Global diversity of amphibians (Amphibia) in freshwater

Transcription

1 Hydrobiologia (2008) 595: DOI /s FRESHWATER ANIMAL DIVERSITY ASSESSMENT Global diversity of amphibians (Amphibia) in freshwater Miguel Vences Æ Jörn Köhler Ó Springer Science+Business Media B.V Abstract This article present a review of species numbers, biogeographic patterns and evolutionary trends of amphibians in freshwater. Although most amphibians live in freshwater in at least their larval phase, many species have evolved different degrees of independence from water including direct terrestrial development and viviparity. Of a total of 5,828 amphibian species considered here, 4,117 are aquatic in that they live in the water during at least one lifehistory stage, and a further 177 species are waterdependent. These numbers are tentative and provide a conservative estimate, because (1) the biology of many species is unknown, (2) more direct-developing species e.g. in the Brachycephalidae, probably depend directly on moisture near water bodies and (3) the accelerating rate of species discoveries and descriptions in amphibians indicates the existence of many more, yet undescribed species, most of which are likely to have aquatic larvae. Regional endemism in Guest editors: E. V. Balian, C. Lévêque, H. Segers & K. Martens Freshwater Animal Diversity Assessment M. Vences (&) Division of Evolutionary Biology, Zoological Institute, Technical University of Braunschweig, Spielmannstr. 8, Braunschweig 38106, Germany m.vences@tu-bs.de J. Köhler Department of Zoology, Hessisches Landesmuseum Darmstadt, Friedensplatz 1, Darmstadt 64283, Germany j.koehler@hlmd.de amphibians is very high, with only six out of 348 aquatic genera occurring in more than one of the major biogeographic divisions used herein. Global declines threatening amphibians are known to be triggered by an emerging infectious fungal disease and possibly by climate change, emphasizing the need of concerted conservation efforts, and of more research, focused on both their terrestrial and aquatic stages. Keywords Amphibia Anura Urodela Gymnophiona Species diversity Evolutionary trends Aquatic species Biogeography Threats Introduction Amphibians are a textbook example of organisms living at the interface between terrestrial and aquatic habitats. They fulfil this role both in an ecological context, with typically a strictly aquatic larval and largely terrestrial adult phase, and in an evolutionary context, representing the intermediate bauplan level between aquatic and fully terrestrial vertebrates ( fishes vs. amniotes). Most amphibians are strictly dependent from water for their larval development, and water for this group of animals is synonym to freshwater. Although a few amphibians are able to tolerate high-salinity levels (Balinsky, 1981), there are no marine representatives of this class. Although existence of an aquatic larval phase is probably the ancestral condition for recent

2 570 Hydrobiologia (2008) 595: amphibians, there are only few amphibian taxa with also fully aquatic adult phases. In contrast, multiple evolutionary trends towards more terrestrial reproduction have led to a plethora of reproductive modes (Duellman & Trueb, 1986) which make it difficult, in some instances, to decide if and to what degree a particular species is indeed strictly dependent on freshwater. Recent amphibians are often named Lissamphibia. They are divided in three orders: frogs (Anura), salamanders (Urodela), and caecilians (Gymnophiona). Dubois (2004) recommended abandoning several other higher taxa names based on arguments of nomenclatural priority. Although these priority rules do not strictly apply to names above the family level, we here follow Dubois (2004) in not using the names Apoda (for caecilians), Caudata (for salamanders), Salientia (for Recent frogs), and Archaeobatrachia and Neobatrachia (for basal and modern frogs). However, we decided to here continue using Lissamphibia for the clade containing all three recent amphibian orders, and we use Archaeobatrachia and Neobatrachia in quotation marks since these established terms make discussion of anuran relationships easier. Due to the large diversity of extinct Paleozoic amphibians, the phylogenetic relationships of lissamphibians relative to amniotes has in the past been questioned. Current evidence converges on their monophyly, based on morphological characters such as, for instance, their pedicellate teeth, special visual cells (green rods) in the retina, or the ear structure (Duellman & Trueb, 1986), and on molecular characters (e.g. Meyer & Zardoya, 2003; San Mauro et al., 2005). The paucity of fossils, especially from the Mesozoic, makes it difficult to trace the early evolution of lissamphibians, but they appear to be a very old group according to molecular clocks which date the separation among the three orders back into Paleozoic times before Pangaean break-up (San Mauro et al., 2005; Roelants & Bossuyt, 2005). Furthermore, deep divergences are also typical for amphibians at the species level. As already noted by Wilson et al. (1974), amphibian species have much larger molecular divergences (and, consequently, ages) than other vertebrates e.g. mammals and birds, and also the large divergences among populations considered to be conspecific are typical for amphibians (e.g. Vences et al., 2005a, b). The (aquatic) larvae of caecilians and salamanders are morphologically largely similar to their adults, except for the presence of external gills which usually are reduced in the adults. In contrasts, the larval stage of frogs, the tadpole, is a larval phase radically different from the adults (Altig & McDiarmid, 1999). Especially the oral and digestive system of tadpoles is composed of numerous features which are not homologous to the corresponding structures in the adult, such as a horny beak, oral papillae and keratinous labial teeth, mainly due to the fact that typically tadpoles are omnivorous suspension feeders, ingesting a high degree of vegetable matter, while adult frogs are strictly carnivorous (with only a single known exception, Xenohyla truncata, a species that also eats fruits). Several excellent resources on amphibians were available over the world wide web at the time of preparation of the present article. The Amphibian Species of the World database (Frost, 2004), hosted by the American Museum of Natural History, continues previous efforts (Frost, 1985; Duellman, 1993) to document from a taxonomic point of view all amphibian taxa. Amphibiaweb (2005, hosted by the University of California at Berkeley, provides a full species list of amphibians, too, but aims at providing also additional information such as distribution, photographs, and biological data. The Global Amphibian Assessment ( has compiled, during , the expertise of regional and taxonomic experts worldwide and provides an estimate of threat status (according to IUCN criteria) and distribution for all amphibian species. In this article, we summarize species diversity and distribution, and zoogeography, of extant amphibians, based on a species list and distributional information compiled from these three online data sources as accessed in December Furthermore, we categorize all amphibian species according to their water dependence on a regional and taxonomical level. For taxonomy we generally follow Frost et al. (2006). In the following sections, we use the definitions of the freshwater diversity assessment project in defining (1) aquatic species as those with at least part of their life cycle in or on the water, (2) water dependent species as those which do not live directly in the water but closely depend on it e.g. for habitat or food, (3) water related species as aquatic plus water dependent

3 Hydrobiologia (2008) 595: species, and (4) nor water related species as those which are neither aquatic nor water dependent. Species diversity A striking aspect of amphibian taxonomy is the increasing rate of new species discoveries (Glaw & Köhler, 1998; Köhler et al., 2005). As of December 2005, a total of 5,828 amphibian species (aquatic+ non-aquatic) were known, but still at the end of 1992, there were only 4,533 species (Duellman, 1993). The absolute number of newly described amphibian species per decade (not only the cumulative number of valid and described species) has been steadily increasing since the decade of the 1960s, with especially steep increases since the 1990s (Glaw & Köhler, 1998; Köhler et al., 2005). The new species are in part known populations of described species that are found to be genetically or bioacoustically distinct, and hence recognized as different species. However, the largest proportion of new species are genuine new discoveries, as exemplified by the recent spectacular findings of a new frog lineage, genus and species in India, Nasikabatrachus (Biju & Bossuyt, 2003; considered to be part of the family Sooglossidae by Frost et al, 2006), and of a new genus and species of plethodontid salamander, Karsenia, in Korea, being the first Asian representative of this family (Min et al., 2005). This taxonomic progress has been made possible by the combination of increased field exploration in tropical regions, together with the application of molecular and bioacoustic techniques becoming routine. A case study in Madagascar indicated that newly discovered species since 1990 are genetically not less distinct from already described species than species discovered in the research periods before, and that the increase in new species is not a sign of taxonomic inflation due to exaggerated splitting approaches (Köhler et al., 2005). Of the total of 5,828 amphibian species considered here, 4,117 are aquatic in that they live in the water during at least one life-history stage, and a further 177 species are water dependent. By the end of 2005, there were a total of 168 species of caecilians, 514 salamanders and 5,146 frogs. While it is relatively straightforward to decide which of these are, under the definitions used here, real aquatics (i.e. with at least part of the life cycle in or on the water), it is more difficult to decide which of the non-aquatic species may be water dependent, i.e. with close/ specific dependence on aquatic habitats (see Figs. 1 and 2). In our categorization, all amphibians characterized by direct development, viviparity with terrestrial birth, or tadpole development in terrestrial jelly or foam nests are considered to be non-aquatic, while species with tadpoles in water-filled leaf axils of plants or tree holes were included in the aquatic category (Figs. 2 and 3). To be able to categorize all species, we have here assumed that species share the reproductive mode of their closest relatives (usually congeneric taxa). Although certainly not fully precise, this estimate should be a relatively reliable approximation to the real situation. In our analysis, species were categorized as unknown with respect to their dependence from freshwater only when no data at all were available to us concerning the life history of the particular genera. This concerns a small portion of 67 amphibian species only (Fig. 2). Several of the non-aquatic species are certainly water dependent. For instance, the South African pyxicephalids of the genera Arthroleptella and Anhydrophryne have direct development but usually live in dense, mossy vegetation around springs and cannot colonize other habitats. Many plethodontid salamanders have direct development and do not live in the water, but are predominantly found along streams because they rely on the humid substrate nourished by the water. These species were placed in category water dependent. Nevertheless, there certainly are species which are fully independent from water, such as the desert-dwelling species of Breviceps which occur far from any water body and depend on air humidity only. Several direct-developing frogs of the genera Craugastor, Euhyas and Eleutherodactylus live in bromeliads, and probably depend on the moisture provided by water-filled phytotelmes. However, since the life habits of very few species in this species-rich genera were studied, we here found it premature to decide which and how many species are non-aquatic but water dependent. As we did in the genera Craugastor, Euhyas and Eleutherodactylus, we considered all non-aquatic species where natural history observations are sparse also as non-water dependent, which is probably true for the majority of these species. However, it is implicit in this procedure that the numbers provided in Tables 1 and 2 will be slight underestimates.

4 572 Hydrobiologia (2008) 595: Fig. 1 Map showing species and genus diversity of water related amphibian species by major zoogeographic divisions. PA, Palearctic; NA, Nearctic; NT, Neotropical; AT, Afrotropical; OL, Oriental; AU, Australasian; PAC, Pacific Oceanic Islands; ANT, Antarctic. Numbers include aquatic amphibian species (with at least one aquatic life-history stage) plus those that are water dependent (e.g. some direct-developing species) Fig. 2 Percentages (rounded) of amphibian species in categories aquatic, water dependent, not water related and unknown according to the definitions used herein. Together, the categories aquatic and water dependent are summed up as Total in Tables 1 and 2 Historical processes: evolutionary trends away from and towards water As outlined above, fully aquatic adult amphibians are rare. Among anurans, only representatives of the family Pipidae qualify as such plus very few species of other families, while among salamanders and caecilians, an aquatic life history is more common. In caecilians, species of the family Caeciliidae are reported to have aquatic adults with viviparous reproduction. However, ecological studies suggest that at least some species may not be strictly aquatic but actually display a semi-aquatic behaviour with resting periods out of water and foraging in aquatic habitat (Moodie, 1978). Among salamanders, fully aquatic families are the Cryptobranchidae, Amphiumidae, Sirenidae, and Proteidae, and aquatic adults also occur in the genus Ambystoma. In many of these species, for example in the well-known Axolotl, Ambystoma mexicanum (Ambystomatidae), the aquatic adults are paedomorphic (neotenic), and retain larval features such as external gills and a fully functional lateral-line system. The most extreme modifications of reproductive modes are those that completely eliminate the freeliving larval phase: direct development, and sometimes viviparity. Viviparous and ovoviviparous amphibians are relative rare, but caecilians are an exception where about 75% of the known species are considered to have a viviparous mode of reproduction (Himstedt 1996). Members of the Neotropical genus Typhlonectes and of the Afrotropical genus Scolecomorphus are exclusively viviparous, and viviparity also occurs in some other genera of the family Caeciliidae. In salamanders, viviparous species are

5 Hydrobiologia (2008) 595: Fig. 3 Percentages (rounded) of frog (A), salamander (B) and caecilian (C) species in categories aquatic, water dependent, not water related and unknown. Together, the categories aquatic and water dependent are summed up as Total in Tables 1 and 2 found exclusively in the Salamandridae (genera Salamandra and Lyciasalamandra). Among frogs, viviparity is known only in the two African bufonid genera Nectophrynoides and Nimbaphrynoides, and in the brachycephalid Eleutherodactylus jasperi (see Wake, 1977; 1989). All viviparous frogs bear fully metamorphosed young and therefore have no aquatic larval stage. Viviparous salamanders can bear metamorphosed young or (aquatic) larvae. Caecilians of the genus Typhlonectes are viviparous with aquatic larvae, whereas other viviparous caecilians have no larval stage. In pipid frogs of the genus Pipa, the eggs are embedded in the dorsum of the (aquatic) female and thus indirectly undergo development in an aquatic environment; in all other cases, direct-developing amphibians lay terrestrial eggs. Direct development has evolved independently in many amphibian lineages: in salamanders once in the family Plethodontidae, which contains almost exclusively direct developers; in caecilians, in the family Caeciliidae; and among anurans in many of the major lineages: (1) in the basal Leiopelmatidae, genus Leiopelma from New Zealand; (2) in sooglossids (genus Sooglossus); (3) in some species of the genus Pipa (Pipidae); (4) in some genera of myobatrachids; (5) in brachycephalids; (6) in Hemiphractidae (genus Hemiphractus); (7) in Cryptobatrachidae (genera Cryptobatrachus, Stefania); (8) in Amphignathodontidae (genera Flectonotus, Gastrotheca); (9) in several genera of bufonids (e.g. Oreophrynella, Osornophryne, Rhamphophryne); (10) in the genus Platymantis (Ceratobatrachidae); (11) in Pyxicephalidae (Arthroleptella and related genera); (12) in at least one dicroglossid species (Limnonectes hascheanus); (13) in several species of one genus of mantellids (Gephyromantis; see Glaw & Vences, 2006); (14) in at least one genus of rhacophorids (Philautus); (15) in brevicipitids (Breviceps and related genera); (16) in Australasian microhylids (subfamily Asterophryinae); (17) in few Neotropical microhylids (Myersiella, Synapturanus); (18) in several arthroleptids (genera Arthroleptis, Schoutedenella). Hence, altogether there are at least 18 independent evolutionary events towards direct development in anurans, while this reproductive mode has evolved only once in salamanders and probably twice in caecilians. It needs to be emphasized that these are minimum estimates, and especially in frogs it is likely that more complete phylogenetic data will provide evidence that direct development evolved even more commonly in parallel.

6 574 Hydrobiologia (2008) 595: Table 1 Numbers of water related amphibian species by order PA NA NT AT OL AU PAC ANT World Order FW WDpt FW WDpt FW WDpt FW WDpt FW WDpt FW WDpt FW WDpt FW WDpt FW WDpt Anura , , , Urodela Gymnophiona Total , , , FW refers to aquatic plus water dependent species, WDpt to water dependent only (of a total of 5828 amphibian species considered; see text for definitions). PA : Palearctic; NA : Nearctic; NT : Neotropical; AT : Afrotropical ; OL : Oriental; AU : Australasian; PAC : Pacific Oceanic Islands, ANT: Antarctic Interestingly, some of the direct-developing amphibian lineages are characterized by a very high species richness. The Brachycephalidae contain approximately 800 direct-developing species, and among salamanders, the largely direct-developing Plethodontidae encompass by far the largest number of species (349 out of a total of 514 salamander species). This may be seen as indication that waterindependence is a particularly successful strategy for amphibians. It could also be a by-product of a putative higher fragmentation into isolated demes in direct developers, which may lead to an increased rate of species formation (Dubois, 2005). Studies that apply comparative methods to test against null models, and population genetic studies of direct developers are necessary to clarify this question. Remarkably, recent phylogenetic evidence indicates that in a number of groups of predominantly direct development, some lineages have reversed their reproductive mode and re-acquired an aquatic larval stage. This appears to be the case in plethodontid salamanders, Desmognathus, as well as in some amphignathodontid frogs, Gastrotheca, and possibly in mantellid frogs as well (Duellman & Hillis, 1987; Vences & Glaw, 2001; Chippindale et al., 2004). These reversed trends emphasize the selective advantage, under at least some evolutionary conditions, of biphasic aquatic-terrestrial life cycles, and the importance of freshwaters for amphibian diversity. Distribution and endemicity Amphibians are in general considered to be poor dispersers, and the strong phylogeographic structure encountered in many amphibian species (e.g. Avise, 2000) appears to support this view. Due to their limited osmotic tolerance, overseas dispersal was long neglected as dispersal mechanism for amphibians, and their zoogeographic patterns explained largely by vicariance and dispersal over land connections (e.g. Duellman & Trueb, 1986). Evidence from molecular clocks and the discovery of endemic amphibians on oceanic islands, such as Mayotte on the Comoros, provide strong support that amphibians are able to colonize landmasses over the sea (e.g. Hedges et al., 1992; Vences et al., 2003; 2004). This concerns frogs, but may also apply to salamanders

7 Hydrobiologia (2008) 595: Table 2 Numbers of amphibian genera including aquatic and water dependent species according to the definitions used herein by order Order PA NA NT AT OL AU PAC ANT World Anura Urodela Gymnophiona Total For the subfamily Hylinae, the classicatory scheme of Faivovich et al. (2005) is followed. However, all other numbers in genera have not been updated and refer to a taxonomy prior to the publication of Frost et al. (2006). Therefore, some genera were supposed to occur in more than one biogeographic region and thus the sums of all regional numbers are higher than the total numbers. PA : Palearctic; NA : Nearctic; NT : Neotropical; AT : Afrotropical; OL : Oriental; AU : Australasian; PAC : Pacific Oceanic Islands, ANT: Antarctic and even caecilians, as indicated by the presence of an endemic caecilian species, Schistometopum thomense, on the fully volcanic São Tomé island in the Gulf of Guinea (Measey et al., 2007). Nevertheless, it remains true that amphibian distributions have certainly largely been shaped by vicariance, as shown by relationships of relict forms such as the Seychellean sooglossid frogs and the Indian Nasikabatrachus (Biju & Bossuyt, 2003). At the deep phylogenetic levels, there are clear distributional trends of salamanders and basal frogs having their centres of diversity in the Holarctis and caecilians and modern frogs in the tropics. Since some phylogenetic reconstructions placed caecilians as sister group of salamanders, and both basal and modern frogs ( archaeobatrachians and neobatrachians ) as monophyletic groups, the distributional patterns of these lineages were interpreted by some phylogeneticists as indicative of vicariance during the break-up of Pangaea into the Laurasia and Gondwana supercontinents, with caecilians and neobatrachians evolving and diversifying on Gondwana, and archaeobatrachians and salamanders diversifying on Laurasia (Feller & Hedges, 1998). Recent phylogenetic evidence, however, does not support this hypothesis. Evidence from complete mitochondrial sequences and nuclear genes indicates that frogs and salamanders, not salamanders and caecilians, are sister groups (Meyer & Zardoya, 2003; San Mauro et al., 2005; Frost et al., 2006). In addition, phylogenetic reconstructions based on different nuclear genes are concordant in establishing paraphyly of basal frogs ( archaeobatrachians ) relative to the monophyletic neobatrachians (Hoegg et al., 2004; San Mauro et al., 2005; Roelants & Bossuyt, 2005), in accordance with morphological hypotheses (e.g. Duellman & Trueb, 1986). Furthermore, also the distributional patterns observed leave room for the assumption that not only causes of vicariance biogeography, but also of ecological requirements have shaped the distribution of the three amphibian orders. Salamanders are almost exclusively distributed on previous Laurasian landmasses to which 9 of the 10 salamander families are restricted, if the presence of a few representatives of salamandrids (Salamandra and Pleurodeles) in northern Africa is disregarded. This pattern is obscured by the fact that one large radiation of one family, the Plethodontidae, has colonized the Neotropics and attains a high species diversity in Mexico and Central America. Indeed, 252 salamander species occur in the Neotropics as defined here, more than in any other biogeographic region. However, only few species of two genera, Bolitoglossa and Oedipina, have penetrated further into South America, leaving no doubts that northern America was the initial centre of diversification of this family. Almost all plethodontids are characterized by direct development, and are therefore less relevant in the present survey of freshwater diversity, most species not being included in Table 1. Salamanders have almost not penetrated into tropical areas of Asia, although there are salamandrids occurring as far to the south as Laos and Vietnam. Caecilians have a distribution fully restricted to the tropics. They are found in the Neotropical, Afrotropical and Oriental regions. Interestingly, although endemic caecilians are present on the Seychelles, they are absent from Madagascar. Caecilians do not occur in southernmost South America, or in southern

8 576 Hydrobiologia (2008) 595: Africa, indicating that the limiting factor for their distribution is indeed the presence of tropical-humid environments. Apart from climate which obviously triggers the current distribution of caecilians, it has been assumed that radiation in caecilians largely took place before Gondwana split into sub-continents. Some families like the South American Rhinatrematidae and the Asian Ichthyophiidae are supposed to represent relict distributions of formerly widespread Gondwanan ancestors (Duellman & Trueb, 1986), whereas some Neotropical members of the Caeciliidae (e.g. Typhlonectes) possibly are the product of subsequent radiations on the already isolated South American continent (Himstedt, 1996). Basal frog lineages ( archaeobatrachians, defined as a paraphyletic group of all extant frogs not belonging to the modern frogs or neobatrachians ) are mainly distributed in the Holarctis, with four notable exceptions, however. (1) Pipids, the only frogs which are fully aquatic also in their adult stage, have a clearly Gondwanan distribution, with genera in Africa (Xenopus, Silurana, Pseudhymenochirus and Hymenochirus; 23 species) and in South America (Pipa; 7 species). (2) Leiopelmatidae: the genus Leiopelma (4 species) occurs in New Zealand, although its closest relative, Ascaphus, is restricted to the Nearctis. (3) The discoglossid genus Barbourula occurs on Borneo and the Philippines, whereas its closest relatives, the genus Bombina, has a Palearctic distribution. And (4) the Megophryidae, with 72 species by far the largest archaeobatrachian family, has radiated in the Oriental region and is common in tropical environments. Based on a robust molecular phylogeny and molecular clock dating, Roelants & Bossuyt (2005) found evidence for three major cladogenetic events between a Laurasia- and a Gondwana-associated lineage, represented by Ascaphus and Leiopelma, Rhinophrynidae and Pipidae and Pelobatoidea and Neobatrachia, respectively, all these splits being very close to the onset of Pangaean break-up at 180 mya. Although this pattern substantiates a high-biogeographic relevance of archaeobatrachians, they altogether make up only a negligible part of overall frog diversity, with a total of 191 of the total of 5,146 frog species (including non-water related taxa). Neobatrachians, with 4,955 species, do not only form the most speciose anuran subgroup, but also include by far more species than all other amphibian groups together. Monophyly of these modern frogs is well established by molecular and morphological characters (Duellman & Trueb, 1986; Hoegg et al., 2004; Roelants & Bossuyt, 2005; San Mauro et al., 2005). Their largest diversity belongs into two subgroups, the hyloids, with a centre of diversity in the Neotropics, and the ranoids, with a centre of diversity in Africa and Asia. Ranoids, according to the scheme of Frost et al. (2006), include families restricted to Africa and/or Madagascar, such as the Arthroleptidae, Hemisotidae, Hyperoliidae, Mantellidae, Phrynobatrachidae, Ptychadenidae and Pyxicephalidae; two mainly Asian families with few African representatives, the Rhacophoridae and Dicroglossidae; two South Asian families, the Nyctibatrachidae and Microxalidae; one South-East Asian family, the Ceratobatrachidae; one family present in Africa and Asia, and that succeeded to colonize also North and South America, the Microhylidae; and the species-rich Ranidae that colonized Europe as well as North and South America. The Pedropeditae sensu Frost et al. (2006) contain African and South Asian species, although these relationships require further corroboration. Hyloids include several families restricted to the Neotropics, such as the Amphignathodontidae, Brachycephalidae, Ceratophryidae, Centrolenidae, Cryptobatrachidae, Cycloramphidae, Dendrobatidae, Hemiphractidae, Leptodactylidae, Thoropidae; one family, the Bufonidae, common in the Neotropis with also many representatives in the Palearctic, Nearctic, Oriental, and Afrotropical regions, including genera endemic to the main biogeographical regions; and one family, the Hylidae, with many species in the Neotropis, which has representatives also in the Nearctic, Palearctic and Australian region. Besides hyloids and ranoids, a number of further neobatrachian families exist. Into this assemblage belong the sooglossids from the Seychelles and southern India, as well as heleophrynids from South Africa, limnodynastids and myobatrachids from Australia and New Guinea and the Batrachophrynidae from South America. While providing a general zoogeographic picture for amphibians or discussing the possible prevalence of vicariance vs. dispersal hypotheses, is clearly beyond the scope of this article, a number of general patterns can still be discerned from the distributions outlined above:

9 Hydrobiologia (2008) 595: (1) At a very fundamental level, the influence of vicariance is very clearly visible in a number of amphibian distributions. Although salamanders and caecilians are certainly limited in their distribution by adaptations to temperate vs. tropical environments, their general patterns of geographic occurrence and the restriction of salamanders to temperate regions of the northern hemisphere make it likely that the basal diversification of salamanders occurred on Laurasia and that of caecilians on Gondwana (Feller & Hedges, 1998). Archaeobatrachians are separated in a number of lineages of alternatingly Laurasian or Gondwanian distribution (Roelants & Bossuyt, 2005). The distribution of basal neobatrachians in the southern hemisphere indicates that they initially had a Gondwanan distribution. And the diversity centres of hyloid vs. ranoid neobatrachian frogs in the New World vs. the Old World (and here, especially Africa) are likely to correspond to the separation of South America and Africa, which also roughly agrees with molecular clock calculations (e.g. San Mauro et al., 2005; Roelants & Bossuyt, 2005). (2) The initial pattern originated by vicariance has been modified extensively by dispersal. The occurrence of some endemic amphibians on oceanic islands is a clear evidence for the possibility of overseas dispersal also in this group. The phylogenetic split of several lineages like the reed frogs, Hyperoliidae, are so young according to molecular clocks that their occurrence in Madagascar and on the Seychelles can only be seen in colonization by ancestors rafting on flotsam over the sea (Vences et al., 2003). The few genera occurring in more than one biogeographic region and continent provide further evidence for the possibility and potential speed of amphibian dispersal. The genus Hoplobatrachus has a number of species in Asia, and has one Afrotropical species that colonized, out of Asia, vast areas of the African savannas in short time spans, as to judge from the low molecular differentiation among Asian and African Hoplobatrachus species (Kosuch et al., 2001). Similar examples can be found in other genera and families as well. The colonization of South America by plethodontid salamanders and ranid frogs, and the colonization of the Palearctis by hylid frogs, almost certainly represent such instances. At the interface between dispersal and vicariance, the hylid subfamilies Pelodryadinae (Australian) and Phyllomedusinae (Neotropical) are sister groups (Hoegg et al., 2004; Frost et al., 2006), and probably are witnesses of a vaster distribution of these tree frogs while South America and Australia were connected over Antarctica in the Early Cenozoic. During this time, probably, the ancestor of these frogs dispersed from South America to Australia, and the two groups evolved in vicariance after the continental connections were severed. (3) A third factor that should not be underestimated is (natural) extinction. There is impressive evidence in current amphibian distributions for formerly larger distribution areas of groups of today relictual occurrence. Among the examples are the relationships between Ascaphus and Leiopelma (Leiopelmatidae), the most basal of the extant anurans, and with two and four species restricted to the Western North America, and to New Zealand respectively. Plethodontid salamanders are today most diverse in the Nearctis and Neotropis, but one genus, Speleomantes, is known from Italy and France. The very recent discovery of the first Asian plethodontid Karsenia by Min et al. (2005) clearly demonstrates that this group had a wider distribution in Asia and Europe before, and probably went extinct over most of its Palearctic distribution area. Again, also the relictual distribution of the basal neobatrachian frogs in southern South America, Australia, South Africa, the Seychelles and India probably witnesses a previous, much wider Gondwanian distribution. Probably, and especially in frogs, successive waves of radiation of more modern groups have largely replaced the more basal groups which survived, if at all, as species-poor relicts in very restricted and fragmented distribution areas. Due to their relatively limited dispersal ability, by far most amphibian genera, and almost all amphibian species, are endemic to single continents or biogeographic units. These units largely correspond to the

10 578 Hydrobiologia (2008) 595: biogeographic regions used here. Some additional subdivisions are obvious (of course not considering introductions): In the Afrotropical region, all species and all genera but one occurring in Madagascar are endemic to the island (with two genera of mantellid frogs also having one species each endemic to the Comoro island of Mayotte); and all genera and species of Seychellean caecilians and frogs are endemic to the archipelago. Sub-Saharan Africa shares no amphibian species with Asia or Europe, and the species-level of endemism of Australia is above 90%. Further islands with a degree of endemism of 100% at the species level are Jamaica, São Tomé and Principe, New Zealand, Fiji and Palau (percentages based on analyses including non-water related amphibian species; see for a more detailed analysis based on the data from the Global Amphibian Assessment). Most of the few genera with distributions extending over more than one of the main zoogeographic regions, all of them frogs, were classical dump bin genera which were recently split into various genera after comprehensive revisions (Faivovich et al., 2005; Frost et al., 2006). For example, the formerly 340 species of Hyla with representatives in the Neotropis, Orientalis, Nearctis and Palearctis have recently been taxonomically revised and were split into 15 genera, with the genus Hyla now being restricted to few species in the Nearctic and Palearctic regions (Faivovich et al. 2005). In addition, only the frog genera Hoplobatrachus (Orientalis with four and Afrotropis with one species) and Ptychadena (one species in Palaearctis as defined herein, all others in Afrotropis) have a distribution across biogeographic regions, and the salamander genus Ambystoma has 15 Nearctic and 14 Neotropical representatives; however, the Neotropical species are restricted to Mexico and the genus did not further disperse into Central or South America. Human related issues: global amphibian declines More new amphibian species are being discovered every year than ever, but at the same time, amphibians are paradoxically declining at a very fast rate (Hanken, 1999). Multi-causal declines have been recorded worldwide. The most obvious and immediate threat to most amphibians in a threatened IUCN category is habitat destruction, and for some species overexploitation (as pets or food) constitutes an imminent danger as well. The Global Amphibian Assessment (Stuart et al., 2004) classified 1856 amphibian species (32.5%) into one of the IUCN threat categories (Vulnerable, Endangered or Critically Endangered), many more than in other groups such as mammals (23%) or birds (12%). About 43% of all species were recorded to experience some sort of population decline. A total of 32 amphibian species have become extinct, and 122 species were considered to be possibly extinct, with no recent sightings. Most alarmingly, so-called enigmatic declines have also been reported from unaltered and largely undisturbed habitats, especially in South America and Australia, but also in North America and Europe (Blaustein et al., 1994). Furthermore, it has been shown that the absence of aquatic larvae in declining anuran populations may significantly alter freshwater ecosystems (Ranvestel et al. 2004). Most likely, emerging infectious diseases, especially chytridiomycosis, play a key role in these declines which in many cases apparently have led to full extinctions already (Daszak et al., 2003). The chytrid fungus Batrachochytrium dendrobatidis especially affects species that are ecologically predisposed in that their natural history (high-altitude occurrence, stream breeding) coincides with the preferences of the pathogen, and if combined with low fecundity and habitat specialization, a species can quickly be driven to extinction (Daszak et al., 2003), and this process may be furthered by climatic change (Pounds et al., 2006). Interestingly, despite high degrees of chytrid infection in the wild, no African frogs have yet been reported to have enigmatically declined, which led Weldon et al. (2004) to hypothesize that the disease may have originated in Africa and spread to other continents by exported clawed frogs, Xenopus, as carriers. The important role of chytridiomycosis in amphibian declines has been asserted, and obvious measures include the control of amphibian introductions into unaffected areas as well as the disinfection of fishing gear and similar equipment by limnologists working on different continents. However, the influence of other agents such as pesticides or increased UV radiation should not be disregarded, and multi-causal

11 Hydrobiologia (2008) 595: hypotheses may well be most powerful to explain declines in some cases. Certainly, the simultaneous dependence of many amphibian species from both aquatic and terrestrial environments make them especially vulnerable to a multitude of factors. The importance of the freshwater environment for the survival of amphibian populations is paramount, and more studies on the specific requirements of amphibians in their aquatic, mostly larval phase are necessary to develop integrated conservation strategies. Acknowledgements We are grateful to Francisco Hita García for his help with updating our amphibian species database, and to Frank Glaw for numerous discussions and comments. References Altig, R. & R. W. McDiarmid, Body plan: development and morphology. In McDiarmid, R. W. & R. Altig (eds), Tadpole: The Biology of Anuran Larvae. University of Chicago Press, Chicago: AmphibiaWeb, Information on Amphibian Biology and Conservation. [web application]. Berkeley, California. AmphibiaWeb. Available: (Accessed, 2005). Avise, J. C., Phylogeography. The History and Formation of Species. Harvard University Press, Cambridge, Massachusetts. Balinsky, J. B., Adaptation of nitrogen metabolism to hyperosmotic environment in Amphibia. Journal of Experimental Zoology 215: Biju, S. D. & F. Bossuyt, New frog family from India reveals an ancient biogeographical link with the Seychelles. Nature 425: Blaustein, A. R., D. B. Wake & W. P. Sousa, Amphibian declines: judging stability, persistence, and susceptibility of populations to local and global extinctions. Conservation Biology 8: Chippindale, P. T., A. S. Baldwin, R. M. Bonett & J. J. Wiens, Phylogenetic evidence for a major reversal of life history evolution in plethodontid salamanders. Evolution 58: Daszak, P., A. A. Cunningham & A. D. Hyatt, Infectious disease and amphibian population declines. Diversity and Distributions 9: Dubois, A., The higher nomenclature of recent amphibians. Alytes 22: Dubois, A., Developmental pathway, speciation and supraspecific taxonomy in amphibians 1. Why are there so many frog species in Sri Lanka. Alytes 22: Duellman, W. E., Amphibian Species of the World: Additions and Corrections, Vol. 21. University of Kansas, Museum of Natural History, Special Publication, pp Duellman, W. E. & D. M. Hillis, Marsupial frogs (Anura: Hylidae: Gastrotheca) of the Ecuadorian Andes: resolution of taxonomic problems and phylogenetic relationships. Herpetologica 43: Duellman, W. E. & L. Trueb, Biology of Amphibians. McGraw-Hill, New York. Faivovich, J., C. F. B. Haddad, P. C. A. Garcia, D. R. Frost, J. A. Campbell & W. C. Wheeler, Systematic review of the frog family Hylidae, with special reference to Hylinae: phylogenetic analysis and taxonomic revision. Bulletin of the American Museum of Natural History 294: Feller, A. E. & S. B. Hedges, Molecular evidence for the early history of living amphibians. Molecular Phylogenetics and Evolution 9: Frost, D. R. (ed.), Amphibian Species of the World. Association of Systematic Collections, Allen Press, Lawrence, Kansas. Frost, D. R., Amphibian Species of the World: an Online Reference. Version 3.0 (22 August, 2004). Electronic Database accessible at American Museum of Natural History, New York, USA. Frost, D. R., T. Grant, J. Faivovich, R. H. Bain, A. Haas, C. F. B. Hadad, R. De Sa, A. Channing, M. Wilkinson, S. C. Donnellan, C. J. Raxworthy, J. A. Campbell, B. L. Blotto, P. Moler, R. C. Drewes, R.A. Nussbaum, J. D. Lynch, D. M. Green & W. C. Wheeler, The amphibia tree of life. Bulletin of the American Museum of Natural History 297: Glaw, F. & J. Köhler, Amphibian species diversity exceeds that of mammals. Herpetological Review 29: Glaw, F. & M. Vences, Phylogeny and genus-level classification of mantellid frogs (Amphibia, Anura). Organisms Diversity & Evolution 6: Hanken, J., Why are there so many new amphibian species when amphibians are declining? Trends in Ecology and Evolution 14: 7 8. Hedges, S. B., C. A. Hass & L. R. Maxson, Caribbean biogeography: molecular evidence for dispersal in west Indian terrestrial vertebrates. Proceedings of the National Academy of Sciences of the USA. 89: Himstedt, W., Die Blindwühlen. Neue Brehm-Bücherei, Vol Hoegg, S, M. Vences, H. Brinkmann & A. Meyer, Phylogeny and comparative substitution rates of frogs inferred from sequences of three nuclear genes. Molecular Biology and Evolution 21: Kosuch, J., M. Vences, A. Dubois, A. Ohler & W. Böhme, Out of Asia: mitochondrial DNA evidence for an oriental origin of tiger frogs, genus Hoplobatrachus. Molecular Phylogenetics and Evolution 21: Köhler, J., D. R. Vietes, R. M. Bonett, F. Hita Garcia, F. Glaw, D. Steinke & M. Vences, New amphibians and global conservation: a boost in species discoveries in a highly endangered vertebrate group. BioScience 55: Measey, G. J., M. Vences, R. C. Drewes, Y. Chiari, M. Melo & B. Bourles, Freshwater paths into the ocean: molecular phylogeny of the frog Ptychadena newtoni

12 580 Hydrobiologia (2008) 595: gives insights into amphibian colonization of oceanic islands. Journal of Biogeography 34: Meyer, A. & R. Zardoya, Recent advances in the (molecular) phylogeny of vertebrates. Annual Reviews of Ecology, Evolution and Systematics 34: Min, M. S., S. Y. Yang, R. M. Bonett, D. R. Vieites, R. A. Brandon & D. B. Wake, Discovery of the first Asian plethodontid salamander. Nature 435: Moodie, G. E. E., Observations on the life history of the caecilian Typhlonectes compressicaudus (Duméril & Bibron) in the Amazon basin. Canadian Journal of Zoology 56: Pounds, J. A., M. R. Bustamante, L. A. Coloma, J. A. Consuegra, M. P. L. Fogden, P. N. Foster, E. LaMarca, K. L. Masters, A. Merino-Viteri, R. Puschendorf, S. R. Ron, G. A. Sánchez-Azofeifa, C. J. Still & B. E. Young, Widespread amphibian extinctions from epidemic disease driven by global warming. Nature 439: Ranvestel, A. W., K. R. Lips, C. M. Pringle, M. R. Whiles & R. J. Bixby, Neotropical tadpoles influence stream benthos: evidence for the ecological consequences of decline in amphibian populations. Freshwater Biology 49: Roelants, K. & F. Bossuyt, Archaeobatrachian paraphyly and Pangaean diversification of crown-group frogs. Systematic Biology 54: San Mauro, D., M. Vences, M. Alcobendas, R. Zardoya & A. Meyer, Initial diversification of living amphibians predated the breakup of Pangaea. The American Naturalist 165: Stuart, S. N., J. S. Chanson, N. A. Cox, B. E. Young, A. S. L. Rodriguez, D. L. Fishman & R. W. Waller, Status and trends of amphibian declines and extinctions worldwide. Science 306: Vences, M. & F. Glaw, Systematic review and molecular phylogenetic relationships of the direct developing Malagasy anurans of the Mantidactylus asper group (Amphibia, Mantellidae). Alytes 19: Vences, M., J. Kosuch, M. -O. Rödel, S. Lötters, A. Channing, F. Glaw & W. Böhme, Phylogeography of Ptychadena mascareniensis suggests transoceanic dispersal in a widespread African-Malagasy frog lineage. Journal of Biogeography 31: Vences, M., M. Thomas, A. van der Meijden, Y. Chiari & D. R. Vieites, 2005a. Comparative performance of the 16S rrna gene in DNA barcoding of amphibians. Frontiers in Zoology 2: article 5. Vences, M., M. Thomas, R. M. Bonett & D. R. Vieites, 2005b. Deciphering amphibian diversity through DNA barcoding: chances and challenges. Philosophical Transactions of the Royal Society London Series B 360: Vences M., D. R. Vieites, F. Glaw, H. Brinkmann, J. Kosuch, M. Veith & A. Meyer, Multiple overseas dispersal in amphibians. Proceedings of the Royal Society of London Series B 270: Wake, M. H., The reproductive biology of caecilians: an evolutionary perspective. In Taylor, D. H. & S. I. Guttman (eds), Reproductive Biology of Amphibians. Plemum Press, New York: Wake, M. H., Phylogenesis of direct development and viviparity in vertebrates. In Wake, D. B. & G. Roth (eds), Complex Organismal Functions: Integration and Evolution in Vertebrates. John Wiley & Sons Ltd.: Weldon, C., L. H. du Preez, A. D. Hyatt, R. Muller & R. Speare, The origin of the amphibian chytrid fungus. Emerging Infectious Diseases 10: Wilson, A. C., L. R. Maxson & V. M. Sarich, Two types of molecular evolution. Evidence from studies of interspecific hybridization. Proceedings of the National Academy of Sciences of the USA 71: