Evolutionary History and Its Relevance in Understanding and Conserving Southern African Biodiversity

Evolutionary history and its relevance in understanding and conserving southern African biodiversity

Thèse de doctorat és Science de la vie (PhD)

Presentée à la

Faculté de Biologie et Médicine de l’Université de Lausanne

Par

Dorothea Pio

Diplômée en Ecologie et Conservation (University of Aberdeen and University of East Anglia)

Jury de Thèse: Prof. Edward E. Farmer, Président Prof. Antoine Guisan, Directeur de thèse Dr Nicolas Salamin, Co-directeur de thèse Dr Richard Grenyer, Expert Prof. Luca Fumagalli, Expert

Lausanne 2010 2 Abstract

Understanding how biodiversity is distributed is central to any conservation effort and has traditionally been based on niche modeling and the causal relationship between spatial distribution of organisms and their environment. More recently, the study of species’ evolutionary history and relatedness has permeated the fields of ecology and conservation and, coupled with spatial predictions, provides useful insights to the origin of current biodiversity patterns, community structuring and potential vulnerability to extinction.

This thesis explores several key ecological questions by combining the fields of niche modeling and phylogenetics and using important components of southern African biodiversity. The aims of this thesis are to provide comparisons of biodiversity measures, to assess how climate change will affect evolutionary history loss, to ask whether there is a clear link between evolutionary history and morphology and to investigate the potential role of relatedness in macro-climatic niche structuring.

The first part of my thesis provides a fine scale comparison and spatial overlap quantification of species richness and phylogenetic diversity predictions for one of the most diverse plant families in the Cape Floristic Region (CFR), the Proteaceae. In several of the measures used, patterns do not match sufficiently to argue that species relatedness information is implicit in species richness patterns.

The second part of my thesis predicts how climate change may affect threat and potential extinction of southern African animal and plant taxa. I compare present and future niche models to assess whether predicted species extinction will result in higher or lower phylogenetic diversity survival than what would be experienced under random extinction processes. I find that predicted extinction will result in lower phylogenetic diversity survival but that this non-random pattern will be detected only after a substantial proportion of the taxa in each group has been lost. The third part of my thesis explores the relationship between phylogenetic and morphological distance in southern African bats to assess whether long evolutionary histories correspond to equally high levels of morphological variation, as predicted by a 3 neutral model of character evolution. I find no such evidence; on the contrary weak negative trends are detected for this group, as well as in simulations of both neutral and convergent character evolution.

Finally, I ask whether spatial and climatic niche occupancy in southern African bats is influenced by evolutionary history or not. I relate divergence time between species pairs to climatic niche and range overlap and find no evidence for clear phylogenetic structuring. I argue that this may be due to particularly high levels of micro-niche partitioning.

4 Résumé

Comprendre la distribution de la biodiversité représente un enjeu majeur pour la conservation de la nature. Les analyses se basent le plus souvent sur la modélisation de la niche écologique à travers l’étude des relations causales entre la distribution spatiale des organismes et leur environnement. Depuis peu, l'étude de l'histoire évolutive des organismes est également utilisée dans les domaines de l'écologie et de la conservation. En combinaison avec la modélisation de la distribution spatiale des organismes, cette nouvelle approche fournit des informations pertinentes pour mieux comprendre l'origine des patterns de biodiversité actuels, de la structuration des communautés et des risques potentiels d'extinction.

Cette thèse explore plusieurs grandes questions écologiques, en combinant les domaines de la modélisation de la niche et de la phylogénétique. Elle s’applique aux composants importants de la biodiversité de l'Afrique australe. Les objectifs de cette thèse ont été 1) de comparer différentes mesures de la biodiversité, 2) d'évaluer l’impact des changements climatiques à venir sur la perte de diversité phylogénétique, 3) d’analyser le lien potentiel entre diversité phylogénétique et diversité morphologique et 4) d’étudier le rôle potentiel de la phylogénie sur la structuration des niches macro-climatiques des espèces.

La première partie de cette thèse fournit une comparaison spatiale, et une quantification du chevauchement, entre des prévisions de richesse spécifique et des prédictions de la diversité phylogénétique pour l'une des familles de plantes les plus riches en espèces de la région floristique du Cap (CFR), les Proteaceae. Il résulte des analyses que plusieurs mesures de diversité phylogénétique montraient des distributions spatiales différentes de la richesse spécifique, habituellement utilisée pour édicter des mesures de conservation.

La deuxième partie évalue les effets potentiels des changements climatiques attendus sur les taux d’extinction d’animaux et de plantes de l'Afrique australe. Pour cela, des modèles de distribution d’espèces actuels et futurs ont permis de déterminer si l'extinction des espèces se traduira par une plus grande ou une plus petite perte de diversité phylogénétique en

5 comparaison à un processus d'extinction aléatoire. Les résultats ont effectivement montré que l'extinction des espèces liées aux changements climatiques pourrait entraîner une perte plus grande de diversité phylogénétique. Cependant, cette perte ne serait plus grande que celle liée à un processus d’extinction aléatoire qu’à partir d’une forte perte de taxons dans chaque groupe.

La troisième partie de cette thèse explore la relation entre distances phylogénétiques et morphologiques d’espèces de chauves-souris de l’Afrique australe. Il s’agit plus précisément de déterminer si une longue histoire évolutive correspond également à des variations morphologiques plus grandes dans ce groupe. Cette relation est en fait prédite par un modèle neutre d'évolution de caractères. Aucune évidence de cette relation n’a émergé des analyses. Au contraire, des tendances négatives ont été détectées, ce qui représenterait la conséquence d'une évolution convergente entre clades et des niveaux élevés de cloisonnement pour chaque clade.

Enfin, la dernière partie présente une étude sur la répartition de la niche climatique des chauves-souris de l’Afrique australe. Dans cette étude je rapporte temps de divergence évolutive (ou deux espèces ont divergé depuis un ancêtre commun) au niveau de chevauchement de leurs niches climatiques. Les résultats n’ont pas pu mettre en évidence de lien entre ces deux paramètres. Les résultats soutiennent plutôt l’idée que cela pourrait être dû à des niveaux particulièrement élevés de répartition de la niche à échelle fine.

6 Aknowledgements

First and foremost I would like to thank my two supervisors Antoine Guisan and Nicolas Salamin, for their guidance, their patience and their competence. Working with them was in more ways than one a privilege and I thank them for all that they have taught me in the last few years. Amongst other things, I greatly admire how they both balance very busy professional and private lives and still manage to be such thoroughly pleasant people.

My gratitude goes to the two examiners Richard Grenyer and Luca Fumagalli who kindly agreed to participate in the reviewing process of this thesis and for all the constructive criticism they already provided during the intermediate evaluation.

Amongst the people without whom this thesis would never have been written I would like to mention Robin Engler and Olivier Broennimann, I owe them a few hundred beers for their time, their patience, their infinite expertise and their British sense of humour.

I would also like to thank Peter Pearman, Julien Pottier, Gwenaëlle Lelay, Christophe Randin, Luigi Maiorano and Blaise Petitpierre for their stimulating conversation, support and advice. Pascal Vittoz, Glenn Litsios, Patricio Pliscoff, Anne Dubuis, Loïc Pellissier, Maryam Zaheri, Charlotte Ndiribe and Anna Kostikova all contributed to making one of the best working environments anyone could hope for. My labwork would not have been possible without the expertise and help of Nadia Bruyndonckx, Nelly di Marco, Dessislava Savova Bianchi, Chloe Andrey, Sabrina Joye, Pascal-Antoine Christin and Guillaume Besnard. I would like to thank the students I had over the past couple of years for helping me discover the pleasure of teaching, for being keen learners and for challenging me.

I would like to express my gratitude to all the external collaborators I had the pleasure of working with during my thesis. In particular, my thanks go to Ara Monadjem, Michael Curran and Mirjam Kopp who accompanied me on tough but fabulous bat catching trips to southern Africa.

Amongst the many people whose friendliness softened the culture shock when I first moved to Switzerland, Christophe Randin, Daniel Croll, Philippe Christe, Nicole Galland and Luc Gigord definitely stand out. My thanks also go to France Pham, Virginie Cantamessa, Felicidad Jaquiéry, Giuseppina Rota, Marinette Donadeo and Corinne Bolle who tirelessly helped me navigate the local bureaucracy and FBM doctoral school requirements.

Thanks to all of the Hotspots project students, for their energy, their drive, their eccentric ways and their quirky sense of humour, which made me feel comfortable from the start. I also thank the Hotspots Consortium for accepting me onto their programme and giving me this amazing opportunity.

A big thanks goes to Tropical Biology Association (TBA) director Rosy Trevelyan, who organized two courses I was lucky enough to go on as part of the Hotspots applied conservation training programme. Her commitment to capacity building for conservation in developing countries, her infinite energy and indomitable character are a huge inspiration. 7 Other people who have inspired me over the last few years are: Koen Meyers, Alfie Alexander, Siti Rachmania, Paul Racey, Frank Clarke and William Sutherland.

Outside the University I would like to thank Carole Revelly, Marielle Fraser, Duncan Fraser, Christine de Luca, Anne-Laure Pernet of Yogaworks, and Sarah Zahno of Khatoon Dance, who have taught me so much and made life in Lausanne that much more enjoyable. Amongst the people who made my stay in Lausanne particularly special I would like to mention Sébastian Gay, Krister Swenson, Marie-Noëlle Wurm, Daniele Fraboulet, Cedric Wurm, Christopher Cianci, Carmen Cianfrani, Federica Sandrone, Paroma Basu and Rajat Mukherjee. Amongst my dearest friends here in Lausanne it would be impossible not to mention Karen Sanguinet. Thanks for sharing pain and joy for the past few years. Saya cinta anda.

I would like to thank my parents, Anna and Julian and my two sisters, Miranda and Carolina, who have come to accept my restlessness over the past 12 years, I thank you for your encouragement, your faith, your patience and most of all for your unconditional and unwavering love.

Finally, I thank Yannick for being so loving, understanding and for being my rock during these last few months.

TABLE OF CONTENTS

Introduction 10

Chapter 1 - Spatial predictions of phylogenetic diversity challenge conservation decision making 27

Chapter 2 - Climate change effects on phylogentic diversity 50

Chapter 3 - Exploring the relationship between morphology and phylogenetic diversity 72

Chapter 4 - No macro-climatic niche conservatism in southern African bats 83

Conclusions 104

Annexes - A recent inventory of the bats of Mozambique with documentation of seven new 112 species to the country

Bats of Borneo: diversity, distributions and representation in protected areas 149

9 Introduction

Studying how the components of diversity are related to each other and spatially distributed is relevant to conservation for several reasons. Firstly, an understanding of the evolutionary mechanisms which have generated and currently rule diversity patterns is essential if we are to ensure their future through conservation. Secondly, knowledge of how particular lineages have responded to challenges in the past may help us understand how they now respond or will soon respond to environmental changes. Thirdly, the way diversity is spatially and climatically distributed can tell us a lot about species requirements, community structuring and potential vulnerability to such changes. This thesis explores the relationship between spatial and phylogenetic patterns of several biodiversity components in southern Africa, a region of high biogeographic and conservation interest.

In this introductory chapter I summarize some of the key findings on the origins of current diversity patterns in the southern African region. I then describe some of our knowledge about past and present biodiversity loss. Finally, I illustrate how the study of species distributions has gradually merged with the study of evolutionary relationships to understand why specific biodiversity patterns establish, why some taxa occupy the niches they do and why certain species may go extinct before others.

10 The southern African biodiversity hotspots Understanding the origins of diversity can assist in its protection by contrasting current and historical patterns. Ultimately such understanding can help preserve the conditions required for the establishment of diverse communities. Why are some areas so much more diverse than others? What are the respective roles of environmental and historical factors in the radiation of particularly diverse clades? We have few satisfactory answers to these and other questions, but we do know that past climate change and refugia locations have had an enormous impact on how diversity is distributed today (Moritz et al., 2005). Furthermore, we know that lineages often differ in their evolutionary responses to the same environmental history, thus complicating the use of one lineage as a surrogate model for another (Moritz et al., 2005).

Southern Africa contains 4 out of 7 biodiversity hotspots identified on the African continent (Myers et al., 2000). These are the coastal forests of eastern Africa, Mapotaland-Pondoland- Albany, the Succulent Karoo and the Cape Floristic Region (CFR). All, by definition, display very high levels of floral species richness, endemism and have lost over 70% of their original extent due to human activities. The flora of the south-western tip of southern Africa is made up of over 9,000 species in an area of 90,000 km2 and is much more speciose than would be expected from its area or latitude (Goldblatt, 1978). Endemism levels of almost 70% are comparable only to those found on islands (Linder, 2003) and most likely accounted for by the ecological and geographical isolation of the CFR (Linder, 2003). Explanations for the high species richness, resulting from extreme radiation of 33 Cape floral clades, however are harder to find (Linder et al., 1992; Linder & Hardy, 2004; Linder, 2005; Linder, 2008; Verboom et al., 2009; Valente et al., 2010).

The historical events underlying the origin of this diversity, as well as the time frame over which it occurred, have been the subject of considerable debate in the literature (Levyns, 1964; Linder et al., 1992; Linder, 2003; Linder et al., 2005). A recent study using succulent karoo- and fynbos-endemic lineages across 17 groups of plants, found that all succulent karoo-endemic lineages are less than 17.5 My old, the majority being younger than 10 My (Verboom et al., 2009). This is largely consistent with suggestions that this biome is the 11 product of recent radiation in the late Miocene (Levyns, 1964). In contrast, the even richer fynbos-endemic lineages were found to display a broader age distribution, with some lineages originating in the Oligocene, but most being more recent (Verboom et al., 2009).

The massive speciation in the Cape flora might be due to genetic isolation because of a topographically and climatically heterogeneous landscape, availability of many pollinators, a long flowering season, as well as a regular fire regime (Goldblatt, 1978; Linder & Ferguson, 1985; Linder, 1995; Bakker et al., 2005). Though all of these factors are likely to have played a role, climate continues to be considered the main trigger for this radiation (Levyns, 1964; Linder, 2003). Levyns (1964) was the first to suggest that the remarkable plant species diversity of the western Cape was the result of elevated speciation following the onset of arid climates in the area, which started around the end of the Miocene. This may have led to widespread extinction, opening a variety of empty niches into which lineages which were pre-adapted to survive summer aridity were able to diversify. However, more recent studies have estimated the start of the origin and radiation of several Cape lineages to be well before the late Miocene (Linder and Hardy, 2004; Bakker et al., 2005; Linder, 2005), when climates were presumably moister than they are at present. It is possible that much radiation may have happened in high-altitude environments which support the greatest fynbos plant species richness as well as the highest concentrations of local endemics, a pattern that may partly be a result of reduced extinction in the past (Cowling & Lombard, 2002). It is also in these environments that most of the region’s palaeoendemic taxa occur (Linder et al., 1992).

Past and present biodiversity loss Both speciation and extinction are heavily affected by climate change (Erwin, 2001; Linder, 2003; Midgley et al., 2005; Barnosky, 2008; Erwin, 2009).

All of the five mass extinction events have been related to large scale climatic changes, such as sea level fluctuations, which resulted from extensive global warming in the first mass extinction and global cooling after bolide impacts in the second mass extinction (Erwin, 2001; Erwin, 2009). During the Late Permian, a combination of drop in atmospheric oxygen and climate warming (supposedly caused by another bolide impact and subsequent volcanic 12 activity) is thought to have induced hypoxic stress and compressed altitudinal ranges to near sea level with consequent habitat fragmentation and population isolation effects (Huey & Ward, 2005).

A period of climatic oscillations that began about 1 Mya, during the Pleistocene, was characterized by glaciations alternating with episodes of glacial melting (Barnosky, 2008). The current episode of global warming can be considered as an extreme and extended interglacial period; however, most geologists treat this period as a separate epoch, the Holocene, which began ~11,000 years ago at the end of the last glaciation. The Holocene extinctions were greater than occurred in the Pleistocene, especially with respect to large terrestrial vertebrates. These are also the only major extinctions that took place when humans were on the planet and occurred during a global warming episode at a time when human populations were rapidly expanding (Fig. 1). Around 20,000 years ago megafauna biomass collapsed at the same time human biomass started increasing exponentially, reached a new lower plateau ~10,000 years ago and has not recovered (Fig. 2). Recent studies suggest that human impacts such as hunting and habitat alteration contributed in many places to extinction events, and that climate change exacerbated them (Barnosky, 2008).

Figure 1: Number of non-human magafauna species that went extinct through time plotted against estimated population growth of humans (from Barnosky, 2008).

Figure 2: Estimated biomass of humans plotted against the estimated biomass of non- human megafauna (from Barnosky, 2008).

The Holocene extinctions take on special significance in understanding the potential outcomes of similar kinds of pressures on biodiversity today: the exponential growth of human populations at the same time as the Earth is warming at unprecedented rates.

The possibility that a new mass extinction spasm is upon us has received much attention. Many scientists argue that we are either entering or in the midst of the sixth great mass extinction and that it may be largely triggered by human activities (Wilson, 1988; Leakey & Lewin, 1995).

Causes of current biodiversity loss The latest update of the IUCN Red List of Threatened Species shows that 17,291 species out of the 47,677 assessed species are threatened with extinction and that 875 are already extinct or extinct in the wild (IUCN, 2009).

The well known causes of present biodiversity loss are multiple, but almost all inextricably linked to poverty and human population growth in developing countries, as well as disproportionately high per capita resource consumption and inadequate technological advancement in wealthier countries. Together with the realisation that local actions anywhere in the world have global repercussions for biodiversity and human survival, climate change 14 has become more and more prevalent in the popular and scientific literature (Biello; Lewis, 2006; Ott et al., 2008; Levi, 2009; Pettorelli et al., 2009; Veron et al., 2009).

Climate change is a major cause of biodiversity loss in southern Africa, partly because it exacerbates the effects of land use change and introductions of exotic species. Temperatures have risen in this region by approximately 1 degree over the past 100 years, which is 0.3 degrees higher than the world average (IPCC, 2007). There is now evidence that many species are disappearing from the northern parts of their ranges. In addition, there is experimental evidence that the recorded expansion of woody invasions into grasslands and savannas may be driven by rising global CO2 concentrations (Millennium Ecosystem Assessment, 2005). The ability of native species to disperse and survive these pressures will be hampered by a severely fragmented landscape (Bomhard et al., 2005; Midgley et al., 2006).

Major losses in many southern African mammal species are predicted in the next 40 to 70 years as a result of climate change, as well as an eastward shift of mammal diversity (Thuiller et al., 2006). These results suggested that the effects of climate change on wildlife communities may be most noticeable not only as substantial loss of species from their current ranges, but also as a fundamental change in community structure, as species associations shift with influxes of new colonisers (Thuiller et al., 2006). The Cape Floristic Region and the Succulent Karoo are also predicted to lose more than 41% of endemic plant species richness and undergo 39% range reduction by 2050 (Broennimann et al., 2006).

The effects of a warming climate are magnified by human landuse. Forests and woodlands are converted to croplands and pastures at a very fast rate. Half of the southern African region consists of drylands, where overgrazing is the main cause of desertification (Millennium Ecosystem Assessment, 2005). The spread of oil palm in the upper limits of southern Africa as well as South-East Asia is another example of landuse with strong effects on local climates. African oil palm, Elaeis guineensis, is grown across more than 13.5 million ha of tropical, low-lying areas, a zone naturally occupied by moist tropical forest, one of the most biologically diverse terrestrial ecosystems on Earth (Corley & Tinker, 2003; MillenniumEcosystemAssessment, 2005). Vegetable oils are among the most rapidly

15 expanding agricultural sectors (EC, 2006), and more palm oil is produced than any other vegetable oil (Corley & Tinker, 2003). Global palm oil production increased by 55% between 2001 and 2006 (http://faostat.fao.org), prompted largely by expanding biofuel markets in the European Union (MillenniumEcosystemAssessment, 2005) and by food demand globally (EC, 2006). Some of the largest multinationals worldwide, including Nestlé, Unilever and Dove, make abundant use of palm oil in their processed food and beauty products, as it is far cheaper than any other oil on the market (Fitzherbert et al., 2008). In palm oil plantations, 85% of the pre-existing vertebrate and invertebrate communities are unable to persist and go locally extinct (Fitzherbert et al., 2008). The species lost include species with the most specialised diets, those reliant on habitat features not found in plantations, those with the smallest range sizes and those of highest conservation concern (Chung et al., 2000; Corley & Tinker, 2003; Aratrakorn S. et al., 2006). Plantation assemblages are typically dominated by a few abundant generalists, non-forest species (including alien invasives) and pests (Chung et al., 2000; Corley & Tinker, 2003; Aratrakorn S. et al., 2006).

Niche modeling meets phylogenetics Niche or species distribution modeling has traditionally been one of the most powerful tools in conservation science (Vaughan et al., 2003; Rushton et al., 2004; Guisan & Thuiller, 2005). These empirical models relate field observations to environmental predictor variables to identify current and future species distributions (Guisan & Zimmermann, 2000; Guisan & Thuiller, 2005). At the core of species distribution models is the concept of the “ecological niche”, the theoretical framework to the quantification of the relationship between species and their environment (Austin et al., 1990; Araujo & Guisan, 2006). The concept of niche as used in niche models was formalized by Hutchinson (1957) as the ensemble of environmental conditions under which populations of a species can maintain a positive growth rate. At this time an important distinction between “fundamental” and “realized” niches was made. In the “fundamental” niche abiotic factors only (such as climate and topography) are taken into account whilst both biotic (such as competition and facilitation) and abiotic factors make up the “realized” niche (Hutchinson, 1957). Since they are calibrated from field observations of species that include the effects of biotic interactions, niche models capture an approximate realized niche (Jimenez-Valverde et al., 2008). 16

Some of the major applications of niche models (Guisan & Thuiller, 2005; Franklin, 2010) include improving the likelihood of identifying the location of rare species (Engler et al., 2004; Guisan et al., 2006; Le Lay et al., in review), predicting the susceptibility of a particular area to invasive species (Thuiller et al., 2005c; Broennimann et al., 2007) and predicting how species will shift their distributions as a result of climate change (Thuiller et al., 2005b; Randin et al., 2009).

There have been substantial improvements to niche models in terms of accounting for dispersal (Engler & Guisan, 2009; Engler et al., 2009) and increasingly for species interactions (Araujo & Luoto, 2007). A major drawback of using niche models to predict future distributions is that they generally assume either no dispersal at all or unlimited dispersal (i.e. the species occupies all potentially suitable habitat; e.g. (Thomas et al., 2004; Engler et al., 2009). Inevitably these two options provide unrealistic scenarios of plant dispersal. Recently, models have started to account for a large number of parameters such as seed dispersal, evolution of a population’s reproductive potential over time, stochastic long distance dispersal events, barriers to dispersal, random population extinctions, vegetative and seed-bank resilience to environmental change and differential dispersal along rivers or roads (Engler & Guisan, 2009; Thuiller et al., 2009b).

Phylogenetics, traditionally used by systematists, is the science of species’ evolutionary relationships and their reconstruction into phylogenetic trees. The explosion of molecular phylogenetics in the last couple of decades has been triggered by the emergence of new molecular methods and statistical techniques and has been used to address a wide array of important evolutionary and ecological questions. Phylogenetics has been used to identify the presence of cryptic species (Bode et al., 2010; Schonhofer & Martens, 2010), to test speciation and biogeographic hypotheses (Moritz et al., 2005; Oliveros & Moyle, 2010; Thinh et al., 2010) and to understand why some species may be better biological invaders than others (Strauss et al., 2006). It has also been useful to retrace switches in evolutionary history, the rise of key adaptations and whether these made single or parallel appearances (Christin et al., 2007; Christin et al., 2008).

The overlap between the science of species distributions and that of species evolutionary relationships has taken several interesting directions. Firstly, niche models and phylogenies are increasingly coupled to study speciation patterns (Hugall et al., 2002; Savolainen et al., 2006; Carnaval et al., 2009; Malay & Paulay, 2009).

Secondly, species distribution data and phylogenies have been used to study the relationship between biodiversity measures. Traditionally, the units in conservation biology have been species, which provide an intuitive measure to compare biodiversity at different sites. However, species are not equivalent in the amount of evolutionary history they contribute to a community and it has been argued by many authors that they should not be considered as equal conservation units. Phylogenetics made one of its first contributions to conservation biology with the introduction of phylogenetic diversity (PD) (Faith, 1992a), a measure of diversity which takes evolutionary relationships into account. The important question of how species richness and phylogenetic diversity patterns compare (and thus whether most of the past conservation efforts based on species richness have intrinsically incorporated evolutionary history or not) has been examined by several authors with different methods. Some studies found a tight relationship between patterns of SR and PD (Rodrigues & Gaston, 2002), while others found significant discrepancies (Rissler et al., 2006; Forest et al., 2007), but in general little attention has been paid to how these two measures overlap spatially.

Thirdly, species distributions and phylogenies have been used to study niche evolution. A large body of literature still disagrees as to whether closely related species strive to partition resources by differentiating their ecological niches or whether they tend to conserve more similar niches (Peterson et al., 1999; Losos et al., 2003; Graham et al., 2004; Knouft et al., 2006; Losos, 2008a; Pearman et al., 2008). Because characters are assumed to evolve following a neutral model, most studies expect close relatives to occupy similar niches (Losos, 2008a). However, theory and practice do not always match and the evidence for this pattern in nature is limited and controversial (Peterson et al., 1999; Losos & Glor, 2003; Rice et al., 2003; Knouft et al., 2006).

Together with life history traits, phylogenies have been related to the level of threat experienced by many species (Purvis et al., 2000; Purvis et al., 2005; Davies et al., 2008; Fritz et al., 2009). Considerable attention has been devoted to investigate rarity patterns and whether extinctions within a particular clade or taxon are generally random or phylogenetically clumped (Purvis et al., 2000; Sakai et al., 2002; Pilgrim et al., 2004; Sjostrom & Gross, 2006; Vamosi & Vamosi, 2007; Vamosi & Wilson, 2008). If extinction risk were indeed mostly phylogenetically clumped as argued for some bird, mammal and plant groups (Purvis et al., 2000; Vamosi & Wilson, 2008) this could have very dramatic consequences on evolutionary history loss, especially within hotspots of diversity. So far estimates have been made for the present, but very little attention has been paid to what consequences climate change may have on future loss of evolutionary histories. Only through spatially explicit niche modeling will this be possible.

Finally, patterns of phylogenetic relatedness within communities have been widely used to infer the importance of different ecological and evolutionary processes during community assembly (Kembel, 2009) and are increasingly used in combination with niche modeling to make powerful predictions in community ecology.

Main aims and thesis structure The general aim of this thesis is to answer several questions relating to diversity patterns and evolutionary history of southern African animal and plant taxa. More specifically, my aims are to provide the first spatial comparison of species richness and phylogenetic diversity predictions, to assess how much phylogenetic diversity may be lost in the future, to ask whether there is a clear link between evolutionary history and morphology and to investigate the structure and stability of climatic niches. The thesis structure is as follows:

Chapter 1: Spatial predictions of phylogenetic diversity challenge conservation decision making I quantify spatial overlap of species richness and phylogenetic diversity predictions in an extremely diversified plant family found in the Cape region: the Proteaceae.

Chapter 2: Climate change effects on phylogenetic diversity I compare present and future predictions for several animal and plant taxa to assess how species extinctions will affect evolutionary history loss.

Chapter 3: Exploring the relationship between phylogenetic diversity and morphology I compare phylogenetic diversity measures to morphological disparity in a diverse bat community to evaluate whether phylogenetic and morphological distances can be thought of as interchangeable.

Chapter 4: Are climatic niches conserved? I present the first species level phylogeny for southern African bats and employ it to determine the extent to which spatial and climatic partitioning is influenced by evolutionary relationships.

Conclusions In this section, I recapitulate the main findings of each chapter and discuss some of the limitations, as well as how an understanding of evolutionary history may best contribute to conservation in the future.

Annexes I include two studies to which I contributed during my doctorate.

20 References

Aratrakorn S., Thunhikorn S. & Donald, P. F. (2006) Changes in bird communities following conversion of lowland forest to oil palm and rubber plantations in southern Thailand. Bird Conservation International, 16, 71-82. Araujo, M. B. & Guisan, A. (2006) Five (or so) challenges for species distribution modelling. Journal of Biogeography, 33, 1677-1688. Araujo, M. B. & Luoto, M. (2007) The importance of biotic interactions for modelling species distributions under climate change. Global Ecology and Biogeography, 16, 743-753. Austin, M. P., Nicholls, A. O. & Margules, C. R. (1990) Measurement of the Realised Qualitative Niche - Environmental Niches of 5 Eucalyptus Species. Ecological Monographs, 60, 161-177. Bakker, F. T., Culham, A., Marais, E. M. & Gibby, M. (2005) In Plant Species-Level Systematics: New Perspectives on Pattern & Process (eds F. T. Bakker, L. W. Chatrou, B. Gravendeel & P. B. Pelser), pp. 75-100. Barnosky, A. D. (2008) Megafauna biomass tradeoff as a driver of Quaternary and future extinctions. Proceedings of the National Academy of Sciences of the United States of America, 105, 11543-11548. Biello, D. Negating "Climategate" Copenhagen talks and climate science survive stolen e- mail debacle. Scientific American, 302, 16-16. Bode, S. N. S., Adolfsson, S., Lamatsch, D. K., Martins, M. J. F., Schmit, O., Vandekerkhove, J., Mezquita, F., Namiotko, T., Rossetti, G., Schon, I., Butlin, R. K. & Martens, K. (2010) Exceptional cryptic diversity and multiple origins of parthenogenesis in a freshwater ostracod. Molecular Phylogenetics and Evolution, 54, 542- 552. Bomhard, B., Richardson, D. M., Donaldson, J. S., Hughes, G. O., Midgley, G. F., Raimondo, D. C., Rebelo, A. G., Rouget, M. & Thuiller, W. (2005) Potential impacts of future land use and climate change on the Red List status of the Proteaceae in the Cape Floristic Region, South Africa. Global Change Biology, 11, 1452-1468. Broennimann, O., Thuiller, W., Hughes, G., Midgley, G. F., Alkemade, J. M. R. & Guisan, A. (2006) Do geographic distribution, niche property and life form explain plants' vulnerability to global change? Global Change Biology, 12, 1079-1093. Broennimann, O., Treier, U. A., Muller-Scharer, H., Thuiller, W., Peterson, A. T. & Guisan, A. (2007) Evidence of climatic niche shift during biological invasion. Ecology Letters, 10, 701-709. Carnaval, A. C., Hickerson, M. J., Haddad, C. F. B., Rodrigues, M. T. & Moritz, C. (2009) Stability Predicts Genetic Diversity in the Brazilian Atlantic Forest Hotspot. Science, 323, 785-789. Christin, P. A., Besnard, G., Samaritani, E., Duvall, M. R., Hodkinson, T. R., Savolainen, V. & Salamin, N. (2008) Oligocene CO2 decline promoted C-4 photosynthesis in grasses. Current Biology, 18, 37-43. Christin, P. A., Salamin, N., Savolainen, V., Duvall, M. R. & Besnard, G. (2007) C-4 photosynthesis evolved in grasses via parallel adaptive genetic changes. Current Biology, 17, 1241-1247.

21 Chung, A. Y. C., Eggleton, P., Speight, M. R., Hammond, P. M. & Chey, V. K. (2000) The diversity of beetle assemblages in different habitat types in Sabah, Malaysia. Bulletin of Entomological Research, 90, 475-496. Corley, R. H. V. & Tinker, P. B. (2003) The Oil Palm, Blackwell Science. Cowling, R. M. & Lombard, A. T. (2002) Heterogeneity, speciation/extinction history and climate: explaining regional plant diversity patterns in the Cape Floristic Region. Diversity and Distributions, 8, 163-179. Davies, T. J., Fritz, S. A., Grenyer, R., Orme, C. D. L., Bielby, J., Bininda-Emonds, O. R. P., Cardillo, M., Jones, K. E., Gittleman, J. L., Mace, G. M. & Purvis, A. (2008) Phylogenetic trees and the future of mammalian biodiversity. Proceedings of the National Academy of Sciences of the United States of America, 105, 11556-11563. EC (2006) An EU Strategy for Biofuels, Commission of the European Communities. Engler, R. & Guisan, A. (2009) MIGCLIM: Predicting plant distribution and dispersal in a changing climate. Diversity and Distributions, 15, 590-601. Engler, R., Guisan, A. & Reichsteiner, L. (2004) Predicting the distribution of rare and endangered species from occurrence and pseudo-absence data Journal of Applied Ecology, 41, 263-274. Engler, R., Randin, C. F., Vittoz, P., Czaka, T., Beniston, M., Zimmermann, N. E. & Guisan, A. (2009) Predicting future distributions of mountain plants under climate change: does dispersal capacity matter? Ecography, 32, 34-45. 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. (2009) Climate as a Driver of Evolutionary Change. Current Biology, 19, R575- R583. Faith, D. P. (1992) Conservation Evaluation and Phylogenetic Diversity. Biological Conservation, 61, 1-10. Fitzherbert, E. B., Struebig, M. J., Morel, A., Danielsen, F., Bruhl, C. A., Donald, P. F. & Phalan, B. (2008) How will oil palm expansion affect biodiversity? Trends in Ecology & Evolution, 23, 538-545. Forest, F., Grenyer, R., Rouget, M., Davies, T. J., Cowling, R. M., Faith, D. P., Balmford, A., Manning, J. C., Proches, S., van der Bank, M., Reeves, G., Hedderson, T. A. J. & Savolainen, V. (2007) Preserving the evolutionary potential of floras in biodiversity hotspots. Nature, 445, 757-760. Franklin, J. (2010) Mapping species distribution: spatial inference and prediction, Cambridge University Press, Cambridge. Fritz, S. A., Bininda-Emonds, O. R. P. & Purvis, A. (2009) Geographical variation in predictors of mammalian extinction risk: big is bad, but only in the tropics. Ecology Letters, 12, 538-549. Goldblatt, P. (1978) Analysis of the flora of southern Africa - Its characteristics, relationships and origins Annals of the Missouri Botanical Garden, 65, 369-436. Graham, C. H., Ron, S. R., Santos, J. C., Schneider, C. J. & Moritz, C. (2004) Integrating phylogenetics and environmental niche models to explore speciation mechanisms in dendrobatid frogs. Evolution, 58, 1781-1793. Guisan, A., Broennimann, O., Engler, R., Vust, M., Yoccoz, N. G., Lehmann, A. & Zimmermann, N. E. (2006) Using niche-based models to improve the sampling of rare species. Conservation Biology, 20, 501-511.

22 Guisan, A. & Thuiller, W. (2005) Predicting species distribution: offering more than simple habitat models. Ecology Letters, 8, 993-1009. Guisan, A. & Zimmermann, N. E. (2000) Predictive habitat distribution models in ecology. Ecological Modelling, 135, 147-186. Huey, R. B. & Ward, P. D. (2005) Hypoxia, global warming, and terrestrial Late Permian extinctions. Science, 308, 398-401. Hugall, A., Moritz, C., Moussalli, A. & Stanisic, J. (2002) Reconciling paleodistribution models and comparative phylogeography in the Wet Tropics rainforest land snail Gnarosophia bellendenkerensis (Brazier 1875). Proceedings of the National Academy of Sciences of the United States of America, 99, 6112-6117. Hutchinson, G. E. (1957) Population studies - Animal Ecology and Demography - Concluding remarks. Cold Spring Harbor Symposia on Quantitative Biology, 22, 415-427. IPCC (2007) Climate Change 2007 - The physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. , Cambridge University Press. Jimenez-Valverde, A., Lobo, J. M. & Hortal, J. (2008) Not as good as they seem: the importance of concepts in species distribution modelling. Diversity and Distributions, 14, 885-890. Kembel, S. W. (2009) Disentangling niche and neutral influences on community assembly: assessing the performance of community phylogenetic structure tests. Ecology Letters, 12, 949-960. Knouft, J. H., Losos, J. B., Glor, R. E. & Kolbe, J. J. (2006) Phylogenetic analysis of the evolution of the niche in lizards of the Anolis sagrei group. Ecology, 87, S29-S38. Le Lay, G., Franc, E., Engler, R. & Guisan, A. (in review) Using habitat-suitability models enhances chances to find rare species in the field Ecography. Leakey, R. & Lewin, R. (1995) The Sixth Extinction: patterns of life and the future of humankind Anchor Books, New York. Levi, M. A. (2009) Copenhagen's Inconvenient Truth How to Salvage the Climate Conference. Foreign Affairs, 88, 92-104. Levyns, M. R. (1964) Migrations and Origin of the Cape Flora. Transactions of the Royal Society of South Africa, 37, 85–107. Lewis, S. L. (2006) Tropical forests and the changing earth system. Philosophical Transactions of the Royal Society B-Biological Sciences, 361, 195-210. Linder, H. P. (1995) Setting conservation priorities - The importance of endemism and phylogeny in the southern African orchid genus Herschelia. Conservation Biology, 9, 585- 595. Linder, H. P. (2003) The radiation of the Cape flora, southern Africa. Biological Reviews, 78, 597-638. Linder, H. P. (2005) Evolution of diversity: the Cape flora. Trends in Plant Science, 10, 536-541. Linder, H. P. (2008) Plant species radiations: where, when, why? Philosophical Transactions of the Royal Society B-Biological Sciences, 363, 3097-3105. Linder, H. P. & Ferguson, I. K. (1985) On the pollen morphology and phylogeny of the Restionales and Poales Grana, 24, 65-76. Linder, H. P. & Hardy, C. R. (2004) Evolution of the species-rich Cape flora. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 359, 1623-1632. Linder, H. P. & Hardy, C. R. (2005) In Plant Species-Level Systematics: New Perspectives on Pattern & Process (eds F. T. Bakker, L. W. Chatrou, B. Gravendeel & P. B. Pelser), pp. 47-73. 23 Linder, H. P., Meadows, M. E. & Cowling, R. M. (1992) In The Ecology of Fynbos: Nutrients, Fire and Diversity (ed R. M. Cowling), pp. 113-134. Oxford University Press, Cape Town. Losos, J. B. (2008) Phylogenetic niche conservatism, phylogenetic signal and the relationship between phylogenetic relatedness and ecological similarity among species. Ecology Letters, 11, 995-1003. Losos, J. B. & Glor, R. E. (2003) Phylogenetic comparative methods and the geography of speciation. Trends in Ecology & Evolution, 18, 220-227. Losos, J. B., Leal, M., Glor, R. E., de Queiroz, K., Hertz, P. E., Schettino, L. R., Lara, A. C., Jackman, T. R. & Larson, A. (2003) Niche lability in the evolution of a Caribbean lizard community. Nature, 424, 542-545. Malay, M. C. D. & Paulay, G. (2009) Peripatric speciation drives diversification and distributional pattern of reef hermit crabs (Decapoda: Diogenidae: Calcinus) Evolution, 64, 634-662. Midgley, G. F., Hughes, G. O., Thuiller, W. & Rebelo, A. G. (2006) Migration rate limitations on climate change-induced range shifts in Cape Proteaceae. Diversity and Distributions, 12, 555-562. Midgley, G. F., Reeves, G. & Klak, C. (2005) In Phylogeny and Conservation (eds A. Purvis, J. L. Gittleman & T. Brookes), pp. 230-242. Cambridge University Press. MillenniumEcosystemAssessment (2005) World Resources Institute. Moritz, C., Hoskin, C., Graham, C. H., Hugall, A. & Moussalli, A. (2005) In Phylogeny and Conservation (eds A. Purvis, J. L. Gittleman & T. Brookes), pp. 243-264. Cambridge University Press. 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. Oliveros, C. H. & Moyle, R. G. (2010) Origin and diversification of Philippine bulbuls. Molecular Phylogenetics and Evolution, 54, 822-832. Ott, H. E., Sterk, W. & Watanabe, R. (2008) The Bali roadmap: new horizons for global climate policy. Climate Policy, 8, 91-95. Pearman, P. B., Guisan, A., Broennimann, O. & Randin, C. F. (2008) Niche dynamics in space and time. Trends in Ecology & Evolution, 23, 149-158. Peterson, A. T., Soberon, J. & Sanchez-Cordero, V. (1999) Conservatism of ecological niches in evolutionary time. Science, 285, 1265-1267. Pettorelli, N., Katzner, T., Gordon, I., Garner, T., Mock, K., Redpath, S. & Gompper, M. (2009) Possible consequences of the Copenhagen climate change meeting for conservation of animals. Animal Conservation, 12, 503-504. Pilgrim, E. S., Crawley, M. J. & Dolphin, K. (2004) Patterns of rarity in the native British flora. Biological Conservation, 120, 161-170. Purvis, A., Agapow, P. M., Gittleman, J. L. & Mace, G. M. (2000) Nonrandom extinction and the loss of evolutionary history. Science, 288, 328-330. Purvis, A., Cardillo, M., Grenyer, R. & Collen, B. (2005) In Phylogeny and Conservation (eds A. Purvis, J. L. Gittleman & T. Brooks), pp. 295-316. Randin, C. F., Engler, R., Normand, S., Zappa, M., Zimmermann, N. E., Pearman, P. B., Vittoz, P., Thuiller, W. & Guisan, A. (2009) Climate change and plant distribution: local models predict high-elevation persistence. Global Change Biology, 15, 1557-1569.

24 Rice, N. H., Martinez-Meyer, E. & Peterson, A. T. (2003) Ecological niche differentiation in the Aphelocoma jays: a phylogenetic perspective. Biological Journal of the Linnean Society, 80, 369-383. Rissler, L. J., Hijmans, R. J., Graham, C. H., Moritz, C. & Wake, D. B. (2006) Phylogeographic lineages and species comparisons in conservation analyses: A case study of california herpetofauna. American Naturalist, 167, 655-666. Rodrigues, A. S. L. & Gaston, K. J. (2002) Maximising phylogenetic diversity in the selection of networks of conservation areas. Biological Conservation, 105, 103-111. Rushton, S. P., Ormerod, S. J. & Kerby, G. (2004) Newparadigms for modelling species distributions? Journal of Applied Ecology, 41, 193-200. Sakai, A. K., Wagner, W. L. & Mehrhoff, L. A. (2002) Patterns of endangerment in the Hawaiian flora. Systematic Biology, 51, 276-302. Savolainen, V., Anstett, M. C., Lexer, C., Hutton, I., Clarkson, J. J., Norup, M. V., Powell, M. P., Springate, D., Salamin, N. & Baker, W. J. (2006) Sympatric speciation in palms on an oceanic island. Nature, 441, 210-213. Schonhofer, A. L. & Martens, J. (2010) Hidden Mediterranean diversity: Assessing species taxa by molecular phylogeny within the opilionid family Trogulidae (Arachnida, Opiliones). Molecular Phylogenetics and Evolution, 54, 59-75. Sjostrom, A. & Gross, C. L. (2006) Life-history characters and phylogeny are correlated with extinction risk in the Australian angiosperms. Journal of Biogeography, 33, 271-290. Strauss, S. Y., Webb, C. O. & Salamin, N. (2006) Exotic taxa less related to native species are more invasive. Proceedings of the National Academy of Sciences of the United States of America, 103, 5841-5845. Thinh, V. N., Mootnick, A. R., Geissmann, T., Li, M., Ziegler, T., Agil, M., Moisson, P., Tilo, N., Walter, L. & Roos, C. (2010) Mitochondrial evidence for multiple radiations in the evolutionary history of small apes. Bmc Evolutionary Biology, 10. Thomas, C. D., Cameron, A., Green, R. E., Bakkenes, M., Beaumont, L. J., Collingham, Y. C., Erasmus, B. F. N., de Siqueira, M. F., Grainger, A., Hannah, L., Hughes, L., Huntley, B., van Jaarsveld, A. S., Midgley, G. F., Miles, L., Ortega-Huerta, M. A., Peterson, A. T., Phillips, O. L. & Williams, S. E. (2004) Extinction risk from climate change. Nature, 427, 145-148. Thuiller, W., Broennimann, O., Hughes, G., Alkemade, J. R. M., Midgley, G. F. & Corsi, F. (2006) Vulnerability of African mammals to anthropogenic climate change under conservative land transformation assumptions. Global Change Biology, 12, 424-440. Thuiller, W., Lafourcade, B., Engler, R. & Araujo, M. B. (2009) BIOMOD - a platform for ensemble forecasting of species distributions. Ecography, 32, 369-373. Thuiller, W., Lavorel, S., Araujo, M. B., Sykes, M. T. & Prentice, I. C. (2005a) Climate change threats to plant diversity in Europe. Proceedings of the National Academy of Sciences of the United States of America, 102, 8245-8250. Thuiller, W., Richardson, D. M., Pysek, P., Midgley, G. F., Hughes, G. O. & Rouget, M. (2005b) Niche-based modelling as a tool for predicting the risk of alien plant invasions at a global scale. Global Change Biology, 11, 2234-2250. Valente, L. M., Reeves, G., Schnitzler, J., Mason, I. P., Fay, M. F., Rebelo, T. G., Chase, M. W. & Barraclough, T. G. (2010) Diversification of the African genus Protea (PROTEACEAE) in the cape biodiversity hotspot and beyond: equal rates in different biomes. Evolution, 64, 745-759.

25 Vamosi, J. C. & Vamosi, S. M. (2007) Body size, rarity, and phylogenetic community structure: insights from diving beetle assemblages of Alberta. Diversity and Distributions, 13, 1-10. Vamosi, J. C. & Wilson, J. R. U. (2008) Nonrandom extinction leads to elevated loss of angiosperm evolutionary history. Ecology Letters, 11, 1047-1053. Vaughan, I. P., Ormerod, S. J. & Kerby, G. (2003) Improving the quality of distribution models for conservation by addressing shortcoming in the field collection of training data Conservation Biology, 17, 1601-1611. Verboom, G. A., Archibald, J. K., Bakker, F. T., Bellstedt, D. U., Conrad, F., Dreyer, L. L., Forest, F., Galley, C., Goldblatt, P., Henning, J. F., Mummenhoff, K., Linder, H. P., Muasya, A. M., Oberlander, K. C., Savolainen, V., Snijman, D. A., van der Niet, T. & Nowell, T. L. (2009) Origin and diversification of the Greater Cape flora: Ancient species repository, hot-bed of recent radiation, or both? Molecular Phylogenetics and Evolution, 51, 44-53. Veron, J. E. N., Hoegh-Guldberg, O., Lenton, T. M., Lough, J. M., Obura, D. O., Pearce- Kelly, P., Sheppard, C. R. C., Spalding, M., Stafford-Smith, M. G. & Rogers, A. D. (2009) The coral reef crisis: The critical importance of < 350 ppm CO2. Marine Pollution Bulletin, 58, 1428-1436. Wilson, E. O. (1988) Biodiversity, National Academy Press, Washington, D.C.

26 1. Spatial predictions of phylogenetic diversity challenge conservation decision making

Dorothea V. Pio1,2, Olivier Broennimann1, Timothy G. Barraclough3, , Gail Reeves4,5,

Anthony G. Rebelo5, Wilfried Thuiller6, Antoine Guisan1*, and Nicolas Salamin1,2

1Department of Ecology and Evolution, University of Lausanne, 1015 Lausanne, Switzerland 2Swiss Institute of Bioinformatics, University of Lausanne, 1015 Lausanne, Switzerland 3Division of Biology and NERC Centre for Population Biology, Imperial College London, Silwood Park Campus, Ascot, Berkshire SL5 7PY, UK 4Jodrell Laboratory, Royal Botanic Gardens, Kew, TW9 3DS, UK 5Protea Atlas Project, South African National Biodiversity Institute, P/Bag X7, Claremont 7735, Cape Town, South Africa 6Laboratoire d'Ecologie Alpine, CNRS, Université Joseph Fourier, BP 53, 38041 Grenoble Cedex 9, France

27 Abstract The inclusion of a measure of evolutionary history and relatedness (phylogenetic diversity) in conservation has long been argued as an important step towards preserving biodiversity in a more meaningful and comprehensive way. Some of the studies that have addressed this issue find that phylogenetic diversity patterns do not differ enough from those of species richness to justify their inclusion in conservation planning. This conclusion, however, is often reached by correlating these two measures across a series of sites without paying much attention to their spatial patterns. Here, we compared fine-scale species richness and phylogenetic diversity predictions of a diverse plant family, the Cape Proteaceae, obtained through individual species distribution models and ten different phylogenetic diversity indices. We examined their correlations, spatial patterns of overlap and performance in a complementarity algorithm. Overlap was found to vary enormously among phylogenetic indices, but discrepancies existed for most measures when considering realistic amounts of land set aside for conservation. Climate explained in part the segregation of particularly species rich versus phylogenetically rich areas. In view of our results, the gradual breakdown in the species concept and an increased availability of molecular data, we encourage conservation prioritization to take advantage of the additional information provided by phylogenetic diversity.

Keywords: phylogenetic diversity, species richness, Proteaceae, spatial overlap, South Africa, conservation planning, predictive modeling, Angiosperms.

Contribution to the project: I carried out the analyses in collaboration with O.B. and N.S., produced figures and wrote the paper. This paper is currently in review

28 Introduction

Allocation of funds for nature conservation relies heavily on prioritisation exercises. The budgets in most environmental organizations are very limited, making the use of the most meaningful criteria a number one priority in the design of protected area networks (Bottrill et al., 2008). In an effort to provide more realistic and comprehensive examples for conservation practice, several authors have argued for the inclusion of costs, ecosystem services, potential human-wildlife conflicts and other socio-economic factors (Moore et al., 2004; Eigenbrod et al., 2009). However, the primary purpose of these prioritization exercises is the identification of the most biologically rich and unique areas still existing today. In this regard, the inclusion of evolutionary history in conservation, through the calculation of phylogenetic diversity has long been argued as an important step towards preserving biodiversity in a more meaningful and comprehensive way (Faith, 1992a). Species are not equal in the amount of evolutionary history they bring to their community and should not, as such, be considered equivalent conservation units. Biodiversity hotspots (Myers et al., 2000) for example contain a higher proportion of species characterized by exceptionally long and unique evolutionary histories (Sechrest et al., 2002). If phylogenetic diversity patterns were found to match those of species richness, there would be no reason to use phylogenetic diversity measures in conservation prioritization, as species richness will always be easier, cheaper and quicker to measure. Some studies have found a tight relationship between patterns of species richness and phylogenetic diversity (Rodrigues & Gaston, 2002; Schipper et al., 2008), while others have found significant discrepancies (Rissler et al., 2006; Forest et al., 2007). However, as a general rule it seems that species richness is a bad surrogate for phylogenetic diversity only when species restricted to species poor areas correspond to the ancient branches of an “unbalanced” tree (i.e. containing long ancient branches which account for a disproportionate amount of phylogenetic diversity; Rodrigues & Gaston, 2002).

Phylogenetic methods have mostly been applied across a limited number of systems and spatial scales, and often at the genus rather than species levels (Rodrigues & Gaston, 2002; Forest et al., 2007; Proches et al., 2009) with notable exceptions (Winter et al., 2009). Many 29 global studies use incomplete and coarse data to identify areas for conservation. While these exercises may be useful for resource allocation at a country level, the conclusions they reach and their use at a finer geographical scale are limited. If we are to incorporate phylogenetic diversity information into practical conservation prioritization efforts, it is of prime importance that we test whether phylogenetic diversity measures are congruent with species richness, using appropriate spatially-explicit species level data. The most recent and one of the most thorough studies examining the relationship between species richness and phylogenetic diversity found a decoupling of taxon richness and phylogenetic diversity for plant genera in the Cape Floristic Region (Forest et al. 2007). Furthermore, by means of a complementarity algorithm, this study illustrated that within a conservation planning context, gains in phylogenetic diversity are poorly matched by gains in taxon richness (Forest et al. 2007).

In this study, our aim was to assess the relationship between spatial predictions of species richness and several phylogenetic diversity indices by investigating how correlated they are and by examining their spatial patterns of overlap. Since there is considerable variation in the way evolutionary history is measured and we wanted this analysis to be as comprehensive as possible, we employed all ten phylogenetic diversity measures recently listed by Schweiger et al (2008). Our aim was also to conduct the first species level analysis (as conservation still mostly operates on this scale) and to relate potential discrepancies between species richness and phylogenetic diversity patterns to environmental gradients present in the study area. As a model group, we use the Proteaceae, an ancient Gondwanan plant family with fossils attributed to extant genera from the mid-Cretaceous (Drinnan et al., 1994; Dettmann & Jarzen, 1996). This group is found in South-Africa’s Cape Floristic Region, a biodiversity hotspot containing one of the highest levels of species richness and endemism of any known tropical or temperate area (Myers et al., 2000; Linder, 2003). These extremely diverse, low- growing shrubs and trees include over 330 species (Cowling & Lamont, 1998), and present a wide variety of pollination and fire survival strategies (Rebelo, 2001). Of the 13 genera occurring in mainland Africa, ten are almost entirely endemic to the fynbos vegetation of the south-western Cape (Barker, 2002). The Cape Floristic Region contrasts with other high- diversity areas, such as tropical forests as it is made up of dissimilar local communities, in

30 which most species are relatively abundant and very few are rare (Latimer et al., 2005). This pattern can be explained by examining migration rates in the fynbos, which are two orders of magnitude lower than in tropical forests, and speciation rates of this vegetation type, which are higher than in any previously studied plant system (Latimer et al., 2005). The interesting evolutionary history, high diversity and excellent quality of both genetic and occurrence data available for Proteaceae in South Africa make this group an ideal model for the study of spatial patterns of overlap between phylogenetic diversity measures and species richness.

Materials and Methods

Predicting species distributions Niche modeling was employed to obtain all-inclusive and wide-ranging predictions of likely species distributions at a fine scale in the Cape Floristic Region. Though the occurrence data for this group is extensive and of excellent quality, its coverage does not include 100% of the regions of the Cape Floristic Region. Niche modeling was therefore necessary to provide a probability distribution of the occurrence of each Proteaceae species over the whole Cape Floristic Region. We built species distribution models at a resolution of 1’ × 1’ (~1.6 × 1.6 km at this latitude) for 168 endemic or near endemic Proteaceae species (the availability of both occurrence and genetic data was necessary in order to include species in the study) and occurring in more than 20 mapping cells. Species distribution data were taken from the Protea Atlas Project (PAP) database, comprising field-determined species presence and absence observations at more than 40,000 geo-referenced locations. Generalised Additive Models (Hastie & Tibshirani, 1990) were calibrated in the Splus-based BIOMOD application (Thuiller, 2003) using seven bioclimatic variables. These variables were derived from the Worldclim database for the Cape region and included annual evapo-transpiration, evapotranspiration of the wettest quarter, annual precipitation, precipitation of the wettest quarter (May to August), precipitation of the driest quarter (November to February), annual temperature and temperature of the coldest quarter (May to August). A random sample of the initial data (70%) and a stepwise selection methodology (forwards and backwards) were employed to identify the best model using the Akaike information criterion (AIC) as a selection criterion. The predictive power of each model was evaluated on the remaining 30% 31 of the initial dataset using the values obtained for the area under the curve (AUC) of a receiver operating characteristic (ROC) plot (Fielding & Bell, 1997).

The probabilities of occurrence were filtered with a measure of anthropogenic disturbance, the “human footprint”, considered as a regionally consistent way to represent land transformation on a global scale (Sanderson et al., 2002).

Predictions for individual species distribution models were summed at each site to obtain species richness predictions, which were in turn used to calculate corresponding values for various phylogenetic diversity indices. Modeled distributions were therefore the basis for both species richness and phylogenetic diversity predictions used throughout this study.

Calculation of phylogenetic diversity indices: A calibrated phylogenetic tree for the Proteaceae family based on 23 genes was assembled from pre-existing data (ITS, Reeves, Barraclough et al, unpublished data) and all other available sequences for the South African (and some Australian) Proteaceae in GenBank (McMahon & Sanderson, 2006). The tree comprising 284 species was built using MrBayes 3.1.2 (Huelsenbeck et al., 2001). Two runs of four Markov chain Monte Carlo chains were run for 10 mio generations using the GTR+Gamma model of DNA evolution (as determined by likelihood ratio tests) and default priors. The convergence of the two runs was assessed using Tracer (Drummond & Rambaut, 2007). The tree with the highest posterior probability was then dated with a penalized likelihood method (Sanderson, 2002) as implemented in the ape package (Paradis et al., 2004) in R using previously described fossils (Sauquet et al., 2009). To check the consistency of the date estimates, we also ran penalized likelihood on 100 randomly sampled trees from the posterior distribution given by MrBayes.

Phylogenetic diversity values for each of the grid cells on the map of the study area were calculated using each of the measures listed in Schweiger et al. (2008). Calculations were carried out with scripts in R based on the ape package (Paradis et al., 2004). These measures include topology indices, which are based on node information only (W and Q) and pairwise-

32 distance (J, F, AvTD, TTD and Dd) as well as minimum-spanning-path indices (PDroot, PDnode, AvPD), which are based on both branch length and node information. Moreover, indices used in this study, can be subdivided into total indices (Q, W, PDnode, PDroot, F, TTD, Dd), which add the evolutionary history of all species present in an area and averaged measures (AvTD, J, AvTD), where total evolutionary history is divided by the number of species present. Details on the mathematical properties of each of these measures can be found in a summary table in Schweiger et al (2008).

Discrepancy values Species richness and phylogenetic diversity indices were first normalized. This consisted for each of the two measures in subtracting the mean value and then dividing it by the standard deviation calculated from each grid cell. Species richness was then subtracted from phylogenetic diversity to obtain discrepancy values. Where these values were above zero, phylogenetic diversity was greater than species richness and where they were below zero, phylogenetic diversity was smaller than species richness.

Comparison between species richness and phylogenetic diversity by correlation, spatial overlap and complementarity algorithm: In order to describe the relationship between species richness and phylogenetic diversity in the study area, a Spearman correlation was used for each phylogenetic diversity measure. In addition, we ran the complementarity algorithm developed by Forest et al. (2007), a traditional approach in reserve selection. This algorithm chooses the most diverse grid cell first and sequentially adds grid cells with the highest complementary diversity (gain) until all diversity is represented. This analysis investigates how gains in phylogenetic diversity or species richness may change as a function of which measure is maximized and whether sites selected by maximizing phylogenetic diversity or species richness overlap spatially. Finally, we quantified the spatial overlap between phylogenetic diversity and species richness measures when considering increasing amounts of land set aside for conservation. For increasing percentages of land considered, the richest grid cells as measured by species richness and phylogenetic diversity indices were identified. The spatial overlap was then calculated as the percent of common grid cells among those identified by both measures,

33 paying particular attention to the values obtained for average amounts of land set aside for conservation (UNEP-WCMC, 2008).

PCA of environmental variables and quadratic regression A PCA of the environmental variables used to predict individual species distributions was calculated in the R package ade4 (Franquet et al., 1995). The scores corresponding to higher normalized phylogenetic diversity or species richness values were highlighted in different colors to identify possible spatial segregation between the two groups of scores. Following this analysis we conducted a polynomial quadratic regression to describe the relationship between altitude and discrepancy values.

Results Correlations, complementarity analysis and discrepancy values Species distribution model accuracy was consistently excellent with an average AUC of 0.98 over all species (range: 0.88-0.99). Spearman rho coefficients of correlations between species richness and phylogenetic diversity varied greatly between phylogenetic diversity indices. Topology measures scored high in their correlation to species richness (0.98 to 0.99 Spearman rho for W and Q respectively), while methods using both node and branch length information showed considerable differences and ranged from -0.75 to 0.94 for minimum- spanning-path methods (0.92, 0.94 and -0.72 for PDnode, PDroot and AvPD respectively) and from -0.06 to 0.99 for pairwise distance methods (-0.06, -0.03, 0.99, 0.98, and 0.7 for AvTD, J, F, TTD and Dd respectively). Graphical checks (data not shown) of these relationships indicated that two of these correlations were linear (W and TTD). Others were upward sloping asymptotically (PDnode, PDroot and Dd), downward sloping asymptotically (W and F) and some showed no correlation to species richness (AvPD, AvTD and J).

The complementarity algorithm showed that gains in different phylogenetic diversity measures did not match each other closely when complementarity in pixels added was maximized for species richness (Fig. 1).

Figure 1 – Complementarity analysis. Gains in species richness and several phylogenetic diversity indices when a complementarity algorithm is run maximizing species richness. The gains for the sites selected are normalized for each measure.

This was also not the case when complementarity was maximized for phylogenetic diversity indices (data not shown). Moreover, when maximizing gains for each measure separately and plotting the sites selected for each measure on a map of the Cape Floristic Region, no overlap existed between species richness and phylogenetic diversity sites (data not shown). Gains in Q (a topology measure) were the only ones to match species richness gains closely. Those for a minimum-spanning-path (PDroot) were poorly predicted by species richness. One of the pairwise-distance indices (Dd) followed an even more unpredictable trend, with a decrease in values when the second and third cells were added. Finally, the two averaged methods showed completely conflicting patterns with species richness with their values decreasing as the number of cells were added (Fig. 1). Normalizing and subtracting species richness from phylogenetic diversity revealed areas of discrepancy. The spatial patterns of discrepancy between species richness and phylogenetic diversity varied considerably depending on the index used (Fig. 2).

Figure 2 - Discrepancy maps between species richeness and different phylogenetic diversity indices. In red are all the areas where phylogenetic diversity is greater than species richness, whilst areas in blue are areas where species richness is greater than phylogenetic diversity (both are normalized). The Spearman correlation (Rho) between species richness and phylogenetic diversity and the percent of grid cells where phylogenetic diversity is greater than species richness are indicated.

Both of the topology indices (W and Q) and the pairwise distance measure F showed the areas harboring unexpectedly high phylogenetic diversity values to be congruent with the most significant species richness hotspot (c in Fig.3). On the other hand, most of the minimum spanning path and pairwise distance indices (e.g. PDroot, AvPD, J, AvTD, and Dd) identified more peripheral areas to the main species richness hotspot as having higher than expected levels of evolutionary history. These peripheral areas included the Koebeeberge mountains in the northern portion of the Cederberg range (Fig. 3a) and the more low-lying areas between Knysna and Port-Elizabeth (Fig. 3d) in particular, as well as the area comprising parts of the Cederberg, KoueBokkeveld and Groot Winterhoek mountains (Fig. 3b.

Figure 3 - Predicted species richness for the Proteaceae of the Cape Floristic Region. Areas of special interest which are identified and discussed throughout this study are: the Koebeeberge mountains, in the northern portion of the Cederberg range (a), parts of the Cederberg, KoueBokkeveld and Groot Winterhoek mountains (b), the Hawekwas, Hottentots Holland and Kogelberg Mountains, the Cape Peninsula and the Agulhas plain (c), the areas between Knysna and Port Elizabeth (d).

37 Spatial overlaps for high diversity areas The overlap between species richness and various measures of evolutionary history varied with the amount of surface area considered, but it did not increase linearly with it (Fig. 4). The highest level of overlap when selecting the average amount of land set aside for conservation, for example, was experienced with the use of the two topology methods, Q and W (93% and 89% respectively) and two of the pairwise distance methods, F and TTD (95% and 93% respectively). Intermediate levels of overlap with species richness patterns were experienced in one of the pairwise distance methods, Dd (53%) and in two of the minimum spanning distance methods, PDroot and PDnode (79% and 78% respectively), and no overlap at all was experienced with two pairwise distance methods and the remaining minimum spanning method (J, AvTD and AvPD respectively).

Figure 4 - Percentage of overlap between species richness and different evolutionary history patterns against amount of land cover considered. The change in overlap between predicted species richness and phylogenetic diversity indices when different percentages of the landscape are set aside for conservation. The overlap values obtained for realistic percentages of the “richest” land cover to be set aside for conservation are highlighted within the black box.

PCA of environmental variables and quadratic regression 38 The PCA-based gradient analysis of the six environmental variables employed to predict species distributions (and consequently species richness and evolutionary history) was used to investigate the differences in climatic features between areas where phylogenetic diversity was higher than species richness and vice-versa (Fig. 5). The scores corresponding to these grid cells are climatically separated along the axes of the PCA (Fig. 5a,b,c). Segregation between these points was more evident in phylogenetic diversity measures which correlate poorly with species richness (e.g. AvTD, AvPD and J, Fig. 5c), and less so in phylogenetic diversity indices which correlate highly with species richness (e.g. TTD, F, W and Q, Fig 5a). In general however scores corresponding to grid cells with higher than expected evolutionary history were associated with higher temperatures and evapo-transpiration and lower rates of precipitation (Fig 5). Altitude explained 31%, 11% and 9% of the variation in the discrepancies between TTD (t(27,243)=-76.82, p<0.001, Fig.6a), PDroot (t(27,243)=- 40.32, p<0.001, Fig.6b) and AvTD (t(27,243)=-28.12, p<0.001, Fig.6c) respectively.

Figure 5 - Principal component analysis of the environmental variables used to predict patterns of species richness and phylogenetic diversity. The seven environmental variables in the analysis are: average annual evapo-transpiration (Evtr0112), average evapo-transpiration between May and August (Evtr0508), average annual temperature (Temp0112), average temperature between May and August (Temp0508), average annual precipitation (Prec0112), average precipitation between June and August (Prec0508) and average precipitation between November and February (Prec1102) (5d). Principal components 1, 2 and 3 explain respectively 74, 17 and 8% of the variability. Scores corresponding to grid cells where phylogenetic diversity is greater than species richness are highlighted in red, whilst those where species richness is greater than phylogenetic diversity are in blue.

Figure 6 – The relationship between altitude and discrepancy values – A quadratic regression using altitude as the explanatory variable and the discrepancy between species richness and three phylogenetic diversity indices (TTD in a, PDroot in b and AvTD in c) as response variables.

Discussion We found considerable variation in how different phylogenetic diversity indices correlated with species richness. The investigation of spatial patterns, both from a complementarity algorithm perspective and a spatial overlap approach revealed that only topology-based phylogenetic diversity indices can be truly considered interchangeable with species richness. Moreover, we find that regions highlighted preferentially by species richness or phylogenetic diversity are, to some extent, segregated climatically and spatially. Given that the Proteaeceae are part of a unique system with low migration rates and high beta-diversity, our results are of particular relevance to this and other similar biodiversity hotspots.

Correlations Highly significant correlations between species richness and phylogenetic diversity were often but not always matched by an equally significant level of overlap in spatial patterns. An important distinction exists amongst functions which calculate the evolutionary history of species living in a particular area. Non-averaged phylogenetic indices sum the evolutionary history across all species, whilst averaged phylogenetic diversity measures divide the sum of evolutionary paths by a function of the number of species present, ultimately representing the mean evolutionary history brought by each species. This property removes the relationship between species richness and these phylogenetic diversity indices (Schweiger et al. 2008). Because species richness is somehow included in non-averaged phylogenetic

41 diversity measures, a positive relationship is to be expected, but the exact nature of this relationship is likely to vary between different classes of non-averaged indices. Although Rodrigues and Gaston (2002) found a linear relationship between genera richness and phylogenetic diversity (using PDroot), our results showed that this may not always be the case. Tree shape, a very influential feature in phylogenetic analyses (Mooers & Heard, 2002) has also been shown to have an effect on the correlation between taxon richness and phylogenetic diversity (Rodrigues et al., 2005) and the addition of a single species from a heavily imbalanced tree, for example, will have a disproportionate effect on phylogenetic diversity. It is also likely that diversification rates among lineages will have a large influence on the correlation with species richness. For example, areas including clades which originated from recent diversification bursts (and thus characterized by many closely related taxa with shorter branches) should show lower expected phylogenetic diversity compared to areas inhabited by a relatively higher proportion of older monotypic clades.

Discrepancies – the roles of climate and space Soil characteristics, seasonal fire and climate regimes and pollinator specificity are amongst some of the favorite candidates used to explain the unusual diversity found in the Cape Floristic Region (Goldblatt & Manning, 2002; Linder, 2003). Colonization and climatic history are regarded as some of the major factors explaining current distribution of diversity within this exceptional region (Engler, 1904; Hedberg, 1965; Hedberg, 1970; Bergh & Linder, 2009). In this study found that areas harboring higher than expected species richness were characterized by higher precipitation and lower rates of evapotranspiration (Fig. 5). These results were strengthened by the observation that higher elevations (normally associated with higher precipitation and lower evapotranspiration regimes) supported unusually high numbers of species (Fig.6).

Mountains in the Cape region have long been described as hotspots of diversity (Myers et al., 2000; Linder, 2003; Linder & Hardy, 2004; Linder, 2005). Cowling and Lombard (2002) for example attributed the high levels of species richness and endemism to reduced summer aridity of these environments (Cowling & Lombard, 2002). Several authors attributed their high diversity to a variety of repeated dispersal and colonization events from regions such as

42 the Mediterranean, the Northern and Southern Hemispheric temperate regions and the Cape Floristic Region itself (Engler, 1904; Hedberg, 1965; Hedberg, 1970). It may be possible to explain these unexpectedly high levels of species richness in mountain areas with the fact that they are characterized by allopatric speciation processes, which took place in particularly stable climates and which underwent very low extinction rates compared to lower lying regions (Dynesius & Jansson, 2000; Lawes et al., 2000). Moreover, mountains provide a sharp environmental gradient, where species are able to take advantage of the shorter migration distances to re-colonize suitable habitats and survive climatic fluctuations (Loarie et al., 2009). Therefore, relatively stable mountain climates (yet diverse from a micro-climate perspective) may have allowed species to persist and evolve through time undisturbed, displaying today most of the original members in each clade. Lower lying regions on the other hand may have been subject to much higher extinction rates, and may contain today only the surviving members of once much more diverse clades, thus accounting for higher amounts of unique evolutionary history.

Spatial patterns of overlap When considering portions of the landscape which may realistically be set aside for conservation there was considerable difference in levels of spatial overlap, with topology indices always scoring high and averaged indices always scoring low. Non-averaged (minimum-spanning-path and pairwise-distance) indices mostly displayed intermediate levels of overlap. The kind of index used can therefore influence greatly both the areas considered to represent large amounts of evolutionary history and our decision to include such measures in conservation planning altogether or to simply resort to species richness.

How to select the right index? We agree with Schweiger et al. (2008) that there is no overall best phylogenetic diversity index, but simply different situations to which certain indices may be better than others. What kinds of factors should we take into account when selecting a phylogenetic diversity index? And what may the advantages and disadvantages be of a high correlation with species richness? Averaged indices were consistently found to have the lowest correlations with species richness and lowest levels of spatial overlap. This was not particularly surprising as

43 their very aim is to eliminate the effect of taxon richness (Schweiger et al. 2008). In theory, a low correlation with species richness is a very desirable property for a phylogenetic diversity measure, as it will maximize the effect of phylogeny and therefore provide an unbiased measure of evolutionary history. From a practical point of view however, if phylogenetic diversity information is to be included in a complementarity algorithm we discourage conservation practitioners from using averaged measures. Our results showed (Fig. 1) that the increase in phylogenetic diversity gained by adding a new site is heavily counter-weighted by the number of new species that will be added ultimately causing a reduction in averaged phylogenetic diversity estimated. The mathematical properties of averaged methods are such that maximising diversity and representing all parts of the phylogenetic tree is impossible through basic complementarity algorithms.

Ultimately, the choice of index and whether to include a phylogenetic diversity index at all depends on what conservation efforts are trying to prioritise. Presently, a large complement of inter-specific genetic diversity as option value for the future is both still a desirable and sensible feature to preserve (Mooers et al., 2005b; Forest et al., 2007; Cadotte et al., 2009). Though feature diversity of very old monotypic lineages is always likely to be higher, it may very well also be more susceptible to going extinct (Purvis et al., 2000) and depending on what is driving extinction, impossible to retain in the longterm. On the other hand, we have evidence that evolutionary rich communities have higher levels of productivity in terms of biomass (Cadotte et al., 2009) suggesting that high phylogenetic diversity not only preserves options for the future, but is also important in maintaining ecosystem function. If, for example, evidence accumulates suggesting that species richness is a main driver for ongoing speciation processes our priorities may change and the focus may shift back to species richness alone. For the time being, however, maximising high species richness and phylogenetic diversity simultaneously seems the safest option. Using a combination of species richness and a non-averaged, maximum-spanning-path or pairwise-distance method index, such as PDroot or Dd will allow us to select for both high species richness and high overall phylogenetic diversity.

44 Choosing a more accurate index if possible is also an important factor to consider. If the data available allow it, indices using branch lengths should be favored, as they are by definition more accurate and complete, compared to measures relying solely on topology (Crozier, 1997; Schweiger et al., 2008). Finally, phylogenetic indices can be combined with other information, for example rarity or endemism in order to identify geographical concentrations of evolutionary history (Rosauer et al., 2009; Cadotte & Davies, 2010).

Conclusions What would the potential consequences for conservation be if discrepancies between species richness and phylogenetic diversity were found to be high in many groups? In the most extreme of cases species-rich areas resulting from recent radiations and representing high levels of phylogenetic clustering would always be spatially segregated from species-poor areas with high phylogenetic diversity and old lineages representing high levels of phylogenetic overdispersion. In this case and given issues of limited resource availability, habitat connectivity and species’ dispersal abilities, conservation practice would have to choose to channel resources using either one or the other of these two measures. This kind of scenario however is likely to be extremely rare and we predict that there should often be some overlap, as was in this case, between areas of high species richness and high phylogenetic diversity.

In light of our results, of the gradual breakdown of the species concept and of common taxonomic uncertainty, as well as the increased availability of molecular data (Mace et al., 2003; Rosauer et al., 2009), we strongly recommend the inclusion of phylogenetic data, not as a replacement but as a complement to species richness in conservation, because more comprehensive, accurate and detailed understanding of diversity patterns can only lead us to better informed decisions.

45 Acknowlegments DP is funded by the Marie-Curie Early Stage Researcher (ESR) Fellowship as part of the EU-Hotspots project (http://www.kew.org/hotspots) to AG and NS. NS is funded by a Swiss National Science Foundation grant (3100AO-116412). We thank the Swiss Institute of Bioinformatics for access to the Vital-IT high performance computing facilities. Thanks go to J. Pottier and R. Engler for helpful suggestions and advice on how to calculate spatial overlap. We are indebted to R. Grenyer and J. Davies for the use of a complementarity script published in Forest et al. (2007). We are very much indebted to 3 anonymous referees who greatly improved a previous version of this paper.

References Barker, G. M. (2002) Phylogenetic diversity: a quantitative framework for measurement of priority and achievement in biodiversity conservation. Biological Journal of the Linnean Society, 76, 165-194. Bergh, N. G. & Linder, H. P. (2009) Cape diversification and repeated out-of-southern- Africa dispersal in paper daisies (Asteraceae-Gnaphalieae). Molecular Phylogenetics and Evolution, 51, 5-18. Bottrill, M. C., Joseph, L. N., Carwardine, J., Bode, M., Cook, C., Game, E. T., Grantham, H., Kark, S., Linke, S., McDonald-Madden, E., Pressey, R. L., Walker, S., Wilson, K. A. & Possingham, H. P. (2008) Is conservation triage just smart decision making? Trends in Ecology & Evolution, 23, 649-654. Cadotte, M. W., Cavender-Bares, J., Tilman, D. & Oakley, T. H. (2009) Using Phylogenetic, Functional and Trait Diversity to Understand Patterns of Plant Community Productivity. Plos One, 4. Cadotte, M. W. & Davies, T. J. (2010) Rarest of the rare: advances in combining evolutionary distinctiveness and scarcity to inform conservation at biogeographical scales. Diversity and Distributions, 16, 376-385. Cowling, R. M. & Lamont, B. B. (1998) On the nature of Gondwanan species flocks: Diversity of Proteaceae in Mediterranean south-western Australia and South Africa. Australian Journal of Botany, 46, 335-355. Cowling, R. M. & Lombard, A. T. (2002) Heterogeneity, speciation/extinction history and climate: explaining regional plant diversity patterns in the Cape Floristic Region. Diversity and Distributions, 8, 163-179. Crozier, R. H. (1997) Preserving the information content of species: Genetic diversity, phylogeny, and conservation worth. Annual Review of Ecology and Systematics, 28, 243- 268. Dettmann, M. E. & Jarzen, D. M. (1996) Pollen of proteaceous-type from latest Cretaceous sediments, southeastern Australia. Alcheringa, 20, 103-160.

46 Drinnan, A. N., Crane, P. R. & Hoot, S. B. (1994) Patterns of Floral Evolution in the Early Diversification of Non-Magnoliid Dicotyledons (Eudicots). Plant Systematics and Evolution, 93-122. Drummond, A. & Rambaut, A. (2007) BEAST: Bayesian evolutionary analysis by sampling trees. Bmc Evolutionary Biology, 7, 8. Dynesius, M. & Jansson, R. (2000) Evolutionary consequences of changes in species' geographical distributions driven by Milankovitch climate oscillations. Proceedings of the National Academy of Sciences of the United States of America, 97, 9115-9120. Eigenbrod, F., Anderson, B. J., Armsworth, P. R., Heinemeyer, A., Jackson, S. F., Parnell, M., Thomas, C. D. & Gaston, K. J. (2009) Ecosystem service benefits of contrasting conservation strategies in a human-dominated region. Proceedings of the Royal Society B- Biological Sciences, 276, 2903-2911. Engler, A. (1904) On the vegetation conditions of Somaliland. Sitzungsberichte Der Koniglich Preussischen Akademie Der Wissenschaften, 355-416. Faith, D. P. (1992) Conservation Evaluation and Phylogenetic Diversity. Biological Conservation, 61, 1-10. Fielding, A. H. & Bell, J. F. (1997) A review of methods for the assessment of prediction errors in conservation presence/absence models. Environmental Conservation, 24, 38-49. Forest, F., Grenyer, R., Rouget, M., Davies, T. J., Cowling, R. M., Faith, D. P., Balmford, A., Manning, J. C., Proches, S., van der Bank, M., Reeves, G., Hedderson, T. A. J. & Savolainen, V. (2007) Preserving the evolutionary potential of floras in biodiversity hotspots. Nature, 445, 757-760. Franquet, E., Doledec, S. & Chessel, D. (1995) Using multivariate analyses for separating spatial and temporal effects within species-environment relationships. Hydrobiologia, 300, 425–431. Goldblatt, P. & Manning, J. C. (2002) Plant diversity of the Cape Region of southern Africa. Annals of the Missouri Botanical Garden, 89, 281-302. Hastie, T. J. & Tibshirani, R. J. (1990) Generalised Additive Models, Chapman & Hall, London. Hedberg, O. (1965) Afroalpine flora elements. Webbia, 19, 519–529. Hedberg, O. (1970) Evolution of the afroalpine flora. Biotropica, 2, 16-23. Huelsenbeck, J., Ronquist, F., Nielsen, R. & Bollback, J. (2001) Bayesian inference of phylogeny and its impact on evolutionary biology. Science, 294, 2310-2314. Latimer, A. M., Silander, J. A. & Cowling, R. M. (2005) Neutral ecological theory reveals isolation and rapid speciation in a biodiversity hot spot. Science, 309, 1722-1725. Lawes, M. J., Eeley, H. A. C. & Piper, S. E. (2000) The relationship between local and regional diversity of indigenous forest fauna in KwaZulu-Natal Province, South Africa. Biodiversity and Conservation, 9, 683-705. Linder, H. P. (2003) The radiation of the Cape flora, southern Africa. Biological Reviews, 78, 597-638. Linder, H. P. (2005) Evolution of diversity: the Cape flora. Trends in Plant Science, 10, 536-541. Linder, H. P. & Hardy, C. R. (2004) Evolution of the species-rich Cape flora. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 359, 1623-1632. Loarie, S. R., Duffy, P. B., Hamilton, H., Asner, G. P., Field, C. B. & Ackerly, D. D. (2009) The velocity of climate change. Nature, 462, 1052-U1111. Mace, G. M., Gittleman, J. L. & Purvis, A. (2003) Preserving the Tree of Life. Science, 300, 1707-1709.

47 McMahon, M. M. & Sanderson, M. J. (2006) Phylogenetic supermatrix analysis of GenBank sequences from 2228 papilionoid legumes. Systematic Biology, 55, 818-836. Mooers, A. H. & Heard, S. B. (2002) Using tree shape. Systematic Biology, 51, 833-834. Mooers, A. O., Heard, S. B. & Chrostowski, E. (2005) In Phylogeny and Conservation (eds A. Purvis, J. L. Gittleman & T. Brooks), pp. 120-138. Moore, J., Balmford, A., Allnutt, T. & Burgess, N. (2004) Integrating costs into conservation planning across Africa. Biological Conservation, 117, 343-350. 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. Paradis, E., Claude, J. & Strimmer, K. (2004) APE: Analyses of Phylogenetics and Evolution in R language. Bioinformatics, 20, 289-290. Proches, S., Forest, F., Veldtman, R., Chown, S. L., Cowling, R. M., Johnson, S. D., Richardson, D. M. & Savolainen, V. (2009) Dissecting the plant-insect diversity relationship in the Cape. Molecular Phylogenetics and Evolution, 51, 94-99. Purvis, A., Agapow, P. M., Gittleman, J. L. & Mace, G. M. (2000) Nonrandom extinction and the loss of evolutionary history. Science, 288, 328-330. Rebelo, A. G. (2001) Proteas, a fieldguide to the Proteas of Southern Africa. , Fernwood Press, Cape Town. Rissler, L. J., Hijmans, R. J., Graham, C. H., Moritz, C. & Wake, D. B. (2006) Phylogeographic lineages and species comparisons in conservation analyses: A case study of california herpetofauna. American Naturalist, 167, 655-666. Rodrigues, A. S. L., Brooks, T. M. & Gaston, K. J. (2005) In Phylogeny and conservation (eds A. Purvis, J.L. Gittleman & T. M. Brooks), pp. 101-119. Cambridge Univ. Press,, Cambridge Rodrigues, A. S. L. & Gaston, K. J. (2002) Maximising phylogenetic diversity in the selection of networks of conservation areas. Biological Conservation, 105, 103-111. Rosauer, D., Laffan, S. W., Crisp, M. D., Donnellan, S. C. & Cook, L. G. (2009) Phylogenetic endemism: a new approach for identifying geographical concentrations of evolutionary history. Molecular Ecology, 18, 4061-4072. Sanderson, E. W., Jaiteh, M., Levy, M. A., Redford, K. H., Wannebo, A. V. & Woolmer, G. (2002) The human footprint and the last of the wild. Bioscience, 52, 891-904. Sanderson, M. J. (2002) Estimating absolute rates of molecular evolution and divergence times: a penalized likelihood approach. Molecular Biology and Evolution, 19, 101-109. Sauquet, H., Weston, P. H., Anderson, C. L., Barker, N. P., Cantrill, D. J., Mast, A. R. & Savolainen, V. (2009) Contrasted patterns of hyperdiversification in Mediterranean hotspots. Proceedings of the National Academy of Sciences of the United States of America, 106, 221-225. Schipper, J., Chanson, J. S., Chiozza, F., Cox, N. A., Hoffmann, M., Katariya, V., Lamoreux, J., Rodrigues, A. S. L., Stuart, S. N., Temple, H. J. & et al (2008) The status of the world's land and marine mammals: Diversity, threat, and knowledge. Science, 322, 225- 230. Schweiger, O., Klotz, S., Durka, W. & Kuhn, I. (2008) A comparative test of phylogenetic diversity indices. Oecologia, 157, 485-495. Sechrest, W., Brooks, T. M., da Fonseca, G. A. B., Konstant, W. R., Mittermeier, R. A., Purvis, A., Rylands, A. B. & Gittleman, J. L. (2002) Hotspots and the conservation of evolutionary history. Proceedings of the National Academy of Sciences of the United States of America, 99, 2067-2071. 48 Thuiller, W. (2003) BIOMOD - optimizing predictions of species distributions and projecting potential future shifts under global change. Global Change Biology, 9, 1353- 1362. UNEP-WCMC (2008) State of the world’s protected areas: an annual review of global conservation progress. UNEP-WCMC, Cambridge, UK. Winter, M., Schweiger, O., Klotz, S., Nentwig, W., Andriopoulos, P., Arianoutsou, M., Basnou, C., Delipetrou, P., Didziulis, V., Hejda, M. & et al. (2009) Plant extinctions and introductions lead to phylogenetic and taxonomic homogenization of the European flora. Proceedings of the National Academy of Sciences of the United States of America, 106, 21721-21725.

49 2. Climate change effects on phylogenetic diversity

Dorothea V. Pio1,2, Nicolas Salamin1,2, Robin Engler1, Ara Monadjem3, Fenton PD Cotterill4, Peter Taylor5, Peter Linder6, Timothy G. Barraclough7, Gail Reeves7,8, Anthony G. Rebelo8, & Antoine Guisan1

1Department of Ecology and Evolution, University of Lausanne, 1015 Lausanne, Switzerland 2Swiss Institute of Bioinformatics, University of Lausanne, 1015 Lausanne, Switzerland 3All Out Africa Research Unit, Department of Biological Sciences, University of Swaziland, Private Bag 4, Kwaluseni, Swaziland 4AEON – African Earth Observatory Network, Departments of Geological Sciences, and Molecular and Cell Biology, University of Cape Town, Rondebosch 7701, South Africa 5Durban Natural Science Museum, P. O. Box 4085, Durban, South Africa and Dept of Ecology, and Resource Management, School of Environmental Sciences, University of Venda, P/Bag X5050, Thohoyandou, 0950, South Africa 6Institute for Systematic Botany, University of Zurich, Zollikerstrasse 107, 8008 Zurich, Switzerland 7Division of Biology and NERC Centre for Population Biology, Imperial College London, Silwood Park Campus, Ascot, Berkshire SL5 7PY, UK 7Jodrell Laboratory, Royal Botanic Gardens, Kew, TW9 3DS, UK 8Protea Atlas Project, South African National Biodiversity Institute, P/Bag X7, Claremont 7735, Cape Town, South Africa

50 Abstract Much attention has been paid to the effects of climate change on species’ extinction rates and range reduction; there is however surprisingly little information on how climate change driven extinctions may impact the tree of life and the loss of phylogenetic diversity (PD). Some plant families and mammal orders have shown non-random extinction patterns, but many other plant families have not. Are these discrepancies due to taxon specific differences or to how many species a group may have lost already? We combine phylogenetic analyses with species distribution modeling on two of the largest plant families in the Cape Floristic Region (Proteaceae and Restionaceae), as well as the second most diverse mammal order in southern Africa (Chiroptera), to answer this question. We model current and future species distributions to assess species threat levels over the next 70 years, and then compare projected to random PD survival. We find that surviving PD is significantly lower under predicted than under random extinction simulations, but only once a very large percentage of the taxa in question have been lost. We suggest that non-random extinction patterns and subsequently lower-than-expected PD survival are likely to occur for many groups, but that detectability of such patterns may depend on the number of species already lost.

Keywords: climate change, phylogenetic diversity, extinction, Proteaceae, Restionaceae, Chiroptera, southern Africa, predictive modeling.

Contribution to the project: I carried out the analyses in collaboration with R.E, produced the figures and wrote the paper. This manuscript is currently in preparation.

51 Introduction The effects of climate change on the distribution, abundance and migration of plant and animal species are already apparent both in terrestrial and marine ecosystems (Parmesan & Yohe, 2003; Perry et al., 2005) and these effects are predicted to become even more striking in the coming decades (Thomas et al., 2004; Thuiller et al., 2005a; Broennimann et al., 2006). Experimental warming confirms these predictions, and suggests that there could be rapid losses of plant diversity and a takeover by competitive neophytes (Gedan & Bertness, 2009).

In southern Africa, climate warming is particularly daunting, as its effects are magnified by human demographic pressure on intact landscapes with widespread and burgeoning landuse change that converts wildlands into agroecolandscapes (Vetter, 2009; Willis, 2009). Together, these two factors have been leading to the gradual desertification of huge portions of the landscape (Vetter, 2009). Subtropical thickets are turning into pseudo-savannas of scattered woody species, where tree mortality rates exceed recruitment. In the Karoo, heavy grazing and drought have led to the loss of palatable shrubs and increased dominance by unpalatable woody species and annuals (Vetter, 2009; Willis, 2009). A significant portion of miombo savanna woodlands across south-central Africa have been converted to agroecolandscapes since the mid 20th century and this trend continues (Campbell, 1996; Willis, 2009).

Over the past decade, several authors have predicted that species richness would be reduced and turnover rates would increase in southern African biota as a result of global change (Midgley et al., 2003; Bomhard et al., 2005; Broennimann et al., 2006; Midgley et al., 2006; Thuiller et al., 2006; Midgley & Thuiller, 2007). However, similarities in life history traits, associations to particular environments, susceptibility to habitat reduction and other anthropogenic pressures mean that species’ cannot be considered as independent entities anylonger. In many cases taxa may well be influenced by their shared evolutionary history and display differential probabilities of extinction between lineages (Purvis et al., 2000). The consequences on future levels of biodiversity are important, yet no attention has been paid to how these processes will affect differential species survival. In particular, an open question is whether climate change will result in random distribution of threat across a phylogenetic tree. 52

If extinction will follow a non-random pattern, we would expect to lose disproportionately high amounts of PD as older branches and deeper nodes are pruned from their respective phylogenetic trees. Equally, if threatened species mostly occurred in ancient monotypic taxa we would also expect a higher than random loss in phylogenetic diversity. Some authors have found evidence for non-random extinction processes in bird, mammal and angiosperm groups (Purvis et al., 2000; Sjostrom & Gross, 2006; Vamosi & Wilson, 2008). For mammals, threat status has been shown to be related both to anthropogenic pressures and life history traits, so it makes sense that extinction may not be random (Davies et al., 2008). As temperature rise is predicted to continue in the next few decades, the kinds of pressures facing species’ survival will increase and may result in complex extinction patterns.

We chose two of the four major components of South Africa's Cape flora (Proteaceae and Restionaceae) and the second most diverse mammal order in southern Africa (Chiroptera) to make predictions on how extinction and PD may be affected by climate change in 70 years time. More specifically, our aims were firstly to identify future hotspots of PD for these three groups. Secondly, we investigated whether the extinction of taxa predicted to become threatened by 2080 would result in higher or lower surviving PD than that observed under random extinction processes, thus providing the first assessment of potential climate change effects on the probability of survival between lineages of different taxonomic groups.

Methods Species and environmental datasets Species occurrence datasets were assembled for three southern African animal and plant groups: Proteaceae, Restionaceae and Chiroptera. Species occurrence data for 66 out of 75 species of Chiroptera were obtained from museum records, as well as from several unpublished surveys and represent the most comprehensive distribution dataset for Southern African bat species existing today with 88% of all species’ sampled. The occurrence dataset which makes up the Protea Atlas was used to model over 230 out of 330 species from this family (69% of total species’). An unpublished dataset for the Restionaceae (Linder

53 et al, unpublished) was used to model 180 out of 297 species of this family (60% of all species). Environmental data for calibrating the Chiroptera models was obtained from the CRU CL 2.0 dataset (New et al., 2000) that has a grid size of 10’ × 10’ (~ 16 × 16 km at this latitude). We used six uncorrelated variables (correlation coefficient < 0.7) representing the major climatic gradients in Africa, namely: mean annual potential evapo-transpiration, annual growing-degree days, minimum temperature of the coldest month, maximum temperature of the warmest month, mean annual temperature and annual sum of precipitation. For the Proteaceae and Restionaceae, we used a finer resolution of 1’ × 1’ (~1.6 × 1.6 km at this latitude), since the precision of the occurrence data allowed for finer scale predictions. We used a set of seven custom variables, adapted from the Worldclim database for the Cape region, namely, annual evapo-transpiration, evapotranspiration of the wettest quarter, annual precipitation, precipitation of the wettest quarter, precipitatation of the driest quarter, annual temperature and temperature of the coldest quarter. Future climate projections representing the potential averaged climate in 2080 were produced by perturbing the current climatic data with anomalies derived from climatic data with anomalies derived from climatic simulations produced by the HadCM3 General Circulation Model using the A1FI IPCC SRES scenarios (Nakicenovic & Swart, 2000) in accordance with globally accepted guidelines for climate impact assessment (IPCC-TGCIA 1999). In this scenario concentrations of CO2 increase from 380 ppm in 2000 to 800 ppm in 2080, and temperature rises by 3.6 K.

Species distribution modeling As our species datasets were lacking true species absences, we generated random pseudo- absences, i.e. randomly selected locations that were considered as species absences (Elith et al., 2006). For each species, pseudo-absences were generated in a number 10 times more abundant than the species' presences. Pseudo-absences were however down-weighted for model calibration as to ensure equal prevalence between presences and pseudo-absences (presences were given a weight of 1, pseudo-absences a weight of 0.1). Species distribution models were calibrated for each species using six different modeling techniques: generalized linear models, GLM (McCullagh & Nelder, 1989); generalized additive models, GAM (Hastie T & R, 1986); boosted regression trees, GBM (Ridgeway,

54 1999); random forest, RF (Breiman, 2001); multivariate adaptive regression splines, MARS (Friedman, 1991); and Classification Tree Analysis, CTA (Breiman et al., 1984). GLMs and GAMs were calibrated using a binomial distribution and a logistic link function. A bi- directional stepwise procedure was used for explanatory variable selection, based on the Akaike information criterion (Akaike, 1974). Up to second-order polynomials (linear and quadratic terms) were allowed for each explanatory variable in the GLMs, and up to third order splines in the GAMs. GBMs were calibrated with a maximum number of trees set to 5000, 5-fold cross-validation procedure to select the optimal numbers of trees to be kept and a value of 5 as maximum depth of variable interactions. Random forest models were fitted by growing 750 trees with half the numbers of available predictors sampled for splitting at each node. MARS models were fitted with a maximum interaction degree equal to 2. All models were calibrated using the BIOMOD package (Thuiller et al., 2009a) in R (Team, 2009).

The predictive power of each individual model was evaluated through a repeated data- splitting procedure (for details see Thuiller et al. 2009). A model was trained on 70% of the data before being evaluated against the remaining 30% using two measures: the area under the receiver operating characteristic curve (AUC; (Hanley JA & BJ, 1982)) and the true skill statistic (TSS; (Allouche et al., 2006). This data-splitting procedure was repeated 10 times and the evaluation values averaged. The final models used to carry-out spatial projections were calibrated using 100% of the data. To avoid using poorly calibrated models, only projections from models with AUC > 0.75 and TSS > 0.5 were considered in all subsequent analyses. Model averaging (ensemble forecast) was performed by weighting the individual model's projections respectively by their AUC or TSS scores and averaging the result, a method shown to be particularly robust (Marmion et al., 2009) Each ensemble forecast model was then reclassified into two different binary projections (i.e. the species is either projected present or absent), using the threshold that would respectively maximize jointly the percentage of presences and absences correctly predicted (43), and maximize the TSS value. Since generating pseudo-absence introduced a random component into our models, the entire modeling procedure was replicated 11 times, each time with a new set of pseudo- absences (pseudo-absence replicates). The final projected distribution of a species was then

55 obtained by combining together the projections from our 11 replicates. Two methods were used: majority and average. In the majority method, binary projections obtained from each replicate are summed and the final value is equal to the majority across the 11 replicates (e.g., if 6 replicates predicted a presence and 5 and an absence, the majority method assigns a presence). In the average method, the continuous ensemble forecasts for each pseudo- absence replicate were averaged. The resulting averaged projection was then reclassified into binary projections using the average reclassification threshold of all 11 replicates.

Phylogenetic tree building We used dated species level phylogenies for each taxonomic group.

Proteaceae: A calibrated phylogenetic tree for the Proteaceae based on 23 genes was assembled from pre-existing data (Reeves, Barraclough et al, unpublished data, (Valente et al., 2010) and all other available sequences for the South African (and some Australian) Proteaceae in GenBank) following the method of McMahon and Sanderson (2007). The tree comprising 284 taxa was built using MrBayes 3.1.2 (Huelsenbeck et al., 2001). Two runs of four Markov chain Monte Carlo chains were run for 10 million generations using the GTR+Gamma model of DNA evolution and default priors. The best-fit model was estimated through likelihood-ratio tests. The convergence of the two runs was assessed using Tracer. The tree with the highest posterior probability was then dated with a penalized likelihood method (Sanderson, 2002) as implemented in the ape package in R (Paradis et al., 2004) using the fossils described in (Sauquet et al., 2009). To check the consistency of the date estimates, we also ran penalized likelihood on 100 randomly sampled trees from the posterior distribution given by MrBayes (see Pio et al., submitted for details).

Restionaceae: This tree was published previously (Hardy et al., 2008).

Chiroptera:

56 A species level phylogeny for 89 species of bats was built using two mitochondrial markers: cytochrome b and 16S. DNA was extracted from fresh skin samples collected in Malawi and Mozambique between August and December 2007, and muscle samples obtained from the Durban Science Museum, South Africa (Annex 2). Both fresh and museum specimens were stored in 95% ethanol, and DNA was extracted using a phenolchloroform-isoamyl procedure (Sambrook et al., 1989). A set of 6 custom primers were used to amplify the two genes (Annex 1), which were sequenced using the Sanger method (Sanger et al., 1977). Existing sequences available in GenBank by 1/10/09 were also integrated in the analysis (Annex 2). Sequences were first aligned with a multiple alignment programme, ClustalX (Larkin et al., 2007) and then adjusted manually with the sequence alignment editor BioEdit (Hall, 1999). Sequences from the two genes were concatenated to form a supermatrix after checking that single gene phylogenetic analyses did not show incongruent nodes supported by more than 80% bootstrap.

Trees were built with both Maximum Likelihood (ML) and Bayesian methods. We used PhyML 2.4.4 (Guindon & Gascuel, 2003) and MrBayes 3.1.2 (Huelsenbeck et al., 2001). The best model for the combined data set was GTR+G+I as determined by likelihood ratio tests. Three mammals, the horse (Equidae, Perissodactyla), the pangolin (Manidae, Pholidota) and the sea lion (Phocidae, Carnivora) were designated as outgroups. ML bootstrap support was determined from 200 replicates with NNI branch swapping (Salamin et al., 2003). The MrBayes tree was obtained by running three independent analyses, each with four Markov chains for 50,000,000 generations with sampling performed every 1000 trees, thereby generating 50,000 sample points per analysis. Burn-in values were determined in Tracer 1.5, thus the first 5000 samples were discarded and a consensus tree with posterior probability values was obtained in TreeAnnotator 1.5.3 by pooling trees from the three independent runs (both extensions of BEAST 1.5.3, see below).

57 and variance adjusted to include the uncertainty associated with the fossil. The first constraint (corresponding to the last shared ancestor of the Chiroptera) was set to 65±5 MYA, at or following the Cretaceous-Tertiary boundary (Eick et al., 2005; Teeling et al., 2005). The second fossil constraint, corresponding to the last known common ancestor shared by the families Rhinolophidae and Hipposideridae was set to 46±5 MYA (Eick et al., 2005; Teeling et al., 2005) and a third was set to 75±10 MYA for the carnivore pangolin split (McKenna & Bell, 1997).

Red list assessment Each species was assigned to an IUCN Red list threat category (Extinct - EX, Critically Endangered - CR, Endangered - EN, Vulnerable – VU) if they met the relevant criteria shown in Table 1 (IUCN, 2009). These taxa are collectively referred to as Red List taxa or threatened taxa. We classified species based on model predictions twice, once for the present and once for 2080. For present classifications into the three threat categories, we used criterion B2a,b(ii) based on the Area of Occupancy (AOO) of a species, whilst for future classifications we used both criterion B2a,b(ii) and A3c (IUCN, 2009). Taxa that disappear from all their localities under the climate change scenarios are listed as Extinct (EX). Here we have followed a previously adopted approach (Thuiller et al., 2005b) and are aware that it represents a simplified way of classifying many species into categories and that it does not necessarily represent their actual status in the IUCN red list of threatened species (Akcakaya et al., 2006). Thus, the information generated from this analysis should be taken in its entirety, with the intention of predicting general trends for the future at a community level.

58 Table 1: IUCN Red List criteria used to classify species into three main threat categories: Critically Endangered (CR), Endangered (EN) and Vulnerable (VU).

CR EN VU CRITERION A3c

Population reduction projected or > 80% > 50% > 30% suspected to be met in the future range loss range loss range loss (up to a maximum of 100 years) based on a decline in area of occupancy (AOO)

B2a,b

Severely fragmented and declining < 10 km2 < 500 km2 <2,000 km2 present geographic range (AOO) occupancy occupancy occupancy

PD calculation Although many PD measures exist (Schweiger et al., 2008), we used rooted PD (Rodrigues & Gaston, 2002), as it is one of the most intuitive and most common ways of accounting for evolutionary history and relatedness between taxa. First, rooted PD (hereafter referred to simply as PD) was calculated for the entire tree. Threatened taxa were thus identified using the predictive models described above and assumed to be the most likely species to eventually become extinct. Threatened taxa were removed from their respective phylogenetic trees. Strictly speaking this means pruning the tree of all the unique evolutionary history represented by the threatened species in question. Following the extinction event, remaining PD was recalculated. Each predicted surviving PD value was compared to a random distribution (10,000 random species pools). Subsequently, we checked whether our observed value fell in the 5% most extreme random values (lowest).

59 Results We obtained present and future (year 2080) ensemble climatic niche models for 167 species of Proteaceae, 147 species of Restionaceae and 50 species of Chiroptera.

Models All models showed good to excellent predictive power with a mean AUC of 0.98 and 0.99 and a mean TSS value of 0.98 and 0.99, for majority and average methods respectively, estimated from the split-sample evaluation procedure. Since the AUC (ROC) and TSS-based projections with both the majority and average methods yielded extremely similar results, we hereafter present only results from the "TSS majority" method.

IUCN threat categories There was a good match between the classification criteria, with species ranking threatened with B (area of occupancy) also ranking threatened with A (predicted reduction in AOO), though A generally resulted in a higher (more endangered) listing and was therefore responsible for category assignment for most species. Though in all three groups the great majority of species were uplisted from less threatened in the present to more threatened categories in the future, only within Proteaceae were species predicted to go extinct.

Climate change effects on spatial patterns of PD Climate change resulted in very serious effects on PD spatial patterns in all three groups. The PD for Proteaceae (Fig. 1a) became concentrated in mountainous areas on the border with the Succulent Karoo (Cederberg, KoueBokkeveld and Groot Winterhoek mountains).

The PD for Restionaceae, which currently occupies most of the coastal zones of the CFR and some of the Cederberg, KoueBokkeveld and Groot Winterhoek mountains, was predicted to disappear almost entirely from the eastern half of the CFR and to concentrate in the western half, except for a small portion of the low-lying areas between Knysna and Port Elizabeth (Fig 1b).

60 Finally, the PD for Chiroptera, which today is widespread all over southern Africa, is predicted to contract to the outermost, coastal portions of this region. The Kalahari Desert is predicted to become very barren in terms of future evolutionary diversity of Chiroptera, as well as portions of the central Zambezian and southern Miombo woodland. The Kalahari Xeric savannah and Acacia woodlands also appear to become increasingly inhospitable for bats. The Maputaland Coastal mosaic, on the other hand, the succulent Karoo, portions of the Namib Desert and eastern African mountain ranges are predicted to accumulate most of the remaining PD for this group.

61 PD present 8.39 Ü PD - SR present 1 0.98 -1.9

0 50 100 200 Km

PD 2080 PD - SR 2080 8.63 1.2

0.98 -2.2

PD - SR present PD present 0.58 Ü 32.06 -0.51 0

02100 00Km 0 100 200 Km PD - SR 2080 PD 2080 b 0.684 31.71 -1 0

0 100 200 Km

Ü Ü

PD present PD - SR present 1296 0.65 05001'000250 Km 1 -0.63

PD 2080 1281 PD - SR 2080 0.81 1 -0.88

Figure 1: Predicted phylogenetic diversity (PD) and normalized discrepancy (PD minus species richness (SR)) are represented for present and future (2080) climatic conditions in Proteaceae (a), Restionaceae (b) and Chiroptera (c).

Effect of loss of threatened species on surviving PD Under the current scenario only Proteaceae species were classified in EN and CR threat categories, the extinction of which would result in an overall loss of 3.59% of species richness and 0.01% of PD. In 2080, the loss of the same threat categories (as well as EX) would result in 79% of species richness loss and 64% of PD loss. Similarly, Restionaceae and Chiroptera would undergo losses of 73% and 82% in species richness and 57% and 76% of PD respectively, though in these two groups no species were predicted to go extinct by 2080.

63 We found non-random PD survival in all three groups, but only after a very substantial portion of the taxon had been lost; 79%, 87% and 82% for Proteaceae, Restionaceae and Chiroptera respectively (Table 2). Instead no effect was found when 3% and 45% of the species were lost in Proteaceae, 17% and 73% for Restionaceae and 38% for Chiroptera.

Table 2: Significance values (for Proteaceae, Restionaceae and Chiroptera) representing the comparison between observed remaining evolutionary history and 10,000 random permutations, when different predicted IUCN threat categories (EX - Extinct, CR – critically endangered, EN – endangered and VU – vulnerable) are pruned from the tree.

present 2080

IUCN categories EX-CR-EN EX-CR EX-CR-EN EX-CR-EN-VU pruned from trees

Proteaceae 0.35 (3.5%) 0.91 (43.1%) 0.02* (79%) 0.03* (84.4%)

no species Restionaceae listed 0.7 (17%) 0.46 (73.4%) 0.02* (87.7%)

no species no additional Chiroptera listed 0.38 (20%) 0.04* (82%) species listed

* statistically significant values

Discussion This investigation predicts that hotspots of PD for Proteaceae, Restionaceae and Chiroptera will contract considerably through the next 70 years (Fig.1). In all three groups, predicted surviving PD in 2080 was smaller than would have been experienced if species had been pruned at random. However, this pattern was only visible once a very large percentage of species had been lost (Table 2).

Our results for the Proteaceae are consistent with climate change simulations previously carried out in a study which assessed the relative potential impacts of future land use and

64 climate change on the threat status of this plant family (Bomhard et al., 2005). However, we found many additional species to go extinct and to be uplisted to higher threat categories, because our simulations extend for a further 60 years into the future. Our results for Chiroptera also agree with previously published work on African terrestrial mammals (Thuiller et al., 2006). A recent study evaluated whether national parks met their mandate under future climate change and land transformation conditions by assessing the sensitivity of 277 (non-volant) mammals on the African continent. The Kalahari region is predicted by previous results to lose the highest number of species on the African continent (Thuiller et al., 2006); similarly our results predicted this region to become very barren in terms of future surviving bat PD.

Although no studies have simulated the effects of predicted climate change on PD, an early study showed that little PD is lost under even catastrophic extinction scenarios, because of the large amount of evolutionary history shared between species (Nee & May, 1997). In one of the simulations in this study, which followed a “field of bullets” model leading to a random pattern of extinction, 81% of the phylogenetic branch length remained even when only 5% of the species survived an extinction episode. Our predictions showed that likely amounts of surviving PD under similarly catastrophic events would be much lower (33.62% of surviving PD if 84% of species were lost for Proteaceae, 18.84% of surviving PD if 90% of the species were lost for bats, 20% of surviving PD if 87.75% of species were lost for Restionaceae). One difference with the Nee & May (1997) study is that extinction probabilities of species are probably not random in our data sets. The different climatic variables used to estimate the predicted habitat have a large phylogenetic component when estimated using the K statistics (Blomberg et al., 2004). The loss of PD under phylogenetically non-random extinction is expected to be much larger because the network of branches that form connections among extinct species will also be lost (Purvis et al., 2000). The consequence is then an increased PD loss, especially if internal branches are more likely to be long than terminal branches, which is the expectation under many branching models. The excessive reduction in PD when compared with random expectations could also indicate that extinctions happen in groups with long terminal branches indicating slower rates of diversification. Another difference can explain the

65 discrepancies between the two results. The shape of the real trees used in this study, which are less balanced than expected under a Yule model, could also explain the difference observed. Nee & May (1997) used coalescent trees to estimate the expected loss of evolutionary history, which have been shown to be more balanced than topologies based on real data (Heard, 19922). The frequency of species-poor lineages with distinct evolutionary histories will thus be larger in real phylogenetic trees (Heard & Mooers, 2002), which leads to a much greater loss of evolutionary history as expected by Cadotte and Davies (2010).

A considerable body of literature exists which tries to assess whether present species’ extinction risk exhibits a pattern of phylogenetic clumping, which would depict a non- random process of extinction (Bennett & Owens, 1997; Purvis et al., 2000; Sakai et al., 2002; Pilgrim et al., 2004; Sjostrom & Gross, 2006; Faith, 2008; Vamosi & Wilson, 2008). Extinctions have received attention from many researchers because they result in both a shift in community structure and interspecies interactions (Williams, 2007; Stralberg, 2009) and the deletion of portions of the history of life. Studying how extinction is distributed phylogenetically ultimately allows us to investigate to what extent communities may be disrupted and what portions of evolutionary history may be eliminated.

Several studies on birds and non-volant mammals, but more recently also on angiosperm families, have shown evidence for non-random extinction processes (Bennett & Owens, 1997; Russell et al., 1998; Purvis et al., 2000; Sjostrom & Gross, 2006; Vamosi & Wilson, 2008). Threat in birds for instance is negatively influenced by increasing body size and decreases in fecundity, which may have predisposed certain lineages to extinction (Bennett & Owens, 1997). Studies on angiosperms found that elevated risk was found in smaller clades (Vamosi & Wilson, 2008). However other studies on regional and global floral datasets found contrasting or mixed evidence for non-random threat or rarity patterns (Schwartz & Simberloff, 2001; Webb & Pitman, 2002; Pilgrim et al., 2004).

Some of the authors who found evidence for monotypic genera being more threatened, have pointed out that at an advanced stage in the extinction process, (once phylogenetically distributed traits have already mediated considerable extinction), then many monotypic

66 genera or families might be the last survivors of once-larger clades. This could lead to a higher proportion than expected of monotypic genera, or species on long phylogenetic branches, being threatened (Purvis et al., 2000). They also found the difference between observed and expected loss (the “extra” loss) to increase with the proportion of species culled (Purvis et al., 2000). Our results suggest that our three groups are therefore in an earlier stage of their extinction process, seen that presently almost no monotypic genera exist and given that we only find non-random patterns once a large percentage of each group has been lost.

Presumably, the degree of phylogenetic clumping varies along a continuum and is therefore varyingly detectable. Here we studied three southern African taxa which are not particularly heavily affected by man. Since no data exists on how future landuse changes will evolve, our models are conservative and only predict range contractions related to intrinsic factors such as the degree of environmental specialization for example (Broennimann et al., 2006). Because of intense overharvesting of bird, primate and carnivore groups, their trees are being pruned at a very fast rate, probably much higher than that of the groups used in our study and their extinction process is therefore likely to be at a much more advanced stage.

Every single new species that goes extinct will cause a greater loss of PD than any previous extinction events (Nee & May, 1997; Steel, 2006). However, the rate of PD loss is expected to reduce with time (Gernhard et al, 2008). This prediction has important consequences on conservation practice as further extinctions occur. None of our groups are actively overharvested enough in the present to show any sign of phylogenetic clumping, yet our predictions for the future reveal that once a significant portion of the taxon has been lost, and with them more ancestral and unique branches in the tree representing large amounts of evolutionary history, non-random processes will be visible. These results reveal that current and future extinctions will have consequences for the loss of evolutionary history and suggest that non-random phylogentic diversity loss may be varyingly detectable; depending on how many extinctions already occurred for the group in question.

67 Acknowledgements We would like to thank the hundreds of volunteers who collected the data which make up the Protea Atlas. We are indebted to Dr Richard Grenyer for comments on an earlier draft. DP is funded by the Marie-Curie Early Stage Researcher (ESR) Fellowship as part of the EU-Hotspots project (http://www.kew.org/hotspots) to AG and NS. NS is funded by the Swiss National Science Foundation grant no 3100A0-116412. We would like to thank the Swiss Institute of Bioinformatics for access to Vital-IT, their high performance computer center.

References

Akaike, H. (1974) A new look at statistical model identification. IEEE Transactions on Automatic Control, 19, 716-722. Akcakaya, H. R., Butchart, S. H. M., Mace, G. M., Stuart, S. N. & Hilton-Taylor, C. (2006) Use and misuse of the IUCN Red List Criteria in projecting climate change impacts on biodiversity. Global Change Biology, 12, 2037-2043. Allouche, O., Tsoar, A. & Kadmon, R. (2006) Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS). . Journal of Applied Ecology, 43, 1223-1232. Bennett, P. M. & Owens, I. P. F. (1997) Variation in extinction risk among birds: Chance or evolutionary predisposition? Proceedings of the Royal Society of London Series B-Biological Sciences, 264, 401-408. Bomhard, B., Richardson, D. M., Donaldson, J. S., Hughes, G. O., Midgley, G. F., Raimondo, D. C., Rebelo, A. G., Rouget, M. & Thuiller, W. (2005) Potential impacts of future land use and climate change on the Red List status of the Proteaceae in the Cape Floristic Region, South Africa. Global Change Biology, 11, 1452-1468. Breiman, L. (2001) Random Forests. . Machine Learning, 45, 5-32. Breiman, L., Friedman, J. H., Olshen, R. A. & Stone, C. J. (1984) Classification and regression trees, New York, NY, US. Broennimann, O., Thuiller, W., Hughes, G., Midgley, G. F., Alkemade, J. M. R. & Guisan, A. (2006) Do geographic distribution, niche property and life form explain plants' vulnerability to global change? Global Change Biology, 12, 1079-1093. Campbell, B. (1996) The Miombo in Transition: Woodlands and Welfare in Africa, CIFOR, Jakarta. Davies, T. J., Fritz, S. A., Grenyer, R., Orme, C. D. L., Bielby, J., Bininda-Emonds, O. R. P., Cardillo, M., Jones, K. E., Gittleman, J. L., Mace, G. M. & Purvis, A. (2008) Phylogenetic trees and the future of mammalian biodiversity. Proceedings of the National Academy of Sciences of the United States of America, 105, 11556-11563. Drummond, A., Ho, S., Phillips, M. & Rambaut, A. (2006) Relaxed Phylogenetics and Dating with Confidence. PLoS Biology, 4, 699-710. Drummond, A. & Rambaut, A. (2007) BEAST: Bayesian evolutionary analysis by sampling trees. Bmc Evolutionary Biology, 7, 8. 68 Eick, G. N., Jacobs, D. S. & Matthee, C. A. (2005) A nuclear DNA phylogenetic perspective on the evolution of echolocation and historical biogeography of extant bats (Chiroptera). Molecular Biology and Evolution, 22, 1869-1886. Elith, J., Graham, C. H., Anderson, R. P., Dudik, M., Ferrier, S., Guisan, A., Hijmans, R. J., Huettmann, F., Leathwick, J. R., Lehmann, A., Li, J., Lohmann, L. G., Loiselle, B. A., Manion, G., Moritz, C., Nakamura, M., Nakazawa, Y., Overton, J. M., Peterson, A. T., Phillips, S. J., Richardson, K., Scachetti-Pereira, R., Schapire, R. E., Soberon, J., Williams, S., Wisz, M. S. & Zimmermann, N. E. (2006) Novel methods improve prediction of species' distributions from occurrence data. Ecography, 29, 129-151. Faith, D. P. (2008) Threatened Species and the Potential Loss of Phylogenetic Diversity: Conservation Scenarios Based on Estimated Extinction Probabilities and Phylogenetic Risk Analysis. Conservation Biology, 22, 1461-1470. Friedman, J. (1991) Multivariate Adaptive Regression Splines. . Annals of Statistics, 19, 1- 67. Gedan, K. B. & Bertness, M. D. (2009) Experimental warming causes rapid loss of plant diversity in New England salt marshes. Ecology Letters, 12, 842-848. Guindon, S. & Gascuel, O. (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic Biology, 52. Hall, T. A. (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl. Acids. Symp. Ser., 41, 95-98. Hanley JA & BJ, M. (1982) The Meaning and Use of the Area under a Receiver Operating Characteristic (Roc) Curve. . Radiology, 143, 29-36. Hardy, C. R., Moline, P. & Linder, H. P. (2008) A phylogeny for the African Restionaceae and new perspectives on morphology's role in generating complete species phylogenies for large clades. International Journal of Plant Sciences, 169, 377-390. Hastie T & R, T. (1986) Generalized Additive Models. Statistical Science, 1, 297-318. Huelsenbeck, J., Ronquist, F., Nielsen, R. & Bollback, J. (2001) Bayesian inference of phylogeny and its impact on evolutionary biology. Science, 294, 2310-2314. IUCN (2009) IUCN Red List of Threatened Species., Version 2009.2002, Downloaded on 2024 February 2010. Larkin, M., Blackshields, G., Brown, N., Chenna, R., McGettigan, P., McWilliam, H., Valentin, F., Wallace, I., Wilm, A., Lopez, R., Thompson, J., Gibson, T. & Higgins, D. (2007) Clustal W and Clustal X version 2.0. Bioinformatics, 23, 2947-2948. Marmion, M., Parviainen, M., Luoto, M., Heikkinen, R. K. & Thuiller, W. (2009) Evaluation of consensus methods in predictive species distribution modelling. Diversity and Distributions, 15, 59-69. McCullagh, P. & Nelder, J. (1989) Generalized Linear Models, Chapman & Hall, London, UK. McKenna, M. C. & Bell, S. K. (1997) Classification of mammals above the species level. Columbia University Press, New York. Midgley, G. F., Hannah, L., Millar, D., Thuiller, W. & Booth, A. (2003) Developing regional and species-level assessments of climate change impacts on biodiversity in the Cape Floristic Region. Biological Conservation, 112, 87-97. Midgley, G. F., Hughes, G. O., Thuiller, W. & Rebelo, A. G. (2006) Migration rate limitations on climate change-induced range shifts in Cape Proteaceae. Diversity and Distributions, 12, 555-562.

69 Midgley, G. F. & Thuiller, W. (2007) Potential vulnerability of Namaqualand plant diversity to anthropogenic climate change. Journal of Arid Environments, 70, 615-628. Nakicenovic, N. & Swart, R. (2000) Emissions Scenarios: A Special Report of Working Group III of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge. Nee, S. & May, R. M. (1997) Extinction and the loss of evolutionary history. Science, 278, 692-694. New, M., Hulme, M. & Jones, P. (2000) Representing twentieth-century space-time climate variability. Part II: Development of 1901-96 monthly grids of terrestrial surface climate. Journal of Climate, 13, 2217-2238. Paradis, E., Claude, J. & Strimmer, K. (2004) APE: Analyses of Phylogenetics and Evolution in R language. Bioinformatics, 20, 289-290. Parmesan, C. & Yohe, G. (2003) A globally coherent fingerprint of climate change impacts across natural systems. Nature, 421, 37-42. Perry, A. L., Low, P. J., Ellis, J. R. & Reynolds, J. D. (2005) Climate change and distribution shifts in marine fishes. Science, 308, 1912-1915. Pilgrim, E. S., Crawley, M. J. & Dolphin, K. (2004) Patterns of rarity in the native British flora. Biological Conservation, 120, 161-170. Purvis, A., Agapow, P. M., Gittleman, J. L. & Mace, G. M. (2000) Nonrandom extinction and the loss of evolutionary history. Science, 288, 328-330. Ridgeway, G. (1999) The state of boosting. Computing Science and Statistics, 31, 172-181. Rodrigues, A. S. L. & Gaston, K. J. (2002) Maximising phylogenetic diversity in the selection of networks of conservation areas. Biological Conservation, 105, 103-111. Russell, G. J., Brooks, T. M., McKinney, M. M. & Anderson, C. G. (1998) Present and future taxonomic selectivity in bird and mammal extinctions. Conservation Biology, 12, 1365- 1376. Sakai, A. K., Wagner, W. L. & Mehrhoff, L. A. (2002) Patterns of endangerment in the Hawaiian flora. Systematic Biology, 51, 276-302. Salamin, N., Chase, M. W., Hodkinson, T. R. & Savolainen, V. (2003) Assessing internal support with large phylogenetic DNA matrices. Molecular Phylogenetics and Evolution, 27, 528-539. Sambrook, J., Fritsch, E. & Maniatis, T. (1989) Molecular cloning: a laboratory manual Cold Spring Harbor Laboratory Press, , New York. Sanderson, M. J. (2002) Estimating absolute rates of molecular evolution and divergence times: a penalized likelihood approach. Molecular Biology and Evolution, 19, 101-109. Sanger, F., Nicklen, S. & Coulsen, A. (1977) DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences, 74, 5463-5467. Sauquet, H., Weston, P. H., Anderson, C. L., Barker, N. P., Cantrill, D. J., Mast, A. R. & Savolainen, V. (2009) Contrasted patterns of hyperdiversification in Mediterranean hotspots. Proceedings of the National Academy of Sciences of the United States of America, 106, 221-225. Schwartz, M. W. & Simberloff, D. (2001) Taxon size predicts rates of rarity in vascular plants. Ecology Letters, 4, 464-469. Schweiger, O., Klotz, S., Durka, W. & Kuhn, I. (2008) A comparative test of phylogenetic diversity indices. Oecologia, 157, 485-495. Sjostrom, A. & Gross, C. L. (2006) Life-history characters and phylogeny are correlated with extinction risk in the Australian angiosperms. Journal of Biogeography, 33, 271-290.

70 Steel, M. (2006) In Reconstructing the Tree of Life: Taxonomy and Systematics of Species Rich Taxa (eds T. R. Hodkinson & J. A. Parnell), CRC Press, USA Stralberg, D., Jongsomjit, D., Howell, C. A., Snyder, M. A., Alexander, J. D., Wiens, J. A. & Root, T. L. (2009) Re-shuffling of species with climate disruption: a no- analog future for California birds? PLoS ONE, 4, e6825. Team, R. D. C. (2009) R: A language and environment for statistical computing. , R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org. Teeling, E. C., Springer, M. S., Madsen, O., Bates, P., O'Brien, S. J. & Murphy, W. J. (2005) A molecular phylogeny for bats illuminates biogeography and the fossil record. Science, 307, 580-584. Thomas, C. D., Cameron, A., Green, R. E., Bakkenes, M., Beaumont, L. J., Collingham, Y. C., Erasmus, B. F. N., de Siqueira, M. F., Grainger, A., Hannah, L., Hughes, L., Huntley, B., van Jaarsveld, A. S., Midgley, G. F., Miles, L., Ortega-Huerta, M. A., Peterson, A. T., Phillips, O. L. & Williams, S. E. (2004) Extinction risk from climate change. Nature, 427, 145-148. Thuiller, W., Broennimann, O., Hughes, G., Alkemade, J. R. M., Midgley, G. F. & Corsi, F. (2006) Vulnerability of African mammals to anthropogenic climate change under conservative land transformation assumptions. Global Change Biology, 12, 424-440. Thuiller, W., Lafourcade, B., Engler, R. & Araújo, M. (2009) BIOMOD - a platform for ensemble forecasting of species distributions. Ecography, 32, 369-373. Thuiller, W., Lavorel, S., Araujo, M. B., Sykes, M. T. & Prentice, I. C. (2005a) Climate change threats to plant diversity in Europe. Proceedings of the National Academy of Sciences of the United States of America, 102, 8245-8250. Thuiller, W., Lavorel, S., Araujo, M. B., Sykes, M. T. & Prentice, I. C. (2005b) Climate change threats to plant diversity in Europe. Proceedings of the National Academy of Sciences of the United States of America, 102, 8245-8250. Valente, L. M., Reeves, G., Schnitzler, J., Mason, I. P., Fay, M. F., Rebelo, T. G., Chase, M. W. & Barraclough, T. G. (2010) Diversification of the African genus Protea (Proteaceae) in the cape biodiversity hotspot and beyond: equal rates in different biomes. Evolution, 64, 745-759. Vamosi, J. C. & Wilson, J. R. U. (2008) Nonrandom extinction leads to elevated loss of angiosperm evolutionary history. Ecology Letters, 11, 1047-1053. Vetter, S. (2009) Drought, change and resilience in South Africa's arid and semi-arid rangelands. South African Journal of Science, 105, 29-33. Webb, C. O. & Pitman, N. C. A. (2002) Phylogenetic balance and ecological evenness. Systematic Biology, 51, 898-907. Williams, J. W. J., S. T. (2007) Novel climates, no-analog communities, and ecological surprises. Frontiers in Ecology and the Environment, 5, 475-482. Willis, J. (2009) Slavery in the Great Lakes Region of East Africa. History, 94, 378-379.

71 3. Exploring the relationship between morphology and phylogenetic diversity

Dorothea Pio1,2, Arrigo Nils1,2, Antoine Guisan1 & Nicolas Salamin1,2

1Department of Ecology and Evolution, University of Lausanne, 1015 Lausanne, Switzerland 2Swiss Institute of Bioinformatics, University of Lausanne, 1015 Lausanne, Switzerland

72 Abstract

The use of phylogenetic diversity (PD) in conservation is often justified with the need to maximize the representation of feature diversity to preserve evolutionary potential. But what is feature diversity exactly and does it embody morphological diversity? We propose to explore the relationship between evolutionary history (as measured by PD) and morphological disparity of the diverse bat fauna of southern Africa, under the expectation that larger subclades (and therefore higher levels of evolutionary history) should correspond to higher variability in morphology. Instead, we find a weak negative relationship between PD and morphological disparity and after confirming this trend with simulations under Brownian motion (BM) and Ornstein-Uhlenbeck (OU) models of character evolution, argue that PD may indeed be decoupled from morphological disparity, irrespective of the speciation processes involved.

Keywords: phylogenetic diversity, southern Africa, Chiroptera, morphology, disparity, neutral evoution.

Contribution to the project: I carried out part of the analyses in collaboration with A.N. and N.S., produced figures and wrote the paper. This manuscript is currently in preparation.

73 Introduction

The importance of biodiversity has often been interpreted in terms of its option value for the future: the greater the complement of contemporary biodiversity, the greater the possibilities for future biodiversity, because of the diverse genetic resource needed to ensure continued evolution (IUCN, 1980; Barker, 2002). Option value is thus associated with richness in the different features expressed by organisms (Faith, 1992b; Faith, 1992a; Weitzman, 1992). An individual species of greater value is one contributing more novel features to a given subset. The assumption that representing feature diversity is important to preserve evolutionary potential has gone hand in hand with the use of phylogenetic diversity (PD) (Faith, 1992a) as a biodiversity measure (Faith, 1992a; Mooers et al., 2005a; Forest et al., 2007; Cadotte & Davies, 2010).

Feature diversity, which could be interpreted as the multitude of adaptations associated with the radiation of a particular clade, is difficult to define. What components make up feature diversity? Does morphology, for example, traditionally used as a surrogate for evolutionary, functional and ecological diversity, reflect the variety of adaptations which make up feature diversity and thus evolutionary history accurately? At its origin, PD was defined as a cladistic measure for branch lengths on phylogenies, constructed using morphological characters (Faith, 1992a; Barker, 2002). PD was thus, by definition, a function of morphology. However, PD values are currently calculated from molecular phylogenies and little attention has been paid to whether PD maintains its original link to morphology, irrespective of the data used to calculate it.

The degree of morphological differentiation among taxa has been described with various disparity measures and been previously used to characterise changes in pattern of morphospace occupation (Ciampaglio et al., 2001). More specifically it has been used to study patterns of morphological variation through time with fossil data showing that morphological disparity usually peaks earlier in a clade’s history than does species diversity (Foote, 1997; Wills, 2001). Disparity measures have also been used to study the tempo and mode of radiation of extant species (Harmon et al., 2003). In one important study conducted to examine how tempo of lineage diversification relates to morphological radiation patterns, 74 taxa whose cladogenesis is concentrated early in their histories were found to partition more of their morphological disparity among, rather than within, subclades (Harmon et al., 2003).

We propose to employ such disparity measures to explore the relationship between evolutionary history (as measured by PD) and morphological variation. We expect that larger subclades (and therefore higher levels of evolutionary history) should correspond to higher variability in morphology, assuming that morphological characters of this taxon evolved following a neutral model. We propose to test this relationship with a diverse mammal order, the Chiroptera, well known for possessing a striking array in the variety of food sources, prey acquisition strategies, flight modes and echolocation design, which can generally be related to morphology (Kalko, 1998). Ear size, for example is typically related to food type and acquisition strategies as larger ear sizes, indicate that species prefer to listen for their prey and often either switch off echolocation completely or emit very weak echolocation calls to glean prey off vegetation or other surfaces (Kalko et al., 1999). Flight morphology may also be considered as a strong indicator of taxonomic membership. Species in the family Molossidae, for example are well known to forage in open spaces, above canopy and to fly at high speed, emitting loud low frequency calls to catch prey in mid-air (Norberg & Rayner, 1987; Mora et al., 2004). Their narrow pointed wings are therefore characterized by high wing loading (high weight to wing surface area ratio) (Norberg & Rayner, 1987), as opposed to species in the family Phyllostomidae for example, which feed within understory vegetation and fly at lower speeds and with much better maneuverability (Kalko et al., 1999).

Methods

Five morphological characters known to have strong associations with varying life history strategies (Norberg & Rayner, 1987; Kalko, 1998; Kalko et al., 1999; Mora et al., 2004) were used to calculate disparity among different bat subclades: forearm length, weight, ear length, wing loading and aspect ratio. Disparity was calculated from average pairwise Euclidean distances, a variance-related method of estimating the dispersion of points in multivariate space that is insensitive to sample size (Ciampaglio et al., 2001). Disparity was calculated for each node of a phylogeny. In order to ensure independence we subtracted the difference between ancestor and descendant for each pairwise euclidean distance value before averaging 75 all the values for one subclade. Evolutionary history was calculated with the 10 PD measures described in a recent study (Schweiger et al., 2008). Finally, linear models were used to describe the relationships between different phylogenetic diversity measures and morphological disparity. We repeated this procedure twice, once with a dated phylogenetic tree and once with a tree with molecular branch lengths (the trees are those described in Chapter 4).

The relationship between PD and disparity is complicated by the measure per branch that we calculate here. A basic expectation is that under Brownian motion (BM) evolution for each character, PD is positively correlated with disparity as the variance in the character increases with added evolutionary history. It is however more difficult to assess this relationship when i) the variance per unit of time increases, and ii) the evolution is not following a simple Brownian motion model. To investigate this aspect, we simulated 1000 data sets of five characters using the topology and branch lengths of the southern African bat tree. The models of evolution used were Brownian motion with variance of 0.1, 1.0 and 10.0 and Ornstein-Uhlenbeck (OU) process with variance of 0.1, 1.0 and 10.0, and strength of selection of 0.1, 1.0 and 10.0. The optimum for OU was set to the real values of the five characters for the bats. The difference in clade PD and disparity between each descendant and its ancestor was calculated for each simulated data set and the slope was estimated using linear regression.

Results

We obtained 20 linear models for observed values of morphological disparity and 10 PD measures (10 for the dated tree and 10 for the tree with molecular branch lengths) and 12,000 simulated linear models under two main models of character evolution (1,000 for each combination of parameters considered). We found observed PD and morphological disparity to have consistently negative but non-significant relationships for all measures, except for averaged ones (AvPD, J, AvTD, Fig.1; only the results from the dated tree are presented, as those from the tree with molecular branch lengths revealed exactly the same patterns). Similarly, all of our BM and OU simulations did not reveal significant relationships between PD and morphological disparity (Table 1). The only simulation showing a 76 constantly positive slope was the OU model which assumed lowest variation (0.1) and lowest strength of selection (0.1). In all but one of the simulations carried out using BM and OU models of character evolution, our observed slopes were outside the 95% confidence interval of simulated distributions, the exception here being the BM model with lowest variation.

77 Q W

0 500 1500 2500 0 102030405060 -40 -20 0 20 40 -40 -20 0 20 40

AvPD PDroot

-0.02 0.02 0.06 0.10

0123

-40 -20 0 20 40 -40 -20 0 20 40

PDnode J Phylogenetic Diversity

0123 0.00 0.10 0.20 0.30

-40 -20 0 20 40 -40 -20 0 20 40

Dd AvTD

01234 0.00 0.10 0.20 0.30 -40 -20 0 20 40 -40 -20 0 20 40 0 F TTD

0 5 10 15 20 0 200 600 1000 140 -40 -20 0 20 40 -40 -20 0 20 40

Morphological disparity

Figure 1: Observed phylogenetic diversity plotted against morphological disparity for each sub-clade in a species level phylogeny of southern African bats. 78

Table 1: Quantiles of the slope for all BM and OU simulations

Model of character evolution selection variance Quantiles 2.5% 50% 97.5%

0.1 -9.54 20.11 60.59 Brownian Motion (BM) 1 -1.02 2.01 5.84 10 -0.09 0.19 0.59

0.1 0.1 0.227369 4.177251 11.08184 0.1 1 -1.65E-06 -1.64E-06 -1.64E-06 0.1 10 -3.23E-17 -3.22E-17 -3.22E-17 1 0.1 -2.639694 -1.035647 0.565959 Ornstein-Uhlenbeck (OU) 1 1 -1.70E-06 -1.64E-06 -1.58E-06 1 10 -3.22E-17 -3.22E-17 -3.22E-17 10 0.1 -3.22E-17 -3.22E-17 -3.22E-17 10 1 -1.83E-06 -1.37E-06 -9.87E-07 10 10 -3.29E-17 -3.22E-17 -3.14E-17

Discussion

We found a negative trend between PD and amount of morphological disparity across all but the three averaged PD measures.

Under a model of continuous Brownian evolution, a positive relationship may be expected between evolutionary history and morphological accumulation, as larger clades should contain both longer histories and more variety in the types of morphologies represented. The distribution of the slope based on the simulated BM model showed a bias towards positive values (median of 20.11; Table 1), although negative slopes were also possible under this model.

Previous paleontological research and molecular systematic analyses suggest that the initial stages of evolutionary radiation often show extensive cladogenesis and rapid ecological and morphological divergence among lineages, often leading to long-term persistence of among- lineage ecomorphological differences (Schluter, 2000). However, increases in

79 ecomorphological disparity are not always limited to the earlier stages of radiation (Foote, 1993; Fortey et al., 1996; Foote, 1997; Jackman et al., 1999; Ruber & Adams, 2001). In taxa that accumulate lineages more slowly, for example, subclades have greater opportunity to diversify morphologically, and thus more of a taxon’s disparity is partitioned within, rather than among, subclades (Harmon et al., 2003). Moreover, extensive ecomorphological convergence may occur among lineages (Jackman et al., 1999; Ruber & Adams, 2001).

Significant ecomorphological convergence in bats is supported by a study on several subclades of the genus Myotis which occur in the Neartic, Neotropical and Palearctic (Ruedi & Mayer, 2001). In these subgenera, there are several pairs of morphologically and ecologically very similar species; however none of these species pairs are phylogenetically related. This study revealed several independent adaptive radiations in the genus Myotis which produced strikingly similar evolutionary solutions in different parts of the world (Ruedi & Mayer, 2001). This study highlights convergence of species occurring in geographical separate regions. Another study assessed the ecomorphological structure of fifteen species of the same bat family (Rhinolophidae) occurring in a Malaysian rainforest, and used principal component analysis to describe the morphological space occupied by these species (Kingston et al., 2000). They found this family to be non-randomly distributed in multivariate space and evidence for morphological overdispersion of the most similar species, suggesting niche differentiation in response to competition (Kingston et al., 2000). These results suggest that patterns of convergent evolution between distantly related taxa (Ruedi & Mayer, 2001) and morphological overdispersion in closely related taxa (Kingston et al., 2000) are both possible within this order.

Our simulations represent simplified versions of neutral evolution and evolution models which converge towards an optimum. In a sense they represent opposite ends of a spectrum and our data almost certainly falls between these two extremes. The lack of relationship between PD and morphological disparity was consistent irrespective of model of evolution, type of variance and strength of selection. These results suggest that the character evolution model present in a particular taxon may have only limited influence on its relationship between morphological disparity and PD.

We conclude that the accumulation of evolutionary history in clades is not well matched by corresponding higher variance in morphological characters for this group and argue that the concept of feature diversity may have to be revised. If functional disparity were to show the same relationship to PD, as displayed by morphological disparity, FD may have to be thought of in terms of only genotypic (and not phenotypic) diversity.

Acknowledgements DP is funded by the Marie-Curie Early Stage Researcher (ESR) Fellowship as part of the EU-Hotspots project (http://www.kew.org/hotspots) to AG and NS. NS is funded by the Swiss National Science Foundation grant no 3100A0-116412. We would like to thank the Swiss Institute of Bioinformatics for access to Vital-IT, their high performance computer center.

References

Barker, G. M. (2002) Phylogenetic diversity: a quantitative framework for measurement of priority and achievement in biodiversity conservation. Biological Journal of the Linnean Society, 76, 165-194. Cadotte, M. W. & Davies, T. J. (2010) Rarest of the rare: advances in combining evolutionary distinctiveness and scarcity to inform conservation at biogeographical scales. Diversity and Distributions, 16, 376-385. Ciampaglio, C. N., Kemp, M. & McShea, D. W. (2001) Detecting changes in morphospace occupation patterns in the fossil record: characterization and analysis of measures of disparity. Paleobiology, 27, 695-715. Faith, D. P. (1992a) Conservation Evaluation and Phylogenetic Diversity. Biological Conservation, 61, 1-10. Faith, D. P. (1992b) Systematics and Conservation - on Predicting the Feature Diversity of Subsets of Taxa. Cladistics-the International Journal of the Willi Hennig Society, 8, 361-373. Foote, M. (1993) Discordance and concordance between morphological and taxonomic diversity Paleobiology, 19, 185-204. Foote, M. (1997) The evolution of morphological diversity. Annual Review of Ecology and Systematics, 28, 129-152. Forest, F., Grenyer, R., Rouget, M., Davies, T. J., Cowling, R. M., Faith, D. P., Balmford, A., Manning, J. C., Proches, S., van der Bank, M., Reeves, G., Hedderson, T. A. J. & Savolainen, V. (2007) Preserving the evolutionary potential of floras in biodiversity hotspots. Nature, 445, 757-760.

81 Fortey, R. A., Briggs, D. E. G. & Wills, M. A. (1996) The Cambrian evolutionary 'explosion': Decoupling cladogenesis from morphological disparity. Biological Journal of the Linnean Society, 57, 13-33. Harmon, L. J., Schulte, J. A., Larson, A. & Losos, J. B. (2003) Tempo and mode of evolutionary radiation in iguanian lizards. Science, 301, 961-964. IUCN (1980) World conservation strategy: living resource conservation for sustainable development., Gland: IUCN-UNEP-WWF. Jackman, T. R., Larson, A., de Queiroz, K. & Losos, J. B. (1999) Phylogenetic relationships and tempo of early diversification in Anolis lizards. Systematic Biology, 48, 254-285. Kalko, E. K. V. (1998) Organisation and diversity of tropical bat communities through space and time. Zoology-Analysis of Complex Systems, 101, 281-297. Kalko, E. K. V., Friemel, D., Handley, C. O. & Schnitzler, H. U. (1999) Roosting and foraging behavior of two Neotropical gleaning bats, Tonatia silvicola and Trachops cirrhosus (Phyllostomidae). Biotropica, 31, 344-353. Kingston, T., Jones, G., Zubaid, A. & Kunz, T. H. (2000) Resource partitioning in rhinolophoid bats revisited. Oecologia, 124, 332-342. Mooers, A., Heard, S. B. & Chrostowski, E. (2005) In Phylogeny and Conservation (eds A. Purvis, J. L. Gittleman & T. Brooks), pp. 120-138. Cambridge University Press, Cambridge. Mora, E. C., Macias, S., Vater, M., Coro, F. & Kossl, M. (2004) Specializations for aerial hawking in the echolocation system of Molossus molossus (Molossidae, Chiroptera). Journal of Comparative Physiology a-Neuroethology Sensory Neural and Behavioral Physiology, 190, 561-574. Norberg, U. M. & Rayner, J. M. V. (1987) Ecological morphology and flight in bats (Mammalia, Chiroptera) - Wing adaptations, flight performance, foraging strategy and echolocation. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 316, 337-419. Ruber, L. & Adams, D. C. (2001) Evolutionary convergence of body shape and trophic morphology in cichlids from Lake Tanganyika. Journal of Evolutionary Biology, 14, 325- 332. Ruedi, M. & Mayer, F. (2001) Molecular systematics of bats of the genus Myotis (vespertilionidae) suggests deterministic ecomorphological convergences. Molecular Phylogenetics and Evolution, 21, 436-448. Schluter, D. (2000) The ecology of adaptive radiation, Oxford University Press, New York. Schweiger, O., Klotz, S., Durka, W. & Kuhn, I. (2008) A comparative test of phylogenetic diversity indices. Oecologia, 157, 485-495. Weitzman, M. L. (1992) ON DIVERSITY. Quarterly Journal of Economics, 107, 363-405. Wills, M. A. (2001) Fossils, Phylogeny, and Form, Kluwer, New York.

82 4. No macro-climatic niche conservatism in the bats of southern Africa

Dorothea V. Pio1,2, Olivier Broennimann1, Ara Monadjem3, Peter Taylor4, Fenton Cotterill5, Geeta Eick6,7, Michael Curran8,9, Mirjam Kopp8, Chloé Andey1, Sabrina Joye1, Antoine Guisan1* & Nicolas Salamin1,2*

1Department of Ecology and Evolution, University of Lausanne, 1015 Lausanne, Switzerland 2Swiss Institute of Bioinformatics, University of Lausanne, 1015 Lausanne, Switzerland 3All Out Africa Research Unit, Department of Biological Sciences, University of Swaziland, Private Bag 4, Kwaluseni, Swaziland 4Durban Natural Science Museum, P. O. Box 4085, Durban, South Africa and Dept of Ecology, and Resource Management, School of Environmental Sciences, University of Venda, P/Bag X5050, Thohoyandou, 0950, South Africa 5AEON – African Earth Observatory Network, Departments of Geological Sciences, and Molecular and Cell Biology, University of Cape Town, Rondebosch 7701, South Africa 6Center for Ecology and Evolutionary Biology 7Howard Hughes Medical Institute, University of Oregon, Eugene, Oregon 97403, USA. 8Institute for Natur-, Landschafts- und Umweltschutz (NLU), University of Basel, St. Johanns-Vorstadt 10, CH-4056, Switzerland 9Institute of Environmental Engineering, ETH Zurich, HIF C 13, Wolfgang-Pauli-Str. 15, Zurich, Switzerland

83 Abstract The idea that closely related species also share similar ecological niches has sparked interesting findings and debates in the last decade. Contrasting results come from using different model species, speciation systems, small sample sizes, but mostly from setting different null hypotheses and not making the distinction between phylogenetic signal and niche conservatism. Using the first complete interfamilial phylogeny for southern African bats we test for phylogenetic signal, by examining climatic niche similarity and equivalency between species pairs. Furthermore, we examine the relationship between divergence age, climatic niche and geographic overlap to assess whether evolutionary history has a role in shaping this group’s distribution and climatic niche partitioning. We find little evidence for phylogenetic structuring (or indeed for phylogenetic conservatism) in this group’s current spatial partitioning and climatic niche occupancy patterns and suggest that a combination of excellent fine-grained resource partitioning and dispersal ability may be responsible for this result.

Keywords: niche modeling, phylogenetic signal, Chiroptera, southern Africa,

Contribution to the project: I carried the analyses, produced figures and wrote the manuscript. This paper is currently in preparation.

84 Introduction Phylogenetics is not only important to uncover the evolutionary relationships between species, but increasingly, coupled with distribution modeling and other kinds of ecological data, provides a powerful tool to test important theories about the origins of biological diversity and the mechanisms of community structuring. For example, methods have been developed that combine species level phylogenies with geographic range data to explore alternative geographic modes of speciation (Barraclough et al., 1998; Barraclough & Vogler, 2000; Mattern & McLennan, 2000; Losos & Glor, 2003). Likewise, the combination of phylogenetic hypotheses and climate based modeling of species’ ranges has been used to infer historical distributions (Hugall et al., 2002) and whether speciation is correlated with ecological shifts (Peterson et al., 1999). Recently community ecologists have become very interested in the idea that closely related species are ecologically similar and now frequently compare the degree of phylogenetic relatedness of community members to that of species source pools to draw inferences about the processes structuring communities (Losos, 2008a). The approaches taken so far to answer this question use different taxa and speciation scenarios (Peterson et al., 1999; Losos et al., 2003; Graham et al., 2004; Knouft et al., 2006) and have come to contrasting conclusions. For example, niche conservatism was found to occur during allopatric speciation of birds, butterflies and mammals on either side of the Isthmus of Tehuantepec in Mexico (Peterson et al., 1999). However, no evidence for niche conservatism was found in a community of Caribbean lizards living in sympatry (Losos et al., 2003; Knouft et al., 2006) or for dendrobatid frogs living in the highly heterogeneous Ecuadorian Andes (Graham et al., 2004). The discrepancies observed between studies may stem from different interpretations of the basic concepts used. On the one hand, there are different null hypotheses underlying the definition of niche conservatism (Warren et al., 2008). Niche conservatism has been described as a continuum between niche equivalency (i.e. when niches of sister species are indistinguishable) and niche similarity (i.e. where niches of sister species are more similar to each other than to those of other species in the same family or phylogeny) where truth often lies somewhere in between (Warren et al., 2008). Others place it at the end of a spectrum which ranges between adaptive radiation and rapid diversification at one end and niche conservatism and slow divergence at the other (Ackerly, 2009). 85 On the other hand, discrepancies in findings can be attributed to the variety in different techniques and approaches investigating phylogenetic relatedness and ecological similarity among species (Losos, 2008b; Pearman et al., 2008; Ackerly, 2009). Phylogenetic signal exists when characters evolve in a Brownian motion-like manner, in which the amount of change in any given interval is generally small and random in direction (Losos, 2008a). Under this model, a relationship should exist between the degree of phylogenetic relatedness, quantified as the time since divergence between pairs of species, and the degree of phenotypic similarity between them; the smaller the amount of time since two species shared a common ancestor (i.e. the more closely related the two species), the less the expected phenotypic difference between them (Blomberg & Garland, 2002). The same could be expected for ecological niche differences. The niche conservatism scenario, on the other hand, suggests that some factor is causing closely related species to be more similar ecologically than would be expected by simple Brownian motion, in other words, it suggests that some process is constraining divergence among closely related species (Losos, 2008a).

Many recent studies assume a priori that phylogenetic signal in ecological characters is widespread and pervasive (Duncan & Williams, 2002; Anderson et al., 2004; Peres-Neto, 2004; Mouillot et al., 2006; Swenson et al., 2006; Johnson & Stinchcombe, 2007) though much of the available evidence does not support this assumption (Bohning-Gaese & Oberrath, 1999; Vitt et al., 1999; Lindeman, 2000; Losos et al., 2003; Graham et al., 2004; Knouft et al., 2006). It is possible, therefore that the universality of ecological phylogenetic signal and phylogenetic niche conservatism is being overstated (Losos, 2008a; Pearman et al., 2008).

Studies have been carried out in the Caribbean, the Americas and Europe (Vitt et al., 1999; Losos et al., 2003; Rice et al., 2003; Graham et al., 2004; Peres-Neto, 2004), though none to date have explored any components of the flora or fauna on the African continent, despite its well known habitat heterogeneity and high diversity levels (Myers et al., 2000; Linder, 2003). Southern Africa hosts around 100 species of bats (Taylor, 2000), one tenth of global diversity (Simmons, 2005), yet compared to their relatives on other continents or to other 86 mammal orders remain poorly studied (Monadjem et al., in review). Many factors may influence how bats are currently distributed. These may include their high dispersal abilities, the history of their speciation and the need and scale at which resources are partitioned. No studies to date have attempted to examine whether phylogeny shapes the way their communities are structured spatially nor climatically.

The aim of this study was to test whether the bats of Southern Africa show evidence for phylogenetic signal, niche conservatism or neither with the first complete species level phylogeny for this group. Our null model predicts that sister species will not occupy significantly more similar niches than those occupied by other species in this group. Pairwise species comparisons are therefore tested against the total environmental space occupied by all other species in the phylogenetic tree.

Methods Phylogenetic reconstruction A species level phylogeny for 89 species of bats was built using two mitochondrial markers: cytochrome b and 16S. DNA was extracted from fresh skin samples collected in Malawi and Mozambique between August and December 2007, and muscle samples obtained from the Durban Science Museum, South Africa (Annex 2). Both fresh and museum specimens were stored in 95% ethanol, and DNA was extracted using a phenolchloroform-isoamyl procedure (Sambrook et al., 1989). A set of 6 custom primers were used to amplify the two genes (Annex 1), which were sequenced using the Sanger method (Sanger et al., 1977). Existing sequences available in GenBank by 1/10/09 were also integrated in the analysis (Annex 2). Sequences were first aligned with a multiple alignment programme, ClustalX (Larkin et al., 2007) and then adjusted manually with the sequence alignment editor BioEdit (Hall, 1999). Sequences from the two genes were concatenated to form a supermatrix after checking that single gene phylogenetic analyses did not show incongruent nodes supported by more than 80% bootstrap.

Trees were built with both Maximum Likelihood (ML) and Bayesian methods. We used PhyML 2.4.4 (Guindon & Gascuel, 2003) and MrBayes 3.1.2 (Huelsenbeck et al., 2001). The 87 best model for the combined data set was GTR + Γ+ Ι as determined by likelihood ratio tests. Three mammals, the horse (Equidae, Perissodactyla), the pangolin (Manidae, Pholidota) and the sea lion (Phocidae, Carnivora) were designated as outgroups. ML bootstrap support was determined from 200 replicates with NNI branch swapping (Salamin et al., 2003). The MrBayes tree was obtained by running three independent analyses, each with four Markov chains for 50,000,000 generations with sampling performed every 1000 trees, thereby generating 50,000 sample points per analysis. Burn-in values were determined in Tracer 1.5 which allowed to discarding the first 5000 samples and a consensus tree with posterior probability values was obtained in TreeAnnotator 1.5.3 by pooling trees from the three independent runs (both extensions of BEAST 1.5.3, see below).

The relaxed Bayesian clock method (Drummond et al., 2006) was obtained to generate a dated tree using BEAST 1.5.3 (Drummond & Rambaut, 2007). As for the previous analyses, the GTR+G+I model was used. Three fossil constraints were introduced on the tree and prior distributions followed a normal distribution with mean taken from the fossil record and variance adjusted to include the uncertainty associated with the fossil. The first constraint (corresponding to the last shared ancestor of the Chiroptera) was set to 65±5 MYA, at or following the Cretaceous-Tertiary boundary (Eick et al., 2005; Teeling et al., 2005). The second fossil constraint, corresponding to the last known common ancestor shared by the families Rhinolophidae and Hipposideridae was set to 46±5 MYA (Eick et al., 2005; Teeling et al., 2005) and a third was set to 75±10 MYA for the carnivore pangolin split (McKenna & Bell, 1997).

Niche modeling Species occurrence data were obtained from museum records all over Africa, North America and Europe and represent the most comprehensive distribution dataset for Southern African bat species existing today. Maxent 3.2.19, a maximum entropy method which employs presence-only data (Phillips et al., 2006) was used to model the actual potential distribution of 20 species of Southern African bats with 6 climatic variables provided by the Climatic Research Unit database (CRU) at the University of East Anglia. These were actual evapo- transpiration, potential evapo-transpiration, minimum temperature, mean temperature,

88 growing degree days (a measure of heat accumulation) and average precipitation. Models were built only for species with 30 or more occurrences. The intersection between the fraction of background predicted and the omission on training samples was used as the threshold to transform prediction probabilities into a binary prediction. This approach minimizes both the predicted area and the number of training sample omissions. Percentage overlap values were calculated in ArcMap by overlaying the binary predictions maps obtained from Maxent 3.2.19.

Tests of niche similarity and niche equivalency Using this complete interfamilial phylogeny we tested the theory of niche conservatism by examining climatic niche similarity and equivalency with pairwise species comparisons. The pairs included in this analysis had to be separate from one another, in other words, the evolutionary paths linking different pairs could not be shared (Maddison, 2000). We calculated environmental niche equivalency and similarity following the methods recently proposed by Warren et al (2008). Species pairs are therefore tested against the total environmental space occupied by all other species in the phylogenetic tree. The nature of the relationship between closely related species and their climatic niches was tested by fitting a linear model between the time since two species shared a common ancestor and the degree of niche overlap experienced.

Results Phylogenetic reconstruction The combined supermatrix comprised 2782 sites for 89 taxa (Annex 2). Of these, 1054 characters were constant, 326 were variable and parsimony-uninformative and 1402 were parsimony-informative. The species in the tree comprise all extant African families and make up more than 80% of the bat fauna of mainland Southern Africa (Taylor, 2000), as well as including the Pteropus and Plecotus clades, which are primarily found in African islands and northern Africa respectively (Annex 2).

89 Mirounga leonina Manis pentadactyla Pteropus rufus Pteropus aldabrensis Pteropus seychellensis Pteropus livingstonii Pteropus voeltzkowi Rousettus aegyptiacus Rousettus lanosus Epomophorus minor Pteropodiformes Epomophorus crypturus Epomophorus wahlbergi Pteropodidae Micropteropus pusillus Epomops franqueti Myonycteris torquata Myonycteris brachycephala Myonycteris relicta Lyssonycteris angolensis Megaloglossus woermanni Eidolon helvum Rhinolophus fumigatus Rhinolophus hildebrandtii Rhinolophus dentii Rhinolophus maendeleo Rhinolophus simulator Rhinolophus alcyone Rhinolophus landeri Rhinolophidae Hipposideros vittatus Rhinolophus blasii Cloeotis percivali Hipposideros jonesi Hipposideros caffer Hipposideros fulminans Hipposideridae Hipposideros beatus Hipposideros ruber Myzopoda schliemanni Myzopoda aurita Myzopodidae Barbastella leucomelas Scotophilus nux Scotophilus leucogaster Scotophilus viridis Scotophilus dinganii Scotophilus borbonicus Glauconycteris beatrix Glauconycteris variegata Glauconycteris argentata Glauconycteris poensis Eptesicus fuscus Eptesicus hottentotus Plecotus christii Plecotus balensis Plecotus austriacus Otonycteris hemprichii Laephotis botswanae Vespertilionidae Laephotis wintonii Neoromicia capensis Neoromicia zuluensis Scotoecus albigula * Scotoecus hindei Pipistrellus rusticus Vespertilioniformes Pipistrellus hesperidus Neoromicia nanus Pipistrellus kuhlii Hypsugo nanus Eptesicus rendallii Neoromicia melckorum Kerivoula argentata Kerivoula lanosa Mimetillus moloneyi * Myotis blythii Myotis tricolor Myotis bocagei Myotis welwitschii Cistugo seabrai Cistugo lesueuri Miniopterus inflatus Miniopterus sororculus Miniopterus petersoni Miniopterus gleni Miniopterus majori Miniopteridae Miniopterus manavi Miniopterus newtonii Miniopterus minor Miniopterus natalensis Miniopterus fraterculus Nycteris macrotis Nycteridae Nycteris grandis Molossus brachypterus Chaerophon chapini Chaerophon pumilus Chaerophon ansorgei Mops condylurus Chaerophon nigeriae Mops midas Molossidae Tadarida fulminans Tadarida lobata Sauromys petrophilus Tadarida aegyptiaca Otomops martissieni Emballonura tiavato Emballonura atrata Emballonuridae Equus caballus substitutions/site 0.06

Figure 1: Maximum likelihood tree (-ln likelihood = -71799.12038) resulting from the analysis of the concatenated matrix using a GTR + Γ+ Ι model of nucleotide evolution. The number of asterisks indicate whether labeled nodes were recovered with >75% bootstrap or >0.95 Bayesian posterior probability by all three (***), two (**), or only one (*) of the three methods of phylogenetic inference used. Branch lengths are proportional to the number of substitutions as indicated by the scale bar. Species are grouped into families and suborders.

Tests of niche similarity, niche equivalency and niche overlap Tests of niche equivalency revealed that in none of the species pairs climatic niches were identical (Table 1). Sister species had on average a higher degree of climatic niche and geographic overlap (43±10 and 26±11 respectively) than non-sister pairs (35±15 and 24±13 respectively) but this difference was not significant (p=0.14 and p=0.70 respectively). Both climatic niche and spatial overlap were high in all species pairs, with an overall mean of 38±13% and 25±12% respectively, though geographic and climatic niche overlap were not significantly correlated (rho: 0.376, p= 0.102). Niche similarity tests revealed that 8 species pairs had one significant p-value and one non significant p-value, indicating that one species’ climatic niche was nested within the other’s (Table 1).

91 Table 1: Results from the similarity and equivalency tests for 8 pairs of sister and 12 pairs of non- sister species against a background including all of the environmental space occupied by the southern African bat species included in this study. One p-value per species is given for each niche similarity test.

Niche Geographical Niche Node Species equivalency overlap overlap age p-value (%) (%) (Mya)

Epomophorus minor 0.039 22.60 0.319 3.09 Epomophorus crypturus 0.118

Eidolon helvum 0.019 18.58 0.447 42.03 Rousettus aegyptiacus 0.019

Rhinolophus denti 0.099 5.75 0.316 11.18 Rhinolophus simulator 0.019

Rhinolophus hildebrandtii 0.138 35.14 0.316 10.99 Rhinolophus fumigatus 0.019

Cloetis percivali 0.019 40.78 0.451 30.28 Rhinolophus blasii 0.019

Hipposideros ruber 0.019 8.30 0.358 21.03 Hipposideros caffer 0.079

Myotis bocagii 0.019 15.00 0.36 20.44 Myotis tricolor 0.019

Kerivoula argentata 0.277 23.52 0.252 24.35 Kerivoula lanosa 0.257

Neoromicia nanus 0.039 28.83 0.436 10.47 Pipistrellus hesperidus 0.039

Neoromicia capensis 0.534 31.94 0.259 18.17 Laephotis botswanae 0.693

Eptesicus hottentotus 0.138 20.37 0.222 59.22 Cistugo lesueuri 0.019

Glauconycteris variegata 0.019 41.60 0.457 41.93 Scotophilus leucogaster 0.019

Scotophilus dinganii 0.138 38.90 0.404 4.59 Scotophilus viridis 0.990

Nycteris grandis 0.039 4.71 0.563 29.38 Nycteris macrotis 0.019

Sauromys petrophilus 0.019 35.33 0.543 7.89 Tadarida aegyptiaca 0.019

Chaerophon chapini 0.019 31.97 0.526 5.79 Chaerophon pumilus 0.019

Miniopterus natalensis 0.019 11.73 0.305 13.26 Miniopterus fraterculus 0.891

Otomops martissieni 0.633 15.74 0.188 9.25 Tadarida fulminans 0.039

Mops condylurus 0.039 24.74 0.252 19.05 Mops midas 0.633

Chaerophon nigeriae 0.019 49.36 0.745 10.33 Chaerophon ansorgei 0.019 92

Phylogenetic signal When considering all taxa together, time since divergence was neither related to niche overlap nor to geographic overlap. (p= 0.621 and p=0.615 respectively, Figs 4a,b). When considering the two suborders, Pteropodiformes and Vespertilioniformes, separately, the former displayed a positive slope whilst the latter a negative slope (Figs 4c,d and 5e,f respectively), but in all but one case (Fig 4c, p=0.007, slope=0.004) these relationships were not significant.

93 a b

Geographic overlap Geographic Climatic niche overlap niche Climatic

10 20 30 40 50 0.2 0.3 0.4 0.5 0.6 0.7

10 20 30 40 50 60 10 20 30 40 50 60

Node Age Node Age

d c

Geographic overlap Geographic overlap niche Climatic 0.32 0.36 0.40 0.44 510203040 10 20 30 40 10 20 30 40 Node Age Node Age

e f

Geographic overlap Geographic

overlap niche Climatic 10 20 30 40 50 0.2 0.3 0.4 0.5 0.6 0.7

10 20 30 40 50 60 10 20 30 40 50 60

Node Age Node Age

Figure 4: Time since divergence plotted against climatic niche overlap (a,c and e) and geographic overlap (b,d and f) obtained by comparing species pairs to all other species. Figs 4a,b represent all species comparisons together, while 4c,d and 4e,f represent Pteropodiformes and Vespertilioniformes respectively. 94 Discussion Not surprisingly, none of the southern African bat species in this study were found to have equivalent macro-climatic niches. Hence, no two species share exactly the same niche. However, the overall levels of both climatic niche and geographic overlap were high amongst both sister and non-sister species pairs. Hence, we found little evidence for generalized phylogenetic signal or indeed niche conservatism in this group, since sister pairs were just as likely to share similar niches as more distantly related species.

Conservatism is the expected pattern during species diversification (Webb et al., 2002) and has been hypothesized to result from active, stabilizing selection (Lord et al. 1995), or from fixation of ancestral traits that limit the potential range of outcomes during niche evolution (Westoby et al., 1995; Webb et al., 2002; Knouft et al., 2006). While the theoretical framework agrees that some level of phylogenetic niche conservatism is to be expected and that it makes evolutionary sense (Wiens & Graham, 2005; Losos, 2008a) examination of specific clades such as the Anolis lizards, the dendrobatid frogs and the Aphelocoma jays yields contrasting findings to this expectation (Losos et al., 2003; Rice et al., 2003; Graham et al., 2004; Knouft et al., 2006). Earlier work examining niche overlap in species pairs separated by a geographic barrier supported the prediction that niche conservatism characterizes evolutionary diversification (Peterson et al., 1999). However, more recent work has suggested that patterns of niche evolution within and beyond sister taxa can be inconsistent and not conserved (Losos & Glor, 2003; Rice et al., 2003; Graham et al., 2004; Knouft et al., 2006). Our results are congruent with these more recent findings.

When all species were considered together our data showed evidence for sympatric speciation (generally thought to be the most common in mammals) since the regression intercept for the relationship between node age and geographic overlap was significantly greater than 0.5 and the slope was negative (Fitzpatrick & Turelli, 2006).

As discussed by other authors, scale is certainly a key aspect to whether patterns of niche conservatism are detected or not (Wiens & Graham, 2005; Losos, 2008b; Wiens, 2008). For example, despite there being no evidence for niche conservatism for Anolis lizards in Cuba, 95 the Anolis genus is restricted to tropical regions and has never been able to colonise cool temperate and desert regions (Conant & Colllins, 1998). Therefore, conclusions reached for a group at a certain scale should be applicable for that group and for that scale only.

When the data for the bats of southern Africa is divided into suborders, there seem to be different trends in phylogenetic signals (Figs 4c-f). Although the small number of data points makes reaching any definite conclusion difficult, these trends suggest again that scale is an important factor to consider in the analysis of phylogenetic conservatism.

The factors that may be influencing the lack of phylogenetic structuring in climatic niche and space for southern African bats are many. Bats are generally considered to be very good at partitioning resources with notable differences in size, echolocation frequency, flight morphology, trophic level membership and food acquisition strategies (Kalko, 1998). Within a few square kilometers of any tropical forest, there are species which feed above the canopy, within the undergrowth and in gaps or vegetation edges. Species feeding in the undergrowth can choose to either capture flying or resting insects and their echolocation design (with a few exceptions in the New World) allows significant partitioning in terms of size and type of prey (Kalko, 1998; Kingston et al., 2000; Aguirre et al., 2002; Pio et al., 2010). This may mean that the pressure to occupy different macro climatic niches and to separate distributional ranges may not be very strong. Fine-grained resource partitioning may indeed be very influential in structuring this community and may also explain why bat diversity can be so exceptionally high in very restricted areas (Kalko, 1998; Struebig et al., 2010).

Another factor which may contribute to the apparent lack of phylogenetic structuring in this community is their ability to disperse. Some bat species travel several kilometers every night just to feed (Corlett, 2009). Some species are well known to migrate hundreds of kilometers every year (Hedenstrom, 2009). With such dispersal capacities, it is easy to imagine how phylogenetic constraints on species’ ranges and climatic niche occupancy may be attenuated by other more significant limiting factors.

Despite methodological and scale differences, our results are broadly congruent with previous studies on dendrobatid frogs, Aphelocoma jays and Anolis lizards (Rice et al., 2003; 96 Graham et al., 2004; Knouft et al., 2006): in all three cases, many instances of closely related taxa that diverge greatly in niche have been discovered. Thus, at the genus, family and order level niches are frequently not phylogenetically conserved, though this does not prevent these same groups from showing phylogentic conservatism at larger or smaller scales.

Conclusions, limitations and further work In conclusion, our results for the bats of southern Africa support the feeling that the pervasiveness of niche conservatism and phylogenetic signal may be overstated (Losos, 2008a; Pearman et al., 2008) and we argue that in this case phylogenetic structuring may not be obvious due to a combination of fine-grained resource partitioning and dispersal abilities.

Pairwise comparisons in phylogenetic studies are attractive as they avoid many frequently unmet assumptions about branch lengths and stochastic models of character evolution (Maddison, 2000). However, the pairwise comparison method is by no means ideal either. The major drawback is the loss of information caused by focusing on only a subset of branches and comparisons (Felsenstein, 1985), and may have low power to detect correlations (Grafen & Ridley, 1996). Recently methods based on rates of character evolution are being revived (Ackerly, 2009) and may help to reach a deeper understanding of when and how niche conservatism or niche partitioning may prevail in structuring communities and patterns of ecological niche occupancy. We plan to implement such methods for comparative purposes and to provide a more robust analysis before submitting this manuscript.

Acknowledgements We would like to thank the Durban Natural Science Museum for providing us with muscle samples and to the Swiss Institute for Bioinformatics (SIB) for access to the Vital-IT cluster. DP is funded by the Marie-Curie Early Stage Researcher (ESR) Fellowship as part of the EU-Hotspots project (http://www.kew.org/hotspots) to AG and NS. NS and DP are members of the Swiss Institute of Bioinformatics.

97 References

Ackerly, D. (2009) Conservatism and diversification of plant functional traits: Evolutionary rates versus phylogenetic signal. Proceedings of the National Academy of Sciences of the United States of America, 106, 19699-19706. Aguirre, L. F., Herrel, A., van Damme, R. & Matthysen, E. (2002) Ecomorphological analysis of trophic niche partitioning in a tropical savannah bat community. Proceedings of the Royal Society of London Series B-Biological Sciences, 269, 1271-1278. Anderson, T. M., Lachance, M. A. & Starmer, W. T. (2004) The relationship of phylogeny to community structure: The cactus yeast community. American Naturalist, 164, 709-721. Barraclough, T. G. & Vogler, A. P. (2000) Detecting the geographical pattern of speciation from species-level phylogenies. American Naturalist, 155, 419-434. Barraclough, T. G., Vogler, A. P. & Harvey, P. H. (1998) Revealing the factors that promote speciation. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 353, 241-249. Blomberg, S. P. & Garland, T. (2002) Tempo and mode in evolution: phylogenetic inertia, adaptation and comparative methods. Journal of Evolutionary Biology, 15, 899-910. Bohning-Gaese, K. & Oberrath, R. (1999) Phylogenetic effects on morphological, life- history, behavioural and ecological traits of birds. Evolutionary Ecology Research, 1, 347- 364. Conant, R. & Colllins, J. T. (1998) A Field Guide to Reptiles and Amphibians. Eastern and Central North America. Corlett, R. T. (2009) Seed Dispersal Distances and Plant Migration Potential in Tropical East Asia. Biotropica, 41, 592-598. Drummond, A., Ho, S., Phillips, M. & Rambaut, A. (2006) Relaxed Phylogenetics and Dating with Confidence. PLoS Biology, 4, 699-710. Drummond, A. & Rambaut, A. (2007) BEAST: Bayesian evolutionary analysis by sampling trees. Bmc Evolutionary Biology, 7, 8. Duncan, R. P. & Williams, P. A. (2002) Ecology - Darwin's naturalization hypothesis challenged. Nature, 417, 608-609. Graham, C. H., Ron, S. R., Santos, J. C., Schneider, C. J. & Moritz, C. (2004) Integrating phylogenetics and environmental niche models to explore speciation mechanisms in dendrobatid frogs. Evolution, 58, 1781-1793. Guindon, S. & Gascuel, O. (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic Biology, 52. Hall, T. A. (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl. Acids. Symp. Ser., 41, 95-98. Hedenstrom, A. (2009) Optimal migration strategies in bats. Journal of Mammalogy, 90, 1298- 1309. Huelsenbeck, J., Ronquist, F., Nielsen, R. & Bollback, J. (2001) Bayesian inference of phylogeny and its impact on evolutionary biology. Science, 294, 2310-2314. Hugall, A., Moritz, C., Moussalli, A. & Stanisic, J. (2002) Reconciling paleodistribution models and comparative phylogeography in the Wet Tropics rainforest land snail Gnarosophia bellendenkerensis (Brazier 1875). Proceedings of the National Academy of Sciences of the United States of America, 99, 6112-6117.

98 Johnson, M. T. J. & Stinchcombe, J. R. (2007) An emerging synthesis between community ecology and evolutionary biology. Trends in Ecology & Evolution, 22, 250-257. Kalko, E. K. V. (1998) Organisation and diversity of tropical bat communities through space and time. Zoology-Analysis of Complex Systems, 101, 281-297. Kingston, T., Jones, G., Zubaid, A. & Kunz, T. H. (2000) Resource partitioning in rhinolophoid bats revisited. Oecologia, 124, 332-342. Knouft, J. H., Losos, J. B., Glor, R. E. & Kolbe, J. J. (2006) Phylogenetic analysis of the evolution of the niche in lizards of the Anolis sagrei group. Ecology, 87, S29-S38. Larkin, M., Blackshields, G., Brown, N., Chenna, R., McGettigan, P., McWilliam, H., Valentin, F., Wallace, I., Wilm, A., Lopez, R., Thompson, J., Gibson, T. & Higgins, D. (2007) Clustal W and Clustal X version 2.0. Bioinformatics, 23, 2947-2948. Lindeman, P. V. (2000) Resource use of five sympatric turtle species: effects of competition, phylogeny, and morphology. Canadian Journal of Zoology-Revue Canadienne De Zoologie, 78, 992-1008. Linder, H. P. (2003) The radiation of the Cape flora, southern Africa. Biological Reviews, 78, 597-638. Losos, J. B. (2008a) Phylogenetic niche conservatism, phylogenetic signal and the relationship between phylogenetic relatedness and ecological similarity among species. Ecology Letters, 11, 995-1003. Losos, J. B. (2008b) Rejoinder to Wiens (2008): Phylogenetic niche conservatism, its occurrence and importance. Ecology Letters, 11, 1005-1007. Losos, J. B. & Glor, R. E. (2003) Phylogenetic comparative methods and the geography of speciation. Trends in Ecology & Evolution, 18, 220-227. Losos, J. B., Leal, M., Glor, R. E., de Queiroz, K., Hertz, P. E., Schettino, L. R., Lara, A. C., Jackman, T. R. & Larson, A. (2003) Niche lability in the evolution of a Caribbean lizard community. Nature, 424, 542-545. Maddison, W. P. (2000) Testing character correlation using pairwise comparisons on a phylogeny. Journal of Theoretical Biology, 202, 195-204. Mattern, M. Y. & McLennan, D. A. (2000) Phylogeny and speciation of felids. Cladistics-the International Journal of the Willi Hennig Society, 16, 232-253. Monadjem, A., Schoeman, M. C., Reside, A., Pio, D. V., Stoffberg, S., Bayliss, J., Cotterill, F. P. D., Curran, M., Kopp, M. & Taylor, P. (in review) A recent inventory of the bats of Mozambique with documentation of seven new species to the country Mouillot, D., Krasnov, B. R., Shenbrot, G. I., Gaston, K. J. & Poulin, R. (2006) Conservatism of host specificity in parasites. Ecography, 29, 596-602. 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. Pearman, P. B., Guisan, A., Broennimann, O. & Randin, C. F. (2008) Niche dynamics in space and time. Trends in Ecology & Evolution, 23, 149-158. Peres-Neto, P. R. (2004) Patterns in the co-occurrence of fish species in streams: the role of site suitability, morphology and phylogeny versus species interactions. Oecologia, 140, 352-360. Peterson, A. T., Soberon, J. & Sanchez-Cordero, V. (1999) Conservatism of ecological niches in evolutionary time. Science, 285, 1265-1267. Phillips, S. J., Anderson, R. P. & Schapire, R. E. (2006) Maximum entropy modeling of species geographic distributions. Ecological Modelling, 190, 231-259.

99 Pio, D. V., Clarke, F. M., Mackie, I. & Racey, P. A. (2010) Echolocation calls of the bats of Trinidad, West Indies: is guild membership reflected in echolocation signal design? Acta Chiropterologica. Rice, N. H., Martinez-Meyer, E. & Peterson, A. T. (2003) Ecological niche differentiation in the Aphelocoma jays: a phylogenetic perspective. Biological Journal of the Linnean Society, 80, 369-383. Salamin, N., Chase, M. W., Hodkinson, T. R. & Savolainen, V. (2003) Assessing internal support with large phylogenetic DNA matrices. Molecular Phylogenetics and Evolution, 27, 528-539. Sambrook, J., Fritsch, E. & Maniatis, T. (1989) Molecular cloning: a laboratory manual Cold Spring Harbor Laboratory Press, , New York. Sanger, F., Nicklen, S. & Coulsen, A. (1977) DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences, 74, 5463-5467. Simmons, N. (2005) in Mammalian Species of the World: A Taxonomic and Geographic Reference, D. E. Wilson, D. M. Reeder, Eds. Johns Hopkins Univ. Press, Baltimore. Struebig, M. J., Christy, L., Pio, D. & Meijaard, E. (2010) Bats of Borneo: diversity, distributions and representation in protected areas. Biodiversity and Conservation, 19, 449-469. Swenson, N. G., Enquist, B. J., Pither, J., Thompson, J. & Zimmerman, J. K. (2006) The problem and promise of scale dependency in community phylogenetics. Ecology, 87, 2418-2424. Taylor, P. (2000) Bats of Southern Africa. University of Natal Press, Pietermaritzburg. Vitt, L. J., Zani, P. A. & Esposito, M. C. (1999) Historical ecology of Amazonian lizards: implications for community ecology. Oikos, 87, 286-294. Warren, D. L., Glor, R. E. & Turelli, M. (2008) Environmental niche equivalency versus conservatism: quantitative approaches to niche evolution Evolution, 62, 2868-2883. Webb, C. O., Ackerly, D. D., McPeek, M. A. & Donoghue, M. J. (2002) Phylogenies and community ecology. Annual Review of Ecology and Systematics, 33, 475-505. Westoby, M., Leishman, M. & Lord, J. (1995) Issues of interpretation after relating comparative datasets to phylogeny. Journal of Ecology, 83, 892-893. Wiens, J. J. (2008) Commentary on Losos (2008): Niche conservatism deja vu. Ecology Letters, 11, 1004-1005. Wiens, J. J. & Graham, C. H. (2005) Niche conservatism: Integrating evolution, ecology, and conservation biology. Annual Review of Ecology Evolution and Systematics, 36, 519-539.

100 Annex 1: Custom primers used in the amplification and sequencing of Cytochrome b and

16S genes.

Cytochrome b a. Forward: 5’-AAY CAC CGT TGT AYT TCA AC-3’ b. Reverse: 5’-TAG AAT NTC AGC TTT GGG TG-3’

16S a. Forward: 5’-GCA CCT AGT TTA CAC CTA GAA GAT T-3’ b. Reverse: 5’-GCC GAG TTC CTT TTA CTT CTT TT-3’ c. Forward: 5’-TCT CTM YTC CTT TCG TAC TG-3’ d. Reverse: 5’-TGT TAA TAT GAG TAA YAA G-3’

Annex 2: Chiropteran and out-group taxa Included in this study

Order Family Species Source ID Chiroptera Emballonuridae Emballobura tiavato GenBank DQ178285 Emballonura atrata GenBank AF203773 Hipposideridae Cloeotis percivali DNSM 8027 Hipposideros beatus GenBank FJ347976 Hipposideros caffer GenBank FJ185182 Hipposideros fuliginosus GenBank EU934 Hipposideros jonesi GenBank EU934472 Hipposideros ruber DNSM MB013 Hipposideros vittatus DP/AM M328 Miniopteridae Miniopterus fraterculus MK/MC MB021 Miniopterus gleni GenBank FJ383146 Miniopterus inflatus MK/MC AY495487 Miniopterus majori GenBank DQ899780 Miniopterus manavi GenBank FJ383130 Miniopterus minor GenBank FJ232803 Miniopterus natalensis MK/MC/DP 410 Miniopterus newtonii GenBank EF363523 Miniopterus petersoni GenBank FJ383131 Miniopterus sororculus GenBank DQ89977 Molossidae Chaerophon ansorgeri DNSM 8607 Chaerophon chapini GE 46 Chaerophon nigeriae GE 45

101 Chaerophon pumilus GE AY495454 Molossus brachypterus MK/MC/DP 532 Mops condylurus DP/AM M323 Mops midas GE 47 Otomops martissieni DNSM 8571 Sauromys petrophilus DNSM 8611 Tadarida aegyptiaca DNSM 5673 Tadarida fulminans DNSM 8619 Tadarida lobata GE 48 Pteropodidae Eidolon helvum MK/MC/DP 462 Epomophorus crypturus MK/MC/DP 388b Epomophorus minor DNSM 8643 Epomophorus wahlbergii MK/MC/DP MB030 Epomops franqueti GenBank DQ445707 Lyssonycteris angolensis MK/MC/DP 454 Megaloglossus woermanni GenBank AF044620 Micropteropus pusillus GenBank AF044617 Myonycteris brachycephala GenBank AF044613 Myonycteris relicta GenBank AF044618 Myonycteris torquata GenBank FJ218483 Pteropus aldabrensis GenBank FJ588907 Pteropus livingstonii GenBank FJ588885 Pteropus rufus GenBank FJ588894 Pteropus seychellensis GenBank FJ588902 Pteropus voeltzkowi GenBank FJ588896 Rousettus aegyptiacus GE AF044627 Rousettus lanosus GenBank AY012139 Rhinolophidae Rhinolophus alcyone GenBank AY395858 Rhinolophus dentii MK/MC/DP 522 Rhinolophus fumigatus DNSM 8567 Rhinolophus hildebrandtii MK/MC/DP 159 Rhinolophus landeri MK/MC MB008 Rhinolophus maendeleo MK/MC NL219 Rhinolophus simulator MK/MC/DP 135 Rhinolphus blasii MK/MC/DP 531 Rhinopomatidae Rhinopoma hardwickii GenBank AF263232 Vespertilionidae Barbastella leucomelas GenBank EF53476 Cistugo lesueuri GE 27 Cistugo seabrai GE 115 Eptesicus fuscus GenBank AF326092 Eptesicus hottentotus GenBank AY495466 Eptesicus rendallii GenBank AY495515 Glauconycteris argentatus GenBank AY49546 Glauconycteris beatrix GenBank AY495469 Glauconycteris poensis GenBank AY495470 Glauconycteris variegata GenBank AY495471

102 Kerivoula argentata DP/AM M322 Kerivoula lanosus MK/MC/DP 584 Laephotis botswanae MK/MC ML092 Laephotis wintonii GenBank AJ841964 Mimetillus moloneyi MK/MC/DP 383b Myotis blythii GenBank DQ120906 Myotis bocagei DNSM 9520 Myotis tricolor MK/MC/DP 591 Myotis welwitchii MK/MC/DP AY495511 Myzopoda aurita GenBank AF345926 Myzopoda schliemanni GenBank DQ178325 Neoromicia capensis GenBank AF203737 Neoromicia melckorum MK/MC/DP 595 Neoromicia nanus MK/MC/DP 389 Neoromicia zuluensis DNSM 8476 Nycteris grandis DNSM 8669 Nycteris macrotis DNSM 8663 Otonycteris hemprichii GenBank AF326103 Pipistrellus hesperidus MK/MC/DP 384 Pipistrellus kuhlii MK/MC/DP 248 Pipistrellus rusticus MK/MC/DP 354 Plecotus austriacus GenBank AF326107 Plecotus balensis GenBank AF513799 Plecotus christii GenBank AY531615 Scotoecus albigula GE EICK Scotoecus hindei DP/AM M308 Scotophilus borbonicus GenBank AY49553 Scotophilus dinganii MK/MC AY495533 Scotophilus leucogaster MK/MC/DP AY395867 Scotophilus nux GenBank AY495535 Scotophilus viridis MK/MC/DP AF326112 Carnivora Phocidae Mirounga leonina GenBank AY377389 Perissodactyla Equidae Equus caballus GenBank AY011180 Pholidota Manidae Manis pentadactyla GenBank AY011188

NOTE - DNSM, Durban Natural Science Museum, South Africa; DP/AM, collected by D.V. Pio and A. Monadjem in Mozambique; MK/MC/DP, collected by M. Kopp, M. Curran and D.V. Pio in Malawi; GE, sequence provided by G. Eick.

103 Conclusions

In this section, I will summarise how my work has contributed to some ongoing debates of key importance for ecological theory and conservation practice, particularly in those fields of overlap between the sciences of species distribution modelling and phylogenetics. I will end by describing how these combined fields may best serve conservation and the preservation of evolutionary potential.

Spatial predictions of evolutionary history challenge decision making The first contribution of this thesis was to quantify the degree of spatial overlap between predictions of species richness and several measures of phylogenetic diversity (PD). I found that in terms of land surface likely to be set aside for conservation, the overlap was high in some measures, though by no means perfect. Despite the fact that the plant family considered in this study possesses a fairly balanced tree, where older branches with few relatives are not geographically isolated, there is still a discrepancy of about 20-30% in richest areas selected. In general, the lower the percentage of land available for conservation, the lower the overlap between measures, thus the larger the amount of land set aside, the higher the chance of selecting sites which incorporate both high species richness and high PD. Despite these results and the fact that generating the data necessary to incorporate PD into conservation planning is becoming easier, quicker and cheaper and is predicted to become even more so in the future (Mace et al., 2003), barriers still exist to conservation managers taking this sort of information into account. Phylogenetics and conservation may be two disciplines which have been working towards common goals in academia for decades now, but their combined efforts have by no means entered mainstream conservation thinking. The best contribution this field could make to practical conservation would be to come up with a simple way of rapidly incorporating PD into conservation planning. Many studies have come up with different measures which incorporate species history and relatedness, lately also coupled with endemism and abundance data (Rosauer et al., 2009; Cadotte & 104 Davies, 2010), but until any of these measures are compiled in user-friendly and fool-proof software (and larger more complete datasets become available), conservation organizations are likely to take little notice of research progress.

Climate change and evolutionary history The second major contribution of this thesis was to assess the potential effects of climatic change in future decades on PD. I predicted that in three diverse southern African animal and plant groups ranges will contract in such a way to threaten a huge proportion of the taxa in each group. Under present climatic conditions species listed in IUCN threat categories are still very few (if any) and their simulated loss revealed no difference in amounts of surviving evolutionary history from random extinction. However, when simulations were carried out under future climatic conditions (2080) involving significant pruning of each tree, surviving PD was lower than would be experienced under random extinction processes. This confirms previous studies which find non-random extinction patterns and provides a likely explanation for studies which did not, by arguing that detectability may simply be a function of what stage of the extinction process a particular group finds itself in.

The use of climate change and species modeling to inform conservation has been criticized for one main reason. Models are intrinsically uncertain. Uncertainty arises because of limited knowledge of current species’ distributions and future climate change, as well as the challenge of predicting species responses to novel climates and novel interactions (Elith & Leathwick, 2009; Ackerly et al., 2010). Despite these limitations, species distribution models remain a very powerful tool to predict trends in species shifts, contractions and extinctions, and have so far been surprisingly precise (Kharouba et al., 2009; Tingley et al., 2009). Moreover, since the future will be impossible to predict with absolute accuracy (Araujo & Rahbek, 2006), modeling tools, which increasingly incorporate information about species interactions and dispersal abilities (Engler et al., 2004; Guisan et al., 2006; Le Lay et al., in review), should remain the primary way of informing and preparing conservation action to predicted changes in climate.

105 Exploring the relationship between morphology and phylogenetic diversity Because diversity is ultimately the product of descent and gradual modification, branch lengths on a phylogenetic tree should predict both morphological and genetic feature diversity (Faith, 1992a; Mooers et al., 2005a). This would be particularly true when considering a model of Brownian motion evolution where characters evolve slowly in a more or less random manner and where more closely related species are predicted to share more similar phenotypes and genotypes (Hansen, 1997). I tested the relationship between morphological disparity and PD for the diverse bat community of southern Africa and found weak but negative trends for almost all of the 10 evolutionary history measures used. I argue that this pattern is likely the result of considerable convergent evolution amongst clades and of differentiation within individual clades. I conclude that feature diversity is probably best described as genotypic variation and not necessarily linked to morphological variation.

This finding reinforces previous studies which question the assumption of Brownian motion model of evolution (Freckleton & Harvey, 2006; Hansen et al., 2008). A recent study provided evidence that the evolution of feeding adaptations in two radiations of warblers was not predicted by Brownian motion, but that instead it showed characteristics expected under certain niche-filling models (Freckleton & Harvey, 2006).

This result does not change the value or importance of maximizing the representation of PD in conservation. Simply, it states that a link with morphology, as would be expected under a Brownian model of evolution, cannot be assumed. Though it remains to be shown, it is possible that additional components of feature diversity, such as reproductive and other functional adaptations, may also display an equally weak link with PD. If this is indeed the case, the way feature diversity is conceptualized and promoted should certainly be changed and its definition should be refocused on maintaining genotypic and not phenotypic options.

106 No macro-climatic niche conservatism in the bats of southern Africa Another contribution of this thesis was to investigate whether phylogenetic relationships determine macro-climatic niche occupancy in a highly diverse, highly mobile mammal community. I compared the macro-climatic niches occupied by southern African bat species pairs against the environmental background occupied by the whole phylogenetic tree and found that more closely related species were not more likely to display higher overlap. Though a negative trend between macro-climatic niche overlap and age at divergence between pairs of species was detected, the relationship was weak, suggesting limited phylogenetic structuring in niche occupancy patterns. I conclude that considerable fine-scale resource partitioning and exceptional dispersal abilities may be in part responsible for the apparent lack of phylogentic structuring. The pervasiveness of niche conservatism and phylogenetic signal is therefore questioned, especially in view of other recent studies which concur with the results presented here (Losos et al., 2003; Rice et al., 2003; Graham et al., 2004).

Future use of phylogenetics and niche modeling in conservation Some authors have argued that too much attention is paid to organisms which are already deemed to go extinct and that conservation efforts should be directed at preserving areas that will facilitate speciation in the future (Myers & Knoll, 2001). They argue that our increasing knowledge on speciation events could help us determine minimum areas and abundance values for speciation to take place, ultimately identifying hotspots of future speciation. However, others (Barraclough & Davies, 2005) have argued that thinking in these terms is simply unfeasible for various reasons. First of all, speciation occurs over thousands to millions of years. It is impossible to think of any continuity of policy and/or management over such timescales (Barraclough & Davies, 2005). Furthermore, our current climatic predictions only extend for a few decades at the most, so it is unthinkable to make

107 predictions about which areas will be likely most important for speciation with the information currently available (Barraclough & Davies, 2005).

Perhaps, the most realistic and immediately achievable contribution, which the combined fields of phylogenetics and niche modeling can make to conservation, is to make reliable predictions about where diversity is now and where it will likely be in the near future, as well as what parts of diversity may be more at risk, based on phylogenetic likelihood of extinction. In this way, targeted efforts can be made to ensure that gradual changes to management and policy match those which are taking place in the environment. Slowing current biodiversity loss by ensuring the maintenance of natural habitat will ultimately ensure evolutionary potential, as long as enough of it is preserved for species to shift through the landscape, as conditions continue to change. But how can we make use of the growing wealth of knowledge gained from the study of evolutionary processes and relationships?

Some authors argue that because the maintenance of adequate genetic variability ultimately facilitates evolutionary processes and maximises adaptive potential, a strong emphasis should be placed on spatial patterns of intraspecific variation (Vandergast et al., 2008). Evolutionary potential can thus be characterised in terms of interpopulation genetic divergence and interpopulation genetic variation (Vandergast et al., 2008). This approach has many obvious advantages, but one drawback common also to biodiversity measures based on species richness alone is the treatment of all species as equal conservation units. Another drawback is that population level data may not be available for the majority of species.

One way to take evolutionary potential into account without losing the information on intra- species relationships is to use molecular rates of evolution to investigate the tempo of diversification (Alfaro et al., 2009). The likelihood of the tempo to be increasing, decreasing or stable can thus be estimated. In view of the fact that phenotypic and genotypic diversity may be de-coupled, trait as well as molecular rates should be used in this estimation. Ideally, clades characterized by particularly high rates of evolution should be prioritized and measures can be devised which combine both evolutionary uniqueness and evolutionary potential to make sure we maximize both current diversity and future evolutionary potential.

108

Another factor thought to play an important role in evolutionary potential is range size. Larger ranges are associated with higher allopatric speciation rates and a decrease in extinction rates (Rosenzweig, 2001). On the other hand the climatically and topographically heterogeneous Cape Floristic Region is an example of extreme speciation within a restricted area. Also, climate fluctuations can periodically and dramatically reduce the geographic range size of even widespread species, thus suddenly increasing their susceptibility to extinction. Therefore the use of climatic niche breadth combined with range size may constitute a more reliable indicator of whether a species or clade has at least one of the basic premises for allopatric speciation to occur. Some authors argue that species will only survive through periods of climatic fluctuations if their individual populations are "threat tolerant," somehow able to persist in spite of the high extinction risk (Waldron, 2010). Threat tolerance is conceptually different from classic extinction resistance, and could have a stronger relationship with diversification rates (Waldron, 2010). Some narrowly distributed species for example have higher threat tolerance than widespread ones, confirming that tolerance is an unusual form of resistance (Waldron, 2010). Our knowledge of the relationship between precise range size, threat tolerance and speciation, which is likely to be taxon specific, is still incomplete, but will in time provide fundamental contributions to the field of conservation science. Whether or not the increasing knowledge gained from evolutionary research is effectively integrated in conservation theory and practice will largely depend on the ability of scientists to make it more accessible.

References Ackerly, D. D., Loarie, S. R., Cornwell, W. K., Weiss, S. B., Hamilton, H., Branciforte, R. & Kraft, N. J. B. (2010) The geography of climate change: implications for conservation biogeography. Diversity and Distributions, 16, 476-487. Alfaro, M. E., Santini, F., Brock, C., Alamillo, H., Dornburg, A., Rabosky, D. L., Carnevale, G. & Harmon, L. J. (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. Araujo, M. B. & Rahbek, C. (2006) How does climate change affect biodiversity? Science, 313, 1396-1397.

109 Barraclough, T. G. & Davies, T. J. (2005) In Phylogeny and Conservation (eds A. Purvis, J. Gittleman & T. Brooks), Cambridge University Press, Cambridge. Cadotte, M. W. & Davies, T. J. (2010) Rarest of the rare: advances in combining evolutionary distinctiveness and scarcity to inform conservation at biogeographical scales. Diversity and Distributions, 16, 376-385. Elith, J. & Leathwick, J. R. (2009) Species Distribution Models: Ecological Explanation and Prediction Across Space and Time. Annual Review of Ecology Evolution and Systematics, 40, 677-697. Engler, R., Guisan, A. & Reichsteiner, L. (2004) Predicting the distribution of rare and endangered species from occurrence and pseudo-absence data Journal of Applied Ecology, 41, 263-274. Faith, D. P. (1992) Conservation Evaluation and Phylogenetic Diversity. Biological Conservation, 61, 1-10. Freckleton, R. P. & Harvey, P. H. (2006) Detecting non-Brownian trait evolution in adaptive radiations. Plos Biology, 4, 2104-2111. Graham, C. H., Ron, S. R., Santos, J. C., Schneider, C. J. & Moritz, C. (2004) Integrating phylogenetics and environmental niche models to explore speciation mechanisms in dendrobatid frogs. Evolution, 58, 1781-1793. Guisan, A., Broennimann, O., Engler, R., Vust, M., Yoccoz, N. G., Lehmann, A. & Zimmermann, N. E. (2006) Using niche-based models to improve the sampling of rare species. Conservation Biology, 20, 501-511. Hansen, T. F. (1997) Stabilizing selection and the comparative analysis of adaptation. Evolution, 51, 1341-1351. Hansen, T. F., Pienaar, J. & Orzack, S. H. (2008) A comparative method for studying adaptation to a randomly evolving environment. Evolution, 62, 1965-1977. Kharouba, H. M., Algar, A. C. & Kerr, J. T. (2009) Historically calibrated predictions of butterfly species' range shift using global change as a pseudo-experiment. Ecology, 90, 2213-2222. Le Lay, G., Franc, E., Engler, R. & Guisan, A. (in review) Using habitat-suitability models enhances chances to find rare species in the field Ecography. Losos, J. B., Leal, M., Glor, R. E., de Queiroz, K., Hertz, P. E., Schettino, L. R., Lara, A. C., Jackman, T. R. & Larson, A. (2003) Niche lability in the evolution of a Caribbean lizard community. Nature, 424, 542-545. Mace, G. M., Gittleman, J. L. & Purvis, A. (2003) Preserving the Tree of Life. Science, 300, 1707-1709. Mooers, A., Heard, S. B. & Chrostowski, E. (2005) In Phylogeny and Conservation (eds A. Purvis, J. L. Gittleman & T. Brooks), pp. 120-138. Cambridge University Press, Cambridge. Myers, N. & Knoll, A. H. (2001) The biotic crisis and the future of evolution. Proceedings of the National Academy of Sciences of the United States of America, 98, 5389-5392. Rice, N. H., Martinez-Meyer, E. & Peterson, A. T. (2003) Ecological niche differentiation in the Aphelocoma jays: a phylogenetic perspective. Biological Journal of the Linnean Society, 80, 369-383. Rosauer, D., Laffan, S. W., Crisp, M. D., Donnellan, S. C. & Cook, L. G. (2009) Phylogenetic endemism: a new approach for identifying geographical concentrations of evolutionary history. Molecular Ecology, 18, 4061-4072.

110 Rosenzweig, M. L. (2001) Loss of speciation rate will impoverish future diversity. Proceedings of the National Academy of Sciences of the United States of America, 98, 5404-5410. Tingley, M. W., Monahan, W. B., Beissinger, S. R. & Moritz, C. (2009) Birds track their Grinnellian niche through a century of climate change. Proceedings of the National Academy of Sciences of the United States of America, 106, 19637-19643. Vandergast, A. G., Bohonak, A. J., Hathaway, S. A., Boys, J. & Fisher, R. N. (2008) Are hotspots of evolutionary potential adequately protected in southern California? Biological Conservation, 141, 1648-1664. Waldron, A. (2010) Lineages that cheat death: surviving the squeeze on range size. Evolution, 64, 2278-2292.

111 Annexes

112 A recent inventory of the bats of Mozambique with documentation of seven new species to the country

Ara Monadjem1,11, M. Corrie Schoeman2, April Reside3, Dorothea V. Pio4, Samantha Stoffenberg5, Julian Bayliss6, F.P.D. (Woody) Cotterill7, Michael Curran,9, Mirjam Kopp8 and Peter J. Taylor10

1All Out Africa Research Unit, Department of Biological Sciences, University of Swaziland, Private Bag 4, Kwaluseni, Swaziland 2School of Biological and Conservation Sciences, University of KwaZulu-Natal, Durban, South Africa 3All Out Africa, P.O. Box 153, Lobamba, Swaziland 4Département d’Ecologie et Evolution, Université de Lausanne, Biophore 1015, Lausanne, Switzerland 5Evolutionary Genomics Group, Department of Botany and Zoology, University of Stellenbosch, Private Bag X1, Matieland, Stellenbosch, South Africa 6Mulanje Mountain Conservation Trust (MMCT), P.O. Box 139, Mulanje, Malawi and Conservation Science Group, Department of Zoology, University of Cambridge, UK 7AEON – African Earth Observatory Network, Departments of Geological Sciences, and Molecular and Cell Biology, University of Cape Town, Rondebosch 7701, South Africa 8Institute for Natur-, Landschafts- und Umweltschutz (NLU), University of Basel, St. Johanns-Vorstadt 10, CH-4056, Switzerland 9Institute of Environmental Engineering, ETH Zurich, HIF C 13, Wolfgang-Pauli-Str. 15 CH-8093 Zurich, Switzerland 10Durban Natural Science Museum, P. O. Box 4085, Durban, South Africa and Dept of Ecology, and Resource Management, School of Environmental Sciences, University of Venda, P/Bag X5050, Thohoyandou, 0950, South Africa

113 Abstract

The bat fauna of Mozambique is poorly documented. We conducted a series of inventories across the country between 2005 and 2009, resulting in the identification of 48 species from 41 sites. Of these, seven species represent new national records that increase the country total to 65 species. These data include results from the first detailed surveys across northern Mozambique, over an area representing almost 50% of the country. We detail information on new distribution records and measurements of these specimens. Special attention is paid to the Rhinolophidae, because these include several taxa that are currently in a state of taxonomic confusion. Furthermore, we also present some notes on taxonomy, ecology and echolocation calls. Finally, we combine modelled distributions to present predicted species richness across the country. Species richness was lowest across the coastal plain, to the east and far north, and is predicted to increase in association with rising altitude and higher topographic unevenness of the landscape.

Key words: Mozambique, Chiroptera, distribution, check-list, conservation

Contribution to the project: I collected part of the data and wrote part of the manuscript. This paper is currently in review.

114 Introduction

The distribution of southern African bat species is poorly known compared with other small mammal taxa such as rodents (Monadjem et al., in press), or compared with other regions such as Europe (Mitchell-Jones et al., 1999). Our knowledge of the bat fauna of Mozambique is particularly inadequate, given that the most recent synopsis is 35 years old (Smithers and Lobão Tello, 1976). At that time, only a single site (Ihla de Mozambique) had been surveyed in the northern provinces (north of the Zambezi river), an area covering almost 50% of the country. The remaining “southern” half of the country was only patchily covered, with 19 species known from single localities and a further nine from just two localities. Hence, prior to this study, a total of only 56 bat species were known to occur in Mozambique, and 28 (50%) of these were known from two or less sites (Smithers and Lobão Tello, 1976) in a country covering over 801,000 km2. A further indication of the under-sampled state of bats in Mozambique is that smaller neighbouring countries had higher species richness: Malawi with 62 species (Happold et al., 1987; Happold and Happold, 1997; Bergmans and van Strein, 2004; Skliba and Petr, 2007) and Zimbabwe with 62 species (Smithers and Wilson, 1979; Cotterill, 1996). To address this lack of distributional data, a series of bat surveys were conducted across Mozambique between 2005 and 2009. We pool the data from these various expeditions to provide a broad overview of bat distributions at a country-wide level. Hence, the aim of this paper is to present information on a number of recent bat expeditions to Mozambique, and to synthesise current and historical records of bat species distributions. The 2003 IUCN assessment of protected areas shows that only 5.7% of Mozambique is designated as protected, and only 2% in “categories I and II” (nature reserves and national parks), compared to the average of 10.9% for sub-Saharan Africa, 15.8% in the USA and 10.5% in the UK (UNEP-WCMC, 2003). Mozambique has 143 vertebrate species threatened with extinction, 11 of these mammals, a further 174 near threatened animal species (resulting in 12% of the total fauna being threatened or near threatened) and 101 that are data deficient (IUCN, 2009). Seven bat species found in Mozambique are listed as near threatened and three as vulnerable (IUCN, 2009). This paucity of knowledge seriously weakens any attempt to assess the status of the bat fauna of Mozambique. With the post- 115 war resettlement of Mozambique, 70% of the human population is rural and reliant on farming (Hanlon, 2007), so it appears that human threats to bat populations in Mozambique will increase. This paper provides vital baseline information on the bat fauna of Mozambique that is essential for both ecological research and applied conservation work.

Methods Study sites Mozambique stretches for approximately 2,770 km along the eastern seaboard of southern Africa, mostly north of the Tropic of Capricorn. Much of the country lies below 200m above sea level, although a few mountain ranges and peaks are scattered in the central and northern parts (Fig. 1). Mozambique straddles sub-tropical Southern Africa to tropical East Africa, and is characterised by its patchily distributed and erratic precipitation with 95% of the annual precipitation concentrated between October and March (Amaral and Sommerhalder, 2004). The mean annual temperature and rainfall varies across the country, this being mostly affected by latitude and altitude. In the far south, the capital city Maputo (39 m above sea level) has a mean monthly temperature of 23.3 °C, and a mean annual rainfall of 769 mm. In the north at Nacala (15 m above sea level), the monthly temperature is 26.2 °C, and the mean annual rainfall is 945 mm. In contrast, Chimoio in central Mozambique (730 m above sea level), has a mean monthly temperature of 21.3 °C, and a mean annual rainfall of 1060 mm. For convenience, we followed the biogeographical divisions denoted by the Bird Atlas (Parker, 1999): 1) southern Mozambique - south of the Save river, 2) central Mozambique - between the Save and Zambezi rivers, and 3) northern Mozambique - north of the Zambezi river.

Sampling and data analysis Bat specimens were trapped and collected from 41 sites. Bats were captured using mistnets, canopy nets and/or harp traps, and sampling intensity varied considerably between sites. Where possible, some bats were caught by hand or hand net within roost sites such as caves and hollow tree trunks. Voucher specimens were collected and preserved in 70% ethanol, and deposited in the Durban Museum of Natural Science (DM), the Transvaal Museum 116 (TM) or the Geneva Museum of Natural History (MHNG). A few specimens from six sites (see Appendix 1) were not collected; for these individuals wing biopsy punches were taken and stored in 99% ethanol before the bat was released. Forearm length (FA) of live bats was measured in the field using digital callipers to the nearest 0.1 mm. Body mass (Bm), to the nearest 1g, was obtained using a Pesola spring scale. These measurements on live individuals are presented under “Field measurements”. All other measurements are based on preserved voucher specimens. For those specimens where skull and dental measurements were necessary for identification (e.g. the Rhinolophidae), the following measurements were taken to the nearest 0.1 mm using digital callipers: greatest length of skull measured dorsally from occiput to anterior point of skull (GSL); condylo-incisive length from occipital condyles to front of premaxillae (CIL); condylo-canine length from occipital condyles to front of canines (CCL); zygomatic width, the greatest distance across the zygoma (ZW); mastoid width, the greatest distance across the lateral projections of the mastoid processes (MW); width of maxilla between outer edges of M3 (M3M3); braincase width measured at dorsal root of zygomatic arches (BCW); least interorbital width between orbits (IOW); upper toothrow length from anterior surface of C to posterior surface of M3 (CM3); greatest width across anterior lateral nasal inflations (NW); length from occipital condyles to front of nasal inflations (NL); and height of nasal inflation directly above the anterior cingulum of M2 (NH). Measurements are summarised as mean ± SD (for external measurements only), range and sample size. Where measurements of males and females differed, these are shown separately for each sex. Where sexes were similar in size or sample sizes were small (e.g., in the case of cranial data), measurements were combined for the sexes. Only measurements of adult bats are shown under “Field measurements” unless otherwise stated. Echolocation calls were recorded from individuals of selected high duty cycle (Hipposideridae and Rhinolophidae) species. Bats were recorded whilst being held by an observer, thus eliminating any possible Doppler shift compensation (Heller and von Helversen, 1989). Echolocation recordings were made using an ANABAT II bat detector (Titley Electronics, Ballina, Australia), or a Pettersson D240x or D980 bat detector (Pettersson Electronik AB, Uppsala, Sweden). ANABAT recordings were analysed with ANALOOK (Chris Corben, version 4.8), and Pettersson recordings with either Raven Pro

117 version 1.3 (Charif, RA, AM Waack, and LM Strickman. 2008, Cornell Laboratory of Ornithology), or with BatSound Pro (v3.20; Pettersson Electronik AB, Uppsala Sweden) software. For ANABAT recordings of high duty cycle bats, we defined peak echolocation frequency as the frequency of the constant frequency (CF) component of the call, F(max), (Monadjem et al., 2007). For Pettersson recordings of high duty cycle bats, peak echolocation frequency was measured from the peak of the power spectrum (Obrist, 1995). The distributions of 46 species of bats occurring in Mozambique were modelled across their southern African range by Monadjem et al. (in press) using the MaxEnt algorithm (version 2.3; Phillips et al., 2006). Nine continuous environmental variables were used as predictors in the final model: altitude, mean annual temperature, maximum temperature of the hottest month, minimum temperature of the coldest month, temperature seasonality, annual precipitation, precipitation of the wettest month, precipitation of the driest month and precipitation seasonality. The MaxEnt model was run with all distribution records (100% training), the regularization multiplier was set to 1 and the maximum number of iterations was set to 500; all other MaxEnt settings were kept unchanged. Only species with ten or more distribution localities in southern Africa were modelled. For further details see Monadjem et al. (in press). The predicted distributions of the 46 bat species within Mozambique were analyzed in ArcView 3.2. Individual distributions were then combined using the Map Calculator function in Spatial Analyst to produce a single map of bat species richness for the country. The IUCN Red List status follows the Global Mammal Assessment (GMA) of African small mammals in January 2004 (IUCN, 2009) and taxonomy follows Wilson and Reeder (2005) unless otherwise stated.

Results and Discussion Over 500 individuals from 48 species and seven families were captured (Table 1) at 41 sites across Mozambique (Fig. 1).

118 Table 1. Summary of 50 bat species captured during this study (2005-2009).

Region* Species Southern Central Northern Pteropodidae Epomophorus crypturus ● Epomophorus labiatus ● Epomophorus wahlbergi ● ● ● Lissonycteris goliath ● Myonycteris relicta ● Rousettus aegyptiacus ● ● Rhinolophidae Rhinolophus blasii ● Rhinolophus clivosus ● Rhinolophus deckenii ● Rhinolophus fumigatus ● Rhinolophus hildebrandtii ● ● ● Rhinolophus landeri ● ● Rhinolophus cf. maendeleo ● Rhinolophus simulator ● Rhinolophus cf. swinnyi ● Hipposideridae Hipposideros caffer ● ● ● Hipposideros ruber ● Hipposideros vitattus ● Triaenops afer ● ● ● Nycteridae Nycteris grandis ● Nycteris hispida ● Nycteris macrotis ● ● ● Nycteris thebaica ● ● Molossidae Chaerephon ansorgei ● Chaerephon pumilus ● ● Mops condylurus ● ● Sauromys petrophilus ● Tadarida aegyptiaca ● Tadarida fulminans ● Vespertilionidae Eptesicus hottentotus ● Glauconycteris variegata ● ● Kerivoula argentata ● Kerivoula lanosa ● Kerivoula cf. phalaena ● Laephotis botswanae ● Myotis bocagii ● Myotis tricolor ● Neoromicia capensis ● Neoromicia nana ● ● ● Neoromicia rendalli ● Neoromicia zuluensis ● ● Nycticeinops schlieffeni ● ● ● Pipistrellus hesperidus ● ● ● Scotoecus hindei/albigula ● ● ● Scotophilus dinganii ● ● Scotophilus leucogaster ● Scotophilus cf. viridis ● ● ● 119 Miniopteridae Miniopterus cf. fraterculus ● Miniopterus inflatus ● ● Miniopterus natalensis ● Species total 20 21 38 *Southern = south of Save river; Central = between Save and Zambezi rivers; Northern = north of Zambezi river (see Fig. 1)

Figure 1. Map of Mozambique showing the sites at which bats were captured during this study (2005-2009). The sampling sites are overlaid on a digital elevation model (altitude in metres). The boundaries of the three biogeographic zones (following Parker 1999) are represented by the two black lines. The southern line follows the Save River, whereas the northern line follows the Zambezi River. The numbers refer to sites and correspond to the site numbers presented in Table 2.

120

Differences in numbers of species collected per site may, at least partially, be attributed to differences in sampling intensity, including the number of nights spent trapping. At five of the eight sites where a single species was captured (Appendix 1), this was due to ad hoc collecting rather than intensive netting and trapping. At 15 sites, five or more bat species were captured, whilst three sites (foothills of Mount Namuli, Mount Namuli and Mount Mabu) yielded more than 10 species of bats (Appendix 1). Thus bat species alpha-richness may well be an underestimate. This is supported by the modelled distribution richness of bats across Mozambique which suggests that mean species richness across most of the country is 9-16 species (Fig. 2).

Figure 2. Map showing the modelled distribution of species richness across Mozambique (only those bat records used in developing the distribution models are presented here, and include some of the localities in Fig. 1., as well as some earlier museum specimens).

121

Based on the modelled species richness, the number of species of bats varied from six to 30 across the country (Fig. 2). Richness was generally lowest in the coastal plain but increased with increasing altitude and in broken terrain. The mountainous regions along the Zimbabwe border, the Tete province in north-west of the central section, and isolated mountains in northern Mozambique (e.g. Mount Namuli) had the highest richness. However, the alluvial plains along the Save river also exhibited very high species richness. Six species represent new country records: Epomophorus labiatus, Myonycteris relicta, Rhinolophus cf. deckenii, Rhinolophus cf. maendeleo, Kerivoula sp., and Laephotis botswanae. Tadarida ventralis was not recorded by Smithers and Lobão Tello (1976) and therefore constitutes a seventh new species. However, this species had been previously collected and overlooked (see Tadarida ventralis account, below). An eighth species Mimetillus thomasi was recorded for the first time by Cotterill (2001a). This brings the country total to 65 species of bats (Appendix 2) which is comparable to other subtropical southern African countries. For example, 65 bat species have been recorded from Zambia, 63 species from Angola, 62 species from Zimbabwe, and 62 species from Malawi (Ansell, 1978; Happold et al., 1987; Crawford-Cabral, 1989; Cotterill, 1996; Happold and Happold, 1997; Bergmans and van Strein, 2004; Skliba and Petr, 2007). Equatorial African countries typically have higher species richness e.g. 85 bat species recorded in Kenya, 99 in Uganda and 116 in the DRC (Happold et al., 1987; Thorn et al., in press).

Species accounts Pteropodidae Epomophorus crypturus Peters 1852 This species was recorded at six localities north of the Zambezi River (Appendix 1), where it was captured in relatively large numbers (e.g. nine individuals at a single site). Field measurements: FA (adult male) 83.7 ± 1.11 (82.7-84.9, 3); Bm (adult male) 107.7 ± 2.52 (105-110, 3); FA (adult female) 78.7 ± 3.13 (75.7-81.6, 4); Bm (adult female) 79.0 ± 8.29 (70-89, 4).

Epomophorus wahlbergi (Sundevall 1846)

122

This species was recorded at ten sites throughout the country, but was scarce or absent in the far north, where E. crypturus was abundant. The two species were captured sympatrically at two sites, Namapa and Mount Namuli (Appendix 1). Field measurements: FA (adult male) 87.2 ± 2.26 (84.7-89.1, 3); Bm (adult male) 110.0 ± 9.17 (102-120, 3); FA (adult female) 82.6 ± 2.93 (78.5-85.8, 6); Bm (adult female) 92.5 ± 8.04 (81-102, 6).

Epomophorus labiatus (Temminck 1837) Five specimens of E. labiatus were captured at Meponda on the banks of Lake Niassa (Appendix 1). This constitutes a new record for this species in Mozambique. Field measurements: FA (adult female) 59.6 ± 2.23 (57.8-62.1, 3); Bm (adult female) 39.0 ± 8.19 (32-48, 3).

Lissonycteris goliath Bergmans 1997 This species was recorded from two sites in northern Mozambique, where it was relatively abundant in the foothills of Mount Namuli (nine specimens were captured at this site). It was previously known from the Zimbabwe border (Smithers and Lobão Tello, 1976) and from Marromeu, central Mozambique (Cotterill, 2001b). Field measurements: FA (adult female) 83.9 ± 2.11 (81.4-87.4, 6); Bm (adult female) 94.7 ± 5.32 (88-101, 6). Four juveniles had FA ranging 77.2-83.5 and Bm 68-87; far larger than that for Myonycteris relicta (below).

Myonycteris relicta Bergmans 1980 A single specimen of this rarely collected species was captured in Chinizuia Forest and constitutes a new species record for Mozambique. The only other record for southern Africa is from south-east Zimbabwe on the Mozambique border at the Haroni-Rusiti confluence; elsewhere the species occurs in coastal forests of East Africa (Bergmans, 1997). Field measurements: FA (juvenile female) 71.1 (1); Bm (juvenile female) 52 (1).

Rousettus aegyptiacus (E. Geoffroy Saint-Hilaire 1810)

123

This species was recorded at four sites in central and northern Mozambique. A large colony was located in a cave system on the Cheringoma plateau (Appendix 1). Field measurements: FA (adult male) 94.6 ± 1.67 (92.3-98.5, 11); Bm (adult male) 135.8 ± 7.08 (122-149, 11); FA (adult female) 96.8 (96.2-97.3, 2); Bm (adult female) 119 (102-136, 2). There was no overlap in measurements of adult R. aegyptiacus and Lissonycteris goliath (above). However, three juvenile R. aegyptiacus had FA ranging 80.7-90.6 and Bm 75-86 which overlaps considerably with the measurements of L. goliath. The insertion of the wing on the toes is a more reliable character than size for discriminating between these two similar species (insertion is on the 1st toe, or midway between the 1st and 2nd toe, in R. aegyptiacus and on the 2nd toe in L. goliath).

Rhinolophidae Rhinolophus blasii Peters 1867 Eight rhinolophid individuals captured in the foothills of Mount Namuli and two from Mt Mabu were ascribed to R. blasii on the basis of the pointed connecting process, minute 1st upper premolar inside the toothrow, cranial measurements (Table 2) and on molecular grounds. However, peak echolocation frequencies ranged between 93.2-95.4 kHz (ANABAT, Pettersson D240x, n = 10), 8-9 kHz higher than previously recorded for R. blasii in Swaziland and South Africa (Monadjem, 2005; Schoeman and Jacobs, 2008; Monadjem et al., 2007). This difference may be due to geographic variation or may indicate cryptic species. Field measurements: FA (adult male) 44.6 ± 1.10 (43.5-46.4, 5); Bm (adult male) 8.6 ± 0.65 (8.0-9.5, 5); FA (adult female) 45.1 (44.7-45.4, 2); Bm (adult female) 8.3 (8-8.5, 2). Mean nose-leaf width was 8.48 ± 0.33 for males (n = 5) and 8.45 for females (n = 2).

Rhinolophus clivosus Cretzschmar 1828 Several rhinolophid individuals captured in the foothills and montane plateau of Mount Namuli and Mount Mabu were ascribed to R. clivosus on the basis of the rounded connecting process, minute 1st upper premolar outside of the toothrow, cranial measurements (Table 2) and molecular grounds. However, peak echolocation frequencies ranged between 79.8-81.0 kHz (ANABAT, Pettersson D240x, n = 12), 10-11 kHz lower than in South Africa (Schoeman and Jacobs, 2008; Monadjem et al., in press). Despite these differences in

124 echolocation frequency, the Mozambican individuals showed no genetic differentiation (mtDNA control region) to South African R. clivosus. Field measurements: FA (adult male) 52.7 ± 0.67 (51.8-53.5, 7); Bm (adult male) 14.5 ± 0.91 (13.0-15.5, 7); FA (adult female) 54.5 (1); Bm (adult female) 16.0 (1). Mean nose-leaf width was 8.13 ± 0.63 for males (n = 7) and 8.0 for the single female.

Rhinolophus cf. deckenii A single male specimen assigned to this species was collected from Chinizuia forest. It had a rounded connecting process, similar to R. clivosus, but the 1st upper premolar was large and partly within the toothrow. Although the location of the premolar was atypical for R. deckenii (in which it is typically outside the toothrow), this character is variable in R. deckenii and occasionally the premolar is located partly within the toothrow (Csorba et al., 2003). The skull had well developed zygomatic arches, sagittal and occipital crests and moderately inflated anterior medial narial inflations. Cranial measurements (Table 2) fell within the range of values given for R. deckenii in Csorba et al. (2003), although greatest skull length was identical to the minimum value of that recorded for R. deckenii. The baculum is characteristic of R. deckenii (as figured in Cotterill, 2002) in both length (3.8mm cf. 3.9 mm in Cotterill, 2002) and shape (Fig. 3). Peak echolocation frequency of a single male was recorded at 72 kHz (ANABAT, n = 1). Field measurements: FA (adult male) 49.9 (1); Bm (adult male) 15.5 (1). Nose-leaf width was 8.9 for the single male.

Figure 3. Ventral (a) and lateral (b) views of the bacculum of Rhinolophus cf. deckenii (DM 8560), from Chinizuia, central Mozambique.

(a) (b)

125

Rhinolophus fumigatus Rüppell 1842 Identified on the basis of wide noseleaf (>9 mm) and hairy sella, three specimens were collected at two sites in northern Mozambique. Peak echolocation call of a single male was 54 kHz (ANABAT). Field measurements: FA (adult male) 51.6 ± 1.97 (50.1-53.8, 3); Bm (adult male) 12.0 ± 1.00 (11.0-12.0, 3). Mean nose-leaf width was 11.03 ± 0.60 for the three males.

Rhinolophus hildebrandtii Peters 1878 This species was recorded at five sites in southern, central and northern Mozambique. A large colony was discovered in a cave system in the Cheringoma plateau (Appendix 1). Peak echolocation frequencies ranged between 35-40 kHz (ANABAT, Pettersson D240X, n = 15). Based on the analysis of two mtDNA genes (cytochrome b and control region), two divergent lineages of R. hildebrandtii are present in Mozambique, one comprising smaller- sized individuals occurring in savanna habitats at lower elevations (Namapa, Niassa Game Reserve, Gerhard’s Cave) and another comprising large-sized individuals from montane habitats (Mts Mabu and Inago) (Stoffberg et al., unpublished). These two forms are morphologically distinct as shown by the non-overlap between them in most cranial measurements (Table 2). Field measurements: For the low elevation taxon, FA (adult male) 63.3 ± 1.40 (60.1- 65.2, 12); Bm (adult male) 30.9 ± 2.57 (28.0-34.5, 12); FA (adult female) 62.2 ± 2.16 (59.6- 64.6, 4); Bm (adult female) 27.88 ± 5.04 (23.5-34.0, 4). Mean nose-leaf width was 14.36 ± 0.49 for males (n = 12) and 14.35 ± 0.72 for the females (n = 4). For two females from Mounts Mabu and Inago (montane form), mean FA was 67.5 (66-69) mm. Nose-leaf width for these two females was 15.0 and 15.1 mm.

Rhinolophus landeri Martin 1838 Two females and a male of this species were recorded in central and northern Mozambique. The male had bright orange axillary tufts in the armpits. Peak echolocation frequencies were 102.2 kHz (female; ANABAT, n = 1) and 104 kHz (male; Pettersson D240x, n = 1).

126

Field measurements: FA (adult male) 46.3 (1); Bm (adult male) 9.5 (1); FA (adult female) 46.3 (1); Bm (adult female) 10.0 (1). Nose-leaf width was 7.9 for the single female.

Rhinolophus cf. maendeleo Kock, Csorba and Howell 2000 Two specimens assigned to this recently described species were recorded from Mount Namuli in northern Mozambique. They all had a rounded connecting process, similar to R. clivosus, but the 1st upper premolar was small and situated in the toothrow (unlike R. clivosus). The skull was slender and narrow in shape with gracile zygomatic arches (and MW greater or equal to ZYW; see Table 2), undeveloped sagittal and lambdoid crests, a long rostrum with bulbous anterior narial inflation in relation to posterior inflations (giving concave rostral profile) as described by Kock et al. (2000); cranial measurements match closely the values for the holotype and paratype of this species recorded by these authors. However, slight differences between the Mt Namuli male (DM 10833) and the R. maendeleo holotype are present in baculum shape (not shown) and the presence of a bony bar closing the infraorbital foramen (open in holotype and paratype of R. maendeleo but only on the right hand side of one Mt Namuli specimen (DM10833) and on neither side in DM10839). These differences warrant further analyses to determine whether these individuals represent an undescribed species, preferably including molecular comparisons of the Mt Namuli specimens with the holotype and/or paratype. Specimens from Mt Gorongosa and Nyika Plateau of Malawi may also be referable to this species and should be examined. Field measurements: FA (adult male) 47.5 (1); FA (adult female) 48.9 (1). Nose-leaf width was 8.7 for the male and 8.3 for the female.

Rhinolophus simulator K. Andersen 1904 A single specimen assigned to this species was collected just south of the Niassa Game Reserve, northern Mozambique. Although it matched R. simulator in most characters (including cranial measurements - Table 2), the anterior upper premolar was tiny and mostly outside the toothrow (inside toothrow in R. simulator) and the ear length was 17mm which is extremely short for this species. Due to technical difficulties its echolocation call could not be recorded.

127

Field measurements: FA (adult male) 42.4 (1); Bm (adult male) 6 (1). Nose-leaf width was 7.8 for the single male.

Rhinolophus cf. swinnyi A series of small rhinolophids was collected from a cave (“Gerhard’s cave”) south of the Save river and from Mt Inago in northern Mozambique. On molecular grounds (cytochrome b) these specimens group with typical R. simulator from which they differ both in morphology (smaller skull lengths) and echolocation, with peak echolocation frequencies ranging between 99-103 kHz (Pettersson D980, n = 10). Peak echolocation frequencies of R. simulator in South Africa are 20 kHz lower (Schoeman and Jacobs, 2008). This suggests that these individuals may represent an, as yet, undescribed species. Field measurements: FA (adult male) 42.5 ± 0.56 (42.0-43.3, 5); Bm (adult male) 6.0 ± 0.82 (5.0-7.0, 4); FA (adult female) 42.5 ± 0.60 (41.6-43.1, 5); Bm (adult female) 6.7 ± 0.97 (5.5-8.0, 5). Mean nose-leaf width was 7.24 ± 0.44 for males (n = 5) and 7.38 ± 0.31 for the females (n = 5). See Table 2 for cranial measurements.

Hipposideridae Hipposideros caffer (Sundevall 1846) This species was recorded at nine sites throughout Mozambique. Hipposideros caffer and H. ruber (below) were not recorded sympatrically although there are no obvious ecological differences in the habitats selected by these two species within Mozambique. Further survey work is needed to determine whether their distributions overlap in northern Mozambique. Peak echolocation frequencies of two individuals were recorded at 145 kHz (ANABAT). Field measurements: FA (adult male) 45.9 ± 1.55 (44.1-48.7, 13); Bm (adult male) 7.4 ± 1.21 (6.0-9.5, 12); FA (adult female) 46.6 ± 1.50 (44.1-47.8, 6); Bm (adult female) 6.5 ± 0.50 (6.0-7.0, 3).

Hipposideros ruber (Noack 1893) This species is represented by just five individuals recorded at two sites in northern Mozambique, which constitute the second record of this species for Mozambique (Fenton, 1986). This species is considerably larger than H. caffer from which it can reliably be

128 distinguished by forearm length (>50 mm in H. ruber and <49 mm in H. caffer), and also the enlarged nasal compartments (Fenton, 1986). Little is known about the ecology of this species in southern Africa and only a few specimens have been collected in the region. Peak echolocation frequencies ranged between 130.5 (Pettersson D240x, n = 1) and 132-136 kHz (ANABAT, n = 4). Field measurements: FA (adult male) 51.7 ± 0.71 (51.2-51.5, 4); Bm (adult male) 10.3 ± 0.50 (10.0-11.0, 4); FA (adult female) 50.0 (1); Bm (adult female) 9.0 (1).

Hipposideros vittatus (Peters 1852) Five specimens of this species were collected at two sites in central Mozambique. A large colony was discovered in a cave system in the Cheringoma plateau (Appendix 1). Peak echolocation frequencies ranged between 64-66 kHz (ANABAT, n = 2). Field measurements: FA (adult male) 97.0 (1); Bm (adult male) 143 (1); FA (adult female) 96.2 (1); Bm (adult female) 100 (1).

Triaenops persicus Dobson, 1871 This species was recorded at seven sites in southern, central and northern Mozambique. Fifteen individuals were captured at the entrance to a cave system in the Cheringoma plateau suggesting that a large roosting colony occurs there. Another large colony occurs just south of the Save river. However, the core of its southern African distribution is central and northern Mozambique, with marginal intrusion into Zambia and Zimbabwe and south of the Save river. Echolocation calls are sexually dimorphic: peak echolocation frequencies of males ranged between 71-75 kHz (n = 7) and those of females between 82-85 kHz (ANABAT, n = 11). Taylor et al. (2005) presented similar evidence for sexual dimorphism in the peak echolocation frequency of this species in Zambia. Field measurements: FA (adult male) 54.1 ± 1.18 (51.7-55.8, 16); Bm (adult male) 12.1 ± 1.00 (10.0-13.0, 11); FA (adult female) 52.4 ± 1.34 (49.8-55.2, 16); Bm (adult female) 10.8 ± 1.42 (8.0-12.5, 9).

Nycteridae Nycteris grandis Peters 1865

129

A single female was collected from within a hollow baobab tree (Adansonia digitata) at Pemba, northern Mozambique. This is only the third record of this species from Mozambique (Monadjem et al., in press). Field measurements: FA (adult female) 61.5 (1); Bm (adult female) 30 (1).

Nycteris hispida (Schreber 1774) Five specimens were recorded from southern and northern Mozambique. Nycteris hispida is considerably smaller than N. thebaica with shorter ears and trifid (not bifid) upper incisors (Monadjem et al., in press). Field measurements: FA (adult male) 39.6 (38.9-40.2, 2); Bm (adult male) 7.5 (7.0-8.0, 2); FA (adult female) 37.7 (35.3-41.7, 3); Bm (adult female) 6.7 (6.0-7.0, 3). Ear length of a single male was 18.5.

Nycteris macrotis Dobson, 1876 Four specimens were collected from southern, central and northern Mozambique. Live individuals can be confused with N. thebaica. But N. macrotis is slightly larger with longer forearm, and is easily distinguished by the semi-lunate tragus. The skull morphology of the two species differs significantly with N. macrotis having a more robust skull with a longer condylo-incisive length (>18.8 mm) (Monadjem et al., in press). Field measurements: FA (adult male) 52.0 ± 1.40 (50.4-53.0, 3); Bm (adult male) 16.7 (15.0-17.5, 3); FA (adult female) 52.2 (1); Bm (adult female) 16.0 (1).

Nycteris thebaica E. Geoffroy, 1813 Six specimens were recorded from four sites in central Mozambique. This species can easily be confused with N. macrotis but is slightly smaller and more gracile (see above), and has a finger-like tragus (Monadjem et al., in press). Field measurements: FA (adult male) 45.9 ± 1.77 (43.9-47.8, 4); Bm (adult male) 9.3 ± 1.50 (8.0-11.0, 4); FA (adult female) 47.1 ± 0.76 (46.6-48.0, 3); Bm (adult female) 10.3 ± 1.04 (9.5-11.5, 3). Ear length of a single male was 29.3.

Molossidae

130

Chaerephon ansorgei (Thomas, 1913) Nine individuals were netted across a small stream in northern Mozambique. Prior to our expeditions there was only a single record of this species from Chiutu, central Mozambique (Monadjem et al., in press). Field measurements: FA (adult male) 45.7 (1); Bm (adult male) 22 (1); FA (adult female) 46.2 ± 0.78 (44.8-47.3, 8); Bm (adult female) 19.1 ± 1.10 (18.0-20.0, 8).

Chaerephon pumilus (Cretzschmar, 1826) This species was recorded from four sites across the country. It has previously been collected widely in southern and central Mozambique (Smithers and Lobão Tello, 1976). Field measurements: FA (subadult male) 38.0 (1); Bm (subadult male) 8.0 (1); FA (adult female) 38.0 ± 0.76 (36.9-39.0, 7); Bm (adult female) 11.2 ± 1.04 (9.5-12.0, 5).

Mops condylurus (A. Smith, 1833) This species was recorded at six sites in southern and central Mozambique. It has not been recorded from northern Mozambique, where the rarely collected Mops niveiventer may occur (although the latter species has yet to be recorded from Mozambique). Field measurements: FA (adult male) 47.4 ± 1.54 (45.4-49.7, 15); Bm (adult male) 26.0 ± 3.25 (21.0-31.0, 15); FA (adult female) 42.2 ± 1.28 (43.3-49.8, 29); Bm (adult female) 27.1 ± 2.99 (21.0-33.0, 24).

Sauromys petrophilus Roberts 1917 Three individuals were recorded at a single site in northern Mozambique. Previously, this species was only known from Chiutu, Tete Province, central Mozambique (Smithers and Lobão Tello, 1976). Field measurements: FA (adult male) 35.9 (1); Bm (adult male) 7.0 (1); FA (adult female) 36.2 (35.5-37.6, 2); Bm (adult female) 7.0 (6.0-8.0, 2).

Tadarida aegyptiaca (E. Geoffroy 1818)

131

Two specimens were netted across a small stream at a single site in northern Mozambique. In Mozambique, T. aegyptiaca appears to be associated with broken, hilly and mountainous terrain and has not yet been recorded from the flat coastal plain. Field measurements: FA (adult male) 49.6 (1); Bm (adult male) 21.0 (1); FA (adult female) 49.8 (1); Bm (adult female) 21.0 (1).

Tadarida fulminans (Thomas 1903) A single specimen was netted across a small stream in northern Mozambique, which constitutes only the third record of this poorly collected species in the country. The species was not reported by Smithers and Lobão Tello (1976), but two specimens collected from the Tete province in central Mozambique are housed in the Smithsonian Institute (Monadjem et al., in press). Field measurements: FA (juvenile female) 59.3 (1); Bm (juvenile female) 36.0 (1).

Vespertilionidae Glauconycteris variegata (Tomes 1861) Two individuals of this species were netted over a small water body in mature woodland along the Save river in southern Mozambique. Five previous specimens were collected from southern and central Mozambique (Smithers and Lobao Tello, 1976; Monadjem et al., in press). Field measurements: FA (adult male) 43.1 (1); FA (adult female) 42.6 (1); Bm (adult female) 15.5 (1).

Kerivoula argentata Tomes, 1861 Two specimens were collected in southern Mozambique. One specimen was captured in mature woodland along the Save river, the other in coastal forest in the Maputo Special Reserve. Previously, five specimens were collected from scattered localities across the country (Smithers and Lobao Tello, 1976). Field measurements: FA (adult male) 36.7 (36.6-36.7, 2); Bm (adult male) 8.0 (7.0-8.0, 2).

132

Kerivoula lanosa (A. Smith, 1847) One specimen was collected in mature woodland along the Save river in southern Mozambique. This constitutes only the second record of this species in the country. Field measurements: FA (adult male) 29.7 (1); Bm (adult male) 4.0 (1).

Kerivoula sp. A single adult male of this species was recorded in mid-altitude forest on Mount Mabu. It is smaller than the two southern African species, K. lanosa and K. argentata. It exhibits a dark brown coloration but lacks the characteristic grizzling. The species may be undescribed, or may represent a new record of an existing species. The two candidate species in Africa that match the overall size and coloration of this enigmatic specimen are Kerivoula africana Dobson 1878 and Kerivoula phalaena Thomas 1912. However, the distinction between these two species is unclear based on the original species descriptions. Both species have similar forearm measurements (28 in K. africana, 28-29.5 (2) in K. phalaena) and ear length (12.7 in K. africana and 13 in K. phalaena). Beyond this, the descriptions contain no further overlapping characters. Only comparisons with museum material (including the holotype of K. africana) can establish the affinities of this specimen, and whether it might represent a new species. In either case, this Mount Mabu specimen represents a most interesting discovery for Mozambique. Field measurements: FA (adult male) 27.5 (1), BM (adult male) 3.5 (1)

Laephotis botswanae Setzer 1971 A single adult male specimen of Laephotis botswanae was recorded from the lower slopes (ca. 550m) of Mount Mabu (MNHG 1971.009). Identification was based on a principal component analysis of cranial measurements with the dataset from Kearney and Seamark (2005). The Mabu specimen fell clearly within specimens attributed by Kearney and Seamark (2005) to L. botswanae (data not shown). Three checklists include L. botswanae in the Mozambique bat fauna (Taylor, 2000; Stuart and Stuart, 2001; Schneider et al., 2005), occurring in the Manica province, near the border to Zimbabwe (Taylor, 2000). However, the validity of this record is in doubt as the record was omitted from later publications (Cotterill, 1996; Bronner et al., 2003; van Cakenberghe and Seamark, 2008; Monadjem et al.,

133 in press). Therefore this specimen vouches for a new distribution record for Mozambique. The paucity of museum records for L. botswanae suggests that this species is rare throughout its range across Southern Africa (Taylor, 2000; Kearney and Seamark, 2005). However, the species has been commonly recorded throughout southern Malawi (Happold et al., 1987; Happold and Happold, 1989, 1997), including numerous records from lowland forest fragments around the base of Mount Mulanje, only ca. 60 km away from Mount Mabu (Curran and Kopp, unpublished data). Hence, the discovery of L. botswanae in Mozambique close to the Malawi border was not surprising. Field measurements: FA (adult male) 35.2 (1), BM (adult male) 6 (1).

Myotis bocagii (Peters 1870) Five specimens were captured at a single site on the shores of Lake Niassa in northern Mozambique. Previously, one specimen was recorded from central Mozambique (Smithers and Lobão Tello, 1976). Field measurements: FA (adult male) 38.9 (1); Bm (adult male) 9.0 (1); FA (adult female) 40.8 ± 1.48 (39.3-42.8, 4); Bm (adult female) 9.8 ± 0.65 (9.0-10.5, 4).

Myotis tricolor (Temminck 1832) This species was collected from Mount Namuli and Mount Chiperone in northern Mozambique. Two prior records exist for central Mozambique (Smithers and Lobão Tello, 1976). Field measurements: FA (adult male) 49.9 (48.0-51.7, 2); Bm (adult male) 13.5 (13.0- 14.0, 2).

Neoromicia capensis (A. Smith 1829) Four specimens were captured in central Mozambique. It was previously widely recorded from southern and central Mozambique (Smithers and Lobão Tello, 1976). Field measurements: FA (adult male) 34.5 (32.7-36.3, 2); Bm (adult male) 7.5 (7.0-8.0, 2); FA (adult female) 36.4 (35.8-37.0, 2); Bm (adult female) 7.8 (7.5-8.0, 2).

Neoromicia nanus (Peters 1852)

134

This common species was netted widely throughout Mozambique, and characteristically found roosting within a rolled-up banana leaf at the base of Mount Mabu. Field measurements: FA (adult male) 30.1 ± 1.90 (26.7-32.5, 22); Bm (adult male) 3.4 ± 0.44 (3.0-4.0, 21); FA (adult female) 30.7 ± 1.50 (27.7-32.0, 22); Bm (adult female) 4.1 ± 1.08 (3.0-6.5, 21).

Neoromicia rendalli (Thomas 1889) A single specimen was captured in southern Mozambique, which constitutes the third record of this species for the country (Monadjem et al., in press). FA (adult female) 37.6 (1); Bm (adult female) 9.5 (1).

Neoromicia zuluensis (Roberts 1924) This species was recorded at three sites in central and northern Mozambique. Field measurements: FA (adult male) 30.6 ± 1.15 (29.7-31.9, 3); Bm (adult male) 4.0 ± 0 (4.0, 3); FA (adult female) 30.0 ± 1.18 (28.8-31.2, 4); Bm (adult female) 4.1 ± 0.25 (4.0- 4.5, 4).

Nycticeinops schlieffeni (Peters 1859) This species was collected from nine sites across the country. It was previously collected from southern and central Mozambique (Smithers and Lobão Tello, 1976). Field measurements: FA (adult male) 30.9 ± 0.42 (30.6-31.4, 3); Bm (adult male) 6.0 ± 0.50 (5.5-6.5, 3); FA (adult female) 31.4 ± 1.16 (29.6-33.2, 13); Bm (adult female) 5.7 ± 0.82 (5.0-7.0, 11).

Pipistrellus hesperidus Temminck 1840 Thirteen specimens were collected in the foothills and montane plateau of Mount Namuli, northern Mozambique and a further individual from the Rio Wanetsi in southern Mozambique (Appendix 1). Two prior specimens were recorded from central and southern Mozambique (Smithers and Lobão Tello, 1976). Hence, the species occurs throughout the country.

135

Field measurements: FA (adult male) 30.9 ± 1.03 (29.8-31.8, 3); Bm (adult male) 5.0 ± 0.50 (4.5-5.5, 3); FA (adult female) 31.5 (31.3-31.6, 2); Bm (adult female) 6.8 (6.5-7.0, 2).

Scotoecus hindei/albigula Thomas 1901, 1909 Six individuals referable to either Scotoecus hindei or S. albigula were collected from four sites across the country. There is confusion regarding the specific status of hindei/albigula (Monadjem et al., in press), and therefore we have chosen to lump these two species until the genus has been revised. Field measurements: FA (adult male) 34.2 ± 1.08 (33.4-35.4, 3); Bm (adult male) 11.7 ± 0.76 (11.0-12.5, 3); FA (adult female) 33.5 ± 0.71 (32.9-34.0, 3); Bm (adult female) 8.0 (1).

Scotophilus dinganii (A. Smith, 1833) Twenty-five individuals were recorded from southern and northern Mozambique. It is possible that the species has been overlooked in central Mozambique since it was recorded there in the past (Smithers and Lobão Tello, 1976). Field measurements: FA (adult male) 52.7 ± 2.21 (50.2-56.6, 7); Bm (adult male) 23.9 ± 2.94 (21.5-29.5, 7); FA (adult female) 53.4 ± 1.99 (50.3-57.2, 18); Bm (adult female) 26.4 ± 5.17 (18.0-37.0, 18).

Scotophilus leucogaster (Cretzschmar 1826) A single individual was collected from mopane woodland near the Kruger National Park boundary in southern Mozambique. This constitutes the first record of this species for the country. Field measurements: FA (adult male) 48.3 (1); Bm (adult male) 23.0 (1).

Scotophilus cf. viridis (Peters, 1852) Twenty-five individuals were recorded from southern and northern Mozambique. The species has probably been overlooked in central Mozambique since it has been recorded there in the past (Smithers and Lobão Tello, 1976). Toward absolving the confusion associated with the status of this taxon, the smallest yellow-bellied species of Scotophilus in southern Africa were assigned S. viridis by Monadjem et al. (in press), distinct from the similar

136 sized, but cream-bellied S. leucogaster. The recent paper by Jacobs and Barclay (2009), assigned the name S. “mhlanganii” for their studied population of this small, yellow-bellied Scotophilus, but this most unfortunate introduction of a nomen nudum into an already confusing taxonomy has created more problems than it solved (Monadjem et al., in press). Field measurements: FA (adult male) 46.4 ± 1.22 (44.1-48.2, 17); Bm (adult male) 19.2 ± 2.90 (16.0-24.5, 16); FA (adult female) 47.7 ± 2.06 (45.4-51.7, 8); Bm (adult female) 26.1 ± 7.06 (17.0-33.0, 6). Miniopteridae Miniopterus cf. fraterculus (Peters 1867) Five adult male bats referable to this species were captured on the lower slopes of Mount Namuli. The two specimens that were collected had greatest skull lengths of 14.3 and 14.8 mm, with head lengths of 16.0 and 16.2 mm, respectively. The three released individuals had head lengths of 15.8, 16.1 and 16.8 mm. These head and skull measurements are significantly shorter than Miniopterus natalensis but within the range of M. fraterculus. The closest population of M. fraterculus to that on Mount Namuli (and the nearby Mount Zomba in neighbouring Malawi) is >1000 km away in South Africa (Monadjem et al., in press) suggesting that this isolated population may be specifically distinct. Field measurements: FA (adult male) 43.7 ± 0.95 (42.1-44.5, 5); Bm (adult male) 8.0 ± 0.35 (7.5-8.5, 5).

Miniopterus inflatus Thomas 1903 Eight individuals were captured at the entrance of a cave system in the Cheringoma plateau (Appendix 1), suggesting that a large colony was roosting within the caves. Additional records are from Mount Chiperone and Mount Namuli in northern Mozambique. These constitute the second record of this species from Mozambique. The greatest skull length of two individuals was 16.4 and 16.6 mm, with head lengths of 18.4 and 18.6 mm, respectively. The other six individuals had head lengths of 17.6-18.3 mm, larger than that of Miniopterus natalensis but within the range of M. inflatus. Field measurements: FA (adult male) 47.3 ± 0.47 (46.7-48.1, 7); Bm (adult male) 15.0 ± (14.0-15.0, 7); FA (adult female) 47.0 (1); Bm (adult female) 13.0 (1).

Miniopterus natalensis (A. Smith 1833) 137

Three individuals were captured at Gerhard’s Cave and Inhambane in southern Mozambique. Molecular analyses based on cytochrome b show that the individuals from Gerhard’s Cave are similar to M. natalensis (AJ841977.1; Stadelmann et al., 2004) from Springbok, South Africa. Field measurements: FA (adult male) 47.1 (46.7-47.5, 2); Bm (adult male) 10.5 ± (9.0- 12.0, 2); FA (adult female) 46.3 (1); Bm (adult female) 10.5 (1).

138 Table 2. Cranial measurements (mean, range and number of specimens) of Rhinolophus specimens collected recently in Mozambique (preserved in the Mammal Collection of the Durban Natural Science Museum). For measurement codes see Sampling and data analysis.

Species GSL CCL ZW MW M3M3 CM3 R. landeri 19.55 (19.4-19.7), 2 17.20 (17.0-17.4), 2 10.30 (10.1-10.5), 2 9.57 (9.5-9.6), 2 7.16 (7.0-7.3), 2 7.36 (7.1-7.6), 2 R. cf. swinnyi 17.73 (17.1-18.2), 5 15.46 (15.1-15.7), 6 8.68 (8.2-9.0), 6 8.58 (7.9-8.8), 6 6.04 (5.8-6.3), 6 6.30 (6.2-6.4), 6 R. fumigatus 22.7 (22.4-23.0), 2 19.67 (19.3-20.3), 3 11.37 (11.1-11.6), 3 10.53 (10.4-10.7), 3 8.22 (8.0-8.3), 3 8.17 (8.0-8.4), 3 R. clivosus 22.6 (22.1-23.0),6 19.74 (19.2-20.0), 7 11.68 (11.1-12.0), 7 10.26 (10.0-10.6), 7 8.18 (7.9-8.4), 7 8.27 (8.2-8.4), 7 R. deckenii 22.55 19.50 12.02 10.70 8.37 8.51 R. cf. maendeleo 19.88 17.28 (17.2-17.3), 2 9.44 (9.4-9.5), 2 9.48 (9.4-9.5), 2 6.7 (6.6-6.8), 2 7.02 (7.01-7.04), 2 R. blasii 18.68 (18.2-19.4), 4 16.0 (15.7-16.1), 4 8.87 (8.8-9.0), 5 8.83 (8.7-8.9), 4 6.07 (5.8-6.2), 5 6.32 (6.2-6.5), 5 R. simulator 18.80 15.90 9.08 8.58 6.41 6.31 R. hildebrandtii Clade1 (large, 29.99 (29.7-30.3), 2 25.58 (25.2-25.9), 2 14.21 (14.1-14.3), 2 13.16 (13.1-13.2), 2 9.82 (9.6-10.1), 2 10.48 (10.3-10.7), 2 montane) Clade2 (small, 27.71 (26.7-28.5), 3 24.17 (23.7-25.0), 3 13.64 (13.2-14.2), 3 12.35 (12.1-12.7), 3 9.47 (9.3-9.7), 3 9.86 (9.7-10.0), 3 savanna)

Table 2 (Cont’d)

Species IOW NW NL NH BCW CIL R. landeri 2.56 (2.5-2.6), 2 5.40 (5.4-5.4), 2 16.45 (16.3-16.6), 2 6.16 (6.0-6.3), 2 9.09 (9.0-9.2), 2 17.25 (17.2-17.3), 2 R. cf. swinnyi 2.26 (2.0-2.6), 6 4.20 (4.0-4.4), 6 14.95 (14.6-15.3), 6 5.31 (5.1-5.5), 6 8.11 (7.5-8.4) 15.78 (15.5-16.1), 3 R. fumigatus 2.96 (2.8-3.2), 3 6.15 (6.0-6.2), 2 19.33 (19.0-19.6), 3 7.49 (7.4-7.9), 3 9.95 (9.8-10.2), 3 20.30 (19.8-20.8), 2 R. clivosus 2.79 (2.6-3.0), 7 6.10 (5.8-6.7), 7 18.76 (18.5-19.0), 7 7.05 (6.8-7.3), 7 10.06 (9.8-10.4), 7 20.22 (19.3-20.9), 6 R. deckenii 2.88 5.48 18.64 7.01 10.40 19.60 R. cf. maendeleo 2.61 (2.5-2.7), 2 4.94 (4.9-5.0), 2 16.59 (16.5-16.6), 2 5.82 (5.80-5.84), 2 8.89 (8.8-8.9), 2 17.65 R. blasii 2.51 (2.4-2.7), 5 4.84 (4.7-4.9), 5 15.84 (15.6-16.1), 4 5.21 (5.0-5.3), 5 8.48 (8.3-8.6), 5 16.76 R. simulator 1.99 4.81 15.90 5.42 8.07 16.60 R. hildebrandtii Clade1 (large, 3.54 (3.5-3.6), 2 7.82 (7.6-8.0), 2 24.58 (24.4-24.6), 2 10.08 (9.6-10.6), 2 12.39 (12.3-12.5), 2 26.83 (26.5-27.2), 2 montane) Clade2 (small, 3.44 (3.4-3.5), 3 7.58 (7.2-8.2), 3 23.46 (22.8-24.3), 3 9.40 (8.9-9.9), 3 11.85 (11.6-12.1) 24.73 (23.8-25.6), 3 savanna)

Conclusions Though still in its infancy, our knowledge of the bat fauna of Mozambique has been much improved by this series of surveys. The number of known bat species has been elevated to 65, a comparable number to other similarly sized countries in Southern Africa. It has shed light on important taxonomic issues that should be addressed with further survey work and molecular analyses. Moreover, in a scientific world increasingly dominated by whole genome biology and dependence on the availability of expensive technology, surveys are often under- valued. However, large-scale surveys provide precious occurrence data which allow us to investigate species’ ecological requirements, vulnerability to extinction, and future distribution patterns using for example predictive modelling techniques.

Acknowledgements

This is the 16th communication of the All Out Africa Research Unit (www.alloutAfrica.org). AM is grateful to Kim Roques from All Out Africa for financial and logistic support and to Julien Cornut for assistance in the field. Specimens collected by JB, MC and MK were collected under the British funded Darwin Initiative Award (15/036) ‘Monitoring and Managing Biodiversity Loss in South-East Africa's Montane Ecosystems’. This is a project coordinated by the Royal Botanic Gardens Kew and BirdLife International (UK) in collaboration with the Mozambique Agricultural Research Institute (IIAM) and the Mulanje Mountain Conservation Trust (MMCT) in Malawi. MC and MK were further funded by a National Geographic Society / Waitt Grants Program Award (W37-08). DVP is grateful to Alexander Arnold for assistance in the field and was supported by a EU Marie-Curie Early Stage Research Training grant: HOTSPOTS MEST-CT-2005-020561. MCS and SS are very grateful to the Toyota Enviro Outreach for logistical support.

140

References Amaral, H. and R. Sommerhalder. 2004. The Limpopo River Basin – Case Study on Science and Politics of International Water Management. ETH, Zurich. Ansell, W.F.H. 1978. The mammals of Zambia. The National Parks and Wildlife Service, Zambia, Chilanga. Bergmans, W. 1997. Taxonomy and biogeography of African fruit bats (Mammalia, Megachiroptera). 5. The genera Lissonycteris Andersen, 1913, Myonycteris Matschie, 1899 and Megaloglossus Pagenstecher, 1885; General remarks and conclusions; annex: Key to all species. Beaufortia, 47: 11-90. Bergmans, W, and N. J. van Strien. 2004. Systematic notes on a collection of bats from Malawi. I. Megachiroptera: Epomophorinae and Rousettinae (Mammalia, Chiroptera). Acta Chiropterologica, 6: 249-268. Bronner, G.N., M. Hoffmann, P.J. Taylor, C.T. Chimimba, P.B. BEST, C.A. Matthee and T.J. Robinson. 2003. A revised systematic checklist of the extant mammals of the southern African subregion. Durban Museum Novitates, 28: 56-106. Cotterill, F. P. D. 1996. New distribution records of insectivorous bats of the families Nycteridae, Rhinolophidae and Vespertilionidae (Microchiroptera: Mammalia) in Zimbabwe. Arnoldia Zimbabwe, 10: 71-89. Cotterill, F.P.D. 2001a. The first specimen of Thomas’ flat-headed bat, Mimetillus moloneyi thomasi (Microchiroptera:Mammalia) in southern Africa from Mozambique. Arnoldia Zimbabwe, 10: 211-218. Cotterill, F. P. D. 2001b. New records for two species of fruit bats (Megachiroptera: Mammalia) in southeast Africa, with taxonomic comments. Durban Museum Novitates, 26: 53-56. Cotterill, F.D.P. 2002. A new species of horseshoe bat (Microchiroptera: Rhinolophidae) from south-central Africa: with comments on its affinities and evolution, and the characterization of rhinolophid species. Journal of Zoology, London, 256: 165-179. Csorba, G., P. Ujheli, and N. Thomas. 2003. Horseshoe Bats of the World (Chiroptera: Rhinolophidae). Alana books, Shropshire, UK, 160 pp. Crawford-Cabral, J. 1989. A list of Angolan Chiroptera with notes on their distributions. Garcia de Orta, Ser. Zool. 13: 7-48. Fenton, M. B. 1986. Hipposideros ruber in Zimbabwe. Journal of Zoology, London, 210: 347- 353 Hanlon, J. 2007. Is poverty decreasing in Mozambique? Proceedings of the Inaugural Conference of the Instituto de Estudos Sociais e Economicos, Maputo. Happold, D.C.D., M. Happold, and J.E. Hill. 1987. The bats of Malawi. Mammalia, 51: 337- 414. Happold, D.C.D., and M. Haappold. 1989. The bats (Chiroptera) of Malawi, central Africa: checklist and keys for identification. Nyala, 14: 89-114. Happold, D.C.D., and M. Happold. 1997. New records of bats (Chiroptera: Mammalia) from Malawi, east-central Africa, with an assessment of their status and conservation. Journal of Natural History, 31: 805-836. Heller, K. –G., and O. von Helversen 1989. Resource partitioning of sonar frequency bands in rhinolophoid bats. Oecologia, 80: 178-186. IUCN. 2009. www.iucnredlist.org

141

Jacobs, D. S. and Barclay, R. M. R. 2009. Niche differentiation in two sympatric sibling bat species Scotophilus dinganii and Scotophilus mhlanganii. Journal of Mammalogy, 90: 879– 887. Kearney, T.C., and E.C. Seamark. 2005. Morphometric analysis if cranial and external characters of Laephotis Thomas, 1901, (Mammalia: Chiroptera: Vespertilionidae) from southern Africa. Annals of the Transvaal Museum, 42: 71-87. Kock, D., Csorba, G. and Howell, K. M. 2000. Rhinolophus maendeleo n. sp. from Tanzania, a horseshoe bat noteworthy for its systematics and biogeography. Senckenbergiana Biologica, 80: 233–239. Mitchell-Jones, A.J., G. Amori, W. Bogdanovicz, B. Krystufek, P.J.H. Reijnders, F. Spitzenberger, M. Stubbe, J.B.M. Thissen, V. Vohralik, and J. Zima. 1999. The atlas of European mammals. Poyser Natural History, London, 484 pp. Monadjem, A. 2005. Recording of the call of the Peak-saddle Horseshoe Bat (Rhinolophus blasii Peters, 1867) from Swaziland. African Bat Conservation News, 3: 5-6. Monadjem, A., A. Reside, and L. Lumsden 2007. Echolocation calls of rhinolophid and hipposiderid bats in Swaziland. South African Journal of Wildlife Research, 37: 9-15. Monadjem, A., P.J. Taylor, F.P.D. Cotterill, and M.C. Schoeman. In press. Bats of Southern Africa: A Biogeographic and Taxonomic Synthesis. University of the Witwatersrand, Johannesburg. Obrist, M.K. 1995 Flexible bat echolocation: the influence of individual, habitat and conspecifics on sonar signal design. Behavioral Ecology and Sociobiology 36: 207-219. Parker, V. 1999. The atlas of the birds of Sol du Save, southern Mozambique. Avian Demography Unit & Endangered Wildlife Trust, Cape Town & Johannesburg, 276 pp. Phillips, S. J., Anderson, R. P. A. and Chapire, R. E. S. 2006. Maximum entropy modeling of species geographic distributions. Ecological Modelling, 190: 231–259. Schneider, M., V. Buramuge, L. Aliasse, and F. Serfontein. 2005. Checklist and Centres of Vertebrate Diversity in Mozambique. Forestry Department (DEF), Eduardo Mondlane University, Maputo, Mozambique. Schoeman, M.C. and D.S. Jacobs 2008. The relative influence of competition and prey defenses on the phenotypic structures of insectivorous bat ensembles in southern Africa. PLoS ONE 3 (11): e3715. doi:10.1371/journal.pone.0003715 Skliba, R. and B. Petr. 2007. Bocage's fruit bat (Lissonycteris angolensis), a new species for Malawi. Nyala, 24: 61-63. Skinner, J.D., and C.T. Chimimba. 2005. The mammals of the southern African subregion. Cambridge University Press, Cambridge, 814 pp. Smithers, R.H.N. and J.L.P. Lobao Tello. 1976. Check list and atlas of the mammals of Mozambique. Museum Memoir, The Trustees of the National Museums and Monuments of Rhodesia, 8: 1-184. Smithers, R.H.N., and V.J. Wilson. 1979. Check list and atlas of the mammals of Zimbabwe Rhodesia. Museum Memoir, The Trustees of the National Museums and Monuments of Rhodesia, 9: 1-193. Stadelmann, B.,D.S. Jacobs, M.C. Schoeman, and M. Ruedi. 2004. Phylogeny of African Myotis bats (Chiroptera, Vespertilionidae) inferred from cytochrome b sequences. Acta Chiropterologica 6: 177–192.

142

Stuart, C., and T. Stuart. 2001. Field guide to the mammals of southern Africa. Struik, Cape Town, 307 pp. Taylor, P.J. 2000. Bats of Southern Africa. Guide to biology, identification and conservation. University of Natal Press, Pietermaritzburg, 206 pp. Taylor, P.J., C. Geiselman, P. Kabochi, B. Agwanda, and S.Turner. 2005. Intraspecific variation in the calls of some African bats (Order Chiroptera). Durban Museum Novitates, 30: 24-37. Thorn, E., J. Kerbis-Peterhans, and J. Baranga. In press. Uganda small mammals other than rodents. Bonner Zoologische Monographien. UNEP-WCMC. 2003. United Nations Environment Programme - World Conservation Monitoring Centre. World Database on Protected Areas (WDPA) Version 6. Cambridge, U.K.:WCMC. Van Cakeberger, V., and E.C.J. Seamark. 2008. African Chiroptera Report 2008. African Chiroptera Project, Pretoria. www.africanbats.org Wilson, D.E. and D.M. Reeder. 2005. Mammal Species of the World. The John Hopkins University Press, Baltimore, 2000 pp.

Appendix 1. Sites at which specimens were collected during this study, bat species recorded at each site and museum(s) in which specimens have been deposited (TM – Transvaal Museum; DM – Durban Natural Science Museum; MHNG - Natural History Museum of Geneva; MM - Maputo Natural History Museum). The sites are arranged roughly from south to north. The site numbers correspond with the numbers in Fig. 1. For each site, species are listed in alphabetical order. 143

Site Site Latitude Longitude Species number Southern Mozambique Ponto D’Ouro 1 -26.84049 32.87484 Chaerephon pumilus, Hipposideros caffer, Mops condylurus Maputo Special 2 -26.34534 32.92919 Epomophorus wahlbergi (TM), Kerivoula Reserve argentata (TM), Scotophilus dinganii (TM) Maputo 3 -25.9653 32.589 Epomophorus wahlbergi, Neoromicia nana, Scotophilus dinganii, Scotophilus cf. viridis Palmiera 4 -25.21716 32.8321 Neoromicia rendalli (DM) Paradise Magoo 5 -25.0146 34.011 Epomophorus wahlbergi, Neoromicia nana, Pipistrellus hesperidus Chidenguele 6 -24.95743 34.18951 Scotophilus cf. viridis (TM) Magude, north of 7 -24.94134 32.54862 Neoromicia nana (DM), Nycticeinops schlieffeni (DM), Scotoecus hindei/albigula (DM), Scotophilus cf. viridis (DM) Inhambane 8 -24.7886 34.31278 Miniopterus natalensis, Mops condylurus, Neoromicia nana, Nycteris hispida, Scotophilus dinganii Rio Wanetsi 9 -24.49252 32.17136 Pipistrellus hesperidus (DM) Chokwe 10 -24.40597 32.88168 Mops condylurus (DM) Mepuze, south of 11 -24.16792 32.80604 Nycteris macrotis (DM) Mepuze 12 -23.20518 32.49864 Hipposideros caffer (DM), Scotophilus leucogaster (DM) Bazaruto 13 -21.661 35.488 Mops condylurus, Nycteris hispida Massangena 14 -21.55513 32.96057 Epomophorus wahlbergi (DM), Hipposideros caffer (DM), Neoromicia nana (DM), Nycticeinops schlieffeni (DM), Scotoecus hindei/albigula (DM), Scotophilus dinganii (DM), Scotophilus cf. viridis (DM) Save River, south 15 -21.17578 34.7465 Glauconycteris variegata (DM), Kerivoula of bridge on EN1 argentata (DM), Kerivoula lanosa (DM), Nycticeinops schlieffeni (DM), Scotophilus dinganii (DM), Scotophilus cf. viridis (DM) Save River 16 -21.34427 34.458217 Nycteris hispida (TM) Gerhard’s Cave 17 -21.66954 34.864403 Hipposideros caffer (TM), Miniopterus natalensis (TM), Nycteris thebaica (TM), Rhinolophus hildebrandtii (DM, TM), Rhinolophus cf. swinnyi (DM), Triaenops afer (DM, TM)

Central Mozambique Buzi River 18 -19.92783 33.82821 Epomophorus wahlbergi (DM), Mops condylurus (DM), Nycticeinops schlieffeni (DM), Triaenops afer (DM) Chimoio 19 -19.1164 33.4833 Neoromicia capensis, Nycteris thebaica Casa Msika 20 -19.04024 33.06507 Chaerephon pumilus (DM), Mops condylurus (DM) Gorongosa 21 -18.97847 34.17577 Hipposideros vittatus (DM), Neoromicia nana National Park (DM), Nycticeinops schlieffeni (DM), Scotoecus

144

hindei/albigula (DM), Scotophilus cf. viridis (DM) Chinizuia forest 22 -18.97724 35.05221 Epomophorus wahlbergi (DM), Myonycteris relicta (DM), Neoromicia capensis (DM), Neoromicia zuluensis (DM), Rhinolophus hildebrandtii (DM), Rhinolophus landeri (DM), Rhinolophus deckenii (DM), Scotophilus cf. viridis (DM) Cheringoma Caves 23 -18.56478 34.87204 Hipposideros caffer (DM), Hipposideros vittatus (DM), Miniopterus inflatus (DM), Nycteris macrotis (DM), Nycteris thebaica (DM), Rhinolophus hildebrandtii (DM), Rousettus aegyptiacus (DM), Triaenops afer (DM) Caia Lodge 24 -17.8471 35.32311 Nycteris thebaica (DM), Nycticeinops schlieffeni (DM)

Northern Mozambique Mount Inago 25 -15.13333 37.65 Epomophorus wahlbergi, Rhinolophus hildebrandtii (DM), Rhinolophus cf. swinnyi (DM) Mount Namuli, 26 -15.46221 37.01918 Epomophorus crypturus (MHNG), Epomophorus foothills wahlbergi (DM), Hipposideros ruber (DM), Lissonycteris goliath (DM), Miniopterus cf. fraterculus (DM), Myotis bocagii (DM), Myotis tricolor (DM), Neoromicia nana (DM), Pipistrellus hesperidus (DM), Rhinolophus blasii (DM), Rhinolophus clivosus (DM), Rousettus aegyptiacus (DM), Scotophilus dinganii (DM) Mount Namuli 27 -15.36925 37.061361 Eptesicus hottentotus (DM), Miniopterus cf. fraterculus (DM), Miniopterus inflatus (DM), Myotis tricolor (DM), Pipistrellus hesperidus (DM), Rhinolophus blasii, Rhinolophus clivosus (DM), Rhinolophus hildebrandtii (DM), Rhinolophus cf. maendeleo (DM), Rhinolophus cf. swinnyi (DM) Mount Mabu 28 -16.286477 36.403013 Epomophorus wahlbergi, Hipposideros ruber (DM), Laephotis botswanae (MHNG), Miniopterus cf. fraterculus (MHNG), Miniopterus natalensis (DM), Myotis tricolor (DM), Kerivoula cf. phalaena (MHNG), Rhinolophus blasii (DM), Rhinolophus clivosus (DM), Rhinolophus hildebrandtii (DM), Rhinolophus landeri (MHNG), Rousettus aegyptiacus (MM) Mount Chiperone 29 -16.50694 35.72583 Miniopterus inflatus (DM), Myotis tricolor (DM) Ribaue, 40km west 30 -14.97082 38.07951 Chaerephon ansorgei (DM), Epomophorus of crypturus (DM), Lissonycteris goliath (DM), Rousettus aegyptiacus (DM), Scotoecus hindei/albigula (DM), Scotophilus dinganii (DM), Tadarida aegyptiaca (DM), Tadarida

145

fulminans (DM) Namapa 31 -13.49489 39.78403 Epomophorus crypturus (DM), Epomophorus wahlbergi (DM), Rhinolophus hildebrandtii (DM), Triaenops afer (DM) Balama Coutada 32 -13.41615 38.03745 Epomophorus wahlbergi (DM), Triaenops afer (DM) Meponda 33 -13.40146 34.87048 Chaerephon pumilus (DM), Epomophorus crypturus (DM), Epomophorus labiatus (DM), Hipposideros ruber (DM), Myotis bocagii (DM), Neoromicia nana (DM), Nycteris hispida (DM), Meponda, 10km 34 -13.37021 34.93798 Rhinolophus fumigatus (DM) east of Meponda, 6km east 35 -13.36474 34.89817 Neoromicia zuluensis (DM) of Pemba 36 -13.00637 40.52368 Chaerephon pumilus (DM), Epomophorus crypturus (DM), Nycteris grandis (DM), Rhinolophus landeri (DM), Niassa Game 37 -12.86879 37.69196 Hipposideros caffer (DM), Neoromicia nana Reserve, south of (DM), Nycteris macrotis (DM), Nycticeinops schlieffeni (DM), Rhinolophus simulator (DM), Sauromys petrophilus (DM), Scotophilus cf. viridis (DM), Triaenops afer (DM) Niassa Game 38 -12.62413 37.65644 Neoromicia nana (DM), Neoromicia zuluensis Reserve, Kiboko, (DM), Scotophilus cf. viridis (DM), Triaenops 23km south of afer (DM) Niassa Game 39 -12.18234 37.55024 Epomophorus crypturus (DM), Hipposideros Reserve, Maputo caffer (DM), Neoromicia nana (DM), Camp Neoromicia zuluensis (DM), Nycticeinops schlieffeni (DM), Rhinolophus fumigatus (DM) Niassa Game 40 -12.16906 38.24283 Hipposideros caffer (DM), Neoromicia nana Reserve, Nkuli (DM), Nycticeinops schlieffeni (DM), Scotophilus Camp dinganii (DM), Scotophilus cf. viridis (DM), Triaenops afer (DM) Niassa Game 41 -12.1319 37.43574 Epomophorus crypturus (DM), Hipposideros Reserve, on caffer (DM) Matondovela Road

146

Appendix 2. Updated checklist and conservation status of 67 species of bats recorded from Mozambique

Not Species Confirmed1 Red List Reference confirmed2 Pteropodidae Eidolon helvum ● Least concern Monadjem et al, (2010) Epomophorus crypturus ● Least concern This study Epomophorus labiatus ● Least concern This study Epomophorus wahlbergi ● Least concern This study Lissonycteris goliath ● Vulnerable This study Myonycteris relicta ● Vulnerable This study Rousettus aegyptiacus ● Least concern This study Rhinolophidae Rhinolophus blasii ● Near threatened This study Rhinolophus clivosus ● Least concern This study Rhinolophus darlingi ● Least concern Monadjem et al, (2010) Rhinolophus deckenii ● Data deficient This study Rhinolophus fumigatus ● Least concern This study Rhinolophus hildebrandtii ● Least concern This study Rhinolophus landeri ● Least concern This study Rhinolophus cf. maendeleo ● Data deficient This study Rhinolophus simulator ● Least concern This study Rhinolophus cf. swinnyi ● Near threatened This study Hipposideridae Cloeotis percivali ● Vulnerable Smithers and Lobão Tello (1976) Hipposideros caffer ● Least concern This study Hipposideros ruber ● Least concern This study Hipposideros vitattus ● Least concern This study Triaenops afer ● Least concern This study Emballonuridae Coleura afra ● Least concern Van Cakenberghe & Seamark (2008) Taphozous mauritianus ● Least concern Monadjem et al, (2010) Taphozous perforatus ● Least concern Smithers and Lobão Tello (1976) Nycteridae Nycteris grandis ● Least concern This study Nycteris hispida ● Least concern This study Nycteris macrotis ● Least concern This study Nycteris thebaica ● Least concern This study Nycteris woodi ● Near threatened Monadjem et al, (2010) Molossidae Chaerephon ansorgei ● Least concern This study Chaerephon bivittatus ● Least concern Smithers and Lobão Tello (1976) Chaerephon pumilus ● Least concern This study Mops brachypterus ● Least concern Smithers and Lobão Tello (1976) Mops condylurus ● Least concern This study Mops niveiventer ● Least concern Smithers and Lobão Tello (1976) Sauromys petrophilus ● Least concern This study Tadarida aegyptiaca ● Least concern This study

147

Tadarida fulminans ● Least concern This study Tadarida ventralis ● Near threatened Smithers and Lobão Tello (1976) Vespertilionidae Eptesicus hottentotus ● Least concern This study Glauconycteris variegata ● Least concern This study Hypsugo anchietae ● Least concern Monadjem et al, (2010) Kerivoula argentata ● Least concern This study Kerivoula lanosa ● Least concern This study Kerivoula cf. phalaena ● Least concern This study Laephotis botswanae ● Least concern This study Mimetillus thomasi ● Least concern Monadjem et al, (2010) Myotis bocagii ● Least concern This study Myotis tricolor ● Least concern This study Myotis welwitschii ● Least concern Smithers and Lobão Tello (1976) Neoromicia nana ● Least concern This study Neoromicia capensis ● Least concern This study Neoromicia rendalli ● Least concern This study Neoromicia zuluensis ● Least concern This study Nycticeinops schlieffeni ● Least concern This study Pipistrellus hesperidus ● Least concern This study Pipistrellus rueppellii ● Least concern Van Cakenberghe & Seamark (2008) Scotoecus albofuscus ● Data deficient Monadjem et al, (2010) Scotoecus hindei/albigula ● Data deficient This study Scotophilus dinganii ● Least concern This study Scotophilus leucogaster ● Least concern This study Scotophilus nigrita ● Near threatened Van Cakenberghe & Seamark (2008) Scotophilus cf. viridis ● Least concern This study Miniopteridae Miniopterus inflatus ● Least concern This study Miniopterus cf. fraterculus ● Least concern This study Miniopterus natalensis ● Near threatened This study Species total 59 8 1Specimens of species were either examined by the authors or by Monadjem et al. (2010). 2These are typically old specimens possibly housed in European institutions and not located or examined by Monadjem et al. (2010), but mentioned in Smithers and Lobão Tello (1976) or Van Cakenberghe & Seamark (2008).

148 Biodivers Conserv (2010) 19:449–469 DOI 10.1007/s10531-008-9482-5

ORIGINAL PAPER

Bats of Borneo: diversity, distributions and representation in protected areas

Matthew J. Struebig Æ Lenny Christy Æ Dorothea Pio Æ Erik Meijaard

Received: 1 April 2008 / Accepted: 17 September 2008 / Published online: 3 October 2008 Ó Springer Science+Business Media B.V. 2008

Abstract Protected areas are valuable in conserving tropical biodiversity, but an insuf- ﬁcient understanding of species diversity and distributions makes it difﬁcult to evaluate their effectiveness. This is especially true on Borneo, a species rich island shared by three countries, and is particularly concerning for bats, a poorly known component of mammal diversity that may be highly susceptible to landscape changes. We reviewed the diversity, distributions and conservation status of 54 bat species to determine the representation of these taxa in Borneo’s protected areas, and whether these reserves complement each other in terms of bat diversity. Lower and upper bound estimates of bat species composition were characterised in 23 protected areas and the proposed boundaries of the Heart of Borneo conservation area. We used lower and upper bound estimates of species composition. By using actual inventories, species representation was highly irregular, and even if some reserves were included in the Heart of Borneo, the protected area network would still exhibit low complementarity. By inferring species presence from distributions, composition between most reserves was similar, and complementarity was much higher. Predicting species richness using abundance information suggested that bat species representation in reserves may lie between these two extremes. We recommend that researchers better sample biodiversity over the island and address the conservation threats faced in Borneo both within and outside protected areas. While the Heart of Borneo Initiative is commendable, it should not divert attention from other conservation areas.

Keywords Heart of Borneo Á Chiroptera Á Gap analysis Á Species richness prediction Á Southeast Asia Á Forest Á Tropical conservation Á Indonesia Á Malaysia Á Brunei

M. J. Struebig (&) School of Biological and Chemical Sciences, Queen Mary University of London, London E1 4NS, UK e-mail: [email protected]

L. Christy Á E. Meijaard The Nature Conservancy, East Kalimantan Program, Jalan Polantas no. 5, Markoni, Balikpapan, East Kalimantan, Indonesia

D. Pio Department of Ecology and Evolution, Biophore, University of Lausanne, 1015 Lausanne, Switzerland 123 450 Biodivers Conserv (2010) 19:449–469

Introduction

The survival of tropical biodiversity is threatened in most regions, so its conservation requires that it receives adequate protection and that habitats are sustainably managed (du- Toit et al. 2004). The majority of protected areas have proved valuable in this regard (Bruner et al. 2001). However, many protected areas were designed for purposes other than conservation, and hence diversity, particularly of low-profile species, is under-represented (Rodrigues et al. 2004). Identifying the species that are represented in reserves and those that are missing is therefore crucial for conservation planning, and the distributions of species and spatial patterns of species richness have become widely used in evaluations of conservation effectiveness (Jennings 2000). These ‘gap’ analyses range from overlays of species distributions and protected area boundaries to generate inventories, to sophisticated assessments using niche modelling. In addition to identifying gaps in protected area networks these analyses can also determine the extent to which reserves complement each other in terms of biodiversity protection, and whether unique diversity remains outside the system. Unfortunately, analyses are often restricted to well-known taxa because species distribution information for most taxa is limited or unavailable. Borneo is well known for its high species richness (Mackinnon et al. 1996; Sodhi et al. 2004), but high deforestation rates are of serious conservation concern and place enormous pressures on the island’s protected area system (Curran et al. 2004; Fuller et al. 2004). Nine percent of Borneo is within protected areas of IUCN categories I–IV (IUCN 1994), administered by Indonesia (in Kalimantan, four provinces), Malaysia (in Sabah and Sara- wak) and Brunei Darussalam. However, these reserves have become increasingly isolated (Curran et al. 2004), leading the three countries to recently commit to the Heart of Borneo Initiative (HoB), which seeks to connect 22 parks and reserves into a continuous 24 million ha transboundary conservation area (WWF 2005). While this initiative is of clear conservation merit, there is little biodiversity information available for much of the area, and no formal assessment has yet determined whether island-wide levels of faunal diversity are adequately represented—particularly for lesser known, low-profile animal groups. Bats exhibit many of the required features of indicators for biodiversity appraisals: they are diverse (at least 93 species in Borneo, Simmons 2005; ca. 40% of the island’s land mammals, Payne et al. 2000); perform ecosystem services (Fujita and Tuttle 1991; Hutson et al. 2001); and comprise assemblages with clear and predictable guild/ensemble structure (Kingston et al. 2003; McKenzie et al. 1995). Many of these animals also depend on forests to some extent and are sensitive to disturbance and landscape changes (Danielsen et al. 2008; Struebig et al. 2008; Zubaid 1993); as a result losses of bat species in Southeast Asia are predicted to exceed 40% by 2100 (Lane et al. 2006). Most of Borneo’s old world fruit bats (Pteropodidae) exhibit dispersal capabilities and generalist feeding habits that support their persistence in disturbed landscapes (Meijaard et al. 2005). Likewise, insectivorous species that forage in forest edges and tree fall gaps (e.g. Vespertilionid sub- families Miniopterinae and Vespertilioninae, Emballonura spp.) or the open spaces above or outside forests (e.g. Molossidae, Taphozous spp.), may be readily adapted to exploit disturbed habitats (Kingston et al. 2003; Meijaard et al. 2005). However, insectivorous species that forage in the forest interior (Hipposideridae, Megadermatidae, Nycteridae, Rhinolophidae, vespertilionid sub-families Kerivoulinae and Murininae) are likely to be particularly sensitive to landscape change because of ecomorphological traits that restrict their use of more open habitats (reviewed in Kingston et al. 2003). Knowledge of bat distributions in Borneo has been historically biased towards easily captured species, with many appearing confined to the north of the island where most 123 Biodivers Conserv (2010) 19:449–469 451 research and specimen collection has been undertaken (reviewed in Payne et al. 2000; Suyanto and Struebig 2007). However, research effort has increased in recent years, and improved communication amongst local and regional experts has led to distributions being substantially updated via the Southeast Asian Mammal Databank (SAMD) (Boitani et al. 2006). As part of the IUCN Global Mammal Assessment, the SAMD team worked with a network of institutions and taxonomic experts to compile and disseminate all information on the distribution, basic ecology and conservation status of Southeast Asian mammals. The main output is an impressive database on mammal diversity that includes distribution maps of the extent of occurrence of each species in a freely available GIS format. How- ever, the database is still limited by a paucity of information for a substantial number of taxa, and will quickly become outdated as more empirical mammal research is undertaken in the region. This is particularly evident for bats, of which much research undertaken in recent years was not considered by the SAMD (e.g. in Borneo, Anwarali et al. 2007; Jayaraj et al. 2006; Struebig et al. 2006a, b; Suyanto and Struebig 2007). We sought to determine the known and predicted representation of bat species in Borneo’s protected areas (including the proposed Heart of Borneo conservation area) using reviewed and updated species distributions. We used lower and upper bound estimates of species composition derived from simple gap analyses to explore the extent to which each reserve protected a unique bat fauna, and whether this would be represented in the Heart of Borneo. We then tested the quality of these estimates and their value in describing true species diversity by predicting the likely species richness at a site based on existing assemblage data. Finally, we identified conservation issues relevant to bats inside and outside protected areas on Borneo to guide future conservation research efforts.

Methods

Data

Analyses were based on a database of bat locality records from all seven administrative units of Borneo, compiled after consultation with international and regional bat specialists. We included records from published and unpublished sources, as well as specimens collected since 1995 in the zoological collections of Bogor, Indonesia; Bandar Seri Begawan, Brunei Darussalam; and Kuching, Malaysia. For species that are notoriously hard to identify, or for which taxonomy is not well defined, we consulted the authors/collectors for additional information to confirm their identity. Because of a general paucity of mammal records in Indonesian territory (Meijaard and Nijman 2003), additional data were obtained from surveys of 11 sites undertaken between 2002 and 2008 in Central and East Kali- mantan (Struebig et al. 2006a, b; Suyanto and Struebig 2007; D. Pio et al., Unpublished data). We considered the only conspicuous species likely to be correctly identified by non- specialists to be the large flying-foxes, Pteropus vampyrus and P. hypomelanus. Hence, we included opportunistic records provided by local people and conservation practitioners for these two species, but only when information was given by more than one source in an area (89 combined records). To our knowledge, and to the agreement of the specialists we consulted, our database represented the most up-to-date knowledge of bat distribution records in Borneo. However, because the majority of records arose from captures in mist- nets or harp-traps set at ground level, some species records remained particularly infre- quent. Thus, subsequent analyses focused on a subset of species that comprised the

123 452 Biodivers Conserv (2010) 19:449–469 majority of records. These included the 17 Bornean fruit bats within the Pteropodidae, and 37 forest interior insectivorous species (sensu Kingston et al. 2003). These species are readily captured in mist-nets (Pteropodidae) or harp-traps (bats of the forest interior) used for most bat diversity surveys on the island. Eleven of these species are currently IUCN red listed (Boitani et al. 2006).

Bat species composition in protected areas

Bat species composition was determined in 23 protected areas distributed throughout the island (Fig. 1), and also within the proposed boundaries of the Heart of Borneo (According to WWF in 2007). We did not consider all protected areas on Borneo since the boundaries of many reserves have changed substantially over time, and not all are recognised by IUCN/WCMC (http://www.wdpa.org/). The reserves we selected comprised a substantial proportion of those in the existing Borneo network, and included the largest parks both

Fig. 1 Locality records of bats on Borneo overlaid with the current protected area network (dark shading), and the proposed Heart of Borneo boundaries (broken line, according to WWF in 2007). Bubble size is proportional to the number of bat species reported at the site, and numbers refer to the protected areas used in analyses and listed in Table 2. Lines indicate state/province boundaries 123 Biodivers Conserv (2010) 19:449–469 453 within and outside of the Heart of Borneo boundaries. Three estimates of species composition within each area were derived, which gave lower and upper bounds of species richness. The first estimate was derived from the actual species records within each area and generated a lower bound of species composition from confirmed inventories. For this lower bound estimate all bat-locality records were assigned a latitude and longitude co-ordinate, entered into Arcview GIS (ESRI 1982–1998), and overlaid with a thematic layer of protected areas (based on WCMC data, but updated with local government data if applicable) to derive species lists. The second estimate determined an upper bound to species composition in an area by identifying species potentially present using a simple gap analysis. We overlaid the species distribution polygons provided by the SAMD with protected area boundaries; all species that had a distribution overlapping with a boundary, regardless of the extent, were included in an inventory. The third estimate used a similar gap analysis, but this time based on species distributions updated or corrected by our bat locality record database. We overlaid our bat locality records with the SAMD species distributions and digitised novel distribution polygons that encompassed new records if appropriate. We also corrected the distributions of species that predominantly roost in caves to exclude them from the coastal wetlands on the South coast. This represents an area of peat swamps and mangroves between the Barito and Kumai Rivers (see Langner et al. 2007 for delineation of these forest types) in which there are no known caves, and in which intensive surveys have not captured cave-roosting bat species (Struebig et al. 2006b). We used the Bray–Curtis dissimilarity index to determine overall differences in bat species composition in protected areas, and hence assess the extent to which each reserve protected a unique subset of the bat species pool. Dissimilarities between reserve inventories based on each of the three estimates were represented using Non-Metric Multidimensional Scaling (NMDS) in PC-ORD (McCune and Meffor 1999). Ordinations were conducted on Bray–Curtis dissimilarity coefficients derived from presence–absence data at each protected area using 500 iterations and 250 runs of both real and randomised data. Two dimensions were chosen based on the reduction of stress from additional axes. We included the Heart of Borneo inventory for inspection of dissimilarity in bat species composition between this area and reserves outside of the proposed boundaries.

Species richness predictions from abundance data

We checked the consistency of our gap-derived estimates of bat species richness by further predicting richness using abundance data from seven standardised harp-trap inventories that provided this information. Given that harp-traps best target forest interior insectivorous species, we restricted our analyses to this bat ensemble (maximum 37 species). We used two multinomial prediction models introduced by Solow and Polasky (1999) and Shen et al. (2003) to predict species richness after an additional sampling effort. Although other methods for predicting species richness exist, they can become unstable or do not behave as expected at high predicted sample sizes (Shen et al. 2003). Furthermore, predictors and estimators have already been tested in Malaysian harp-trap inventories by Kingston (2009), and all but two led to substantial overestimation of bat species richness. Hence we used the two reliable predictors to create upper and lower bounds of species richness. We used a total predicted sample effort of 1,000 individuals because most species in this ensemble were represented by this effort in Malaysia (Kingston 2009). 123 454 Biodivers Conserv (2010) 19:449–469

Results

We amassed 1,811 independent bat-locality records of 96 species representing all families in Borneo, and at least 634 of these were additional to those available to the SAMD. Records per species ranged from only one for Hypsugo impricatus (previously Pipistrellus impricatus) to 89 records for Cynopterus brachyotis (Table 1). The majority of records represented the 17 frugivorous species within the Pteropodidae (576 records or 31.8% of total), while a further 774 records (42.7%) represented the 37 forest interior insectivorous species. The remainder (452 records, 25.5%) were of patchily recorded and/or poorly known species of the families Vespertilionidae, Emballonuridae or Molossidae (42 species). With 662 bat-locality records, Sabah remained the most studied part of Borneo, but by accounting for area the overall coverage in Brunei was more complete (0.87 records per 100 km2 in Sabah vs. 2.17 for Brunei; Table 1; Fig. 1). Geographic coverage in Kali- mantan remained poor compared to Sabah, Sarawak and Brunei: the least studied state was South Kalimantan with only 0.05 records per 100 km2, followed by West Kalimantan with 0.06 records per 100 km2.

Bat species composition in protected areas

Our three estimates of bat species composition yielded strikingly different reserve inventories (Table 2). Lower bound estimates derived from actual survey data in each reserve produced largely incomplete inventories compared to the bat species pool. The most species-rich protected area (according to the subset of 54 species) was Gunung Kinabalu in Sabah with 32 recorded species (59% of total). Danum Valley in Sabah and Sangkulirang in East Kalimantan also had relatively high inventories (29 species; 54% of total), followed by Ulu Temburong in Brunei (27 species, 50% of total). Conversely we could ﬁnd only one species record (\2% total) for two of the reserves, Gunung Gading and Loagan Bunut in Sarawak, and the lists for Bukit Baka-Bukit Raya in West/Central Ka- limantan and Kutai in East Kalimantan were limited to two and three species respectively (\6% total). Species poor inventories were often characterised by depauperate lists of forest interior insectivorous bats, particularly in Kalimantan and Sarawak where the majority of inventories listed\30% of the species pool. The high contrast between species- poor and species-rich inventories resulted in large differences between the protected areas in their representation of bat diversity (Fig. 2a). Wide scatter of protected areas in the NMDS plot highlighted large differences between the reserves, and also between the Heart of Borneo and reserves far outside the boundaries (e.g. No. 15 in Fig. 1; Table 2: Gunung Palung). By this evaluation the Heart of Borneo would support 81% of the bat species we considered (44 species), but complementarity between this area and reserves outside the boundaries would be low because these reserves represented depauperate faunas. Upper bound estimates from SAMD-derived species distributions resulted in larger bat inventories in all protected areas (Table 2). According to this estimate, Kayan-Mentarang and Maliau Basin were predicted to contain almost 75% of the bat species pool (40 species each; Table 2), though the species composition of each reserve was markedly different (Fig. 2b). Much of the variation in species composition among reserves was accounted for by these protected areas and others in Sabah and Sarawak included within the Heart of Borneo boundaries, suggesting that this conservation area would protect substantial bat diversity. Indeed the Heart of Borneo would protect almost 95% of the pool of bat species (51 species; Table 2) according to this estimate. However, large distances on the NMDS plot between reserves within the Heart of Borneo and those in the south of the island 123 idvr osr 21)1:4–6 455 19:449–469 (2010) Conserv Biodivers Table 1 Characteristics of the 54 bat species used in analyses and the number of independent locality records for these taxa in each state/province of Borneo Family/species (previous Foraging Red list Rangec Roost No. independent locality records species name) strategya statusb ecologyd P Sabah Saraw Brunei W.Kal C.Kal S.Kal E.Kal

Pteropodidae Aethalops aequalis Bf 1 f? 15 9 2 1 27 (alecto) Balionycteris maculata Bf NT 3 f 16 18 6 2 8 8 58 Chironax melanocephalus Bf 4 f, c 2 2 1 1 3 3 12 Cynopterus brachyotis Bf 6 f, c 36 23 3 5 10 12 89 C. horsﬁeldii Bf 4/5 c, f 10 12 1 1 5 29 C. minutus Bf 3 f 5 2 7 Dyacopterus spadiceus Cf NT 3 c, f? 4 10 2 2 1 3 22 Eonycteris major Cf 1 C 8 10 2 1 1 22 E. spelaea Cf 6 C 13 13 2 1 1 4 34 Macroglossus minimus Bf 6 f 33 14 6 5 4 1 6 69 Megaerops ecaudatus Bf 3 f 10 10 4 2 3 3 32 M. wetmorei Bf VU 2 f 4 2 1 7 Penthetor lucasi Bf/Cf 3 C, f 13 19 5 4 4 7 52 Pteropus hypomelanus Cf 5 F 5 3 2 10 P. vampyrus Cf VU 5 F 16 13 7 2 4 38 80 Rousettus amplexicaudatus Cf 6 C 8 5 2 4 1 20 R. spinalatus Cf VU 2 C 2 3 1 6 Hipposideridae Coelops robinsoni Ni 3 f, c 1 3 1 5

123 Hipposideros ater Ni 6 c 8 5 1 3 3 20 H. bicolor Ni 6 C 3 5 1 2 3 14 5 idvr osr 21)19:449–469 (2010) Conserv Biodivers 456 123 Table 1 continued Family/species (previous Foraging Red list Rangec Roost No. independent locality records species name) strategya statusb ecologyd P Sabah Saraw Brunei W.Kal C.Kal S.Kal E.Kal

H. cervinus Ni/Ei 6 C 17 13 1 3 13 11 58 H. cineraceus Ni 6 C 7 2 2 1 5 17 H. coxi Ni DD 1 c 4 3 7 H. diadema Ni/Ei 6 C, f 14 6 5 2 8 1 5 41 H. doriae (sabanus)NiNT3f?311 2 18 H. dyacorum Ni 2 C, f 8 7 1 1 1 5 23 H. galeritus Ni/Ei 6 c 11 7 1 2 2 1 7 32 H. larvatus Ni/Ei 6 C 5 4 2 8 19 H. ridleyi Ni VU 2 f 4 2 2 1 1 10 Megadermatidae Megaderma spasma Ni 6 f, c 16 5 1 2 2 1 5 32 Nycteridae Nycteris tragata Ni 4/5 f, c 6 10 1 5 1 3 26 (javanica) Rhinolophidae Rhinolophus acuminatus Ni 5 ? 13 1 1 1 16 R. afﬁnis Ni/Ei 5 C 1 5 2 3 1 7 19 R. arcuatus Ni 3 C? 2 2 1 5 R. borneensis Ni 5 C 18 11 1 5 10 1 13 59 R. creaghi Ni 4 C 11 1 1 9 22 R. luctus Ni 6 f, c 9 8 3 7 1 1 29 R. philippinensis Ni 6 C 4 7 1 2 1 15 R. pusillus Ni 3 C 2 6 8 idvr osr 21)1:4–6 457 19:449–469 (2010) Conserv Biodivers Table 1 continued Family/species (previous Foraging Red list Rangec Roost No. independent locality records species name) strategya statusb ecologyd P Sabah Saraw Brunei W.Kal C.Kal S.Kal E.Kal

R. sedulus Ni NT 2 f 11 5 4 2 13 2 37 R. trifoliatus Ni 6 f 16 4 5 2 10 7 44 Vespertilionidae (Kerivoulininae) Kerivoula hardwickii Ni 6 f 7 3 1 1 5 4 22 K. intermedia Ni NT 2 f 10 5 1 4 3 23 K. minuta Ni NT 2 f 12 2 1 2 8 25 K. papillosae Ni 6 f 12 10 6 7 5 40 K. pellucida Ni 3/4 f 8 3 2 1 4 1 19 K. whiteheadii Ni 3 f 1 2 3 Phoniscus atrox Ni NT 3 f 5 3 1 9 P. jagorii Ni 4 f 1 2 3 Vespertilionidae (Murininae) Harpiocephalus harpia Ni VU 6 f 3 3 (mordax) Murina aenea Ni VU 2 f 4 3 7 Murina cyclotis Ni 6 f 6 2 3 7 18 Murina rozendaali Ni VU 2 f 7 1 1 1 10 Murina suilla Ni 3 f 10 3 8 5 26 Total records 459 312 89 63 182 12 231 1,348 Other bat records 203 77 36 26 69 5 47 463 123 5 idvr osr 21)19:449–469 (2010) Conserv Biodivers 458 123 Table 1 continued Family/species (previous Foraging Red Rangec Roost No. independent locality records species name) strategya list ecologyd P statusb Sabah Saraw Brunei W.Kal C.Kal S.Kal E.Kal

Coverage (records per 100 km2) 0.87 0.31 2.17 0.13 0.16 0.05 0.06 0.23

Taxonomy follows Simmons (2005) a Foraging strategies assigned based on wing morphology (Kingston et al. 2003; M. Struebig, Unpublished data) following those described in McKenzie et al. (1995): Ni, insectivorous species that forage in narrow-spaces or clutter (Strategy 1); Ei, insectivorous species that forage in partially cluttered edges and/or canopy edges (Strategy 2); Bf, frugivorous or nectarivorous species that forage in clutter below forest canopy (Strategy 4); Cf, frugivorous or nectarivorous species that forage in open areas and over large distances (Strategy 5) b According to the SAMD (2006): DD data deficient; NT near threatened; VU vulnerable c Species distribution according to the SAMD (2006) and categorised as (1) endemic to Borneo; (2) restricted to Borneo and one neighbouring land mass (e.g. Peninsular Malaysia, Sumatra or Philippines); (3) restricted to Borneo and two neighbouring land masses; (4) patchy distribution in three or more land masses; (5) large and covers majority or all of Southeast Asia; (6) large and extends beyond Southeast Asia d Known roost types classified as f, tree cavities or foliage; and c, caves or other large permanent structures such as buildings and mines. Capitalisation indicates colony size: lower case, typically 1–100 individuals; upper case, [100 individuals. ? indicates that limited roost information is available e The taxonomy of Kerivoula papillosa is being revised (K. Helgen, Personal Communication), and some individuals currently assigned to this species in Borneo may represent K. lenis, K. flora, or an undescribed cryptic species. Thus, for the purposes of consistency we assigned all records of these taxa to K. papillosa idvr osr 21)1:4–6 459 19:449–469 (2010) Conserv Biodivers Table 2 Bat species richness in protected areas in Borneo according to the observed species richness from inventories, and two upper bound estimates from gap analyses No.a Protected area Habitatsb Approx. Species richness, Sc area (km2) Pteropodidae Forest interior Rhinolophidae Hipposideridae Kerivoulinae/ (max. 17 spp.) (max. 37 spp.) (max. 10 spp.) (max. 12 spp.) Murinae (max. 13 spp.)

Obs SAMD New Obs SAMD New Obs SAMD New Obs SAMD New Obs SAMD New

1 Gunung Gading NP, Sarawak H 41 1 11 14 0 22 25 0 5 7 0 10 9 0 5 7 2 Kubah NP, Sarawak L, H 22 7 12 13 11 22 26 4 5 7 4 10 10 1 5 7 3 Bako NP, Sarawak L, M 27 6 11 13 17 22 26 4 5 7 8 10 10 3 5 7 4 Batang Ai NP, Sarawak H, U 1,688 2 12 14 0 21 23 0 6 6 0 8 8 0 5 7 5 Niah NP, Sarawak K 31 7 12 14 8 20 23 2 5 6 3 8 8 1 5 7 6 Loagan Bunut NP, Sarawak S, L 107 1 13 15 0 21 24 0 5 6 0 8 8 0 6 8 7 Gunung Mulu NP, Sarawak K 528 8 15 16 13 23 24 3 6 6 6 8 8 3 7 8 8 Ulu Temburong NP, Brunei L, H 489 11 15 16 16 20 24 3 5 6 7 5 8 5 8 8 9 Crocker Range NP, Sabah H 1,399 8 13 14 7 23 27 3 5 5 2 9 9 1 7 11 10 Gunung Kinabalu NP, Sabah L, H, U 754 11 12 14 21 25 25 5 6 5 7 8 9 7 9 9 11 Malinau Basin CF, Sabah H, L 390 4 13 15 7 27 29 4 8 8 2 8 8 1 9 11 12 Danum Valley CF, Sabah L, H 427 7 11 14 22 28 29 5 7 7 6 9 9 10 10 11 13 Tabin WR, Sabah L 1,120 5 10 13 15 26 27 4 7 7 5 9 9 5 8 9 14 Gunung Nyiut NP, West Kal. H, U 1,800 6 11 14 4 22 25 2 5 7 2 10 10 0 5 6 15 Gunung Palung NP, West Kal. M, S, L, H 900 3 11 13 6 20 22 3 5 5 3 9 9 0 4 6 16 Tanjung Puting NP, Cent. Kal. M, S, L 3,550 6 12 11 17 20 19 3 6 4 2 10 3 10 4 10 17 Sebangau NP, Cent. Kal. M, S 5,678 4 12 11 7 21 18 2 5 3 0 9 5 5 5 8 18 Bukit Baka-Raya NP, West/Cent. H, U, L 1,810 1 10 12 1 20 24 1 5 5 0 8 9 0 5 8

123 Kal. 6 idvr osr 21)19:449–469 (2010) Conserv Biodivers 460 123 Table 2 continued No.a Protected area Habitatsb Approx. Species richness, Sc area (km2) Pteropodidae Forest interior Rhinolophidae Hipposideridae Kerivoulinae/ (max. 17 spp.) (max. 37 spp.) (max. 10 spp.) (max. 12 spp.) Murinae (max. 13 spp.)

Obs SAMD New Obs SAMD New Obs SAMD New Obs SAMD New Obs SAMD New

19 Betung Kerihun NP, West Kal. H, U, L 8,000 5 12 15 6 21 25 3 5 6 3 9 10 0 5 7 20 Kayan Mentarang NP, East Kal. H, U, L 16,000 7 15 16 1 25 28 1 6 8 0 10 9 0 7 9 21 Bukit Soeharto WR, East Kal. L, H 618 4 11 14 4 20 25 1 5 7 0 9 9 3 4 7 22 Kutai NP, East Kal. L, H 1,986 2 12 15 1 20 25 1 7 7 0 8 9 0 3 7 23 Sangkulirang [proposed], East Kal. K 2,000 7 11 14 22 22 27 5 6 8 8 9 9 7 5 8 24 HEART OF BORNEO H, L, U, K 240,000 13 17 17 31 34 34 7 9 9 10 11 11 12 12 12 a Protected status: NP, national park; WR, wildlife reserve (Cagar Alam); CF, conservation forest b Habitat types: M, mangrove; S, peat or riverine swamp forest; L, lowland Dipterocarp or heath forest; H, hill dipterocarp; U, montane forest; K, karst c Three estimates of species richness: Obs., a lower bound from species-locality records in a reserve; SAMD, an upper bound derived from gap analysis using species distributions from the Southeast Asian Mammal Databank (SAMD); New, an upper bound derived from gap analysis using revised species distributions based on new species- locality records. Each estimate is partitioned for the main families/sub-families with forest interior insectivorous bats (Hipposideridae, Megadermatidae, Nycteridae, Rhinolophidae and vespertilionids of the sub-families Kerivoulinae and Murininae) shown separately. The total species richness can be derived by summing Pteropodidae and forest interior bats (max. = 54 species) Biodivers Conserv (2010) 19:449–469 461

Fig. 2 Bat assemblage composition in protected areas in Borneo according to lower and upper bound inventories: a lower bound using observed data from biodiversity studies [stress = 0.23]; b upper bound prediction using overlays of species distribution polygons from the SAMD [stress = 0.13]; c upper bound prediction using distribution polygons revised as part of this study [stress = 0.09]. Each point is a non- metric multidimensional scaling (NMDS) representation of bat species composition (maximum 54 species) in a protected area, and distances between points reflect dissimilarity using the Bray–Curtis coefficient. Stress is a measure of the poorness-of-fit of the NMDS, with values \0.20 considered to indicate more reliable ordinations (McCune and Meffor 1999). Numbers on the plots refer to the protected areas in Table 2 (not all labelled for clarity), and bubble size is proportional to species richness. Black bubbles indicate protected areas that are within the current Heart of Borneo boundaries suggested that species composition was still markedly different within and outside the Heart of Borneo boundaries (Fig. 2b). After reviewing species-locality records we updated the SAMD data for 29 bat species (54% of total), and the majority of these changes were to extend distributions southwards following new records from Kalimantan. The subsequent upper bound estimates resulted in bat inventories that were higher than those from both other estimates in all but two of the protected areas—Tanjung Puting and Sebangau in Central Kalimantan (Table 2). The two distribution-based estimates identified the same reserves as the most species rich, but other areas far outside of the Heart of Borneo boundaries were shown to also exhibit high bat species richness when the updated distributions were used (e.g. Sangkulirang, East Kali- mantan, 41 species; Gunung Nyiut, West Kalimantan, 39 species). Bat species representation in the Heart of Borneo remained high using this estimate (51 species, 95%). Furthermore, the scatter of points on the NMDS plot was greatly reduced when using inventories based on this estimate (Fig. 2c) suggesting that protected areas complemented each other in bat species they contained. Most reserves, both inside and outside of the Heart of Borneo were clustered together on the plot indicating strong similarities in bat species composition. However, the bat species composition of the two parks with depauperate bat faunas, was considerably different to that of other sites (Fig. 2c).

Species richness predictions from abundance data

Predicting bat species richness using species abundance data revealed that all but two harp- trap inventories (Tanjung Puting in Kalimantan, and Bau in Sarawak) were incomplete, and that the number of trappable species represented in the predicted inventories exhibited much variation between sites (Table 3). Bat inventories in the karst forests around Bau and the oligotrophic forests of Tanjung Puting remained depauperate subsets of the larger faunas of other sites. Barito Ulu, at the geographic centre of Borneo, supported the most 123 6 idvr osr 21)19:449–469 (2010) Conserv Biodivers 462 123

Table 3 Observed and predicted number of insectivorous bat species susceptible to capture by four-bank harp-trap at seven sites in Borneo, based on two multinomial species richness prediction models used on abundance data n Species richness, S

Obs. Solow and Polasky (1999) predictor Shen et al. (2003) predictor

Predicted Sa Inventory Trappable species Predicted Sa Inventory Trappable species completeness (%, max. = 37)c completeness (%)b (%, max. = 37)c (%)b

Bau, Sarawak (Mohd-Azlan et al. 2005) 83 13 14 93 35–38 14 93 35–38 Danum Valley, Sabah (T. Kingston, 147 17 23 74 46–62 19 90 46–51 Unpublished data) Sepilok, Sabah (Francis 1990) 442 17 24 71 46–65 23 74 46–62 Barito Ulu, Cent. Kalimantan (D. Pio, Unpublished data) 636 26 31 84 70–84 30 87 70–81 Tanjung Puting, Cent. Kalimantan 575 17 18 94 46–49 18 94 46–49 (Struebig et al. 2006b) Lampanut, Cent. Kalimantan 219 14 22 64 38–59 21 67 38–57 (I. Maryanto and M. Struebig, Unpublished data) Lesan, East Kalimantan (M. Struebig, Unpublished data) 310 18 25 72 49–68 24 75 46–65

a Species richness was predicted to a total sample size of 1,000 individuals with a cut-off for rare species set at 10 individuals. Decimal fractions were rounded b Proportion of predicted species richness present in that observed c Proportion of all insectivorous species susceptible to capture by harp-trap (max. = 37 species) that are represented in the observed and predicted inventories Biodivers Conserv (2010) 19:449–469 463 forest-interior bat species with up to 85% of the maximum fauna represented in the predicted inventory. Forests of Lesan in East Kalimantan (68% complete), Sepilok (65%) and Danum Valley (62%) in Sabah also potentially supported a substantial number of species. Predicted species richness varied considerably between adjacent sites, and also at the same site when compared to the distribution-based estimates. While some neighbouring sites were predicted to have quite similar species richness (Danum Valley, 19–23 species; Sepilok, 23–24; 90 km apart in Sabah), assemblages at other adjacent sites were predicted to be quite different (Barito Ulu, 30–31 species; Lampanut, 21–22 species; 95 km apart in Central Kalimantan). Species richness derived from prediction models was also similar to that estimated from revised species distribution estimates at Tanjung Puting (18 vs. 19 species), but was much lower than the estimates for Danum Valley (19–23 vs. 29 species).

Discussion

Representation of bat species in Borneo’s protected areas

Our conclusions of how bat species are represented in protected areas depended greatly on which species composition data we used in the assessment. By using actual species lists, representation was highly irregular over the Bornean protected area network, and even if some reserves were included in the Heart of Borneo, the protected area network would still exhibit low complementarity. By inferring species composition from distribution ranges, the representation of bat species in reserves was significantly improved, but differences between reserves inside and outside of the Heart of Borneo were still apparent. Finally, when species distributions were revised to include up-to-date records, the reserve network achieved high complementarity in representing bat species richness, and species composition between reserves inside and outside the Heart of Borneo was much more similar. Species inventories are often used by conservation practitioners to identify priorities (e.g. Meijaard and Nijman 2003; Sujatnika et al. 1995) or to promote the value of an area for protection (e.g. WWF 2005), yet the efficacy of these assessments depends critically on the quality of the inventory data available. By focusing on bats we demonstrate that using incomplete inventories is likely to grossly underestimate the faunal composition of most protected areas in Borneo. Inferring composition from species distributions suggests that species representation is more uniform than is currently considered, at least at the macro- scale of this analysis. The quality of the distribution data is important, and conservation biologists and practitioners should continue to work together to improve our knowledge of this. Our simple gap analyses should be equally as valuable in assessing potential species representation in protected areas of other large islands, which in Southeast Asia include Java, Sumatra, Sulawesi and Papua. Despite our positive forecast of the efficacy of the reserve system and the Heart of Borneo in protecting biodiversity, there are important limitations to our approach that should be considered. Firstly, our assessment was restricted to a subset of bat species for which distribution information is available, and biogeographic patterns of these taxa may not reflect those of other, poorly known, bat species, or indeed other vertebrates. Some bat species excluded from our analyses are considered endemic to a few localities in Borneo (e.g. Myotis gomantongensis), which gives the protection of these sites clear conservation priority. However, since the species we excluded are likely to forage in edges and open areas and 123 464 Biodivers Conserv (2010) 19:449–469 have considerable dispersal capabilities (Meijaard et al. 2005), they are more likely to persist in disturbed habitats and exhibit wider ranges than currently recognised. Because these species are notoriously difficult to capture (Kingston et al. 2003), this paucity of records in Borneo most likely reflects poor sampling effort rather than actual rarity or endemism. Secondly, our distribution based estimates should be considered as upper bounds to species composition in an area, and in reality some species are likely to be absent from inventories because habitats may not meet their ecological needs. Our abundance-based species richness predictions exhibited much variation between sites, reflecting landscape and region-wide heterogeneity of assemblage composition. While we could confidently account for the lack of cave-roosting species in oligotrophic forests of Southern Borneo, our overall knowledge of habitats over the island is quite limited, and as a result distributions may overestimate species composition in some areas. A simple gap analysis overlooks variation in the availability and quality of habitats in the protected areas assessed, and it is likely that much of the habitat throughout Borneo is unsuitable for all species. Many forest-interior insectivorous species are likely to be restricted to forest, and some of these species (e.g. Murina aenea, M. rozendaali, Phoniscus jagorii), have mostly been recorded in undisturbed habitats. Predictive modelling of species distributions has the potential to improve the accuracy of our estimates (Rodrı´guez et al. 2007), but our bat database for Borneo is still quite limited; sampling bias, unstandardised sampling effort, and low geographic and environmental coverage, all limit the use of biodiversity databases for detailed analyses (Hortal et al. 2007). Therefore, although our findings imply high complementarity within Borneo’s protected area system, we suggest that the Heart of Borneo should not divert attention away from other areas of conservation importance until a better understanding of biodiversity value can be used to assess all areas.

Major threats to bats in Borneo

Of the five main categories of threat to island bat populations identified by Wiles and Brooke (2008), and the four prioritised for Southeast Asia by Kingston (2008), three are major concerns in Borneo: habitat alteration and loss, cave disturbance, and hunting. A large number of this island’s bat species depend on forests for food and shelter and deforestation can therefore have major impacts by reducing foraging habitats and roosting sites. Deforestation rates in Indonesia and Malaysia continue to be high (Langner et al. 2007), but forest loss in Brunei has slowed since 1990 (Koh 2007). Uncontrolled forest fires have destroyed vast areas of Kalimantan, and millions of hectares continue to be lost by illegal logging and land clearance (Fuller et al. 2004). Of most concern is that many reserves, in Kalimantan at least, are not fulfilling their protective roles. Illegal logging has been rampant in national parks such as Bukit Baka-Bukit Raya, Tanjung Puting and Gunung Palung (Curran et al. 2004; Jepson et al. 2001). Over 50% of reserves in East Kalimantan have already been significantly degraded by logging, encroachment and fire, and at least five, including Kutai National Park and Bukit Soeharto Reserve, have almost completely been destroyed (Jepson et al. 2002). A large proportion of deforested areas are converted to plantations of timber or commodities such as rubber, cocoa and palm oil. Although the bat diversity value of these plantations is far from clear, preliminary studies across vertebrate groups (including bats) suggest that species diversity is lower than in natural forests, and that biodiversity value depends on plantation type, with rubber and cocoa plantations sometimes supporting more species than oil palm (Fitzherbert et al. 2008). Borneo’s remaining forests are becoming increasingly fragmented and isolated 123 Biodivers Conserv (2010) 19:449–469 465

(Curran et al. 2004), a process that leaves bats that roost in tree cavities or foliage particularly vulnerable to population declines and local extinction (Struebig et al. 2008). Most forest has been disturbed to some extent, and the loss of tree hollows and exfoliating bark in logged over forest is predicted to have significant impacts on some forest-roosting species (Meijaard et al. 2005). Sustained disturbance to caves has had long-term negative impacts on bat populations worldwide (Hutson et al. 2001), and Borneo is likely to be no exception to this trend. At least 45% of Borneo’s bat species use caves to some extent (Payne et al. 2000), and some formations have historically hosted very large populations—Bau, Mulu and Niah in Sar- awak; Gomantong and Madai in Sabah; and the Sangkulirang peninsula in East Kalimantan (Abdullah et al. 2007; Hutson et al. 2001; Mohd-Azlan et al. 2005; Suyanto and Struebig 2007). Since cave formations are patchily distributed over Borneo (Mackinnon et al. 1996), the protection of these landscapes is of immediate importance to bat conservation efforts. The Meratus mountains and Sangkulirang peninsula in Kalimantan represent the largest areas of karst limestone in Borneo, and indeed much of Southeast Asia (Mackinnon et al. 1996), yet they remain understudied and are protected only for watershed management, not for biodiversity. Efforts to include Sangkulirang as a world heritage site have so far been unsuccessful, but Mulu has achieved this status and is protected as a national park; as is Niah. However, land clearance and fires around caves, quarrying of limestone for cement, and the collection of guano and swiftlet nests all significantly threaten resident bat colonies. Oil palm and cocoa plantations now surround most limestone karst outcrops in Sabah and Sarawak (Hobbs 2004), and fire has caused visible damage to formations at Sangkulirang (Suyanto and Struebig 2007). Anecdotal reports have implicated the activities of guano and swiftlet nest collectors to huge bat population declines at caves in Sabah and Sarawak (Hutson et al. 2001), and populations in caves with evidence of nest collection at Sangkulirang were small or non-existent (Suyanto and Struebig 2007). Nest collection rates in large caves in Sabah and Sarawak remain unsustainable, and issues of ownership, corruption and outside control of the trade need to be addressed if activities are to be regulated and cave visitation minimised (Hobbs 2004). Hunting is a major threat to bat populations on tropical islands, especially the larger fruit bats, and Borneo is no exception (Wiles and Brooke 2008). On Borneo a few taxa are the target of opportunistic collection, for example those species that can be collected easily from bamboo culms (possibly Tylonycteris spp. or Glischropus tylopus), furled banana leaves (Myotis muricola) (Mohd-Azlan et al. 2003); or under palm fronds in village gardens (Cynopterus spp., M. Struebig, Unpublished data). These are typically common and widespread species, which are unlikely to undergo significant long-term population declines. The hunting of Pteropus vampyrus, however, is much more extensive, and in some places could be causing catastrophic population declines. Hunting is unlikely to have a major impact in areas where no commercial market exists, such as remote settlements (e.g. Bulungan in East Kalimantan, Puri 2001) or those with predominantly muslim inhabitants. However, preliminary studies of Pteropus hunting in Sarawak (Gumal 2000) and in the protected Sebangau catchment in Kalimantan (Struebig et al. 2007) suggest intensive harvests exist to meet market demand, and populations show no signs of recovery. The potential long distance movements of flying-foxes (Smith 2007) obscure the true impact of hunting and the efficacy of reserves in protecting Pteropus populations, but in the meantime the listing of all Pteropus species on Appendix 2 of CITES does little to curb domestic trade (Mohd-Azlan et al. 2001), and the quotas introduced in Indonesia (A. Suyanto, Personal Communication) are likely to be ignored.

123 466 Biodivers Conserv (2010) 19:449–469

Improving bat conservation research in Borneo

Bats are relatively understudied compared to most other animals groups and considerable geographic and methodological biases are evident in mammal research on Borneo. Our review of species-locality records highlights the paucity of research effort in Kalimantan compared to Malaysia and Brunei, and illustrates how this geographic bias has resulted in a distorted perception of species distributions on the island and hence species representation in protected areas. Although Sabah has hosted most bat studies (Payne et al. 2000), research effort in Sarawak and Brunei has increased in recent years (e.g. Abdullah et al. 2007; Kofron 2002), though a focus on a single sampling technique (mist-nets) has resulted in low inventories in some reserves compared to better studied sites. Fruit bats dominate species records (Table 1), and many forest-interior taxa (Kerivoulinae and Murininae in particular) are missing from protected area inventories (Table 2). This is particularly interesting for Brunei, which has the greatest coverage of bat-locality records overall, but most of these records are for fruit bats and hipposiderids (Table 1). Using a range of sampling techniques in surveys should improve skewed bat inventories, and standardising effort across studies will allow for reliable comparisons between sites. It is commendable that more researchers are now using harp-traps (T. Abdullah et al. Personal Communication), though the species with the fewest locality records are those that forage in open areas where most capture techniques are likely to be ineffective. Using bat detectors may help to detect these species by their echolocation calls, but only if methodological biases are considered in survey design, data is interpreted appropriately, and there is sufficient information regarding intraspecific call variation (Fenton 2003). Unfortunately, there is still a long way to go in Southeast Asia before reliable data will be available. Our upper bound estimates of species compositions may prove valuable in evaluating the completeness of biodiversity inventories providing distribution information is accurate and up-to-date. Immediate priorities for bat diversity studies identified by our colleagues in Borneo include the limestone karsts of the Meratus mountains and Sangk- ulirang peninsula in Kalimantan; islands, particularly those off the coasts of Sabah, East, South and West Kalimantan, and Western Sarawak; disturbed landscapes throughout the island; and previously unsurveyed areas within the proposed Heart of Borneo. As researchers we need to collaborate more in our biodiversity studies, standardise our efforts, and share our findings to a global audience, but we also need to apply more of our efforts to the conservation threats that our study species face. With a plethora of high profile mammals consuming the activities of Borneo’s conservation community, improving public and governmental support for bats on this island is a big challenge, and one that bat researchers should tackle united. Research questions that need to be addressed to help promote and undertake bat conservation are similar to those for other animal groups in Southeast Asia (Meijaard and Sheil 2007), but specific topics for bats include: quantifying prey consumption by large colonies of insectivorous bats to advocate their protection in agricultural areas by plantation managers; describing roost types available to forest bats so that mitigation strategies can be developed by timber companies; defining a standardised roost count protocol to describe and monitor Borneo’s bat colonies and determine where conservation intervention is needed; and identifying areas where intensive Pteropus hunting is taking place so that relationships between the movements of colonies over the island, hunting pressures and land-use changes can be quantified. Finally, we note that protected area management will only be able to safeguard ca. 15% of Borneo’s remaining forests, with an additional 20% included in forests that are protected for watershed protection. The remainder of Borneo’s forest area is likely to be under 123 Biodivers Conserv (2010) 19:449–469 467 extractive forest management. On Borneo, this generally means that a limited number of trees (mostly \10/ha) are extracted every 25–30 years, theoretically allowing the forest to regenerate. Such forests have considerable value for wildlife (Meijaard et al. 2005). Both Indonesia and Malaysia are making slow progress towards more sustainable forest management (Dennis et al. 2008), and we expect that the importance of these managed forests for wildlife conservation will increasingly be recognised. It is important to further promote the role of well managed timber concessions in wildlife conservation, because these timber concessions can not only provide important habitats for forest-dependent bats, but they can also provide links between strictly protected areas thereby retaining the overall ecological connectivity of Borneo’s forest habitats. The near future will tell whether present trends of clearing degraded forest areas for monocultural plantations (Fitzherbert et al. 2008) will continue, or whether the value of forests will be recognised and further loss, both within and outside protected areas, will be prevented.

Acknowledgements We are most grateful to all the people who kindly contributed bat records to our database, and/or provided useful discussion on bat conservation issues in Borneo and elsewhere in Southeast Asia. These include Mohd. Tajuddin Abdullah, Mohd. Azlan J. Abd. Gulam Azad, Stephen Brend, Anne Brooke, Susan Cheyne, Louise Craig, Rona Dennis, Fanasudi, Charles Francis, Marty Fujita, Antonia Gorog, Melvin Gumal, Les Hall, Elery Hamilton-Smith, Mark Harrison, Kris Helgen, Robert Hodgkison, Jason Hon, Simon Husson, Faisal Ali Bin Anwarali Khan, Tigga Kingston, Christopher Kofron, David Lane, Suwido Limin, Ibnu Maryanto, Ellen McArthur, Kim McKonkey, Helen Morrogh-Bernard, Cahyo Rahmadi, Juliana Senawi, Indrawati Sendow, Martua Sinaga, Tony Start, Agustinus Suyanto, Gary Wiles and Shigeki Yasuma. Thanks also to Markus Radday for sharing the digitised boundaries of the Heart of Borneo. Bat surveys in Kalimantan were funded by The Nature Conservancy (in part through a grant from the Sall Foundation) and grants awarded to either MS or DP for The Kalimantan Bat Conservation Project from Bat Conservation International, British Ecological Society, Gilchrist Educational Trust, Percy Sladen Memorial Fund, Lubee Bat Conservancy, People’s Trust for Endangered Species, Project Barito Ulu, Royal Geo- graphical Society (Institute of Biology and Rio Tinto Inc.), The Orangutan Foundation and University of East Anglia. Permission to conduct bat research in Kalimantan to MS and DP was kindly granted by the Indonesian Institute of Sciences (LIPI).

References

Abdullah MT, Hall LH, Tissen OB, Tuuga A, Earl of Cranbrook (2007) The large bat caves of Malaysian Borneo. Bat Res News 48:99–100 Anwarali FA, Piksin SNS, Ketol B, Japning JRR, Julaihi AM, Hall LS et al (2007) Survey of bats in the tropical lowland dipterocarp forest of Bako National Park, Sarawak, Malaysian Borneo. Sarawak Mus J 83:1–29 Boitani L, Catullo G, Marzetti I, Masi M, Rulli M, Savini S (2006) The Southeast Asian mammal databank. A tool for conservation and monitoring of mammal diversity in Southeast Asia. Instituto di Ecologia Applicata, Rome. http://www.ieaitaly.org/samd/. Cited 31 March 2008 Bruner AG, Gullison RE, Rice RE, da Fonseca G (2001) Effectiveness of parks in protecting tropical biodiversity. Science 291:125–128. doi:10.1126/science.291.5501.125 Curran LM, Trigg SN, McDonald AK, Astiani D, Hardiono YM, Siregar P et al (2004) Lowland forest loss in protected areas of Indonesian Borneo. Science 303:1000–1003. doi:10.1126/science.1091714 Danielsen F, Beukema H, Burgess N, Parish F, Bruhl CA, Donald PF et al (2008) Biofuel plantations on forested lands: double jeopardy for biodiversity and climate. Conserv Biol (in press) Dennis RA, Meijaard E, Nasi R, Gustafsson L (2008) Biodiversity conservation in SE Asian timber concessions: an overview of the implementation of guidelines and recommendations. Ecol Soc 13. http:// www.ecologyandsociety.org/vol13/iss1/art25/ du-Toit JT, Walker BH, Campbell BM (2004) Conserving tropical nature: current challenges for ecologists. Trends Ecol Evol 19:12–17 Fenton MB (2003) Science and the conservation of bats: where to next? Wildl Soc Bull 31:6–15 Fitzherbert EB, Struebig M, Morel A, Danielsen F, Bru¨hl C, Donald PF et al (2008) How will oil palm expansion affect biodiversity? Trends Ecol Evol 23:538–545. doi:10.1016/j.tree.2008.06.012 123 468 Biodivers Conserv (2010) 19:449–469

Francis CM (1990) Trophic structure of bat communities in the understorey of lowland dipteropocarp rain forest in Malaysia. J Trop Ecol 6:421–431 Fujita MS, Tuttle MD (1991) Flying foxes (Chiroptera: Pteropodidae): threatened animals of key ecological importance. Conserv Biol 5:455–463. doi:10.1111/j.1523-1739.1991.tb00352.x Fuller DO, Jessup TC, Salim A (2004) Loss of forest cover in Kalimantan, Indonesia, since the 1997–1998 El Ninõ. Conserv Biol 18:249–254. doi:10.1111/j.1523-1739.2004.00018.x Gumal MT (2000) Ecology and conservation of a fruit bat in Sarawak, Malaysia. Dissertation, University of Cambridge, p 234 Hobbs JJ (2004) Problems in the harvest of edible birds’ nests in Sarawak and Sabah, Malaysian Borneo. Biodivers Conserv 13:2209–2226. doi:10.1023/B:BIOC.0000047905.79709.7f Hortal J, Lobo JM, Jimenez-Valverde A (2007) Limitations of biodiversity databases: case study on seed- plant diversity in Tenerife, Canary Islands. Conserv Biol 21:853–863. doi:10.1111/j.1523- 1739.2007.00686.x Hutson AM, Mickleburgh SP, Racey PA (2001) Microchiropteran bats: global status survey and conservation action plan. IUCN/SSC Chiroptera Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK IUCN (1994) Guidelines for protected area management categories. IUCN Commission on National Parks/ World Conservation Monitoring Centre, The World Conservation Union, Gland Jayaraj VK, Ketol B, Khan FAA, Hall LS, Abdullah MT (2006) Bat survey of Mount Penrisen and notes on the rare Kerivoula minuta, Kerivoula intermedia and Hipposideros coxi in Sarawak, Borneo. J Biol Sci 6:711–716 Jennings MD (2000) Gap analysis: concepts, methods, and recent results. Landsc Ecol 15:5–20. doi:10.1023/A:1008184408300 Jepson P, Jarvie JK, MacKinnon K, Monk KA (2001) The end of Indonesia’s lowland forests? Science 292:859–861. doi:10.1126/science.1061727 Jepson P, Momberg F, van-Noord H (2002) A review of the efficacy of the protected area system of East Kalimantan province, Indonesia. Nat Areas J 22:28–42 Kingston T (2008) Research priorities for bat conservation in Southeast Asia: a consensus approach. Bio- divers Conserv. doi:10.1007/s10531-008-9458-5 Kingston T (2009) Analysis of species diversity of bat assemblages. In: Kunz TH, Parsons S (eds) Behavioral and ecological methods for the study of bats. Smithsonian Institution, Washington (in press) Kingston T, Francis CM, Zubaid A, Kunz TH (2003) Species richness in an insectivorous bat assemblage from Malaysia. J Trop Ecol 19:67–79. doi:10.1017/S0266467403003080 Kofron CP (2002) The bats of Brunei Darussalam, Borneo. Mammalia 66:259–274 Koh LP (2007) Impending disaster or sliver of hope for Southeast Asian forests? The devil may lie in the details. Biodivers Conserv 16:3935–3938. doi:10.1007/s10531-007-9177-3 Lane DJW, Kingston T, Lee BPY-H (2006) Dramatic decline in bat species richness in Singapore, with implications for Southeast Asia. Biol Conserv 131:584–593. doi:10.1016/j.biocon.2006.03.005 Langner A, Miettinen J, Siegert F (2007) Land cover change 2002–2005 in Borneo and the role of fire derived from MODIS imagery. Glob Chang Biol 13:2329–2340. doi:10.1111/j.1365- 2486.2007.01442.x Mackinnon K, Hatta G, Halim H, Mangalik A (1996) The ecology of Kalimantan. Periplus Editions, Singapore McCune B, Meffor MJ (1999) Multivariate analysis of ecological data. PcORD version 4.25. MJM Software Designs, Gleneden Beach, OR McKenzie NL, Gunnell AC, Yani M, Williams MR (1995) Correspondence between flight morphology and foraging ecology in some Paleotropical bats. Aust J Zool 43:241–257. doi:10.1071/ZO9950241 Meijaard E, Nijman V (2003) Primate hotspots on Borneo: predictive value for general biodiversity and the effects of taxonomy. Conserv Biol 17:725–732. doi:10.1046/j.1523-1739.2003.01547.x Meijaard E, Sheil D (2007) Is wildlife research useful for wildlife conservation in the tropics? A review for Borneo with global implications. Biodivers Conserv 16:3053–3065. doi:10.1007/s10531-007-9161-y Meijaard E, Sheil D, Nasi R, Augeri D, Rosenbaum B, Iskandar D et al (eds) (2005) Life after logging: reconciling wildlife conservation and production forestry in Indonesian Borneo. Center for Interna- tional Forestry Research, Bogor Mohd-Azlan J, Zubaid A, Kunz TH (2001) Distribution, relative abundance and conservation status of the large flying fox: Pteropus vampyrus, in peninsular Malaysia: a preliminary assessment. Acta Chiropt 3:149–162

123 Biodivers Conserv (2010) 19:449–469 469

Mohd-Azlan J, Maryanto I, Kartono AP, Abdullah MT (2003) Diversity, relative abundance and conservation of chiropterans in Kayan Mentarang National Park, East Kalimantan, Indonesia. Sarawak Mus J 58:251–265 Mohd-Azlan J, Neuchlos J, Abdullah MT (2005) Diversity of chiropterans in limestone forest area, Bau, Sarawak. Malaysian. Appl Biol 34:59–64 Payne J, Francis CM, Phillips K, Kartikasari SN (2000) Mamalia di Kalimantan, Sabah, Sarawak dan Brunei Darussalam. Wildlife Conservation Society Indonesia Program, Jakarta Puri RK (2001) Bulungan ethnobiology handbook. A field manual for biological and social science research on the knowledge and use of plants and animals amongst 18 indigenous groups in northern East Kalimantan. Centre for International Forestry Research (CIFOR), Bogor Rodrigues ASL, Andelman SJ, Bakarr ML, Boitani L, Brooks TM, Cowling RM et al (2004) Effectiveness of the global protected area network in representing species diversity. Nature 428:640–643. doi:10.1038/nature02422 Rodrı´guez JP, Brotons L, Bustamante J, Seoane J (2007) The application of predictive modelling of species distribution to biodiversity conservation. Divers Distrib 13:243–251 Shen TJ, Chao A, Lin JF (2003) Predicting the number of new species in further taxonomic sampling. Ecology 84:798–804. doi:10.1890/0012-9658(2003)084[0798:PTNONS]2.0.CO;2 Simmons N (2005) Chiroptera. In: Wilson DE, Reeder DM (eds) Mammal species of the world: a taxonomic and geographic reference. John Hopkins University Press, Baltimore, pp 312–529 Smith C (2007) Use of satellite telemetry to study movement of the Malayan flying fox (Pteropus vampyrus): implications for conservation and public health. In: Wildlife health in a shrinking world: ecology, management and conservation. Proceedings of the wildlife disease association international conference, Cairns, Australia, 26 June–1 July 2005 Sodhi NS, Koh LP, Brook BW, Ng PKL (2004) Southeast Asian biodiversity: an impending disaster. Trends Ecol Evol 19:654–659. doi:10.1016/j.tree.2004.09.006 Solow AR, Polasky S (1999) A quick estimator for taxonomic surveys. Ecology 80:2799–2803 Struebig M, Benton-Browne A, Rachmad A, Yusliati N, Atmoko T, Rustam W, Fredriksson G, Meijaard E (2006a) A bat survey in Sungai wain protection Forest, East Kalimantan, Indonesia. Malay Nat J 59:189–196 Struebig MJ, Galdikas BMF, Suatma (2006b) Bat diversity in oligotrophic forests of Southern Borneo. Oryx 40:447–455. doi:10.1017/S0030605306001190 Struebig MJ, Harrison ME, Cheyne SM, Limin S (2007) Intensive hunting of large flying-foxes Pteropus vampyrus natunae in Central Kalimantan, Indonesian Borneo. Oryx 41:390–393. doi:10.1017/ S0030605307000310 Struebig MJ, Kingston T, Zubaid A, Mohd-Adnan A, Rossiter SJ (2008) Conservation value of forest fragments to Palaeotropical bats. Biol Conserv 141:2112–2126. doi:10.1016/j.biocon.2008.06.009 Sujatnika JP, Soehartono T, Crosby MJ, Mardiastuti A (1995) Melestarikan keanekaragaman hayati Indo- nesia: pendekatan daerah burung endemik (Conserving Indonesian biodiversity: the endemic bird area approach). Perlindungan Hutan dan Pelestarian Alam and Birdlife International Indonesia Programme, Jakarta Suyanto A, Struebig MJ (2007) Bats of the Sangkulirang limestone karst formations, East Kalimantan— a priority region for Bornean bat conservation. Acta Chiropt 9:67–95. doi:10.3161/1733- 5329(2007)9[67:BOTSLK]2.0.CO;2 Wiles GJ, Brooke AP (2008) Conservation threats to bats in the tropical pacific islands and insular Southeast Asia. In: Fleming TH, Racey P (eds) Island bats: evolution, ecology and conservation. University of Chicago, Chicago (in press) WWF (2005) Borneo—treasure island at risk. WWF Germany, Frankfurt am Main Zubaid A (1993) A comparison of the bat fauna between a primary and fragmented secondary forest in Peninsular Malaysia. Mammalia 57:201–206

123