Mada-gascan poison frogs (Mantella) and their skin alkaloids

Marta Andriantsiferana

A total of 232 alkaloids, representing 21 structural classes were detected in skin extracts from the dendrobatid poison frog Oophaga pumilio, collected from 53 different populations from over 30 years of research. The highly toxic pumiliotoxins and allopumiliotoxins, along with 5,8-disubstitiuted and 5,6,8-trisubstituted indolizidines, all of which are proposed to be of dietary mite origin, were common constituents in most extracts. One decahydroquinoline (DHQ), previously shown be of ant origin, occurred in many extracts often as a major alkaloid, while other DHQs occurred rather infrequently. Histrionicotoxins, thought to be of ant origin, did not appear to possess a specific pattern of occurrence among the populations, but when present, were usually found as major components. Certain 3,5-disubstituted pyrrolizidines and indolizidines, known to be of ant origin, did occur in extracts, but infrequently. Alkaloid composition differed with regard to geographic location of frog popula...

Abstract: We studied the diet of a population of free-ranging Mantella aurantiaca, an alkaloid-containing poison frog from Madagascar. As in other poison frogs, this species is thought to sequester alkaloids from arthropod prey. Among prey, mites and ants are known to regularly contain alkaloids and mites appear to be a major source of dietary alkaloids in poison frogs. We predicted that mites and ants would constitute the most important prey item for these frogs.

Poison frogs acquire chemical defenses from the environment for protection against potential predators. These defensive chemicals are lipophilic alkaloid toxins that are sequestered by poison frogs from dietary arthropods and stored in skin glands. Despite decades of research focusing on identifying poison frog toxins, we know relatively little about how environmental variation and subsequent arthropod availability impacts toxicity in poison frogs. We investigated how seasonal environmental variation influences poison frog toxin profiles through changes in the diet of the Climbing Mantella (Mantella laevigata). We collected M. laevigata females on the Nosy Mangabe island reserve in Madagascar during the wet and dry seasons and tested the hypothesis that seasonal differences in rainfall is associated with changes in the diet and skin toxin profiles of M. laevigata. The arthropod diet of each frog was characterized into five groups (i.e. ants, termites, mites, insect larvae, or other)...

AMERICAN MUSEUM Novtautes PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 3177, 34 pp., 12 figures August 23, 1996 Madagascan Poison Frogs (Mantella) and Their Skin Alkaloids JOHN W. DALY,1 NIRINA RABE ANDRIAMAHARAVO,2 MARTA ANDRIANTSIFERANA,2 AND CHARLES W. MYERS3 CONTENTS Abstract ....................................... Introduction ....................................... Subfamily Mantellinae Laurent, 1946 ....................................... .................... Genus Mantella Boulenger, 1882 ................... Mantella aurantiaca Mocquard, 1900 ....................................... ....................................... Mantella baroni Boulenger, 1888 . Mantella betsileo (Grandidier, 1872) ....................................... ....................................... Mantella cowanii Boulenger, 1882 . Mantella crocea Pintak and B6hme, 1990 ...................................... Mantella expectata Busse and Bohme, 1992 ..................................... Mantella laevigata Methuen and Hewitt, 1913 .................................. .................... Mantella pulchra Parker, 1925 ................... Mantella viridis Pintak and B6hme, 1988 ...................................... Summary of Alkaloids ....................................... Acknowledgments ....................................... .................... Appendix 1: Voucher Specimens ................... Appendix 2: Alkaloids Identified in Mantella Skin .................... ............ References ....................................... 2 2 3 7 13 17 19 20 21 21 22 22 24 24 28 28 30 31 ' Research Associate in Herpetology, American Museum of Natural History; Chief, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health. 2 Laboratoire de Chimie Organique "Product Naturals," Universite d'Antananarivo, Antananarivo, 1001, Mad- agascar. 3 Chairman and Curator, Department of Herpetology and Ichthyology, American Museum of Natural History. Copyright K American Museum of Natural History 1996 ISSN 0003-0082 / Price $6.00 2 AMERICAN MUSEUM NOVITATES NO. 3177 ABSTRACT Mantella is demonstrably convergent with aposematic Neotropical poison frogs of the family Dendrobatidae in presence of a variety of similar skin alkaloids, and variously convergent in a few other traits. Mantella cowanii is the best example of color pattern convergence to a Dendrobates. Inasmuch as resemblances are primarily to derived features of phylogenetically advanced dendrobatids, the old notion ofrelationship, although intriguing, is not supported by current understanding. But the phylogenetic placement of Mantella is nonetheless unclear and even its relationship to other "mantellines" requires corroboration. Alkaloid profiles in skin extracts of 9 of 11 recognized species of Mantella are reported. Levels of alkaloids ranged from relatively low (M. betsileo, M. expectata, M. laevigata) to moderate (M. aurantiaca, M. crocea, M. pulchra) to high (M. baroni (madagascariensis, auctorum), M. cowanii, M. viridis). However, there was variation in levels for different populations of the same species. The major alkaloids in most species and populations were of the pumiliotoxin class, consisting mainly of pumiliotoxins and allopumiliotoxins, although homopumiliotoxins and other pumiliotoxin-class alkaloids also occurred. A decahydroquinoline and related "dimeric" alkaloids were prominent in certain populations of M. betsileo, M. pulchra, and M. laevigata, but were absent in other populations and in other species. "Izidine" alkaloids, namely disubstituted pyrrolizidines, indolizidines, and quinolizidines, were prominent in all populations of M. baroni, but were usually relatively minor or absent in other species. Although the monophyly of Mantella is supported by presence of lipophilic skin alkaloids, these compounds may be sequestered from arthropod prey and, if so, are unlikely to be phylogenetically informative within the genus. Interspecific and interpopulational differences in alkaloid profiles may be a reflection of dietary differences correlated with habitat diversity. Mantella habitats include upland swamp forest (M. aurantiaca, M. crocea), boggy areas within forest (M. pulchra), upland riverine forest (M. baroni), both disturbed and relatively undisturbed lowland forest (M. betsileo), and semiarid streambeds (M. expectata, M. viridis). Although most species are terrestrial, M. laevigata is semiarboreal in forest and in bamboo groves. Excluding published conjectures based on frogs without provenance (from commercial dealers), assumptions of extraordinary intraspecific or intrapopulational variability in Mantella color patterns have proven unfounded (unlike the situation in Dendrobates), and most species are readily identified in life by coloration alone. The name Mantella baroni Boulenger, 1888, is resurrected from synonymy for a widespread upland species currently known as M. madagascariensis (Grandidier, 1872)-a nomen dubium not certainly applicable to any known species (but conceivably based on M. pulchra Parker, 1925). Use of the name M. betsileo (Grandidier, 1872) for a widespread lowland species is accepted with reservation because the type locality falls outside the known distribution and habitat. INTRODUCTION All amphibians have cutaneous granular mary in Myers et al., 1995: 17), with over ("poison") glands, that in most species se- 200 lipophilic "dendrobatid alkaloids" havcrete diverse pharmacologically active com- ing now been reported (Daly et al., 1987,1993: pounds -including lipophilic alkaloids, which 276). Prior to 1984, most lipophilic frog alkahave a peculiar taxonomic distribution in amphibian skin (Daly et al., 1 978, 1987, 1993: loids were known in nature only from dendrobatids (e.g., Daly et al., 1978) until the 276). The first amphibian lipophilic alkaloids to discovery of some in the South American be chemically elucidated, the samandarines bufonid genus Melanophryniscus (Daly et al., 1984; Garraffo et al., 1993a), in Australian from the European fire salamander (Sch6pf, 1961), remain known only from the Old myobatrachids of the genus Pseudophryne World genus Salamandra (for comment on (Daly et al., 1984, 1990), and in presumptive ranids, rhacophorids, or mantellids of the a contrary report, see Daly et al., 1993: 198). Different lipophilic alkaloids were subse- Madagascan genus Mantella (Daly et al., quently found to characterize a monophyletic 1984; Garraffo et al., 1993b). Lipophilic algroup of several genera of poison frogs within kaloids have not been detected in skin exthe Neotropical family Dendrobatidae (sum- tracts from some 70 other genera of amphib- 1 996 DALY ET AL.: MADAGASCAN POISON FROGS ians. A few structurally distinct classes of alkaloids-batrachotoxins, epibatidines, gephyrotoxins, and histrionicotoxins-are still not known to occur naturally in amphibians other than dendrobatids (but see comment under Mantella pulchra on histrionicotoxins in a pet-trade specimen). Homobatrachotoxin, however, was recently identified in skin and feathers of a toxic bird from Papua New Guinea (Dumbacher et al., 1992). Although they represent four families, anurans with lipophilic skin alkaloids share a number of attributes. All are small (1-5 cm SVL), primarily terrestrial frogs that lay terrestrial eggs. The great majority are brilliantly colored (and often boldly patterned) and all except the Australian Pseudophryne are diurnal. Many are toothless. Lipophilic skin alkaloids appear to function defensively and the ability to synthesize and/or uptake and sequester them seems to have evolved at least five times in the Amphibia. The samandarine alkaloids are currently thought to be synthesized by the salamanders (comment in Daly et al., 1993: 199), and the possibility that dendrobatid frogs also synthesize some of their diverse alkaloids cannot be disproved at this time. Nonetheless, evidence is accumulating that many "dendrobatid alkaloids" are sequestered from some of the small arthropods on which they feed. This discovery originated in the observation that skin extracts of captiveraised dendrobatids contained no alkaloids (Daly et al., 1980, 1992). It was then demonstrated that alkaloids fed to captive-raised frogs were effectively sequestered into skin (Daly et al., 1994a). The ability to sequester alkaloids is absent in the dendrobatid Colostethus, which lacks lipophilic alkaloids in the wild. Finally, feeding leaf-litter insects to dendrobatids being raised in terraria led to establishment of significant levels of many, but not all, of the alkaloids present in wildcaught frogs from the leaf-litter locality (Daly et al., 1994b). What arthropods might be the source of alkaloids not detected in dendrobatids that were fed leaf-litter insects is unknown. Indeed, it is still possible that such alkaloids (pumiliotoxins, histrionicotoxins, decahydroquinolines) may be synthesized by the anuran in the wild, but not when raised in captivity. 3 Sequestering of lipophilic alkaloids provided in the frogs' diet has recently been demonstrated for Mantella, which resembles dendrobatids in that frogs reared in captivity on Collembola, Drosophila, and Acheta do not contain detectable alkaloids (Daly et al., in manuscript). Although it is not known whether the other non-dendrobatid genera (Melanophryniscus and Pseudophryne) with skin alkaloids have similar sequestering systems, it is now conceivable that many, if not, all frog alkaloids are "second-hand" chemicals that are not manufactured by the frogs themselves. This helps explain some, albeit not all, otherwise puzzling instances of interand intraspecific variation in skin toxins, and gives a clearer perspective on the taxonomic usefulness of toxins. Myers et al. (1995), in a preliminary assessment of the systematic implications, pointed out that the underlying genetic mechanisms for alkaloid sequestering provide useful synapomorphies, but that differences in alkaloid profiles of related species may for the most part reflect dietary differences. Profiles of lipophilic skin alkaloids differ remarkably in different species and populations of frogs containing them (Daly et al., 1984, 1987, 1990; Garraffo et al., 1993a, 1993b; Myers et al., 1995). If sequestering systems and diet are the prime determinants, then alkaloid profiles ought to show strong correlations with habitats and kinds of prey. With this in mind, the occurrence of skin alkaloids in nine species of Mantella are presented here in conjunction with general observations on habitats. Specifics of the diet of Mantella spp. and of possible interspecific differences in the genetic base of their sequestering systems remain open questions. SUBFAMILY MANTELLINAE LAURENT, 1946 The correct familial placement of the Mantellinae is uncertain, as discussed below under Mantella. The content of the group varies somewhat according to author, but always includes at least the genera Mantella and Mantidactylus (s.l.). Noble (1931, and in Noble and Parker, 1926: 4n) proposed a close relationship between these frogs on the basis of certain anatomical similarities and a Madagascan distribution. Laurent (1946) erected a subfamily to include the Madagascan gen- 4 AMERICAN MUSEUM NOVITATES NO. 3177 M. viridis (9E) M. laevigata (7F) M. betsileo (6C) M. betsileo (6D), M. Iaevigata (7DE), M. pulchra (9AB) M. baroni (8D), M. pulchra (9C) i(7AB), M. baroni (8B) 1. baroni (8C) M. baroni (8E) M. baroni (8F) M. expectata (7C) Tolagnaro 0 -. L- 100 Km - - Fig. 1. Madagascar, showing collecting localities for Mantella spp. discussed in this paper. The numberAetter combinations within parentheses match the population designations in figures 6-9 (gas chromatograms) and in appendix 2. era Mantella, Mantidactylus, Gephyroman- tis, Trachymantis (= Laurentomantis), and the Sri Lankan Pseudophilautus. Except to designate Mantella as type genus of the Mantellinae, Laurent's work (1 943a, 1 943b, 1946) makes no mention of the genus and one assumes that he accepted at face value Noble's proposal that it is related to Mantidactylus. Since then, a few authors (notably Liem, 1970, and Guibe, 1978) have not rec- 1 996 DALY ET AL.: MADAGASCAN POISON FROGS ognized the subfamily but neither have they been concerned with the relationships of Mantella itself. Otherwise, except for doubts concerning the inclusion of Pseudophilautus (see Frost, 1985: 439), no one seems to have explicitly questioned the naturalness of the Mantellinae, which has been uncritically accepted as a unit for various phylogenetic analyses (see below). Half a century after Noble first proposed the relationship, BlommersSchlosser (1979: 65) asserted that: The genus Mantella is very closely related to the genus Mantidactylus (especially to the wittei, depressiceps and pulcher groups, see Table IV for comparison). The cytogenetic data agree with those of these three species groups (cf. Blommers-Schl6sser, 1978). The tadpole resembles those of the wittei and depressiceps groups (cf. Arnoult, 1966). The lateral metatarsalia are entirely connected. The omosternum is widely forked posteriorly and the stemum is slightly forked anteriorly in Mantella cowanji (cf. Guibe, 1978) and the femoral glands are diffuse. In habitus and general color pattern, some Mantella are also reminiscent of the several bicolored species of the Mantidactylus albofrenatus species group (see p. 160 and color photographs 11 1-1 13 in Glaw and Vences, 1994). In fact, it is not clear whether Mantidactylus (s.l.) possesses any synapomorphies that would exclude the monophyletic Mantella. Nonetheless, the significance of any phenotypic similarity remains to be elaborated. Convincing synapomorphies still have not been presented to link Mantella with any other genus and the monophyly of the Mantellinae must simply be questioned as a matter of course. The only explicit mantelline synapomorphy proposed by Blommers-Schlosser (1993: 210, 212, fig. 2) is a loosely defined behavioral trait: Abbreviated mating contact ... no real amplexus; the male covers the head and shoulders of the female with his thighs (femoral glands) in arboreal species or embraces the female very loosely, inguinally or axillary, in ground dwelling species, which lasts very short, from seconds to a few minutes. This behavior is known only from the Mantellinae. But, as noted by Blommers-Schlosser (1993: 210), femoral glands occur in at least a half-dozen [petropedetine and ranine] genera other than mantellines, and one might 5 suspect that the glands correlate with similar mating behavior at least in some cases. Femoral glands, shown as a derived character in two places on her cladogram (her fig. 2, with character 28 as part of nodes D and J), were considered to be a primitive feature in the Mantellinae (Blommers-Schlosser, 1993: 212), although the supposedly correlated mating behavior was proposed as a synapomorphy for the subfamily. Femoral glands in Mantella have not been figured or described to our knowledge. Blommers-Schl6sser (1979: 65) said that the glands are diffiuse and Blommers-Schl6sser and Blanc (1991: 263) indicated that femoral glands are a male character in Mantella. Glaw and Vences (1994: 70) indicated that the glands are indistinct or absent in male Mantella and absent in females. Cursory examination of some of our voucher material (appendix 1) suggests that the purported glands are coexistent with the patch of granular skin on the underside of the thigh; the granular patch occurs in both males and females, but sexual dimorphism seems evident only in M. baroni and M. betsileo. In female baroni, the granular thigh patch is pale brownish, whereas it is paler and has a more sharply defined perimeter in males. In female betsileo, the patch is black and consists of large angular "granules," whereas it is pale brownish and smoother in the males. In these species, male and female specimens could be separated by the naked eye on the basis of differences in thigh granulation. There were no pigmentation differences in a few M. crocea, although granulation is perhaps stronger in the female, and there were no obvious sexual differences in M. pulchra, M. expectata, M. /aevigata, and M. viridis (unskinned males were not available for M. aurantiaca or M. cowan/i). However, in addition to species differences, there is notable individual variation in the granular thigh area at least in M. expectata. Some male and female expectata have the ventral thigh patch appearing "normally" granular, whereas others have deep parallel grooving in a superficially smoother patch. Strongly granular skin underneath the thigh is commonplace among frogs, but this area seems exceptionally variable in Mantella. Femoral "glands" in Mantella, when they occur (e.g., in betsileo and baroni as described AMERICAN MUSEUM NOVITATES 6 A B C D NO. 3177 Fig. 2. A, B. Mantella aurantiaca from the Andasibe region; orange and red variants from different demes (AMNH 136892, 2 23 mm SVL, and 136921, 9 22 mm SVL, respectively). C. Mantella baroni (M. madagascariensis, auctorum), from An'Ala (AMNH 140556, 8 24 mm SVL). D. Mantella betsileo from Ambavala (AMNH 140573, 2 23 mm SVL). above), bear little superficial resemblance to the often strikingly well-defined glands in Mantidactylus (s.l.), some of which have a central pore,4 and the homology might therefore be questioned. However, BlommersSchlosser (1979: 35, 44, 65) indicated a similarity in stating that some Mantidactylus have the femoral glands "diffused" or not always 4Blommers-Schlosser's (1979: 66) schematic draw- ings of femoral gland variation in Mantidactylus (s.l.) are variously misleading in either not explicitly distinguishing the glands from the area of coarsely granular skin on the underside of the thigh or in implying the absence or restriction of such granular skin. visible. Glaw and Vences (1994: 70) tabulated presence-absence and distribution by sex of the glands in the subgenera or species groups of Mantidactylus, and histological comparisons with Mantella would be useful. The possibility of seasonal variation should also be investigated. Alkaloids were lacking in three species of Mantidactylus (Garraffo et al., 1 993b: 1017), although present in the nine species of Mantella examined. The possibility that Mantidactylus (s.l.) is paraphyletic with respect to Mantella is mentioned above, even though no one has made a convincing case for relating these genera. We neither assume nor DALY ET AL.: MADAGASCAN POISON FROGS 1996 B A C 7 F D Fig. 3. A. Mantella cowanii from unknown locality (AMNH 140550, 2 26 mm SVL). B. Mantella crocea from 14-18 km N Andasibe (AMNH 136897, adult 8, 17.5 mm SVL). C. M. expectata from Massif Isalo (AMNH 136938, 9 26 mm SVL). D. Mantella laevigata from Ambavala (AMNH 140564, 9 24 mm SVL). reject the monophyly of the Mantellinae in this paper, which focuses on Mantella without further reference to Mantidactylus (s.l.) or other putative mantelline genera. Mantella probably is monophyletic as defined by Guibe (1978: 81-82) and Blommers-Schl6sser (1979: 61), and we here suggest that the genetic mechanism leading to presence of diverse lipophilic alkaloids in the skin (whether from sequestering or biosynthesis) is an additional synapomorphy for the genus. But the aposematic frogs of the Neotropical Dendrobatidae are similarly defined (e.g., Myers et al., 1995) and so the questions of relatedness or convergence are revisited below. Genus Mantella Boulenger, 1882 The three oldest species names assigned to this genus were first placed in Dendrobates (betsileo and madagascariensis of Grandidier, 1872, and ebenaui Boettger, 1880). Boulenger (1882: 141) erected the genus Mantella in the family Dendrobatidae only for these three nominal species, basing his concept primarily on available specimens of M. betsileo (which, therefore, was appropriately designated as type species by Liem, 1970: 100). Boulenger (1882: 471-472) named Mantella cowanii as an addendum to his new genus. Although the placement within the Den- 8 AMERICAN MUSEUM NOVITATES drobatidae was followed by a few workers into the 20th century (e.g., Werner, 1901; Mocquard, 1909: 65; Hewitt, 1911), Mantella subsequently became aligned not with New World dendrobatids but with Old World "tree frogs." Noble (1931: 524-526) allied Mantella with other Madagascan genera that he placed in the redefined Polypedatidae (= Rhacophoridae), concluding that "Mantella may be considered a terrestrial tree frog, for its pads, although small, agree with those of Polypedates." Laurent (1946) broke away from this view by assigning Mantella to the Ranidae, which he later (1951) expanded to comprise additional subfamilies (including the Rhacophorinae). Since then, Mantella (alone or as part of the Mantellinae) has been shuffled about between the Ranidae (e.g., Guibe, 1964; Blommers-Schlosser, 1979; Dubois, 1984) and the Rhacophoridae (e.g., Liem, 1970; Guibe, 1978; Channing, 1989), and the Mantellinae also have been elevated to familial rank (Blommers-Schl6sser and Blanc, 1991). More recent analyses include those of Blommers-Schl6sser (1993), who returned to Laurent's (1951) subfamilial classification and explicitly treated the Mantellinae and Rhacophorinae as related subfamilies (sister groups) within the Ranidae, and Ford (e.g., 1993: fig. 3), who considered the Ranidae as being nonmonophyletic and regarded the mantellines as being nested within the family Rhacophoridae. To summarize, the mantellines have been on a kind of taxonomic seesaw but have not fallen off definitively on one side or the other. One ofthe few things implicit in taxonomic discourse of the last six decades is that mantellines are not closely related to dendrobatids, even though general resemblances between Mantella and Dendrobates still receive comment. Cited anatomical differences include external and internal structure of the digits (supradigital scutes in dendrobatids, and intercalary cartilages and Y-shaped terminal phalanges in Mantella), and thigh musculature (distal tendon of the m. semitendinous pierces that of the m. gracilis complex prior to insertion in dendrobatids but inserts deeper than the m. gracilis in Mantella). The aforesaid dendrobatid character states are NO. 3177 synapomorphic compared with the more widely distributed conditions in Mantella. Despite some striking similarities in aposematic colorations (see below), coloration does not provide any obvious support for hypothesizing relationship between dendrobatids and Mantella. The most plesiomorphic aposematic dendrobatids are striped, whereas the least colorful species of Mantella are bicolor, with brown to green dorsa and darker brown or blackish sides. Mantella does approach a correlated condition of the tympanum and jaw musculature that is considered synapomorphic for the Dendrobatidae. In dendrobatids always and in Mantella frequently, the posterodorsal (and, in dendrobatids, sometimes dorsal) part of the tympanum is externally concealed. In Mantella, the partial tympanic concealment is externally correlated with a weak to strong supratympanic fold, which largely follows the underlying anterior edge of the m. depressor mandibulae. In dendrobatids, a similarly aligned supratympanic fold varies from primitively present (e.g., in Aromobates) to a vague swelling (e.g., in Epipedobates trivittatus) to completely absent (e.g., in Dendrobates pumilio). But whether the supratympanic fold is present or absent, the partial concealment of the tympanum ultimately reflects the fact that the large external slip' of the m. depressor mandibulae partially overlaps the tympanum in dendrobatids (Myers et al., 1991: fig. 5), and, similarly overlaps or else skirts the edge of the tympanic ring in Mantella. The usual condition in Mantella (i.e., in the nine species in appendix 1) is for s The m. depressor mandibulae is similarly configured in dendrobatids and Mantella. A large superficial slip originates broadly from the dorsal fascia, completely concealing a smaller, deeper slip originating from the long otic ramus of the squamosal bone. In Mantella, a third slip, not concealed ventrally, originates from the posterior and posteroventral margin of the tympanic ring. In dendrobatids, the third slip is more variable and often poorly defined, sometimes virtually absent or not separable from the slip that originates mainly from the otic ramus (e.g., Myers et al., 1984: 8). Such divisions of the m. depressor mandibulae occur in certain other anurans, but the overlapping of the tympanum by the largest slip may be less common. DALY ET AL.: MADAGASCAN POISON FROGS 1996 A 9 .2 ffL. -x B C Fig.4. A, B. Mantellapulchra from An'Ala (AMNH 136907,621 mm SVL,136908,922 mm SVL). C. Mantella viridis from Montagne des Fran9ais (AMNH 140580, 2 26 mm SVL). AMERICAN MUSEUM NOVITATES 10 NO. 3177 ov -io 0000,. Fig. 5. Comparisons in dorsal and ventral views of Mantella baroni (AMNH 140556) on left side, and Mantella pulchra (AMNH 140552) on right. Both from 2 km SW An'Ala. These specimens (d left, Q right) are equivalent in size (24 mm SVL), but, taking sexual dimorphism into account, M. baroni is the larger species. 1 996 DALY ET AL.: MADAGASCAN POISON FROGS the depressor muscle to slightly overlap the posterodorsal part of the tympanic ring, but there is no overlap at all in several skinned carcasses ofM. pulchra, in which any external concealment is due simply to muscle bulk raising the skin away from the tympanum. Another, even more striking similarity between Mantella and the "advanced" (aposematic) dendrobatids is of course the subject of this paper-the shared presence of lipophilic alkaloids. This trait, however, also is shared with Australian Pseudophryne and South American Melanophryniscus, as first reported by Daly et al. in 1984. Maxson and Myers (1985: 54) noted that dendrobatid antisera showed no cross reactivity with Mantella albumins or Pseudophryne antisera, which was taken to imply that "any phylogenetic association could be no more recent than some 100-120 million years ago." Clearly, much remains to be learned about the comparative anatomy and phylogenetic relationships of major anuran groups. Nonetheless, present evidence suggests that aposematic dendrobatids and Mantella are monophyletic groups that are derived relative to other dendrobatids and other purported mantellines. All the similarities therefore may be postulated as striking examples of convergence beyond ranoid (or hyloid and ranoid) synapomorphies. More than 60 years ago, G. K. Noble penned the following note to himself: It is queer that loss of teeth often runs to iridescent colors. Note Dendrobates and Mantella. (From Genera Salientia [n.d.], a bound volume of typed literature excerpts and hand-written notes, in the AMNH Dept. Herpetology library.) We can now suggest a few pieces to Noble's puzzle (but in recognition of widespread toothlessness, in dull toads for example, we should reverse his statement-iridescent colors often run to loss of teeth). The bright hues are warning colors advertising the existence of noxious skin alkaloids. Loss of teeth in these cases probably reflects specialization on very small prey, some of which may provide the alkaloids sequestered and accumulated by the frogs. It is an ecological niche entered more than once. In color patterns, as well as in accumula- I1I tion of skin alkaloids, Mantella is convergent with aposematic dendrobatids in a general way. Thus, the unicolored Mantella aurantiaca (fig. 2A) is comparable to a few Phyllobates (although the ancestral patterns are very different). Bright flash marks in concealed parts of the limbs of some Mantella have their counterparts not only in the dendrobatid genus Epipedobates but in a variety of other frogs as well. As long recognized, some Mantella are convergent mainly with Dendrobates because the iridescent coloring is displayed as rounded markings and even limb bracelets. Mantella cowanii (fig. 3A), which resembles some populations of Dendrobates histrionicus, is the best example of a Dendrobates-like pattern. Species of Dendrobates, however, are highly variable in color pattern both within and among populations. Such variation has not yet been clearly demonstrated for any Mantella (see below). Some species of Dendrobates (e.g., Panamanian populations of D. pumilio) do not uniformly occupy available habitat but occur as dense populations of well-separated demes. It is an intriguing possibility that at least Mantella baroni may exhibit similar population structure (see observation by John Cadle on page 18). If so, one might also expect similar territorial behavior (which might also correlate with lack of pronounced sexual dimorphism in size). The number of recognized species of Mantella has about tripled in the last decade (since Frost, 1985), in part due to discovery of new species concurrent with increased fieldwork in Madagascar, and in part due to resurrection of valid species names from synonymy. Much of the taxonomic confusion has involved belief that at least one species is extraordinarily variable (like some Dendrobates)-a concept for which the name Mantella "madagascariensis," has been most recently used, even though Guibe (1964) set this name aside as being unattributable. Glaw and Vences (1992a, 1994) seem to have renewed the belief that extreme variability in color pattern is expected in single populations of Mantella, first (1 992a: 166) in their announcement of a "variable colour morph," and then by resurrection (in quotation marks) of the unused name Mantella NO. 3177 AMERICAN MUSEUM NOVITATES 12 M. aurantiaca (B) 3231B 3A 392 323C/ 235C 265F 434 (D) 211D I 247C | 251 D / 384A 384B M. betsileo (E) ID 307G I Fig. 6. Gas chromatographic traces showing alkaloid profiles from Mantella spp. A. M. aurantiaca (10 skins, approx. 14-18 km N Andasibe, upland swamp forest, Nov. 1989). B. M. aurantiaca (5 skins, same region and habitat as preceding, Jan. 1993). C. M. betsileo (11 skins, Antanambaobe, inland clove tree forest, Dec. 1990). D. M. betsileo (8 skins, Ambavala, remnant forest around a bamboo grove, Jan. 1994). E. M. betsileo (12 skins, Farakaraina, disturbed coastal forest, Dec. 1993). F. M. cowanii (3 skins, from dealer, Dec. 1993). Boldface designations of alkaloids match those in appendix 2. The OV-1 column is programmed to 280°C at 10°C per minute from an initial temperature of 150°C; other conditions are as given in Daly 1996 DALY ET AL.: MADAGASCAN POISON FROGS loppei (from Roux, 1935) for some highly diverse specimens (1994: 185, black and white figs. 335-337 and color photos 58-60). But most, if not all, of these seem to be frogs from the animal trade and there is no telling how many populations or demes or species might be represented-nor can the possibility of captive-produced hybrids be ruled out a priori. Commercial and other exportation of Mantella (see Glaw and Vences, 1994: 31326), especially to Europe, has beneficially stimulated much of the recent work on these frogs but also is introducing a new kind of confusion into the literature. Localities, if available at all, are sometimes based on second- or third-hand information that is automatically suspect until verified in the field, and we especially urge authors to indicate clearly any locality data provided by dealers (who understandably might be unwilling to share their sources of income). The number of species of Mantella will increase with investigation of possibly unnamed species (e.g., see mention under M. betsileo and M. laevigata) and discovery of others. But, excluding Mantella loppei (see above), the 11 named species of Mantella that seem clearly recognizable at this time are as follow: Mantella aurantiaca Mocquard, 1900 Mantella baroni Boulenger, 1888 (M. madagascariensis, auctorum) Mantella bernhardi Vences et al., 1994 Mantella betsileo (Grandidier, 1872) Mantella cowanii Boulenger, 1882 Mantella crocea Pintak and Bohme, 1990 Mantella expectata Busse and Bohme, 1992 Mantella haraldmeieri Busse, 1981 Mantella laevigata Methuen and Hewitt, 1913 Mantella pulchra Parker, 1925 Mantella viridis Pintak and B6hme, 1988 6 We agree with these authors that controlled exportation and especially habitat management are probably better conservation solutions for Mantella than attempted "protection" through CITES listing, which often has the unintentional effect of making research so difficult as to be impractical. 13 All these have been investigated for skin alkaloids except Mantella bernhardi and M. haraldmeieri, which are both from southeastern Madagascar. The nine species sampled are shown in figures 2-5 and the localities sampled are mapped in figure 1. The following accounts summarize field observations7 and results ofalkaloid analyses for each of the nine species. Gas chromatographic profiles are shown in figures 6-9. The occurrence of individual alkaloids of various classes is presented in appendix 2 and the structures of representative alkaloids are shown in figures 10-12. Mantella aurantiaca Mocquard, 1900 Figure 2A, B This easily recognized unicolored frog appears to occupy a very small range in the uplands of east-central Madagascar, where its distribution correlates with wet montane forests (Glaw and Vences, 1992a, 1994; Zimmerman et al., 1990). Specimens used for alkaloid analysis were collected with the help of local people in upland swamp-forest habitats in the region north of Andasibe. The widely disjunct swamp forest localities (about 1000 m elev.) were separated mainly by cultivated areas comprising open fields and forest of pine or eucalyptus. The swamp forests themselves contained varying amounts of standing water, grassy hillocks and a variety of trees, including Pandanus. Frogs were active throughout such habitat. The general area is located in the region of the type locality, "une foret entre Beforona et Moramanga" (Mocquard, 1900: 111). Demes from different swamp forests differed somewhat in coloration, varying from yellow through orange to red, any given color being nearly uniform dorsally and ventrally (including palms and soles), with a red calf spot in the concealed part of the shank that 7Fieldwork was accomplished by the first two authors and local guides, with help at times from the third author and A. Rabemanantsoa. et al. (1993). Emergent temperatures can differ somewhat with different columns and variations in flow rates. Some typical emergent temperatures for some common mantelline alkaloids are: 217B 1660, 251D 1720, 267C 1900, 307A 2160, 323A 2300. . AMERICAN MUSEUM NOVITATES 14 M. crocea (A) M. crocea (B) 307A NO. 3177 307A 323B I - M. expectata (C) M. Iaevigata (D) 195C>195A 307A 323D M. laevigata (E) L193D 307G M. laevigata (F) 195C, 195A, 193D V 249A 30,17G Fig. 7. Gas chromatographic traces showing alkaloid profiles from Mantella spp.; conditions as in figure 6. A. M. crocea (10 skins, approx. 14-18 km N Andasibe, upland swamp forest, Nov. 1989). B. M. crocea (35 skins, same region and habitat as preceding, Jan. 1993). C. M. expectata (20 skins, Massif Isalo, semiarid streambed, Jan. 1993). D. M. laevigata (6 skins, Ambodimanga, bamboo grove, Dec. 1990). E. M. laevigata (5 skins, Varary, bamboo grove, Jan. 1994). F. M. laevigata (6 skins, Nosy Mangabe, insular forest, Dec. 1993). 1996 15 DALY ET AL.: MADAGASCAN POISON FROGS B 307G < 325A D 217A>217B 309A 291 E -, 307F Fig. 8. Gas chromatographic traces showing alkaloid profiles from Mantella baroni (M. madagascariensis, auctorum); conditions as in figure 6. A. 10 skins, approx. 14-18 km N Andasibe, stream-side forest, Nov. 1989. B. 6 skins, approx. 12 km SE Andasibe by road, stream-side forest, Dec. 1993 (see Garraffo et al., 1993b, fig. 2B, for a gas chromatogram of another sample of 10 skins taken at the same stream; compare samples 8B and 8B' in appendix 2). C. 10 skins, 30-35 km S Moramanga, stream-side forest, Nov. 1989. D. 3 skins, An'Ala, stream-side forest, Dec. 1993. E. 10 skins, Ranomafana, streamside forest, Nov. 1989. F. 17 skins, Sahavondrona, disturbed stream-side forest, Jan. 1993. AMERICAN MUSEUM NOVITATES 16 M. pulchra (B) M. pulchra (A) I1I NO. 3177 267H 293B \ 384A 384B 382, 384A, 384B / M. viridis (E) Fig. 9. Gas chromatographic traces showing alkaloid profiles from Mantella spp.; conditions as in figure 6. A. M. pulchra, 6 skins, near Ambavala, boggy ridge forest, Dec. 1990. B. M. pulchra, 5 skins, same locality as preceding, Jan. 1994. C. M. pulchra, 5 skins, An'Ala, boggy stream-side forest, Jan. 1993. D. M. viridis, 30 skins, Montagne des Fran9ais, semiarid streambed forest and nearby drainage areas, Jan. 1994. E. M. viridis, 5 skins, approx. 13 km S Antsiranana, semiarid streambed forest, Jan. 1994. F. M. viridis, 1 skin, "region of Antsiranana," from dealer, Nov. 1989. 1 996 DALY ET AL.: MADAGASCAN POISON FROGS 17 is most evident on yellow frogs. We have seen red-colored specimens of unknown provenance with a black ear spot. The uniformly bright coloration seems highly derived, but the dorsum and limbs of metamorphs have dark markings according to Arnoult (1966: 938, fig. 4), possibly representative of an ancestral pattern. Skin extracts from seven samples of M. aurantiaca all contained mainly pumiliotoxins and allopumiliotoxins. Levels ranged from very high to quite low (fig. 6A,B and data not shown; see also fig. 6A of Daly et al., 1984). Homopumiliotoxins also occurred, but in minor amounts. Varying, but always minor amounts of the alkaloids 235C and 233F, proposed to be dehydrohomopumiliotoxins (Garraffo et al., 1993b), were present. One trace alkaloid 323C appears to represent a new subclass of pumiliotoxins and will be referred to as an "isopumiliotoxin" until its structure can be defined. No decahydroquinolines were detected. "Izidine" alkaloids occurred rarely and only as minor or trace constituents. A 5,8-disubstituted indolizidine 239C and a putative 1,4-disubstituted quinolizidine 265L were present in two samples obtained from commercial dealers. Two minor alkaloids 392 and an 0-acetyl derivative 434 represent a new class of alkaloids. Alkaloid 392 has an empirical formula of C22H36N204 and contains the equivalent of six rings or double bonds. It affords a mass spectral base peak at m/z 252 (C14H22NO3). If the skin alkaloids of M. aurantiaca originate from dietary sources, then the swamp forest habitat must provide arthropod prey affording large amounts of pumiliotoxins and allopumiliotoxins, only small amounts of the "izidine" alkaloids, and none of the decahydroquinolines. tween 1882 and 1981 (cowanii, haraldmeieri, pulchra), leading to an erroneous belief in a Mantella baroni Boulenger, 1888 (Mantella madagascariensis, auctorum) 8 After writing the above, we obtained the second edition of Glaw and Vences' admirable Fieldguide, which was published in December 1994. These authors (p. 403) designated a lectotype of Dendrobates madagascariensis because "the paralectotype differs largely from the lectotype and probably belongs to a different species ... Both specimens are in a very bad state of conservation . . . Some colour patterns are present on the legs (which are separated from the body) [of the lectotype].... These colour patterns partly indicate an attribution to the colour morph which in the past was attributed to Mantella madagascariensis, and partly also are similar to ... Mantella pulchra. No unequivocal attribution ofthe mad- Figures 2C, 5 The name Mantella baroni is here resurrected for a species that, as currently recognized, occurs in upland forest over a long geographic range in eastern Madagascar. Based primarily on studies of preserved specimens, the distinctive coloration and color pattern of M. baroni has been confused with the patterns of other valid species named be- species of extraordinary variability-the composite species having been known until now as M. cowanii (sensu Guibe, 1964, 1978) and M. madagascariensis (sensu Busse, 1981; Blommers-Schl6sser and Blanc, 1991; Glaw and Vences, 1992a, 1994). Guibe (1978: 84) had properly considered Dendrobates madagascariensis Grandidier (1872) as an "espece douteuse," that is, as a species indeterminata or a nomen dubiuma species not identifiable from the original publication or a name not certainly applicable to any known species of frog. He based this decision on the poor condition of the syntypes and on the inadequate original description, which reads as follows (Grandidier, 1872: 10-11): d D'un noir bleuatre uniforme. Abdomen seme de taches d'un bleu clair-, cuisses et face interne des jambes d'un beau rouge. 9 D'un beau noir mat avec une tache d'un vert clair veloute a la naissance et sur I'avant de chacun des quatre membres. Abdomen seme de taches d'un bleu ciel. Face interne des jambes d'un beau rouge. La peau de ces Dendrobates est finement chagrinee. Long. du corps, Om,022; des membres poster., Om,032. Habit.: Foret d'Ambalavatou, entre Mananzarine et Fianarantsoua. Grandidier's emphasis on coloration of the male's thighs and, in both sexes, the concealed part of the lower leg ("thighs and inside surface of the legs a beautiful red") seems more applicable to Mantella pulchra than to Mantella baroni (cf. colors in fig. 5), but, unless new information is forthcoming, we suggest that the name Dendrobates madagascariensis Grandidier be left unapplied, as a nomen dubium as intended by Guibe (1978).8 18 AMERICAN MUSEUM NOVITATES Although the type locality of M. baroni is only "Madagascar," Boulenger's (1888: 106, pl. 6, fig. 2) description and illustration seem clearly applicable to the present species. Our samples, which span a north-south distance of nearly 300 km (figs. 2C, 5 left), show little variability in color pattern. We know baroni as a black frog with a pale, sharply defined canthal-supraocular line, lacking a labial line, a large green lateral blotch that runs onto the arm, and a green inguinal blotch extending onto the posterior flank and continuous with an elongated green blotch covering most of the upper thigh; in vivid contrast to all the green on black, the shank and foot are totally orange with black markings. The shank and foot also are orange below, but all the rest of the ventral surfaces are black with blue markings (fig. 5, left). The green markings may run to yellowish green, and there is said to be a morph with yellow markings but it was not seen by us. Samples of skins were obtained at six localities (fig. 1), with an additional sample of unknown orgin from a commercial dealer. The habitat varied somewhat, although all specimens collected were found along streams in upland forest in an elevational range of about 800-1000 m above sea level. Collections were made along fairly rapidly flowing streams in relatively undisturbed forest at sites east ofAndasibe and near Ranomafana. Other collections were made along small meandering streams in forest exhibiting varying degrees of disturbance at the sites north of Andasibe, south of Moramanga, and near An'Ala. One population was found by a large agascariensis lectotype to any described form is possible [emphasis added]; we suggest to continue using the name madagascariensis for the morph figured on cp. 61. The synonyms Mantella baroni Boulenger, 1888, and Phrynomantis maculatus Thominot, 1889, belong to this morph and we therefore continue considering these names as synonyms of madagascariensis." Their own conclusions argue against the last action. Except in the rare case of objective synonyms (based on the same type specimen), junior synonyms cannot be attributed to an older name of an unidentifiable species. Because of the confusion associated with the name madagascariensis, nothing is to be gained by petitioning for its conservation. NO. 3177 stream in second-growth forest near Sahavondrona. John E. Cadle (in litt., August 6, 1995) confirmed the above observations on apparent habitat preference of M. baroni but also emphasized that the species is sometimes found in unlikely places: In the Ranomafana region (Ranomafana National Park and surrounding areas), Mantella baroni is found in a variety of habitats. Although most specimens are found near streams, actual stream-side habitats vary; the recorded elevation in this region is about 7001200 m. I would characterize most situations as riparian primary montane rainforest; the streams usually have a rocky (as opposed to sandy) substrate, and are generally fast-flowing white-water streams. Another streamside habitat where they were abundant in one area was dense grass (up to > 1 m tall). Mantella baroni was also found in swampy areas of little relief with meandering stream courses, and characterized by arborescent Pandanus species, many with stilt roots; these areas were often boggy and with sandy substrate. One specimen was collected in primary montane rainforest away from streams. Finally, near the village of Sahavondrona, M. baroni was seemingly common on a hillside with a very dry aspect and chaparral-like shrubby vegetation (very likely disturbed); although there were some streams in this area, several mantellas were found out in the open away from streams, and I was surprised to find them in this seemingly stressful environment (they were the only frogs found in that habitat). One thing that impresses me overall about the distribution of Mantella baroni in the Ranomafana area is the very local nature ofpopulations. They are abundant in small areas, then apparently absent over wide stretches of apparently suitable habitat. This is all the more surprising given the habitat breadth for the species in the general region. Skin extracts from all samples had moderate to high levels of alkaloids (fig. 8A-F; see also fig. 2B in Garraffo et al., 1 993b). The alkaloid profiles varied among populations, but pumiliotoxins and/or allopumiliotoxins were always major alkaloids. "Izidine" alkaloids were also prominent components. The 5,8-disubstituted indolizidine 217B and the 1 ,4-disubstituted quinolizidines 217A and 231A were present in all samples either as major or minor alkaloids. Other indolizidines and quinolizidines were also present. Homopumiliotoxins often were present as minor components. Other alkaloids occurred in one or more samples. Thus, the skin samples from relatively undisturbed forest by a meandering stream near An'Ala (fig. 8D) and by a cascading rocky stream reached by road east of Andasibe (fig. 8B) contained a major 1 996 DALY ET AL.: MADAGASCAN POISON FROGS alkaloid 281F (C17H31N02) that was proposed to be a dihydropumiliotoxin (Garraffo et al., 1993b). The 8-deoxypumiliotoxins represent a new subclass of alkaloids described from dendrobatid frogs (Jain et al., 1995). The sample from east of Andasibe contained an 8-deoxypumiliotoxin 291E (C19H33NO) and a previously undetected alkaloid 251P (C16H29NO) of unknown structure. Alkaloid 251P had a mass spectral base peak at m/z 136 (CqH14N+ ) and a major fragment at m/z 122 (C8H12N+). The skin sample from near Sahavondrona (fig. 8F) contained the pyrrolizidine oximes 236 and 252A, possibly of millipede origin, and an alkaloid 281G (C17H31NO2) that is proposed to be an 8-deoxypumiliotoxin. The skin sample of unknown origin (from a dealer) had high levels ofallopumiliotoxins 323B and 325A (gas chromatograph not shown). It was lacking the quinolizidines 217A and 231A found in all other M. baroni samples. It had a previously undetected 3,5-disubstituted indolizidine 21 1E (C13H2_NO), a putative 1 ,4-disubstituted quinolizidine 265L, and an alkaloid 392 (C22H36N204) of unknown structure. An early report (Daly et al., 1984) on alkaloids of M. madagascariensis and another (Garraffo et al., 1993b) on M. sp., cf. madagascariensis, both apply to Mantella pulchra and are discussed under that name. Mantella betsileo (Grandidier, 1872) Figure 2D The name Mantella betsileo is being used for a species with an orangish (or yellowish) brown dorsum set off by black sides, with a pale labial line extending to the arm; the specimen shown (fig. 2D) was overall black ventrally (including palms and soles) with a blue line around the lower lip and small, irregular pale blue markings on the venter and under the hind limbs. As currently recognized, Grandidier's M. betsileo appears to have a wide, nearly circumcoastal range (Kuchling, 1993, map), at elevations from sea level to about 500 m (Glaw and Vences, 1994: 177). However, the type locality, "Pays des Betsileos" [land of the Betsileo Tribe], is inland and upland on 19 the southeastern plateau-west ofthe eastern forest zone on the escarpment and therefore well inland from the eastern lowlands (maps in Grandidier, 1893; Labatut and Raharinarivonirina, 1969; Methuen and Hewitt, 1913; Sibree, 1880)-which causes us considerable concern and leads us to question whether the name M. betsileo is being correctly applied. A search for existing Mantella habitat along Grandidier's route would seem worthwhile. Grandidier (1893) provided a detailed route map of his 1869-1870 expedition, including travel in the Betsileo territory. A contemporary account of the territory and its limits was given by Shaw (1875). Two of the localities reported here are inland from the northeast coast (fig. 1), but still at low elevations. Our collections were made in three quite different habitats, one from an inland clove tree forest (about 100-200 m elev.), one from a disturbed patch of inland forest and second growth around a small bamboo grove (about 100-200 m), and one from disturbed, very open coastal forest (50 m). The inland collections were in the Mananara Reserve area near Sandrakatsy. M. betsileo was microsympatric with the semiarboreal M. laevigata at the bamboo grove. M. betsileo was also seen in microsympatry with M. pulchra in nearby, relatively undisturbed forest. One specimen (AMNH 140574) of Mantella from that forest seemed in the field somewhat intermediate between microsympatric M. betsileo and M. pulchra; the possibility of hybridization was considered, inasmuch as the specimen had the dorsal coloring of betsileo and the lateral body pattern of pulchra. However, it had golden yellow flash marks in the groin and in the calf and lacked bright coloring on the rear of the thigh (compare hind leg colors in pulchra, fig. 5, right). AMNH 140574 may represent an unrecognized species, but additional collections are needed; color notes are provided at the end of appendix 1, under Mantella sp. Extracts of M. pulchra (fig. 9B), but not of M. betsileo, were obtained from this site. All three skin extracts from M. betsileo had relatively low levels of alkaloids (fig. 6C, D, E), but alkaloid profiles differed significantly among the three populations. The inland populations (fig. 6C, D) contained a decahy- droquinoline (195A), accompanied by un- AMERICAN MUSEUM NOVITATES 20 saturated analogs (193D, 189) and apparent dimers 384A and 384B. The "dimers" had empirical formulas of C26H44N2. Pumiliotoxins, allopumiliotoxins, and homopumiliotoxins were minor components of the inland populations. The inland sample from the bamboo grove had a 5,8-disubstituted indolizidine 217B, and a 1,4-disubstituted quinolizidine 217A as minor components (fig. 6D). A related quinolizidine 219F was present as a minor component in the clove tree forest population (fig. 6C). An alkaloid, 195C, was present as a major component in one inland population (the clove tree forest), but was absent in the other (bamboo grove). The structure of195C remains uncertain (see Garraffo et al., 1 993b). Thus, the two inland populations were similar in containing the decahydroquinoline 195A as a major component along with "dimers" 384A and 384B, and with pumiliotoxins, allopumiliotoxins, and "izidine" alkaloids as minor components. But the nature of the "izidine" alkaloids was quite different for the two inland populations. In contrast, the coastal population had the decahydroquinoline 195A as a minor component and the "dimers" 384A and 384B were not detected (fig. 6E). Pumiliotoxins, allopumiliotoxins, a homopumiliotoxin, pyrrolizidines, indolizidines, and a quinolizidine were present in the coastal population. An alkaloid 251Q (C16H29NO) of unknown structure was present as a minor component in the coastal population and as a trace component in the inland bamboo grove population. It afforded a mass spectral base peak at m/z 122 (C8Hl2N+). Different sets of small arthropods available as prey in such diverse habitats might be expected to contribute to the marked differences in alkaloid profiles of M. betsileo. Mantella cowanii Boulenger, 1882 Figure 3A This very distinctive species of the eastcentral uplands has vivid orange-red limb markings and axillary/groin spots on a pure black body; pale labial and canthal lines are lacking; ventral surfaces are black with a dozen or so pale blue spots. Despite easy rec- NO. 3177 ognition of M. cowanii, its taxonomic status has been greatly confused, mainly because, until recently, workers have not seen living specimens, but have attempted to include the pattern of preserved specimens into presumptive variation of other species. Busse (1981) considered it a synonym of "madagascariensis," whereas other recent authors have used cowanii as a valid name (usually spelled "cowani") but have misapplied it to other species. Bohme et al. (1993) cleared up some of the nomenclatural confusion and provided photographs in dorsal and ventral view of a syntype of M. cowanii, concluding that it might be distinct from "madagascariensis." The type locality is East Betsileo. Specimens recently found their way into the animal trade, reputedly from upland forest (about 1500 m elev.) near Ambositra (Andrew Clark, personal commun.). And Vences et al. (1994) have reported rediscovery of the species near both Ambatolampy and Fianarantsoa, in the same uplands as Ambositra. Our specimens also were provided in 1993 by a commercial dealer, who gave an improbable (lowland) locality (Anosibe An'Ala) as source, for which reason we consider the specimens as lacking specific locality data. All the above localities are indicated on the map (fig. 1). The major skin alkaloid in M. cowanii was pumiliotoxin 251D, which was present at very high levels (fig. 6F). The only other Mantella species that had pumiliotoxin 251D as a major alkaloid was M. expectata, a frog from semiarid regions of western Madagascar. In M. cowanii other pumiliotoxins and allopumiliotoxin 267A were present as minor components. All alkaloids appeared to be pumiliotoxins/allopumiliotoxins and deoxypumiliotoxins. No decahydroquinolines or "izidine" alkaloids were detected even as trace components. The predominance of pumiliotoxin/allopumiliotoxins in the skin of M. cowanii suggests that dietary prey available to or targeted by this species at the presumed upland forest habitat contain mainly pumiliotoxin 251D and related alkaloids. It also suggests that arthropods containing decahydroquinolines or "izidine" alkaloids are absent from the microhabitat. 1 996 DALY ET AL.: MADAGASCAN POISON FROGS Mantella crocea Pintak and B6hme, 1990 Figure 3B This endemic of northeastern Madagascar is, like M. aurantiaca, known only from upland swamp forests (about 1000 m elev.) north of Andasibe. The coloration seems fairly constant and diagnostic (fig. 3B): A black face mask is sharply defined dorsolaterally until midbody, where it falls off obliquely to the venter; a pale labial line extends to the arm. The dorsal surfaces above the black mask and the posterior flanks are contrasting yellowish to orangish brown, ranging from nearly uniform to finely black-stippled, with or without a blackish vertebral line and sometimes also with a weak, grayish hour-glass marking on the back (AMNH 136896). The limbs are colored much like the dorsum, often with faint dark bands on the thigh and shank. The groin and rear of thigh are bright reddish orange, this color also predominating underneath the shank and, variably, on the adjacent ventral surfaces of thigh and foot. The throat and venter are black with very pale bronzy yellow or whitish markings (ventral coloration can be seen in Pintak and B6hme's 1990 description, and in Zimmerman et al., 1990 [as Mantella sp.]). Four separate samples of skins ofM. crocea were obtained with the help of local collectors. Pumiliotoxins and allopumiliotoxins were major alkaloids as they had been in another other swamp-forest Mantella (aurantiaca) from this region. Levels ranged from low to moderate (fig. 7A, B and data not shown). Furthermore, as in M. aurantiaca, the putative dehydrohomopumiliotoxins 235C and 233F were minor components and "isopumiliotoxin" 323C and alkaloids 392 and 434 were detected. However, all samples of M. crocea contained a putative 5,8-disubstituted indolizidine 295B (CI9H33NO2) as a minor component. This alkaloid, which has a mass spectral base peak at m/z 154 (C9HI6NO+ ), was not detected in any sample of M. aurantiaca. The marked similarity between alkaloid profiles in M. crocea and M. aurantiaca is not unexpected if arthropod prey from similar swamp forest habitats are the source of their skin alkaloids. However, further fieldwork on 21 the nature of the swamp forests inhabited in this region by M. aurantiaca and M. crocea are needed. It is not known if they ever occur together. Mantella expectata Busse and B6hme, 1992 Figure 3C This recently described species appears to occur widely in semiarid western Madagascar, where it probably has been confused with M. betsileo (see Busse and B6hme, 1992: 60). The dorsum is greenish yellow to yellowish brown, sometimes with a reddish brown suffusion posteriorly, the dorsal color being sharply set off from uniformly black sides; a somewhat variable pale labial stripe extends to the arm. The limbs may be a vivid light blue (photos in Busse and Bohme, 1992; Glaw and Vences, 1 992a), varying through grayish blue (fig. 3C) to light or dark brown, with blue coloring remaining on rear of thigh and concealed calf area.9 The ventral surfaces are overall heavily marbled or spotted pale blue on black, with a tendency for a blue horseshoe shaped marking on the chin. The type locality of M. expectata is 20 km SE Toliara; the frog has also been reported from near Morondava (as Mantella sp., Meier, 1986, color photo), with other localities being added by Glaw and Vences (1994: 177). Our collections were made in southwestern Madagascar in the Massif Isalo near Ranohira. The frog occurred in streambeds (about 800 m elev.) that were relatively dry with only standing pools of water. Vegetation along the streambed consisted mainly of grassy hillocks. There were infrequent small trees. The frogs were said to be abundant after infrequent heavy rains, but even calling males were difficult to locate under the dry conditions that pertained when collections were made in January 1993 and January 1994. Both samples were from the same streambed. This 9 A few brown-limbed specimens were seen in the sampled population and the limbs of captives changed from blue to blackish brown. Busse and Bohme (1992: 60) noted that this change does not affect the areas of blue on the hidden parts of the hind limbs. 22 AMERICAN MUSEUM NOVITATES drying bed led to a wet, spring-fed forested drainage area where, quite remarkably, there seemed to be relatively fewer frogs than in the open streambed. Skin extracts of M. expectata contained several pumiliotoxins including 251D, 237A and 307A. Levels were relatively low (fig. 7C and data not shown). One of the pumiliotoxins (323D) was a previously undetected member of the class and its structure is uncertain. Several "izidine" alkaloids were present including a 3,5-disubstituted indolizidine 275C, 5,8-disubstituted indolizidines 207A and 219F, and a 1,4-disubstituted quinolizidine 207C. A tricyclic alkaloid 261C (C18H31N) of unknown structure was present. Another previously undetected alkaloid 197C (C12H23NO) may prove to be a hydroxylated 5,8-disubstituted indolizidine. It gave a base peak at m/z 154 (CgH16NO+) and a major fragment ion at m/z 96 (C6H1ON+). The occurrence of pumiliotoxins/allopumiliotoxins as major alkaloids both in a frog from a swamp forest (M. aurantiaca) and one from a semiarid streambed (M. expectata) suggests that putative alkaloid-containing prey with such alkaloids can occupy a wide range of microhabitat niches. Mantella laevigata Methuen and Hewitt, 1913 Figure 3D This species occurs in the northeastern coastal region and is the only Mantella treated in this paper that is semiarboreal in habit, the others being entirely terrestrial. The relatively large finger discs (fig. 3D), emphasized in the original description, are correlated with its scansorial abilities. Glaw and Vences (1994: 180, color photo 48) indicated the existence of a similarly colored, possibly unnamed, species (at Marojezy) that occurs sympatrically with M. laevigata but lacks the enlarged finger discs. The anterior dorsal surfaces of M. laevigata are green or yellowish green, in marked contrast to the rest of the black body and limbs, except the rear of the thigh, which was pale blue in the specimen illustrated (fig. 3D); pale labial and canthal lines are lacking. The venter and undersides of the limbs (including palms and soles) were overall black with pale NO. 3177 blue markings. Pale markings are weak or absent on the grayish throat and chest, which in preserved specimens are blackish brown to light brown, sometimes but not always in sharp contrast to the black belly. Three samples of skins were obtained, two from similar inland bamboo groves (about 100 m elev.) in the Mananara reserve near Sandrakatsy and the third from forest at about 100 m on a small coastal island, Nosy Mangabe. Levels of alkaloids from all populations were relatively low (fig. 7D, E, F). All three populations contained pumiliotoxin 307G as a minor alkaloid. The populations from the inland bamboo groves contained as minor alkaloids the decahydroquinoline 195A and related unsaturated congeners and trace amounts of "dimers" 384A and 384B (fig. 7D, E). The island population did not have the decahydroquinoline class of compounds (fig. 7F). "Izidine" alkaloids were present, but limited in number. The bicyclic alkaloid 195C of unknown structure was present as a minor component in extract from one bamboo grove population, while the 5,6,8-trisubstituted indolizidine 223A and the 1,4-disubstituted quinolizidine 223J were present as minor components in extract from the other bamboo grove population. A 3,5-disubstituted indolizidine 249A was a minor component in extract from the island population. Mantella laevigata is an arboreal species that breeds in tree holes (Glaw and Vences, 1992b). Most of the specimens seen by us were 1 to 2 meters off the ground, but a few were on the ground. If its alkaloids derive from its food, then only the inland bamboo grove sites appear to provide prey with decahydroquinoline and related alkaloids. Many alkaloids were shared between the M. laevigata and M. betsileo collected together at and around a bamboo grove, but the profiles were different (compare fig. 6D and 7E). Mantella pulchra Parker, 1925 Figures 4A, B, 5 This species has a relatively wide distribution in eastern Madagascar, where it seems to occur from an elevation of about 200 m up to about 1000 m, becoming sympatric at some upland sites with M. baroni. Mantella pulchra is a bronzy brown or 1 996 DALY ET AL.: MADAGASCAN POISON FROGS blackish frog with an ill-defined bronze canthal-supraocular line, no labial line, a large green lateral blotch running onto the upper arm, and a smaller green inguinal blotch confluent with an elongated green blotch covering most of the upper thigh; there is a vivid red or orange flash mark'0 on the concealed part of the thigh and calf. There may be a few dark red spots underneath the thigh, but otherwise the ventral surfaces (including palms and soles) are black overall, nearly uniform or with blue markings overall, including blue edging around the lower lip. The coloration of M. pulchra therefore is similar to the larger M. baroni, but the two are immediately distinguished by hind-limb pattern-a distinct flash mark in pulchra vs. the uniform orange and black coloring in baroni (fig. 5). There is a variant in the commercial trade having blue rather than green blotches (Glaw and Vences, 1994: color fig. 63). Busse (1981: 29) and Blommers-Schl6sser and Blanc (1991) considered pulchra a synonym of M. "madagascariensis." Glaw and Vences (1 992a: 164-165) tentatively treated it as a color morph of "madagascariensis," even though they described call differences in sympatry (at Andasibe). In the same year, Andreone (1992) confirmed call and other differences in sympatry at the same general locality (Perinet forest) and recognized that two taxa were involved-M. "madagascariensis" (baroni) and another taxon for which he used the name Mantella cowani pulchra. M. pulchra is recognized as a species in subsequent literature. We here report microsympatry of M. pulchra and the larger M. baroni along a meandering forest stream near An'Ala, at an elevation of about 900 m. Mantella baroni occurred throughout the collection area along the stream, whereas M. pulchra was more 10 There is a curious differential fading in the orange or red flash marks of preserved M. pulchra. For example, after 1 /2 years in preservative, the thigh/calf mark of the specimen in figure 5 changed to yellowish white with vivid orange edging along the ventrolateral edge of the shank, and the dull reddish markings under the thighs became a bright vivid orange. This explains Parker's (1925) otherwise puzzling original description, in which the flash marking is described as a bicolor "brilliant yellow" and "bright crimson." 23 abundant in stream-side areas that were somewhat boggy. M. pulchra was also found about 50 m up the forested slopes in another boggy area. Another very localized population of M. pulchra was sampled in a boggy area of ridge forest at Ambavala, in the Mananara Reserve near Sandrakatsy, about 300 km northeast of An'Ala. Additional specimens from a dealer were said to have come from Anosibe An'Ala, some 70 km southwest of An'Ala. Levels of alkaloids varied in three population samples of skin of M. pulchra (fig. 9A, B, C), being quite high in the An'Ala sample (fig. 9C) and relatively low in two samples from Ambavala (9A, B). Profiles also differed considerably in the three samples. The population of M. pulchra from An'Ala shared many alkaloids with the microsympatric M. baroni, including pumiliotoxins and allopumiliotoxins, 1 ,4-disubstituted quinolizidines 217A and 231A, and the putative deoxypumiliotoxin 291E. It differed from microsympatric baroni in having an "isopumiliotoxin" 267H, two 3,5-disubstituted pyrrolizidines 223M and 2510, and an alkaloid 239N (C17H21N) of unknown structure. Skins of M. pulchra from a swampy ridge forest near Ambavala contained decahydroquinoline 195A and unsaturated analogs and "dimers" 384A/384B, pumiliotoxin 267C and the 3,5disubstituted pyrrolizidine 267H (fig. 9A, B). The higher molecular-weight pumiliotoxins present as major alkaloids in the An'Ala population were absent in the Ambavala populations. The reputed "Anosibe An'Ala" sample (from a dealer) had very low levels of alkaloids (gas chromatograph not shown). The two major alkaloids in that sample were the 3,5-disubstituted pyrrolizidine 223M and the pumiliotoxin 251D. Profiles in M. pulchra differ considerably, perhaps reflecting different habitats and diet. Based on limited observation, M. pulchra seems to prefer forest that is slightly swampy or boggy underfoot, in contrast to our samples of M. baroni, which came from streamside forest. Two previous reports on skin alkaloids in M. pulchra were published under the names Mantella madagascariensis (in Daly et al., 1984) and Mantella sp., cf. madagascariensis (in Garraffo et al., 1993b). The first report 24 AMERICAN MUSEUM NOVITATES was based on an extract from a specimen of M. pulchra (then in the synonymy of madagascariensis sensu Busse, 1981) obtained from a commercial dealer in the early 1980s. Results from that single specimen (AMNH 114047) are not in accord with alkaloid profiles from any wild population subsequently sampled, in that it contained a set of histrionicotoxins (283A, 285A, 285C) and an apparent acyclic alkaloid 241B-none ofwhich has been detected even in trace amounts in any species of Mantella collected on four trips to Madagascar. Thus, we are cautious and believe that the presence ofhistrionicotoxins or acyclic alkaloids in Mantella must be discounted unless authenticated with data from field-collected specimens. It is conceivable that the specimen obtained from a commercial dealer in the United States had been fed with New World arthropods containing histrionicotoxins and other alkaloids. Mantella viridis Pintak and B6hme, 1988 Figure 4C This frog appears to be endemic to the northern tip of Madagascar. Dorsal surfaces and flanks are greenish yellow, with or without vague grayish lines. A black face mask is sharply defined dorsolaterally to just past the arm, where it falls off obliquely to the venter (fig. 4C); a pale labial line extends to the arm. The rear of the thigh is pale blue. The ventral surfaces were overall black (including palms and soles) with pale blue markings, including a tendency for a horseshoe-shaped mark on the lowerjaw. In preserved specimens, at least, the throat and chest are brown, sometimes in obvious contrast to the black belly (as in M. laevigata). Samples of skins of M. viridis were obtained in 1994 from two sites in the region of the type locality (south of Antsiranana). One was inland, from a nearly dry streambed 13 km south of Antsiranana. The other was from a semiarid stream drainage area on Montagne des Franqais,II overlooking the 11 A report of M. betsileo from Montagne des Francais (Blommers-Schlosser and Blanc, 1991: 371) could not be confirmed and may be a case of misidentification. Only M. viridis was found and was relatively common, more than 40 individuals being seen in about four hours of collecting. NO. 3177 coast. Both streambeds were forested, but both, especially the more heavily forested inland site, were disturbed by clearings. High levels of pumiliotoxins 307A and 323A were present in the Montagne des Francais sample (fig. 9D). In addition, an isomer of pumiliotoxin 307A of unknown structure was present. There were trace amounts of an 8-deoxypumiliotoxin 289C and a homopumiliotoxin 223G. The inland sample, in contrast, contained a wide variety of alkaloids (fig. 9E). Pumiliotoxins 307A and 323A, the former at high levels, were present along with several other pumiliotoxins and high levels of homopumiliotoxin 223G. There were also several 5,8-disubstituted indolizidines, three of which, namely 193E, 217B, and 2211, were present at moderately high levels. A previously undetected alkaloid 207M (C14H25N) of unknown structure was present. It afforded major mass spectral fragment ions at m/z 164 and 140. An earlier (1993) sample of one skin of M. viridis, obtained from a dealer and stated to be from Antsiranana, had very high levels of pumiliotoxins 307A and 323A along with minor amounts of 5,8-disubstituted indolizidines 217B and 237H (fig. 9F). Prey that are the putative source of pumiliotoxins would appear to have been present at both 1994 sites of collection of M. viridis, but prey containing "izidine" and other alkaloids would seem to be available only at the more forested, inland site. SUMMARY OF ALKALOIDS This survey of alkaloids in skin extracts of 9 of 11 species currently recognized in Mantella reveals the presence of about 100 compounds. The structures of many of these were known, having been previously isolated and structurally defined from New World dendrobatid frogs (Daly et al., 1993). However, about 30 alkaloids from Mantella represent new compounds of unknown or uncertain structure. Characterization of these alkaloids will be presented elsewhere. The major classes of alkaloids found in mantelline skin can be summarized as follows: 1. Pumiliotoxin class, which includes pumiliotoxins, allopumiliotoxins, homopumiliotoxins; the structures of most of these are known. In addition, this class contains less welldefined compounds, i.e., the putative dehydro- 25 DALY ET AL.: MADAGASCAN POISON FROGS 1996 Pumiliotoxins (PTX) OH CH3 OH3 267"OH "5OH 251 D 267C OH , C 3OH3 H OH3 ,H - 307F 307A OH3 ""| 8tOH 307G OH OH3 OH OCH H3 N 30'OH H 23 323A 309A Allopumiliotoxins (aPTX) Homopumiliotoxins (hPTX) OH OH3 OH3 H"OH 223G 321 B Fig. 10. Representative pumiliotoxin-class alkaloids from Mantella spp. homopumiliotoxins, the 8-deoxypumiliotoxins, the 9,10-dihydropumiliotoxins, and the "isopumiliotoxins," most of which are as yet unique to Mantella. Compounds of the pumiliotoxin class occur as major alkaloids in all nine species of Mantella sampled. 2. Decahydroquinolines and related compounds. These occur together in several species and consist of a decahydroquinoline 195A, which is well known from dendrobatid frogs, and, in addition, various unsaturated congeners and the apparent "dimers" 384A 26 AMERICAN MUSEUM NOVITATES NO. 3177 Fig. 11. Representative dehydrohomopumiliotoxins, a 9,1 0-dihydropumiliotoxin and an 8-deoxypumiliotoxin from Mantella spp. The structures are tentative. and 384B of unknown structure. This grouping of related compounds has been found in only three of the nine species studied (M. betsileo, M. pulchra, M. laevigata). 3. "Izidine" alkaloids, which consist of 3,5-disubstituted pyrrolizidines, 3,5-disubstituted indolizidines, 5,8-disubstituted indolizidines and 1,4-disubstituted quinolizidines. These occur in varying amounts and types in most populations of the nine species. In M. baroni, indolizidines and quinolizidines are major alkaloids often occurring at high levels. In addition to these three general classes, various bicyclic and tricyclic alkaloids of unknown structure occur in Mantella. Some seem likely to represent new structural class- es, hitherto undetected during 30 years of investigation of alkaloids from New World dendrobatid frogs. Conversely, certain "dendrobatid alkaloids," in particular the batrachotoxins, the histrionicotoxins, the gephyrotoxins, and epibatidine, were not detected in any of the recent mantelline extracts. Three histrionicotoxins were present as minor alkaloids in a single Mantella skin obtained a decade ago from a dealer in the United States (Daly et al., 1984). However, as concluded under M. pulchra, these alkaloid identifications-from a skin of a frog of unknown provenance and maintained for an unknown length of time in the New Worldmust be discounted unless verified from fieldcaught specimens. DALY ET AL.: MADAGASCAN POISON FROGS 1996 27 Decahydroquinoline-Class (DHQ) OH3 OH3 N H N H 193D cis-195A 3,5-Disubstituted Pyrrolizidines (Pyr) HO 267H' 3,5-Disubstituted Indolizidines (3,5-1) 249A 275C 5,8-Disubstituted Indolizidines (5,8-1) OH3 NJ lNJ9 217B 219F C9H160H (sec, Z) 243C 279D 1,4-Disubstituted Quinolizidines (Q) OH3 217A 231A Fig. 12. Representative decahydroquinoline-class and "izidine"-class alkaloids from Mantella spp. The structure of the decahydroquinoline-class alkaloid 193D is tentative. It is of interest that an extract of M. baroni from near Sahavondrona contained pyrrolizidine oximes, that at least in dendrobatid frogs are suspected to have a dietary origin from small millipedes (Daly et al., 1 994b). Other possible dietary sources for frog skin alkaloids are ants, some of which do contain 3,5-disubstituted pyrrolizidines and 3,5-di- 28 AMERICAN MUSEUM NOVITATES substituted indolizidines (Jones and Blum, 1983) and beetles, some which contain tricyclic coccinellines (op. cit.). The only known ant alkaloids that were detected in Mantella were the pyrrolizidines 223H and 251K. No beetle alkaloids were detected. Thus, other unknown dietary sources for most of the mantelline alkaloids must be considered. The present study demonstrating pumiliotoxinclass alkaloids as major components in alkaloid fractions from nine species of Mantella suggests that, if the diet is the source, prey containing these alkaloids must be widely distributed in forest habitats ranging from semiarid lowland to swampy upland. NO. 3177 ACKNOWLEDGMENTS We thank Drs. John E. Cadle, Darrel R. Frost, H. Martin Garraffo, and Richard G. Zweifel for reading and commenting on the manuscript, and Dr. Garraffo for help in many aspects of data analysis. Fieldwork in Madagascar was supported by a grant from the Ministry of Education, Science and Culture of Japan, and we are indebted to Drs. Terumi Nakajima and Yoichiro Kuroda for allowing J.W.D. to participate in a joint MadagascanJapanese program on biologically active natural products. APPENDIX 1: VOUCHER SPECIMENS The following Madagascan specimens are na road ca. 12 km (airline) SE Andasibe, along vouchers only for specimens relevant to this surstream 0.5 km S ofroad, AMNH 133656-133664; vey of skin alkaloids in Mantella. Brief color de11 km WSW Ranomafana, 2 km SE Sahavonscriptions are given in the text and representative drona, ca. 1000 m (21°16'S, 47°22'E), AMNH specimens were photographed at the American 136898-136900; Parc Ranomafana, ca. 10 km Museum (see figs. 2-5). All specimens are alcohol (airline) SW Ranomafana (town on Fianarantsoapreserved in the Amphibian Collection of the Ifanadiana road), AMNH 133667. American Museum of Natural History. At least one to several specimens in each series are stanMantella betsileo dard museum specimens, the others being carAntanambaobe, Mananara Avaratra National casses of some of the frogs skinned for alkaloid Park, ca. 100-200 m (ca. 16°16'S, 49°40'E), AMNH extraction. The carcasses permit easy access to an136940-136942; Farakaraina Forest Station, ca. atomical characters and some will be useful for 10 km E Maroantsetra (on coast), 30 m (1 5°26'S, cleared and stained preparations. 49°52'E), AMNH 140566-140570; ca. 7 km SE Sandrakatsy, near village Ambavala, AMNH Mantella aurantiaca 140572-140573; ca. 8 km SE Sandrakatsy, near village Ambavala, Mananara Avaratra National No specific data (from dealers), AMNH 114042Park, ca. 200 m (ca. 16°23'S, 49°44'E), AMNH 114046, 123693-123695; 14-18 km N Andasibe, ca. 1000 m (18°45'S, 48°25'E), AMNH 136889136892; region of Andasibe, AMNH 136921; near Andasibe (Perinet), AMNH 133611-133641. Mantella baroni 14-18 km N Andasibe, ca. 1000 m (18°45'S, 48025'E), AMNH 136888; near Andasibe (Perinet), AMNH 133668-133681; 7 km SE Andasibe, 1 1 km (by rd) E Andasibe, ca. 1000 m (18°58'S, 48028'E), AMNH 136887; ca. 2 km SW An'Ala, 7-8 km SEAndasibe, ca. 1000 m (18057'S, 48°27'E), AMNH 136901-136902; 2 km SW An'Ala, AMNH 140555-140556; 30-35 km (airline) S Moramanga, near km 45 on Moramanga-Anosibe An'Ala road, along stream ca. 2 km W of road, AMNH 133665-133666; Moramanga-Toamasi- 140571. Mantella cowanii No specific data (dealer claimed "Anosibe An'Ala"), AMNH 140546-140550. Mantella crocea Near Andasibe (Perinet), AMNH 133642-133655; 14-18 km N Andasibe, ca. 1000 m (18°45'S, 48025'E), AMNH 136893-136897. Mantella expectata Massif Isalo, ca. 10 km (by rd) SW Ranohira, 500 [of] road, ca. 800 m (22°38'S, 45°22'E), AMNH 136922-136939, 140617. m south 1996 DALY ET AL.: MADAGASCAN POISON FROGS Mantella laevigata Ambodimanga, Mananara Avaratra National Park, ca. 100 m (ca. 16°22'S, 49°47'E), AMNH 136945136946; Nosy Mangabe, >100 m (15°30'S, 49046'E), AMNH 140557-140562; ca. 7 km SE Sandrakatsy, near village Ambavala, AMNH 140563-140565. Mantella pulchra No specific data (from dealer), AMNH 114047114048; 2 km SW An'Ala, AMNH 140551140552; ca. 2 km SW An'Ala, 7-8 km SE Andasibe, ca. 1000 m (18°57'S, 48°27'E), AMNH 136903-136916; ca. 8 km SE Sandrakatsy, near village Ambavala, Mananara Avaratra National Park, ca. 200 m (ca. 16°23'S, 49°44'E), AMNH 140553-140554; ca. 8 km SE Sandrakatsy, village Ambavala, Mananara Avaratra National Park, ca. 100-200 m (16°23'S, 49°44'E), AMNH 136943136944; region of Anosibe An'Ala (from dealer) (ca. 19°26'S, 48°12'E), AMNH 136918. Mantella viridis No specific data (from dealer), AMNH 133682133683; "region of Antsiranana" (from dealer, AMNH 136919-136920; 13 km S Diego Suarez (=Antsiranana) (12°12'S, 4901 6'E), AMNH 29 140581; Montagne des Fran9ais, 8 km (by rd) southeast from Diego Suarez (=Antsiranana) (12°19'S, 49°20'E), AMNH 140575-140580. Mantella species About 8 km SE Sandrakatsy, near village Ambavala, Mananara Avaratra National Park, ca. 200 m (ca. 16°23'S, 49°44'E), AMNH 140574. See text discussion under Mantella betsileo regarding this specimen. It apparently is an immature female, 21 mm SVL. Color snapshots taken before formalin fixation show an orange-brown dorsum and limbs, with green at the upper arm insertion and a large green patch covering the anterior and most of the dorsal thigh surface. Ventral surfaces were black overall, with bright blue edging around the lower lip and small, irregularly shaped bright blue spots on throat, venter, and limbs. There was a small golden yellow spot in the groin and a sharply demarcated golden yellow flash mark below the knee in the concealed part of the shank, these markings being white in preservative; the inguinal mark is small and the calf marking large, occupying the proximal 60 percent of the shank. In lateral view (not visible in the color photographs), a pale (presumably green) anterior lateral blotch is confluent with the pale limb insertions, and a similar blotch posteriorly on the flanks is confluent with the large (green) anterodorsal thigh patch. 30 AMERICAN MUSEUM NOVITATES NO. 3177 APPENDIX 2: ALKALOIDS IDENTIFIED IN MANTELLA SKIN Flame-ionization gas chromatographic profiles are shown in figures 6-9. Structures of relatively alkaloids are shown in figures 10-12. Abbreviations ofalkaloid classes are as follows: Pumiliotoxins (PTX), allopumiliotoxins (aPTX), homopumiliotoxins (hPTX), dehydrohomopumiliotoxins (dehydrohPTX), 6,1 0-dihydropumiliotoxins (dihydroPTX), 8-deoxypumiliotoxins (deoxyPTX), decahydroquinolines and related alkaloids (DHQ), 3,5-disubstituted pyrrolizidines (3,5-P), 3,5disubstituted indolizidines (3,5-I), 5,8-disubstituted indolizidines (5,8-I), 5,6,8-trisubstituted indolizidines (5,6,8-I), 1,4-disubstituted quinolizidines (Q), 2,5-disubstituted pyrrolidines (Pyr), tricyclics (Tri), pyrrolizidine oximes (Oximes), unclassified (Unclass). Classification and structures for some alkaloids remains tentative. Some alkaloids exist as diastereomers, i.e., PTX 307F often occurs in Mantella as a mixture of diastereomeric 307F' and 307F" as do PTX 307G and 3,5-P 267H (see Garraffo et al., 1993b), but only the gross designation, i.e., 307F, 307G, 267H, followed by the number of diastereomers in parentheses, is given here. An asterisk (*) means that a given sample has been previously described (in Garraffo et al., 1993b). Number designations for each population (6A, 6B, etc.) refer to figures 69 except for 8B'(see Garraffo et al., 1993b, fig. 2B for this sample). common Mantella Species Population Date No. skins Major alkaloids Minor alkaloids Trace alkaloids M. aurantiaca 6A. N Andasibe Nov. 1989 10 skins* PTX 323A. PTX 267C, 307A, aPTX 323B; dehydrohPTX 233F, 235C. PTX 305B, 307B; dehydrohPTX 221F, 265F. 6B. N Andasibe Jan. 1993 5 skins PTX 323A. aPTX 323B; dehydrohPTX 265F; "isoPTX" 323C; Unclass 392, 434. PTX 267C, 277B, 307G, 309A, hPTX 337; dehydrohPTX 233F, 235C; 3,5-I 249A; Q 217A. 8A. N. Andasibe Nov. 1989 10 skins* PTX 309A; Q 217A. PTX 307A, 307F(2), aPTX 325A; dihydroPTX 281F; 5,8-I 217B; 235B, 241F, 243C, 253B, 279D; Q 231A, 233A; Unclass 281F. PTX 251D, hPTX 249F; dehydrohPTX 233F, 235C; 5,8-I 243B, 245B. 8B. E Andasibe Dec. 1993 10 skins PTX 309A; Q 217A, 231A. PTX 307G, 307F(2); aPTX 325A; deoxyPTX 291E, 293D; dihydroPTX 281F; 5,8-I 217B; Unclass 251P. PTX 237A, 251D; hPTX 323E, 337; 5,8-I 241F, 243C, 253B; Q 233A. 8B'. E Andasibe Nov. 1989 10 skins* PTX 309A; dihydroPTX 281F; Q 217A, 231A. PTX 307A; aPTX 325A; 3,5-I 249A, 275C; 5,8-I 217B, 241F; Unclass 293B hPTX 235J; 5,8-I 243C, 279D; Q 273A. 8C. S Moramanga Nov. 1989 10 skins* Q 217A, 231A. PTX 307A, 309A; aPTX 325A; 3,5-I 249A; 5,8-I 217B; 5,6,8-I 223A. PTX 251D, 307F; 3,5-I 275C; 5,8-I 205A, 243D, 245C; Unclass 251L, 293G. 8D. An'Ala Dec. 1993 3 skins PTX 309A; Q 217A, 231A. PTX 307F(2), 323A; deoxyPTX 291E; dihydroPTX 281F; 3,5-I 275C; 5,8-I 217B, 243C; Q 273A. PTX 237A, 307A; aPTX 325A; hPTX 323E; deoxyPTX 293D; 3,5-I 249A; 5,84I 241F, 279D; 8E. Ramomnafana Nov. 1989 10 skins* PTX 309A; Q 217A. PTX 237A, 307F(2); 5,8-I 217B; 243D, 245C; Q 231A, 233A; Tri 207J; Unclass 293B. PTX 251D; 5,8-I 203A; Unclass 205C. 8F. Sahavondrana Jan. 1993 17 skins PTX 309A. 6C. Antanambaobe Dec. 1990 11 skins* DHQ 384A, 384B; Unclass 195C. PTX 307G(2); aPTX 323B; DHQ 195A; 5,84I 219F. PTX 251D, 307A; aPTX 321C, hPTX 223G, 317; DHQ 189, 193D; 5,8-I 217B; Q 217A, 231A, 249C; Tri 207J, 235K; Unclass 211D. 6D. Ambavala Jan. 1994 8 skins DHQ 195A; Unclass 195C. PTX 251D; DHQ 193D, 384A, 384B; 3,5-I 247C; 5,8-I 217B; Q 217A; Unclass 211D PIX 237A, 307G; aPTX 323B; hPTX 321B, 337; DHQ 382, 3,5-I 205A, 247C; 5,84I 251B, 253B; Q 249H, Tri 235M, 265M, Unclass 251Q, 269E, 271B. M. baroni Unclass 275D. M. betsileo PTX 251D, 307F; aPTX 325A; hPTX 251R, 265N; dehydrohPTX 265F; deoxyPTX 281G; 5,8-I 217B; Q 231A, 273A(2). PTX 237A, 307A; 3,5-I 271F, 275C; 5,8-I 241F; Oxime 236, 252A. 1996 DALY ET AL.: MADAGASCAN POISON FROGS 31 6E. Farakaraina Dec. 1993 12 skins PTX 307G; 3,5-P 223H. PTX 251D; hPTX 223G; DHQ 195A; 3,5-P 251K; 5,84I 219F; Q 231A; Unclass 195C. PTX 237A, 267C, 293E; aPTX 323B; hPTX 321B, 335; deoxyPTX 291E; dehydrohPTX 265F; DHQ 181D, 189, 193D, 382, 384A, 384B; 3,54I 247C; Tri 265M; Unclass 237K, 269E. M. cowanii 6F. Dec. 1993 3 skins PTX 251D. PTX 265G, 307G(2), aPTX 7-epi-267A. PTX 267C(2), 323A. M. crocea 7A. N. Andasibe Nov. 1989 10 skins* PTX 267C, 307A, 323A, aPTX 323B. dehydrohPTX 233F, 235C; 5,8-I 295B; Unclass 392, 434. PTX 265G, 305B, 307B; DHQ 195A; 3,5-P 223H(2); Q 231A. 7B. N. Andasibe PTX 267C, 307A, 323A; aPTX 323B. PTX 265G; Unclass 392, 434. PTX 277B, 305B, 309A; dehydrohPTX 233F, 235C; "isoPTX" 267N, 323C; 3,5-P 2390; Unclass 293G. Jan. 1993 35 skins M. expectata 7C. Massif Isalo Jan. 1993 20 skins PTX 251D; 5,8-I 219F; Unclass 261C. PTX 307A; 5,8^ 1193E; Unclass 323D. PTX 267C(2), 305B; aPTX 323B; 3,5-I 223AB, 275C; 5,8-I 197C, 217B, 219J, 241F; Unclass 195C, 233I, 2350. M. laevigata 7D. Ambodimanga Dec. 1990 6 skins* Unclass 195C. PTX 307G, DHQ 193D, 195A. hPTX 223G; DHQ 189, 384A, 384B; Q 207I; Unclass 161, 211C, 211D, 223A. 7E. Varary Jan. 1994 5 skins PTX 307G. DHQ 193D, 195A; 5,8-I 223J; 5,6,8-I 223A; Unclass 195C. hPTX 321B; deoxyPTX 291E; DHQ 189, 382, 384A, 384B, Unclass 275D. 7F. Nosy Mangabe Dec. 1993 6 skins PTX 307G. 3,5-1 249A. PTX 237A, 305B; deoxyPTX 291E; 3,5-P 223H; 3,5-I 247C, 275C; 5,8-I 219F; Q 217A; Pyr 223N, 225H. 9A. Ambavala Dec. 1990 6 skins* 3,5-P 267H(2). PTX 267C; DHQ 384A, 384B; Unclass 161, 293B. PTX 265G; aPTX 325A; 3,5-P 239K(2), 265H(2); DHQ 189; 193D, 195A, Unclass 195C, 211D. 9B. Ambavala Jan. 1994 S skins PTX 267C. PTX 265G; DHQ 189, 193D(2), 382, 384A, 384B; 3,5-P 267H. PTX 307A; DHQ 195A; 3,5-P 223H; Unclass 211D, 271B, 341C, 392, 434. 9C. An'Ala Jan. 1993 5 skins PTX 309A; Q 217A. PTX 251D, 307A, 307F, 325B; aPTX 325A; dihydroPTX 281F; 3,5-P 223M; 3,5-I 249A, 275C; 5,8-1279D; Q 231A(2). PTX 237A, 307G; aPTX 323B; hPTX 337; deoxyPTX 291E, 293D; "isoPTX" 267N; 3,5-P 2510; 5,8-I 245B, 249J; Unclass 239N. 9D. Montagne des Franlais Jan. 1994 30 skins 8PTX 307A. PTX 307D, 323A. PTX 277B, 289C, 305D, 323F. 9E. S. Antsiranana Jan. 1994 5 skins PIX 307A; hPTX 223G. PTX 289C, 323A; 5,8-I 193E, 217B(2), 2211, 237J, Unclass 207M, 2230, 235L(2), 265K, 267M. PTX 209F, 305D; hPTX 239M; "isoPTX" 339C; 3,5-I 223AB; 5,8-I 231C, 245F; Q 207I. 9F. Antsiranana Nov. 1989 1 skin* PTX 307A. PTX 307G, 321A (artefact), 323A; 5,8-I 217B, 237H. PTX 265G, 309B, 307H; hPTX 223G; 3,5-I 223AB; Q 217A, 231A, 233A. M. pulchra M. viridis REFERENCES Andreone, Franco 1992. Syntopy of Mantella cowani Boulenger and Mantella madagascariensis (Grandidier) in central-eastern Madagascar, with notes on the colouration in the genus Mantella (Anura: Mantellidae). Boll. Mus. Regionale Sci. Nat. Torino 10(2): 421-450. Arnoult, J. 1966. Contribution a l'etude des batrachiens de Madagascar. Ecologie et developpement de Mantella aurantiaca Mocquard, 1900. Bull. Mus. Natl. Hist. Nat. Paris, ser. 2, 37: 931-940. Blommers-Schlosser, Rose M. A. 1978. Cytotaxonomy of the Ranidae, Rhacophoridae, Hyperoliidae (Anura) from Madagascar with a note on the karyotype of two amphibians of the Seychelles. Genetica 48(1): 23-40. 32 AMERICAN MUSEUM NOVITATES 1979. Biosystematics of the Malagasy frogs. I. Mantellinae (Ranidae). Beaufortia 29(352): 1-77. 1993. Systematic relationships of the Mantellinae Laurent 1946 (Anura Ranoidea). Ethol. Ecol. Evol. 5: 199-218. Blommers-Schlosser, R. M. A., and Ch. P. Blanc 1991. Amphibiens (premiere partie). Faune de Madagascar 75(1): 379 pp. + 12 pls. Boettger, Oskar 1880. Diagnoses reptilium et batrachiorum novorum a Carolo Ebenau in insula Nossi-Be Madagascariensi lectorum. Zool. Anz. 3(57): 279-283. Bohme, Wolfgang, Klaus Busse, and Frank Glaw 1993. The identity of Mantella cowani Boulenger, 1882 and Mantella haraldmeieri Busse, 1981 (Anura: Mantellinae). Amphib.-Reptilia 14: 269-273. Boulenger, George Albert 1882. Catalogue of the Batrachia Salientia s. Ecaudata in the collection of the British Museum. London, British Museum, xvi + 503 pp. + 30 pls. 1888. Descriptions of new reptiles and batrachians from Madagascar. Ann. Mag. Nat. Hist., ser. 6, 1(2): 101-107 + pls. 5-6. Busse, Klaus 1981. Revision der Farmuster-Variabilitait in der madagassischen Gattung Mantella (Salientia: Ranidae). Amphib.-Reptilia 2: 25-42. Busse, Klaus, and Wolfgang Bohme 1992. Two remarkable frog discoveries of the genera Mantella (Ranidae: Mantellinae) and Scaphiophryne (Microhylidae: Scaphiophryninae) from the west coast of Madagascar. Rev. Francaise d'Aquariol. 19(1/2): 57-64. Channing, A. 1989. A re-evaluation of the phylogeny ofOld World treefrogs. S. Afr. J. Zool. 24(2): 116-131. Daly, John W., George B. Brown, Monica Mensah-Dwumah, and Charles W. Myers 1978. Classification of skin alkaloids from Neotropical poison-dart frogs (Dendrobatidae). Toxicon 16(10): 163-188. Daly, John W., Charles W. Myers, Jordan E. Warnick, and Edson X. Albuquerque 1980. Levels ofbatrachotoxin and lack of sensitivity to its action in poison-dart frogs (Phyllobates). Science 208: 1383-1385. Daly, John W., Robert J. Highet, and Charles W. Myers 1984. Occurrence of skin alkaloids in nondendrobatid frogs from Brazil (Bufoni- NO. 3177 dae), Australia (Myobatrachidae) and Madagascar (Mantellinae). Toxicon 22(6): 905-919. Daly, John W., Charles W. Myers, and Noel Whittaker 1987. Further classification of skin alkaloids from Neotropical poison frogs (Dendrobatidae), with a general survey of toxic/noxious substances in the Amphibia. Toxicon 25(10): 1023-1095. Daly, John W., H. Martin Garraffo, L. K. Pannell, Thomas F. Spande, C. Severini, and V. Erspamer 1990. Alkaloids from Australian frogs (Myobatrachidae): pseudophrynamines and pumiliotoxins. J. Nat. Prod. 53(2): 407421. Daly, John W., Sherrie I. Secunda, H. Martin Garraffo, Thomas F. Spande, Anthony Wisnieski, Charles Nishihira, and John F. Cover, Jr. 1992. Variability in alkaloid profiles in Neotropical poison frogs (Dendrobatidae): genetic versus environmental determinants. Toxicon 30: 887-898. Daly, John W., H. Martin Garraffo, and Thomas F. Spande 1993. Amphibian alkaloids. In G. A. Cordell (ed.), The alkaloids, vol. 43, pp. 185288. San Diego, Ca.: Academic Press. Daly, John W., Sherrie I. Secunda, H. Martin Garraffo, Thomas F. Spande, Anthony Wisnieski, and John F. Cover, Jr. 1994a. An uptake system for dietary alkaloids in poison frogs (Dendrobatidae). Toxicon 32(6): 657-663. Daly, John W., H. Martin Garraffo, Cesar Jaramillo, and A. Stanley Rand 1994b. Dietary source for skin alkaloids of poison frogs (Dendrobatidae)? J. Chem. Ecol. 20(4): 943-955. Dubois, Alain 1984. La nomenclature supragenerique des amphibiens anoures. Mem. Mus. Natl. d'Hist. Nat., ser. A, Zool., 131: 1-64. Dumbacher, John P., Bruce M. Beehler, Thomas F. Spande, H. Martin Garraffo, and John W. Daly 1992. Homobatrachotoxin in the genus Pitohui: chemical defense in birds. Science 258: 799-801. Ford, Linda S. 1993. The phylogenetic position of the dartpoison frogs (Dendrobatidae) among anurans: an examination of the competing hypotheses and their characters. Ethol. Ecol. Evol. 5: 219-231. 1996 DALY ET AL.: MADAGASCAN POISON FROGS Frost, Darrel R. 1985. Amphibian species of the world. Lawrence, KS: Allen Press and Assoc. Syst. Coll., vii + 732 pp. Garraffo, H. Martin, Thomas F. Spande, John W. Daly, A. Baldessari, and Eduardo G. Gros 1993a. Alkaloids from bufonid toads (Melanophryniscus): decahydroquinolines, pumiliotoxins and homopumiliotoxins, indolizidines, pyrrolizidines, and quinolizidines. J. Nat. Prod. 56(3): 357-373. Garraffo, H. Martin, Janet Caceres, John W. Daly, Thomas F. Spande, N.R. Andriamaharavo, and M. Andriantsiferana 1993b. Alkaloids in Madagascan frogs (Mantella): pumiliotoxins, indolizidines, quinolizidines, and pyrrolizidines. J. Nat. Prod. 56(7): 1016-1038. Glaw, Frank, and Miguel Vences 1992a. A fieldguide to the amphibians and reptiles of Madagascar. Cologne, Germany: privately printed, 331 pp. + 16 color pls. 1992b. Zur Biologie, Biometrie und Fiirbung bei Mantella laevigata Methuen and Hewitt, 1913. Sauria (Berlin) 14(4): 25-29. 1994. A fieldguide to the amphibians and reptiles of Madagascar. Second edition, including mammals and freshwater fish. Cologne, Germany: privately printed, 480 pp. + 48 color pls. Grandidier, Alfred 1872. Description de quelques reptiles nouveaux decouverts a Madagascar en 1870. Ann. Sci. Nat. Zool. (Paris), ser. 5, 15(20): 6-11. 1893. Les voyageurs francais a Madagascar pendant les trente demieres annees. Bull. Soc. Geogr., ser. 7, 14: 289-300 + 4 foldout maps. Guibe, Jean 1964. Revison des especes du genre Mantella (Amphibia, Ranidae). Senckenb. Biol. 45(3/5): 259-264. 1978. Les batraciens de Madagascar. Bonner Zool. Monogr. 11: 144 pp. + 82 pls. Hewitt, John 1911. A comparative review ofthe amphibian faunas of South Africa and Madagascar, with some suggestions regarding their former lines of dispersal. Ann. Transvaal Mus. 3(1): 29-39. Jain, Poonan, H. Martin Garraffo, Thomas F. Spande, Herman J. C. Yeh, and John W. Daly 1995. A new subclass ofalkaloids from a dendrobatid poison frog: a structure for 33 deoxypumiliotoxin 251F. J. Nat. Prod. 58(1): 100-104. Jones, Tappey H., and Murray S. Blum 1983. Arthropod alkaloids: distribution, functions, and chemistry. In S. William Pelletier (ed.), Alkaloids: chemical and biological perspectives. vol. 1, pp. 33-84. New York: Wiley. Kuchling, Gerald 1993. Zur Verbreitung und Fortpflanzung von Mantella betsileo in Westmadagaskar. Salamandra 29(3/4): 273-276. Labatut, F., and R. Raharinarivonirina 1969. Madagascar. Etud historique. Paris: Editions Fernand Nathan, 224 pp. Laurent, Raymond F. 1943a. Sur la position systematique et l'osteologie du genre Mantidactylus Boulenger. Bull. Mus. R. Hist. Nat. Belgique 19(5): 8 pp. + 1 pl. 1943b. Contribution a l'osteologie et a la systematique des rhacophorides non Africains. Ibid. 19(28): 16 pp. + 2 pls. 1946. Mises au point dans la taxonomie des ranides. Rev. Zool. Bot. Afr. 39(4): 336338. 1951. Sur la necessite de supprimer la famille des Rhacophoridae mais de creer celle des Hyperoliidae. Ibid. 45(1): 116-122. Liem, Sioe Sing [David S.] 1970. The morphology, systematics, and evolution of the Old World treefrogs (Rhacophoridae and Hyperoliidae). Fieldiana, Zool. 57: vii + 145 pp. Maxson, Linda R., and Charles W. Myers 1985. Albumin evolution in tropical poison frogs (Dendrobatidae): a preliminary report. Biotropica 17(1): 50-56. Meier, Harald 1986. Neues fur den Terrarianer iiber die Gattung Mantella auf Madagaskar. Herpetofauna 8(41): 9-14. Methuen, Paul A., and John Hewitt 1913. On a collection of Batrachia from Madagascar made during the year 1911. Ann. Transvaal Mus. 4(2): 49-64 + pls. 910. Mocquard, Francois 1900. Nouvelle contribution a la faune herpetologique de Madagascar. Bull. Soc. Philom. Paris, ser 9, 2(4): 93-111 + pl. 2. 1909. Synopsis des familles, genres et especes des reptiles ecailleux et des batraciens de Madagascar. Nouv. Archiv. Mus. Hist. Nat., Paris, ser. 5, 1: 1-110. Myers, Charles W., John W. Daly, and Victor Martinez 34 AMERICAN MUSEUM NOVITATES 1984. An arboreal poison frog (Dendrobates) from western Panama. Am. Mus. Novitates 2783: 20 pp. Myers, Charles W., Alfredo Paolillo O., and John W. Daly 1991. Discovery of a defensively malodorous and nocturnal frog in the family Dendrobatidae: phylogenetic significance of a new genus and species from the Venezuelan Andes. Am. Mus. Novitates 3002: 33 pp. Myers, Charles W., John W. Daly, H. Martin Garraffo, Anthony Wisnieski, and John F. Cover, Jr. 1995. Discovery ofthe Costa Rican poison frog Dendrobates granuliferus in sympatry with Dendrobates pumilio, and comments on taxonomic use of skin alkaloids Am. Mus. Novitates 3144: 21 pp. Noble, Gladwyn Kingsley 1931. The biology ofthe Amphibia. New York: McGraw-Hill, xiii + 577 pp. Noble, Gladwyn Kingsley, and H. W. Parker 1926. A synopsis of the brevicipitid toads of Madagascar. Am. Mus. Novitates 232: 21 pp. Parker, H. W. 1925. New and rare reptiles and batrachians from Madagascar. Ann. Mag. Nat. Hist. ser. 9, 16: 390-394. Pintak, Thomas, and Wolfgang Bohme 1988. Mantella viridis sp. n. (Anura: Ranidae: Mantellinae) aus Nord-Madagaskar. Salamandra 24(2/3): 119-124. 1990. Mantella crocea sp. n. (Anura: Ranidae: Mantellinae) aus dem mittleren OstMadagaskar. Ibid. 26(1): 58-62. NO. 3177 Roux, Jean 1935. Sur un nouveau batracien de Madagascar (Mantella loppei n.sp.). Bull. Soc. Zool. France 60: 441-443. Schopf, Clemens 1961. Die Konstitution der Salamander-Alkaloide. Experientia (Basel) 17(7): 285295. Shaw, George A. 1875. The Betsileo: country and people. Antananarivo Ann. and Madagascar Mag. 1(4): 334-347. Sibree, James 1880. The great African island. Chapters on Madagascar. London: Triibner and Co., Ludgate Hill, xii + 372 pp. + frontisp., 3 pls., 2 foldout maps. Vences, Miguel, Frank Glaw, Andre Peyrieras, Wolfgang Bohme, and Klaus Busse 1994. Der Mantella-madagascariensis-Komplex: Wiederentdeckung von Mantella cowani und Beschreibung von Mantella bernhardi n. sp. Datz Aquar. Terrar. 47: 390-393. Werner, Franz 1901. Beschreibung neuer Dendrobatiden. Mit einer Revision dieser Batrachier-Familie. Verhandl. K. K. Zool.-Bot. Gesell., Vienna, for 1901, pp. 627-634. Zimmermann, Helmut, Elke Zimmermann, and Peter Zimmermann 1990. Feldstudien im Biotop vom Goldfr6schchen, Mantella aurantiaca, im tropischen Regenwalds Ost-Madagaskars. Herpetofauna 12(64): 21-24. Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates, Bulletin, and Anthropological Papers published during the last five years are available free of charge. Address orders to: American Museum of Natural History Library, Department D, Central Park West at 79th St., New York, N.Y. 10024. TEL: (212) 769-5545. FAX: (212) 7695009. E-MAIL: scipubs@amnh.org L- ® This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

RELATED PAPERS

RELATED TOPICS

Log In

Mada-gascan poison frogs (Mantella) and their skin alkaloids

Mada-gascan poison frogs (Mantella) and their skin alkaloids

Related Papers

RELATED PAPERS

RELATED TOPICS