenetics and Evolution, 53, 703 715. influence from the uplift of the Andes and from Pleistocene Baker, A.J., Pereira, S.L., Haddrath, O.P. & Edge, K.A. (2006) glacial cycles in the Andes (Weir, 2006; Price, 2008; Thomas Multiple gene evidence for expansion of extant penguins out et al., 2008). The beginning of the uplift of the Andes probably of Antarctica due to global cooling. Proceedings of the Royal also promoted the generic diversification of Neotropical Society B: Biological Sciences, 273, 11 17. parrots (see above), and mountain uplift has indeed been Bininda Emonds, O.R.P., Cardillo, M., Jones, K.E., MacPhee, suggested to have driven the diversification of Pionus parrots R.D.E., Beck, R.M.D., Grenyer, R., Price, S.A., Vos, R.A., (Ribas et al., 2007). Gittleman, J.L. & Purvis, A. (2007) The delayed rise of In summary, we found the diversification of parrots to be present day mammals. Nature, 446, 507 512. correlated with extrinsic mechanisms, namely climatic and Brock, C.D., Harmon, L.J. & Alfaro, M.E. (in press) Testing for geological events, but also to be associated with the evolution temporal variation in diversification rates when sampling is of a key innovation. It thus corresponds to a taxon pulse model incomplete and nonrandom. Systematic Biology. (Erwin, 1981; Halas et al., 2005). Initial vicariance events Brown, J.M. & Lemmon, A.R. (2007) The importance of data caused by continental break up were followed by dispersal, partitioning and the utility of Bayes factors in Bayesian range expansion and local radiations. Vegetation shifts in phylogenetics. Systematic Biology, 56, 643 655. combination with mountain orogenesis may have acted as a Brown, J.W., Payne, R.B. & Mindell, D.P. (2007) Nuclear DNA catalyst for the diversification of parrots by providing new does not reconcile ‘rocks’ and ‘clocks’ in Neoaves: a com ecological opportunities and challenges that facilitated ecolo ment on Ericson et al.. Biology Letters, 3, 257 259. gical speciation, as well as by causing habitat fragmentation Brown, J.W., Rest, J.S., Garcia Moreno, J., Sorenson, M.D. & that resulted in vicariance events. To better understand the Mindell, D.P. (2008) Strong mitochondrial DNA support

Journal of Biogeography 15 ª 2011 Blackwell Publishing Ltd M. Schweizer et al.

for a Cretaceous origin of modern avian lineages. BMC Feduccia, A. (2003) ‘Big bang’ for tertiary birds? Trends in Biology, 6, 6. Ecology and Evolution, 18, 172 176. Burnham, K.P. & Anderson, D.R. (2002) Model selection and Fjeldsa˚, J. & Bowie, R.C.K. (2008) New perspectives on the mulitimodel inference, 2nd edn. Springer Verlag, New York. origin and diversification of Africa’s forest avifauna. African Christidis, L., Schodde, R., Shaw, D.D. & Maynes, S.F. (1991) Journal of Ecology, 46, 235 247. Relationships among the Australo Papuan parrots, lorikeets, Fleming, T.H. & Muchhala, N. (2008) Nectar feeding bird and and cockatoos (Aves, Psittaciformes): protein evidence. bat niches in two worlds: pantropical comparisons of ver Condor, 93, 302 317. tebrate pollination systems. Journal of Biogeography, 35, Clarke, J.A., Tambussi, C.P., Noriega, J.I., Erickson, G.M. & 764 780. Ketcham, R.A. (2005) Definitive fossil evidence for the Forshaw, J.M. (1973) Parrots of the world. Landsdowne Press, extant avian radiation in the Cretaceous. Nature, 433, Melbourne. 305 308. Gaina, V., Roest, W.R., Muller, R.D. & Symonds, P. (1998) The Collar, N.J. (1998) Family Psittacidae (parrots). Handbook of opening of the Tasman sea: a gravity anomaly animation. the birds of the world, Vol. 4 (ed. by J. del Hoyo, A. Elliott American Meteorological Society, AMS Journals Online, Earth and J. Sargatal), pp. 280 477. Lynx Edicions, Barcelona. Interactions, 2, 1 23. Cooper, A. & Penny, D. (1997) Mass survival of birds across Garzione, C.N., Hoke, G.D., Libarkin, J.C., Withers, S., the Cretaceous Tertiary boundary: molecular evidence. MacFadden, B., Eiler, J., Ghosh, P. & Mulch, A. (2008) Rise Science, 275, 1109 1113. of the Andes. Science, 320, 1304 1307. Cooper, A., Lalueza Fox, C., Anderson, S., Rambaut, A., Goldberg, J., Trewick, S.A. & Powlesland, R.G. (2011) Popu Austin, J. & Ward, R. (2001) Complete mitochondrial lation structure and biogeography of Hemiphaga pigeons genome sequences of two extinct moas clarify ratite evolu (Aves: Columbidae) on islands in the New Zealand region. tion. Nature, 409, 704 707. Journal of Biogeography, 38, 285 298. Couvreur, T.L.P., Franzke, A., Al Shehbaz, I.A., Bakker, F.T., Groombridge, J.J., Jones, C.G., Nichols, R.A., Carlton, M. & Koch, M.A. & Mummenhoff, K. (2010) Molecular phylo Bruford, M.W. (2004) Molecular phylogeny and morpho genetics, temporal diversification, and principles of evolu logical change in the Psittacula parakeets. Molecular Phy tion in the mustard family (Brassicaceae). Molecular Biology logenetics and Evolution, 31, 96 108. and Evolution, 27, 55 71. Guntert, M. (1981) Morphologische Untersuchungen zur Cracraft, J. (2001) Avian evolution, Gondwana biogeography adaptiven Radiation des Verdauungstraktes bei Papageien and the Cretaceous Tertiary mass extinction event. Pro (Psittaci). Zoololgische Jahrbucher. Abteilung fur Anatomie ceedings of the Royal Society B: Biological Sciences, 268, 459 und Ontogenie der Tiere, 106, 471 526. 469. Hackett, S.J., Kimball, R.T., Reddy, S., Bowie, R.C.K., Braun, Cusimano, N. & Renner, S.S. (2010) Slowdowns in diversifi E.L., Braun, M.J., Chojnowski, J.L., Cox, W.A., Han, K.L., cation rates from real phylogenies may not be real. Sys Harshman, J., Huddleston, C.J., Marks, B.D., Miglia, K.J., tematic Biology, 59, 458 464. Moore, W.S., Sheldon, F.H., Steadman, D.W., Witt, C.C. & Drummond, A.J. & Rambaut, A. (2007) BEAST: Bayesian Yuri, T. (2008) A phylogenomic study of birds reveals their evolutionary analysis by sampling trees. BMC Evolutionary evolutionary history. Science, 320, 1763 1768. Biology, 7, 214. Haddrath, O. & Baker, A.J. (2001) Complete mitochondrial Drummond, A.J., Ho, S.Y.W., Phillips, M.J. & Rambaut, A. DNA genome sequences of extinct birds: ratite phylo (2006) Relaxed phylogenetics and dating with confidence. genetics and the vicariance biogeography hypothesis. PLOS Biology, 4, 699 710. Proceedings of the Royal Society B: Biological Sciences, 268, Ericson, P.G.P., Anderson, C.L., Britton, T., Elzanowski, A., 939 945. Johansson, U.S., Kallersjo, M., Ohlson, J.I., Parsons, T.J., Halas, D., Zamparo, D. & Brooks, D.R. (2005) A historical Zuccon, D. & Mayr, G. (2006) Diversification of Neoaves: biogeographical protocol for studying biotic diversifica integration of molecular sequence data and fossils. Biology tion by taxon pulses. Journal of Biogeography, 32, 249 Letters, 2, 543 547. 260. Erwin, T.L. (1981) Taxon pulses, vicariance, and dispersal: an Hall, R. (1998) The plate tectonics of Cenozoic SE Asia and the evolutionary synthesis illustrated by carabid beetles. Vicari distribution of land and sea. Biogeography and geological ance biogeography: a critique (ed. by G. Nelson and D.E. evolution of SE Asia (ed. by R. Hall and J.D. Holloway), pp. Rosen), pp. 159 196. Columbia University Press, New York. 99 131. Backhuys Publishers, Leiden. Farıás, M., Charrier, R., Carretier, S., Martinod, J., Fock, A., Hall, R. (2002) Cenozoic geological and plate tectonic evolu Campbell, D., Caceres, J. & Comte, D. (2008) Late Miocene tion of SE Asia and the SW Pacific: computer based high and rapid surface uplift and its erosional response in reconstructions, model and animations. Journal of Asian the Andes of central Chile (33 35 S). Tectonics, 27, T1005. Earth Sciences, 20, 353 431. doi:10.1029/2006TC002046. Hall, T.A. (1999) BioEdit, a user friendly biological sequence Feduccia, A. (1995) Explosive evolution in Tertiary birds and alignment editor and analysis program for Windows 95/98/ mammals. Science, 267, 637 638. NT. Nucleic Acids Symposium Series, 41, 95 98.

16 Journal of Biogeography ª 2011 Blackwell Publishing Ltd Macroevolutionary patterns in the diversiﬁcation of parrots

Harmon, L.J., Weir, J.T., Brock, C.D., Glor, R.E. & Challenger, comparative anatomy. II. Analysis and discussion. Zoological W. (2008) Geiger: investigating evolutionary radiations. Journal of the Linnean Society, 149, 1 95. Bioinformatics, 24, 129 131. Maddison, W.P. & Maddison, D.R. (2009) Mesquite: a modular Harshman, J., Braun, E.L., Braun, M.J., Huddleston, C.J., system for evolutionary analysis. Version 2.72. Available at: Bowie, R.C.K., Chojnowski, J.L., Hackett, S.J., Han, K.L., http://mesquiteproject.org. Kimball, R.T., Marks, B.D., Miglia, K.J., Moore, W.S., Martin, H.A. (2006) Cenozoic climatic change and the devel Reddy, S., Sheldon, F.H., Steadman, D.W., Steppan, S.J., opment of the arid vegetation in Australia. Journal of Arid Witt, C.C. & Yuri, T. (2008) Phylogenomic evidence for Environments, 66, 533 563. multiple losses of flight in ratite birds. Proceedings of the Mata, H., Fontana, C.S., Mauricio, G.N., Bornschein, M.R., de National Academy of Sciences USA, 105, 13462 13467. Vasconcelos, M.F. & Bonatto, S.L. (2009) Molecular phy Hedges, S.B., Parker, P.H., Sibley, C.G. & Kumar, S. (1996) logeny and biogeography of the eastern tapaculos (Aves: Continental breakup and the ordinal diversification of birds Rhinocryptidae: Scytalopus, Eleoscytalopus): cryptic diversi and mammals. Nature, 381, 226 229. fication in Brazilian Atlantic Forest. Molecular Phylogenetics Hillis, D.M. & Bull, J.J. (1993) An empirical test of boot and Evolution, 53, 450 462. strapping as a method for assessing confidence in phyloge Mayr, G. (2005) The Paleogene fossil record of birds in Eur netic analysis. Systematic Biology, 42, 182 192. ope. Biological Reviews, 80, 515 542. Ho, S.Y.W. & Phillips, M.J. (2009) Accounting for calibration Mayr, G. (2008) The phylogenetic affinities of the parrot taxa uncertainty in phylogenetic estimation of evolutionary Agapornis, Loriculus and Melopsittacus (Aves: Psittacifor divergence times. Systematic Biology, 58, 367 380. mes): hypotarsal morphology supports the results of Huelsenbeck, J.P. & Ronquist, F. (2001) MRBAYES: Bayesian molecular analyses. Emu, 108, 23 27. inference of phylogenetic trees. Bioinformatics, 17, 754 755. Mayr, G. (2009) Paleogene fossil birds. Springer Verlag, Berlin. Hugall, A.F., Foster, R. & Lee, M.S.Y. (2007) Calibration Moyle, R.G. (2005) Phylogeny and biogeographical history of choice, rate smoothing, and the pattern of tetrapod diver Trogoniformes, a pantropical bird order. Biological Journal sification according to the long nuclear gene RAG 1. Sys of the Linnean Society, 84, 725 738. tematic Biology, 56, 543 563. Muller, K. (2005) SeqState: primer design and sequence sta Jacobs, B.F., Kingston, J.D. & Jacobs, L.L. (1999) The origin of tistics for phylogenetic DNA data sets. Applied Bioinfor grass dominated ecosystems. Annals of the Missouri Botan matics, 4, 65 69. ical Garden, 86, 590 643. Muller, K. (2006) Incorporating information from length Jonsson, K.A., Irestedt, M., Fuchs, J., Ericson, P.G.P., Christi mutational events into phylogenetic analysis. Molecular dis, L., Bowie, R.C.K., Norman, J.A., Pasquet, E. & Fjeldsa, J. Phylogenetics and Evolution, 38, 667 676. (2008) Explosive avian radiations and multi directional Nee, S., Holmes, E.C., May, R.M. & Harvey, P.H. (1994) dispersal across Wallacea: evidence from the Campephagi Extinction rates can be estimated from molecular phylo dae and other crown Corvida (Aves). Molecular Phylo genies. Philosophical Transactions of the Royal Society B: genetics and Evolution, 47, 221 236. Biological Sciences, 344, 77 82. Jonsson, K.A., Bowie, R.C.K., Nylander, J.A.A., Christidis, L., Newton, I. (2003) The speciation and biogeography of birds. Norman, J.A. & Fjeldsa˚, J. (2010) Biogeographical history of Academic Press, London. cuckoo shrikes (Aves: Passeriformes): transoceanic coloni Nishihara, H., Maruyama, S. & Okada, N. (2009) Retroposon zation of Africa from Australo Papua. Journal of Biogeog analysis and recent geological data suggest near simultaneous raphy, 37, 1767 1781. divergence of the three superorders of mammals. Proceedings Kass, R.E. & Raftery, A.E. (1995) Bayes factors. Journal of the of the National Academy of Sciences USA, 106, 5235 5240. American Statistical Association, 90, 773 795. Nylander, J.A.A. (2004) MrModeltest v2. Evolutionary Biology de Kloet, R.S. & de Kloet, S.R. (2005) The evolution of the Centre, Uppsala University, Sweden. spindlin gene in birds: sequence analysis of an intron of the Nylander, J.A.A., Wilgenbusch, J.C., Warren, D.L. & Swofford, spindlin W and Z gene reveals four major divisions of the D.L. (2008) AWTY (are we there yet?): a system for Psittaciformes. Molecular Phylogenetics and Evolution, 36, graphical exploration of MCMC convergence in Bayesian 706 721. phylogenetics. Bioinformatics, 24, 581 583. Leeton, P.R.J., Christidis, L., Westerman, M. & Boles, W.E. Olson, S.L. & Feduccia, A. (1980) Relationships and evolution (1994) Molecular phylogenetic affinities of the night parrot of flamingos (Aves: Phoenicopteridae). Smithsonian Con (Geopsittacus occidentalis) and the ground parrot (Pezoporus tributions to Zoology, 316, 1 73. wallicus). The Auk, 111, 833 843. Pacheco, M.A., Battistuzzi, F.U., Lentino, M., Aguilar, R., Li, Z.X. & Powell, C.M. (2001) An outline of the palaeogeo Kumar, S. & Escalante, A.A. (2011) Evolution of modern graphic evolution of the Australasian region since the birds revealed by mitogenomics: timing the radiation and beginning of the Neoproterozoic. Earth Science Reviews, 53, origin of major orders. Molecular Biology and Evolution, 28, 237 277. 1927 1942. Livezey, B.C. & Zusi, R.L. (2007) Higher order phylogeny of Patane, J.S.L., Weckstein, J.D., Aleixo, A. & Bates, J.M. (2009) modern birds (Theropoda, Aves: Neornithes) based on Evolutionary history of Ramphastos toucans: molecular

Journal of Biogeography 17 ª 2011 Blackwell Publishing Ltd M. Schweizer et al.

phylogenetics, temporal diversification, and biogeography. Ribas, C.C., Gaban Lima, R., Miyaki, C.Y. & Cracraft, J. (2005) Molecular Phylogenetics and Evolution, 53, 923 934. Historical biogeography and diversification within the Pereira, S.L. & Baker, A.J. (2006) A mitogenomic timescale for Neotropical parrot genus Pionopsitta (Aves: Psittacidae). birds detects variable phylogenetic rates of molecular evo Journal of Biogeography, 32, 1409 1427. lution and refutes the standard molecular clock. Molecular Ribas, C.C., Moyle, R.G., Miyaki, C.Y. & Cracraft, J. (2007) Biology and Evolution, 23, 1731 1740. The assembly of montane biotas: linking Andean tectonics Pereira, S.L. & Baker, A.J. (2008) DNA evidence for a Paleo and climatic oscillations to independent regimes of diver cene origin of the Alcidae (Aves: Charadriiformes) in the sification in Pionus parrots. Proceedings of the Royal Society Pacific and multiple dispersals across northern oceans. B: Biological Sciences, 274, 2399 2408. Molecular Phylogenetics and Evolution, 46, 430 445. Ribas, C.C., Miyaki, C.Y. & Cracraft, J. (2009) Phylogenetic Pereira, S.L., Johnson, K.P., Clayton, D.H. & Baker, A.J. (2007) relationships, diversification and biogeography in Neo Mitochondrial and nuclear DNA sequences support a cre tropical Brotogeris parakeets. Journal of Biogeography, 36, taceous origin of Columbiformes and a dispersal driven 1712 1729. radiation in the Paleogene. Systematic Biology, 56, 656 672. Richardson, K.C. & Wooller, R.D. (1990) Adaptation of the Phillips, M.J., Gibb, G.C., Crimp, E.A. & Penny, D. (2010) alimentary tract of some Australian lorikeets to a diet of Tinamous and moa flock together: mitochondrial genome pollen and nectar. Australian Journal of Zoology, 38, 581 sequence analysis reveals independent losses of flight among 586. ratites. Systematic Biology, 59, 90 107. Ricklefs, R.E., Losos, J.B. & Townsend, T.M. (2007) Evolu Pratt, R.C., Gibb, G.C., Morgan Richards, M., Phillips, M.J., tionary diversification of clades of squamate reptiles. Journal Hendy, M.D. & Penny, D. (2009) Toward resolving deep of Evolutionary Biology, 20, 1751 1762. Neoaves phylogeny: data, signal enhancement, and priors. Ronquist, F. & Huelsenbeck, J.P. (2003) MrBayes 3: Bayesian Molecular Biology and Evolution, 26, 313 326. phylogenetic inference under mixed models. Bioinformatics, Price, T. (2008) Speciation in birds. Roberts and Company, 19, 1572 1574. Greenwood Village, CO. Rowley, I. (1998) Family Cacatuidae (cockatoos). Handbook of Pybus, O.G. & Harvey, P.H. (2000) Testing macro evolution the birds of the world (ed. by J. del Hoyo, A. Elliott and J. ary models using incomplete molecular phylogenies. Pro Sargatal), pp. 246 279. Lynx Edicions, Barcelona. ceedings of the Royal Society B: Biological Sciences, 267, 2267 Ruddiman, W.F. & Kutzbach, J.E. (1990) Late Cenozoic pla 2272. teau uplift and climate change. Transactions of the Royal Rabosky, D.L. (2006a) Laser: a maximum likelihood toolkit for Society of Edinburgh: Earth Sciences, 81, 301 314. detecting temporal shifts in diversification rates from Schodde, R. (2006) Australia’s bird fauna today origins and molecular phylogenies. Evolutionary Bioinformatics Online, evolutionary development. Evolution and biogeography of 2, 257 260. Australasian vertebrates (ed. by J.R. Merrick, M. Archer, Rabosky, D.L. (2006b) Likelihood methods for detecting G.M. Hickey and M.S.Y. Lee), pp. 413 458. Auscipub, temporal shifts in diversification rates. Evolution, 60, 1152 Sydney. 1164. Schweizer, M., Seehausen, O., Guntert, M. & Hertwig, S.T. Rabosky, D.L. (2010) Extinction rates should not be estimated (2010) The evolutionary diversification of parrots supports a from molecular phylogenies. Evolution, 64, 1816 1824. taxon pulse model with multiple trans oceanic dispersal Rabosky, D.L., Donnellan, S.C., Talaba, A.L. & Lovette, I.J. events and local radiations. Molecular Phylogenetics and (2007) Exceptional among lineage variation in diversifica Evolution, 54, 984 994. tion rates during the radiation of Australia’s most diverse Shevenell, A.E., Kennett, J.P. & Lea, D.W. (2004) Middle vertebrate clade. Proceedings of the Royal Society B: Biological Miocene Southern Ocean cooling and Antarctic cryosphere Sciences, 274, 2915 2923. expansion. Science, 305, 1766 1770. Rambaut, A. (2008) Figtree 1.2. Published by the author. Simmons, M.P. & Ochoterena, H. (2000) Gaps as characters in Available at: http://tree.bio.ed.ac.uk/software/figtree/. sequence based phylogenetic analyses. Systematic Biology, Rambaut, A. & Drummond, A.J. (2007) Tracer v1.4. Available 49, 369 381. at: http://beast.Bio.Ed.Ac.Uk/tracer. Slack, K.E., Jones, C.M., Ando, T., Harrison, G.L., Fordyce, Ree, R.H. & Sanmartin, I. (2009) Prospects and challenges for R.E., Arnason, U. & Penny, D. (2006) Early penguin fossils, parametric models in historical biogeographical inference. plus mitochondrial genomes, calibrate avian evolution. Journal of Biogeography, 36, 1211 1220. Molecular Biology and Evolution, 23, 1144 1155. Ree, R.H. & Smith, S.A. (2008) Maximum likelihood inference Smith, G.A. (1975) Systematics of parrots. Ibis, 117, 18 68. of geographic range evolution by dispersal, local extinction, Smith, G.A. (1981) The origin of Philippine parrots and the and cladogenesis. Systematic Biology, 57, 4 14. relationships of the guaiabero Bolbopsittacus lunulatus. Ibis, Ree, R.H., Moore, B.R., Webb, C.O. & Donoghue, M.J. (2005) 123, 345 349. A likelihood framework for inferring the evolution of geo Smith, T.B., Schneider, C.J. & Holder, K. (2001) Refugial graphic range on phylogenetic trees. Evolution, 59, 2299 isolation versus ecological gradients. Genetica, 112, 383 398. 2311.

18 Journal of Biogeography ª 2011 Blackwell Publishing Ltd Macroevolutionary patterns in the diversiﬁcation of parrots

Smith, T.B., Calsbeek, R., Wayne, R.K., Holder, K.H., Pires, D. Zachos, J., Pagani, M., Sloan, L., Thomas, E. & Billups, K. & Bardeleben, C. (2005) Testing alternative mechanisms of (2001) Trends, rhythms, and aberrations in global climate evolutionary divergence in an African rain forest passerine 65 Ma to present. Science, 292, 686 693. bird. Journal of Evolutionary Biology, 18, 257 268. Zhisheng, A., Kutzbach, J.E., Prell, W.L. & Porter, S.C. (2001) Stamatakis, A. (2006) RAxML VI HPC: maximum likelihood Evolution of Asian monsoons and phased uplift of the based phylogenetic analyses with thousands of taxa and Himalaya Tibetan plateau since late Miocene times. Nature, mixed models. Bioinformatics, 22, 2688 2690. 411, 62 66. Stamatakis, A., Hoover, P. & Rougemont, J. (2008) A rapid bootstrap algorithm for the RAxML web servers. Systematic SUPPORTING INFORMATION Biology, 57, 758 771. Suchard, M.A., Weiss, R.E. & Sinsheimer, J.S. (2001) Bayesian Additional supporting information may be found in the online selection of continuous time Markov chain evolutionary version of this article: models. Molecular Biology and Evolution, 18, 1001 1013. Appendix S1 Species sampled, museum and collection Swafford, D.L. (2001) PAUP*: phylogenetic analysis using par numbers, GenBank accession numbers for the three genes simony (*and other methods), version 4.0. Sinauer Associates, analysed, and sample type. Sunderland, MA. Appendix S2 Laboratory methods. Tavares, E.S., Baker, A.J., Pereira, S.L. & Miyaki, C.Y. (2006) Appendix S3 Sequence characteristics and model parameters. Phylogenetic relationships and historical biogeography of Appendix S4 Maximum parsimony tree of parrots and other Neotropical parrots (Psittaciformes: Psittacidae: Arini) avian taxa. inferred from mitochondrial and nuclear DNA sequences. Systematic Biology, 55, 454 470. As a service to our authors and readers, this journal provides Tennyson, A.J.D. (2010) The origin and history of New Zea supporting information supplied by the authors. Such mate land’s terrestrial vertebrates. New Zealand Journal of Ecology, rials are peer reviewed and may be re organized for online 34, 6 27. delivery, but are not copy edited or typeset. Technical support Thomas, G.H., Orme, C.D.L., Davies, R.G., Olson, V.A., issues arising from supporting information (other than Bennett, P.M., Gaston, K.J., Owens, I.P.F. & Blackburn, missing ﬁles) should be addressed to the authors. T.M. (2008) Regional variation in the historical components of global avian species richness. Global Ecology and Bioge ography, 17, 340 351. Trewick, S.A. & Gibb, G.C. (2010) Vicars, tramps and assembly BIOSKETCHES of the New Zealand avifauna: a review of molecular phylo genetic evidence. Ibis, 152, 226 253. Manuel Schweizer is a PhD student at the Natural History Upchurch, P. (2008) Gondwanan break up: legacies of a lost Museum of Bern and the University of Bern, Switzerland. His world? Trends in Ecology and Evolution, 23, 229 236. main interests include the molecular systematics, biogeogra Waters, J.M. & Craw, D. (2006) Goodbye Gondwana? New phy, and diversiﬁcation and speciation patterns of parrots and Zealand biogeography, geology, and the problem of circu birds in general. larity. Systematic Biology, 55, 351 356. Weir, J.T. (2006) Divergent timing and patterns of species Ole Seehausen is a professor of ecology and evolution at the accumulation in lowland and highland Neotropical birds. University of Bern and EAWAG, Switzerland. He is interested Evolution, 60, 842 855. in the processes and mechanisms implicated in the origins, Wilgenbusch, J.C., Warren, D.L. & Swofford, D.L. (2004) maintenance and loss of species diversity and adaptive AWTY: a system for graphical exploration of MCMC con diversity. vergence in Bayesian phylogenetic inference. Available at: Stefan T. Hertwig is head curator of the vertebrate animals http://ceb.csit.fsu.edu/awty (accessed 31 January 2011). department of the Natural History Museum of Bern. He is Wright, T.F., Schirtzinger, E.E., Matsumoto, T., Eberhard, J.R., interested in the evolution, phylogeny and systematics of Graves, G.R., Sanchez, J.J., Capelli, S., Mueller, H., Scharp vertebrate taxa and works with morphological and molecular egge, J., Chambers, G.K. & Fleischer, R.C. (2008) A mul data. tilocus molecular phylogeny of the parrots (Psittaciformes): support for a Gondwanan origin during the Cretaceous. Molecular Biology and Evolution, 25, 2141 2156. Editor: Michael Patten

Journal of Biogeography 19 ª 2011 Blackwell Publishing Ltd SUPPORTING INFORMATION Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations Manuel Schweizer, Ole Seehausen and Stefan T. Hertwig Journal of Biogeography

Appendix S1 Species sampled, museum and collection number, GenBank accession numbers for the three genes analysed, and sample type The treatment of Pelecaniformes and Suliformes follows the fifty-first supplement to the American Ornithologist Union’s Check -list of North American Birds (Chesser et al., 2010) and Scopus is included in Pelecaniformes after the results of Hackett et al. (2008)

Species Museum/Collection Collection nb. c-mos Rag-1 Zenk type

Neognathae Psittaciformes Ara macao NMBE 1056952 JF807951 JF807979 JF807965 Tissue Agapornis canus NMBE 1056201 GQ505083 GQ505191 GQ505138 Tissue Agapornis fischeri NMBE 1056202 GQ505084 GQ505192 GQ505139 Tissue Agapornis lilianae NMBE 1056205 GQ505087 - - Tissue Agapornis nigrigenis NMBE 1056203 GQ505085 GQ505193 GQ505140 Tissue Agapornis roseicollis NMBE 1056204 GQ505086 GQ505194 GQ505141 Tissue Alisterus chloropterus NMBE 1056207 GQ505091 GQ505199 GQ505145 Tissue Alisterus scapularis NMBE 1056206 GQ505090 GQ505198 GQ505144 Tissue Aprosmictus jonquillaceus NMBE 1056208 GQ505092 GQ505200 GQ505146 Tissue Amazona aestiva NMBE 1056947 JF807952 JF807980 JF807966 Tissue Amazona dufresniana NMBE 1056948 JF807953 JF807981 JF807967 Tissue Amazona pretrei NMBE 1056949 JF807954 JF807982 JF807968 Tissue Barnardius zonarius NMBE 1056210 GQ505095 GQ505203 GQ505149 Tissue Bolbopsittacus lunulatus FMNH 48738 JF807955 JF807983 JF807969 Tissue Cacatua galerita fitzroyi NMBE 1056235 GQ505120 GQ505231 GQ505173 Tissue Cacatua moluccensis NMBE 1056236 GQ505121 GQ505232 GQ505174 Tissue Calyptorhynchus funereus NMBE 1056233 GQ505118 GQ505229 - Tissue Calyptorhynchus latirostris NMBE 1056234 GQ505119 GQ505230 GQ505172 Tissue Charmosyna pulchella NMBE 1056241 GQ505126 GQ505237 GQ505179 Tissue Coracopsis nigra GQ505114 GQ505224 - Feathers Coracopsis vasa UWBM 85986/2004-001 GQ505113 GQ505223 GQ505167 Tissue Cyanoramphus auriceps NMBE 1056221 GQ505104 GQ505213 GQ505158 Tissue Cyanoramphus novaezelandiae NMBE 1056220 GQ505103 GQ505212 GQ505157 Tissue Cyclopsitta diophthalma GQ505130 - - Feathers Deroptyus accipitrinus NMBE 1056951 JF807956 JF807984 JF807970 Tissue Eclectus roratus NMBE 1056248 GQ505135 GQ505244 GQ505187 Tissue Eos cyanogenia NMBE 1056237 GQ505122 GQ505233 GQ505175 Tissue Eunymphicus cornutus cornutus NMBE 1056223 GQ505106 GQ505215 GQ505159 Tissue Eunymphicus cornutus uvaeensis NMBE 1056224 GQ505107 GQ505216 GQ505160 Tissue Guarouba guarouba NMBE 1056950 JF807957 JF807985 JF807971 Tissue Lathamus discolor NMBE 1056219 GQ505102 GQ505211 GQ505156 Tissue Loriculus catamene Z MUC 115584 GQ505088 GQ505195 GQ505142 Tissue Loriculus galgulus UWBM 73841/2002-006 GQ505089 GQ505196 - Tissue Loriculus philippensis Z MUC 130608 - GQ505197 GQ505143 Tissue Lorius garrulus NMBE 1056240 GQ505125 GQ505236 GQ505178 Tissue Melopsittacus undulatus UWBM 60748/1998-068 - GQ505222 GQ505166 Tissue Micropsitta finschii tristrami UWBM 66040/2000-022 GQ505128 GQ505240 GQ505182 Tissue Micropsitta pusio UWBM 67905/2001-054 GQ505129 GQ505241 GQ505183 Tissue Neophema chrysogaster NMBE 1056227 GQ505110 GQ505219 GQ505163 Tissue Neophema chrysostoma NMBE 1056228 GQ505111 GQ505220 GQ505164 Tissue Neophema pulchella NMBE 1056226 GQ505109 GQ505218 GQ505162 Tissue Neophema splendida NMBE 1056225 GQ505108 GQ505217 GQ505161 Tissue Neopsephotos bourkii UWBM 57542/1996-109 GQ505112 GQ505221 GQ505165 Tissue Nestor notabilis NMBE 1056242 JF807958 GQ505238 GQ505180 Tissue Northiella haematogaster NMBE 1056954 JF807959 JF807986 JF807972 Tissue Pionus menstruus NMBE 1056955 JF807960 JF807987 JF807973 Tissue Platycercus caledonicus NMBE 1056212 GQ505097 GQ505205 GQ505151 Tissue Platycercus eximius NMBE 1056213 - GQ505206 GQ505152 Tissue Platycercus flaveolus NMBE 1056215 GQ505099 GQ505208 - Tissue Platycercus venustus NMBE 1056214 GQ505098 GQ505207 GQ505153 Tissue Poicephalus gulielmi FMNH 390740 JF807961 JF807988 JF807974 Tissue Poicephalus meyeri FMNH 363077 - JF807989 JF807975 Tissue Poicephalus rufiventris NMBE 1056230 GQ505116 GQ505226 GQ505169 Tissue Poicephalus senegalus NMBE 1056231 - GQ505227 GQ505170 Tissue Polytelis alexandrae NMBE 1056209 GQ505093 GQ505201 GQ505147 Tissue Polytelis anthopeplus NMBE 1056657 GQ505094 GQ505202 GQ505148 Tissue Prioniturus discurus NMBE 1056246 GQ505133 - GQ505186 Tissue Prioniturus luconensis NMBE 1056247 GQ505134 - - Tissue Prioniturus montanus FMNH 454964 JF807962 JF807990 JF807976 Tissue Prosopeia tabuensis NMBE 1056252 GQ505105 GQ505214 - Tissue Psephotus chrysopterygius NMBE 1056953 JF807963 JF807991 JF807977 Tissue Psephotus dissimilis NMBE 1056218 GQ505101 GQ505210 GQ505155 Tissue Psephotus varius NMBE 1056217 GQ505100 GQ505209 GQ505154 Tissue Psittacella brehmii KU 12931/91954 JF807964 JF807992 JF807978 Tissue Psittacula eupatria NMBE 1056250 GQ505137 GQ505246 GQ505189 Tissue Psittaculirostris desmarestii NMBE 1056244 GQ505131 GQ505242 GQ505184 Tissue Psittaculirostris edwardsii NMBE 1056245 GQ505132 GQ505243 GQ505185 Tissue Psittacus erithacus erithacus NMBE 1056229 GQ505115 GQ505225 GQ505168 Tissue Psitteuteles goldiei NMBE 1056239 GQ505124 GQ505235 GQ505177 Tissue Psittinus cyanurus NMBE 1056251 - GQ505247 GQ505190 Tissue Psittrichas fulgidus NMBE 1056243 GQ505127 GQ505239 GQ505181 Tissue Purpureicephalus spurius NMBE 1056211 GQ505096 GQ505204 GQ505150 Tissue Tanygnathus megalorhynchus NMBE 1056249 GQ505136 GQ505245 GQ505188 Tissue Trichoglossus johnstoniae NMBE 1056238 GQ505123 GQ505234 GQ505176 Tissue Triclaria malachitacea NMBE 1056232 GQ505117 GQ505228 GQ505171 Tissue Anseriformes Anas AF478185 AF143729 EU738887 Anser AY994065 DQ137227 EU738899 Charadriiformes Larus U88419 AY228799 EU738965 Coraciiformes Coracias caudata AY056916 AF143737 EU738931 Todus angustirostris AF441636 DQ111790 EU738886

Gaviiformes Gavia U88423 DQ137228 EU738953 Galliformes Gallus AY056925 AF143730 EU738891 Falconiformes Falco AY447974 AY461399 AF490155 Passeriformes Acanthisitta chloris AY056903 AY056975 EU738893 Corvus corone AY056918 AY056989 EF568306 Picathartes gymnocephalus AY056950 AY057019 EF568314 Pipra coronata AY056951 AY057020 AF492518 Pitta AY056952 AY057021 EF568299 Pelecaniformes Scopus umbretta AF339323 DQ881830 EU739024 Pelecanus AF339325 DQ881819 EU738992 Suliformes Phalacrocorax carbo AF339332 DQ881821 EU738997 Phoenicopteriformes Phoenicopterus AF339336 DQ881823 EU739001 Podicipediformes Podiceps AF339334 DQ881825 EU739009 Procellariiformes Puffinus U88421 DQ881827 AF490146 Sphenisciformes Eudyptes pachyrhynchus (Spheniscidiae) U88420 DQ137231 Eudyptula minor (Spheniscidiae) EU738944 Palaeognathae Struthioniformes Struthio camelus U88429 AF143727 EU738886

REFERENCES

Chesser, R.T., Banks, R.C., Barker, F.K., Cicero, C., Dunn, J.L., Kratter, A.W., Lovette, I.J., Rasmussen, P.C., Remsen, J.V., Rising, J.D., Stotz, D.F. & Winker, K. (2010) Fifty-first supplement to the American Ornithologists' Union Check -list of North American Birds. The Auk , 127, 726-744. Hackett, S.J., Kimball, R.T., Reddy, S., Bowie, R.C.K., Braun, E.L., Braun, M.J., Chojnowski, J.L., Cox, W.A., Han, K.L., Harshman, J., Huddleston, C.J., Marks, B.D., Miglia, K.J., Moore, W.S., Sheldon, F.H., Steadman, D.W., Witt, C.C. & Yuri, T. (2008) A phylogenomic study of birds reveals their evolutionary history. Science, 320, 1763-1768.

Appendix S2 Laboratory methods

Frozen tissues of parrot taxa collected in the wild or from captivity and feathers sampled from captive parrot taxa were used for the isolation of genomic DNA. We isolated total genomic DNA using peqGOLD tissue DNA Mini Kit following the manufacturer’s rule (Peqlab, Erlangen, Germany). Partial sequences of the three nuclear genes analysed, c-mos, RAG-1 and Zenk (second exon), were amplified with polymerase chain reaction (PCR). We used the primers F944 (Cooper & Penny, 1997), R1550or05 (Overton & Rhoads, 2004) and R1550hb99 (Hughes & Baker, 1999) for c-mos, for RAG-1 we used the primers R8, R11B, R17 and R18 (Groth & Barrowclough, 1999) and for Zenk the primers Z1F and Z9R (Chubb, 2004). PCR reaction volumes consisted of 20 µL containing 10 µL PCR- Master-Mix S (c-mos) or PCR-Master-Mix Y (RAG1, ZENK) (Peqlab), 2-3 µL genomic DNA, 2 µL of each primer with a concentration of 10 µM, and 3-4 µL ddH2O. PCR was performed on a Techne TC- 512 thermo-cycler. C-mos was amplified under the following parameters: initial denaturation at 94 °C for 2 min followed by 33 cycles of denaturation at 90 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 1 min, with a final extension at 72 °C for 5 min. For RAG1 we used the PCR reaction profile published by Groth & Barrowclough (1999) with the initial denaturation step reduced to 2 min. For Zenk, the PCR profile of Chubb (2004) was used, with the annealing temperature set to 53.5 °C. PCR products were examined by gel electrophoresis to confirm the amplification of the target fragment. PCR products were either excised from gels and cleaned using the Wizard® SV Gel and PCR Clean-UP System (Promega, Madison, Wisconsin, USA) or purified directly with the above mentioned kit or the peqGOLD MicroSpin Cycle-Pure Kit (Peqlab). To increase the quantity of DNA for problematic samples, we put the products of two independent PCR runs together before cleaning or performed a second PCR after cleaning. Sequencing was carried out with Microsynth AG (Balgach, Switzerland) using the same primers as for amplification. All three genes were sequenced from both sides leading to complete overlapping fragments for Rag-1 and c-mos and an overlapping fragment of about 600 bp for Zenk. We checked sequencing files with Chromas (Technelysium Pty. Ltd., Shannon, Co. Clare, Ireland) and ambiguities were assigned standard IUB codes.

REFERENCES Chubb, A.L. (2004) New nuclear evidence for the oldest divergence among neognath birds: the phylogenetic utility of zenk (i). Molecular Phylogenetics and Evolution, 30, 140-151. Cooper, A. & Penny, D. (1997) Mass survival of birds across the Cretaceous-Tertiary boundary: molecular evidence. Science, 275, 1109-1113. Groth, J.G. & Barrowclough, G.F. (1999) Basal divergences in birds and the phylogenetic utility of the nuclear RAG-1 gene. Molecular Phylogenetics and Evolution, 12, 115-123. Hughes, J.M. & Baker, A.J. (1999) Phylogenetic relationships of the enigmatic hoatzin (Opisthocomus hoazin) resolved using mitochondrial and nuclear gene sequences. Molecular Biology and Evolution, 16, 1300-1307. Overton, L.C. & Rhoads, D.D. (2004) Molecular phylogenetic relationships based on mitochondrial and nuclear gene sequences for the todies (Todus, Todidae) of the Caribbean. Molecular Phylogenetics and Evolution, 32, 524-538. Appendix S3 Sequence characteristics and model parameters Sequence characteristics and model parameters for the three genes and their different partitions analysed. Base substitution rates and frequencies, proportion of invariable sites and alpha parameters are mean values obtained using MRBAYES v. 3.1.

c-cmos RAG-1 Zenk comb. data set

1 pos 1&2 pos 2 pos 3 pos 1 pos 1&2 pos 2 pos 3 pos 1 pos 1&2 pos 2 pos 3 pos 1 pos 1&2 pos 2 pos 3 pos

Si ze

(bp) 603 1461 1158 3222

%A 0.219800 0.252651 0.286124 0.121382 0.325463 0.326503 0.335419 0.337005 0.298126 0.257298 0.308027 0.270448 0.249529 0.211619 0.279518 0.303892 0.086512 0.296346 0.228649

%C 0.269647 0.216262 0.217350 0.347393 0.213158 0.222136 0.206329 0.198660 0.229682 0.343907 0.250863 0.337257 0.416465 0.341887 0.269483 0.233179 0.361313 0.281223 0.292378

%G 0.302995 0.381224 0.303146 0.300304 0.217079 0.267182 0.233299 0.209460 0.195756 0.172751 0.212827 0.182470 0.160932 0.192508 0.219039 0.262002 0.044418 0.202517 0.215481

%T 0.207558 0.149863 0.193380 0.230921 0.244300 0.184179 0.224953 0.254876 0.276435 0.226045 0.228283 0.209825 0.173074 0.253986 0.231959 0.200926 0.091279 0.219913 0.263492

Model GTR+I+G HKY+G GTR+I+G K80+G GT R+G GTR+I+G GTR+I+G GT R+I+G GTR+I+G HKY+G GTR+I+G HKY+I+G GTR+I+G GT R+G GT R+G GTR+I+G GTR+I+G GTR+I+G GTR+I+G GT R+G

Ti/Tv ratio - 2.2774 - 2.9915 - - - - - 3.9306 - 2.9181 ------

A-C 0.073289 0.083916 0.082563 0.069933 0.121893 0.096924 0.054797 0.046393 0.086133 0.1278 03 0.048844 0.062715 0.105565 0.086512 0.070582 0.057965

A-G 0.312924 0.300355 0.431352 0.336512 0.276665 0.319014 0.339043 0.586470 0.480900 0.356003 0.590412 0.407683 0.350923 0.361313 0.363821 0.462922

A-T 0.038483 0.033927 0.061340 0.042458 0.054611 0.042575 0.039827 0.039197 0.046870 0.064308 0.045860 0.042683 0.044912 0.044418 0.049776 0.048031

C-G 0.060780 0.089728 0.037412 0.071951 0.066888 0.112473 0.166408 0.049706 0.088640 0.226509 0.035188 0.064943 0.073500 0.091279 0.124485 0.045876

C-T 0.446359 0.417884 0.330858 0.430789 0.430270 0.383070 0.353184 0.222930 0.223325 0.193732 0.238047 0.363181 0.355242 0.350400 0.344877 0.332356

G-T 0.068164 0.074192 0.056474 0.048356 0.049672 0.045944 0.046741 0.055304 0.074132 0.031644 0.041649 0.058794 0.069859 0.066078 0.046459 0.052849

Prop

Inv 3.176824 - 0.467575 - - 0.280943 0.461377 0.405256 0.465930 - 0.479239 0.553861 0.691722 - - 0.322691 0.473278 0.471075 0.062123 -

Alpha 0.492227 0.186758 0.467697 0.189851 1.966053 1.072995 14.490920 1.505734 0.593492 1.495754 0.836498 12.908233 1.492818 0.183171 1.161594 0.885879 1.389695 0.077483 0.556412 1.271899

Appendix S4 Maximum parsimony tree of parrots and other avian taxa Strict consensus tree from the maximum parsimony analysis of the combined data set of parrots and other avian taxa with gaps treated as a fifth character state. Bootstrap values above 50 are indicated at each node.

Chapter 5

NECTARIVORY IN PARROTS IS A KEY INNOVATION THAT TRIGGERED PARALLEL ADAPTATIONS AND SPECIES PROLIFERATION THROUGH A NONADAPTIVE RADIATION

Manuel Schweizera, Marcel Günterta, Ole Seehausenb,c, Christoph Leuenbergerd, Stefan T. Hertwiga

Manuscript

97 Evolutionary diversification and biogeography of parrots

Abstract

Specialization to nectarivory is associated with several radiations within different bird groups including the parrots. One of them, the Australasian lories, were shown to be unexpectedly species-rich. Their shift to nectarivory may have created an ecological opportunity promoting species proliferation. Several morphological specializations of the feeding tract to a nectarivorous diet have been described for parrots. However, these putative adaptations have never been assessed in a quantitative framework taking phylogenetic non-independence into account. Using a phylogenetic comparative approach based on a broad taxon sampling and 15 continuous characters of the digestive tract, we revealed the nectarivorous species to differ in several traits from the remaining parrots. These trait-changes may have allowed them to effectively feed on nectar and may indicate a phenotype-environment correlation and parallel evolution by natural selection. The ecological expansion to nectarivory did not lead to a change in the rate of morphological evolution in all nectarivorous species, but was associated with significant rate-shifts in morphological evolution at the base of the radiation of the lories as shown by a newly developed statistical approach. Their diet shift can be considered as an evolutionary key innovation which promoted significant non-adaptive lineage diversification through allopatric partitioning of the same new niche. The lack of increased rates of cladogenesis in other nectarivorous parrots indicates that evolutionary innovation do not have to be associated 1 to 1 with diversification events.

Keywords: comparative methods, digestive tract, key innovation, morphological adaptations, nectarivory, nonadaptive radiation, parallel evolution, Psittaciformes, rates of evolution

98 Chapter 5

Introduction

Although most flowering plants are pollinated by insects, a considerable number of tropical angiosperms are pollinated by birds and bats specialized on nectarivorous diets (Bawa 1990; Sekercioglu 2006; Fleming and Muchhala 2008). The associated ecological specializations resulted in several radiations of nectarivorous birds and bats in the tropics and subtropics (Fleming and Muchhala 2008). Nectarivory has evolved convergently in several groups of birds with the Neotropical hummingbirds (Trochilidae, 325-340 species), the Australasian honeyeaters (Meliphagidae, 182 species) and the sunbirds (Nectariniidae, 132 species) of Africa and Australasia, representing the major radiations of nectarivorous birds. Additionally, nectarivory can also be found in several groups of parrots (Psittaciformes). Parrots represent one of the most species-rich groups of birds. While they feed mainly on seeds and fruits, the chiefly Australasian Loriinae (lories) are specialized on a nectarivorous diet (Collar 1998; Rowley 1998). The lories consist of 53 species (Collar 1998) and are considered generalized flower visitors with eucalypts being a particularly important nectar source for them (Fleming and Muchhala 2008). Besides the lories, the swift parrot Lathamus discolor of Australia, the genus Loriculus of Australasia and Indo- Malaysia as well as the genus Brotogeris of the Neotropics are all supposed to depend on nectar as food (Homberger 1980; Güntert 1981; Forshaw 1989; Collar 1998). Their specialization to nectarivory has evolved in convergence to that of the lories (Wright et al. 2008; Schweizer et al. 2010; Schweizer et al. in press). Nectar is a liquid food source rich in sugars, which account for almost 100% of its dry weight (e.g. Lüttge 1976; Gartrell 2000). However, it contains only a small amount of amino acids, which is too low to satisfy the nitrogen requirements of a bird (Martínez del Rio 1994). Therefore nectarivorous birds have to rely on other nitrogen sources like insects or pollen (Richardson and Wooller 1990; Brice 1992; van Tets and Nicolson 2000; Nicolson and Fleming 2003). Several morphological and physiological specializations to nectarivory have been described in the different nectarivorous bird groups (Schlamowitz et al. 1976; Brown et al. 1978; Richardson and Wooller 1986; Casotti and Richardson 1993; Casotti et al. 1998; Schuchmann 1999; Gartrell 2000; Gartrell et al. 2000; Nicolson and Fleming 2003; Downs 2004) and the specialized bill structure of some hummingbirds is even considered a result of coevolution with the morphology of pollinated flowers (Feinsinger and Colwell 1978; Temeles and Kress 2003). The use of a new food source like nectar may be considered as an evolutionary key innovation in the sense of creating an ecological opportunity and of promoting species proliferation associated with an expansion in the use of energy (Vermeij 1995; Yoder et al. 2010). In nectarivorous parrots, this may be particularly true for the lories, which were found to be unexpectedly species-rich given their age compared to the remaining parrot lineages (Schweizer et al. in press). Ecological opportunities may lead to strong directional selection and fast adaptation associated in some cases with adaptive radiations (Hunter 1998; Kassen 2009; Yoder et al. 2010). Indeed, several morphological specializations of the feeding tract to a nectarivorous diet have been described for parrots, which may have been essential for them to effectively feed on nectar and pollen. The lories and Lathamus have muscular tongues with a brush tip allowing them to rapidly harvest nectar (Churchill and Christensen 1970; Güntert and Ziswiler 1972; Richardson and Wooller 1990; Gartrell and Jones 2001). For the lories, it was further reported that they have shortened intestines, size-reduced gizzards with reduced muscularity and koilin layers as well as special adaptations in the esophagus,

99 Evolutionary diversification and biogeography of parrots proventriculus and intestine (Güntert 1981; Richardson and Wooller 1990). The Loriculus species analyzed so far and Lathamus shared some of the adaptations of the lories in the oesophagus, proventriculus and intestine (Güntert 1981). However, they were found to have comparatively more muscular gizzards and longer intestines probably allowing them to feed on hard food like insects or seeds (Güntert and Ziswiler 1972; Güntert 1981; Gartrell 2000; Gartrell et al. 2000). Thus, within parrots, especially the lories appeared to have gastrointestinal tracts highly adapted to nectarivory (Gartrell and Jones 2001). The morphological similarities among the different nectarivorous groups may indicate parallel or convergent evolution driven by natural selection under similar ecological conditions. However, matching morphological and ecological traits shared among the members of a closely related group of species, such as the lories, may reflect phylogenetic contingency rather than a high level of adaptation (e.g. Felsenstein 1985; Freckleton 2009). Data on the morphometrics of the digestive tract of parrots were never analyzed correcting for phylogenetic non-independence and the putative adaptations to nectarivory as described above have hence never been statistically assessed. Additionally, the comparisons of Richardson & Wooller (1990) and Gartrell et al. (2000) were based on a small taxon sampling. In the present study, we tested if the morphological variation of the nectarivorous parrots can be considered as a phenotype-environment correlation and indeed reflects dietary adaptations. We therefore applied a phylogenetic comparative approach using generalized least squares (PGLS) ANCOVA implementing diet as a covariate in the model to test for phenotype-environment correlations predicted by the adaptation formulated by previous authors. Moreover, we investigated if diet shifts to nectarivory in parrots influenced the rate of morphological evolution. If further specialization followed the ecological expansion, this would lead to an increased morphological disparity and increased evolutionary rates. We therefore tested for each trait if the nectarivorous species have different rates of morphological evolution than the remaining parrots by estimating maximum likelihood Brownian rate parameters on the basis of a time-calibrated molecular phylogeny. If the ecological expansion was not followed by further morphological specializations, we would expect only an increased rate of trait evolution at the beginning of the diet shift. The lories apparently show the strongest dependence on nectar. We thus additionally tested, if the origin of nectarivory was associated with significantly hastened morphological evolution at the base of their radiation. For this purpose, we developed a new statistical approach to test if a subset of species in a phylogenetic tree shows a shift in the rate of trait evolution. All analyses were based on 15 continuous characters of the digestive tract and a broad taxon sampling of 78 parrot species consisting of representatives of all major groups. A phylogenetic hypothesis was obtained using three nuclear exons and one mitochondrial gene.

Material and Methods

Dissection and morphological measurements Measurements of the digestive tract were taken from 354 individuals of parrots (Appendix 1). The data corresponds to measurements presented in Güntert (1981) complemented with 15 additional species consisting of 19 Individuals. Body mass was calculated for every species as the mean of the fresh dead or frozen and thawed specimens dissected. All weights were taken to the nearest 0.1 g. In Micropsitta

100 Chapter 5 finschi and Loriculus philippensis body mass values were either taken directly from the literature (Mayr 1931; Rand and Rabor 1960) or an average value was calculated combining literature data and the fresh weight of other dissected specimens not used in this analysis. All measurements either were taken to the nearest 0.1 mm using calipers under a dissecting microscope for some or to the nearest mm using dividers for longitudinal measurements (length of esophagus, glandular stomach, and intestine). All the digestive organs were eventually fixed in buffered formalin (4%) and are stored in the vertebrate collection of the Natural History Museum Bern. The digestive tract was removed from specimens and spread out by cutting through the mesenteria under a watery solution of 0.75% NaCl (isotonic for birds). When specimens had been preserved in formalin before dissection, it was no longer possible to straighten out the intestine, and the length had to be measured by means of a sewing thread, with which all the curvatures of the intestinal loops could be followed exactly. As parrots lack Brunner’s glands (glandulae duodenales) and caeca (Ziswiler and Farner 1972; Güntert 1981), it is not possible to subdivide the intestine into different sections. Length of intestine was measured from the pyloric orifice of the gizzard to the rectal widening into the cloaca.

Figure 1. Longitudinal section through the epithelium at the border (arrow) between the esophagus (right) and the proventriculus (left) of Psittrichas fulgidus. The compound glands (CG) of the proventriculus can be clearly distinguished from the mucous glands (MG) of the lower part of the esophagus.

The esophagus is tripartite in parrots, consisting of a pars cervicalis (beginning at the posterior end of the larynx), the ingluvies (crop) and a pars thoracica that leads into the glandular stomach. Esophageal glands are restricted to the caudal area of pars thoracica. Length of the esophagus was defined as the distance from the caudal rim of the larynx to the border between esophageal and gastric glands (Fig. 1). To determine the extension of esophagus glands and the transition between the glandular part of the glandular stomach and its intermediate zone (see below), the digestive tube was cut open longitudinally. The proventriculus or glandular stomach contains the gastric compound glands. This glandular part is followed by an intermediate zone (zona intermedia), lined with mucous glands. Total length of the proventriculus was measured from the first

101 Evolutionary diversification and biogeography of parrots compound glands visible through the wall of the organ to the entrance into the gizzard. The caudal measuring point was the cranial groove, situated on the pyloric side of the proventricular tube (Fig. 2). The extent of the intermediate zone was computed as the difference of total length minus the glandular part (distance between the first and the most caudal compound glands).

Figure 2. Schematic illustration of a caudal view and a transverse section along the median plain between the tendineal centres (right) of the gizzard. The measurements taken in this study are indicated. Modified from Ziswiler (1967).

The gizzard or muscular stomach has two opposing pairs of antagonistic muscles (Fig. 2, 3). Its inner surface is lined with a tough koilin membrane (Akester 1986), cuticle of McLelland (1979), formed by mucosal glands. As external dimensions, we measured gizzard height (distance between the cranial and caudal groove), gizzard depth (minimum distance between tendineal centres on the two flat sides of the gizzard) and gizzard width (maximum distance at right angles of gizzard height) (Fig. 2). Maximum height at main muscles (MHM, thick muscle pair) and maximum height at thin muscles (MHT) were measured along the maximal extension of each muscle pair. Width of caudoventral thin muscle (WTM) was measured from the caudal groove to its outermost muscle bundles on the opposite side. Width of the lumen plus koilin layer (LWiK, including the distinctly visible tunica mucosa) was measured along the axis of the maximum height at main muscle MHM. Thickness of the two thick muscles (MMT) was calculated as the difference between MHM and LWiK.

102 Chapter 5

Lumen height (MLT) was quantified as the maximum distance between the opposite walls of the cranial and caudal sac.

Figure 3. Schematic illustration of a transverse section through the gizzards along the median plain between the tendineal centres of a nectarivorous (Vini australis, left) and a granivorous parrot (Neophema chrysostoma, right).

Phylogenetic analyses To control for phylogenetic non-independence among species, we used a phylogenetic reconstruction based on partial sequences of the three nuclear genes c-mos, RAG-1 and Zenk (second exon) and of the mitochondrial gene NADH dehydrogenase 2 (ND2) (Appendix 2). Pitta and Falco were used as outgroups and the trees were rooted with the latter taxon, but both were subsequently pruned from the tree for statistical analyses. Sequences of the three nuclear genes were taken from Schweizer et al. (2010) or newly generated following the laboratory protocol described in that study. ND2 sequences were taken from GenBank or generated using the primers MetL and ASNH for PCR amplification and sequencing from both sides (Tavares et al. 2006). The laboratory methods followed Schweizer et al. (2010) using the PCR Protocol of Tavares et al. (2006) for ND 2 with the annealing temperature set to 53°C. The alignment of the sequences was done manually with BioEdit 7.0.5.2 (Hall 1999). We checked individual sequences and the whole alignment further for quality by searching for apparent stop codons after the translation of sequences into amino acids. The final alignment was 4254 bp in length with 603 bp from c-mos, 1461 bp from RAG-1, 1149 bp from Zenk and 1041 bp from ND2. It contained one indel of four amino acids for c-mos, one indel of three amino acids and one indel of one amino acid

103 Evolutionary diversification and biogeography of parrots for RAG-1, while for Zenk, there were four indels of one amino acid and one indel of two amino acids. There were no ambiguously aligned amino acids. Phylogenetic analyses were conducted with maximum likelihood (ML) using RAxML 7.0.4 (Stamatakis 2006). The program was run on the Web-server with 100 rapid bootstrap inferences with all free model parameters estimated by the software (Stamatakis et al. 2008). We tested different biologically relevant parameter settings for our concatenated dataset corresponding to separate models of nucleotide substitution for genes and/or codon positions as estimated by RAxML. Akaike's information criterion (AIC) was used as a heuristic indicator for the fit of the different parameter settings (Akaike 1974), and using separate models of nucleotide substitution for the three codon positions of each gene separately was found to be the best-fitting model. This best-scoring ML tree with branch lengths of the best parameter setting was then used for further analyses. We moreover generated a time-calibrated phylogeny in order to estimate maximum likelihood rate parameters for each morphological trait (see below) using Beast 1.6.1 (Drummond and Rambaut 2007). A relaxed molecular clock with uncorrelated lognormal distribution of branch lengths and a Yule tree prior was applied. We used a calibration on the basis of a molecular dating analysis of the parrots which was calibrated with well-accepted avian fossils outside the parrots (Schweizer et al. in press). Following the results of this study, we used a mean estimate of 58 Ma (millions of years ago) with a normal distribution (95% +/- about 20 million years or SD = 10) for the initial split within the crown group parrots separatingNestor (and Strigops) from the remaining taxa. This time interval includes similar estimates from other studies for this split (Pratt et al. 2009; Pacheco et al. 2011; White et al. 2011). Default prior distributions were chosen for all other parameters and the MCMC was run for 25 million generations with sampling every 1000 generations. Tracer 1.5 (Rambaut and Drummond 2007) was used to confirm appropriate burn-in and the adequate effective sample sizes of the posterior distribution. The resulting maximum clade credibility tree was visualized with FigTree 1.2.1 (Rambaut 2008).

Statistical Analyses, Phylogenetic generalized least squares (PGLS) ANCOVA PGLS ANCOVA (Grafen 1989; Garland and Ives 2000) was used to estimate the relationships between our dependent and independent variables. This method incorporates a matrix of variance and covariance into the calculation of regression parameters based on the pattern of relatedness among species. The variance- covariance matrix was calculated using the best-scoring ML tree. To asses the strength of phylogenetic signal in our data, we incorporated the parameter Ā in our model which varies between 0 and 1 (Pagel 1997, 1999; Freckleton et al. 2002). Values of Ā close to 1 imply that traits covary as assumed by a Brownian motion model with the original tree recovered, while values of Ā close to 0 imply that there is almost no phylogenetic signal in the trait data with the phylogenetic tree for the trait having a single polytomy at the basal node (Freckleton et al. 2002; Blomberg et al. 2003; Freckleton 2009). Ā can be interpreted as having one component of the residuals evolving under a Brownian motion model while another additive component has no phylogenetic correlation (Housworth et al. 2004; Lavin et al. 2008). Freckleton et al. (2002) showed that Ā is statistically powerful in detecting if the data shows a phylogenetic signal, robust to incomplete phylogenetic information and that it performs better than Grafen’s (Grafen 1989, 1992) Ȁ transformation. We implemented body mass as the independent variable, because gut measurements of birds are known to be allometrically related to body mass (e.g. Ricklefs 1996; Lavin et al. 2008), and

104 Chapter 5 the morphometric distances described above were used as dependent variables. For all statistical analyses, both the independent and dependent variables were natural-log transformed. To test for significant phenotype-environment correlations between the different traits in the digestive tract and nectarivory, we used diet (nectarivory of the lories, Brotogeris Lathamus and Loriculus versus the more general diets of other parrots) as covariate in the model. We considered increasingly complex models, beginning with simple allometry between the dependent and independent variables, then diet was included as covariate using ANCOVA with different intercepts but the same slope and finally using ANCOVA with different intercepts and different slopes. PGLS ANCOVA including the parameter Ȁ were fitted by maximizing the restricted log- likelihood. To check for heteroscedasticity, plots of the residuals versus the fitted values were investigated. Akaike's information criterion (AIC) was used as a heuristic indicator for the fit of the different models (Akaike 1974) and we considered an increase in model-fit as significant when the reduction in AIC score in a more complex model was Ā 4 (Burnham and Anderson 2002). We also compared the AIC of the models accounting for phylogenetic non-independence with normal general least square approaches which imply full independence of the data. All statistical analyses were performed in R 2.9.0 (R Development Core Team 2009) using the ape and nlme packages (Paradis et al. 2004; Pinheiro et al. 2009).

Evolutionary rates To test if morphological trait evolution differed between nectarivorous and non- nectarivorous parrots, the approach of Price et al. (2011) was applied. We estimated maximum likelihood Brownian rate parameters for each morphological trait implementing a code written for R by Liam Revell (available from http://anolis.oeb.harvard.edu/~liam/R-phylogenetics/) based on O'Meara et al. (2006). We used stochastic character mapping (Huelsenbeck et al. 2003) as implemented in SIMMAP 1.0 (Bollback 2006) to reconstruct the history of nectarivory in parrots. The maximum clade credibility tree from the analysis with Beast was used and character histories were sampled over this tree. Parameter estimates were then integrated over these sampled histories. We tested two models for each morphological trait. The first model fitted a single Brownian rate of morphological evolution on the whole tree implying the same rate for nectarivory and non-nectarivory. The second model was a two-rate model allowing nectarivorous species to have different rates of morphological evolution than non-nectarivorous species. The modified Akaike's information criterion (AICc) was used as a heuristic indicator for the fit of the different models following Price et al. (2011). To incorporate uncertainty in the history of the evolution of nectarivory in parrots, differences in average AICc score were calculated across 100 random character histories. As all morphological traits were associated with body mass, we used size-corrected values for each trait (cf. Price et al. 2010). We therefore calculated independent contrasts for each natural-log transformed trait based on the maximum likelihood phylogeny and performed a regression through the origin of the contrasts of each trait against natural-log transformed body mass contrasts. The slope was then fitted to the original natural-log transformed data and the residuals from the regression line were then taken for further calculations (cf. Revell 2009). We developed a new statistical approach to additionally test if the origin of nectarivory was associated with significantly hastened morphological evolution at the base of the lory radiation (Appendix 3). In general, this method tests if a subset of

105 Evolutionary diversification and biogeography of parrots species in a phylogenetic tree shows a shift in the rate of trait evolution. It is based on a test statistic whose distribution is known and which thus allowed us to determinep - values in the context of a significance test. The basic model is a generalized linear model (GLM), similar to the one described in Martins & Hansen (1997) and in Garland & Ives (2000). The vector y in equation (1) (see Appendix 3) represents, for each of the N species, the logarithmic values of the measured traits which are considered as dependent variables. The matrix A consists of a column of ones (yielding the intercept of the regression line) and a column of logarithmic weights. The linear dependence between the logarithms of weight and measured trait for each species models the assumption of an allometric relationship between these quantities. The normally distributed error term Ȁ reflects the hypothesis that under absence of selection the measured trait evolves according to a geometric Brownian motion (and thus the logarithm of the trait value follows a standard Brownian motion). The estimated phylogenetic relationship between the species induces a strong correlation in the error term. In fact, the (i,j)-th element in the variance-covariance matrix Ȁ, corresponding to two species i and j, is proportional to the total length of the shared branches in the phylogenetic tree from root to the last common ancestor of i and j. The unknown factor of proportionality Ā represents the drift speed of the Brownian movement and has to be estimated from the data. A Ā-value of e.g. two would indicate that the relative rate of trait-change under neutral evolution is twice as fast as the rate of change in the genes used to estimate the variance-covariance matrix based on a phylogenetic tree. Selection pressure on a subset of K species will result in a disproportionate change of the intercept value for these species, a change that cannot be explained by Brownian drift (neutral evolution) alone. The additive term in equation (1) containing the parameter 3 models this possibility of selection pressure for the given subset of species. The null-hypothesis of no selection can be written as 3 = 0 and the alternative hypothesis corresponds to 3 Ā 0. Large values of 3 will support the alternative hypothesis and result in large absolute values of the test statistic defined in (3). Under the null-hypothesis (absence of selection) the distribution of the test statistic is a t-distribution, see Theorem 1. Its proof is based on the technique of restricted least squares; the only non-standard feature in our situation is the presence of correlation, called heteroscedasticity, in the error term. In order to reduce our model to the standard situation in restricted least squares, we have first to de-couple the error terms. The known distribution of the test statistic allowed us to report two- sided p-values. As always in significance testing, low p-values support a rejection of the null-hypothesis.

106 Chapter 5

Figure 4. Best-scoring maximum likelihood tree including bootstrap values above 70% indicated at nodes. This tree was used for the formulation of the hypothesis of the variance-covariance matrix of the error terms in the regression analyses. Nectarivorous species are shown in red.

Results

Phylogenetic relationships Our phylogenetic tree from the maximum likelihood analyses was in good agreement with Schweizer et al. (2010) (Fig. 4). The maximum clade credibility tree of the analysis with Beast did not differ from the maximum likelihood tree in any supported node (Fig. 7). The lories were revealed as a robustly supported monophyletic clade clustering as the sister group of Melopsittacus undulatus. The relationships within them were also highly supported with the exception of the monophyly of the genus Trichoglossus and the position of Lorius and Chalcopsitta. The division of the lories into two clusters is in agreement with Wright et al. (2008). The position of Loriculus as the sister group of Agapornis was also robustly resolved and in agreement with other studies (Wright et al. 2008; Schweizer et al. 2010; Schweizer et al. in press). Within Platycercini, the sister group relationship of Lathamus to a clade consisting of Prosopeia, Cyanoramphus and Eunymphicus was highly supported in congruence with other

107 Evolutionary diversification and biogeography of parrots studies (Schweizer et al. 2010; Joseph et al. 2011; Schweizer et al. in press). The position of Brotogeris within Arini was in congruence with Tavares et al. (2006) and Wright et al. (2008), but not supported in our data. Unlike Schweizer et al. (2010), the cluster consisting of Psittacidae except Arini and Psittacini was robustly supported as was the clade consisting of Coracopsis and Psittrichas. The monophyly of a clade consisting of Loricoloriinae, Platycercini, Psittaculini and Micropsittini was not robustly supported. Within Platycercini, the position of Prosopeia tabuensis, Northiella haematogaster, Psephotus haematonotus and of the species within the genus Platycercus could not be resolved with high support values. In congruence with Wright et al. (2008), Arini were split into two clades, however, the position of Forpus could not be resolved. With the exception of the position of Forpus, the topology within the Neotropical taxa was also in agreement with the results of Tavares et al. (2006). The genus Amazona was found not to be monophyletic with Amazona xanthops not clustering with the other species of this genus in accordance with results from other molecular studies (Russello and Amato 2004; Tavares et al. 2006).

Morphology For linear body dimensions, a scaling with the 0.33 power of the body mass can be expected (Schmidt-Nielsen 1984). Indeed, we found scaling factors for all linear dimensions lying around this expected value ranging between 0.18 and 0.46. With the exception of gizzard lumen width including koilin layer (LWiK), all other measurements showed a significant phylogenetic signal as indicated by lower AIC when incorporating phylogenetic information in the model as compared to a generalized least squares regression (Table 1). For length of proventriculus, gizzard width and MLT, a different ANCOVA model was favored by PGLS ANCOVA compared to a normal general least squares approach. For eight traits, a model including diet as a covariate with different intercepts but the same slope was considered as the best-fitting model, while a simple allometric model explained best the data for the remaining traits (Table 1, 2, Fig. 5, 6). The specialization to nectarivory led to a decrease of the extension of the esophagus glands, though the trait-values of Lathamus were seemingly more similar to the non- nectarivorous parrots and there was some variation within the lories. Furthermore, the length of the intermediate zone was found to be prolonged in nectarivorous parrots. The nectarivorous species clearly differed from the remaining parrots in the measurements of the gizzard, with the exception of: width of whole gizzard and its lumen, width at the caudoventral thin muscle (WTM) and the maximum lumen at thin muscles (MLT). Brotogeris and Lathamus were seemingly more similar to the non- nectarivorous parrots for all gizzard traits. This was apparently also the case for Loriculus galgulus in gizzard mass, gizzard height, gizzard thickness at main muscles (MMT) and maximum gizzard height at main muscle (MHM). Moreover, there was some variation within lories especially for gizzard mass and depth and gizzard thickness at main muscles (MMT). The gizzard measurements of Psittrichas fulgidus showed a clear tendency to be more closely associated with the nectarivorous species than with the non-nectarivorous parrots. In contrast, including diet as covariate in the model did not improve model-fit for the length of the esophagus and of the intestine. While an intermediate value of Ā was revealed for the esophagus length, the length of the intestine showed a high phylogenetic signal with a value of Ā close to one (Table 1). When diet was included in the model for the length of the proventriculus, the AIC was slightly lower for an

108 Chapter 5

ANCOVA with the same slopes and different intercepts compared to simple allometry (Fig. 5, Table 1). However, the more complex model was not substantially supported. Hence, the length of the proventriculus is best explained by a simple allometry.

Table 1. Akaike's information criterion (AIC) for standard regressions (GLS) and phylogenetic generalized least squares (PGLS) including the fitted Ā-values for the different models considered for each trait of the digestive tract. Increasingly complex models were tested, beginning with simple allometry between the dependent and independent variables, then diet (nectarivory) was included as covariate using ANCOVA with different intercepts but the same slope (Mass + Food) and finally using ANCOVA with different intercepts and different slopes (Mass x Food).

trait model AIC (gls) AIC (pgls) Ā

length of intestine simple allometry 38 906 -10 934 0 940 Mass + Food 40 249 -6 783 0 942 Mass x Food 43 976 -5 138 0 952 length of esophagus simple allometry -116 920 -125 082 0 511 Mass + Food -113 514 -119 100 0 498 Mass x Food -109 109 -112 824 0 497 extension of esophagus glands simple allometry 91 273 45 734 0 977 Mass + Food 48 695 40 746 0 952 Mass x Food 46 092 42 497 0 615 length of intermediate zone simple allometry 60 20076 20 7205 0 816 Mass + Food 21 142 11 659 0 803 Mass x Food 25 609 15 720 0 810 length of proventriculus simple allometry -72 200 -81 457 0 558 Mass + Food -81 509 -84 026 0 558 Mass x Food -76 821 -80 204 0 526 gizzard height simple allometry -45 0961 -78 692 0 766 Mass + Food -80 978 -84 985 0 602 Mass x Food -75 401 -79 171 0 602 gizzard width simple allometry -15 961 -54 317 0 774 Mass + Food -52 258 -58 246 0 652 Mass x Food -47 040 -52 899 0 661 gizzard depth simple allometry 40 475 -21 107 0 914 Mass + Food -27 413 -34 577 0 804 Mass x Food -25 417 -31 727 0 769 maximum gizzard height at main muscles (MHM) simple allometry 10 205 -28 377 0 763 Mass + Food -33 212 -36 807 0 589 Mass x Food -27 998 -31 641 0 593 gizzard thickness at main muscles (MMT) simple allometry 169 006 102 775 0 938 Mass + Food 103 458 89 283 0 898 Mass x Food 106 376 91 087 0 917 gizzard lumen width including koilin layer (LWiK) simple allometry 0 110 0 104 0 334 Mass + Food 5 594 4 28 0 395 Mass x Food 10 084 8 877 0 398 gizzard width at caudoventral thin muscle (WTM) simple allometry -37 786 -47 244 0 680 Mass + Food -33 593 -41 924 0 635 Mass x Food -32 005 -37 503 0 667 maximum gizzard height at thin muscle (MHT) simple allometry -37 340 -89 566 0 846 Mass + Food -82 851 -93 727 0 759 Mass x Food -76 966 -88 706 0 747 maximum gizzard lumen at thin muscle (MLT) simple allometry -30 160 -69 159 0 791 Mass + Food -62 319 -70 648 0 699 Mass x Food -56 805 -65 538 0 697 gizzard mass simple allometry 130 144 75 418 0 880 Mass + Food 75 552 66 876 0 787 Mass x Food 79 187 70 643 0 781

109 Evolutionary diversification and biogeography of parrots

Table 2. Regression parameters including p-values of the best-fitting model for the different traits of the digestive tract.

trait model value SE t-value p-value

length of intestine simple allometry intercept 4 251 0 251 16 957 0 000 slope 0 393 0 033 11 772 0 000 length of esophagus simple allometry intercept 2 649 0 083 31 743 0 000 slope 0 356 0 014 25 711 0 000 extension of esophagus glands Mass + Food intercept 0 002 0 313 0 006 0 995 slope 0 321 0 044 7 227 0 000 intercept 0 627 0 141 4 431 0 000 length of intermediate zone Mass + Food intercept 0 258 0 256 1 008 0 317 slope 0 456 0 037 12 482 0 000 intercept -0 221 0 115 -4 149 0 000 length of proventriculus simple allometry intercept 1 404 0 120 11 717 0 000 slope 0 389 0 019 20 414 0 000 gizzard height Mass + Food intercept 0 984 0 116 8 481 0 000 slope 0 322 0 018 17 775 0 000 intercept 1 199 0 053 4 000 0 000 gizzard width simple allometry intercept 1 155 0 158 7 302 0 000 slope 0 336 0 024 14 166 0 000 gizzard depth Mass + Food intercept 0 434 0 188 2 305 0 024 slope 0 303 0 027 11 301 0 000 intercept 0 864 0 085 5 071 0 000 maximum gizzard height at main muscles (MHM) Mass + Food intercept 1 115 0 159 7 027 0 000 slope 0 298 0 025 11 988 0 000 intercept 1 433 0 074 4 310 0 000 gizzard thickness at main muscles (MMT) Mass + Food intercept -0 236 0 473 -0 499 0 619 slope 0 182 0 063 2 871 0 005 intercept 0 713 0 211 4 501 0 000 gizzard lumen width including koilin layer (LWiK) simple allometry intercept 0 606 0 170 3 568 0 001 slope 0 354 0 030 11 831 0 000 gizzard width at caudoventral thin muscle (WTM) simple allometry intercept 0 447 0 155 2 885 0 005 slope 0 370 0 024 15 300 0 000 maximum gizzard height at thin muscle (MHT) Mass + Food intercept 1 195 0 122 9 776 0 000 slope 0 320 0 018 17 920 0 000 intercept 1 391 0 055 3 542 0 000 maximum gizzard lumen at thin muscle (MLT) simple allometry intercept 1 061 0 145 7 293 0 000 slope 0 348 0 022 16 081 0 000 gizzard mass Mass + Food intercept -4 505 0 365 -12 352 0 000 slope 0 870 0 053 16 559 0 000 intercept -3 854 0 165 3 947 0 000

110 Chapter 5

Figure 5. Natural-log transformed values of the different independent variables against natural-log transformed body masses including the regression lines of the best-fitting model. For all these traits, the data was best explained by an allometric relationship between the dependent and independent variables.

111 Evolutionary diversification and biogeography of parrots

Figure 6. Natural-log transformed values of the different independent variables against natural-log transformed body masses including the regression lines of the best-fitting models. For all these traits, a model including the nectarivorous diet as a covariate with different intercepts but the same slope was considered as the best-fitting model.

112 Chapter 5

Evolutionary rates One randomly chosen possible reconstruction of the evolution of nectarivory (green branches) generated with SIMMAP is shown (Fig. 7). The 100 stochastic character reconstructions were basically stable and the shifts to nectarivory usually occurred on the lineages leading to the different nectarivorous parrots (cf. Fig. 7). Only exceptionally, the shifts were reconstructed on more basal branches (e.g. on the lineage leading to the cluster consisting of Loriinae, Cyclopsittacini and Melopsittacus) and then included additional changes back to non-nectarivory in the lineages leading to the non-nectarivorous species. A single rate model with the same rate of morphological evolution for nectarivorous and non-nectarivorous species better fit the data than a two-rate model in all traits (Table 3). Hence, diet shifts to nectarivory did not generally lead to changes in the rate of trait evolution in nectarivorous parrots. In contrast, we found for ten traits a significant rate-shift in trait evolution at the base of the lory radiation (Table 4). These included all traits for which a model with diet as covariate was the best-fitting model in the PGLS ANCOVA’s. Additionally to these traits, a significant rate-shift in trait evolution was also found at the base of the lories for gizzard width and MLT.

Table 3. Model parameters from the one- and two-rate Brownian motion models. A one-rate model with the same rate of morphological evolution for nectarivorous and non-nectarivorous species better fit the data than a two-rate model in all traits.

one-rate model two-rate model

log BM rate log likelihood AICc Non-nectarivorous BM rate Nectarivorous BM rate likelihood AICc

length of intestine mean 5 22 -32 12 68 37 5 63 2 64 -31 03 68 32 sd 2 57 0 00 0 00 2 76 1 30 0 09 0 18 length of esophagus mean 3 49 -16 41 36 95 2 72 9 94 -13 98 34 21 sd 1 71 0 00 0 00 1 35 5 32 0 84 1 68 extension of esophagus glands mean 0 59 53 10 -102 07 0 60 0 49 53 21 -100 16 sd 0 29 0 00 0 00 0 30 0 24 0 04 0 07 length of intermediate zone mean 7 83 -47 95 100 02 8 10 6 20 -47 81 101 88 sd 3 85 0 00 0 00 4 00 2 97 0 08 0 16 length of proventriculus mean 4 67 -27 79 59 70 4 89 3 33 -27 48 61 21 sd 2 30 0 00 0 00 2 40 1 66 0 07 0 14 gizzard height mean 11 01 -61 23 126 59 10 96 11 33 -61 23 128 71 sd 5 41 0 00 0 00 5 37 5 71 0 01 0 01 gizzard width mean 17 99 -80 40 164 92 18 49 14 94 -80 29 166 83 sd 8 85 0 00 0 00 9 06 7 52 0 05 0 11 gizzard depth mean 8 40 -50 71 105 54 8 46 8 09 -50 70 107 65 sd 4 13 0 00 0 00 4 16 3 95 0 01 0 01 maximum gizzard height at main muscles (MHM) mean 31 35 -102 06 208 24 31 56 30 17 -102 05 210 35 sd 15 42 0 00 0 00 15 44 15 27 0 02 0 04 gizzard thickness at main muscles (MMT) mean 50 57 -120 70 245 52 50 79 49 22 -120 69 247 64 sd 24 86 0 00 0 00 24 93 24 47 0 01 0 02 gizzard lumen width including koilin layer (LWiK) mean 40 82 -112 35 228 83 41 08 39 32 -112 34 230 93 sd 20 07 0 00 0 00 20 12 19 79 0 01 0 03 gizzard width at caudoventral thin muscle (WTM) mean 5 86 -36 65 77 43 5 72 6 71 -36 59 79 42 sd 2 88 0 00 0 00 2 81 3 34 0 03 0 07 maximum gizzard height at thin muscle (MHT) mean 8 64 -51 80 107 72 8 44 9 88 -51 74 109 73 sd 4 25 0 00 0 00 4 14 4 96 0 03 0 05 maximum gizzard lumen at thin muscle (MLT) mean 13 33 -68 71 141 54 13 15 14 49 -68 68 143 62 sd 6 56 0 00 0 00 6 44 7 29 0 01 0 03 gizzard mass mean 49 02 -119 49 243 10 48 30 53 48 -119 46 245 17 sd 24 10 0 00 0 00 23 69 26 71 0 01 0 03

113 Evolutionary diversification and biogeography of parrots

Table 4. P-values and drift speed of the test for rate-shifts in morphological evolution at the base of the lory radiation. A significant rate-shift in trait evolution was found for ten traits.

Trait p-value Ā-value (drift speed)

length of intestine n s 0 9667 length of esophagus n s 0 5011 extension of esophagus glands 2 26E-04 1 2912 length of intermediate zone 1 30E-04 1 0891 length of proventriculus n s 0 6343 gizzard height 0 0058 0 6319 gizzard width 0 021 0 756 gizzard depth 2 99E-04 0 8304 maximum gizzard height at main muscles (MHM) 0 0254 0 9091 gizzard thickness at main muscles (MMT) 2 00E-06 1 7508 gizzard lumen width including koilin layer (LWiK) n s 1 2635 gizzard width at caudoventral thin muscle (WTM) n s 0 8016 maximum gizzard height at thin muscle (MHT) 0 0011 0 5496 maximum gizzard lumen at thin muscle (MLT) 0 0091 0 6638 gizzard mass 6 86E-05 1 5585

Discussion

Phenotype-environment correlations The gastrointestinal tract may be considered as one of the major interfaces between an organism and its environment and it mediates their interactions (Karasov 1990). Several interspecific studies on other bird groups have shown that the size of the gut is related to diet and that the morphology of the gastrointestinal tract often reflects the birds feeding strategies (Ricklefs 1996; Battley and Piersma 2005; Caviedes-Vidal et al. 2007; Lavin and Karasov 2008). Moreover, a direct influence of feeding strategies on structures, functionality and physiology of the digestive tract has been shown in other vertebrates like mammals (e.g. Schieck and Millar 1985; Korn 1992; Lovegrove 2010), amphibians and reptiles (e.g. Stevens and Hume 1995; O'Grady et al. 2005) as well as in fish (e.g. German and Horn 2006; Wagner et al. 2009). However, it has to be considered that biological structures are not only fine-tuned to their functional demands by natural selection, they are also influenced by phylogenetic history and biochemical and mechanical constraints (Raia et al. 2010). Within parrots, we showed that nectarivory was associated with reduced extension of the glands in the lower part of the esophagus below the crop (Pars thoracica) as compared to the non-nectarivorous parrots species. These glands produce a mucous secretion, which helps hard ingesta to glide through the glandular stomach (proventriculus) and reduces the risk of mechanical damage to the latter (Güntert 1981). It can be expected that parrot species eating exclusively soft or liquid food evolve reductions of these glands within the esophagus, and our data corroborate this. On the other hand, we could not find any indication that the nectarivorous parrots had a longer esophagus as was proposed for the lories by Güntert (1981). Between the glandular stomach (proventriculus) and the gizzard (muscular stomach, ventriculus), there is an intermediate zone characterized by the absence of compound glands of the former and the koilin layer of the latter (Ziswiler and Farner 1972). This intermediate zone has the function of a storage space, where the proteolytic enzyme pepsin from the proventriculus can react with ingesta (Güntert 1981). We found the nectarivorous parrots to have a longer intermediate zone compared to the remaining

114 Chapter 5 parrots, even though the trait-values of one nectarivorous taxon,Brotogeris , were more similar to the non-nectarivorous species. The intermediate zone seems to play an important role in the digestion of pollen (Güntert 1981). Pollen grains have a high content of proteins with their interior (protoplast) consisting of diverse amino acids (van Tets and Hulbert 1999; Gartrell and Jones 2001). Acidifications of pollen grains in the proventriculus may be important so that their contents can be extruded and digested, whereas mechanical break-up of pollen grains in the gizzard does not seem to be important (Gartrell and Jones 2001). The amount of energy extracted from meals can be enhanced by increasing the retention time (McWhorter et al. 2009). A prolongation of the intermediate zone can thus increase the rate of protein digestion as it increases the retention time of pollen grains, which seems to be the case in nectarivorous parrots. Pollen ingestion may require less energy than feeding on insect as an additional amino acid supply, because nectarivorous birds will encounter pollen while feeding on nectar (Nicolson and Fleming 2003). The gizzard functions as an organ of mechanical digestion beside being a storage organ, the site of preliminary acid proteolytic digestion and a filter for indigestible material (Ziswiler and Farner 1972). During contraction of the gizzard, the thick muscles close up, narrowing the lumen to a thin cleft and forcing the contents into two pouches (cranial and caudal sac, cf. Fig. 2) that lie under the thin muscles (McLelland 1979). Species feeding on soft diet do not need the grinding function to break down their food, and can be expected to evolve reduced gizzard musculature (Steinbacher 1934; McLelland 1979). Indeed, the nectarivorous parrots differed from the remaining parrots by having less well developed gizzard muscles. The variation found among different species within the lories for some traits may be explained by the fact that several species of this group can feed at least complementary on seeds (Homberger 1980). In contrast, a simple allometric relationship explained best the width of the whole gizzard and its lumen as well as its width at the caudoventral thin muscle and the maximum lumen at the thin muscles. The thin muscles act antagonistically to the main muscles and have no grinding function. Therefore, they are not expected to be developed more strongly in species relying on the grinding function of the gizzard. In congruence with our results, Richardson & Wooller (1990) also found two species of lories (Glossopsitta porphyrocephala, Trichoglossus haematodus) to have smaller and less muscular gizzards than four non-nectarivorous parrot species of Australia (Melopsittacus undulatus, Barnardius zonarius, Neopsephotus bourkii, Platycercus icterotis). Interestingly, we found that the reportedly mainly frugivorous Pesquet’s parrot Psittrichas fulgidus (Collar 1998) shared a similar reduced muscularity with the nectarivorous parrots. On the other hand, the blue-crowned hanging-parrot Loriculus galgulus did not show an overall reduced muscularity. This may correspond to the higher amount of seeds in its diet compared to L. philippensis (Homberger 1980) and indicates a correlation between sister species in food and trait values. Similarly, gizzard measurements of Lathamus discolor and Brotogeris jugularis clustered with the non-nectarivorous parrots and these two species were not found to have an overall reduced muscularity of this organ either. Other studies also found Lathamus discolor to have retained the muscular gizzard of a granivorous species (Güntert and Ziswiler 1972; Gartrell et al. 2000). This may allow this species to feed on harder food when nectar and pollen are rare (Gartrell et al. 2000). Smaller gizzards with a reduced muscularity were also found in the nectarivorous honeyeaters (Meliphagidae) compared to similar-sized passerines (Richardson and Wooller 1986), however, phylogenetic non-independence was not controlled for in this study. Gizzards dimensions also vary in other passerines according to diet with longer gizzards in

115 Evolutionary diversification and biogeography of parrots seed- than in fruit- and insect-eaters and thicker muscular and glandular layers in insect- compared to fruit- and seed-eaters (Ricklefs 1996). Chemical digestion of food principally takes place in the intestine (Ziswiler & Farner 1972). Richardson & Wooller (1990) found two species of loriesGlossopsitta ( porphyrocephala, Trichoglossus haematodus) to have shorter intestines compared with four non-nectarivorous parrot species of Australia (Melopsittacus undulatus, Barnardius zonarius, Neopsephotus bourkii, Platycercus icterotis). This was explained as a consequence of sugars in nectar needing less processing in the intestine than other food. In contrast, we could not find any indication for shorter intestines of nectarivorous parrots. In a wide comparative study of birds, Lavin et al. (2008) did not find any significant effect of diet on small intestine length either (Lavin et al. 2008). However, this results is only partly comparable with ours, because we measured the whole lengths of the intestine due to the difficulty of clearly distinguishing the small and the large intestine in parrots due to lack of caeca (Ziswiler and Farner 1972; Güntert 1981). In addition to lengths, intestine function depends among others on volume, surface area, villi and microvilli area as well as enzymatic activity (Ricklefs 1996; Lavin et al. 2008; McWhorter et al. 2009). Moreover, the efficiency of digestion in the intestine may be influenced by the passive absorption of hydrosoluble compounds through the paracellular pathway. This is prominent in birds and may especially be important for nectarivores because they have to deal with large amounts of sugar in their diet (Karasov and Cork 1994; Napier et al. 2008; McWhorter et al. 2009). In general, birds have a lower nominal surface area of the intestine and a shorter small intestine as well as shorter digestive retention times than mammals, however, their higher passive absorption compared to mammals may compensate for this (McWhorter et al. 2009). This may render predictions about the intestine dimensions in relation to diet more difficult. In conclusion, our analyses showed that the nectarivorous parrot species differed, after correction for phylogenetic non-independence, from the remaining parrots in several traits of the digestive tract. The similarity of some trait features among the different nectarivorous groups is an indication for parallel or convergent evolution under the same or similar environmental conditions, i.e. the change to a nectarivorous diet, and implies that natural selection was the main driving force (cf. Losos et al. 1998; Schluter et al. 2004; Colosimo et al. 2005). Hence we uncovered significant phenotype-environment correlations. Functional considerations suggest that some of these differences may allow nectarivorous parrots to effectively feed on nectar implying evidence for trait utility here (Schluter 2000). In contrast, some of the differences between nectarivorous and the remaining parrots reported by previous authors could not be corroborated with our data.

Phenotypic flexibility of the gastrointestinal tract The size, structure and functional characteristics of the gastrointestinal tract of birds can reversibly change (phenotypic flexibility, sensu Piersma & Drent (2003)) as a fast adaptive response to current functionality demands caused by environmental changes or circannual endogenous control (Starck 1999b, a; Piersma and Drent 2003; Starck and Rahmaan 2003; Battley and Piersma 2005; McWhorter et al. 2009). Such phenotypic flexibility in combination with plasticity (irreversible environmentally induced variation sensu Piersma & Drent (2003)) could create an opportunity for selection during dietary differentiation and potentially facilitate diversification (Wagner et al. 2009). As pointed out by Lavin et al. (2008), comparative studies as ours have the limitation that species were not analyzed under common-garden

116 Chapter 5 conditions and thus it is not possible to asses to what extent the variation found among species is influenced by phenotypic flexibility and plasticity at the individual level. However, the inclusion of several individuals for most species and the wide range of body sizes among species considered certainly minimized the effect of intraspecific variation. Further, the individuals analyzed in this study all stem from captivity where more stable conditions than in nature can be expected mirroring a common-garden experiment. There is additionally some evidence that phenotypic flexibility of the gastrointestinal tract is limited in parrots (cf. Güntert 1981). All specimens of Lathamus analyzed in this study were fed with a nectar-alternative, but they retained the partly muscular gizzard similar of a granivorous species, and their features of the gastrointestinal tract seemingly did not differ from wild specimens analyzed by Gartrell et al. (2000). Nevertheless, further studies preferably on wild specimens have to document the interplay between natural selection, plasticity and potential flexibility in features of the digestive tract in different parrot species.

Evolutionary rates and species proliferation Natural selection is most likely responsible for the parallel evolution of similar features of the digestive tract in the distantly related groups of nectarivorous parrots. We found that the diet shift of the lories to nectarivory was associated with a significant rate-shift in morphological evolution at the base of their radiation for several characters chiefly of the gizzard. The new statistical approach applied here allowed us to explicitly infer based on a traditional hypothesis-testing approach one of the relevant questions in our study, namely if a subset of species in a phylogenetic tree shows a change in the rate of trait evolution. When the other nectarivorous groups were added to the analyses and the overall rate of trait evolution considered, we could not find any indication for different rates in nectarivorous compared to non- nectarivorous parrots. Although rapid adaptations may have thus been prevalent at the beginning of the ecological expansion into a nectarivorous lifestyle at least in the lories, the diet shift did not lead to an overall increased disparity and morphological specialization in traits of the digestive tract within all nectarivorous parrots. The lories have diversified into an exceptionally species-rich clade (Schweizer et al. in press) and their diet shift can thus be considered as an evolutionary key innovation. Nectar may have provided a spatially widespread underutilized niche and this may have allowed the lories to expand their ranges and to colonize even remote oceanic islands, which may have fostered allopatric speciation. Even today, congeneric species of the lories do generally not overlap geographically (Collar 1998). Sympatry within genera is found apart from eastern Australia around New Guinea and Wallacea, regions with a complex and composite geological history with several potential vicariance opportunities in the past (Hall 2002; Esselstyn et al. 2009; Deiner et al. 2011). However, this ecological expansion was not followed by further significant ecological specializations within the radiation of the lories. Similar to honeyeaters, the other highly nectarivorous and species-rich bird-group of Australasia (Newton 2003), lories are generalized flower visitors and their ecological relationships with plants are not as specialized as they are in hummingbirds or sunbirds (Fleming and Muchhala 2008). Avian pollinator assemblages differ regionally and the evolutionary specialization between nectar-feeding birds and their food-plants is highest in the Neotropics and decreases through Africa and South Asia to Southeast Asia and Australasia (Fleming and Muchhala 2008). The evolution of specialized plant- pollinator relationship takes time. While the hummingbirds split from their closest relatives in the Eocene or even earlier (Brown et al. 2008; Pratt et al. 2009), the Lories

117 Evolutionary diversification and biogeography of parrots split from Melopsittacus only in the middle Miocene (Schweizer et al. in press). The hummingbirds had certainly more time to ecologically specialize than the lories. However, while the sunbirds are likely to be younger than the hummingbirds, the evolutionary diversification of the honeyeaters started at a similar age (Eocene) (Barker et al. 2004). Thus, other explanations than time may account for the low ecological specialization of Australasian nectarivorous birds compared to sunbirds and hummingbirds. Interactions between plants and pollinators are only likely to evolve when floral resources are spatially and temporally predictable (e.g. Waser et al. 1996). While this seems to be the case in the Neotropics, flowering of eucalypts in Australia varies in space and time, and trees in lowland and montane Papua New Guinea have commonly non-annual flowering patterns (Fleming and Muchhala 2008). This may account for the low specialization in plant-pollinator relationships of Australasian nectarivorous birds and may explain the lack of ecological specialization of the lories after their shift to a nectarivorous diet. In conclusion, the key innovation of the lories allowed an expansion into a new adaptive zone and we hypothesize that the subsequent species proliferation may have essentially been non-adaptive through allopatric speciation. The lories may thus be considered as an example of a nonadaptive radiation (Rundell and Price 2009). The ecological opportunity provided by their key innovation seems not to have been sufficient to trigger an adaptive radiation due to the unpredictable nature of the new resource. The key innovation nevertheless promoted significant lineage diversification through allopatric partitioning of the same new niche. Although other parrot groups changed to a nectarivorous diet, this did not increase their diversification rates and species richness compared to other parrots (Schweizer et al. in press). Various factors may have inhibited an increased rate of cladogenesis in Lathamus, Brotogeris and Loriculus following their change to a nectarivorous diet. Such factors can include developmental or genetic constraints, but also ecological circumstances like interspecific competition or the lack of opportunities for allopatric speciation. Hence, an evolutionary innovation does not necessarily lead to increased diversification (Vermeij 2001; Price et al. 2010). However, it is unclear which factors hampered increased species proliferation in other nectarivorous birds than the lories.

Acknowledgements We especially thank the Silva Casa Foundation for financial support of this work. We are grateful to S. Birks (University of Washington, Burke Museum), R. Burkhard, Federal Veterinary Office Bern (FVO), A. Fergenbauer-Kimmel, J. Fjeldså and J.-B. Kristensen (Zoological Museum, University of Copenhagen), H. Gygax, H. Rosenberger, D. Ruess, P. Sandmeier, T. and P. Walser, D. Willard (Field Museum of Natural History) and G. Weis for kindly providing us with specimens, tissue or feather samples. We further thank the following people for valuable support: S. Bachofner, B. Blöchlinger, M. Hohn, B. Kurz, L. Lepperhof, S. A. Price, L. J. Revell, M. Rieger, T. Roth, A. Stamatakis and C. Sherry.

118 Chapter 5

Figure 7. Time-calibrated phylogeny of the parrots corresponding to the maximum clade credibility tree as estimated with Beast. One possible reconstruction of nectarivory (green branches) generated with stochastic character mapping using SIMMAP 1.0 is shown (Ma: millions of years ago).

119 Evolutionary diversification and biogeography of parrots

Literatur cited

Akaike, H. 1974. New look at statistical-model identification. Ieee Transactions on Automatic Control AC19:716-723. Akester, A. R. 1986. Structure of the glandular layer and koilin membrane in the gizzard of the adult domestic fowl (Gallus gallus domesticus). J. Anat. 147:1- 25. Barker, F. K., A. Cibois, P. Schikler, J. Feinstein, and J. Cracraft. 2004. Phylogeny and diversification of the largest avian radiation. Proceedings of the National Academy of Sciences of the United States of America 101:11040-11045. Battley, P. F., and T. Piersma. 2005. Adaptive interplay between feeding ecology and features of the digestive tract in birds in J. M. Starck, and T. Wang, eds. Physiological and ecological adaptations to feeding in vertebrates. Science Publishers Inc., Enfield. Bawa, K. S. 1990. Plant-pollinator interactions in tropical rain-forests. Annual Review of Ecology and Systematics 21:399-422. Blomberg, S. P., T. Garland, and A. R. Ives. 2003. Testing for phylogenetic signal in comparative data: behavioral traits are more labile. Evolution 57:717-745. Bollback, J. P. 2006. SIMMAP: stochastic character mapping of discrete traits on phylogenies. BMC Bioinformatics 7:7. Brice, A. T. 1992. The essentiality of nectar and arthropods in the diet of the Anna hummingbird (Calypte anna). Comp. Biochem. Physiol. A-Physiol. 101:151- 155. Brown, J. H., W. A. Calder, and A. Kodricbrown. 1978. Correlates and consequences of body size in nectar feeding birds. American Zoologist 18:687-700. Brown, J. W., J. S. Rest, J. Garcia-Moreno, M. D. Sorenson, and D. P. Mindell. 2008. Strong mitochondrial DNA support for a Cretaceous origin of modern avian lineages. BMC Biology 6. Burnham, K. P., and D. R. Anderson. 2002. Model selection and mulitimodel inference. Second edition. Springer-Verlag, New York, USA. Casotti, G., C. A. Beuchat, and E. J. Braun. 1998. Morphology of the kidney in a nectarivorous bird, the Anna's hummingbird Calypte anna. Journal of Zoology 244:175-184. Casotti, G., and K. C. Richardson. 1993. A qualitative analysis of the kidney structure of meliphagid honeyeaters from wet and arid environments. J. Anat. 182:239- 247. Caviedes-Vidal, E., T. J. McWhorter, S. R. Lavin, J. G. Chediack, C. R. Tracy, and W. H. Karasov. 2007. The digestive adaptation of flying vertebrates: high intestinal paracellular absorption compensates for smaller guts. Proceedings of the National Academy of Sciences of the United States of America 104:19132- 19137. Churchill, D. M., and P. Christensen. 1970. Observations on pollen harvesting by brush-tongued lorikeets. Aust J Zool 18:427-437. Collar, N. J. 1998. Family Psittacidae (parrots). Pp. 280-477 in J. del Hoyo, A. Elliott, and J. Sargatal, eds. Handbook of the birds of the world, vol. 4. Lynx Edicions, Barcelona. Colosimo, P. F., K. E. Hosemann, S. Balabhadra, G. Villarreal, M. Dickson, J. Grimwood, J. Schmutz, R. M. Myers, D. Schluter, and D. M. Kingsley. 2005. Widespread parallel evolution in sticklebacks by repeated fixation of ectodysplasin alleles. Science 307:1928-1933.

120 Chapter 5

Deiner, K., A. R. Lemmon, A. L. Mack, R. C. Fleischer, and J. P. Dumbacher. 2011. A passerine bird's evolution corroborates the geologic history of the island of New Guinea. PLoS One 6:15. Downs, C. T. 2004. Some preliminary results of studies on the bill and tongue morphology of Gurney's sugarbird and some southern African sunbirds. Ostrich 75:169-175. Drummond, A. J., and A. Rambaut. 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary Biology 7. Esselstyn, J. A., R. M. Timm, and R. M. Brown. 2009. Do geological or climatic processes drive speciation in dynamic archipelagos? The tempo and mode of diversification in Southeast Asian shrews. Evolution 63:2595-2610. Feinsinger, P., and R. K. Colwell. 1978. Community organization among Neotropical nectar-feeding birds. American Zoologist 18:779-795. Felsenstein, J. 1985. Phylogenies and the comparative method. American Naturalist 125:1-15. Fleming, T. H., and N. Muchhala. 2008. Nectar-feeding bird and bat niches in two worlds: pantropical comparisons of vertebrate pollination systems. Journal of Biogeography 35:764-780. Forshaw, J. M. 1989. Parrots of the world. Landsdowne Press, Melbourne. Freckleton, R. P. 2009. The seven deadly sins of comparative analysis. J. Evol. Biol. 22:1367-1375. Freckleton, R. P., P. H. Harvey, and M. Pagel. 2002. Phylogenetic analysis and comparative data: a test and review of evidence. American Naturalist 160:712- 726. Garland, T., and A. R. Ives. 2000. Using the past to predict the present: confidence intervals for regression equations in phylogenetic comparative methods. American Naturalist 155:346-364. Gartrell, B. D. 2000. The nutritional, morphologic, and physiologic bases of nectarivory in Australian birds. Journal of Avian Medicine and Surgery 14:85- 94. Gartrell, B. D., and S. M. Jones. 2001. Eucalyptus pollen grain emptying by two Australian nectarivorous psittacines. Journal of Avian Biology 32:224-230. Gartrell, B. D., S. M. Jones, R. N. Brereton, and L. B. Astheimer. 2000. Morphological adaptations to nectarivory of the alimentary tract of the swift parrot Lathamus discolor. Emu 100:274-279. German, D. P., and M. H. Horn. 2006. Gut length and mass in herbivorous and carnivorous prickleback fishes (Teleostei: Stichaeidae): ontogenetic, dietary, and phylogenetic effects. Mar. Biol. 148:1123-1134. Grafen, A. 1989. The phylogenetic regression. Philos. Trans. R. Soc. Lond. Ser. B- Biol. Sci. 326:119-157. Grafen, A. 1992. The uniqueness of the phylogenetic regression. J. Theor. Biol. 156:405-423. Güntert, M. 1981. Morphologische Untersuchungen zur adaptiven Radiation des Verdauungstraktes bei Papageien (Psittaci). Zool. Jahrb. Abt. Anat. Ontog. Tiere 106:471-526. Güntert, M., and V. Ziswiler. 1972. Konvergenzen in der Struktur von Zunge und Verdauungstrakt nektarfressender Papageien. Rev. Suisse Zool. 79:1017-1026. Hall, R. 2002. Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based reconstructions, model and animations. J. Asian Earth Sci. 20:353-431.

121 Evolutionary diversification and biogeography of parrots

Hall, T. A. 1999. BioEdit, a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41:95-98. Homberger, D. G. 1980. Funktionell- morphologische Untersuchungen zur Radiation der Ernährungs- und Trinkmethoden der Papageien (Psittaci). Bonn. Zool. Monogr. 13:1-92. Housworth, E. A., E. P. Martins, and M. Lynch. 2004. The phylogenetic mixed model. American Naturalist 163:84-96. Huelsenbeck, J. P., R. Nielsen, and J. P. Bollback. 2003. Stochastic mapping of morphological characters. Systematic Biology 52:131-158. Hunter, J. P. 1998. Key innovations and the ecology of macroevolution. Trends in Ecology & Evolution 13:31-36. Joseph, L., A. Toon, E. E. Schirtzinger, and T. F. Wright. 2011. Molecular systematics of two enigmatic genera Psittacella and Pezoporus illuminate the ecological radiation of Australo-Papuan parrots (Aves: Psittaciformes). Molecular Phylogenetics and Evolution 59:675-684. Karasov, W. H. 1990. Digestion in birds: chemical and physiological determinants and ecological implications. Stud Avian Biol 391-415. Karasov, W. H., and S. J. Cork. 1994. Glucose-absorption by a nectarivorous bird - the passive pathway is paramount. Am. J. Physiol. 267:G18-G26. Kassen, R. 2009. Toward a general theory of adaptive radiation: insights from microbial experimental evolution. Pp. 3-22. Year in Evolutionary Biology 2009. Blackwell Publishing, Oxford. Korn, H. 1992. Intestine lengths of Southern Africa savanna rodents and insectivores - intraspecific and interspecific comparisons. Journal of Zoology 228:455-460. Lavin, S. R., and W. H. Karasov. 2008. Allometry of paracellular absorption in birds. Physiol. Biochem. Zool. 81:551-560. Lavin, S. R., W. H. Karasov, A. R. Ives, K. M. Middleton, and T. Garland. 2008. Morphometrics of the avian small intestine compared with that of nonflying mammals: a phylogenetic approach. Physiol. Biochem. Zool. 81:526-550. Losos, J. B., T. R. Jackman, A. Larson, K. de Queiroz, and L. Rodriguez-Schettino. 1998. Contingency and determinism in replicated adaptive radiations of island lizards. Science 279:2115-2118. Lovegrove, B. G. 2010. The allometry of rodent intestines. J. Comp. Physiol. B- Biochem. Syst. Environ. Physiol. 180:741-755. Lüttge, U. 1976. Chemical composition of nectars. Pp. 63-95 in U. Lüttge, and M. G. Pitman, eds. Transport in plants II. Tissue and organs. Part B. Sringer-Verlag, Berlin, Germany. Martínez del Rio, C. 1994. Nutritional ecology of fruit-eating and flower-visiting birds and bats. Pp. 103-127 in D. Chivers, and P. Langer, eds. The digestive system in mammals: food form and function. Cambridge University Press, Cambridge. Martins, E. P., and T. F. Hansen. 1997. Phylogenies and the comparative method: a general approach to incorporating phylogenetic information into the analysis of interspecific data. American Naturalist 149:646-667. Mayr, E. 1931. Birds collected during the Whitney South Sea expedition. XIII. Amer. Mus. Novit. No 486: 1-29. McLelland, J. 1979. Digestive system. Pp. 70-181 in A. S. King, and J. McLelland, eds. Form and function in birds. Academic Press, London.

122 Chapter 5

McWhorter, T. J., E. Caviedes-Vidal, and W. H. Karasov. 2009. The integration of digestion and osmoregulation in the avian gut. Biol. Rev. 84:533-565. Napier, K. R., T. J. McWhorter, and P. A. Fleming. 2008. Mechanism and rate of glucose absorption differ between an Australian honeyeater (Meliphagidae) and a lorikeet (Loriidae). J. Exp. Biol. 211:3544-3553. Newton, I. 2003. The speciation and biogeography of birds. Academic Press, London. Nicolson, S. W., and P. A. Fleming. 2003. Nectar as food for birds: the physiological consequences of drinking dilute sugar solutions. Plant Syst. Evol. 238:139- 153. O'Grady, S. P., M. Morando, L. Avila, and M. D. Dearing. 2005. Correlating diet and digestive tract specialization: examples from the lizard family Liolaemidae. Zoology 108:201-210. O'Meara, B. C., C. Ane, M. J. Sanderson, and P. C. Wainwright. 2006. Testing for different rates of continuous trait evolution using likelihood. Evolution 60:922-933. Pacheco, M. A., F. U. Battistuzzi, M. Lentino, R. Aguilar, S. Kumar, and A. A. Escalante. 2011. Evolution of modern birds revealed by mitogenomics: timing the radiation and origin of major orders. Molecular Biology and Evolution 28:1927-1942. Pagel, M. 1997. Inferring evolutionary processes from phylogenies. Zoologica Scripta 26:331-348. Pagel, M. 1999. Inferring the historical patterns of biological evolution. Nature 401:877-884. Paradis, E., J. Claude, and K. Strimmer. 2004. APE: analyses of phylogenetics and evolution in R language. Bioinformatics 20:289-290. Piersma, T., and J. Drent. 2003. Phenotypic flexibility and the evolution of organismal design. Trends in Ecology & Evolution 18:228-233. Pinheiro, J., D. Bates, S. DebRoy, D. Sarkar, and the R Development Core Team. 2009. nlme: linear and nonlinear mixed effects models. Pratt, R. C., G. C. Gibb, M. Morgan-Richards, M. J. Phillips, M. D. Hendy, and D. Penny. 2009. Toward resolving deep Neoaves phylogeny: data, signal enhancement, and priors. Molecular Biology and Evolution 26:313-326. Price, S. A., R. Holzman, T. J. Near, and P. C. Wainwright. 2011. Coral reefs promote the evolution of morphological diversity and ecological novelty in labrid fishes. Ecol. Lett. 14:462-469. Price, S. A., P. C. Wainwright, D. R. Bellwood, E. Kazancioglu, D. C. Collar, and T. J. Near. 2010. Functional innovations and morphological diversification in parrotfish. Evolution 64:3057-3068. Raia, P., F. Carotenuto, C. Meloro, P. Piras, and D. Pushkina. 2010. The shape of contention: adaptation, history, and contingency in ungulate mandibles. Evolution 64:1489-1503. Rambaut, A. 2008. FigTree 1.2., published by the author. Rambaut, A., and A. J. Drummond. 2007. Tracer v1.5, available from http://beast.bio.ed.ac.uk/Tracer. Rand, A. L., and D. S. Rabor. 1960. Birds of the Philippine Islands: Siquijor, Mount Malindang, Bohol, and Samar. Fieldiana: Zool. 35:221-441. Revell, L. J. 2009. Size-correction and principal components for interspecific comparative studies. Evolution 63:3258-3268.

123 Evolutionary diversification and biogeography of parrots

Richardson, K. C., and R. D. Wooller. 1986. The structures of the gastrointestinal tracts of honeyeaters and other small birds in relation to their diets. Aust. J. Zool. 34:119-124. Richardson, K. C., and R. D. Wooller. 1990. Adaptation of the alimentary tract of some Australian lorikeets to a diet of pollen and nectar. Aust. J. Zool. 38:581- 586. Ricklefs, R. E. 1996. Morphometry of the digestive tracts of some passerine birds. Condor 98:279-292. Rowley, I. 1998. Family Cacatuidae (cockatoos). Pp. 246-279in J. Del Hoyo, A. Elliott, and J. Sargatal, eds. Handbook of the birds of the world. Lynx Edicions, Barcelona. Rundell, R. J., and T. D. Price. 2009. Adaptive radiation, nonadaptive radiation, ecological speciation and nonecological speciation. Trends in Ecology & Evolution 24:394-399. Russello, M. A., and G. Amato. 2004. A molecular phylogeny of Amazona: implications for Neotropical parrot biogeography, taxonomy, and conservation. Molecular Phylogenetics and Evolution 30:421-437. Schieck, J. O., and J. S. Millar. 1985. Alimentary-tract measurements as indicators of diets of small mammals. Mammalia 49:93-104. Schlamowitz, R., F. R. Hainsworth, and L. L. Wolf. 1976. Tongues of sunbirds. Condor 78:104-107. Schluter, D. 2000. The ecology of adaptive radiation. Oxford University Press, Oxford. Schluter, D., E. A. Clifford, M. Nemethy, and J. S. McKinnon. 2004. Parallel evolution and inheritance of quantitative traits. American Naturalist 163:809- 822. Schmidt-Nielsen, K. 1984. Scaling: why is animal size so important? Cambridge University Press, Cambridge. Schuchmann, K. L. 1999. Family Trochilidae (hummingbirds) in J. del Hoyo, A. Elliott, and J. Sargatal, eds. Handbook of the birds of the world. Lynx Edicions, Barcelona. Schweizer, M., O. Seehausen, M. Güntert, and S. T. Hertwig. 2010. The evolutionary diversification of parrots supports a taxon pulse model with multiple trans- oceanic dispersal events and local radiations. Molecular Phylogenetics and Evolution 54:984-994. Schweizer, M., O. Seehausen, and S. T. Hertwig. in press. Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations. Journal of Biogeography. Sekercioglu, C. H. 2006. Increasing awareness of avian ecological function. Trends in Ecology & Evolution 21:464-471. Stamatakis, A. 2006. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22:2688- 2690. Stamatakis, A., P. Hoover, and J. Rougemont. 2008. A rapid bootstrap algorithm for the RAxML web servers. Systematic Biology 57:758-771. Starck, J. M. 1999a. Phenotypic flexibility of the avian gizzard: rapid, reversible and repeated chances of organ size in response to changes in dietary fibre content. J. Exp. Biol. 202:3171-3179. Starck, J. M. 1999b. Structural flexibility of the gastro-intestinal tract of vertebrates - implications for evolutionary morphology. Zoologischer Anzeiger 238:87-101.

124 Chapter 5

Starck, J. M., and G. H. A. Rahmaan. 2003. Phenotypic flexibility of structure and function of the digestive system of Japanese quail. J. Exp. Biol. 206:1887- 1897. Steinbacher, G. 1934. Zur Kenntnis des Magens blütenbesuchender Pagageien. Orn. Monatsber. 42. Stevens, C. E., and I. D. Hume. 1995. Comparative physiology of the vertebrate digestive system. Cambridge University Press, Cambridge. Tavares, E. S., A. J. Baker, S. L. Pereira, and C. Y. Miyaki. 2006. Phylogenetic relationships and historical biogeography of Neotropical parrots (Psittaciformes: Psittacidae: Arini) inferred from mitochondrial and nuclear DNA sequences. Systematic Biology 55:454-470. Temeles, E. J., and W. J. Kress. 2003. Adaptation in a plant-hummingbird association. Science 300:630-633. van Tets, I. G., and A. J. Hulbert. 1999. A comparison of the nitrogen requirements of the eastern pygmy possum, Cercartetus nanus, on a pollen and on a mealworm diet. Physiol. Biochem. Zool. 72:127-137. van Tets, I. G., and S. W. Nicolson. 2000. Pollen and the nitrogen requirements of the lesser double-collared sunbird. Auk 117:826-830. Vermeij, G. J. 1995. Economics, volcanos, and phanerozoic revolutions. Paleobiology 21:125-152. Vermeij, G. J. 2001. Innovation and evolution at the edge: origins and fates of gastropods with a labral tooth. Biological Journal of the Linnean Society 72:461-508. Wagner, C. E., P. B. McIntyre, K. S. Buels, D. M. Gilbert, and E. Michel. 2009. Diet predicts intestine length in Lake Tanganyika's cichlid fishes. Funct. Ecol. 23:1122-1131. Waser, N. M., L. Chittka, M. V. Price, N. M. Williams, and J. Ollerton. 1996. Generalization in pollination systems, and why it matters. Ecology 77:1043- 1060. White, N. E., M. J. Phillips, M. T. P. Gilbert, A. Alfaro-Nunez, E. Willerslev, P. R. Mawson, P. B. S. Spencer, and M. Bunce. 2011. The evolutionary history of cockatoos (Aves: Psittaciformes: Cacatuidae). Molecular Phylogenetics and Evolution 59:615-622. Wright, T. F., E. E. Schirtzinger, T. Matsumoto, J. R. Eberhard, G. R. Graves, J. J. Sanchez, S. Capelli, H. Mueller, J. Scharpegge, G. K. Chambers, and R. C. Fleischer. 2008. A multilocus molecular phylogeny of the parrots (Psittaciformes): support for a Gondwanan origin during the Cretaceous. Molecular Biology and Evolution 25:2141-2156. Yoder, J. B., E. Clancey, S. Des Roches, J. M. Eastman, L. Gentry, W. Godsoe, T. J. Hagey, D. Jochimsen, B. P. Oswald, J. Robertson, B. A. J. Sarver, J. J. Schenk, S. F. Spear, and L. J. Harmon. 2010. Ecological opportunity and the origin of adaptive radiations. J. Evol. Biol. 23:1581-1596. Ziswiler, V. 1967. Vergleichende morphologische Untersuchungen am Verdauungstrakt körnerfressender Singvögel zur Abklärung ihrer systematischen Stellung. Zool. Jb. Syst. 94:427-520. Ziswiler, V., and D. S. Farner. 1972. Digestion and the digestive system. Pp. 343-430 in D. S. Farner, and J. R. King, eds. Avian biology. Academic Press, New York.

125 Evolutionary diversification and biogeography of parrots

126 Chapter 5

Appendix 1. Number of individuals analyzed for each morphological trait.

gizzard maximum lumen gizzard gizzard width gizzard maximum maximum extension height at thickness including width at gizzard gizzard of length of length of main at main koilin caudoventral height at lumen at length of length of esophagus intermediate pro- gizzard gizzard gizzard muscles muscles layer thin muscle thin muscle thin muscle gizzard Species intestine esophagus glands zone ventriculus height width depth (MHM) (MMT) (LWiK) (WTM) (MHT) (MLT) mass

Agapornis canus 4 3 3 3 3 4 4 4 4 4 4 4 4 4 4 Agapornis fischeri 11 10 8 8 8 9 9 9 9 9 9 9 9 9 9 Agapornis lilianae 6 7 7 7 7 7 7 7 7 6 6 6 7 7 7 Agapornis nigrigenis 2 2 1 2 2 2 2 2 2 2 2 2 2 2 1 Agapornis roseicollis 5 5 3 4 4 4 4 4 4 4 4 4 4 4 4 Agapornis personatus 4 4 3 4 4 4 4 4 4 4 4 2 4 4 4 Alisterus chloropterus 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 Alisterus scapularis 6 6 6 5 5 6 6 6 6 6 6 6 6 6 5 Amazona aestiva 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Amazona dufresniana 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Amazona pretrei 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Amazona xanthops 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Anodohynchus hyacinthinus 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Aprosmictus jonquillaceus 4 5 4 3 3 4 4 4 4 4 4 4 4 4 4 Ara ararauna 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Aratinga leucophthalmus 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Ara macao 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Aratinga solstitialis 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Barnardius zonarius 6 6 6 6 6 5 5 5 5 5 5 5 5 5 5 Barnardius barnardi 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Brotogeris jugularis 5 5 5 5 5 5 5 5 5 5 5 4 5 5 4 Cacatua goffini 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Cacatua moluccensis 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Cacatua sulphurea 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Chalcopsitta cardinalis 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Charmosyna pulchella 7 7 6 7 7 6 6 6 6 6 6 6 6 6 6 Coracopsis vasa 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Cyanoramphus auriceps 5 4 6 6 6 6 6 6 6 6 6 6 6 6 5 Cyanoramphus novaezelandiae 3 2 2 2 2 3 3 3 3 3 3 3 3 3 3 Cyclopsitta diophthalma 4 5 4 4 4 4 4 4 4 4 4 4 4 4 4 Deroptyus accipitrinus 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Eclectus roratus 6 6 6 5 5 6 6 6 6 6 6 6 6 6 6 Eos cyanogenia 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Eunymphicus (cornutus) cornutus 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Eunymphicus (cornutus) uvaeensis 5 6 6 6 6 6 6 6 6 6 6 6 6 6 6 Forpus passerinus 11 11 10 10 11 11 11 11 11 11 11 11 11 11 10 Guaruba guarouba 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Lathamus discolor 5 5 3 5 5 3 3 3 3 3 3 3 3 3 2 Loriculus galgulus 3 3 3 2 2 3 3 3 2 2 3 3 3 2 2 Loriculus philippensis 4 4 4 3 4 1 1 1 1 1 1 2 1 1 2 Lorius garrulus 5 5 5 4 4 4 4 4 4 4 4 4 4 4 4 Melopsittacus undulatus 6 6 6 4 4 7 7 7 7 7 7 6 7 7 6 Micropsitta finschii 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Neophema chrysostoma 4 3 3 2 2 3 3 3 3 3 3 3 3 3 3 Neophema pulchella 5 7 7 7 7 6 6 6 6 6 6 6 6 6 6 Neophema splendida 10 10 12 12 12 11 11 11 11 11 11 10 11 11 11 Neopsephotos bourkii 7 7 8 7 7 8 8 8 8 8 8 7 8 8 8 Nestor notabilis 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Northiella haematogaster 10 13 13 13 13 13 13 13 13 12 12 13 13 13 13 Phigys solitarius 7 8 5 7 7 2 2 2 2 2 2 2 2 2 3 Pionus maximiliani 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Pionus menstruus 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Platycercus caledonicus 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Platycercus eximius 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 Platycercus flaveolus 3 2 2 2 2 3 3 3 3 3 3 2 2 2 2 Platycercus venustus 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Poicephalus gulielmi 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

127 Evolutionary diversification and biogeography of parrots

Appendix 1. Continued.

Poicephalus senegalus 9 9 9 9 9 8 8 8 8 8 8 8 8 7 8 Polytelis alexandrae 6 7 6 6 6 7 7 7 7 7 7 6 7 7 6 Polytelis anthopeplus 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Prioniturus luconensis 2 1 2 1 1 2 2 2 2 2 2 2 2 2 2 Prosopeia tabuensis 10 9 11 10 10 12 12 12 12 12 12 10 12 12 11 Psephotus chrysopterygius 5 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Psephotus dissimilis 9 9 10 9 9 9 9 9 9 9 9 9 9 9 8 Psephotus haematonotus 4 3 3 4 4 4 4 4 4 4 4 3 4 4 4 Psittacula eupatria 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Psephotus varius 4 5 5 5 5 5 5 5 5 5 5 4 5 5 4 Psittaculirostris desmarestii 2 3 2 2 2 2 2 2 2 2 2 2 2 2 2 Psittacus erithacus 6 5 6 6 6 6 6 6 6 6 6 6 6 6 6 Psitteuteles goldiei 5 5 5 6 6 4 4 4 4 5 4 4 4 4 5 Psittinus cyanurus 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Psittrichas fulgidus 4 3 4 4 4 3 3 3 3 3 3 3 3 3 3 Purpureicephalus spurius 4 4 3 4 4 4 4 4 4 4 4 4 4 4 4 Tanygnathus megalorhynchus 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Trichoglossus haematodus 16 16 16 16 16 12 12 12 12 12 12 12 12 12 11 Trichoglossus johnstoniae 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Triclaria malachitacea 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Vini australis 10 8 7 9 9 9 9 9 9 9 9 9 9 9 6

128 Chapter 5

Appendix 2. Species sampled, Museum and collection number, GenBank Accession numbers for the four genes analyzed. (xxx: GenBank Accession numbers will be requested by the publication of this chapter).

Species Museum/Collection Collection nb. c-mos Rag-1 Zenk ND2

Agapornis canus NMBE 1056201 GQ505083 GQ505191 GQ505138 - Agapornis fischeri NMBE 1056202 GQ505084 GQ505192 GQ505139 - Agapornis lilianae NMBE 1056205 GQ505087 - - - Agapornis nigrigenis NMBE 1056203 GQ505085 GQ505193 GQ505140 - Agapornis roseicollis NMBE 1056204 GQ505086 GQ505194 GQ505141 EU327596 Agapornis personatus NMBE 1059098 xxx xxx - xxx Alisterus chloropterus NMBE 1056207 GQ505091 GQ505199 GQ505145 - Alisterus scapularis NMBE 1056206 GQ505090 GQ505198 GQ505144 - Amazona aestiva NMBE 1056947 JF807952 JF807980 JF807966 AY194434 Amazona dufresniana NMBE 1056948 JF807953 JF807981 JF807967 - Amazona pretrei NMBE 1056949 JF807954 JF807982 JF807968 - Amazona xanthops NMBE 1059099 xxx xxx - DQ143316 Anodohynchus hyacinthinus - DQ143329 - DQ143311 Aprosmictus jonquillaceus NMBE 1056208 GQ505092 GQ505200 GQ505146 - Ara ararauna - DQ143340 - DQ143315 Aratinga leucophthalmus - DQ143331 - DQ143298 Ara macao NMBE 1056952 JF807951 JF807979 JF807965 EU327601 Aratinga solstitialis - xxx - DQ143317 Barnardius zonarius NMBE 1056210 GQ505095 GQ505203 GQ505149 xxx Barnardius barnardi NMBE 1059100 - xxx xxx xxx Brotogeris jugularis - - - EU327604 Cacatua goffini - DQ143355 - DQ143323 Cacatua moluccensis NMBE 1056236 GQ505121 GQ505232 GQ505174 - Cacatua sulphurea - - EU738913 EU327605 Chalcopsitta cardinalis - - EU738919 - Charmosyna pulchella NMBE 1056241 GQ505126 GQ505237 GQ505179 xxx Coracopsis vasa UWBM 85986/2004-001 GQ505113 GQ505223 GQ505167 EU327612 Cyanoramphus auriceps NMBE 1056221 GQ505104 GQ505213 GQ505158 xxx Cyanoramphus novaezelandiae NMBE 1056220 GQ505103 GQ505212 GQ505157 xxx Cyclopsitta diophthalma GQ505130 - - EU327616 Deroptyus accipitrinus NMBE 1056951 F807956 JF807984 JF807970 EU327617 Eclectus roratus NMBE 1056248 GQ505135 GQ505244 GQ505187 EU327619 Eos cyanogenia NMBE 1056237 GQ505122 GQ505233 GQ505175 xxx Eunymphicus (cornutus) cornutus NMBE 1056223 GQ505106 GQ505215 GQ505159 xxx Eunymphicus( cornutus) uvaeensis NMBE 1056224 GQ505107 GQ505216 GQ505160 xxx Forpus passerinus - - - EU327625 Guaruba guarouba NMBE 1056950 JF807957 JF807985 JF807971 EU327628 Lathamus discolor NMBE 1056219 GQ505102 GQ505211 GQ505156 xxx Loriculus galgulus UWBM 73841/2002-006 GQ505089 GQ505196 - EU327631 Loriculus philippensis ZMUC 130608 - GQ505197 GQ505143 - Lorius garrulus NMBE 1056240 GQ505125 GQ505236 GQ505178 xxx Melopsittacus undulatus UWBM 60748/1998-068 - GQ505222 GQ505166 EU327633 Micropsitta finschii UWBM 66040/2000-022 GQ505128 GQ505240 GQ505182 EU327634 Neophema chrysostoma NMBE 1056228 GQ505111 GQ505220 GQ505164 xxx Neophema pulchella NMBE 1056226 GQ505109 GQ505218 GQ505162 xxx Neophema splendida NMBE 1056225 GQ505108 GQ505217 GQ505161 xxx Neopsephotos bourkii UWBM 57542/1996-109 GQ505112 GQ505221 GQ505165 EU327639 Nestor notabilis NMBE 1056242 JF807958 GQ505238 GQ505180 EU327641 Northiella haematogaster NMBE 1056954 JF807959 JF807986 JF807972 xxx Phigys solitarius - - - EU327646 Pionus maximiliani - DQ143347 - EF517657 Pionus menstruus NMBE 1056955 JF807960 JF807987 JF807973 EU327650 Platycercus caledonicus NMBE 1056212 GQ505097 GQ505205 GQ505151 EU407679 Platycercus eximius NMBE 1056213 - GQ505206 GQ505152 EU407711 Platycercus flaveolus NMBE 1056215 GQ505099 GQ505208 - EU407696 Platycercus venustus NMBE 1056214 GQ505098 GQ505207 GQ505153 xxx Poicephalus gulielmi FMNH 390740 JF807961 JF807988 JF807974 - Poicephalus senegalus NMBE 1056231 - GQ505227 GQ505170 - Polytelis alexandrae NMBE 1056209 GQ505093 GQ505201 GQ505147 EU327653 Polytelis anthopeplus NMBE 1056657 GQ505094 GQ505202 GQ505148 EU407716 Prioniturus luconensis NMBE 1056247 GQ505134 - - EU327654 Prosopeia tabuensis NMBE 1056252 GQ505105 GQ505214 - EU327656 Psephotus chrysopterygius NMBE 1056953 JF807963 JF807991 F807977 xxx Psephotus dissimilis NMBE 1056218 GQ505101 GQ505210 GQ505155 xxx Psephotus haematonotus NMBE 1059101 xxx xxx xxx xxx Psittacula eupatria NMBE 1056250 GQ505137 GQ505246 GQ505189 - Psephotus varius NMBE 1056217 GQ505100 GQ505209 GQ505154 xxx Psittaculirostris desmarestii NMBE 1056244 GQ505131 GQ505242 GQ505184 - Psittacus erithacus NMBE 1056229 GQ505115 GQ505225 GQ505168 EU327661 Psitteuteles goldiei NMBE 1056239 GQ505124 GQ505235 GQ505177 xxx Psittinus cyanurus NMBE 1056251 - GQ505247 GQ505190 - Psittrichas fulgidus NMBE 1056243 GQ505127 GQ505239 GQ505181 EU327662 Purpureicephalus spurius NMBE 1056211 GQ505096 GQ505204 GQ505150 xxx Tanygnathus megalorhynchus NMBE 1056249 GQ505136 GQ505245 GQ505188 - Trichoglossus haematodus NMBE 1059102 - xxx - EU327671 Trichoglossus johnstoniae NMBE 1056238 GQ505123 GQ505234 GQ505176 xxx Triclaria malachitacea NMBE 1056232 GQ505117 GQ505228 GQ505171 AY669486 Vini australis - - - EU327672 Falco AY447974 AY461399 AF490155 EU196361 Pitta AY056952 AY057021 EF568299 GQ369692

129 &KDSWHU

Appendix 3. Mathematical appendix

We consider the general linear model y = Aβ + bβ + . 12 3 (1)

Here, y = (y1,...,yN) are the logarithmic trait values, A is the N × 2-matrix ⎛ ⎞ ⎜ ⎟ ⎜ 1 x1 ⎟ ⎜ ⎟ ⎜ 1 x2 ⎟ A = ⎜ . . ⎟ , ⎜ . . ⎟ ⎝⎜ ⎠⎟ 1 xN where x1,...,xN are the logarithmic size values. The vector

b = (1, 1,...,1, 0,...,0) contains ones at the K entries corresponding to the subgroup in question and zeros β = β ,β β elsewhere. 12 ( 1 2) and 3 are the unknown model parameters, and the error term ∼N(0,σ2Σ) is non-degenerate multivariate normal with covariance matrix σ2Σ where σ2 is an unknown factor whereas the matrix Σ is known. Setting X = [A|b] and β = (β1,β2,β3) , we can rewrite the model (1) more conveniently as

y = Xβ + . (2)

We have 2 2 Theorem 1. Let βˆ3 the GLM estimator of β3 and s the GLM estimator of σ . Moreover, set R = (0, 0, 1).

Under the null hypothesis H0 : β3 = 0 (no selection on the subgroup) the test statistic βˆ Tˆ := 3 (3) s R(XΣ−1X)−1R has the tN−3-distribution. Proof. Since Σ is positive deﬁnite and symmetric it has a symmetric and positive def- inite square root Q, i.e. Q2 = Σ. Multiplying both sides of (2) with Q−1 we get the homoskedastic model y0 = X0β + 0, (4) −1 −1 2 where y0 = Q y, X0 = Q X, and 0 ∼N(0,σ I). For this we have the usual estima- tors βˆ = X X −1X y = XΣ−1X −1XΣ−1y ( 0 0) 0 0 ( ) and 1 s2 = y I − X (X X )−1X y N − 3 0 0 0 0 0 0 1 = yΣ−1y − yΣ−1X(XΣ−1X)XΣ−1y . N − 3

(YROXWLRQDU\GLYHUVL¿FDWLRQDQGELRJHRJUDSK\RISDUURWV

It follows from the theory in Ruud 2000, Chapter 11, that the test statistic 1 −1 Fˆ := R(X X )−1R (Rβˆ)2 s2 0 0 for the model (4) has the F1,N−3-distribution under the null-hypothesis

H0 : Rβ = β3 = 0.

It is easy to check that −1 R(XΣ−1X)−1R Fˆ = βˆ2. s2 3 ˆ = ˆ1/2 ∼ 2 Since T F and F1,N−3 tN−3, the claim follows. A similar test statistic can be derived if not only a shift of intercept, but also a change of slope of the regression line for the subgroup is considered in the alternative hypothesis. In this case, we consider the general linear model y = Aβ + bβ + cβ + . 12 3 4 (5)

Notation is as above, the only new ingredient is

c = (x1, x2,...,xK, 0,...,0) where the non-zero entries correspond to the subgroup in question. The parameter β4 models a possible change in slope for the subgroup. Setting X = [A|b|c] and β = (β1,...,β4) , we can again rewrite the model (6) more conveniently as

y = Xβ + . (6)

An argument analogous to the one given above yields Theorem 2. Let βˆ the GLM estimator of β in (6) and s2 the GLM estimator of σ2. Moreover, set R = 0010 . 0001

Under the null hypothesis H0 : β3 = β4 = 0 (or equivalently Rβ = 0) the test statistic 1 −1 Fˆ := βˆ R R(XΣ−1X)−1R Rβˆ (7) 2s2 has the F2,N−4-distribution.

Literature cited

Ruud P. A. 2000. An Introduction to Classical Econometric Theory. Oxford University Press, Oxford, New York.

Chapter 6

OUT OF THE BASSIAN PROVINCE: HISTORICAL BIOGEOGRAPHY OF THE AUSTRALASIAN PLATYCERCINE PARROTS

Manuel Schweizera, Marcel Günterta, Stefan T. Hertwiga

aNaturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, CH 3005 Bern, Switzerland

Manuscript

135 Evolutionary diversification and biogeography of parrots

Aim The Australian continent was predominantly covered by rainforest in the past, until aridification especially from mid-Miocene onwards led to a fragmentation of mesic biomes and expansion of arid habitats. This had a major influence on the diversification of terrestrial organisms, and the general direction of their radiation is supposed to have been from rainforests into drier habitats. This has also been hypothesized for the platycercine parrots (Platycercini). They occur in a variety of habitats in Australia and occupy also islands in Melanesia and in New Zealand. We aim at providing a picture of the radiation of Platycercini based on a spatial and temporal framework. We tested the hypothesis that mesic biomes should map as ancestral in their phylogeny and examined colonization patterns in Melanesia and New Zealand.

Location Australia, Melanesia, New Zealand

Methods We used partial sequences of three nuclear and two mitochondrial genes to examine phylogenetic relationships in Platycercini. We investigated temporal and spatial patterns of their diversification using a Bayesian relaxed molecular clock approach and model-based biogeographic reconstructions.

Results Diversification within two of the three robustly supported platycercine clades started in the middle Miocene coinciding with the beginning of aridification in Australia. Mesic biomes were confirmed as the ancestral area of Platycercini and the Bassian biota were found to be their center of origin. Colonization of Pacific island proceeded from Australia over Melanesia to New Zealand. The phylogenetic relationships revealed contradict partially with traditional taxonomic treatment which was mainly based on non-molecular traits.

Conclusion The aridification of Australia from mid-Miocene onwards may have catalyzed the diversification of Platycercini. Ecological speciation caused by adaptation to the precursors of modern arid environments and adaptation to aridity in situ may both have played a role. The diversification events within Australian non-arid platycercine parrots seem to have been primarily shaped by vicariance across proposed faunal barriers due to habitat fragmentation following the spread of arid environments. The small oceanic islands of Melanesia contributed as stepping stones for the colonization of New Zealand providing further evidence that small islands are not just biodiversity sinks.

Keywords: Australasia, aridification, colonization, dispersal, historical biogeography, phylogeny, Platycercini, Psittaciformes, vicariance

136 Chapter 6

Introduction

The avifauna of the Australasian region and especially of its oceanic islands has played an important role for the development of theories on speciation and biogeography as well as the evolution of continental biota (Wallace, 1867; Cracraft, 1986; Mayr & Diamond, 2001; Filardi & Moyle, 2005; Mcdowall, 2008). The core of the Australasian bird fauna has a Gondwana origin and has radiated in isolation during the Paleogene under the influence of climate and environmental change as well as tectonic acitivities (Schodde, 2006; Christidis & Norman, 2010; Trewick & Gibb, 2010). The aridification of the Australian continent beginning in the middle Miocene or perhaps even earlier seems to have had a major influence on the diversification of the Australian bird fauna (Schodde, 2006; Byrne et al., in press). Then, a major change of vegetation began in Australia with shrinkage of rainforests, fragmentation of mesic biomes and an expansion of drier, more open sclerophyll forests. Today, arid and semi-arid habitats cover much of the Australian continent (Jacobs et al., 1999; Martin, 2006; Schodde, 2006). The general direction of radiation for land birds and other taxonomic groups in Australasia has been suggested to be from rainforests into drier and more open habitats (Schodde, 2006; Byrne et al., 2008; Byrne et al., in press). Ancestors of recent arid taxa may have either diverged from their closest relatives before the onset of aridification and adapted to arid environment in situ or originated from mesic ancestors through multiple diversification and colonization events of arid regions (reviewed in Byrne et al., 2008). The fragmentation of mesic biomes moreover led to localized endemism and phylogeographic structure in taxa specialized on these habitats (cf. e.g. Schodde, 2006; Byrne et al., in press). Parrots are one of the dominant elements of the Eyrean avifauna of the arid parts of Australia (Schodde, 2006), and especially members of the cockatoos and of the broad- tailed or platycercine parrots (Platycercini) are today adapted to dry environments (Collar, 1998; Rowley, 1998). The diversification of cockatoo genera occurred during the early Miocene to Pliocene, and the contemporaneous aridification may have promoted their radiation and the convergent evolution of specific morphological adaptations such as plumage color, body size, wing shape and bill structure across different lineages of cockatoos (White et al., 2011). The beginning of the radiation of the platycercine genera which comprise species adapted to arid habitats also seems to have coincided with the onset of the aridification of Australia (Schweizer et al., in press). Today, the Platycercini occur in a variety of habitats in the Australian continent with the exception of the Tumbunan and Irian rainforests and occupy also several islands in Melanesia and the South Pacific including New Zealand (Collar, 1998; Schodde, 2006). Christidis et al. (1991) proposed two successive radiations of Platycercini, a first continent-wide from mid-Miocene involving the divergence of Pezoporus, Melopsittacus, Neopsephotus, Neophema and the ‘core Platycercini’ (Platycercus, Barnardius, Northiella, Purpureicephalus, Psephotus and Lathamus) and the spread of ancestors of Prosopeia, Cyanoramphus and Eunymphicus to Pacific islands. A second radiation concerned the diversification of the ‘core Platycercini’ and was supposed to have been centered in the eucalypt-dominated Bassian biota found today in southern Australia with subsequent colonization of arid regions and the monsoon woodlands in the north. Reconstructions of the direction of radiations are heavily based on phylogenetic hypothesis and are only as good as the phylogenetic reconstruction which they are based upon (Schodde, 2006). The scenario formulated by Christidis et al. (1991) was based on allozyme data and did not rely on an explicit temporal and spatial framework. In the meantime, the results of several recent

137 Evolutionary diversification and biogeography of parrots molecular phylogenetic, but also morphological studies have rearranged the relationships of Platycercini and other Australasian parrots with some controversial issues still remaining. Importantly, Melopsittacus does not belong to Playcercini and is closely related to the lories, while Lathamus seems to be more closely related to Cyanoramphus, Prosopeia and Eunymphicus than to the ‘core Platycercini’ (Wright et al., 2008; Schweizer et al., 2010; Joseph et al., 2011; Schweizer et al., in press). In the present study, we aimed at reconstructing the radiation of Platycercini based on a spatial and temporal framework. We tested the hypothesis on their radiation formulated by Christidis et al. (1991) and the assumption by Byrne et al. (in press) that mesic biomes should be ancestral in Australian taxa containing both mesic and arid/monsoon zone representatives. We therefore constructed a molecular phylogeny of the platycercine parrots using three nuclear and two mitochondrial markers. We used a relaxed molecular clock approach and model-based ancestral area reconstructions to uncover historical and biogeographic patterns.

Material and Methods

Sampling and laboratory methods We obtained tissue or feather samples from 35 species of old world parrots complemented with sequences from GenBank (Table 1). We included all species of the ‘core Platycercini’ with the exception of the almost certainly extinct paradise parrot Psephotus pulcherrimus (Collar, 1998) and four species of the well-defined monophyletic genus Platycercus (Joseph et al., 2008). Moreover, we included samples of all remaining platycercine genera and potentially affiliated Australasian taxa, and Nestor notabilis was chosen as outgroup (cf. De Kloet & De Kloet, 2005; Wright et al., 2008; Schweizer et al., 2010; Schweizer et al., in press). The taxonomy of Collar (1998) is followed here, though we include ‘Geopsittacus’ in Pezoporus (Joseph et al., 2011) and follow the recommendations of Boon et al. (2001) for Cyanoramphus. Total genomic DNA was isolated using peqGOLD tissue DNA Mini Kit (Peqlab) or DNeasy Blood & Tissue Kit (Qiagen) following manufacturer’s rule. Partial sequences of the three nuclear genes c-mos, RAG-1 and Zenk (second exon) and the two mitochondrial genes ND2 (NADH2) and Cytb (Cytochrome b) were amplified with Polymerase chain reaction (PCR) using published primer sequences (Table 2). PCR reaction volumes were 20 µl containing 10µl PCR-Master-Mix S (c-mos), PCR- Master-Mix Y (RAG-1, Zenk, ND2) or HotStart-Master-Mix (Cytb) (Peqlab), 2-3µl genomic DNA, 2µl of each primer with a concentration of 10µM and 3-4µl ddH2O. PCR was performed on a Techne TC-512 thermo-cycler. Amplification of c-mos was performed with the following parameters: initial denaturation of 94°C for 2 min followed by 33 cycles of denaturation at 90°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 1 min, with a final extension at 72°C for 5 min. The PCR reaction profile published by Groth and Barrowclough (1999) was used for RAG-1 with the initial denaturation step reduced to 2 min. For Zenk, the PCR profile of Chubb (2004) was used with the annealing temperature set to 53.5°C. The PCR Protocol of Tavares et al. (2006) with the annealing temperature set to 53° was applied for ND2. Amplification of Cytb was performed with an initial denaturation of 94°C for 3 min followed by 35 cycles of denaturation at 95°C for 30 s, annealing at 54.5°C for 30 s, and extension at 72°C for 1 min, with a final extension at 72°C for 7 min. PCR products were examined by gel electrophoresis to confirm the amplification

138 Chapter 6 of the target fragment. PCR products were either excised from gels and cleaned using the Wizard® SV Gel and PCR Clean-UP System (Promega) or directly purified with the above mentioned kit or with the peqGOLD MicroSpin Cycle-Pure Kit (Peqlab). To increase the quantity of DNA for problematic samples, the products of two independent PCR runs were put together before the cleaning. Sequencing was carried out with Microsynth AG (Balgach, Switzerland). Cytb was sequenced with the primer L15132, while the other four genes were sequenced from both sides using the same primers as for amplification leading to complete overlapping fragments for RAG-1 and c-mos and an overlapping fragment of about 600 bp for Zenk and of about 500 bp for ND2. Sequencing files were checked with Chromas (Technelysium Pty. Ltd.) and ambiguities were assigned standard IUB codes. The alignment of the sequences was done manually after translation of the sequences into amino acids with BioEdit 7.0.5.2 (Hall, 1999). We checked individual sequences and the whole alignment further for apparent stop codons and for indels that were not a multiple of three bases.

Table 1 Species sampled, Museum and collection number, GenBank Accession numbers for the five genes analyzed. (xxx: GenBank Accession numbers will be requested by the publication of this chapter)

Species Museum/Collection Collection nb c-mos RAG-1 Zenk ND2 Cytb

Agapornis roseicollis NMBE (except ND2) 1056204 GQ505086 GQ505194 GQ505141 EU327596 xxx Barnardius zonarius NMBE 1056210 GQ505095 GQ505203 GQ505149 xxx xxx Barnardius barnardi NMBE 1059100 - xxx xxx xxx xxx Charmosyna pulchella NMBE 1056241 GQ505126 GQ505237 GQ505179 xxx xxx Cyanoramphus auriceps NMBE 1056221 GQ505104 GQ505213 GQ505158 xxx xxx Cyanoramphus novaezelandiae NMBE 1056220 GQ505103 GQ505212 GQ505157 xxx xxx Eclectus roratus NMBE (except ND2) 1056248 GQ505135 GQ505244 GQ505187 EU327619 xxx Eos cyanogenia NMBE 1056237 GQ505122 GQ505233 GQ505175 xxx xxx Eunymphicus cornutus cornutus NMBE 1056223 GQ505106 GQ505215 GQ505159 xxx xxx Eunymphicus cornutus uvaeensis NMBE 1056224 GQ505107 GQ505216 GQ505160 xxx xxx Lathamus discolor NMBE 1056219 GQ505102 GQ505211 GQ505156 xxx xxx Loriculus galgulus UWBM (except ND2) 73841/2002-006 GQ505089 GQ505196 - EU327631 xxx Lorius garrulus NMBE 1056240 GQ505125 GQ505236 GQ505178 xxx xxx Melopsittacus undulatus UWBM (except ND2) 60748/1998-068 - GQ505222 GQ505166 EU327633 xxx Micropsitta finschii tristrami UWBM (except ND2) 66040/2000-022 GQ505128 GQ505240 GQ505182 EU327634 xxx Neophema chrysostoma NMBE 1056228 GQ505111 GQ505220 GQ505164 xxx xxx Neophema pulchella NMBE 1056226 GQ505109 GQ505218 GQ505162 xxx xxx Neophema splendida NMBE 1056225 GQ505108 GQ505217 GQ505161 xxx xxx Neopsephotus bourkii UWBM (except ND2) 57542/1996-109 GQ505112 GQ505221 GQ505165 EU327639 xxx Nestor notabilis NMBE (except ND2) 1056242 JF807958 GQ505238 GQ505180 EU327641 xxx Northiella haematogaster NMBE 1056954 JF807959 JF807986 JF807972 xxx xxx Pezoporus wallicus ANWC 45982 HQ316825 HQ316839 HQ316854 HQ316878 HQ324422 Platycercus caledonicus NMBE (except ND2) 1056212 GQ505097 GQ505205 GQ505151 EU407679 xxx Platycercus eximius NMBE (except ND2) 1056213 - GQ505206 GQ505152 EU407711 xxx Platycercus flaveolus NMBE (except ND2) 1056215 GQ505099 GQ505208 - EU407696 xxx Platycercus venustus NMBE 1056214 GQ505098 GQ505207 GQ505153 xxx xxx Polytelis alexandrae NMBE (except ND2) 1056209 GQ505093 GQ505201 GQ505147 EU327653 xxx Prosopeia tabuensis NMBE (except ND2) 1056252 GQ505105 GQ505214 - EU327656 xxx Psephotus chrysopterygius NMBE 1056953 JF807963 JF807991 JF807977 xxx xxx Psephotus dissimilis NMBE 1056218 GQ505101 GQ505210 GQ505155 xxx xxx Psephotus haematonotus NMBE 1059101 xxx xxx xxx xxx xxx Psephotus varius NMBE 1056217 GQ505100 GQ505209 GQ505154 xxx xxx Psittacella brehmii KU 12931/91954 JF807964 JF807992 JF807978 xxx xxx Psitteuteles goldiei NMBE (except Cytb) 1056239 GQ505124 GQ505235 GQ505177 xxx AF346343 Purpureicephalus spurius NMBE 1056211 GQ505096 GQ505204 GQ505150 xxx xxx

139 Evolutionary diversification and biogeography of parrots

Table 2 Primers used for the amplification of the five genes analyzed in this study.

Gene Primer name Reference

c-mos F944 (Cooper & Penny 1997) R1550or05 (Overton & Rhoads 2004) R1550hb99 (Hughes & Baker 1999) RAG-1 R8 (Groth & Barrowclough 1999) R11B (Groth & Barrowclough 1999) R17 (Groth & Barrowclough 1999) R18 (Groth & Barrowclough 1999) Zenk Z1F (Chubb 2004) Z9R (Chubb 2004) ND2 MetL (Tavares et al 2006) ASNH (Tavares et al 2006) Cytb L15132 (Boon et al 2000) H16065 (Boon et al 2000)

Phylogenetic analyses Nuclear and mitochondrial markers were analysed combined and separately throughout. Bayesian inference was conducted in MrBayes 3.2 (Huelsenbeck & Ronquist, 2001; Ronquist & Huelsenbeck, 2003) using a mixed-model approach. We evaluated alternative biologically relevant parameter settings corresponding to separate models with varying base frequencies, rate matrix, shape parameters and proportion of invariable sites for the concatenated data and the different genes and/or their codon positions. Nucleotide substitution models were evaluated with MrModeltest 2.3 (Nylander, 2004) using the Akaike information criterion (Akaike, 1974). We performed two independent runs of Metropolis-coupled Markov chain Monte Carlo analyses for each parameter setting, each run consisting of one cold chain and three heated chains with a default temperature of 0.2. The chains were run for 10 million generations with sampling every 100 generations. We verified that the average standard deviation of split frequencies converged towards zero and calculated the length of ‘burn-in’ by monitoring the change in cumulative split frequencies using AWTY (Wilgenbusch et al., 2004; Nylander et al., 2008). We then discarded the first 25% of samples as burn-in (25,000 trees). Moreover, we compared likelihoods and posterior probabilities of all splits and parameters to assess convergence among the two independent runs using AWTY and Tracer 1.5 (Rambaut & Drummond, 2007). We used the Bayes Factor (BF) to evaluate the relevance of the different parameter settings (Kass & Raftery, 1995; Brown & Lemmon, 2007). The harmonic mean calculated by MrBayes was used as an estimation of the marginal likelihood of the data (Kass & Raftery, 1995). A more complex model (i.e. more parameters) was considered to be more supported compared to a simpler model if 2lnBF was greater than 10 (Brown & Lemmon, 2007). We employed RAxML 7.0.4 (Stamatakis, 2006) for the maximum likelihood (ML) search. The software was run on the web server with 100 rapid bootstrap inferences (Stamatakis et al., 2008) with all free model parameters estimated by the software (substitution rates, gamma shape parameter, base frequencies) using the best parameter setting as obtained with MrBayes. For the

140 Chapter 6 maximum parsimony analyses we conducted heuristic searches in PAUP* (Swafford, 2001) (1000 random taxon-addition replicates, TBR branch swapping) with gaps treated as a fifth character state. Nodal support was estimated with a maximum parsimony bootstrap analysis (1000 pseudo-replicates, heuristic search, 10 random taxon-addition replicates, TBR branch swapping, number of max trees limited to 100). We considered clades as being supported when the clade credibility values of the Bayesian inference were ≥ 0.95 (Huelsenbeck & Ronquist, 2001) and when bootstrap values were ≥ 70 in the maximum likelihood and maximum parsimony analyses (Hillis & Bull, 1993).

Molecular dating Divergence times were estimated using Beast 1.6.1 (Drummond & Rambaut, 2007). We applied a relaxed molecular clock with uncorrelated lognormal distribution of branch lengths and a Yule tree prior with the best parameter setting found via the MrBayes analyses and the respective nucleotide substitution models as evaluated in MrModeltest. As there are no representative fossil parrots which could provide prior distributions for primary calibrations (cf. Mayr, 2009; Schweizer et al., in press), the calibration adopted was based on a molecular dating analysis of the parrots which was calibrated with well-accepted avian fossils outside the parrots (Schweizer et al., in press). According to the results of this study, we used a mean estimate of 58 Ma (millions of years ago) with a normal distribution (95% +/- about 20 million years or SD = 10) for the initial split within the crown group parrots separating Nestor (and Strigops) from the remaining taxa. This time interval also includes similar estimates of other studies for the initial split within crown group parrots based on fossil calibrations (Pratt et al., 2009; Pacheco et al., 2011; White et al., 2011). Moreover, the ‘core Platycercini’ (excluding Lathamus, see below) were defined as being monophyletic and their initial split was used as a second calibration point with a mean estimate of 20 Ma and a normal distribution (95% +/- about 9 million years or SD = 4.5) again based on the results of Schweizer et al. (in press). Default prior distributions were chosen for all other parameters and the MCMC was run for 25 million generations with sampling every 1000 generations. We used Tracer 1.5 (Rambaut & Drummond, 2007) to confirm appropriate burn-in and the adequate effective sample sizes of the posterior distribution. Two independent chains were run and we compared likelihoods and posterior probabilities of all parameters to assess convergence among the runs using Tracer. The resulting maximum clade credibility tree and the 95% highest posterior density (HPD) distributions of each estimated node were visualized with FigTree 1.2.1 (Rambaut, 2008).

Biogeographic reconstruction We used the likelihood approach of the software Lagrange to reconstruct the biogeographic history of the platycercine parrots (Snapshot version for web configuration tool at http://www-reelab.net/lagrange) (Ree et al., 2005; Ree & Smith, 2008). The reconstruction was based on the maximum clade credibility tree obtained in Beast. We were primarily interested in the biogeographic history of Platycercini, thus all other taxa were pruned from the tree. Lagrange treats dispersal and local extinctions as stochastic processes incorporating a continuous-time model for geographic range evolution through dispersal, extinction and cladogenesis (DEC model) (Ree et al., 2005; Ree & Smith, 2008; Ree & Sanmartin, 2009). According to their current distribution (Collar, 1998), the zoogeographic delimitation of Merrick et

141 Evolutionary diversification and biogeography of parrots al. (2006) an the classification of Schodde (2006), the taxa were assigned to the following biogeographic provinces (avifaunulas): Melanesia (Fiji, Tonga, New Caledonia), New Zealand as well as Bassian, Eyrean and Torresian regions. Bassian taxa occur in sclerophyllous forests and woodlands of south, south-west and southeast Australia as well as Tasmania, while Eyrean taxa occupy the dry areas of inland and western Australia (mulga woodlands, spinifex dune deserts and dry stony ranges). Torresian species are confined to monsoonal eucalypt and paperbark woodlands in northern Australia (Schodde, 2006) (Fig. 4). Pezoporus wallicus occurs in Bassian biota, Pezoporus occidentalis, its sister taxon (Joseph et al., 2011), is an Eyrean species (Schodde, 2006). Therefore both biogeographic provinces were assigned to Pezoporus. As distribution of parrots may have spanned over more than one biogeographical area in the past as in some species today, the maximum range size was not restricted and all combinations of areas were allowed in the adjacency matrix. Baseline rates of dispersal and local extinction were estimated by the software.

Results

Phylogenetic analyses The final alignment consisted of 5121 bp with 603 stemming from c-mos, 1458 from RAG-1, 1143 from Zenk, 876 from Cytb and 1041 from ND2. It contained one indel of two and one indel of three amino acids for RAG-1, and one indel of one amino acid for Zenk. Different models of nucleotide evolution were selected for the different partitions of our data set (Table 3). Combined analyses: the parameter setting using different models of sequence evolution for the five genes and three codon positions was chosen as the best-fitting model after calculation of the Bayes Factors. Based on this parameter setting, the analysis was rerun for 20 million generations with the temperature set to 0.1. Most basal nodes were not supported in the resulting 50% majority rule consensus tree (Fig. 1), and especially the position of Psittacella was not resolved. The monophyly of Platycercini was not robustly supported. Within Platycercini, a cluster consisting of Pezoporus, Neophema and Neopsephotus bourkii was the sister group to the remaining taxa. The latter were divided into two well supported clades: the first comprised Prosopeia, Eunymphicus and Cyanoramphus as well as Lathamus discolor, while the second consisted of the remaining ‘core Platycercini’. Within the former clade, Lathamus discolor split first from the remaining taxa followed by Prosopeia. This is in congruence with Joseph et al. (2011), though the position of Prosopeia was not robustly resolved here. Within the ‘core Platycercini’, the sister group relationship between Barnardius and Platycercus was robustly supported. Purpureicephalus spurius clustered together with Psephotus varius, P. dissimilis and P. chrysopterygius. Psephotus haematonotus did not cluster with the remaining species of its genus. However, its position, as well as that of Northiella haematogaster, was not robustly supported. The topology of the best-scoring maximum likelihood tree was similar to the 50% majority rule consensus tree of the Bayesian inference (Fig. 1). However, Purpureicephalus spurius and Psephotus varius were revealed as sister taxa, though this clade was not supported. Two trees were found to be equally parsimonious in maximum parsimony analyses. The strict consensus tree revealed Platycercini to be monophyletic, but this clade was not supported. The topology within this taxon was very similar to the Bayesian tree. However, Northiella haematogaster and Psephotus haematonotus were the sister group to the remaining ‘core Platycercini’ and the

142 Chapter 6 position of Purpureicephalus spurius and Psephotus varius was not resolved in their cluster with P. dissimilis and P. chrysopterygius. Further, Lathamus discolor and Prosopeia formed a basal polytomy with the clade consisting of Eunymphicus and Cyanoramphus. The topology of the Beast tree (Fig. 3) differed from the 50% majority rule consensus tree of MrBayes in the position of Psittacella and the clade consisting of Agapornis and Loriculus. Furthermore, Purpureicephalus spurius clustered with Psephotus varius as in the maximum likelihood tree. When the nuclear and mitochondrial markers were analyzed separately using MrBayes, the following parameter setting were chosen: different models of sequence evolution for the two genes and three codon positions in the mitochondrial markers, and for the three genes and first and second vs. third codon position for the nuclear markers. Compared with the combined analyses, there were a smaller number of supported nodes in the resulting trees for both datasets (Fig. 2), however, there was no conflict between supported nodes.

Table 3 Sequence evolution models for the five different genes and their different partitions analyzed.

model

number of bp all 1 pos 1&2 pos 2 pos 3 pos

c-mos 603 GTR+G HKY HKY+G K80+I HKY+G RAG-1 1458 GTR+I+G HKY+G GTR+I+G GTR+I HKY+G Zenk 1143 HKY+I HKY+I HKY+I HKY HKY+G Cytb 876 GTR+I+G SYM+G GTR+I+G GTR+I+G GTR+G ND2 1041 GTR+I+G GTR+I+G GTR+I+G HKY+I+G GTR+I+G combined data set 5121 GTR+I+G nuclear markers 3204 HKY+I+G mitochondrial markers 1917 GTR+I+G

Molecular dating The comparison of the two independent runs for the Beast analyses revealed high convergence among the different parameters. The two runs were then combined with 2.5 million generations of burn-in for each. Effective samples sizes were >214 for all parameters and the maximum clade credibility tree was then estimated from 50003 trees. Divergence time estimates dated the initial split within the crown group parrots at 46.54 Ma (95% highest posterior density HPD: 30.80-62.82 Ma) below the mean value of our prior distribution (Fig. 3, Table 4). This is close to the results of White et al. (2011) and Pratt et al. (2009), whose mean age estimate placed the same split at 47.4 Ma or 50.38 Ma respectively based on complete mitochondrial genomes and external calibration points. The initial split within Platycercini occurred at 30.96 Ma (95% highest posterior density HPD: 20.43-41.43), while the diversification of the ‘core Platycercini’ started around 22.33 Ma (95% highest posterior density HPD: 14.88-29.73 Ma) above the mean value of the prior distribution. The diversification of the main clades of Platycercini started around 15 Ma in the middle Miocene. The mean values of the time estimates for the splits outside Platycercini were found to be older (about four millions of years, split between Loriculus and Agapornis; about seven millions of

143 Evolutionary diversification and biogeography of parrots years, split of Melopsittacus from Lorini) or younger (about five millions of years, split of Micropsitta from Psittaculini) compared to our previous study (Schweizer et al., in press). These divergence time estimates may have been influenced by incomplete taxon sampling outside Platycercini. On the other hand, this should not have influenced the divergence estimates for the splits within Platycercini as these were constrained by our prior distributions.

Table 4 Estimated divergence dates for different nodes using the Bayesian relaxed molecular clock approach of Beast. The mean values and the 95% highest posterior density (HPD) distributions are given. Node numbers refer to Fig. 3.

Node Number Mean 95% HPD Millions of years ago

1 49 54 30 80-62 82 2 30 96 20 43-41 43 3 22 33 14 88-29 73 4 27 65 17 91-37 26 5 22 34 14 24-30 47 6 13 49 8 26-18 81 7 8 4 48-11 6 8 16 05 10 51-21 76 9 14 09 9 2-19 38 10 11 32 7 11-15 61 11 5 01 2 93-7 11 12 2 55 1 25-4 06 13 2 49 1 25-3 82 14 1 40 0 68-2 24 15 12 29 7 73-16 9 16 10 64 6 63-15 17 6 88 3 91-10 02 18 13 18 8 34-18 31 19 11 45 7 05-15 97 20 7 97 4 62-11 42 21 0 94 0 39-1 6 22 0 22 0 03-0 47

Biogeographic reconstruction The global ML at the root node was -45.33 (dispersal rate = 0.01168, extinction rate = 0.007814). The reconstruction of the biogeographic history for each node, based on the split between areas with the highest likelihood, found the origin of Platycercini in the Bassian province. We revealed that the Eyrean province was colonized six times from the Bassian province (Fig. 4). A derivation of Torresian from Bassian taxa was found twice. Colonization of the Pacific islands first comprised dispersal from the Bassian region to Melanesia followed by colonization of New Zealand. For several nodes, alternative splits were found within two log-likelihood units of the maximum for each node (Table 5), however, they did not change the general picture of dispersal and colonization events.

144 Chapter 6

Table 5 Biogeographic reconstruction with Lagrange for the different area splits within Platycercini. All reconstructions within two log-likelihood units of the maximum for each node are shown, with Rel.Prob. being the relative probability (fraction of the global likelihood) of a split. The numbering of the nodes corresponds to Fig. 4. The split format reads as follows: [left|right], where 'left' and 'right' are the ranges inherited by each descendant branch with 'left' corresponding to the upper branch, and 'right' to the lower branch in the phylogenetic tree in Fig. 4.

Node Split lnL Rel Prob

145 Evolutionary diversification and biogeography of parrots

Figure 1 50% majority-rule consensus tree of the Bayesian inference using MrBayes. Clade credibility values and bootstrap values above 50 of the maximum likelihood inference with RAxML are indicated at each node. Platycercini and a notation used in the text for a group within this clade are indicated to the right of the tree.

146 Chapter 6

Figure 2 Pie charts indicate at each node the support values in the separate analyses of the mitochondrial and nuclear markers for the Bayesian inference, maximum likelihood and maximum parsimony analyses. The tree topology corresponds to the 50% majority-rule consensus tree of the Bayesian inference using MrBayes.

147 Evolutionary diversification and biogeography of parrots

Figure 3 Maximum clade credibility tree of the dating analysis using Beast with nodes corresponding to estimated mean ages. 95% highest posterior density (HPD) distributions are shown for the nodes in the tree that had a posterior probability greater than 0.5. Node numbers refer to Table 4.

Discussion

Phylogenetic relationships The phylogenetic relationships within the ‘core Platycercini’ revealed in this study are in some cases in contradiction with traditional thinking and taxonomic treatment which is mainly based on non-molecular traits. We especially show that shared plumage features may not always be a sign of a close phylogenetic relationship, but may instead indicate a plesiomorphic pattern or convergent evolution. We uncovered the clade consisting of Pezoporus, Neopsephotus bourkii and Neophema as the sister group of the remaining Platycercini and found no indication for a close relationship between the former clade and Agapornis-Loriculus (cf. Wright et al., 2008). However, the monophyly of Platycercini was not robustly supported throughout. Similarly, Joseph et al. (2011) concluded that the Pezoporus- Neopsephotus-Neophema clade was most likely the sister group of the remaining Platycercini although this relationship was never robustly supported in their data. The phylogenetic position of Pezoporus-Neopsephotus-Neophema also remains unclear from morphological data (Mayr, 2010), thus more data need to be analyzed to robustly

148 Chapter 6 resolve their phylogenetic relationship. The same is true for Psittacella: like other molecular phylogenetic studies (Joseph et al., 2011; Schweizer et al., in press), we could not robustly resolve its phylogenetic position, but confirm that it has neither close affinity to Psittaculini nor to Platycercini. We corroborated that Lathamus discolor is not part of the ‘core Platycercini’ (cf. Schweizer et al., 2010; Joseph et al., 2011; Schweizer et al., in press). The relationships within the latter group are in conflict with their current generic treatment. Importantly, the genus Psephotus was found to be polyphyletic. Forshaw (1981) considered Psephotus haematonotus and P. varius as closely allied due to their similar plumage patterns and marked sexual dimorphism. However, we consistently found P. varius in a clade with Purpureicephalus spurius and the two closely related taxa Psephotus chrysopterygius and P. dissimilis (cf. Joseph et al. 2011). Although the relationships of Psephotus haematonotus in combination with Northiella haematogaster could not be resolved with robust support throughout, they were found to be closer related to Platycercus and Barnardius than to the clade consisting of the other Psephotus species and P. spurius. Joseph et al. (2011) could neither robustly resolve the position of P. haematonotus (and N. haematogaster), but in congruence with our results, it never clustered with the remaining Psephotus species analyzed there (P. dissimilis, P. varius). The revealed relationship of P. spurius is remarkable. It differs from the remaning Platycercini among others by a narrow projecting bill and by a narrowing of the anterior of the cranium (Forshaw, 1981). Early authors have placed it near Platycercus and Barnardius (e.g. Finsch, 1867; Salvadori, 1891) or even together with Eunymphicus (Cain, 1955), while Homberger (1980) described similarities in the auditory region of the skull between Purpureicephalus, Psephotus and Barnardius. Generic treatment will certainly have to be revised in the ‘core Platycercini’, however, before the relationships of N. haematogaster and P. haematonotus are consistently resolved, we refrain from proposing any provisional revision. Based on the results of Joseph et al. (2011) and this study, an assembly of all ‘core Platycercini’ in the genus Platycercus under priority rules would be the best solution in order to avoid the creation of multiple phylogenetically uninformative monotypic genera. The two clades within the ‘core Platycercini’ indicated in our study could also support a division into the two genera Platycercus (including Platycercus, Barnardius, Psephotus haematonotus and Northiella haematogaster) and Purpureicephalus (including the remaining Psephotus species and Purpureicephalus spurius).

Biogeography, spatial and temporal diversification patterns The mesic biome was confirmed as the ancestral area of Platycercini. According to our biogeographic reconstruction, the group has its centre of diversity in the Bassian biota and seems to have originated there around 32 Ma. All dry-adapted Eyrean platycercine parrots, the Torresian as well as the Melanesian and New Zealand taxa thus derive from Bassian ancestors (cf. Christidis et al., 1991; Schodde, 2006). The mesic biomes of the northern, eastern and southern coastline of Australia today were more widespread until the mid-Tertiary and stem from the ancestral Mesozoic Gondwanan forests (Schodde, 2006). The Australian landscape only changed during the Neogene from one dominated by rainforest to one dominated by open vegetation. Present-day communities had developed to some degree only by the end of the Pliocene with subsequent major changes during the Pleistocene (Kershaw et al., 1994) and ever-wet forests are now restricted to small scattered areas along the east coast (Byrne et al., in press). Although no platycercine parrots occur in the rainforest

149 Evolutionary diversification and biogeography of parrots remains of Australia today, we cannot exclude that this was the case when this biome was more widespread in the past. There is evidence that rainforest communities suffered extinction during the Neogene (Byrne et al., in press) and this may have also been the case in platycercine parrots. The divergence between the ‘core Platycercini’ and the last common ancestor of Lathamus, Prosopeia, Eunymphicus and Cyanoramphus took place in the beginning of the Miocene around 23 Ma. Both groups started only to diversify from middle Miocene onwards coinciding with the beginning of severe aridification in Australia (Kershaw et al., 1994; Jacobs et al., 1999; Martin, 2006; Byrne et al., 2008). Thus, early divergence within Platycercini and the radiation of the ‘core Platycercini’ occurred earlier than proposed by Christidis et al. (1991). The Eyrean taxa which occupy the dry areas of Australia were not found to cluster together and were all independently derived from Bassian ancestors. We thus found no indication for speciation within the arid zone, although we cannot exclude that arid lineages went extinct especially during the dramatic climate changes of the Pleistocene (cf. Byrne et al., 2008). The lineage of the strictly Eyrean taxa Neopsephotus bourkii split rather early from its closest relative (Neophema) in the late Oligocene/early Miocene (around 24 Ma). This divergence occurred before the onset of the aridification of Australia, and N. bourkii may have adapted to arid landscapes in situ. In contrast, Northiella haematogaster became a separate evolutionary lineage in the middle Miocene (around 16 Ma) contemporaneous with the start of severe aridification. Other arid-adapted lineages within Platycercini concern Psephotus varius and both Barnardius species. They not only occur in Eyrean, but also in Bassian biota, and colonized arid biota from there without subsequent speciation. However, there is evidence for differentiation along a forest to dry open habitat environmental gradient in Barnardius (Joseph & Wilke, 2006). We found the divergence between the two Barnardius species to have occurred around three Ma in accordance with Joseph & Wilke (2006), who inferred that ancestral populations of the Barnardius-complex were first split at around two Ma based on data of the mitochondrial ND2 gene. They proposed that vicariance across the Eyrean barrier could have caused this divergence. This barrier marks the division between eastern and western Bassian and Eyrean biota at the head of Spencer Gulf (Joseph & Wilke, 2006; Schodde, 2006). An east-west Bassian disjunction is also apparent between Neophema splendida and N. pulchella. However, the split between these two taxa occurred earlier around eight Ma during a period of major climate change termed the ‘Hill Gap’ (Byrne et al., 2008). This interval, which lasted from ten to six Ma, marked the beginning of extreme aridity in central Australia (Byrne et al., 2008). Purpureicephalus spurius is restricted to southwest Australia where it inhabits a variety of wooded habitats and is closely associated with marri Eucalyptus calophylla (Forshaw, 1981; Schodde, 2006). Southwest Australia is rich in endemic taxa and recognized as a global biodiversity hotspot (Myers et al., 2000; Hopper & Gioia, 2004). Schodde (2006) hypothesized, that P. spurius became isolated in drier glacial times during the Quaternary when forests became isolated in refugial pockets there. Such a scenario has been found in cockatoos, where the south-western endemic species Calyptorhynchus baudinii and C. latirostris split from the eastern species C. funereus at around 1.3 Ma (White et al., 2011). However, we found that P. spurius split from Psephotus varius already in the middle Miocene around 11 Ma and the isolation of P. spurius may have rather been caused by the expansion of arid landscapes in this time period (Kershaw et al., 1994; Jacobs et al., 1999; Martin, 2006; Byrne et al., 2008).

150 Chapter 6

The separation of the Torresian ‘Psephotellus’ taxa (Psephotus chrysopterygius and P. dissimilis) from the Bassian Pesphotus varius and Purpureicephalus spurius occurred at around 13 Ma. The divergence of the last common ancestor of these taxa into a northern and southern form may have coincided with the contraction and fragmentation of mesic biomes at this time and the development of separate northern and southern parts of sclerophyll habitats into Torresian and Bassian biota (cf. Schodde, 2006). The distribution areas of Psephotus chrysopterygius and P. dissimilis are divided by the Carpentarian faunal barrier, and their split occurred in the late Miocene or early Pliocene at the end of the ‘Hill Gap’ (see above). The Carpentarian barrier or gap acted as a serial barrier to trans-Torresian connections (Cracraft, 1986; Schodde, 2006; Bowman et al., 2009), and temporal patterns of disjunctions along it were found to be rather variable. In two honeyeater species, they ranged from around four Ma (Melithreptus albogularis) to around 0.7 Ma (M. gularis) (Toon et al., 2010), while grassfinch species (Poephila cincta and P. acuticauda/hecki) diverged at 0.6 Ma (Jennings & Edwards, 2005) and divergence times around 0.27 Ma were estimated in the Fairy Wren Malurus melanocephalus (Lee & Edwards, 2008). This rather high variation in the temporal divergence between taxa on either side of the Carpentarian barrier seems to reflect a long and dynamic history influenced by Plio-Pleistocene climatic cycles and different ecological conditions with varying influences on taxa with different niches and habitat requirements (Bowman et al., 2009; Toon et al., 2010). A common ancestor of Prosopeia, Eunymphicus and Cyanoramphus split from the Bassian taxon Lathamus discolor at around 14 Ma and colonized Melanesia and later New Zealand. L. discolor is one of very few migratory parrots today (Collar, 1998). If a migratory behavior or a predisposition for it was already present in the common ancestor of this clade, this could have greatly facilitated dispersal and colonization (Joseph et al., 2011). The genus Prosopeia is restricted to Fiji and Tonga (Collar, 1998) and forms the sister taxon to Eunymphicus and Cyanoramphus. While the former occurs on New Caledonia, the latter is distributed throughout the South Pacific including New Caledonia (Collar, 1998). While Mayr (1939) proposed that Prosopeia has a Papuan origin, Rinke (1989) considered it as the only New Zealand element in the resident Fijian avifauna. In contrast, our reconstruction revealed Prosopeia to be of Australian origin. Most other species of the Fijian avifauna are thought to have originated in Papua, while the few Australian taxa are considered as recent immigrants (Mayr, 1939). Prosopeia clearly is derived from an Australian stock, however, the taxon may have once been more widely distributed in the region and we cannot exclude that it also occurred in New Guinea. The region of Melanesia has a complex tectonic history (cf. Hall, 2002) and New Guinea existed in separated fragments north of Australia at the beginning of the Miocene and began to form its present state only four to five Ma (Schodde, 2006). Intermittent land connections along the Melanesian arc could have facilitated dispersal between New Guinea and Fiji (Irestedt et al., 2008; Jonsson et al., 2010). While Fleming (1980) proposed that New Zealand was the centre of origin and dispersal in Cyanoramphus, we found it to have its origin in New Caledonia in congruence with other studies (Boon et al., 2001; Chambers et al., 2001). Cyanoramphus may have dispersed to New Zealand, probably via Norfolk Island, after having diverged from Eunymphicus in New Caledonia (Chambers et al., 2001). The New Zealand avifauna is generally most closely related to that of Australia (Trewick & Gibb, 2010), but the Melanesian island group of New Caledonia has recently been described as a major source of different taxa dispersing to New Zealand.

151 Evolutionary diversification and biogeography of parrots

A close biogeographic connection between New Zealand and New Caledonia could also be found in geckos and different genera of cicadas (Chambers et al., 2001; Heads, 2010). While the biogeographic patterns observed in the geckos seem to have been caused by vicariance, the cicada apparently colonized New Zealand by over- water dispersal or via intermittent land connections (Chambers et al., 2001). Although Heads (2010) argued that the origin of Cyanoramphus could be explained by vicariance, our time estimates support the hypothesis of Chambers et al. (2001) that Cyanoramphus must have colonized New Zealand via over-water dispersal. New Zealand was long separated from Australia when the diversification of Platycercini started in the middle Miocene (Trewick & Gibb, 2010) and possible island chains which could have facilitated dispersal from New Caledonia to New Zealand in the early and middle Miocene (Herzer et al., 1997; Chambers et al., 2001) predated the split between Eunymphicus and Cyanoramphus. It has to be taken into account that current endemic lineages may not have evolved in the area where they are found today and may have been more widely distributed in the past (Goldberg et al., 2011). This can influence the reconstruction of colonization patterns especially among island groups. Nevertheless, our data clearly show, that the Melanesian islands Fiji, Tonga and New Caledonia provided stepping stones for parrots in their colonization of the ‘mini continent’ New Zealand. Colonization of islands from continents or larger landmasses have long been considered as one-way journeys with islands representing evolutionary dead ends whose diversity is balanced between colonization, speciation and extinction (cf. Bellemain & Ricklefs, 2008). However, in congruence with our results, several recent studies have provided evidence for ‘reverse colonization’ from oceanic islands to continents and larger landmasses suggesting that islands may contribute to the build-up of biota on larger landmasses (Filardi & Moyle, 2005; Bellemain & Ricklefs, 2008; Jonsson et al., 2010).

Conclusions The platycercine parrots seem to have their center of origin in the Bassian biota and their subsequent diversification was shaped by the aridification of Australia from mid- Miocene onwards and by long-distance colonization of islands in the Pacific. The independent derivation of Eyrean lineages from Bassian ancestors did not result in a radiation within the arid zone. Ecological speciation caused by adaptation to the precursors of modern arid environments may have played a role for the Eyrean taxa Northiella haematogaster which split from its closest relatives when the aridification of Australia started. The other Eyrean taxa apparently adapted to dry environments after they had become separate evolutionary lineages either in situ (Neopsephotus bourkii) or probably by colonization of arid habitats from Bassian biota without subsequent speciation (Psephotus varius, Barnardius). The diversification events within non-Eyrean platycercine parrots seem to have been primarily caused by vicariance due to habitat fragmentation on continental Australia following the spread of arid environments or by allopatric speciation after the colonization of islands in the Pacific. The lack of temporal congruence between potential vicariance, but also between colonization events implies a dynamic history with varying influences on different groups. New Zealand and the South Pacific were colonized by platycercine parrots not directly from Australia, but via dispersal over several small Melanesian islands acting as stepping stones. Thus, contrary to traditional biogeographic paradigms, we provide further evidence that the colonization of oceanic islands may not just be considered as one-way journeys.

152 Chapter 6

Figure 4 Biogeographic reconstruction of the area splits at the different nodes within Platycercini. Only the most likely splits are shown and node numbers refer to Table 5. The different biogeographic provinces are represented in different colors. The beginning of the aridification is indicated on the timescale at the bottom of the tree. Numbers in black circles illustrate hypothesized vicariance and long-distance dispersal events. 1: Torresian/Bassian divergence. 2: east/west Bassian divergence. 3: divergence across Carpentarian faunal barrier. 4: colonization of Melanesia from Bassian region. 5: colonization of New Zealand from Melanesia.

153 Evolutionary diversification and biogeography of parrots

Acknowledgements We especially thank the Silva Casa Foundation for financial support of this work. We are grateful to S. Birks (University of Washington, Burke Museum), R. Burkhard, A.T. Peterson, R.G. Moyle, M.B. Robbins and A. Nyari (The University of Kansas, Natural History Museum and Biodiversity Research Center) as well as P. Sandmeier for kindly providing us with tissue samples. We further thank the following people for valuable support: S. Bachofner, B. Blöchlinger, L. Joseph, M. Rieger, O. Seehausen, C. Sherry and T. F. Wright.

Literature cited

Akaike, H. (1974) New look at statistical-model identification. Ieee Transactions on Automatic Control, AC19, 716-723 Bellemain, E. & Ricklefs, R.E. (2008) Are islands the end of the colonization road? Trends in Ecology & Evolution, 23, 461-468 Boon, W.M., Daugherty, C.H. & Chambers, G.K. (2001) The Norfolk Island green parrot and New Caledonian red-crowned parakeet are distinct species. Emu, 101, 113-121 Boon, W.M., Kearvell, J.C., Daugherty, C.H. & Chambers, G.K. (2000) Molecular systematics of New Zealand Cyanoramphus parakeets: conservation of orange-fronted and forbes' parakeets. Bird Conservation International, 10, 211-239 Bowman, D.M.J.S., Brown, G.K., Braby, M.F., Brown, J.R., Cook, L.G., Crisp, M.D., Ford, F., Haberle, S., Hughes, J., Isagi, Y., Joseph, L., Mcbride, J., Nelson, G. & Ladiges, P.Y. (2009) Biogeography of the Australian monsoon tropics. Journal of Biogeography, 37, 201-216 Brown, J.M. & Lemmon, A.R. (2007) The importance of data partitioning and the utility of Bayes Factors in Bayesian phylogenetics. Systematic Biology, 56, 643-655 Byrne, M., Steane, D.A., Joseph, L., Yeates, D.K., Jordan, G.J., Crayn, D., Aplin, K., Cantrill, D.J., Cook, L.G., Crisp, M.D., Keogh, J.S., Melville, J., Moritz, C., Porch, N., Sniderman, J.M.K., Sunnucks, P. & Weston, P.H. (in press) Decline of a biome: evolution, contraction, fragmentation, extinction and invasion of the Australian mesic zone biota. Journal of Biogeography Byrne, M., Yeates, D.K., Joseph, L., Kearney, M., Bowler, J., Williams, M.A.J., Cooper, S., Donnellan, S.C., Keogh, J.S., Leys, R., Melville, J., Murphy, D.J., Porch, N. & Wyrwoll, K.H. (2008) Birth of a biome: insights into the assembly and maintenance of the Australian arid zone biota. Molecular Ecology, 17, 4398-4417 Cain, A.J. (1955) A revision of Trichoglossus haematodus and of the Australian platycercine parrots. Ibis, 97, 432-479 Chambers, G.K., Boon, W.M., Buckley, T.R. & Hitchmough, R.A. (2001) Using molecular methods to understand the Gondwanan affinities of the New Zealand biota: three case studies. Australian Journal of Botany, 49, 377-387 Christidis, L. & Norman, J.A. (2010) Evolution of the Australasian songbird fauna. Emu, 110, 21-31 Christidis, L., Schodde, R., Shaw, D.D. & Maynes, S.F. (1991) Relationships among the Australo-Papuan parrots, lorikeets, and cockatoos (Aves, Psittaciformes) - protein evidence. Condor, 93, 302-317

154 Chapter 6

Chubb, A.L. (2004) New nuclear evidence for the oldest divergence among neognath birds: the phylogenetic utility of Zenk (i). Molecular Phylogenetics and Evolution, 30, 140-151 Collar, N.J. (1998) Family Psittacidae (parrots). Handbook of the birds of the world, vol. 4 (ed. by J. Del Hoyo, A. Elliott and J. Sargatal), pp. 280-477. Lynx Edicions, Barcelona. Cooper, A. & Penny, D. (1997) Mass survival of birds across the Cretaceous-Tertiary boundary: Molecular evidence. Science, 275, 1109-1113 Cracraft, J. (1986) Origin and evolution of continental biotas - speciation and historical congruence within the Australian avifauna. Evolution, 40, 977-996 De Kloet, R.S. & De Kloet, S.R. (2005) The evolution of the spindlin gene in birds: sequence analysis of an intron of the spindlin W and Z gene reveals four major divisions of the Psittaciformes. Molecular Phylogenetics and Evolution, 36, 706-721 Drummond, A.J. & Rambaut, A. (2007) Beast: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary Biology, 7 Filardi, C.E. & Moyle, R.G. (2005) Single origin of a pan-Pacific bird group and upstream colonization of Australasia. Nature, 438, 216-219 Finsch, O. (1867) Die Ppapageien. E. J. Brill, Leiden. Fleming, C.A. (1980) The geological history of New Zealand and its life. Auckland University Press, Auckland. Forshaw, J.M. (1981) Australian parrots, Second revised edition edn. Dominion Press, North Blackburn. Goldberg, J., Trewick, S.A. & Powlesland, R.G. (2011) Population structure and biogeography of Hemiphaga pigeons (Aves: Columbidae) on islands in the New Zealand region. Journal of Biogeography, 38, 285-298 Groth, J.G. & Barrowclough, G.F. (1999) Basal divergences in birds and the phylogenetic utility of the nuclear RAG-1 gene. Molecular Phylogenetics and Evolution, 12, 115-123 Hall, R. (2002) Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based reconstructions, model and animations. Journal of Asian Earth Sciences, 20, 353-431 Hall, T.A. (1999) Bioedit, a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series, 41, 95-98 Heads, M. (2010) Biogeographical affinities of the New Caledonian biota: a puzzle with 24 pieces. Journal of Biogeography, 37, 1179-1201 Herzer, R.H., Chaproniere, G.C.H., Edwards, A.R., Hollis, C.J., Pelletier, B., Raine, J.I., Scott, G.H., Stagpoole, V., Strong, C.P., Symonds, P., Wilson, G.J. & Zhu, H. (1997) Seismic stratigraphy and structural history of the Reinga Basin and its margins, southern Norfolk Ridge system. New Zealand Journal of Geology and Geophysics, 40, 425-451 Hillis, D.M. & Bull, J.J. (1993) An empirical-test of bootstrapping as a method for assessing confidence in phylogenetic analysis. Systematic Biology, 42, 182- 192 Homberger, D.G. (1980) Funktionell- morphologische Untersuchungen zur Radiation der Ernährungs- und Trinkmethoden der Papageien (Psittaci). Bonn. Zool. Monogr., 13, 1-92

155 Evolutionary diversification and biogeography of parrots

Hopper, S.D. & Gioia, P. (2004) The Southwest Australian floristic region: evolution and conservation of a global hot spot of biodiversity. Annual Review of Ecology Evolution and Systematics, 35, 623-650 Huelsenbeck, J.P. & Ronquist, F. (2001) MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics, 17, 754-755 Irestedt, M., Fuchs, J., Jonsson, K.A., Ohlson, J.I., Pasquet, E. & Ericson, P.G.P. (2008) The systematic affinity of the enigmatic Lamprolia victoriae (Aves: Passeriformes) - an example of avian dispersal between New Guinea and Fiji over Miocene intermittent land bridges? Molecular Phylogenetics and Evolution, 48, 1218-1222 Jacobs, B.F., Kingston, J.D. & Jacobs, L.L. (1999) The origin of grass-dominated ecosystems. Annals of the Missouri Botanical Garden, 86, 590-643 Jennings, W.B. & Edwards, S.V. (2005) Speciational history of Australian grass finches (Poephila) inferred from thirty gene trees. Evolution, 59, 2033-2047 Jonsson, K.A., Bowie, R.C.K., Moyle, R.G., Christidis, L., Norman, J.A., Benz, B.W. & Fjeldsa, J. (2010) Historical biogeography of an Indo-Pacific passerine bird family (Pachycephalidae): different colonization patterns in the Indonesian and Melanesian archipelagos. Journal of Biogeography, 37, 245-257 Joseph, L., Dolman, G., Donnellan, S., Saint, K.M., Berg, M.L. & Bennett, A.T.D. (2008) Where and when does a ring start and end? Testing the ring-species hypothesis in a species complex of Australian parrots. Proceedings of the Royal Society B-Biological Sciences, 275, 2431-2440 Joseph, L., Toon, A., Schirtzinger, E.E. & Wright, T.F. (2011) Molecular systematics of two enigmatic genera Psittacella and Pezoporus illuminate the ecological radiation of Australo-Papuan parrots (Aves: Psittaciformes). Molecular Phylogenetics and Evolution, 59, 675-684 Joseph, L. & Wilke, T. (2006) Molecular resolution of population history, systematics and historical biogeography of the australian ringneck parrots Barnardius: are we there yet? Emu, 106, 49-62 Kass, R.E. & Raftery, A.E. (1995) Bayes Factors. Journal of the American Statistical Association, 90, 773-795 Kershaw, A.P., Martin, H.A. & Mcewen-Mason, J.C. (1994) The Neogene: a time of transition. History of the Australian vegetation: Cretaceous to recent. (ed. by R.S. Hill), pp. 299-327. Cambridge University Press, Cambridge Lee, J.Y. & Edwards, S.V. (2008) Divergence across Australia's Carpentarian barrier: statistical phylogeography of the red-backed fairy wren (Malurus melanocephalus). Evolution, 62, 3117-3134 Martin, H.A. (2006) Cenozoic climatic change and the development of the arid vegetation in Australia. Journal of Arid Environments, 66, 533-563 Mayr, E. (1939) The origin and history of the bird fauna of Polynesia. Pac. Sci. Congr. Proc, 6, 197-216 Mayr, E. & Diamond, J. (2001) The birds of Northern Melanesia. Speciation, ecology & biogeography. Oxford University Press, Oxford. Mayr, G. (2009) Paleogene fossil birds. Springer-Verlag, Berlin Heidelberg. Mayr, G. (2010) Parrot interrelationships - morphology and the new molecular phylogenies. Emu, 110, 348-357 Mcdowall, R.M. (2008) Process and pattern in the biogeography of New Zealand - a global microcosm? Journal of Biogeography, 35, 197-212

156 Chapter 6

Merrick, J.R., Archer, M., Hickey, G.M. & Lee, M.S.Y. (2006) Perspectives. Evolution and biogeography of Australasian vertebrates. (ed. by J.R. Merrick, M. Archer, G.M. Hickey and M.S.Y. Lee). Auscipub, Sidney. Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G.A.B. & Kent, J. (2000) Biodiversity hotspots for conservation priorities. Nature, 403, 853-858 Nylander, J.A.A. (2004) MrModeltest v2. Evolutionary Biology Centre, Uppsala University Uppsala, Sweden Nylander, J.A.A., Wilgenbusch, J.C., Warren, D.L. & Swofford, D.L. (2008) AWTY (are we there yet?): a system for graphical exploration of MCMC convergence in Bayesian phylogenetics. Bioinformatics, 24, 581-583 Overton, L.C. & Rhoads, D.D. (2004) Molecular phylogenetic relationships based on mitochondrial and nuclear gene sequences for the todies (Todus, Todidae) of the Caribbean. Molecular Phylogenetics and Evolution, 32, 524-538 Pacheco, M.A., Battistuzzi, F.U., Lentino, M., Aguilar, R., Kumar, S. & Escalante, A.A. (2011) Evolution of modern birds revealed by mitogenomics: timing the radiation and origin of major orders. Molecular Biology and Evolution, 28, 1927-1942 Pratt, R.C., Gibb, G.C., Morgan-Richards, M., Phillips, M.J., Hendy, M.D. & Penny, D. (2009) Toward resolving deep Neoaves phylogeny: data, signal enhancement, and priors. Molecular Biology and Evolution, 26, 313-326 Rambaut, A. (2008) Figtree 1.2., published by the author. Rambaut, A. & Drummond, A.J. (2007) Tracer v1.5, available from http://beast.Bio.Ed.Ac.Uk/tracer. Ree, R.H., Moore, B.R., Webb, C.O. & Donoghue, M.J. (2005) A likelihood framework for inferring the evolution of geographic range on phylogenetic trees. Evolution, 59, 2299-2311 Ree, R.H. & Sanmartin, I. (2009) Prospects and challenges for parametric models in historical biogeographical inference. Journal of Biogeography, 36, 1211-1220 Ree, R.H. & Smith, S.A. (2008) Maximum likelihood inference of geographic range evolution by dispersal, local extinction, and cladogenesis. Systematic Biology, 57, 4-14 Rinke, D. (1989) The relationships and taxonomy of the Fijian parrot genus Prosopeia. Bull. B. O. C., 109, 185-195 Ronquist, F. & Huelsenbeck, J.P. (2003) MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics, 19, 1572-1574 Rowley, I. (1998) Family Cacatuidae (cockatoos). Handbook of the birds of the world (ed. by J. Del Hoyo, A. Elliott and J. Sargatal), pp. 246-279. Lynx Edicions, Barcelona. Salvadori, T. (1891) Catalogue of the Psittaci, or parrots, in the collection of the British Museum, vol. 20. Br. Mus. (Nat. Hist.), London. Schodde, R. (2006) Australia's bird fauna today - origins and development. Evolution and biogeography of Australasian vertebrates. (ed. by J.R. Merrick, M. Archer, G.M. Hickey and M.S.Y. Lee). Auscipub, Sidney. Schweizer, M., Seehausen, O., Güntert, M. & Hertwig, S.T. (2010) The evolutionary diversification of parrots supports a taxon pulse model with multiple trans- oceanic dispersal events and local radiations. Molecular Phylogenetics and Evolution, 54, 984-994 Schweizer, M., Seehausen, O. & Hertwig, S.T. (in press) Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations. Journal of Biogeography

157 Evolutionary diversification and biogeography of parrots

Stamatakis, A. (2006) RAxML-vi-hpc: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics, 22, 2688- 2690 Stamatakis, A., Hoover, P. & Rougemont, J. (2008) A rapid bootstrap algorithm for the RAxML web servers. Systematic Biology, 57, 758-771 Swafford, D.L. (2001) PAUP*: phylogenetic analysis using parsimony (*and other methods), version 4.0. Sinauer Associates, Sunderland, MA Tavares, E.S., Baker, A.J., Pereira, S.L. & Miyaki, C.Y. (2006) Phylogenetic relationships and historical biogeography of Neotropical parrots (Psittaciformes: Psittacidae: Arini) inferred from mitochondrial and nuclear DNA sequences. Systematic Biology, 55, 454-470 Toon, A., Hughes, J.M. & Joseph, L. (2010) Multilocus analysis of honeyeaters (Aves: Meliphagidae) highlights spatio-temporal heterogeneity in the influence of biogeographic barriers in the Australian monsoonal zone. Molecular Ecology, 19, 2980-2994 Trewick, S.A. & Gibb, G.C. (2010) Vicars, tramps and assembly of the New Zealand avifauna: a review of molecular phylogenetic evidence. Ibis, 152, 226-253 Wallace, A.R. (1867) The geographical distribution of animals. Harper, New York. White, N.E., Phillips, M.J., Gilbert, M.T.P., Alfaro-Nunez, A., Willerslev, E., Mawson, P.R., Spencer, P.B.S. & Bunce, M. (2011) The evolutionary history of cockatoos (Aves: Psittaciformes: Cacatuidae). Molecular Phylogenetics and Evolution, 59, 615-622 Wilgenbusch, J.C., Warren, D.L. & Swofford, D.L. (2004) AWTY: a system for graphical exploration of MCMC convergence in Bayesian phylogenetic inference. http://ceb.csit.fsu.edu/awty. Wright, T.F., Schirtzinger, E.E., Matsumoto, T., Eberhard, J.R., Graves, G.R., Sanchez, J.J., Capelli, S., Mueller, H., Scharpegge, J., Chambers, G.K. & Fleischer, R.C. (2008) A multilocus molecular phylogeny of the parrots (Psittaciformes): support for a Gondwanan origin during the Cretaceous. Molecular Biology and Evolution, 25, 2141-2156

158

Chapter 7

PHYLOGENY AND BIOGEOGRAPHY OF THE PARROT GENUS Prioniturus (Aves: Psittaciformes)

Manuel Schweizera*, Marcel Günterta, Stefan T. Hertwiga

aNaturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, CH 3005 Bern, Switzerland

*Corresponding author: Naturhistorisches Museum der Burgergemeinde Bern, Bernastrasse 15, CH 3005 Bern, Switzerland, Telephone: +41 31 350 72 95, Fax: +41 31 350 74 99, E- mail: [email protected]

Submitted to: Journal of Zoological Systematics and Evolutionary Research

161 Evolutionary diversification and biogeography of parrots

Abstract

The parrot genus Prioniturus occurs in the oceanic Philippines, Palawan and Wallacea, a geologically dynamic region with a complex history of land and sea. The described taxa of Prioniturus have been variously placed in different assemblages, and different numbers of species have been recognized. However, a phylogenetic framework is so far lacking. This would be the prerequisite to reconstructing dispersal and colonization patterns of Prioniturus across and within Wallacea and the Philippines. Following our robustly supported phylogenetic hypothesis based on two mitochondrial genes, we propose to treat P. mindorensis comb. nov. as well as P. montanus and P. waterstradti as separate species. In P. d. discurus and P. discurus whiteheadi further studies using additional data and specimens are necessary to clarify their taxonomic status. This result is congruent with other studies demonstrating that alpha diversity of the Philippine avifauna is strongly underestimated. According to our biogeographic reconstruction, Prioniturus has diversified by a complex combination of colonization of islands and subsequent divergence in allopatry among and within island groups. Dispersal between Sulawesi/Wallacea and the Philippines occurred twice and documents a rare case of faunal exchange between these two regions.

Keywords: Aves: Psittaciformes: Prioniturus, biogeography, colonization, dispersal, mitochondrial genes, molecular phylogeny

162 Chapter 7

Introduction

Insular biota have played an important role in the formulation of evolutionary, biogeographic and ecological theory in the past (MacArthur and Wilson, 1967, Wallace, 1867, Darwin, 1859). However, understanding their evolutionary history is still a major issue and a prerequisite to explaining biogeographic and macroecological patterns in a broader context (Bellemain and Ricklefs, 2008, Jonsson et al., 2010a, Filardi and Moyle, 2005). The insular part of Southeast Asia is usually divided into the three biogeographic zones of Sundaland (Malay Peninsula, Borneo, Java, Sumatra and Palawan), the Philippines and Wallacea (Esselstyn et al., 2009). All three regions harbour an exceptional concentration of endemic species and are considered to be biodiversity hotspots (Myers et al., 2000). Wallacea is situated between Wallace’s line to the west and Lydekker’s line to the east. It is today dominated by Sulawesi, but also includes the Moluccas and Lesser Sunda Islands (Moss and Wilson, 1998). Wallacea has played an important role in the biotic exchange between the Australasian and the Indo-Malayan regions. The Australian continent approached Asia from the south during the Paleogene and reached its present position around 20-25 Ma (million years ago) (Hall, 2002, Li and Powell, 2001). Then, the proximity of these two landmasses, in combination with emergent land and squeezed island arcs serving as stepping stones, facilitated biotic exchange between Australasia and the Indo-Malayan region over Wallacea (Jonsson et al., 2008, Moss and Wilson, 1998). While the fauna and flora of Sundaland are still clearly of Asian origin, Wallacea harbours a mix of Asian and Australian fauna and flora and can be considered as a transitional area between these biota (Moss and Wilson, 1998). Sulawesi, as a major part of Wallacea, probably consisted in the past of distinct islands that have been recently unified. These may never have been connected to continental landmasses, and the biota of Sulawesi are most probably the result of over-water colonization (Hall, 1998, Hall, 2002, Hall, 2001, Esselstyn et al., 2009). The Philippine Islands are situated to the north of Sulawesi on the other side of the Wallace’s line. They consist of about 7000 islands and lie close to a continental shelf on one side, but are surrounded by deep oceans on all other sides. They are characterized as a fringing archipelago which lies adjacent to a continental shelf but has never been connected to it (Jones and Kennedy, 2008). The Philippine archipelago as it is known today formed a cohesive region only from the late Miocene onwards (Hall, 2002). Its fauna can be considered as being intermediate between that of an oceanic island and land-bridge islands (Diamond and Gilpin, 1983), and Dickerson (1928) proposed four different colonization routes of the Philippines: 1) dispersal (over a submerged land bridge) from present-day Taiwan (northern route), 2) across Palawan or 3) the Sulu archipelago from Borneo, and 4) dispersal (over a submerged land bridge) from Sulawesi (southern route). While only a few examples suggest a northern or southern colonization route, an important contribution to the Philippine fauna seems to stem from colonization events from Sundaland with a significant contribution from Borneo across the Sulu archipelago (e.g. Jones and Kennedy, 2008, Esselstyn et al., 2009, Evans et al., 2003, Esselstyn and Oliveros, 2010, Oliveros and Moyle, 2010). The islands of Palawan, the Calamianes group and Balabac are politically part of the Philippines; however, they were likely connected to Borneo in the past and are biologically best considered as a transition zone between the Sunda Shelf and the oceanic Philippines (Hall, 1998, Jones and Kennedy, 2008, Diamond and Gilpin, 1983, Esselstyn et al., 2010). Parrots originated in Australasia and have colonized the Indo-Malayan region several times independently from there, mainly after these two regions came into contact (Schweizer et al., 2010, Schweizer et al., in press). However, the exact pattern of colonization of Sundaland, the Philippines and Wallacea could not so far be unambiguously inferred. While members of 11 parrot genera occur in Wallacea (one Cacatuidae, five Loriinae, five Psittacinae each), only three colonized Sundaland and mainland Asia (Psittacinae), and six are distributed in the

163 Evolutionary diversification and biogeography of parrots

Philippines today (one Cacatuidae, one Loriinae, four Psittacinae) (Collar, 1998, Rowley, 1998). Of these, Bolbospittacus is a monotypic genus endemic to this island group, while Prioniturus and Tanygnathus are chiefly restricted to Wallacea and the Philippines. Additionally, Prioniturus platenae also occurs on Palawan, and Tanygnathus lucionensis on Palawan and Malaysian islands off NE Borneo. Prioniturus and Tanygnathus can be considered examples of the few bird genera that occur in the Philippines but not in the main part of Sundaland (cf. Diamond and Gilpin, 1983). While four species of Tanygnathus are currently recognized, Prioniturus is more species-rich and consists of eight to nine different species (Collar, 1998, Dickinson, 2003). As implied by the English name racket-tailed parrots, Prioniturus is characterized by elongated central tail-feathers formed by bare shafts widening to spatulate tips (Forshaw, 1973, Smith, 1975). It is traditionally included in the Psittaculini (Collar, 1998, Smith, 1975) and was found to be the sister group of an assemblage consisting of the genera Eclectus, Geoffroyus Psittacula, Psittinus and Tanygnathus, by recent molecular phylogenetic work (Schweizer et al., 2010, Wright et al., 2008, Schweizer et al., in press). While one of the 13 to 14 recognized taxa of Prioniturus occurs on Palawan and five are restricted to Wallacea, the majority (seven to eight) occur in the oceanic Philippines. The different taxa of Prioniturus have been variously arranged in different groups or subgroups with various numbers of species being recognized (Forshaw, 1973, Collar, 1998, Juniper and Parr, 1998, Forshaw, 2006, Peters, 1937, Dickinson, 2003) (Table 1). Three species of Prioniturus are considered to be threatened and a further three are categorized as near threatened (BirdLifeInternational, 2011). Conservation strategies usually focus on the preservation of distinct species (Mace, 2004). Given the discordant taxonomic history of species recognition within Prioniturus, an accurate species-level taxonomy based on a robust phylogenetic hypothesis is needed for conservation planning. This may be especially important for Philippine bird species as the alpha diversity on this archipelago currently seems highly underestimated (Oliveros and Moyle, 2010, Lohman et al., 2010, Peterson, 2006). The aim of the present study was thus to establish a phylogenetic hypothesis of Prioniturus using genetic sequence data and to test if traditional groupings based on morphology are confirmed by molecular data. On the basis of this phylogenetic hypothesis, we reconstructed dispersal and colonization patterns of Prioniturus across and within Wallacea and the Philippines. Moreover, we tested whether the different radiations in the Philippines and Sulawesi/Wallacea represent monophyletic clades, implying single colonization events followed by in situ speciation, or if they are the results of multiple colonization events with subsequent allopatric speciation.

Material and Methods

Sampling and laboratory methods We obtained samples from all Prioniturus taxa, with the exception of Prioniturus platurus sinerubris from Taliabu in the Sula Islands (Table 2). Taxonomy follows Collar (1998, cf. Table 1). Moreover, we included samples of 27 old world parrot species including all genera belonging to the Psittaculini with the exception of Alisterus (Table 3). Nestor notabilis was chosen as outgroup according to the results of several molecular studies (Schweizer et al., 2010, Wright et al., 2008, de Kloet and de Kloet, 2005, Schweizer et al., in press). Total genomic DNA of frozen tissue or blood samples was isolated using peqGOLD tissue DNA Mini Kit (Peqlab), DNeasy Blood & Tissue Kit (Qiagen) or Wizard® SV Genomic DNA Purification System (Promega) following the manufacturer’s instructions. For the isolation of DNA from toe pads sampled from skins, sbeadex® forensic kit (LGC Genomics) was used. DNA extraction and the preparation of PCR reactions for toe pad samples took place in a

164 Chapter 7 separate lab to avoid potential contamination. Moreover, laboratory equipment used for preparing these samples was treated with DNA AwayTM (Molecular BioProducts) solution and irradiated with ultraviolet light prior to use. Partial sequences of the two mitochondrial genes ND2 (NADH2) and Cytb (Cytochrome b) were amplified with Polymerase chain reaction (PCR) using published primer sequences for the analyses of tissue and blood samples (Table 4). For DNA extracted from toe pad samples, we amplified three (Cytb) or four (ND2) overlapping fragments using modified published primers or primers specially designed for this study (Table 4). Negative controls were used to check for potential contaminations. PCR reaction volumes were 25 Āl containing 12.5Āl GoTaq® Hot Start Green Master Mix (Promega), PCR-Master-Mix Y or HotStart-Master-Mix (Peqlab), 2-3Āl genomic DNA, 2Āl of each primer with a concentration of 10ĀM and 5.5-6.5 Āl ddH2O. PCR was performed on a Techne TC-512 thermo-cycler. For ND2, we applied the PCR protocol of Tavares et al. (2006) with the annealing temperature set to 53° for the whole fragment and to 56°C for the smaller overlapping fragments. Amplifications of Cytb were performed with an initial denaturation of 94°C for 3 min followed by 35 cycles of denaturation at 95°C for 30 s, annealing for 30 s at 54.5°C for the whole fragment and at 57°C for the smaller overlapping fragments, and extension at 72°C for 1 min, with a final extension at 72°C for 7 min. PCR products were examined by gel electrophoresis to confirm amplification of the target fragment. They were then excised from gels and cleaned using the Wizard® SV Gel and PCR Clean-UP System (Promega). To increase the quantity of DNA, the products of two independent PCR runs were put together before the cleaning. Sequencing was carried out by Microsynth AG (Balgach, Switzerland) or by Macrogen Inc. (Seoul, Korea). We checked individual sequences by searching for apparent stop codons after the translation of sequences into amino acids.

Table 1. Overview of some different taxonomic treatments of Prioniturus.

Peters (1937) Collar (1998) Dickinson (2003) Forshaw (2006)

Prioniturus flavicans montanus Prioniturus montanus Prioniturus montanus montanus Prioniturus montanus Prioniturus discurus Prioniturus montanus Prioniturus waterstradti malindangensis malindangensis malindangensis Prioniturus discurus waterstradti Prioniturus waterstradti Prioniturus montanus waterstradti Prioniturus waterstradti waterstradti Prioniturus discurus platenae Prioniturus platenae Prioniturus platenae Prioniturus platenae Prioniturus luconensis Prioniturus luconensis Prioniturus luconensis Prioniturus luconensis Prioniturus discurus discurus Prioniturus discurus discurus Prioniturus discurus discurus Prioniturus discurus discurus Prioniturus discurus suluënsis Prioniturus discurus whiteheadi Prioniturus discurus whiteheadi Prioniturus discurus whiteheadi Prioniturus discurus mindorensis Prioniturus discurus mindorensis Prioniturus discurus mindorensis Prioniturus discurus mindorensis Prioniturus flavicans verticalis Prioniturus verticalis Prioniturus verticalis Prioniturus verticalis Prioniturus flavicans flavicans Prioniturus flavicans Prioniturus flavicans Prioniturus flavicans Prioniturus platurus platurus Prioniturus platurus platurus Prioniturus platurus platurus Prioniturus platurus platurus Prioniturus platurus sinerubris Prioniturus platurus sinerubris Prioniturus platurus sinerubris Prioniturus platurus talautensis Prioniturus platurus talautensis Prioniturus platurus talautensis Prioniturus platurus talautensis Prioniturus mada Prioniturus mada Prioniturus mada Prioniturus mada

165 Evolutionary diversification and biogeography of parrots

Table 2. Prioniturus taxa sampled, collection locality, collection number and GenBank Accession numbers for the two genes analyzed in this study. (xxx: GenBank Accession numbers will be requested by the publication of this chapter).

GenBank accession number

Taxon Locality Collection number Cytb ND2 Sample type

Prioniturus discurus discurus Philippines: Mindanao, Bukidnon, Cabanglasan ZMUC136355 xxx xxx toe pad Prioniturus discurus discurus Philippines: Mindanao, Zamboanga, Curuan, Sapamoro farm ZMUC136358 xxx xxx toe pad Prioniturus discurus discurus Philippines: Mindanao, Agusan, Agay ZMUC1357 xxx xxx toe pad Prioniturus discurus discurus Philippines: Mindanao, Agusan, Lake Mainit ZMUC136356 xxx xxx toe pad Prioniturus discurus mindorensis Philippines: Mindoro AMNH620956 xxx xxx toe pad Prioniturus discurus mindorensis Philippines: Mindoro, Ilong Peak, Halcon ZMUC22 12 1955 128 xxx xxx toe pad Prioniturus discurus whiteheadi Philippines: Sibuyan, Romblon Prov, Romblon Prov, Mt Guitinguitin FMNH344943 xxx xxx tissue Prioniturus discurus whiteheadi Philippines: Samar, Muncipality of Taft, Brgy San Rafael KU14237 xxx xxx tissue Prioniturus discurus whiteheadi Philippines: Samar, Paranas BMNH1897 5 13 502 xxx xxx toe pad Prioniturus discurus whiteheadi Philippines: Pagyabunan, Basis Negros Oriental ZMUC22 12 1955 130 xxx xxx toe pad Prioniturus flavicans Sulawesi: between Menado & Arakan AMNH620995 xxx xxx toe pad Prioniturus luconensis Philippines: Luzon NMBE1056247 xxx xxx tissue Prioniturus mada Buru: Wa Fehat AMNH621031 xxx xxx toe pad Prioniturus montanus Philippines: Luzon, Nueva Vizcaya, Quezon Municipality FMNH454964 xxx xxx tissue Prioniturus waterstradti Philippines: Mindanao, Bukidnon, Mt Kitanglad Range FMNH357417 xxx xxx tissue Prioniturus waterstradti Philippines: Bukidnon, Mindanao, Kaatoan, Katanglad Volcano ZMUC1 10 1953 1295 xxx xxx toe pad Prioniturus waterstradti Philippines: Bukidnon, Mindanao, Kaatoan, Katanglad Volcano ZMUC1 10 1953 1297 xxx xxx toe pad Prioniturus waterstradti Philippines: Mindanao Gingoog City, Brgy Lumotan, Sitio San Isidro, Mt Balatukan KU19120 xxx xxx tissue Prioniturus platenae Philippines: Palawan, Iwahig BMNH1911 11 16 112 xxx xxx toe pad Prioniturus platurus platurus Sulawesi: Indrulaman AMNH621028 xxx xxx toe pad Prioniturus platurus platurus Sulawesi: Butung BMNH1926 12 3 1 xxx xxx toe pad Prioniturus platurus talautensis Talaud: River Essang, Karakelang ZMUC146427 xxx xxx blood Prioniturus platurus talautensis Talaud: Liung Islans AMNH621000 xxx xxx toe pad Prioniturus verticalis Philippines: Tawi Tawi AMNH620985 xxx xxx toe pad Prioniturus verticalis Philippines: Jarawakam, Tawi Tawi Island ZMUC26 6 1976 27 xxx xxx toe pad Prioniturus verticalis Philippines: Lapid Lapid, Tawi Tawi Island ZMUC26 6 1976 28 xxx xxx toe pad

Table 3. Species sampled outside Prioniturus, museum and collection number, GenBank Accession numbers for the two genes analyzed in this study. (xxx: GenBank Accession numbers will be requested by the publication of this chapter).

GenBank accession number

Species Museum/Collection Collection nb ND2 Cytb

Agapornis roseicollis NMBE (except ND2) 1056204 EU327596 xxx Amazona xanthops AY669485 AY669441 Aprosmictus erythropterus EU327599 AB177959 Barnardius zonarius NMBE 1056210 xxx xxx Cacatua sulphurea EU327605 AB177976 Charmosyna pulchella NMBE 1056241 xxx xxx Cyanoramphus novaezelandiae NMBE 1056220 xxx xxx Eclectus roratus NMBE (except ND2) 1056248 EU327619 xxx Eos cyanogenia NMBE 1056237 xxx xxx Geoffroyus heteroclitus EU327626 xxx Lathamus discolor NMBE 1056219 xxx xxx Loriculus galgulus UWBM (except ND2) 73841/2002-006 EU327631 xxx Lorius garrulus NMBE 1056240 xxx xxx Melopsittacus undulatus UWBM (except ND2) 60748/1998-068 EU327633 xxx Micropsitta finschii tristrami UWBM (except ND2) 66040/2000-022 EU327634 xxx Neophema splendida NMBE 1056225 xxx xxx Neopsephotos bourkii UWBM (except ND2) 57542/1996-109 EU327639 xxx Nestor notabilis NMBE (except ND2) 1056242 EU327641 xxx Platycercus caledonicus NMBE (except ND2) 1056212 EU407679 xxx Polytelis alexandrae NMBE (except ND2) 1056209 EU327653 xxx Prosopeia tabuensis NMBE (except ND2) 1056252 EU327656 xxx Psephotus varius NMBE 1056217 xxx xxx Psittacella brehmii KU 12931/91954 xxx xxx Psittacula columboides EU327659 AY220108 Psitteuteles goldiei NMBE (except Cytb) 1056239 xxx AF346343 Psittinus cyanurus NMBE 1056251 xxx xxx Psittrichas fulgidus EU327662 U89177 Tanygnatgus megalorhynchus NMBE 1056249 xxx xxx

166 Chapter 7

Table 4. Primers used for the amplification of the two genes analyzed in this study.

Gene Primer name Source

ND2 MetL 5’-GGCCCATACCCCGAAAATGA-3’ (Tavares et al 2006) ASNH 5’-GATCRAGGCCCATCTGTCTAG-3’ (Tavares et al 2006) L5394 PRI1 5’-CTCAAAATCCCACCACCCACGAGC-3’ L5574 PRI 5’-TTCCACTTYTGATTCCCAGAAGTCCT-3’ L5745 PRI 5’-AACCAAACACAAACCCGAAAAAT-3’ L5932 PRI 5’-TAACCTCATGAACCAAAACCCCCA-3’ H5636 PRI 5’-GKAGTTTTATTRCTGTTGATAGGAG-3’ H5823 PRI1 5’-TTCARCCTAGGTGKGAGATGGATGA-3’ H6005 PRI1 5’-CCTGTTARTGGGGGGAGGCCTGCTA-3’ H6180 PRI 5’-GTGTGATTGTTGARCAGTATGCTAG-3’ Cytb L15132 5’-CGAACCGTACAATACGGATGGYTAATC-3’ (Boon et al 2000) H16065 5’-TCATCTCCGGTTTACAAGAC-3’ (Boon et al 2000) CBL15298PRI 5’-TGAGGYCAAATATCATTYTGAGGRGC-3’ CBL15462PRI 5’-YACMRGCCTARTYCTCGTCC-3’ PriL1 5’-CACTTCCTYCTCCCATTCATAA-3’ PriL1 3 5’-TTCCTYCTCCCATTYATAACACCAG-3’ L15643 PRI 5’-GCCCTATTCTCACCAAACCTMCT-3’ L15643 PRI2 5’-GCYCTATTYTCACCAAACCTCCT-3’ CBH15516PRI 5’-TTTGATGARATGCCTAGRGGG-3’ H15706 5’-GGCAAATAGGAARTATCATTC-3’ H15706Pri 5’-GCGAATAGRAARTATCATTCTGGTTT-3’ CBH15902 PRI2 5’-TGGTTGGCTTCCHACTCAKGTTA-3’ CBH15902 PRI 5’-TGGYTGGCTTCCHACYCATGTTA-3’ CBH15902 PRI3 5’-TGGRTGTTCGACTGGTTGGCTTCC-3’

Phylogenetic analyses Bayesian inference (BI) was conducted using MrBayes 3.2 (Huelsenbeck and Ronquist, 2001, Ronquist and Huelsenbeck, 2003). Alternative biologically relevant parameter settings were evaluated for our concatenated data corresponding to separate models with varying base frequencies, rate matrix, shape parameters and proportion of invariable sites for the different genes and/or their codon positions. Models were selected with MrModeltest 2.3 (Nylander, 2004) using the Akaike information criteria (Akaike, 1974). Base frequencies, rate matrix, shape parameter and proportion of invariable sites were allowed to vary between partitions. We performed two independent runs of Metropolis-coupled Markov chain Monte Carlo analyses for each parameter setting and each run consisted of one cold chain and three heated chains with a default temperature of 0.2. The chains were run for 10 million generations with sampling every 100 generations. We checked that the average standard deviation of split frequencies converged towards zero and estimated the length of the ‘burn-in’ by monitoring the change in cumulative split frequencies using AWTY (Nylander et al., 2008, Wilgenbusch et al., 2004) and by visually inspecting trace files with Tracer 1.4.1 (Rambaut and Drummond, 2007). We then discarded the first 25% of samples as burn-in (25,000 trees) well after the chains had reached stationarity. Moreover, we compared likelihoods and posterior probabilities of all splits to assess convergence among the two independent runs using AWTY. We used the Bayes Factor (BF) to evaluate the relevance of the different parameter settings (Kass and Raftery, 1995, Brown and Lemmon, 2007). The harmonic mean calculated by MrBayes was used as an estimation of the marginal likelihood of the data (Kass and Raftery, 1995). A more complicated model (i.e. more partitions) was considered to be more supported than a simpler model if 2lnBF was greater than 10 (Brown and Lemmon, 2007). RAxML 7.0.4 (Stamatakis, 2006) was used for the maximum likelihood (ML) search and the software was run on the web server with 100 rapid bootstrap inferences. All free model parameters were estimated by the software (substitution rates, gamma shape parameter, base frequencies) (Stamatakis et al., 2008) based on the best parameter settings found by MrBayes. Clades were considered to be supported when the clade credibility values of the BI were Ā 0.95 (Huelsenbeck and Ronquist, 2001) and when bootstrap values were Ā 70 in the ML analyses (Hillis and Bull, 1993). Pairwise genetic distances among focal taxa were estimated

167 Evolutionary diversification and biogeography of parrots

with Geneious 5.4 (Drummond et al., 2011) using the models of sequence evolution selected with MrModeltest 2.3.

Biogeography and colonization patterns We chose a maximum likelihood approach to infer colonization patterns and the biogeographic history of Prioniturus based on the maximum clade credibility tree from the analyses with MrBayes. We reconstructed for each node the area assignment that maximizes the probability of arriving at the observed areas in the tip taxa given a stochastic Markov model of evolution (Lewis, 2001), as implemented in Mesquite 2.72 (Maddison and Maddison, 2006, Maddison and Maddison, 2009). All other taxa of Psittaculini and Micropsitta were included as outgroup and the current distribution areas of the different species were categorized in the following discrete character states according to distributional data given in Collar (1998): Australasia, Indo-Malaysia, Palawan, oceanic Philippines (Philippines excluding Palawan and Sulu), Wallacea and the Sulu archipelago. As only one character state can be assigned to a given taxon in this likelihood approach, we assigned Eclectus to the Australasian region, as this area covers most of the distribution of this taxon. Moreover, Tanygnathus was assigned to Wallacea, as most species of this genus occur there. However, an assignment of Tanygnathus to the Philippines had no influence on the final results.

Results

Phylogenetic analyses The final alignment was 1287 bp in length, consisting of 558 bp for cytb and 729 bp for ND2 (Table 5). The translation into amino acids did not reveal any unexpected stop codons and the aligned sequences contained no indels. MrModeltest selected different models of nucleotide evolution for the different partitions of our data set (Table 5). After calculation of the Bayes Factors for the different priors in the Bayesian analyses, the parameter setting using different models of sequence evolution for the two genes and three codon positions was chosen as the best-fitting model. Based on this parameter setting, we reran the analyses with MrBayes for 20 million generations with sampling every 100 generation (Figure 1, 2). The topology of the best-scoring tree obtained by the maximum likelihood analyses with RAxML was identical to the Bayesian maximum clade credibility tree for Prioniturus with the exception of the resolution of clades within P. discurus discurus. However, it differed in topology at five unsupported nodes at the base of the tree outside the Psittaculini-Micropsitta clade (Figure 1).

Table 5. Sequence characteristics and model parameters for the three genes and their different partitions analyzed.

Cytb ND2 comb. data set 1 pos 1&2 pos 2 pos 3 pos 1 pos 1&2 pos 2 pos 3 pos 1 pos 1&2 pos 2 pos 3 pos

Size 558 729 1287 %A 0 3013 0 2550 0 2125 0 2013 0 3688 0 3256 0 3764 0 2749 0 1667 0 4319 0 3456 0 3153 0 2442 0 1810 0 3996 %C 0 4491 0 3561 0 3197 0 2924 0 4913 0 4105 0 3087 0 3397 0 3652 0 4443 0 4376 0 3281 0 3369 0 3343 0 4698 %G 0 0901 0 1950 0 1595 0 1054 0 0286 0 0551 0 1219 0 1025 0 0901 0 0244 0 0541 0 1671 0 1317 0 0976 0 0263 %T 0 1596 0 1939 0 3082 0 4009 0 1114 0 1987 0 1930 0 2829 0 3780 0 0994 0 1627 0 1895 0 2872 0 3871 0 1043 Model GTR+I+G GTR+G GTR+I+G GTR+I GTR+G GTR+I+G GTR+I+G GTR+I+G HKY+I+G GTR+G GTR+I+G GTR+I+G GTR+I+G GTR+I+G GTR+G Ti/Tv ------7 7612 - - - - Prop -0 4574 - 0 6202 0 8074 - 0 1177 0 3517 0 4241 0 4621 - 0 3807 0 4309 0 4791 0 5855 - Alpha 0 8418 0 1734 0 6882 - 1 5504 0 4927 1 1665 0 7939 0 5336 1 8233 0 8083 1 1056 0 6634 0 5582 1 7503

168 Chapter 7

Figure 1. 50% majority-rule consensus tree of the Bayesian inference. Clade credibility values and bootstrap values of the ML likelihood tree are indicated at nodes of all taxa analyzed; Prioniturus taxa are collapsed.

The Psittaculini (including Prioniturus) formed a monophyletic clade together with Micropsitta finschii. Prioniturus was consistently revealed to be monophyletic with robust support. Its sister group was formed by a cluster consisting of the Psittaculini-taxa Eclectus, Tanygnathus, Psittacula, Psittinus and Geoffroyus (Figure 1). The different taxa of Prioniturus were divided into three clades (Figure 2). Clade 1 formed the sister group to the remaining Prioniturus taxa and comprised the Wallacean taxa P. mada and both subspecies of P. platurus analyzed. A clade consisting of both specimens analyzed of P. platurus talautensis was robustly supported, but the cluster of the two individuals of P. platurus platurus gained no support. Clade 2 consisted of oceanic Philippine species and P. platenae from Palawan. P. waterstradti was revealed to be the sister group of a clade consisting of P. montanus, P. discurus mindorensis and P. platenae. The two taxa P. waterstradti and P. montanus which have often been lumped in a single species were not found to be monophyletic. Clade 3 comprised three oceanic Philippine taxa (P. luconensis, P. d. discurus, P. discurus whiteheadi), P. flavicans from Sulawesi as well as P. verticalis from the Sulu archipelago. The latter formed the sister group to the remaining taxa, and P. flavicans was the sister group to the Philippine taxa. P. discurus whiteheadi and P. discurus discurus were each found to form two robustly supported clades; however their relationship as sister taxa was not robustly supported throughout.

169 Evolutionary diversification and biogeography of parrots

Genetic distances between P. discurus whiteheadi and P. d. discurus ranged from 3.7-7-5% (3.4-4.3% uncorrected) for Cytb and from 2.1-3.7% (2-2.4% uncorrected) for ND2. For P. montanus and P. waterstradti, distances ranged from 5.7-6% (3.8-4% uncorrected) for Cytb and from 5.6-5.7% (4.5% uncorrected) for ND2.

Biogeography and colonization patterns Our biogeographic reconstruction revealed that five major dispersal events need to be invoked to explain the current distribution pattern of Prioniturus (Figure 3), although the initial colonization events could not be inferred unambiguously. A last common ancestor of Prioniturus colonized either Wallacea or the oceanic Philippines from Australasia. This was followed by dispersal either from the oceanic Philippines to Wallacea or from Wallacea to the oceanic Philippines leading to the ancestors of clade 1 and clade 2/3 respectively. Two further dispersal events occurred within clade 3, namely the colonization of the Sulu archipelago from the oceanic Philippines by the ancestor of P. verticalis and the colonization of Sulawesi from the oceanic Philippines by the ancestor of P. flavicans. Within clade 2, Palawan was colonized by an ancestor of P. platenae from the oceanic Philippines.

Discussion

Phylogeny and taxonomy Insular biota with many allopatric populations as in the case of Prioniturus, have always presented a challenge for taxonomists, especially when applying a biological species concept (e.g. Winker, 2010, Helbig et al., 2002). Based on morphological features, several authors have proposed a grouping of the Wallacean species P. platurus and P. mada which they termed ‘superspecies’ (e.g. Salomonsen, 1953, Collar, 1998, White and Bruce, 1986). This corresponds to the well supported clade 1 in our analyses, which formed the sister group to the remaining Prioniturus taxa. As an additional ‘superspecies’, Collar (1998) and White & Bruce (1986) proposed a clade including P. montanus, P. waterstradti together with P. platenae and debatably with P. verticalis. The first three were revealed to form a supported clade with the inclusion of P. discurus mindorensis (clade 2). This group is restricted to the oceanic Philippines and Palawan. Another supposed ‘superspecies’ comprising P. luconensis and P. discurus, probably P. flavicans and debatably P. verticalis was proposed by Collar (1998) and White & Bruce (1986). A cluster consisting of these species, with the exception of P. discurus mindorensis, corresponding to our clade 3 was supported in our analyses. Interestingly, this clade comprised species from the oceanic Philippines, Wallacea and the Sulu archipelago. The different subspecies of P. discurus were revealed not to form a monophyletic entity. While P. d. discurus and P. discurus whiteheadi formed a monophyletic, though not robustly supported, group within clade 3, Prioniturus discurus mindorensis was found to cluster in the distantly related clade 2 (together with P. platenae, P. montantus and P. waterstradti). P. discurus mindorensis is restricted to Mindoro and differs from the other subspecies of P. discurus by the combination of a deeper blue crown and a more extensively green forehead (Forshaw, 2006, Salomonsen, 1953). P. discurus whiteheadi, which occurs from Luzon south to Bohol (cf. Fig 2), has some green on the forehead unlike P. d. discurus, and the blue crown is paler and more restricted than in P. d. discurus from Mindanao, Basilan, Sulu and Olutanga (e.g. Salomonsen, 1953, Collar, 1998). According to Salomonsen (1953) the populations on Samar-Leyte are intermediate between P. d. discurus and P. discurus whiteheadi. However, the specimens from Samar in our analyses clustered with the remaining samples from P. discurus whiteheadi.

170 Chapter 7

The allopatrically distributed P. montanus and P. waterstradti have frequently been regarded as conspecific (e.g. Dickinson, 2003, Dickinson et al., 1991), but others have treated them as different species (Collar, 1998, Forshaw, 2006, Peterson, 2006). We found these two taxa to be non-monophyletic: P. waterstradti formed the sister group of a clade consisting of the disjunct taxa P. montanus, P. platenae and P. discurus mindorensis. We recommend treating P. mindorensis comb. nov. as well as P. montanus and P. waterstradti as separate species under a Phylogenetic Species Concept (PSC) (Cracraft, 1989) and a Monophyletic Species Concept (MSC) (Mishler and Donoghue, 1982), given that they are morphologically discretely differentiated and also genetically distinct. All three taxa are also valid species under the integrative guidelines for assessing species rank provided by Helbig et al. (2002), which require levels of divergence comparable to those between sympatric species in order for allopatric taxa to be recognized as species. In addition, the genetic distances found between P. d. discurus and P. discurus whiteheadi may qualify them as separate species in combination with their morphological distinctiveness. However, we refrain from taking any final taxonomic decision until more samples are analyzed from the potentially intermediate populations between P. d. discurus and P. discurus whiteheadi from Samar-Leyte. These findings thus indicate that, depending on the taxonomic source used, one to two, or perhaps even three additional species should be recognized in the Philippines. This is congruent with other studies demonstrating that the diversity of the Philippine avifauna is strongly underestimated (Oliveros and Moyle, 2010, Lohman et al., 2010, Peterson, 2006). A similar pattern has been found in reptiles (e.g. Siler et al., 2011). Species usually form the basis for conservation assessment and strategies (Mace, 2004) and the lack of recognition of distinct evolutionary lineages as distinct species can have severe consequences due to the omission of conservation measures. This may particularly be the case in susceptible island populations (Oliveros and Moyle, 2010, Sangster, 2000, Lovette et al., 1999, Hazevoet, 1996).

Biogeography The parrot genus Prioniturus has evolved in a geologically dynamic region whose complex history of land and sea has created multiple opportunities for allopatric speciation. Accordingly, as in other bird groups in the same area (e.g. Oliveros and Moyle, 2010), our biogeographic reconstruction revealed a complicated pattern of multiple colonization events between the different island groups followed by divergence in allopatry. Five major dispersal events need to be invoked to explain the current distribution pattern of Prioniturus. While the radiation of Prioniturus on Sulawesi/Wallacea was found not to be monophyletic, the Philippine taxa were the result of a single colonization event followed by in-situ speciation and subsequent dispersal to Sulawesi, Palawan and the Sulu archipelago. The last common ancestor of Prioniturus split from its sister group about 25 Ma in Australasia (Schweizer et al., in press) and then either colonized the Philippines or Sulawesi. As the geological history and the distribution of land and sea in the region of Wallacea and the Philippines is complex (e.g. Hall, 2002, Hall, 1998), it is difficult to infer which of these two scenarios is more probable. At the time of the emergence of Prioniturus, the margin of north Australia approached Sulawesi and the Halmahera Arc. This could have created a discontinuous land connection from Australia via the island arcs of Halmahera and the Philippines to Sulawesi, when parts of the Philippines were located further south than today (Hall, 1998). An initial colonization of the Philippines followed by a colonization of the region of Sulawesi and/or adjacent islands would then seem more probable. A colonization of the Philippines from the Australasian region has also been proposed for cuckoo-shrikes (Campephagidae) (Jonsson et al., 2010b) and whistlers (Pachycephalida) (Jonsson et al., 2010a).

171 Evolutionary diversification and biogeography of parrots

Figure 2. 50% majority-rule consensus tree of the Bayesian inference with clade credibility values and bootstrap values of the ML likelihood tree indicated at nodes of the different Prioniturus taxa analyzed. The distribution areas of the taxa in the three inferred clades are indicated in different colors. Distributional information was taken from Forshaw (2006).

172 Chapter 7

Figure 3. Markov-ML reconstruction of ancestral areas. Pie charts at each node represent the proportion of the total likelihood assigned to each biogeographic region as the ancestral area of a given clade. The present distribution of the taxa is indicated by the color of the circles at the node of their last common ancestor, with colors referring to the map on top left. Inferred colonization events are shown with arrows and numbers referring to the corresponding nodes in the tree. Ambiguously inferred colonization events with an alternative reconstruction are given in grey.

173 Evolutionary diversification and biogeography of parrots

A further colonization event took place when the ancestor of P. flavicans dispersed to Sulawesi from the oceanic Philippines. Wallacea and especially Sulawesi have been considered for a long time as holding a source of species to colonize the oceanic Philippines and vice versa (Dickerson, 1928). Dispersal between these two regions could have been facilitated by island hopping over the Sangihe-Talaud Islands, but reports of faunal exchanges among the Philippines and Sulawesi are rare (Evans et al., 2003, Jones and Kennedy, 2008). A colonization of the oceanic Philippines from Sulawesi has only been proposed for one other bird species (Phylloscopus trivirgatus), gastropods (Oncomelania) and insects (Apis) (Smith et al., 2000, Woodruff et al., 1999, Jones and Kennedy, 2008). Dispersal to Sulawesi from the oceanic Philippines and vice versa has, in addition, been described for fanged frogs Limnonectes (Evans et al., 2003), and a close relationship between shrews from Sulawesi and Palawan may exist (Esselstyn et al., 2009). Within the two major clades containing Philippine species (clade 2 and 3), all taxa are allopatric with the exception of a partial overlap in the distribution areas of P. discurus whiteheadi and P. luconensis. Taxa that occur in sympatry and belong to the two different major clades (P. luconensis/P. montanus and P. d. discurus/P. waterstradti) are probably ecologically separated by different habitats and/or altitudinal zones (Collar, 1998). The contemporary diversity of Prioniturus on the oceanic Philippines is thus most probably the result of a single colonization event followed by dispersal among islands or island groups and subsequent divergence in allopatry. Several studies have revealed that the phylogenetic patterns of some Philippine animal groups can at least partly be explained by temporarily fused islands due to lower sea levels during the Pleistocene (Pleistocene aggregate island complexes) (e.g. Evans et al., 2003, Brown and Guttman, 2002, Esselstyn et al., 2004). However, other studies have refuted a major role of Pleistocene geography in diversification patterns in the Philippines (Jones and Kennedy, 2008, Siler et al., 2011, Oliveros and Moyle, 2010, Esselstyn et al., 2009). This seems also to be the case in Prioniturus, as the current distributions of taxa and clades do not correspond to Pleistocene islands. Palawan and the Sulu archipelago were both colonized from the oceanic Philippines. Both island chains may have played an important role as a colonization route into the oceanic Philippines from Sundaland. However, the importance of Palawan as a bridge between the Philippines and Borneo has probably been overestimated (Oliveros and Moyle, 2010, Jones and Kennedy, 2008). Instead, Palawan, with its complex biogeographic history, can additionally be considered an extension of Sundaland as well as a component of the Philippines (Esselstyn et al., 2010). In the case of Prioniturus, Palawan and the Sulu archipelago may represent ‘dead-ends’, as this genus colonized the island groups from the oceanic Philippines, but apparently never successfully established itself on Borneo. It must always be considered that island taxa in particular may have been more widely distributed in the past (cf. Goldberg et al., 2011), and we therefore cannot exclude the possibility that extinct Prioniturus species previously occurred on Sundaland.

Acknowledgements

We especially thank the Silva Casa Foundation for financial support of this project. We are grateful to the following people and institutions for kindly providing us with tissue or toe pad samples: J. Fjeldså and J.-B. Kristensen (Zoological Museum, University of Copenhagen), Paul Sweet (American Museum of Natural History), R. Prys-Jones and M. Adams (British Museum of Natural History), D. Willard (Field Museum of Natural History), A.T. Peterson, R.G. Moyle, M.B. Robbins and A. Nyari (The University of Kansas, Natural History Museum and Biodiversity Research Center), S. Birks (University of Washington, Burke Museum). We

174 Chapter 7 further thank the following people for valuable support: R. Ayé, S. Bachofner, B. Blöchlinger, H. Murray, M. Rieger, H. Schnittker, and C. Sherry.

References

Akaike H (1974) New look at statistical-model identification. Ieee Transactions on Automatic Control, AC19:716-723. Bellemain E, Ricklefs RE (2008) Are islands the end of the colonization road? Trends in Ecology & Evolution, 23:461-468. BirdLifeInternational (2011) Species factsheet. Downloaded from http://www.birdlife.org on 15/03/2011. Boon WM, Kearvell JC, Daugherty CH, Chambers GK (2000) Molecular systematics of New Zealand Cyanoramphus parakeets: conservation of orange-fronted and Forbes' parakeets. Bird Conservation International, 10:211-239. Brown JM, Lemmon AR (2007) The importance of data partitioning and the utility of Bayes Factors in Bayesian phylogenetics. Systematic Biology, 56:643-655. Brown RM, Guttman SI (2002) Phylogenetic systematics of the Rana signata complex of Philippine and Bornean stream frogs: reconsideration of Huxley's modification of Wallace's Line at the Oriental-Australian faunal zone interface. Biological Journal of the Linnean Society, 76:393-461. Collar NJ (1998) Family Psittacidae (Parrots). In: del Hoyo J, Elliott A, Sargatal J (eds), Handbook of the birds of the world, vol. 4. Lynx Edicions, Barcelona, pp. 280-477. Cracraft J (1989) Speciation and its ontology: the empirical consequences of alternative species concepts for understanding patterns and processes of differentiation. In: Otte D, Endler JA (eds), Speciation and its consequences. Sinauer Associates, Sunderland, MA, pp. 28-59. Darwin C (1859) On the origin of species by means of natural selection. Murray, London. de Kloet RS, de Kloet SR (2005) The evolution of the spindlin gene in birds: sequence analysis of an intron of the spindlin W and Z gene reveals four major divisions of the Psittaciformes. Molecular Phylogenetics and Evolution, 36:706-721. Diamond JM, Gilpin ME (1983) Biogeographic umbilici and the origin of the Philippine avifauna. Oikos, 41:307-321. Dickerson RE (1928) Distribution of life in the Philippines. Philippine Bureau of Science, Manila. Dickinson ED, Kennedy RS, Parkes KC (1991) The birds of the Philippines. British Ornithologists' Union, Tring, U.K. Dickinson EDE (2003) The Howard and Moore complete checklist of the birds of the world. Princeton University Press, Princeton, New Jersey. Drummond AJ, Ashton B, Buxton S, et al. (2011) Geneious v5.4. Available from . Esselstyn JA, Oliveros CH (2010) Colonization of the Philippines from Taiwan: a multi-locus test of the biogeographic and phylogenetic relationships of isolated populations of shrews. Journal of Biogeography, 37:1504-1514. Esselstyn JA, Oliveros CH, Moyle RG, Peterson AT, McGuire JA, Brown RM (2010) Integrating phylogenetic and taxonomic evidence illuminates complex biogeographic patterns along Huxley's modification of Wallace's Line. Journal of Biogeography, 37:2054-2066. Esselstyn JA, Timm RM, Brown RM (2009) Do geological or climatic processes drive speciation in dynamic archipelagos? The tempo and mode of diversification in Southeast Asian shrews. Evolution, 63:2595-2610.

175 Evolutionary diversification and biogeography of parrots

Esselstyn JA, Widmann P, Heaney LR (2004) The mammals of Palawan Island, Philippines. Proceedings of the Biological Society of Washington, 117:271-302. Evans BJ, Brown RM, McGuire JA, et al. (2003) Phylogenetics of fanged frogs: testing biogeographical hypotheses at the interface of the Asian and Australian faunal zones. Systematic Biology, 52:794-819. Filardi CE, Moyle RG (2005) Single origin of a pan-Pacific bird group and upstream colonization of Australasia. Nature, 438:216-219. Forshaw JM (1973) Parrots of the world. Landsdowne Press, Melbourne. Forshaw JM (2006) Parrots of the world: an identification guide. Princeton University Press, Princeton. Goldberg J, Trewick SA, Powlesland RG (2011) Population structure and biogeography of Hemiphaga pigeons (Aves: Columbidae) on islands in the New Zealand region. Journal of Biogeography, 38:285-298. Hall R (1998) The plate tectonics of Cenozoic SE Asia and the distribution of land and sea. In: Hall R, Holloway JD (eds), Biogeography and geological evolution of SE Asia. Backhuys Publishers, Leiden, pp. 99-131. Hall R (2001) Cenozoic reconstructions of SE Asia and the SW Pacific: changing patterns of land and sea. In: Metcalfe I, Smith JMB, Morwood M, Davidson ID (eds), Faunal and floral migrations and evolution in SE Asia-Australasia. A.A. Balkema (Swets & Zeitlinger Publishers) Lisse, pp. 35-56. Hall R (2002) Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based reconstructions, model and animations. Journal of Asian Earth Sciences, 20:353-431. Hazevoet CJ (1996) Conservation and species lists: taxonomic neglect promotes the extinction of endemic birds, as examplified by taxa from eastern Atlantic islands. Bird Conservation International, 6:181-196. Helbig AJ, Knox AG, Parkin DT, Sangster G, Collinson M (2002) Guidelines for assigning species rank. Ibis, 144:518-525. Hillis DM, Bull JJ (1993) An empirical-test of bootstrapping as a method for assessing confidence in phylogenetic analysis. Systematic Biology, 42:182-192. Huelsenbeck JP, Ronquist F (2001) MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics, 17:754-755. Jones AW, Kennedy RS (2008) Evolution in a tropical archipelago: comparative phylogeography of Philippine fauna and flora reveals complex patterns of colonization and diversification. Biological Journal of the Linnean Society, 95:620-639. Jonsson KA, Bowie RCK, Moyle RG, et al. (2010a) Historical biogeography of an Indo- Pacific passerine bird family (Pachycephalidae): different colonization patterns in the Indonesian and Melanesian archipelagos. Journal of Biogeography, 37:245-257. Jonsson KA, Bowie RCK, Nylander JAA, Christidis L, Norman JA, Fjeldsa J (2010b) Biogeographical history of cuckoo-shrikes (Aves: Passeriformes): transoceanic colonization of Africa from Australo-Papua. Journal of Biogeography, 37:1767-1781. Jonsson KA, Irestedt M, Fuchs J, et al. (2008) Explosive avian radiations and multi- directional dispersal across Wallacea: evidence from the Campephagidae and other crown Corvida (Aves). Molecular Phylogenetics and Evolution, 47:221-236. Juniper T, Parr M (1998) Parrots: a guide to the parrots of the world. Pica Press, Sussex. Kass RE, Raftery AE (1995) Bayes Factors. Journal of the American Statistical Association, 90:773-795. Lewis PO (2001) A likelihood approach to estimating phylogeny from discrete morphological character data. Systematic Biology, 50:913-925.

176 Chapter 7

Li ZX, Powell CM (2001) An outline of the palaeogeographic evolution of the Australasian region since the beginning of the Neoproterozoic. Earth-Science Reviews, 53:237- 277. Lohman DJ, Ingram KK, Prawiradilaga DM, et al. (2010) Cryptic genetic diversity in "widespread" Southeast Asian bird species suggests that Philippine avian endemism is gravely underestimated. Biological Conservation, 143:1885-1890. Lovette IJ, Bermingham E, Ricklefs RE (1999) Mitochondrial DNA phylogeography and the conservation of endangered Lesser Antillean Icterus orioles. Conservation Biology, 13:1088-1096. MacArthur RH, Wilson EO (1967) The theory of island biogeography. Princeton University Press, Princeton. Mace GM (2004) The role of taxonomy in species conservation. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 359:711-719. Maddison WP, Maddison DR (2006) StochChar: a package of Mesquite modules for stochastic models of character evolution. Maddison WP, Maddison DR (2009) Mesquite: a modular system for evolutionary analysis. Version 2.72. http://mesquiteproject.org. Mishler BD, Donoghue MJ (1982) Species concepts: a case for pluralism. Systematic Zoology, 31:491-503. Moss SJ, Wilson MEJ (1998) Biogeographic implications of the Tertiary palaeogeographic evolution of Sulawesi and Borneo. In: Hall R, Holloway JD (eds), Biogeographic and geological evolution of SE Asia. Backhuys Publishers, Leiden, pp. 133-163. Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity hotspots for conservation priorities. Nature, 403:853-858. Nylander JAA (2004) MrModeltest v2. Evolutionary Biology Centre, Uppsala University Uppsala, Sweden. Nylander JAA, Wilgenbusch JC, Warren DL, Swofford DL (2008) AWTY (are we there yet?): a system for graphical exploration of MCMC convergence in Bayesian phylogenetics. Bioinformatics, 24:581-583. Oliveros CH, Moyle RG (2010) Origin and diversification of Philippine bulbuls. Molecular Phylogenetics and Evolution, 54:822-832. Peters JL (1937) Checklist of the birds of the world. 3. Harvard University Press, Cambridge. Peterson AT (2006) Taxonomy is important in conservation: a preliminary reassessment of Philippine species-level bird taxonomy. Bird Conservation International, 16:155-173. Rambaut A, Drummond AJ (2007) Tracer v1.5, Available from http://beast.bio.ed.ac.uk/Tracer. Ronquist F, Huelsenbeck JP (2003) MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics, 19:1572-1574. Rowley I (1998) Family Cacatuidae (Cockatoos). In: Del Hoyo J, Elliott A, Sargatal J (eds), Handbook of the birds of the world. Lynx Edicions, Barcelona, pp. 246-279. Salomonsen F (1953) Taxonomic notes on the genus Prioniturus. Videnskabelige Meddelelser fra Dansk Naturhistorisk Forening, 115:222-235. Sangster G (2000) Taxonomic stability and avian extinctions. Conservation Biology, 14:579- 581. Schweizer M, Seehausen O, Güntert M, Hertwig ST (2010) The evolutionary diversification of parrots supports a taxon pulse model with multiple trans-oceanic dispersal events and local radiations. Molecular Phylogenetics and Evolution, 54:984-994. Schweizer M, Seehausen O, Hertwig ST (in press) Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations. Journal of Biogeography.

177 Evolutionary diversification and biogeography of parrots

Siler CD, Diesmos AC, Alcala AC, Brown RM (2011) Phylogeny of Philippine slender skinks (Scincidae: Brachymeles) reveals underestimated species diversity, complex biogeographical relationships, and cryptic patterns of lineage diversification. Molecular Phylogenetics and Evolution, 59:53-65. Smith DR, Villafuerte L, Otis G, Palmer MR (2000) Biogeography of Apis cerana F. and A. nigrocincta Smith: insights from mtDNA studies. Apidologie, 31:265-279. Smith GA (1975) Systematics of parrots. Ibis, 117:18-68. Stamatakis A (2006) RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics, 22:2688-2690. Stamatakis A, Hoover P, Rougemont J (2008) A rapid bootstrap algorithm for the RAxML web servers. Systematic Biology, 57:758-771. Tavares ES, Baker AJ, Pereira SL, Miyaki CY (2006) Phylogenetic relationships and historical biogeography of Neotropical parrots (Psittaciformes: Psittacidae: Arini) inferred from mitochondrial and nuclear DNA sequences. Systematic Biology, 55:454-470. Wallace AR (1867) The geographical distribution of animals. Harper, New York. White CMN, Bruce MD (1986) The birds of Wallacea: (Sulawesi, the Moluccas and Lesser Sunda Islands, Indonesia). An annotated checklist. British Ornithologists' Union. Wilgenbusch JC, Warren DL, Swofford DL (2004) AWTY: a system for graphical exploration of MCMC convergence in Bayesian phylogenetic inference. http://ceb.csit.fsu.edu/awty. Winker K (2010) Is it a species? Ibis, 152:679-682. Woodruff DS, Carpenter MP, Upatham ES, Viyanant V (1999) Molecular phylogeography of Oncomelania lindoensis (Gastropoda: Pomatiopsidae), the intermediate host of Schistosoma japonicum in Sulawesi. Journal of Molluscan Studies, 65:21-31. Wright TF, Schirtzinger EE, Matsumoto T, et al. (2008) A multilocus molecular phylogeny of the parrots (Psittaciformes): support for a Gondwanan origin during the Cretaceous. Molecular Biology and Evolution, 25:2141-2156.

Phylogenie und Biogeographie der Papageiengattung Prioniturus (Aves: Psittaciformes)

Zusammenfassung Die Papageien der Gattung Prioniturus sind auf den Philippinen und den Inseln der Wallacea verbreitet. Diese geologisch dynamische Region hat eine komplexe Geschichte der Verteilung von Land und Meer. Einigen der bisher beschriebenen Taxa innerhalb von Prioniturus wurde jeweils ein unterschiedlicher taxonomischer Rang zuerkannt. Obwohl eine phylogenetische Hypothese der Verwandtschaftsverhältnisse dieser Gruppe bisher nicht vorliegt, wurden zudem unterschiedliche Artengruppen vorgeschlagen. Anhand unserer Analysen von Sequenzdaten zweier mitochondrialer Gene konnten wir die Ausbreitungsgeschichte der verschiedenen Prioniturus-Taxa zwischen den Inseln der Philippinen und der Wallacea sowie der Radiationen innerhalb dieser beiden biogeografischen Zonen rekonstruieren. Nach den Ergebnissen unserer gut aufgelösten phylogenetischen Hypothese sollten P. mindorensis comb. nov. sowie P. montanus und P. waterstradti in den Artrang erhoben werden. Bei P. d. discurus und P. discurus whiteheadi sind allerdings weitere Untersuchungen an umfangreicherem Material notwendig, um den Status beider Formen zu klären. Dieses Ergebnis stimmt mit anderen Studien überein, welche zeigen konnten, dass die Alpha- Diversität der Vögel der Philippinen stark unterschätzt wird. Gemäss unserer biogeographischen Rekonstruktion diversifizierte Prioniturus durch eine Folge von mehreren Kolonisierungsereignissen und anschliessender allopatrischer Artbildung zwischen und innerhalb von Inselgruppen. Die nachgewiesene Besiedlung von Sulawesi/Wallacea

178 Chapter 7 ausgehend von den Philippinen bzw. umgekehrt dokumentiert einen sonst nur selten nachgewiesenen Faunenaustausch zwischen diesen beiden Regionen.

179

Chapter 8

DISPARITY VERSUS DIVERSITY AND THE ROLE OF ECOLOGICAL OPPORTUNITY IN THE EVOLUTION OF NEOTROPICAL PARROTS

Manuel Schweizera, Stefan T. Hertwiga, Marcel Günterta, Ole Seehausenb,c

Manuscript

183 Evolutionary diversification and biogeography of parrots

Abstract

Avian diversity is higher in South America than in all other tropical regions combined. Accordingly, the Neotropical parrots (Arini) were found to be most probably more diverse compared to the remaining parrots. They colonized South America from Antarctica after they split in the late Eocene/early Oligocene from a clade consisting of African parrots (Psittacus, Poicephalus) today. The newly invaded Neotropics may have presented under-utilized adaptive zones providing several ecological opportunities facilitating diversification. However, diversification of Arini may have also been driven by range fragmentation and ecological changes caused by Andean uplift and climate change from the Miocene onwards. We therefore tested whether the Neotropical parrots diversified in an adaptive radiation after their colonization of South America or if their diversification occurred later and was influenced by historical processes from the Miocene onwards. We used a time- calibrated phylogeny and morphological measurements to analyze lineage diversification and phenotypic evolution of Neotropical parrots. The signature of an adaptive radiation was found in their early history with increased diversification rates, a concentration of morphological size evolution and partitioning of size niches. However, we found no signs for a decrease in speciation rates through time and Arini may have not yet reached their equilibrium diversity. Historical processes from the Miocene onwards may have continuously provided new niche spaces not limiting the number of parrot species in the Neotropics.

Keywords: adaptive radiation, Arini, colonization, diversification, ecological opportunity, morphological disparity, Psittaciformes

184 Chapter 8

Introduction

Species richness is not randomly and not uniformly distributed on earth (Gaston 2000). Especially the latitudinal diversity gradient is so prevalent (e.g. Mittelbach et al. 2007; Lomolino et al. 2010) that it has been called earth’s first-order biodiversity pattern (Krug et al. 2009). But diversity also varies between continents within similar latitudes, and among the tropics, species richness is highest in many taxa in the Neotropics (e.g. Rosenzweig 1995). South America contains approximately one third of all bird species and avian diversity is higher than in all other tropical regions combined; a similar pattern can also be found in other groups of organisms (Weir 2006; Price 2008; Hoorn et al. 2010). Several hypotheses have been formulated to explain this unusually high number of species in the Neotropics. Prominent theories have focused either on low extinction rates associated with long-term environmental stability through the Cenozoic, or on high speciation rates associated with recurrent episodes of habitat fragmentation and re-connection caused by Pleistocene climatic fluctuations (reviewed e.g. in Weir 2006; Hoorn et al. 2010). A recent synthesis paper concluded that the unusually high species diversity of the Neotropics is most probably caused by a combination of several mechanisms associated with spatially and temporally heterogeneous landscape and climate dynamics (Hoorn et al. 2010). Pleistocene speciation events importantly contributed to the species richness of Andean birds, but lowland and especially lowland Amazonian bird diversity patterns originated before the Pleistocene (Weir 2006; Hoorn et al. 2010). The uplift of the Andes first peaking in late Oligocene/early Miocene times with subsequent intensive orogenesis in the late Miocene/early Pliocene may have been a crucial driver of Neotropical species diversification by providing ecological opportunities and promoting vicariance (Weir 2006; Brumfield and Edwards 2007; Farias et al. 2008; Garzione et al. 2008; Thomas et al. 2008; Hoorn et al. 2010). Not surprisingly, the Neotropical Arini are the most species-rich group within the circumtropically distributed parrots (Psittaciformes) (Collar 1998; Rowley 1998). The Arini are a monophyletic assemblage most closely related to the African Psittacini (Schweizer et al. 2010; Schweizer et al. in press). A common ancestor of Arini and Psittacini most probably lived in Antarctica and colonized South America and Africa from there in the late Eocene or early Oligocene when continental ice sheets expanded rapidly on Antarctica as a result of climate change toward cooler conditions (Schweizer et al. in press). While the African Psittacini consist of only ten species today, the Neotropical Arini have radiated into a clade of 148 species which is likely unexpectedly species-rich given its age with an overall increased diversification rate compared to the remaining parrot groups (Schweizer et al. in press). The newly colonized, previously parrot-free Neotropics may have presented an under-utilized adaptive zone providing ecological opportunities that would have facilitated diversification (Schweizer et al. 2010; Schweizer et al. in press). Ecological opportunity may lead to strong directional selection within, and divergent selection between habitats, and can initiate adaptive radiations (Losos 2010; Yoder et al. 2010). Thus, the Arini may be an example of an adaptive radiation that evolved after the colonization of South America and resulted in the diverse and ecologically disparate clades now observed. The process of adaptive radiation is defined as the evolution of ecological and phenotypic diversity within a rapidly multiplying lineage (Simpson 1953; Schluter 2000). The pattern associated with adaptive radiations is hence characterized by rapid accumulation of lineages early in clade history in combination with an increase of ecological and

185 Evolutionary diversification and biogeography of parrots morphological disparity among lineages caused by divergent adaptations to diverse resources (Schluter 2000; Harmon et al. 2003; Harmon et al. 2010; Slater et al. 2010). To the extent that diversification is driven by ecological opportunity in initially underexploited adaptive zones, an early burst of lineage diversification and morphological evolution should be followed by a slowdown as a consequence of density dependence when ecological niche space becomes filled (Harmon et al. 2003; Seehausen 2006; Rabosky and Lovette 2008; Harmon et al. 2010). Although probably influenced by sampling artefacts, there seems to have been a time lag between the arrival of parrots in South America at around 35 Ma (millions of years ago) and the beginning of their diversification in the late Oligocene about 25 Ma (Schweizer et al. in press). Thus, the initial diversification of Neotropical parrots would have occurred contemporaneously with the first peak of Andean orogenesis (cf. Hoorn et al. 2010). Andean uplift has indeed been invoked as a driver of speciation in Neotropical parrots during the Neogene (Ribas et al. 2005; Ribas et al. 2007a). Moreover, in addition to Andean orogenesis, shrinking and fragmentation of rain forests caused by climate change with a subsequent increase in areas of open vegetation in the middle/late Miocene could have further driven parrot diversification via niche expansion and ecological speciation on one hand, and by promoting vicariance events on the other (Tavares et al. 2006; Schweizer et al. in press). In this study, we used phylogenetic comparative methods to ask whether the Neotropical parrots indeed diversified in an adaptive radiation, and if this was triggered by their colonization of South America or rather by later orogenetic and climatic processes of landscape remodeling especially during the Neogene. In the first case, we would expect a slowdown in lineage diversification through time after an early burst of speciation and ecological evolution with little morphological variation found within subclades. The alternative hypothesis predicts an increase in lineage accumulation rates and ecomorphological diversification in the Neogene, i.e. well after the arrival of parrots in South America. Moreover, we investigated if the high diversity of Neotropical parrots was a result of high net diversification rates in particular clades. We reconstructed a timeframe for the evolution of the Neotropical Arini based on a relaxed molecular clock approach using secondary calibration points derived from another study (Schweizer et al. in press). We used overall size and relative bill length as characters which are likely to represent ecological differences among parrot species (Collar 1998).

Material & Methods

Divergence times and phylogenetic inference Nucleotide sequences of 99 species of Arini and of Psittacus erithacus as outgroup were extracted from Genbank. We used partial sequences of the comparatively slowly evolving nuclear exon RAG-1 (Groth and Barrowclough 1999) and of the more variable mitochondrial genes COI (cytochrome oxidase subunit I), Cytb (cytochrome b) and ND2 (NADH dehydrogenase subunit 2) (Appendix 1). Our sampling included more than 66% of all described Arini species and all of the 30 recognized genera except Ognorhynchus and Psilopsiagon (taxonomy follows Collar (1998)). We used Beast v. 1.6.1 (Drummond and Rambaut 2007) to simultaneously estimate divergence times and establish a phylogenetic hypothesis for the Neotropical Arini. According to the results of Tavares et al. (2006) and Wright et al. (2008), we defined a cluster of the following genera as being monophyletic in the prior settings of Beast:

186 Chapter 8

Anodorhynchus, Ara, Aratinga, Cyanoliseus, Cyanopsitta, Deroptyus, Diopsittaca, Enicognathus, Guarouba, Leptosittaca, Nandayus, Orthopsittaca, Pionites, Propyrrhura, Pyrrhura, Rhynchopsitta. We applied a relaxed molecular clock with uncorrelated lognormal distribution of branch lengths and a Yule tree prior. Because of missing data in some of the species, genes were analyzed in a concatenated dataset using a GTR + I + G model of sequence evolution. We incorporated a secondary calibration point in our dating analyses according to the results of Schweizer et al (in press). We used a mean estimate of 35 Ma based on a normal distribution with a standard deviation of 6.5 for the split between Arini and Psittacini. The 95% interval of this normal distribution included the 95% highest posterior density (HPD) for this split as estimated in Schweizer et al. (in press). Default prior distributions were chosen for all other parameters and we ran three independent chains of MCMC with 25 million generations sampled every 1000 generations. Tracer v. 1.5 (Rambaut and Drummond 2007) was used to confirm appropriate burn-in, the adequate effective sample sizes of the posterior distribution for each run and to assess convergence among runs by comparing likelihoods and posterior distributions of all parameters. The resulting trees and the 95% highest posterior density (HPD) distributions of each estimated node were analyzed with FigTree v. 1.2.1 (Rambaut 2008). Psittacus erithacus was pruned from the trees for further analyses.

Lineage diversification The R packages ‘Laser’ v. 2.3 (Rabosky 2006a) and ‘Geiger’ v. 1.3.1 (Harmon et al. 2008) were used for all comparative data modeling. Temporal variation in diversification rates was visualized with semilogarithmic lineage-through-time plots (LTT). We used the maximum clade credibility tree from the Beast analyses and plotted the 95% confidence interval of the LTT plots from the 1000 last trees of the Beast analyses with the root node set to 26.548 My (mean value recovered, see below). To test whether diversification rates have decreased over time, we calculated the gamma (Ā) statistic of Pybus & Harvey (2000). The constant rate test of the gamma (Ā) statistic can incorrectly reject the null hypothesis (type-I error) because of incomplete taxon sampling and thus the Monte Carlo constant-rates test is usually applied (Pybus and Harvey 2000). However, this procedure assumes that missing taxa are randomly distributed in the tree and if this assumption is violated, this can increase type-I error unless > 80% of species in a particular clade have been sampled (Cusimano and Renner 2010). Our taxon sampling is probably biased towards inclusion of more deeply diverging lineages (phylogenetically overdispersed sampling). Thus, we assessed the sensitivity of the results of the Monte Carlo constant-rates test to overdispersed sampling using the NRS test developed by Brock et al. (2011). We used birth and death rates as estimated in the BDL approach (see below) and implemented different values of the scaling parameter Ā between 0 and 10. For Ā = 0, pruning of taxa from a phylogenetic tree is done completely randomly in the NRS test, whereas a higher proportion of taxa with shorter tip branches are pruned from the tree when values of Ā are higher (Brock et al. 2011). To infer whether particular clades within Arini are unexpectedly species-rich with regard to their age, we applied the method implemented in MEDUSA (Alfaro et al. 2009). It uses a time-calibrated phylogeny (diversity tree) with a species richness value assigned at each tip of the tree as bases. The maximum clade credibility tree from the Beast analyses was used as diversity tree and pruned to 29 tips mostly representing genera where species missing in our sampling could be incorporated (Fig 2). Ognorhynchus icterotis was included in the clade of Ara and the two Psilopsiagon

187 Evolutionary diversification and biogeography of parrots species in the clade of Bolborhynchus (cf. Forshaw 1989; Collar 1998). In a stepwise procedure, MEDUSA first fits a single birth-death model to the diversity tree on the bases of the likelihood function of Rabosky et al. (2007). Next, a model with two birth rates and two death rates, including a shift location parameter on a branch of the phylogeny (five parameters), is fitted to the diversity tree and its AIC score is compared with that of the first two-parameter model. If this new model produces a significant improvement in AIC scores, it is retained and compared with a eight- parameter model (cf. Alfaro et al. 2009; Slater et al. 2010). This process is continued until the improvement in AIC score of a more complex model with additional rate- shifts and rate parameters is less than a defined threshold number. We considered an increase in model fit as significant in the whole study when the reduction in AIC score in a more complex model was Ā 4 (Burnham and Anderson 2002) To test for temporal shifts in diversification rates in the radiation of Arini as a whole, we used the BDL approach of Rabosky (2006b) based on the maximum clade credibility tree from the Beast analyses. This approach compares the Akaike information criterion (AIC) of the best-fitting rate-constant diversification model with the best-fitting rate-variable diversification model. We tested two rate-constant models (1 and 2) and two rate-variable models (3 and 4): 1) a pure birth (Yule) model (one parameter Ȁ (speciation), Ā (extinction) set to 0); 2) a rate-constant-birth-death (bd) model (two parameters Ȁ and Ā); 3) a pure birth rate-variable model with the speciation rate Ȁ1 shifting to Ȁ2 at a time t (yule2rate, three parameters Ȁ1, Ȁ2 and t); 4) a pure birth rate-variable model with two rate-shifts and three speciation rates (yule3rate, five parameters Ȁ1, Ȁ2, Ȁ3, and t1, t2).

Morphological disparity and models of morphological evolution Morphological measurements were compiled from Forshaw (1989). For each trait and species an average trait value was calculated from the mean values for the two sexes. We used total length and tarsus length as indication of overall size. We took bill length, measured as a straight line from the anterior edge of the cere to the tip of the upper mandible (Forshaw 1989). All measurements were natural-log (ln) transformed prior to analyses. As bill length correlates with body mass or body length, we implemented size-corrected values for this measurement. For this, we calculated independent contrasts (Felsenstein 1985) based on the maximum clade credibility tree of the Beast analyses and performed a regression through the origin of the contrasts of ln transformed bill length against contrasts of ln transformed body length. The slope from this regression was then fitted to the original ln transformed data and the residuals from the regression line were used for further analyses (cf. Revell 2009). To investigate the history of morphological evolution, we calculated mean subclade disparity through time (Harmon et al. 2003) for all morphological measurements, with disparity calculated from average squared Euclidean distances among species traits based on the maximum clade credibility tree of the Beast analyses. Disparity values close to zero imply that most variation is partitioned as differences between subclades, while values close to one indicated that a substantial amount of variation is contained within subclades (Harmon et al. 2003). Observed values were compared with a null distribution from 1000 simulations of trait evolution over our tree under a Brownian motion process. Disparity through time was visualized by plotting observed and simulated mean subclade disparity against node age (DTT plot). Additionally, we calculated mean subclade disparity from the 1000 last trees of the Beast analyses and plotted each value relative to the timescale of the maximum clade credibility tree. Moreover, we calculated the MDI (morphological disparity index) statistic, which is

188 Chapter 8 the area between the DTT plot for the data and the mean of the simulations. Negative MDI values are characteristic for adaptive radiations and indicate a lower subclade disparity than expected under Brownian motion evolution (Harmon et al. 2003; Slater et al. 2010). We tested six models of continuous character evolution by fitting likelihood models to our data based on the maximum clade credibility tree of the Beast analyses. 1) Brownian motion (BM); 2) Ornstein-Uhlenbeck (OU) process, a random walk model with a central tendency proportional to the parameter alpha (Butler and King 2004); 3) early-burst (EB) model, with the rate of evolution decreasing (or increasing) exponentially through time (Harmon et al. 2010); 4) Pagel's lambda, where all internal branch lengths of the tree are multiplied by lambda, with values of lambda close to 1 implying that traits covary as assumed by Brownian motion (Pagel 1999); 5) Pagel’s delta, where all node depths are raised to the power delta, if delta is < 1, morphological change is concentrated early in the tree and then slows down, whereas values of delta > 1 indicate that evolution speeds up and is concentrated towards the tips (Pagel 1999); 6) Pagel’s kappa, where all branch lengths are raised to the power kappa, with values of kappa close to 0 indicating a speciational model of trait evolution (Pagel 1999). The first three models are formulated in distinct and alternative evolutionary processes, namely continuous gradualism, bounded evolution and density dependent evolutionary change respectively (Harmon et al. 2010). The last three models on the other hand, estimate best-fitting model coefficients, and depending on the value of the coefficients, they describe different evolutionary processes (Pagel 1999). Therefore, they can not be considered as biological alternatives of the first three models and we compared AICc scores (Akaike's information criterion corrected for sample size) among models 1) -3) and models 4) - 6).

Results

Divergence times and phylogenetic inference The final alignment consisted of 5506 bp and contained no indels. The comparison of the three independent runs with Beast revealed high convergence among the different parameters and effective sample sizes were >1108 for all parameters. The three runs were combined with 2.5 million generations burn-in for each. The maximum clade credibility tree was then estimated from 67503 trees (Fig. 1). After the split from the African Psittacini at a mean value of 30.1 Ma (see Table 1 for 95% highest posterior density (HPD)), the diversification of the Neotropical parrots started around 26.5 Ma in the late Oligocene. Thus, although much shorter than in Schweizer et al. (in press), there is still a time lag of about 3.5 million years between the divergence of the last common ancestor of Psittacini and Arini and the start of the diversification of the Neotropical parrots. The Neotropical parrots then rapidly diversified into four clades (nodes 2-4 in Fig. 1). The diversification within these clades basically started in the early to middle Miocene, while most of the extant genera originated in the middle Miocene. The diversification within genera started in the late Miocene, e.g. 11.5 Ma in Amazona, 10.5 Ma in Pionopsitta (excluding P. pileata), 8.5 Ma in Brotogeris, 8.2 Ma in Pionus, 8.0 Ma in Ara, 7.6 Ma in Pyrrhura.

189 Evolutionary diversification and biogeography of parrots

Table 1. Divergence dates for different nodes estimated with a Bayesian relaxed-clock approach. The mean values and the 95% highest posterior density (HPD) distributions are given. Node numbers refer to those in Fig. 1.

Node number Mean 95% HPD (Ma)

1 30 068 16 183-44 360 2 26 548 13 553-39 985 3 25 766 12 835-38 686 4 24 945 12 597-37 487 5 21 570 10 768-32 549 6 18 214 8 840-28 016 7 17 570 8 656-26 477 8 21 564 10 405-32 766 9 10 507 5 002-16 053 10 8 209 3 867-12-778 11 11 521 5 602-17 552 12 8 530 4 054-13 200 13 7 569 3 530-11 914 14 8 047 3 890-12 561

The topology of the maximum clade credibility tree from the Beast analyses was compared with recently published molecular phylogenies of Arini. The phylogenetic relationships found among genera were highly congruent with the results of Tavares et al. (2006), although their sampling was more restricted. All supported nodes in the Bayesian inferences in the latter study were also revealed in our analyses with the exception of the position of Forpus. While we found Forpus as the sister taxon of a clade consisting of Bolborhynchus, Touit and Nannopsittaca, Tavares et al. (2006) revealed it as the sister group of taxa congruent with our clade at node 7 (Fig. 1). However, the latter study did not include Touit. All robustly resolved clades of the Bayesian analyses of Wright et al. (2008) were also found in our study except that we found no cluster consisting of Cyanoliseus, Deroptyus, Pyrrhura, Rhynchopsitta and Enicognathus. Pionopsitta turned out to be polyphyletic in congruence with Tavares et al. (2006) and Ribas et al. (2005). As in the latter study, Hapalopsittaca was found to be the sister group to the remaining taxa of Pionopsitta (excluding P. pileata). The relationships within the latter group as found by us were identical to those in Ribas et al. (2005). With the exception of the position of Aratinga leucophthalmus, the relationships of the analyzed species of Aratinga were in congruence with Ribas & Miyaki (2004) and Nandayus nenday was found to cluster within Arantinga. Within Brotogeris, we revealed the same clades as Ribas et al. (2009). Moreover, the relationships found within Pionus were congruent with the supported nodes in Ribas et al. (2007a). Amazona xanthops did not cluster with the other members of this genus as also observed in other studies (Russello and Amato 2004; Tavares et al. 2006). The relationships within Amazona were found to be overall consistent with the results of Russello & Amato (2004) with the exception of the position of A. amazonica, A. farinosa and the clade consisting of A. dufresniana and A. rhodocorytha. The topology within Pyrrhura was also broadly congruent with the results of Ribas et al. (2006) for supported nodes.

Lineage diversification The gamma (Ā) statistic did not detect any evidence for a significant decrease in diversification rate over time for any value of in the NRS sensitivity assessment of the Monte Carlo constant-rates test.

190 Chapter 8

MEDUSA did not uncover any rate-shift within Arini, thus no clade was found to be unexpectedly species-rich given its age compared to the remaining Neotropical parrot clades. The net diversification rate (r = - Ā) as estimated by MEDUSA was 0.13 and the relative extinction rate (r = Ā/ ) very low at 4.79e-07. The BDL approach revealed the rate-variable models to fit the data better than the rate-constant models. A yule3rate model with three speciation rates and two shift location parameters was identified as the best-fitting model (Table 2). A first rate-shift to a reduced speciation rate occurred very early in the diversification of the Neotropical parrots at 25.75 Ma. The second rate-shift, again to a further reduced speciation rate, happened shortly before the present at 1.11 Ma. However, the reduction in AIC score compared to a yule2rate model was below the threshold for a significant increase in model fit. The latter model identified just one shift to a reduced speciation rate at 1.11 Ma. This rate-shift shortly before the present in both the yule2rate and the yule3rate model is most probably influenced by incomplete lineage sampling, with species missing in several genera (cf. Ricklefs et al. 2007). Hence, we have an indication that the speciation rate decreased early in the history of Arini. This was also supported by the visual inspection of the LTT plot for the maximum clade credibility tree and the 95% confidence interval from the last 1000 trees of the Beast analyses (Fig. 3).

Table 2. Log likelihood (lnL) and AIC scores of the comparison between rate-constant with rate- variable diversification models. The rate-variable models fit the data better than the rate-constant models and a yule3rate model with three speciation rates and two shift location parameters was identified as the best-fitting model ( =speciation, Ā=extinction, t=time of rate-shift in millions of years ago).

Model pure birth bd yule2rate yule3rate

Parameters =0 1176856 -Ā=0 1176856 1=0 1326796 1=1 2482412

Ā=0 2=0 01848098 2=0 13017762

t=1 105685 Ȁ3=0 01848098

t1=25 753449

t2=1 105685 Ln L 49 98442 49 98442 57 67432 62 52323 AIC -97 96883 -95 96883 -109 34863 -110 77394 dAIC 12 805107 14 805107 1 425308 0

191 Evolutionary diversification and biogeography of parrots

Figure 1. Maximum clade credibility tree of the dating analysis of Neotropical parrots (Arini) using Beast. Mean node ages and the 95% highest posterior density distributions are shown. Posterior summaries were only calculated for the nodes in the tree that had a posterior probability greater than 0.5. Node numbers refer to Table 1.

192 Chapter 8

Morphological disparity and models of morphological evolution Subclade disparity through time was lower than expected under the null model of Brownian motion for total length and tarsus length throughout the radiation of Arini (Fig. 3). This visually revealed result is quantitatively supported by a negative MDI value of -0.168 for total length and of -0.170 for tarsus length. However, there was a clear secondary burst in the evolution of subclade disparity approximately between 15 and 13 Ma in the middle Miocene. Two smaller increases of subclade disparity were found at the beginning of the radiation of Arini and close to the present. On the other hand, subclade disparity for relative bill length was overall close to the null expectation with an MDI value of 0.056. It showed a burst of evolution shortly after the start of Arini radiation with values well above the expectation under the null model followed by lower values until a second burst at about 15 Ma in the middle Miocene (Fig. 3). Afterwards, values were close to the expectation until a last increase happened close to the present. When comparing the first three models of morphological evolution (BM, OU, EB), a Brownian motion process was revealed as the best-fitting model for all three traits (Table 3). When comparing models 3) - 6), Pagel’s delta was found to be the best- fitting model of morphological evolution (Table 3). The inferred values for delta of 0.256 for total length and 0.358 for tarsus length suggest that morphological evolution was concentrated early in the evolutionary history of Arini and slowed down as their radiation progressed. In contrast to the two body size measurements, the revealed delta value of 2.0 for relative bill length indicates that morphological evolution was concentrated nearer the tips of the tree.

Table 3. AICc scores for the six models of morphological evolution analyzed.

Model

Ornstein- Early- Brownian Lambda Delta Kappa Uhlenbeck burst

Total length -50 75 -48 61 -48 61 -48 61 -57 467 -48 61

Tarsus length -112 01 -109 88 -109 88 -109 88 -114 68 -109 88

Relative bill length -153 28 -152 63 -152 63 -152 19 -153 68 -151 21

193 Evolutionary diversification and biogeography of parrots

Figure 2. Diversity tree used for the MEDUSA analysis. The maximum clade credibility tree from the Beast analyses was pruned to 29 tips mostly representing genera where species missing in our sampling could be incorporated. The number of species for each tip lineage is given. Figures (not to scale) were taken from Forshaw (2010).

194 Chapter 8

Figure 3. Lineage and disparity-through-time plots for Neotropical parrots (Arini) from the dated phylogeny using Beast. a) Semi-logarithmic lineage-through-time plot. Points are derived from the maximum clade credibility tree, while the area shaded in grey represents the 95% confidence interval from the last 1000 trees of the Beast analyses with the age of the initial split fixed at the mean value of the posterior distribution. b) - d) Disparity-through-time plot for b) total length, c) tarsus length and d) relative bill length. The solid line indicates the mean subclade disparity through time based on the maximum clade credibility tree of the Beast analyses, while the dashed line shows the median expected subclade disparity derived from 1000 simulations under Brownian motion evolution. The grey points are the mean subclade disparity values from the 1000 last trees of the Beast analyses plotted relative to the timescale of the maximum clade credibility tree.

Discussion

Neotropical parrots are more species-rich compared to other parrot groups (Schweizer et al. in press). Here we showed that this exceptional diversity is not due to one or two particularly diverse clades within them, but is instead a general property of Arini. We found indications for an early rapid diversification of the Neotropical parrots as a whole, which was probably associated with an increase in speciation rates relative to the African sister clade (cf. also Schweizer et al. in press). Short branch lengths early in a clade’s history may either be caused by rapid ecological speciation or by a more or less simultaneous fragmentation of populations into multiple geographically isolated lineages with subsequent speciation primarily by non-ecological modes (Price 2008; Rundell and Price 2009). Hence, ecology and biogeography of extant taxa have to be consulted to tell whether a rapid radiation was ‘adaptive’ or ‘nonadaptive’ (Rundell and Price 2009). Our analyses revealed that the early diversification of Neotropical parrots was coupled with rapid diversification in body size, likely reflecting a partitioning of the ecological niche space. This would suggest that ecological processes might have played an important role. Parrots, including Arini, show exceptionally large between-species variation in size compared to other bird groups and different body sizes may reflect divergent

195 Evolutionary diversification and biogeography of parrots adaptation to different ecological niches (Collar 1998). Body size evolution was concentrated at the beginning of the radiation of Neotropical parrots and the observed mean body size disparity in subclades was always below the values expected under Brownian motion evolution. Niche dimensions where body size matters thus seem to have been partitioned early in the evolutionary diversification of Arini probably leaving limited opportunity for morphological diversification in body sized afterwards within subclades. These results correspond to a model of rapid evolution within a clade after its origination (Harmon et al. 2003; Smith et al. 2011), and declining rates subsequently. Analyses of body size evolution over several lineages of birds also revealed patterns consistent with an early adaptive radiation in body size (Harmon et al. 2010). The lineages of Arini that emerged in the early speciation burst have largely overlapping geographical ranges. Taken together, these observations suggest that the early diversification was an adaptive radiation. In contrast, Arini showed an overall level of subclade disparity for relative bill length close to the expectation under Brownian motion evolution. There were three bursts of morphological evolution with elevated subclade disparity values indicating that subclades evolved into overlapping morphospace (cf. Harmon et al 2003). Values of subclade disparity above the expectation can be the result of evolution occurring within bounds (Smith et al. 2011). However, was constrained morphological evolution a major factor, we had expected to find support for an Ornstein-Uhlenbeck model of bill size evolution, whereas bill size evolution did hardly differ from a Brownian motion process or may have speeded up through time. Differences in bill features between parrot species often are dietary adaptations (Homberger 1980; Collar 1998), and bursts of morphological diversification probably reflecting radiations in trophic adaptations, seem to have happened recurrently within body size radiations. There was apparently an early burst of body size evolution followed by a slowdown, however, it is not clear whether bill size evolution started later or if it happened in parallel and did not slow down. It is thus unclear how the sequence in trait diversification corresponds to the progress of niche diversification and how it is related to divergence within or among communities and macrohabitats (cf. Ackerly et al. 2006). This has to be further investigated by using more extensive measurements of bill structure and other morphological aspects associated with feeding and habitat specializations. Evolutionary patterns are hence consistent with an adaptive radiation process having been prevalent in the early evolution of Neotropical parrots as indicated by low body size disparity and an early burst in diversifications rates. This adaptive radiation may have been causally linked with the arrival of Arini in South America. Signs of rapid speciation and morphological evolution after invasion into new areas are classically associated with adaptive radiations on isolated islands and in lakes (e.g. Givnish 1999; Schluter 2000) but have also been postulated in other organisms following colonization of continents (e.g. Simpson 1950; Webb 1991; Rowe et al. 2008; Burbrink and Pyron 2009). The previously parrot-free South American continent may have presented itself as an unexploited or underexploited adaptive zone with ecological opportunities promoting diversification of the initial colonizing population of parrots. However, the diversification of the Neotropical parrots did not start immediately after their divergence from their last common ancestor with the African Psittacini. The timing of the split between Psittacini and Arini did not necessarily happen contemporaneously with the colonization of South America by the latter. Their divergence may have instead taken place in Antarctica with subsequent extinction of the Arini lineage in Antarctica. On the other hand, the initial

196 Chapter 8 diversification of Arini is temporally coincident with the first peak of Andean uplift in the late Oligocene/early Miocene (Hoorn et al. 2010). This early stage of Andean orogenesis was probably important in shaping the evolution of Neotropical avian species richness (Thomas et al. 2008) and could also have driven speciation in parrots. Some formulations of the ecological theory of adaptive radiation predict that diversification rates decline in the course of a radiation as a consequence of available ecological niche space becoming filled and leaving fewer opportunities for speciation (Schluter 2000; Rabosky and Lovette 2008). Such a pattern, reflecting density dependent speciation rates, has been detected in several groups of organisms (Harmon et al. 2003; Kozak et al. 2005; Ruber and Zardoya 2005; Weir 2006; Phillimore and Price 2008; Rabosky and Lovette 2008; Burbrink and Pyron 2009; Brock et al. 2011). However, we found no overall statistical support that the early occupation of niche space was followed by such density dependent effects in the speciation rates of Neotropical parrots. Similarly, there is no evidence for a slowdown in speciation rates for Neotropical birds as a whole (Price 2008). It is possible that the expected density dependent slowdown of speciation was compensated (and the resulting pattern obscured) by the de-novo formation from the Miocene onwards of new adaptive zones through increasing Andean uplift and climate change (Farias et al. 2008; Garzione et al. 2008; Thomas et al. 2008; Hoorn et al. 2010). Again, new habitats due to mountain orogenesis and new areas of open vegetation in combination with drainage of wetlands (Hoorn et al. 2010) may have promoted habitat fragmentation and vicariance on one hand, and ecological opportunity associated with competitive release in new and vacant niche space on the other hand. In Neotropical parrots, the importance of the latter mechanism is indicated by a burst of morphological evolution in the middle Miocene with increased subclade disparity in all morphological features analyzed. This time period had been postulated as important in shaping diversity patterns in Neotropical parrot groups like Pionopsitta, Pionus and to a lesser degree Brotogeris, with varying effects on the different groups of Andean orogenesis and extension of open vegetation in the lowlands (Ribas et al. 2005; Ribas et al. 2007a; Ribas et al. 2009). Moreover, dispersal from lowlands to highlands happened primarily in the late Miocene/early Pliocene in several Neotropical bird groups (Weir 2006). However, we have no indication for increased lineage diversification rates in our data during this time period. Weir (2006) revealed contrasting diversification patterns between lowland and highland taxa in different Neotropical bird groups. While speciation rates for lowland taxa slowed down through time, they cumulated in the late Pleistocene in highland species. Lowland diversity thus seems to approach its carrying capacity, whereas diversity in the highlands may not have reached carrying capacity yet (Weir 2006). In line with this, Ribas et al. (2007a) hypothesized, that the diversification of Pionus parrots in the Andes occurred mainly in the Pleistocene, whereas diversification of tropical lowland Pionopsitta taxa occurred before the Pleistocene (Ribas et al. 2005). However, there are also recent speciation events in lowland Neotropical parrots which may have been influenced by Pleistocene climate changes and associated habitat alterations through dynamic shifts between wet and dry forests, isolation of open dry areas and speciation along forest/open-area ecotones (Ribas and Miyaki 2004; Ribas et al. 2007b; Ribas et al. 2009). Such temporal and spatial variation in diversification patterns may explain the absence of overall change in diversification rates of Neotropical parrots as a whole from the Miocene onwards. Clade age has often been proposed as an important predictor of species richness (e.g. Wiens 2011). The contrasting diversity of the same-aged Neotropical Arini and African Psittacini is thus striking. However, the lack of a relationship between clade

197 Evolutionary diversification and biogeography of parrots age and species richness has been found in many groups, and radiations on different continents rarely unfold in the same manner (e.g. Magallon and Sanderson 2001; Seehausen 2006; Ricklefs 2007; Ricklefs et al. 2007; Losos 2010). Ecological limits might lead to a decoupling of clade age and species richness, and can regulate species diversity (Rabosky 2009). While Arini might not have reached their equilibrium diversity yet in the Neotropics as a whole, the situation might be different for Psittacini on the African continent. Factors which might have limited the diversification of parrots in Africa and the influence of extinction remain to be identified. As pointed out by Hoorn et al. (2010), species accumulation in Amazonia and the Neotropical region as a whole was driven by more than a single mechanism and happened over a prolonged period. Although Neotropical parrots show patterns consistent with an adaptive radiation early in their evolutionary history, a complex array of different factors influenced their further diversification since.

References

Ackerly, D. D., D. W. Schwilk, and C. O. Webb. 2006. Niche evolution and adaptive radiation: testing the order of trait divergence. Ecology 87:S50-S61. Alfaro, M. E., F. Santini, C. Brock, H. Alamillo, A. Dornburg, D. L. Rabosky, G. Carnevale, and L. J. Harmon. 2009. Nine exceptional radiations plus high turnover explain species diversity in jawed vertebrates. Proceedings of the National Academy of Sciences of the United States of America 106:13410- 13414. Brock, C. D., L. J. Harmon, and M. E. Alfaro. 2011. Testing for temporal variation in diversification rates when sampling is incomplete and nonrandom. Systematic Biology 60:410-419. Brumfield, R. T., and S. V. Edwards. 2007. Evolution into and out of the Andes: a Bayesian analysis of historical diversification in Thamnophilus antshrikes. Evolution 61:346-367. Burbrink, F. T., and R. A. Pyron. 2009. How does ecological opportunity influence rates of speciation, extinction, and morphological diversification in New World ratsnakes (tribe Lampropeltini)? Evolution 64:934-943. Burnham, K. P., and D. R. Anderson. 2002. Model selection and mulitimodel inference. Second edition. Springer-Verlag, New York. Butler, M. A., and A. A. King. 2004. Phylogenetic comparative analysis: a modeling approach for adaptive evolution. American Naturalist 164:683-695. Collar, N. J. 1998. Family Psittacidae (parrots). Pp. 280-477 in J. del Hoyo, A. Elliott, and J. Sargatal, eds. Handbook of the birds of the world, vol. 4. Lynx Edicions, Barcelona. Cusimano, N., and S. S. Renner. 2010. Slowdowns in diversification rates from real phylogenies may not be real. Systematic Biology 59:458-464. Drummond, A. J., and A. Rambaut. 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary Biology 7. Farias, M., R. Charrier, S. Carretier, J. Martinod, A. Fock, D. Campbell, J. Caceres, and D. Comte. 2008. Late Miocene high and rapid surface uplift and its erosional response in the Andes of central Chile (33 degrees-35 degrees S). Tectonics 27:22.

198 Chapter 8

Felsenstein, J. 1985. Phylogenies and the comparative method. American Naturalist 125:1-15. Forshaw, J. M. 1989. Parrots of the world. Landsdowne Press, Melbourne. Forshaw, J. M. 2010. Parrots of the world: a field guide. Christopher Helm, London. Garzione, C. N., G. D. Hoke, J. C. Libarkin, S. Withers, B. MacFadden, J. Eiler, P. Ghosh, and A. Mulch. 2008. Rise of the Andes. Science 320:1304-1307. Gaston, K. J. 2000. Global patterns in biodiversity. Nature 405:220-227. Givnish, T. J. 1999. Adaptive radiation, dispersal, and diversification of the Hawaiian lobeliads. Pp. 67-99 in M. Kato, ed. The biology of biodiversity. Springer- Verlag, Tokyo. Groth, J. G., and G. F. Barrowclough. 1999. Basal divergences in birds and the phylogenetic utility of the nuclear RAG-1 gene. Molecular Phylogenetics and Evolution 12:115-123. Harmon, L. J., J. B. Losos, T. J. Davies, R. G. Gillespie, J. L. Gittleman, W. B. Jennings, K. H. Kozak, M. A. McPeek, F. Moreno-Roark, T. J. Near, A. Purvis, R. E. Ricklefs, D. Schluter, J. A. Schulte, O. Seehausen, B. L. Sidlauskas, O. Torres-Carvajal, J. T. Weir, and A. O. Mooers. 2010. Early bursts of body size and shape evolution are rare in comparative data. Evolution 64:2385-2396. Harmon, L. J., J. A. Schulte, A. Larson, and J. B. Losos. 2003. Tempo and mode of evolutionary radiation in iguanian lizards. Science 301:961-964. Harmon, L. J., J. T. Weir, C. D. Brock, R. E. Glor, and W. Challenger. 2008. GEIGER: investigating evolutionary radiations. Bioinformatics 24:129-131. Homberger, D. G. 1980. Funktionell- morphologische Untersuchungen zur Radiation der Ernährungs- und Trinkmethoden der Papageien (Psittaci). Bonn. Zool. Monogr. 13:1-92. Hoorn, C., F. P. Wesselingh, H. ter Steege, M. A. Bermudez, A. Mora, J. Sevink, I. Sanmartin, A. Sanchez-Meseguer, C. L. Anderson, J. P. Figueiredo, C. Jaramillo, D. Riff, F. R. Negri, H. Hooghiemstra, J. Lundberg, T. Stadler, T. Sarkinen, and A. Antonelli. 2010. Amazonia through time: Andean uplift, climate change, landscape evolution, and biodiversity. Science 330:927-931. Kozak, K. H., A. A. Larson, R. M. Bonett, and L. J. Harmon. 2005. Phylogenetic analysis of ecomorphological divergence, community structure, and diversification rates in dusky salamanders (Plethodontidae: Desmognathus). Evolution 59:2000-2016. Krug, A. Z., D. Jablonski, J. W. Valentine, and K. Roy. 2009. Generation of earth's first-order biodiversity pattern. Astrobiology 9:113-124. Lomolino, M. V., B. R. Riddle, R. J. Whittaker, and J. H. Brown. 2010. Biogeography, fourth edition. Sinauer Associates, Inc., Sunderland. Losos, J. B. 2010. Adaptive radiation, ecological opportunity, and evolutionary determinism. American Naturalist 175:623-639. Magallon, S., and M. J. Sanderson. 2001. Absolute diversification rates in angiosperm clades. Evolution 55:1762-1780. Mittelbach, G. G., D. W. Schemske, H. V. Cornell, A. P. Allen, J. M. Brown, M. B. Bush, S. P. Harrison, A. H. Hurlbert, N. Knowlton, H. A. Lessios, C. M. McCain, A. R. McCune, L. A. McDade, M. A. McPeek, T. J. Near, T. D. Price, R. E. Ricklefs, K. Roy, D. F. Sax, D. Schluter, J. M. Sobel, and M. Turelli. 2007. Evolution and the latitudinal diversity gradient: speciation, extinction and biogeography. Ecol. Lett. 10:315-331.

199 Evolutionary diversification and biogeography of parrots

Pagel, M. 1999. Inferring the historical patterns of biological evolution. Nature 401:877-884. Phillimore, A. B., and T. D. Price. 2008. Density-dependent cladogenesis in birds. PLoS. Biol. 6:483-489. Price, T. 2008. Speciation in birds. Roberts and Company, Greenwood Village, CO. Pybus, O. G., and P. H. Harvey. 2000. Testing macro-evolutionary models using incomplete molecular phylogenies. Proceedings of the Royal Society B- Biological Sciences 267:2267-2272. Rabosky, D. L. 2006a. LASER: a maximum likelihood toolkit for detecting temporal shifts in diversification rates from molecular phylogenies. Evol Bioinform Online 2:257-260. Rabosky, D. L. 2006b. Likelihood methods for detecting temporal shifts in diversification rates. Evolution 60:1152-1164. Rabosky, D. L. 2009. Ecological limits and diversification rate: alternative paradigms to explain the variation in species richness among clades and regions. Ecol. Lett. 12:735-743. Rabosky, D. L., S. C. Donnellan, A. L. Talaba, and I. J. Lovette. 2007. Exceptional among-lineage variation in diversification rates during the radiation of Australia's most diverse vertebrate clade. Proceedings of the Royal Society B- Biological Sciences 274:2915-2923. Rabosky, D. L., and I. J. Lovette. 2008. Density-dependent diversification in North American wood warblers. Proceedings of the Royal Society B-Biological Sciences 275:2363-2371. Rambaut, A. 2008. FigTree 1.2., published by the author. Rambaut, A., and A. J. Drummond. 2007. Tracer v1.5, available from http://beast.bio.ed.ac.uk/Tracer. Revell, L. J. 2009. Size-correction and principal components for interspecific comparative studies. Evolution 63:3258-3268. Ribas, C. C., R. Gaban-Lima, C. Y. Miyaki, and J. Cracraft. 2005. Historical biogeography and diversification within the Neotropical parrot genus Pionopsitta (Aves: Psittacidae). Journal of Biogeography 32:1409-1427. Ribas, C. C., L. Joseph, and C. Y. Miyaki. 2006. Molecular systematics and patterns of diversification in Pyrrhura (Psittacidae), with special reference to the picta- leucotis complex. Auk 123:660-680. Ribas, C. C., and C. Y. Miyaki. 2004. Molecular systematics in Aratinga parakeets: species limits and historical biogeography in the 'solstitialis' group, and the systematic position of Nandayus nenday. Molecular Phylogenetics and Evolution 30:663-675. Ribas, C. C., C. Y. Miyaki, and J. Cracraft. 2009. Phylogenetic relationships, diversification and biogeography in Neotropical Brotogeris parakeets. Journal of Biogeography 36:1712-1729. Ribas, C. C., R. G. Moyle, C. Y. Miyaki, and J. Cracraft. 2007a. The assembly of montane biotas: linking Andean tectonics and climatic oscillations to independent regimes of diversification in Pionus parrots. Proceedings of the Royal Society B-Biological Sciences 274:2399-2408. Ribas, C. C., E. S. Tavares, C. Yoshihara, and C. Y. Miyaki. 2007b. Phylogeny and biogeography of yellow-headed and blue-fronted parrots (Amazona ochrocephala and Amazona aestiva) with special reference to the South American taxa. Ibis 149:564-574.

200 Chapter 8

Ricklefs, R. E. 2007. Estimating diversification rates from phylogenetic information. Trends in Ecology & Evolution 22:601-610. Ricklefs, R. E., J. B. Losos, and T. M. Townsend. 2007. Evolutionary diversification of clades of squamate reptiles. J. Evol. Biol. 20:1751-1762. Rosenzweig, M. L. 1995. Species diversity in time and space. Cambridge University Press, Cambridge, U.K. Rowe, K. C., M. L. Reno, D. M. Richmond, R. M. Adkins, and S. J. Steppan. 2008. Pliocene colonization and adaptive radiations in Australia and New Guinea (Sahul): multilocus systematics of the old endemic rodents (Muroidea: Murinae). Molecular Phylogenetics and Evolution 47:84-101. Rowley, I. 1998. Family Cacatuidae (cockatoos). Pp. 246-279 in J. Del Hoyo, A. Elliott, and J. Sargatal, eds. Handbook of the birds of the world. Lynx Edicions, Barcelona. Ruber, L., and R. Zardoya. 2005. Rapid cladogenesis in marine fishes revisited. Evolution 59:1119-1127. Rundell, R. J., and T. D. Price. 2009. Adaptive radiation, nonadaptive radiation, ecological speciation and nonecological speciation. Trends in Ecology & Evolution 24:394-399. Russello, M. A., and G. Amato. 2004. A molecular phylogeny of Amazona: implications for Neotropical parrot biogeography, taxonomy, and conservation. Molecular Phylogenetics and Evolution 30:421-437. Schluter, D. 2000. The ecology of adaptive radiation. Oxford University Press, Oxford. Schweizer, M., O. Seehausen, M. Güntert, and S. T. Hertwig. 2010. The evolutionary diversification of parrots supports a taxon pulse model with multiple trans- oceanic dispersal events and local radiations. Molecular Phylogenetics and Evolution 54:984-994. Schweizer, M., O. Seehausen, and S. T. Hertwig. in press. Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations. Journal of Biogeography. Seehausen, O. 2006. African cichlid fish: a model system in adaptive radiation research. Proceedings of the Royal Society B-Biological Sciences 273:1987- 1998. Simpson, G. G. 1950. History of the fauna of Latin America. American Scientist 38:361-389. Simpson, G. G. 1953. The major features of evolution. Columbia University Press, New York. Slater, G. J., S. A. Price, F. Santini, and M. E. Alfaro. 2010. Diversity versus disparity and the radiation of modern cetaceans. Proceedings of the Royal Society B- Biological Sciences 277:3097-3104. Smith, K. L., L. J. Harmon, L. P. Shoo, and J. Melville. 2011. Evidence of constrained phenotypic evolution in a cryptic species complex of agamid lizards. Evolution 65:976-992. Tavares, E. S., A. J. Baker, S. L. Pereira, and C. Y. Miyaki. 2006. Phylogenetic relationships and historical biogeography of Neotropical parrots (Psittaciformes: Psittacidae: Arini) inferred from mitochondrial and nuclear DNA sequences. Systematic Biology 55:454-470. Thomas, G. H., C. D. L. Orme, R. G. Davies, V. A. Olson, P. M. Bennett, K. J. Gaston, I. P. F. Owens, and T. M. Blackburn. 2008. Regional variation in the

201 Evolutionary diversification and biogeography of parrots

historical components of global avian species richness. Global Ecology and Biogeography 17:340-351. Webb, S. D. 1991. Ecogeography and the great American interchange. Paleobiology 17:266-280. Weir, J. T. 2006. Divergent timing and patterns of species accumulation in lowland and highland Neotropical birds. Evolution 60:842-855. Wiens, J. J. 2011. The causes of species richness patterns across space, time, and clades and the role of “ecological limits". Q. Rev. Biol. 86:75-96. Wright, T. F., E. E. Schirtzinger, T. Matsumoto, J. R. Eberhard, G. R. Graves, J. J. Sanchez, S. Capelli, H. Mueller, J. Scharpegge, G. K. Chambers, and R. C. Fleischer. 2008. A multilocus molecular phylogeny of the parrots (Psittaciformes): support for a Gondwanan origin during the Cretaceous. Molecular Biology and Evolution 25:2141-2156. Yoder, J. B., E. Clancey, S. Des Roches, J. M. Eastman, L. Gentry, W. Godsoe, T. J. Hagey, D. Jochimsen, B. P. Oswald, J. Robertson, B. A. J. Sarver, J. J. Schenk, S. F. Spear, and L. J. Harmon. 2010. Ecological opportunity and the origin of adaptive radiations. J. Evol. Biol. 23:1581-1596.

202 Chapter 8

Appendix 1. Species sampled and GenBank Accession numbers for the four genes analyzed for this study.

COI Cytb ND2 RAG1

Amazona aestiva EU340705 AY194404 AY194434 JF807980 Amazona albifrons AY301427 AY283506 - - Amazona amazonica AY194399 AY283516 AY194466 - Amazona arausiaca - AY283465 - - Amazona autumnalis AY194379 AY283456 AY194446 - Amazona barbadensis AY301436 AY283462 - - Amazona brasiliensis AY301437 - - - Amazona collaria AY301438 AY283493 - - Amazona dufresniana AY301439 AY283454 - JF807981 Amazona farinosa AY301443 FJ899168 AY194461 DQ143346 Amazona festiva AY301444 - - - Amazona finschi AY301445 AY283461 - - Amazona guildingii AY301446 AY283460 - - Amazona imperialis AY301447 AY283458 - - Amazona leucocephala AY301449 AY283518 - - Amazona ochrocephala ochrocephala AY194397 AY194411 AY194460 - Amazona pretrei AY301457 - - JF807982 Amazona rhodocorytha AY301458 - - - Amazona tucumana AY301459 - - - Amazona ventralis AY301460 U89178 - - Amazona versicolor AY301461 AY283466 - - Amazona vinacea AY301462 - - - Amazona viridigenalis AY301463 - EU327598 - Amazona vittata AY301464 AY283514 - - Amazona xanthops AY301465 DQ143293 AY669485 DQ143345 Anodorhynchus hyacinthinus AF370738 AF346370 DQ143311 DQ143329 Anodorhynchus leari - AF370763 AY669446 - Ara ararauna AY301467 DQ150994 DQ143315 - Ara chloroptera - AF346367 - - Ara macao EU621598 AF346366 EU327601 DQ143340 Ara severa AF370748 AF370766 - - Aratinga aurea AF370746 - DQ143321 DQ143341 Aratinga auricapilla AY301466 AY208239 - - Aratinga cactorum AF370750 AY281254 - - Aratinga jandaya - AY208244 - - Aratinga leucophthalmus AF370749 AY281253 Q143298 DQ143331 Aratinga nana - - FJ361233 - Aratinga pertinax EU621597 - EU327600 - Aratinga solstitialis - DQ143285 DQ143317 DQ143330 Aratinga weddelli - AY669401 AY669445 - Bolborhynchus lineola - DQ143291 DQ143319 DQ143350 Brotogeris chiriri - FJ652859 FJ652922 DQ143348 Brotogeris chrysoptera - FJ652893 FJ652957 - Brotogeris cyanoptera - FJ652871 FJ652938 - Brotogeris jugularis EU621601 FJ652910 FJ652973 - Brotogeris pyrrhopterus - FJ652865 FJ652928 - Brotogeris sanctithomae - FJ652902 FJ652965 - Brotogeris tirica - FJ652849 FJ652911 - Brotogeris versicolurus AF370755 FJ652853 FJ652917 - Cyanoliseus patagonus GQ232191 GQ232267 EU327613 DQ143334 Cyanopsitta spixii EU621610 - EU327614 DQ143336 Deroptyus accipitrinus EU621613 DQ150992 EU327617 DQ143338 Diopsittaca nobilis AF370752 AF370769 EU327618 DQ143335 Enicognathus leptorhynchus EU621616 - EU327620 DQ143332 Forpus crassirostris - - DQ143300 DQ143325 Forpus passerinus EU621621 - EU327625 - Graydidascalus brachyurus AY301468 AY669439 AY669484 DQ143344 Guarouba guarouba AF370741 AF346371 DQ143309 DQ143333 Pionopsitta barrabandi AY661216 AY669424 AY669468 DQ143349 Pionopsitta haematotis AY661225 AY669433 AY669477 - Pionopsitta pulchra AY661231 AY669427 AY669472 - Pionopsitta pyrilia - AY669426 AY669470 - Hapalopsittaca amazonina EU621625 AY669443 EU327629 -

203 Evolutionary diversification and biogeography of parrots

Appendix 1. Continued.

COI Cytb ND2 RAG1

Leptosittaca branickii EU621626 - EU327630 - Myiopsitta monachus EU621631 AF346378 EU327635 DQ143328 Nandayus nenday EU621632 AY219915 AY274066 DQ143326 Nannopsittaca panychlora EU621633 - EU327637 - Orthopsittaca manilata EU621640 DQ150991 EU327644 DQ143337 Pionites leucogaster - DQ143288 DQ143312 DQ143342 Pionites melanocephala EU621644 - EU327648 - Pionopsitta caica AY661221 - EU327649 - Pionopsitta pileata AY661227 AY669438 AY669479 DQ143351 Pionopsitta vulturina AY661213 AY669412 AY669453 - Pionus chalcopterus AY661232 - EF517656 - Pionus fuscus - EF517636 EF517671 - Pionus maximiliani - EF517622 EF517662 DQ143347 Pionus menstruus AY301469 EF517611 EF517637 JF807987 Pionus senilis - EF517618 EF517653 - Pionus sordidus - EF517634 EF517669 - Pionus tumultuosus - - EF517649 - Propyrrhura auricollis AF370753 DQ150995 DQ143302 DQ143343 Propyrrhura couloni AF370762 AF370780 - - Propyrrhura maracana AF370754 AF370775 - - Pyrrhura albipectus - AY751640 - - Pyrrhura cruentata - AY751658 - - Pyrrhura frontalis - AY751643 - AY233360 Pyrrhura hoffmanni - AY751654 - - Pyrrhura leucotis - AY751633 DQ143307 DQ143353 Pyrrhura lepida - AY751645 - - Pyrrhura melanura - AY751652 - - Pyrrhura molinae - AY751642 - - Pyrrhura orcesi - AY751636 - - Pyrrhura perlata - AY751646 - - Pyrrhura picta EU621660 AY751602 AY669444 - Pyrrhura rhodocephala - AY751638 - - Pyrrhura rupicola - AY751656 - - Rhynchopsitta pachyrhyncha EU621661 DQ143297 DQ143303 DQ143339 Touit batavica EU621666 - EU327670 - Triclaria malachitacea - - AY669486 DQ143327 Psittacus erithacus EU621657 AY082076 GU816826 EF517674

204

Acknowledgements

ACKNOWLEDGEMENTS

I am grateful to my supervisor Marcel Güntert who gave me the opportunity to write my PhD on the evolutionary diversification of birds, a subject I have always wanted to work on. I would like to thank him for his thorough support, dedication and enthusiasm.

This work would not have been possible without the help of Stefan T. Hertwig. I am indebted to his dedication, guidance and support. He always had time for discussions and an open ear for my ideas.

Ole Seehausen, my co-supervisor, was a continuous support and most chapters profited from his tremendous knowledge in evolutionary biology and thought- provoking comments and ideas.

I especially thank the Silva Casa Foundation for financial support of the whole project.

I had the fortune to work at the Naturhistorisches Museum der Burgergemeinde Bern which provided a very fruitful and stimulating working environment.

I would like to thank the following people or institutions for kindly providing me with specimens, tissue or feather samples: S. Birks (University of Washington, Burke Museum), R. Burkhard, Federal Veterinary Office Bern (FVO), A. Fergenbauer- Kimmel, J. Fjeldså and J.-B. Kristensen (Zoological Museum, University of Copenhagen), H. Gygax, A. T. Peterson, R. G. Moyle, M. B. Robbins and A. Nyari (The University of Kansas, Natural History Museum and Biodiversity Research Center), H. Gygax, R. Prys-Jones and M. Adams (British Museum of Natural History), D. Ruess, P. Sandmeier, P. Sweet (American Museum of Natural History), T. and P. Walser, G. Weis, D. Willard (Field Museum of Natural History).

The following people helped in various ways during the work on my thesis and they are highly acknowledged: R. Ayé, S. Bachofner, H. Baur, B. Blöchlinger, C. D. Brock, R. Burri, L. Cathrow, V. de Pietri, J. Fischer, H. Frick, N. Gruner, L. J. Harmon, M. Hohn, L. Joseph, S. Klopfstein, C. Kropf, B. Kurz, S. Kurz, S. Lauper, L. Lepperhof, C. Leuenberger, M. Maan, R. Markwalder, G. Mayr, R. Morales-Hojas, H. Murray, S. A. Price, L. J. Revell, L. & A. Rieger, T. Roth, H. Schnitker, C. Sherry, D. Strahm, M. Troxler, P. Walser, T. F. Wright, Y. Zahn.

Last but not least, I am grateful to Mirjam Rieger and to my parents, Therese and Franz Schweizer, for their understanding, continuous support, patience and love.

207

Erklarung

gemass Art. 28 Abs. 2 RSL 05

NameN orname:

Matrikelnummer:

Studiengang:

Bachelor D Master D Dissertation D

Titel der Arbeit:

Leiterln der Arbeit:

lch erklare hiermit, dass ich diese Arbeit selbstandig verfasst und keine anderen als die angegebenen Quellen benutzt habe. Alie Stellen, die wortlich oder sinngemass aus Quellen entnommen wurden, habe ich als solche gekennzeichnet. Mir ist bekannt, dass andernfalls der Senat gemass Artikel 36 Absatz 1 Buchstabe o des Gesetztes vom 5. September 1996 Ober die Universitat zum Entzug des auf Grund dieser Arbeit verliehenen Titels berechtigt ist.

Ort/Datum

Unterschrift

Curriculum vitae

CURRICULUM VITAE

Manuel Schweizer

Naturhistorisches Museum der Burgergemeinde Bern Bernastrasse 15 CH 3005 Bern Switzerland (tel) +41(0)31 350 72 86 (e-mail) [email protected]

Date of birth: 10.05.1979 Nationality: Swiss

Education

- 2008- Dissertation at the Department of Vertebrates of the Natural History Museum of Bern, and the University of Bern, Institute of Ecology and Evolution. Title: The evolutionary diversification and biogeography of parrots (Aves: Psittaciformes): an integrative approach. Supervision: Prof. Dr. Marcel Güntert, Prof. Dr. Ole Seehausen.

- 2006-2007 Certificate for high school teaching in biology at the Institut Sekundarstufe II, Pädagogische Hochschule Bern.

- 2004-2005 Master thesis in population genetics at the University of Bern, Zoological Institute. Title: Dispersal and fine-scale genetic structure in the common vole Microtus arvalis. Supervision: PD. Dr. Gerald Heckel, Prof. Dr. Laurent Excoffier.

- 2000-2003 Undergraduate studies at the University of Bern.

- 1995-1999 Matura Typus C at the Mathematisch-Naturwissenschaftliches Gymnasium Bern-Neufeld.

Professional experiences and memberships

- 2005- Scientific assistant at the Department of Vertebrates of the Natural History Museum of Bern (with an interruption between July 2006 and Aug 2007).

- 2005-2010 Work on a field guide to the birds of Central Asia for A&C Black Publishers, London: Aye, R., M. Schweizer, and T. Roth. in press. Birds of Central Asia. Helm Field Guides. Helm/A&C Black, London. (ISBN 978 07136 7038 X).

211 Evolutionary diversification and biogeography of parrots

- 2007-2008 Part-time high school biology teacher at Gymnasium Hofwil, Münchenbuchsee.

- 2002- Guidance of altogether 18 ornithological study trips in 12 different countries for Liberty Bird GmbH, RB Reiseberatung AG, Bern.

- 2002- Member of the Swiss Bird Rarities Committee.

- 2000- Collaboration in various bird monitoring project fort he Swiss Ornithological Institution, Federal Office for the Environment FOEN, Hintermann & Weber AG, Büro Mosimann & Strebel Landschafts- pflege & Faunistik, Groupe d'étude et de gestion de la Grande Caricaie, Urbasol S.A.

- 1997 - 2008 Member of the executive board of the society of bird protection of Bern (Bernische Gesellschaft für Vogelkunde und Vogelschutz, Berner ALA).

Publications

Books

R. Aye, M. Schweizer, and T. Roth. Birds of Central Asia. Helm Field Guides. Helm/A&C Black, London. (ISBN 978 07136 7038 X). Published January 2012.

Scientific Articles

Schweizer, M., O. Seehausen, and S. T. Hertwig. in press. Macroevolutionary patterns in the diversification of parrots - effects of climate change, geological events and key innovations. Journal of Biogeography. Armbruster G. F. J., M. Schweizer, and D. T. Vogt. in press. A database on visible diurnal spring migration of birds (Central Europe: Lake Constance). Ecology. Streitberger. K., M. Schweizer, R. Kropatsch, G. Dekomien, O. Distl, M. S. Fischer, J. T. Epplen, and S. T. Hertwig. in press. Rapid genetic diversification within dog breeds as evidenced by a case study on Schnauzers. Animal Genetics. Hertwig, S. T., I. Das, M. Schweizer, R. Brown and A. Haas. in press. Phylogenetic relationships of the Rhacophorus everetti-group and implications for the evolution of reproductive modes in Philautus (Amphibia: Anura: Rhacophoridae). Zoologica Scripta. Schweizer M., O. Seehausen, M. Güntert, and S. T. Hertwig. 2010. The evolutionary diversification of parrots supports a taxon pulse model with multiple trans- oceanic dispersal events and local radiations. Molecular Phylogenetics and Evolution 54:984-994. Ayé, R., S. T. Hertwig, and M. Schweizer. 2010. Discovery of a breeding area of the enigmatic large-billed reed warbler Acrocephalus orinus. Journal of Avian Biology 41:452-459.

212 Curriculum vitae

Vallotton, L., B. Piot, M. Schweizer, P. Knaus. 2010. Seltene Vogelarten und ungewöhnliche Vogelbeobachtungen in der Schweiz im Jahre 2009. 19. Bericht der Schweizerischen Avifaunistischen Kommission. [Rare birds and unusual observations in Switzerland in 2009]. Der Ornithologische Beobachter 107:221-246. Haag, C., M. Schweizer, D. Hercigonja, P. Lustenberger, and A. Fasel. 2010. Première observation du Goubmouche à Demi-Collier Ficedula semitorquata en Suisse. Nos Oiseaux 57:58-62. Hertwig, S. T., M. Schweizer, S. Stepanow et al. 2009. Regionally high rates of hybridization and introgression in German wildcat populations (Felis silvestris, Carnivora, Felidae). Journal of Zoological Systematics and Evolutionary Research 47:283-297. Schweizer, M., and M. Thoma. 2009. Seltene Vogelarten und ungewöhnliche Vogelbeobachtungen in der Schweiz im Jahre 2008. 18. Bericht der Schweizerischen Avifaunistischen Kommission. [Rare birds and unusual observations in Switzerland in 2008]. Der Ornithologische Beobachter 106:377-400. Schweizer, M., M. Thoma, and B. Piot. 2009. Oiseaux rares et observations inhabituelles en Suisse en 2008. 18e rapport de la Commission de l'avifaune suisse. [Rare birds and unusual observations in Switzerland in 2008]. Nos Oiseaux 56:237-255. Kirwan, G. M., M. Schweizer, R. Ayé, and A. Grieve. 2009. Taxonomy, identification and status of desert sparrows. Dutch Birding 31:139-158. Haag, C., M. Schweizer, P. Lustenberger & A. Fasel 2009. Identification of the first semi-collared flycatcher for Switzerland. Birding World 22:246-248. Schweizer, M., and M. Mitropolskiy. 2009. Erratum: Schweizer, M. & M. Mitropolskiy 2008. The occurrence of the crested honey buzzard Pernis ptilorhyncus in Uzbekistan and Tajikistan and its status in Central Asia. Sandgrouse 20, 161-163. Sandgrouse 31:79. Aubry, S., and M. Schweizer. 2008. Le Goéland Pontique Larus cachinnans en images. Statut et détermination en Suisse. Nos Oiseaux 55:211-225. Vallotton, L., B. Piot, and M. Schweizer. 2008. Seltene Vogelarten und ungewöhnliche Vogelbeobachtungen in der Schweiz im Jahre 2007. 17. Bericht der Schweizerischen Avifaunistischen Kommission. [Rare birds and unusual observations in Switzerland in 2007]. Der Ornithologische Beobachter 105:305-328. Schweizer, M., and M. Mitropolskiy. 2008. The occurrence of the crested honey buzzard Pernis ptilorhyncus in Uzbekistan and Tajikistan and its status in Central Asia. Sandgrouse 20:161-163. Schweizer, M., and M. Güntert. 2008. Die erste dokumentierte Feststellung einer Raubmöwe Stercorarius sp. in der Schweiz. [The earliest documented record of a jaeger Stercorarius sp. in Switzerland.]. Der Ornithologische Beobachter 105:145-152. Schweizer M., L. Excoffier, and G. Heckel. 2007. Fine-scale genetic structure and dispersal patterns in the common vole Microtus arvalis. Molecular Ecology 16:2463-2473. Ritschard, M., and M. Schweizer. 2007. Identification of Asian Glaucidium owlets. Birding Asia 7:39-47. Schweizer, M., and R. Aye. 2007. Greater sand plover in Tunisia in November- December 2005 and November 2006. Dutch Birding 29:302-303.

213 Evolutionary diversification and biogeography of parrots

Schweizer, M., 2007. Seltene Vogelarten und ungewöhnliche Vogelbeobachtungen in der Schweiz im Jahre 2006. 16. Bericht der Schweizerischen Avifaunistischen Kommission. [Rare birds and unusual observations in Switzerland in 2006]. Der Ornithologische Beobachter 104:241-262. Schweizer, M., and L. Vallotton. 2007. Oiseaux rares et observations inhabituelles en Suisse en 2006. 16e rapport de la Commission de l'avifaune suisse. [Rare birds and unusual observations in Switzerland in 2006]. Nos Oiseaux 54:225-241. Schweizer, M., and R. Aye. 2007. Identification of the pale sand martin Riparia d. diluta in Central Asia. Alula 13:152-158. Aye, R., and M. Schweizer. 2006. Identification of red-mantled Carpodacus rhodochlamys and Blyth's rosefinch C. grandis. Dutch Birding 28:201-213. Schweizer, M., R. Aye, and R. Burri. 2006. Afghan babbler - a little known Western Palearctic bird. Birding World 19:69-70. Schweizer, M. 2005. Seltene Vogelarten und ungewöhnliche Vogelbeobachtungen in der Schweiz im Jahre 2004. 14. Bericht der Schweizerischen Avifaunistischen Kommission. [Rare birds and unusual observations in Switzerland in 2004.]. Der Ornithologische Beobachter 102:251-270. Schweizer, M., and B. Piot. 2005. Oiseaux rares et observations inhabituelles en Suisse en 2004. 14e rapport de la Commission de l'avifaune suisse. [Rare birds and unusual observations in Switzerland in 2004]. Nos Oiseaux 52:245-260. Schweizer, M. 2005. Hybridization beween the blue-headed Motacilla flava and the ashy-headed wagtail M. cinerocapilla in Switzerland. Dutch Birding 27:235- 241. Armbruster, G., D. Renz, and M. Schweizer. 2005. Eine dreijährige Feldstudie zum sichtbaren Frühjahreszug am Bodensee (Süddeutschland). Vogelwarte 43:171- 178. Roth, T., R. Aye, R. Burri, and M. Schweizer. 2005. Bird observations from Iran in February – March 2001, including a new species for the Middle East. Sandgrouse 27:63-68. Aye, R., and M. Schweizer. 2003. Wintering hybrids pine bunting x yellowhammer in northern Iran. Dutch Birding 25:40-43. Schweizer, M. 2003a. Die Postjuvenilmauser der Mittelmeermöwe Larus michahellis. Limicola 17:169-187. Schweizer, M. 2003b. A European stonechat Saxicola rubicola in northern Iran. Sandgrouse 25:146-147. Schweizer, M. 2003c. Seltene Vogelarten und ungewöhnliche Vogelbeobachtungen in der Schweiz im Jahre 2002. 12. Bericht der Schweizerischen Avifaunistischen Kommission. [Rare birds and unusual observations in Switzerland in 2002]. Der Ornithologische Beobachter 100:293-314. Schweizer, M., and L. Maumary. 2003. Oiseaux rares et observations inhabituelles en Suisse en 2002. 12e rapport de la Commission de l'avifaune suisse. [Rare birds and unusual observations in Switzerland in 2002]. Nos Oiseaux 50:263-278. Fischer, J., and M. Schweizer. 2003. Ein adulter Weissbürzelstrandläufer Calidris fuscicollis im Chablais de Cudrefin/Fanel. [An adult white-rumped sandpiper Calidris fuscicollis at Lake Neuchatel]. Der Ornithologische Beobachter 100:247-250. Schweizer, M. 2000. Der Buschspötter Hippolais caligata , eine neue Art für die Schweiz. [The booted warbler Hippolais caligata, a new species for Switzerland.]. Der Ornithologische Beobachter 97:249-257.

214 Curriculum vitae

Schweizer, M. 1999. Der Erstnachweis der Nominatform der Weisskopfmöwe Larus c. cachinnans in der Schweiz. [First record of nominate subspecies of the yellow-legged gull Larus c. cachinnans in Switzerland]. Der Ornithologische Beobachter 96:131-134.