Amphibia-Reptilia 33 (2012): 503-520
Description of a new Malagasy treefrog (Boophis) occurring
syntopically with its sister species, and a plea for studies on
non-allopatric speciation in tropical amphibians
Miguel Vences1,∗ , Marcelo Gehara1 , Jörn Köhler2 , Frank Glaw3
Abstract. Based on concordant differences in male advertisement call, tadpole morphology, and absence of haplotype sharing
in mitochondrial and nuclear DNA markers, we describe a new species of treefrog from Ranomafana National Park in the
southern central east of Madagascar. In its adult stage Boophis narinsi sp. n. is highly similar to its sister species, Boophis
majori, but appears to differ in having longer hindlimbs. The genetic divergences between these two species (2.5-3.3% in a
fragment of the 16S rRNA gene, depending on fragment length and individual haplotype analyzed) are below the threshold
typically characterizing distinct species of anurans. Together with their relatively small and largely overlapping ranges and
their sympatric occurrence in Ranomafana National Park, this indicates that they potentially could have originated rather
recently by adaptive speciation under parapatric or sympatric conditions. Most studies on amphibian speciation have so far
by default assumed vicariant speciation. We suggest that alternative speciation scenarios should be considered in future works
and characterize settings in which more reliable assessments of adaptive parapatric or sympatric speciation could be carried
out.
Keywords: Amphibia, Anura, Boophis majori, Boophis narinsi sp. n., Madagascar, Mantellidae, Ranomafana National Park,
speciation.
Introduction
Allopatric (vicariant) speciation has for decades
been considered the dominant process leading to the species diversity of animals and
plants (Via, 2009), and its prevalence has been
strongly advocated by researchers (e.g., Mayr,
1982). However, since the pioneering works of
Bush (1969), empirical and theoretical evidence
has accumulated demonstrating that species formation in sympatric or parapatric settings exists
(e.g., Schliewen et al., 1994; Via, 2001; Barluenga et al., 2006; Bolnick and Fitzpatrick,
2007; Seehausen et al., 2008). In the last years,
the focus of research has broadened from purely
analyzing the geography of speciation towards
deciphering adaptive versus non-adaptive spe-
1 - Division of Evolutionary Biology, Zoological Institute,
Technical University of Braunschweig, Mendelssohnstr.
4, 38106 Braunschweig, Germany
2 - Hessisches Landesmuseum Darmstadt, Friedensplatz 1,
64283 Darmstadt, Germany
3 - Zoologische Staatssammlung München, Münchhausenstr. 21, 81247 München, Germany
∗ Corresponding author; e-mail: [email protected]
© Koninklijke Brill NV, Leiden, 2012.
ciation mechanisms, and the influence of ecological divergence versus sexual selection (Via,
2001; Butlin et al., 2008).
In amphibians, sympatric scenarios of species
formation have only been rarely considered
(summarized in Vences and Wake, 2007). Several well-studied examples of adaptive speciation focus on how sexual selection causes
vocalizations to diverge, but the call differences observed were among different populations, thus in allopatry (e.g., Hoskin et al.,
2005; Boul et al., 2007). In European fire salamanders, Steinfartz et al. (2007) demonstrated
ecological differentiation (pond vs. stream larvae deposition) driving genetic differentiation
within one forest in western Germany, leading to largely reproductively isolated lineages
that, however, have not yet completed speciation. Wollenberg et al. (2011) found among
mantellid frogs of Madagascar evidence for speciation in geographical proximity, and identified
several young pairs of sister species occurring
in full sympatry, each species characterized by
very small ranges. This fragmentary data indiDOI:10.1163/15685381-00002856
504
cate that the possibility of sympatric, adaptive
speciation in amphibians merits more attention.
Among the pairs of sister species in mantellid frogs flagged by Wollenberg et al. (2011) as
showing a distribution concordant with a putative divergence in sympatry, two small treefrog
species of the genus Boophis are unusual in
that they are genetically less differentiated than
the average of mantellid sister species, yet occur in syntopy, and have drastical differences
in tadpole morphology and advertisement calls
(Schmidt et al., 2008; Vieites et al., 2009; Grosjean et al., 2011). Boophis is a genus of treefrogs
belonging to the Madagascan-Comoran endemic family Mantellidae. It is the sole genus
in the subfamily Boophinae, and most species
are specialized to rainforest habitat. Besides a
clade of basal pond-breeders, Boophis typically
deposit eggs into lotic water where their exotrophic tadpoles develop. Compared with the
representatives of the mantellid subfamily Mantellinae, the genus Boophis is characterized by a
low diversity in reproductive modes (Andreone
et al., 2002; Cadle, 2003; Glaw and Vences,
2006) and karyotypes (Aprea et al., 2004), despite a high number of morphologically cryptic but bioacoustically and genetically distinct
species (e.g., Glaw et al., 2010). Taking into
account the latest species descriptions, Boophis
currently contains 74 nominal species but numerous additional candidate species still await
description (Vieites et al., 2009).
The target species pair of the present study
consists of Boophis majori and a confirmed candidate species that has previously been named
B. sp. aff. majori “Ranomafana” by Glaw and
Vences (2007), B. sp. aff. majori “long calls” by
Schmidt et al. (2008), and B. sp. 35 by Vieites
et al. (2009) and Wollenberg et al. (2011). A detailed analysis of the differentiation of these two
forms in adult morphology, bioacoustics and
genetics is still missing. We here fill this gap
and find congruence of independent taxonomic
characters supporting this candidate species being an independent evolutionary lineage. Consequently we describe it under the name B. nar-
M. Vences et al.
insi sp. n., and will use this name throughout the
paper, although the formal description is found
only at the end of the Results section. Based
on the new data, we argue that these frogs provide an intriguing example of species that possibly have adaptively diverged in sympatry. At
the same time they exemplify the difficulties of
ascertaining such a speculative hypothesis in a
tropical ecosystem, due to the scarcity of basic
data on their general biology and on their past
and present distribution.
Materials and methods
Frogs were collected at night by opportunistic searching
of calling males, using torches and head lamps. Specimens
were euthanized in a chlorobutanol solution, fixed in 95%
ethanol, and preserved in 70% ethanol. Locality information was recorded with GPS receivers. Specimens were deposited in the collection of Université d’Antananarivo, Département de Biologie Animale, Antananarivo (UADBA),
Zoologisches Forschungsmuseum Alexander Koenig, Bonn
(ZFMK), Zoölogisch Museum Amsterdam (ZMA), and
the Zoologische Staatssammlung München (ZSM). FGMV,
FGZC and ZCMV refer to F. Glaw and M. Vences field
numbers and BMNH refers to the Natural History Museum of London. Terminology for biogeographic regions of
Madagascar follows Boumans et al. (2007).
Morphological measurements (in millimetres) were all
done by M. Vences with a digital caliper (precision
0.01 mm) to the nearest 0.1 mm. Used abbreviations are:
SVL (snout-vent length), HW (greatest head width), HL
(head length), ED (horizontal eye diameter), END (eyenostril distance), NSD (nostril-snout tip distance), NND
(nostril-nostril distance), TD (horizontal tympanum diameter), TL (tibia length), HAL (hand length), HIL (hindlimb
length), FOL (foot length), FOTL (foot length including tarsus), FORL (forelimb length), and RHL (relative
hindlimb length). Terminology and description scheme
follow Glaw et al. (2010). Webbing formulae follow
Blommers-Schlösser (1979). Statistical analyses were performed with Statistica software (Statsoft Corp., Tulsa,
USA).
Vocalizations were recorded in the field using different
types of tape recorders (in most cases Sony WM-D6C) and
external microphones (Vivanco EM 238), or with an Edirol
R-09 digital recorder with internal microphones and saved
as uncompressed files. Recordings were sampled (or resampled) at 22.05 kHz and 16-bit resolution and computeranalysed using the software CoolEdit 98. Frequency information was obtained through Fast Fourier Transformation
(FFT; width 1024 points). Spectrograms were obtained at
Hanning window function with 256 bands resolution. Temporal measurements are given as range, with mean ± standard deviation in parentheses. Terminology in call descriptions follows Köhler (2000).
505
New Boophis from Madagascar
Table 1. Primer sequences and PCR conditions used in the present study. PCR conditions start with temperature (in °C) of
each step followed by the time in seconds.
Gene
Primer
name
Sequence (5 → 3 )
Source
Cycling profile
16S
16Sar
CGCCTGTTTATCAAAAACAT
Palumbi et al.
(1991)
{(94°C/120 ), [(94°C/20 ),
(53°C/50 ), (72°C/180 )] × 45,
(72°C/600 )}
16S
16Sbr
CCGGTYTGAACTCAGATCAYGT
RAG1
Amp F2
ACNGGNMGICARATCTTYCARCC
Palumbi et al.
(1991)
see Chiari et al.
(2004)
RAG1
Amp R2
GGTGYTTYAACACATCTTCCATYTCRTA
POMC
POMC
DRVF1
ATATGTCATGASCCAYTTYCGCTGGAA
POMC
POMC
DRVR1
GGCRTTYTTGAAWAGAGTCATTAGWGG
We determined DNA sequences of 22 samples assigned
to Boophis majori or B. narinsi, and 9 samples of two
closely related species, namely B. picturatus (5) and B.
marojezensis (4) (supplementary table S1). Genomic DNA
extractions were carried out following a standard salt extraction protocol (Bruford et al., 1992). Fragments of the
mitochondrial 16S rRNA gene (16S) and two fragments
of nuclear DNA (recombination activating gene 1, RAG1;
and pro-opiomelanocortin, POMC) were amplified via polymerase chain reaction (PCR). The sequences of the primers
used with the respective aligning temperature and cycling
profile are shown in table 1. Reactions were performed in a
final volume of 12.5 μl using the following concentration of
reagents: 0.24 μM of each primer, 200 μM of dNTP, 1xPCR
buffer, and 0.4 units of GoTaq DNA polymerase (Promega,
Mannheim, Germany). PCR products were cleaned with enzymatic purification: 0.15 units of Shrimp Alkaline Phosphatase (SAP) and 1 unit of Exonuclease I (New England Biolabs, Frankfurt am Main, Germany) incubated for
15 min at 37°C followed by 15 min at 80°C. Purified PCR
products were sequenced on an automated DNA sequencer
(Applied Biosystems ABI 3130xl). Sequencing reactions
(10 μl) contained 0.2 or 0.3 μl of PCR product 0.5 μl
of BigDye 3.1 (Applied Biosystems, Darmstadt, Germany)
and 0.3 μM of primer. The mitochondrial fragments were
sequenced using the forward primer while nuclear fragments were sequenced in both directions.
Sequences were edited and aligned in the software
CodonCode Aligner 3.7.1 (Codon Code Corporation, Dedham, MA, USA). All newly determined sequences were
submitted to GenBank (accession numbers JX863575JX863657). A detailed list of voucher specimens and GenBank accession numbers including those compiled from
previous studies, as well as a table assigning haplotypes
to individual samples is available as supplementary tables S1 and S2, and was submitted to the Dryad data repository (doi:10.5061/dryad.qj153). After editing the sequences,
see Chiari et al.
(2004)
Vieites et al.
(2007)
{(94°C/120 ), [(94°C/45 ),
(53°C/50 ), (72°C/180 )] × 45,
(72°C/600 )}
{(95°C/120 ), [(95°C/45 ),
(62°C/50 −1°/cycle),
(72°C/80 )] × 9, [(95°C/45 ),
(52°C/50 ), (72°C/80 )] × 30,
(72°C/600 )}
Vieites et al.
(2007)
alignments of the nuclear gene fragments resulted in lengths
of: 1240 bp for RAG1 and 483 bp for POMC. Haplotypes
of the nuclear genes were determined using the PHASE algorithm (Stephens et al., 2001; Stephens and Sheet, 2005)
implemented in DnaSP software 5.10.3 (Librado and Rozas,
2009). For the 16S analyses we added previously generated
sequences from Vieites et al. (2009) giving a total of 51
sequences of 396 bp in length. We calculated uncorrected
16S p-distances between Boophis majori and Boophis narinsi using MEGA 5 (Tamura et al., 2011). For all gene
fragments Median-Joining haplotype networks (Bandelt et
al., 1999) were constructed using the software Network 4.6
(www.fluxus-engineering.com).
Results
Differentiation in bioacoustic characters
Call recordings of B. majori have been published by Vences et al. (2006) from Maharira
and Ranomafana (see fig. 1 for localities). Call
recordings of B. narinsi were published by the
same authors but wrongly identified as a third
call type of a species referred to as Boophis
sp. aff. rhodoscelis (Ranomafana), on Track 57.
Call types 1 and 2 attributed to this candidate
species were later assigned to B. piperatus, a
species described by Glaw et al. (2010).
These published data were here complemented with additional recordings from various sites in the Ranomafana region (fig. 1) and
506
M. Vences et al.
Figure 1. Distribution map showing all known localities for Boophis majori and B. narinsi. The precise locality at Andringitra
was a forest locally known as Imaloka, close to Ambalamarina village. The map on the right shows the detailed known
distribution of the two species at Ranomafana National Park. Localities at Ranomafana are as follows. B. majori: 1,
Ranomena; 2, Vohiparara; 3, Maharira; 5, Kidonavo. B. narinsi: 4, Andranoroa river; 5, Kidonavo; 6, near Ambatolahy;
7, Sakaroa (type locality). Topography map of Ranomafana courtesy of Brian Gerber. This figure is published in colour in the
online version.
the Andringitra massif. The advertisement calls
analysed originate from six localities. Among
all vocalizations studied, two different advertisement calls are evident, strongly differing in
their temporal parameters, mainly note duration and note repetition rate (fig. 2, table 2).
Calls from Ambalamarina (Andringitra), Ranomena and Maharira (both latter localities in
Ranomafana area) consist of series of short
notes repeated at a high rate, whereas calls from
Sakaroa and Andranoroa (both Ranomafana
area) consist of series of much longer notes, repeated at a distinctly lower rate. Furthermore,
calls from the two latter localities are strongly
pulsatile in nature, with many pulses repeated
at a high rate within notes (app. 300 pulses/s), a
feature not present in the other calls analysed.
Detailed call descriptions are provided in the
species accounts below. The two different types
of calls were heard in syntopy at the locality
Kidonavo (table 2), although recording quality
from this site was rather low and therefore no
sonagrams or oscillograms are shown.
The differences found among the calls studied are far beyond the range of interspecific call
variation in anurans and given that there is at
least one qualitative character and two quantitative parameters without any overlap sorting the
calls, bioacoustics provide a supportive line of
evidence for the presence of two related but separate species occurring in close sympatry.
Differentiation in molecular characters
The mitochondrial haplotype network based on
a fragment of the 16S rRNA gene (396 bp)
shows two distinct mitochondrial lineages and
no haplotype sharing between B. majori and B.
narinsi with 10 mutational steps between them
(fig. 3). The mean uncorrected p-distance between these two lineages is 2.7% (11 substitutions) while mean distances within B. majori
507
New Boophis from Madagascar
Figure 2. Comparative spectrograms and oscillograms of the advertisement calls of populations of Boophis majori and
Boophis narinsi. Names in parentheses denote recording localities. Calls from Sakaroa were recorded from the holotype
ZSM 294/2006.
Table 2. Comparative parameters of advertisement calls of different populations of Boophis majori and Boophis narinsi.
Values refer to regular note series only. Range followed by mean ± standard deviation in parentheses. NRR = note repetition
rate.
Species
Locality
B. majori
B. majori
B. majori
B. majori
B. narinsi
B. narinsi
B. narinsi
Ambalamarina, Andringitra
Maharira, Ranomafana
Ranomena, Ranomafana
Kidonavo, Ranomafana
Sakaroa, Ranomafana
Andranoroa, Ranomafana
Kidonavo, Ranomafana
Note
duration [ms]
Inter-note interval
duration [ms]
NRR
[1/s]
Dominant
frequency [Hz]
19-33 (29 ± 3)
18-40 (30 ± 6)
18-33 (27 ± 5)
23-53 (33 ± 10)
189-246 (216 ± 16)
187-233 (216 ± 17)
189-236 (214 ± 12)
108-130 (115 ± 6)
98-124 (109 ± 8)
107-120 (114 ± 5)
102-205 (133 ± 31)
211-569 (310 ± 101)
277-431 (316 ± 52)
196-280 (233 ± 25)
6.3
7.3
6.9
5.5
1.6-2.1
1.7-2.0
2.1-2.3
3300-3650
3080-3250
3220-3350
3000-3250
2880-3170
3100-3500
2950-3250
and B. narinsi are 0.3% and 0.5% respectively.
All individuals for which either calls were recorded or tadpole morphology was studied were
correctly assigned to their respective haplotype
lineage.
The network based on a fragment of the nuclear RAG1 gene (1240 bp) is concordant with
the mitochondrial network and demonstrates the
absence of haplotype sharing also for this independent marker. The second network based on a
nuclear marker, a fragment of the POMC gene
(483 bp) shows some degree of haplotype sharing between species (fig. 3), but most specimens
had distinct haplotypes in agreement with their
species assignment based on 16S. The disagreement between the two nuclear markers might be
due to the difference in sequence length, given
that the RAG1 fragment is more than twice as
long as the POMC fragment, with a higher probability of diagnostic substitutions to occur.
508
M. Vences et al.
Figure 3. Median-Joining haplotype networks. Branches without numbers correspond to one mutational step. Branches
with more than three mutational steps are not proportional. Numbers close to branches indicate mutational steps. Blue:
B. majori; yellow: B. narinsi; vertical lines: B. picturatus; crossed lines: B. marojezensis. The exact attribution of haplotypes
to specimens is summarized in supplementary table S1. Colours in the nuclear networks are based on the attribution of
specimens on the basis of their mtDNA (16S) haplotype. This figure is published in colour in the online version.
New Boophis from Madagascar
The concordance between two independent
markers (16S and RAG1), and partial concordance with a third independent marker (POMC)
is in agreement with the hypothesis that the
two forms are independent evolutionary lineages with very restricted or absent inter-lineage
gene flow.
Differentiation in morphological characters
and identity of Boophis majori
The congruent differentiation in two independent molecular markers and bioacoustics under conditions of sympatry provides strong
evidence for the existence of two independent
evolutionary lineages that clearly merit recognition as distinct species (de Queiroz, 2007; Padial et al., 2010). This is further confirmed by
their strongly divergent tadpole morphologies
as described by Schmidt et al. (2008) and also
mentioned in Vieites et al. (2009). We have in
the previous sections of this paper already attributed the names B. majori and B. narinsi to
these two species to make it easier to refer to tables and figures herein. However, it still is necessary to provide evidence that the nomen majori is correctly applied, given the high morphological similarity of the two species in their
adult stages.
Boophis majori was described as Rhacophorus majori by Boulenger (1896), based on a
series of syntypes from Ambohimitombo forest. Later, Blommers-Schlösser (1979) designated the male specimen BMNH 1947.2.7.67
as lectotype of the species. Other specimens
(from Mandraka) assigned to the species by
Blommers-Schlösser (1979) were subsequently
transferred to B. marojezensis by Glaw et al.
(2001) who in turn attributed specimens from
the Andringitra massif to B. majori (the same
population for which we analyzed advertisement calls; table 2). To maintain consistency
with Glaw et al. (2001) we have since assigned
the name B. majori to the species with short note
duration (Vences et al., 2006; Glaw and Vences,
2007; Vieites et al., 2009) as typical for the Andringitra population.
509
Unfortunately, no call recordings or fresh tissue materials are available from the type locality
of B. majori, Ambohimitombo, which makes it
impossible to confirm the identity of this species
using bioacoustic or molecular methods. But
even with such material, a confirmation of the
absence of any of the two species in Ambohimitombo would be extremely difficult, considering that they occur sympatrically in the
Ranomafana area and that the precise historical collecting site at Ambohimitombo is unknown. We therefore here rely on morphological and biogeographic arguments to support
that the name Boophis majori has been correctly
applied to the species characterized by short
note duration:
(1) The type locality of B. majori, Ambohimitombo, is at a relatively high elevation of
ca. 1200 m a.s.l. (Glaw et al., 2001). Similarly the collecting locality of the population at Andringitra that we attribute to
this species is located at 1450 m a.s.l. or
higher, our collecting site at Maharira is at
1248 m a.s.l., and the lowest record at Kidonavo is at 1150 m a.s.l. On the contrary,
populations that we attribute to B. narinsi
have been found at maximum elevations of
1150 m a.s.l. (Kidonavo bridge near Vohiparara). Although the differences in altitudinal distribution are weak, the elevation
at the type locality of B. majori appears to
be in slightly better agreement with the elevation of locations where the short-note
species, attributed by us to B. majori, has
been collected.
(2) Some morphological characters also indicate slightly stronger similarities of the B.
majori types to the specimens with a shortnote call: in all individuals of this species,
the hindlimbs are rather short (table 3);
when limbs are adpressed along the body,
the tibiotarsal articulation at most reaches
the snout tip (in the majori types, it only
reaches between eye and nostril as in a
specimen from Andringitra). On the contrary, all measured specimens of the species
Field no.
–
–
–
–
–
–
ZCMV 8510
ZCMV 2583
ZCMV 3308
FGMV 2002.270
ZCMV 2976
ZCMV 2977
ZCMV 2978
ZCMV 2979
ZCMV 2842
Catalogue no.
B. majori
BMNH 1947.2.7.67
BMNH 1947.2.7.66
ZFMK 57394
ZFMK 57395
ZFMK 57396
ZFMK 57397
ZSM 469/2009
ZSM 260/2006
ZSM 240/2006
ZSM 675/2003
B. narinsi
ZSM 294/2006
ZSM 295/2006
ZSM 296/2006
ZSM 297/2006
ZSM 261/2006
Sakaroa
Sakaroa
Sakaroa
Sakaroa
near Ambatolahy
Ambohimitombo
Ambohimitombo
Ambalamarina
Ambalamarina
Ambalamarina
Ambalamarina
near Vohiparara
near Vohiparara
Ranomena
Samalaotra
Locality
HT
PT
PT
PT
PT
LT
PLT
–
–
–
–
–
–
–
Status
M
M
M
M
M
M
F
M
M
M
M
M
M
M
M
Sex
22.7
23.0
22.0
21.4
23.0
23.7
29.2
24.5
23.8
23.2
24.0
21.9
21.9
22.7
21.9
SVL
8.3
8.5
8.4
8.8
8.6
8.5
10.0
8.8
8.4
8.4
8.3
8.3
8.3
8.8
8.2
HW
8.4
8.4
8.8
8.8
8.6
8.8
10.0
9.0
9.0
8.5
8.5
8.5
8.8
8.9
8.5
HL
1.6
1.8
1.9
1.9
1.6
1.6
1.6
1.4
1.5
1.1
1.3
1.6
1.8
1.8
1.8
TD
3.4
3.2
3.5
3.5
3.6
3.3
3.6
3.6
3.4
3.0
3.2
3.2
3.2
3.3
3.3
ED
2.1
1.9
2.0
2.1
2.4
2.0
2.0
1.7
1.8
2.0
1.8
1.8
2.2
1.7
1.8
END
2.1
2.2
2.1
2.1
2.0
1.6
1.7
2.0
1.7
2.0
1.8
2.0
1.8
1.7
1.9
NSD
2.4
2.5
2.6
2.2
2.6
2.1
2.0
2.1
2.2
2.5
2.4
2.2
2.2
2.0
2.5
NND
15.3
16.1
15.6
15.3
16.0
15.1
18.5
16.1
15.3
15.5
14.6
15.1
14.1
15.7
15.4
FORL
7.5
7.6
7.5
7.5
7.7
7.4
8.5
7.5
7.2
7.6
7.6
7.0
7.0
7.4
7.8
HAL
41.3
41.8
42.6
41.7
42.2
39.0
47.7
40.9
41.4
41.0
41.0
36.4
36.8
40.3
29.2
HIL
17.6
18.5
17.6
17.8
17.9
17.0
20.2
17.8
17.0
17.2
17.3
16.4
16.3
17.7
17.3
FOTL
10.1
10.5
10.4
10.4
10.9
10.0
11.9
10.2
10.3
n.m.
10.3
9.8
9.6
10.3
10.3
FOL
12.7
13.2
13.3
13.0
13.3
n.m.
n.m.
n.m.
n.m.
n.m.
n.m.
11.9
12.0
12.6
11.4
TIL
6
6
6
7
6
2
2
3
5
5
2
3
3
3
1
RHL
Table 3. Morphometric measurements (all in mm) of examined voucher specimens of Boophis majori and B. narinsi. For abbreviations of morphometric measurements and collection
acronyms see Materials and methods. Additional abbreviations: HT, holotype; PT, paratype; LT, lectotype; PLT, paralectotype; M, male; F, female. RHL (relative hindlimb length) is coded
as follows: when hindlimb is adpressed along body, tibiotarsal articulation reaches (1) the anterior corner of the eye, (2) between eye and nostril, (3) nostril, (4) between nostril and snout
tip, (5) snout tip, (6) beyond snout tip, (7) well beyond snout tip. Note that ZSM 297/2006 was not sequenced but its identity ascertained through its calls in the field. ZMA and UADBA
paratypes of B. narinsi were not available for analysis during the preparation of this study. FOL of ZFMK 57396 was apparently given wrongly as 16.5 mm in Glaw et al. (2001).
510
M. Vences et al.
511
New Boophis from Madagascar
with a long-note call have the tibiotarsal
articulation reaching beyond the snout tip.
A second possible difference could be in
relative head width: the B. majori lectotype
has a head slightly longer than wide, similar to all eight other males of the species
emitting short notes. On the contrary, in the
species emitting long notes only two out
of five specimens have a head longer than
wide, one has wider than long and two have
equally wide as long heads.
Based on these biogeographic and morphological arguments, and in an effort to minimize
taxonomic change, we continue assigning the
species characterized by short notes in advertisement calls to Boophis majori (Boulenger,
1896). Junior synonyms of Boophis majori do
not exist. Consequently, no name is available for
the species with longer notes in advertisement
calls, and we therefore in the following provide
its formal description.
Boophis narinsi sp. n. (fig. 4)
Holotype
ZSM 294/2006 (field number ZCMV 2976),
adult male, from a place locally known as
Sakaroa, near Talatakely in Ranomafana National Park, south-eastern Madagascar,
21°15.928 S, 47°24.743 E, 1048 m above sea
level, collected on 24 February 2006 by M.
Vences, Y. Chiari, T. and E. Rajeriarison, P.
Bora, and T. Razafindrabe. DNA sequences of
the holotype are deposited in GenBank under
accession numbers FJ559151 (16S), JX863593
(RAG1), JX863650 (POMC).
Paratypes
Five adult males, ZSM 295-297/2006 (field
numbers ZCMV 2977-2979) and UADBA uncatalogued (ZCMV 2972 and ZCMV 2973),
with same locality and collection data as holotype; one adult male, ZSM 261/2006 (ZCMV
2842) from near Ambatolahy village at the
border of Ranomafana National Park (coordinates of precise collection locality not taken
but close to 21°14.632 S, 47°25.573 E, 915 m
above sea level), collected on 21 February 2006
by M. Vences, P. Bora, E. and T. Rajeriarison, T. Razafindrabe, C. Weldon, O. Verneau
and L. du Preez. Five adult males, ZMA
20071 (ZCMV 319), ZMA 20072 (ZCMV
321), ZMA 20073 (ZCMV 322), UADBA
24440 (ZCMV 323), UADBA 24454 (ZCMV
337), from along Andronoroa river near Ranomafanakely, 21°14.872 S, 47°22.580 E,
1138 m a.s.l., collected on 28 January 2004 by
M. Vences and I. De la Riva.
Etymology
The species name is a patronym for Peter M.
Narins, to whom we are pleased to dedicate
this new species in recognition of his inspiring
works on the acoustic communication of anurans.
Diagnosis
Assigned to the genus Boophis based on the
presence of an intercalary element between ultimate and penultimate phalanges of fingers and
toes (verified by external examination), presence of nuptial pads and absence of femoral
glands in males, and overall similarity to other
Boophis species. Within Boophis, the species
can be assigned to the Boophis majori group
based on its small size (male SVL < 25 mm),
absence of green dorsal coloration, ventral
skin not translucent, red colour on webbing
in preservative, absence of a tubercle or spine
at heel and tibiotarsal articulation, and presence of webbing on the hands. The B. majori
group is known to be non-monophyletic but is
a useful phenetic unit to facilitate morphological diagnoses within the species-rich genus
Boophis, and we keep this grouping until a wellsupported phylogenetic hypothesis for the entire
genus becomes available. Within the B. majori
group, the new species is distinguished from
B. arcanus by a uniformly yellow iris (vs. iris
512
M. Vences et al.
Figure 4. Specimens of Boophis narinsi and B. majori in life. (a, b) Holotype (ZSM 294/2006) and (c) paratype (ZSM
295/2006) of Boophis narinsi from Sakaroa in Ranomafana National Park in life (identification of the individuals in these
photographs has been done with a slight uncertainty); (d) adult male (ZCMV 3091) of B. majori during call emission,
photographed at Ranomena in Ranomafana National Park; (e) adult male of B. majori from near Ambalamarina, Andringitra
Massif. This figure is published in colour in the online version.
513
New Boophis from Madagascar
with distinct vertical stripes), smaller body size
(male SVL 21-23 vs. 24.3-26.6 mm), and a low
note repetition rate in advertisement calls (1.62.3 vs. 13.7 notes/s); from B. blommersae by
the presence of red colour ventrally on webbing (vs. absence), and by advertisement call
(a series of slowly repeated notes vs. a fast,
pulsed trill); from B. feonnyala by smaller size
(male SVL 21-23 vs. 25 mm), presence of red
colour on webbing (vs. absence), uniform iris
colour (vs. typically a distinct pattern of some
dark stripes and markings in iris), and advertisement call (note repetition rate 1.6-2.3 vs. 5.3
notes/s); from B. haematopus by smaller size
(male SVL 21-23 vs. 26-32 mm), uniformly
yellowish iris (vs. silvery-beige), and advertisement call (note repetition rate 1.6-2.3 vs. ca. 5
notes/s); from B. marojezensis by presence of
red colour ventrally on webbing (vs. absence),
and by advertisement call (pulsed vs. melodious notes); from B. miniatus by a smaller size
(male SVL 21-23 vs. 23-27 mm), uniformly yellowish iris colour (vs. reddish outer iris area),
and a low note repetition rate in advertisement
calls (1.6-2.3 vs. 11-12 notes/s); from B. picturatus by blue-green colour of iris periphery
(vs. blue), smaller size (male SVL 21-23 vs.
23-33 mm), and advertisement call (note repetition rate 1.6-2.3 vs 2.5-4 notes/s, note duration 187-246 vs. 49-131 ms); from B. piperatus by a smaller size (male SVL 21-23 vs. 2830 mm), rounded snout (vs. slightly pointed),
and advertisement call (with distinct silent intervals between notes in a note series vs. almost no recognizable silent intervals); from B.
pyrrhus by smaller size (male SVL 21-23 vs.
26-32 mm), presence of red colour ventrally on
webbing (vs. absence), uniformly yellowish iris
(vs. silvery-white with blue periphery), and advertisement call (series of pulsed notes vs. series
of notes that start with a pulsed component and
end with a melodious component; note repetition rate 1.6-2.3 vs. ca. 4 notes/s); from B. vittatus by absence of dark dorsolateral stripes (vs.
presence), uniformly yellowish iris (vs. silvery
beige with orange), presence of red colour on
the webbing (vs. absence), and advertisement
call (pulsed notes of 187-246 ms duration vs.
unpulsed notes of 16-65 ms duration).
Furthermore the new species is distinguished
from all other Boophis except B. majori by
a genetic differentiation of >5% uncorrected
pairwise distance in the 16S gene; the distance
to B. majori is 2.5-3.3% in the 16S fragment
studied herein.
The new species is morphologically clearly
most similar to Boophis majori, but differs by
a different advertisement call (longer notes repeated at a rate of 1.6-2.3 vs. 5.5-7.3 notes/s),
tadpole mouthparts (comparatively low degree
of deviation from generalized Boophis tadpole
with one undivided upper keratodont row vs.
transformed upper beak with medial convexity
and all upper tooth rows divided), and possibly
by slightly longer hindlimbs and broader head
(see above and table 3).
Description of the holotype
Adult male, SVL 22.7 mm (fig. 4). For morphometric measurements, see table 3. Body moderately slender; head almost as wide as long,
clearly wider than body; snout rounded in dorsal
view, slightly truncate in lateral view; nostrils
directed laterally, at similar distance from tip
of snout and eye; canthus rostralis slightly concave in dorsal view, loreal region very slightly
concave; tympanum distinct, rounded, TD 47%
of ED; supratympanic fold distinct (more so
on the left side of the head), straight; vomerine odontophores distinct, well-separated in two
small, rounded patches positioned very close
to each other and posteromedial to choanae;
choanae small, rounded. Tongue posteriorly bifid, free. Arms slender, subarticular tubercles
single, round; inner and outer metacarpal tubercles weakly expressed, barely recognizable; fingers with lateral dermal fringes; webbing not
recognizable between fingers 1, 2 and 3; weakly
webbed between fingers 3 and 4; webbing formula: 3e(2), 4(1.5). relative length of fingers
1 < 2 < 4 < 3 (finger 2 distinctly shorter
than finger 4); finger discs enlarged. Hindlimbs
514
slender; tibiotarsal articulation reaching clearly
beyond tip of snout when hindlimb is adpressed along body; lateral metatarsalia separated by webbing; inner metatarsal tubercle
small, indistinct, elongated; no outer metatarsal
tubercle; toes broadly webbed; webbing formula 1(0.75), 2i(1), 2e(0.25), 3i(1.75), 3e(0.25),
4i(1.75), 4e(1.75), 5(0.5); relative length of toes
1 < 2 < 5 = 3 < 4; toe discs enlarged. Skin
smooth on dorsal surfaces, throat, chest, and
ventral surface of thighs, coarsely granular on
belly and especially in cloacal region including
proximal ventral surface of thigh. A tissue sample was removed from the right thigh.
After five years in preservative, ground
colour of the upper surface of the head, dorsum and limbs pink, with small black spots scattered; dark brown patches on the snout around
and above the nostril, above and between the
eyes, in the upper part of the head and in the
tympanic region; a somewhat x-like dark patch
going from the head, at about the eye region, to
the middle of the body; dorsal surface of thigh,
shank, tarsus, and external toe, as well as lower
arm, hand and external finger with distinct dark
crossbands; cloacal region with dark patches
and white spots and tubercles; flanks and ventral surface of the body creamy beige with small
pale brown mottling along the jaw and a pink
patch just below the tip of the jaw; anterior and
posterior surface of the upper arm and anterior
surface of the lower arm creamy beige; anterior
surface of the thigh and tarsus, as well as posterior and ventral surface of the shank clear pink;
ventral and posterior surface of the thigh pink
with brown mottling from the proximal region
till the middle; posterior surface of the tarsus
and external toe dark with some whitish spots.
In life (fig. 4), ground colour of upper surface
of the head, dorsum and legs yellowish brown
with scattered small black spots; snout region
and upper part of the head slightly darker;
a somewhat x-like dark patch going from the
head, at about the eye region, to the middle of
the body; dorsal surface of thigh, shank, tarsus,
and external toe, as well as lower arm, hand and
M. Vences et al.
external finger with dark crossbands; cloacal region whitish, with whitish tubercles and dark
spots; flanks grey with whitish mottling; ventral part grey with a big white patch in the middle, surrounded by whitish spots; throat green
bluish, upper and lower lips whitish; anterior
and posterior surface of the upper arm and anterior surface of the lower arm grey; posterior
surface of the lower arm with a whitish patch
and small dark spots; ventral part of the hand
and feet as well as anterior surface of the thigh
and tarsus and posterior surface of the shank
dark brown; ventral and posterior surface of the
thigh, shank and tarsus dark brown with whitish
spots; outer iris almost uniformly yellow surrounded by a black ring; inner iris ring yellow
brownish, surrounded by an irregular thin black
line.
Variation
All paratypes are very similar to the holotype in
general morphology and coloration. All specimens examined have changed their coloration
in preservative into an intense pinkish-red dorsal ground colour that strongly differs from the
life coloration. For measurements, see table 3.
Larval morphology
The larvae of B. narinsi have been described and
documented in detail by Schmidt et al. (2008).
They are generalized exotrophic tadpoles of
the ranoid type, with few modifications of the
oral disc such as M-shaped posterior keratodont
rows, whereas the tadpoles of B. majori show
additional derived character states such as an
interrupted first anterior keratodont row and a
medial convexity in its upper jaw sheath.
Distribution
The new species is so far known only from the
Ranomafana area, from sites within and directly
adjacent to Ranomafana National Park. As summarized in fig. 1, reliable (genetically identified) specimens are known from the following
515
New Boophis from Madagascar
sites: Andranoroa river near Vohiparara, Kidonavo bridge, Sakaroa, and near Ambatolahy village.
Natural history
Calling males were heard from low and moderately high perches, from about 1 to 3 m
in the vegetation along relatively slow-moving
streams with a partially sandy bottom. At
Sakaroa, specimens were heard calling at a distance of about 5-10 m from the stream, on
bushes near a swampy area. Near Vohiparara,
a chorus was observed in March 1996 at ca.
10 m distance from a slow-moving section of
a stream. After capturing, a gravid female and
a male from this locality (both probably belonging to B. narinsi rather than to B. majori) started
mating and laid a fertile clutch with dark eggs
on 5 March 1996.
Vocalization
Advertisement calls of B. narinsi recorded on
24 February 2006 at the type locality Sakaroa
include calls of the holotype and two further
males, fig. 2, table 2. The call consists of single notes of 189-246 ms (n = 23) duration
repeated at somewhat irregular intervals (211569 ms; n = 23) within long series. Note repetition rate within series varies from 1.6-2.1
notes/s. Within notes, a strong pulsatile character is evident, with pulses being repeated in
very fast succession at a rate of approximately
300 pulses/s. Furthermore, notes show overall
amplitude modulation with highest energy at
the beginning, then decreasing towards the end.
Frequency is mainly distributed between 2000
and 4500 Hz showing some parallel bands of
higher energy (due to fast pulses), with a maximum call energy at 2880-3170 Hz.
Calls recorded on 28 January 2004 at Andranoroa river near Ranomafanakely/Vohiparara
differ only very slightly from those recorded at
the type locality, with note durations of 187233 ms (n = 16), inter-note intervals of 277431 ms (n = 15), a note repetition rate of
1.7-2.0 notes/s, and maximum call energy at
3100-3500 Hz. The latter calls were included
by Vences et al. (2006) as those of Boophis sp.
aff. rhodoscelis (Ranomafana) (CD1, Track 57,
Cut 3).
Calls recorded at the Kidonavo bridge near
Vohiparara on 3-4 March 1996 and originally
considered as another note type of B. majori
were very similar as well, with note durations
of 189-236 ms (n = 25), inter-note intervals of
196-280 ms (n = 24), and a note repetition rate
of 2.1-2.3 notes/s. Note series consisted of up
to >17 notes and had a duration of >8000 ms.
Frequency ranged from 2250-4150 Hz, with the
dominant frequency band at 2950-3250 Hz. The
intensity of each note decreased significantly towards the end. In this population a second note
type was recorded which was emitted either as a
single isolated click note or at the end of a note
series. One such note had a duration of 21 ms
(frequency 2500-3950 Hz, dominant frequency
band 3150-3350 Hz).
Recording temperature of all recordings has
not been precisely measured but can be estimated to range around 20-25°C.
Comparative call data
When comparing the advertisement call of B.
narinsi (fig. 2) to those of its sister species B.
majori (fig. 2), differences are most evident in
temporal parameters, whereas frequency varies
within the same general range in both species.
Calls of B. majori in general consist of single
notes repeated at regular intervals in long series. Note duration in calls of B. majori is distinctly shorter, varying from 18-53 ms among
four populations analysed. Furthermore, note
repetition rate is distinctly higher in call series
of B. majori, with rates of 5.5-7.3 notes/s vs.
1.6-2.3 notes/s in calls of B. narinsi (table 2).
Notes emitted by B. majori are either composed
of 2-3 pulses (populations at Ambalamarina and
Ranomena) repeated at an approximate rate of
90 pulses/s, or pulses are rather indistinct (population at Maharira), but never show the strong
pulsatile nature with many very short and fast
516
repeated pulses as evident in notes of B. narinsi
(fig. 2).
Discussion
Candidate species and stepwise taxonomic
progress
Our study adds one additional species with a
cryptic adult morphology to the taxonomic inventory of Madagascar’s amphibian fauna and
exemplifies how stepwise accumulating evidence can be used to delimit and describe
species in an integrative taxonomy framework
(Padial et al., 2010). Due to the highly conserved adult morphology of B. majori and B.
narinsi, no suspicions about their specific distinctness did initially arise, although we collected the first specimens of B. narinsi as early
as 1996 (specimen ZFMK 62672 from Vohiparara) and noted the existence of different note types in the calls of Boophis majorilike individuals. Only about ten years later,
after collecting trips in 2004 and 2006, we understood that two different mitochondrial lineages of B. majori-like specimens occur in the
Ranomafana area, but the data were insufficient to link these haplotypes to the bioacoustic differences. Based on this preliminary evidence and on constant differences in tadpole
morphology of the two mitochondrial lineages
(Schmidt et al., 2008) we concluded that a confirmed candidate species (due to apparent concordance of independent taxonomic characters)
was present in the Ranomafana area (Vieites et
al., 2009). However, defining a confirmed candidate species is not identical to the delimitation of a distinct species since candidate species
are by definition preliminary units. First, the differences in the taxonomic characters could have
well been explained by the presence of two call
types in a single species, the presence of adaptive plasticity in the tadpoles of a single species,
and by the presence of different deep mitochondrial lineages within a species which is not unlikely given the relatively low amount of diver-
M. Vences et al.
gence (see Fouquet et al., 2007; Hauswaldt et
al., 2011). Second, because detailed analyses of
adult morphology were lacking, the nomenclatural status of the two forms named B. majori
and B. sp. 35 in Vieites et al. (2009) was uncertain. A careful comparison with historical type
specimens and subsequent evaluation of possibly available names is a necessary prerequisite
when assigning a formal Linnean name to a candidate species. In the case of B. narinsi, only
a few years were needed to confirm its species
status. One important evidence that supported
our species hypothesis and consequently accelerated this process was the convincing concordance among independent molecular markers.
In other cases it might take much longer periods before a taxonomic conundrum can be reliably solved, and usage of preliminary candidate
species names can in these cases be a useful tool
to refer consistently, although in a preliminary
way, to the different forms identified.
Endemic amphibians in the Ranomafana area
and their conservation status
The description of Boophis narinsi adds to the
amphibian fauna of Ranomafana National Park
(RNP) one more species which appears to be
microendemic to the Park and its immediate
surrounding. Summarizing recent descriptions
and data of Glaw and Vences (2007), Vieites et
al. (2009) and Strauß et al. (2010), RNP harbors over 110 species and candidate species
of frogs, and of the described species, at least
eight are potential microendemics: Anodonthyla
moramora, A. emilei, Boophis lilianae, B. sandrae, B. piperatus, B. narinsi, Gephyromantis
enki, and G. runewsweeki. It is probable that
several of these species will in the future be discovered in other rainforests of the southern central east or south east of Madagascar, but we
assume that several others will turn out to be
restricted to the area. This highlights the importance of RNP for the conservation of a substantial part of Madagascar’s biodiversity, given
that unprotected forests in the same general area
New Boophis from Madagascar
suffer from ongoing logging and slash-and-burn
agriculture.
For other putative microendemics of the
Ranomafana region such as Gephyromantis
runewsweeki, Anodonthyla emilei and A. moramora a Red List status of Endangered has been
proposed, because these species are known from
fewer than five locations and there is probably continuing decline in the extent and quality of much of their habitat (see Andreone et
al., 2005, 2008 for a general Red List assessment of Madagascar’s amphibian fauna). However, these species are relatively easy to recognize by their characteristic calls and also by
morphology. Their absence from other regions
could therefore be ascertained with some reliability. This is not the case for B. narinsi which
is morphologically similar not only to B. majori
but also to many other species of the B. majori
group (as revised in Glaw et al., 2001). Therefore, we prefer to apply caution and propose a
Red List status as Data Deficient (DD) for this
species, although we assume that its range is indeed restricted to a small part of the southern
central east of Madagascar.
Which data are missing to claim a role for
non-allopatric speciation in tropical
amphibians?
Wollenberg et al. (2011) in their comprehensive
statistical analysis of mantellid sister species
pairs found indications that range sizes of
species included in the sister species comparisons increased with evolutionary age, as did
range size differences between sister species.
Almost all of the youngest species pairs included in the analysis were characterized by
small ranges and small range size differences
(Wollenberg et al., 2011). These data suggest
that particularly microendemic species speciate,
and reject a predominance of peripatric speciation in which a peripheral isolate population
of a widespread species diverges to become
an independent lineage and, hence, a distinct
species. With a male body size below 25 mm
SVL, apparently a restricted distribution area
517
in a narrow latitudinal and altitudinal stretch
of Madagascar’s eastern rainforest, especially
Boophis narinsi fits the pattern of a microendemic species well. The genetic differentiation
between these two species is relatively low compared to other amphibians. Their age of divergence has been estimated to 7.2 million years
ago (Wollenberg et al., 2011), but we speculate that this may be an overestimate due to
“oversmoothing” artefacts during time tree reconstruction and that these taxa can indeed be
considered as relatively young, i.e., of PliocenePleistocene origin.
On the other hand, the high divergence of
these two species in tadpole morphology and
advertisement calls is striking, and much larger
than in other mantellid sister species of comparable genetic distance but fully allopatric distribution. This indicates that non-neutral evolution
in sympatry might have played a role in shaping
the current differentiation of these frogs, either
by adaptive divergence or by character displacement upon secondary contact. Reinforcement
might be plausible for the differentiation in advertisement calls while the differences in tadpole morphology could be explainable by preferences for different breeding sites. Such a putative ecological specialization is supported by the
low incidence of syntopic occurrence, but cannot be easily inferred from the known habitats.
For instance, we have observed B. majori both
along noisy and fast-flowing streams as in Ambalamarina and along very shallow and almost
non-flowing tiny rivulets near Kidonavo.
Thoroughly testing such hypothetical scenarios of adaptive divergence or character displacement/reinforcement in such a poorly known
species pair distributed in a small part of a continuous rainforest band will be, however, extremely difficult. It is clear that the time interval since their initial divergence has provided
numerous occasions for range shifts. Although
such range shifts probably occurred over a small
spatial scale only, they could have created situations of reproductive isolation of populations in
allopatry, even if the populations remained spa-
518
tially proximate to each other. Spatial modelling
could help reconstructing such range shifts but
its application is inhibited by the incomplete
and patchy knowledge on the actual distribution
of these frogs, caused by difficulties accessing
many rainforest areas in Madagascar.
Studying adaptive sympatric or parapatric
speciation in amphibians will require finding
more suitable locations and taxa. Such settings
might be encountered both in temperate and
tropical regions. Certainly, in temperate regions
the early stages of adaptive differentiation can
be more convincingly identified. Due to Pleistocene climatic changes, temperate species often experience re-colonization and range expansions after the last glaciations which have
in several cases erased most of their ancestral polymorphism. Examples such as Canadian and European sticklebacks therefore provide compelling evidence for very fast and parallel within-lake differentiation (McKinnon and
Rundle, 2002). The works of Steinfartz et al.
(2007) indicate that similar processes might be
ongoing in Palearctic salamanders. However, to
discover species of amphibians that have potentially completed the speciation process under sympatric or allopatric conditions, tropical
systems offer a greater potential. Future studies should increasingly attempt to identify pairs
of sister species, or small monophyletic radiations, that are of young age compared with other
amphibian species and range-restricted to particular areas. The cases will be more convincing
if range shifts and immigrations can be reasonably excluded, such as in putatively stable areas (Graham et al., 2006; Carnaval and Moritz,
2008), especially if in small isolated mountain
ranges or islands, and if the species are phenotypically not favored to undergo rapid range expansions. Van Bocxlaer et al. (2010) assessed
for bufonid toads a number of traits associated
with larger distribution areas and thus with the
colonization of vast new areas by these anurans: among others, large body size, large clutch
size, presence of parotoid poison glands, exotrophic larvae, independence from moist con-
M. Vences et al.
ditions, and flexibility in the choice of water bodies for reproduction. Wollenberg et al.
(2011) identified body size as a main predictor of range size in mantellid frogs. Considering
these results, we suggest that small-sized amphibian species of endotrophic development and
dependent on constant water or moist conditions
are not dispersal-prone, and young sister species
with these traits would therefore be ideal models to study adaptive speciation if encountered
sympatrically in isolated areas of high historical
habitat stability.
In the Madagascar model system (Vences et
al., 2009) such a setting might be found in the
isolated rainforest of the Montagne d’Ambre
massif in northern Madagascar. Here, numerous
microendemic amphibian and reptile species
occur, and in several cases these appear to
have diversified genetically within the massif.
This applies to day geckos (Phelsuma dorsivittata; Rocha et al., 2010), ground chameleons
(Townsend et al., 2009), and miniaturized microhylid frogs of the genus Stumpffia (Köhler
et al., 2010). We recommend future studies on
adaptive speciation and diversification in Madagascar, not restricted to amphibians and reptiles,
should target this massif, and we predict that
additional pairs of species or diverging populations will be found that provide convincing evidence rather than just intriguing indications for
adaptive speciation.
Acknowledgements. We are grateful to Ylenia Chiari, Parfait Bora, Ignacio De la Riva, Emile Rajeriarison, Theo
Rajoafiarison, Tokihery Razafindrabe, Axel Strauß, David
R. Vieites, and Katharina C. Wollenberg for their help in
the field. Roger-Daniel Randrianiaina provided crucial data
on tadpoles. Meike Kondermann and Gabriele Keunecke
helped with lab work. Brian Gerber kindly allowed the
use of an elevational map of the Ranomafana area. This
study was made possible by collaboration agreements of
the author’s institutions with the Université d’Antananarivo
Département de Biologie Animale (UADBA) and the Association Nationale pour la Gestion des Aires Protegées.
We are grateful to the staff of UADBA for their continuous support, and to the Malagasy authorities for research and export permits. This research was supported by
grants of the Volkswagen Foundation to MV and FG, of
the Deutsche Forschungsgemeinschaft DFG to MV (grant
New Boophis from Madagascar
number VE247/2-1), and of the Katholischer Akademischer
Austauschdienst (KAAD) to MG.
References
Andreone, F., Cadle, J.E., Cox, N., Glaw, F., Nussbaum,
R.A., Raxworthy, C.J., Stuart, S.N., Vallan, D., Vences,
M. (2005): Species review of amphibian extinction risks
in Madagascar: conclusions from the Global Amphibian
Assessment. Conserv. Biol. 19: 1790-1802.
Andreone, F., Cox, N., Glaw, F., Köhler, J., Rabibisoa,
N.H.C., Randriamahazo, H., Randrianasolo, H., Raxworthy, C.J., Stuart, S.N., Vallan, D., Vences, M. (2008):
Update of the Global Amphibian Assessment for Madagascar in light of new species discoveries, nomenclature
changes, and new field information. Monogr. Mus. Reg.
Sci. Nat. Torino 45: 419-438.
Andreone, F., Vences, M., Guarino, F.M., Glaw, F., Randrianirina, J.E. (2002): Natural history and larval morphology of Boophis occidentalis (Anura: Mantellidae:
Boophinae) provide new insights into the phylogeny and
adaptive radiation of endemic Malagasy frogs. J. Zool.
London 257: 425-438.
Aprea, G., Andreone, F., Capriglione, T., Odierna, G.,
Vences, M. (2004): Evidence for a remarkable stasis of
chromosome evolution in Malagasy treefrogs (Boophis,
Mantellidae). Ital. J. Zool. Suppl. 2: 237-243.
Bandelt, H.-J., Forster, P., Röhl, A. (1999): Median-joining
networks for inferring intraspecific phylogenies. Mol.
Biol. Evol. 16: 37-48.
Barluenga, M., Stoelting, K.N., Salzburger, W., Muschick,
M., Meyer, A. (2006): Sympatric speciation in
Nicaraguan crater lake cichlid fish. Nature 439: 719723.
Blommers-Schlösser, R.M.A. (1979): Biosystematics of the
Malagasy frogs. II. The genus Boophis (Rhacophoridae).
Bijdr. Dierk. 49: 261-312.
Bolnick, D.I., Fitzpatrick, B.M. (2007): Sympatric speciation: models and empirical evidence. Annu. Rev. Ecol.
Syst. 38: 459-487.
Boul, K., Funk, W.C., Darst, C.R., Cannatella, D.C., Ryan,
M.J. (2007): Sexual selection drives speciation in an
Amazonian frog. Proc. Roy. Soc. Lond. B 274: 399-406.
Boulenger, G.A. (1896): Descriptions of new batrachians in
the British Museum. Ann. Mag. Nat. Hist. 17: 401-406,
pl. XVII.
Boumans, L., Vieites, D.R., Glaw, F., Vences, M. (2007):
Geographical patterns of deep mitochondrial differentiation in widespread Malagasy reptiles. Mol. Phylogenet.
Evol. 45: 822-839.
Bruford, M.W., Hanotte, O., Brookfield, J.F.Y., Burke, T.
(1992): Single-locus and multilocus DNA fingerprint.
In: Molecular Genetic Analysis of Populations: A Practical Approach, p. 225-270. Hoelzel, A.R., Ed., IRL Press,
Oxford.
Bush, G.L. (1969): Sympatric host race formation and
speciation in frugivorous flies of the genus Rhagoletis
(Diptera, Tephritidae). Evolution 23: 237-251.
519
Butlin, R.K., Galindo, J., Grahame, J.W. (2008): Sympatric,
parapatric or allopatric? The most important way to
classify speciation. Phil. Trans. Roy. Soc. B 363: 29973007.
Cadle, J.E. (2003): Boophis. In: The Natural History of
Madagascar, p. 916-919. Goodman, S.M., Benstead,
J.P., Eds, University of Chicago Press, Chicago.
Carnaval, A.C., Moritz, C. (2008): Historical climate modelling predicts patterns of current biodiversity in the
Brazilian Atlantic forest. J. Biogeogr. 35: 1187-1201.
Chiari, Y., Vences, M., Vieites, D.R., Rabemananjara, F.,
Bora, P., Ramilijaona Ravoahangimalala, O., Meyer, A.
(2004): New evidence for parallel evolution of colour
patterns in Malagasy poison frogs (Mantella). Mol. Ecol.
13: 3763-3774.
de Queiroz, K. (2007): Species concepts and species delimitation. Syst. Biol. 56: 879-886.
Fouquet, A., Gilles, A., Vences, M., Marty, C., Blanc, M.,
Gemmell, N.J. (2007): Underestimation of species richness in Neotropical frogs revealed by mtDNA analyses.
PLoS One 2: e1109.
Glaw, F., Köhler, J., De la Riva, I., Vieites, D.R.,
Vences, M. (2010): Integrative taxonomy of Malagasy
treefrogs: combination of molecular genetics, bioacoustics and comparative morphology reveals twelve additional species of Boophis. Zootaxa 2383: 1-82.
Glaw, F., Vences, M. (2006): Phylogeny and genus-level
classification of mantellid frogs. Org. Divers. Evol. 6:
236-253.
Glaw, F., Vences, M. (2007): A Field Guide to the Amphibians and Reptiles of Madagascar, 3rd Edition. Vences
and Glaw Verlag, Cologne, 496 pp.
Glaw, F., Vences, M., Andreone, F., Vallan, D. (2001): Revision of the Boophis majori group (Amphibia: Mantellidae) from Madagascar, with descriptions of five new
species. Zool. J. Linn. Soc. 133: 495-529.
Graham, C.H., Moritz, C., Williams, S.E. (2006): Habitat
history improves prediction of biodiversity in rainforest
fauna. Proc. Natl. Acad. Sci. USA 103: 632-636.
Grosjean, S., Randrianiaina, R.D., Strauß, A., Vences, M.
(2011): Sand-eating tadpoles in Madagascar: morphology and ecology of the unique larvae of the treefrog
Boophis picturatus. Salamandra 47: 63-76.
Hauswaldt, J.S., Ludewig, A.-K., Vences, M., Pröhl, H.
(2011): Widespread co-occurrence of divergent mitochondrial haplotype lineages in a Central American
species of poison frog (Oophaga pumilio). J. Biogeogr.
38: 711-726.
Hoskin, C.J., Higgie, M., McDonald, K., Moritz, C. (2005):
Reinforcement drives rapid allopatric speciation. Nature
437: 1353-1356.
Köhler, J. (2000): Amphibian diversity in Bolivia: a study
with special reference to montane forest regions. Bonn.
Zool. Monogr. 48: 1-243.
Köhler, J., Vences, M., D’Cruze, N., Glaw, F. (2010): Giant
dwarfs: discovery of a radiation of large-bodied ’stumptoed frogs’ from karstic cave environments of northern
Madagascar. J. Zool. London 282: 21-38.
Librado, P., Rozas, J. (2009): DnaSP v5: A software for
comprehensive analysis of DNA polymorphism data.
Bioinformatics 25: 1451-1452.
520
Mayr, E. (1982): Processes of speciation in animals. In:
Mechanisms of Speciation, p. 1-19. Liss, A.R.I., Ed.,
Alan R. Liss Inc., New York.
McKinnon, J.S., Rundle, H.D. (2002): Speciation in nature:
the threespine stickleback model systems. Trends Ecol.
Evol. 17: 480-488.
Padial, J.M., Miralles, A., De la Riva, I., Vences, M. (2010):
The integrative future of taxonomy. Front. Zool. 7:
article 16.
Palumbi, S.R., Martin, A., Romano, S., McMillan, W.O.,
Stice, L., Grabowski, G. (1991): The Simple Fool’s
Guide to PCR. Version 2.0. Privately published, Univ.
Hawaii.
Rocha, S., Rösler, H., Gehring, P.-S., Glaw, F., Posada, D.,
Harris, D.J., Vences, M. (2010): Phylogenetic systematics of day geckos, genus Phelsuma, based on molecular and morphological data (Squamata: Gekkonidae).
Zootaxa 2429: 1-28.
Schliewen, U.K., Tautz, D., Pääbo, S. (1994): Sympatric
speciation suggested by monophyly of crater lake cichlids. Nature 368: 629-632.
Schmidt, H., Strauß, A., Reeve, E., Letz, A., Ludewig,
A.-K., Neb, D., Pluschzick, E., Randrianiaina, R.-D.,
Reckwell, D., Schröder, S., Wesolowski, A., Vences,
M. (2008): Descriptions of the remarkable tadpoles of
three treefrog species, genus Boophis, from Madagascar.
Herpetol. Notes 1: 49-57.
Seehausen, O., Terai, Y., Magalhaes, I.S., Carleton, K.L.,
Mrosso, H.D.J., Miyagi, R., van der Sluijs, I., Schneider,
M.V., Maan, M.E., Tachida, H., Imai, H., Okada, N.
(2008): Speciation through sensory drive in cichlid fish.
Nature 455: 620-626.
Steinfartz, S., Weitere, M., Tautz, D. (2007): Tracing the
first step to speciation – ecological and genetic differentiation of a salamander population in a small forest.
Mol. Ecol. 16: 4550-4561.
Stephens, M., Scheet, P. (2005): Accounting for decay
of linkage disequilibrium in haplotype inference and
missing data imputation. Am. J. Hum. Gen. 76: 449462.
Stephens, M., Smith, N.J., Donnelly, P. (2001): A new statistical method for haplotype reconstruction from population data. Am. J. Hum. Gen. 68: 978-989.
Strauß, A., Reeve, E., Randrianiaina, R., Vences, M., Glos,
J. (2010): The world’s richest tadpole communities show
functional redundancy and low functional diversity: ecological data on Madagascar’s stream-dwelling amphibian larvae. BMC Ecol. 10: article 12.
M. Vences et al.
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M.,
Kumar, S. (2011): MEGA5: Molecular evolutionary genetics analysis using maximum likelihood, evolutionary
distance, and maximum parsimony methods. Mol. Biol.
Evol. 28: 2731-2739.
Townsend, T.M., Vieites, D.R., Glaw, F., Vences, M.
(2009): Testing species-level diversification hypotheses
in Madagascar: the case of microendemic Brookesia leaf
chameleons. Syst. Biol. 58: 461-656.
Van Bocxclaer, I., Loader, S.P., Roelants, K., Biju, S.D.,
Menegon, M., Bossuyt, F. (2010): Gradual adaptation
toward a range-expansion phenotype initiated the global
radiation of toads. Science 327: 679-682.
Vences, M., Glaw, F., Márquez, R. (2006): The Calls of the
Frogs of Madagascar. 3 Audio CDs and booklet, 44 pp.
Alosa-Fonozoo, Barcelona.
Vences, M., Wake, D.B. (2007): Speciation, species boundaries and phylogeography of amphibians. In: Amphibian
Biology, Vol. 6, Systematics, p. 2613-2669. Heatwole,
H.H., Tyler, M., Eds, Surrey Beatty and Sons, Chipping
Norton, Australia.
Vences, M., Wollenberg, K.C., Vieites, D.R., Lees, D.C.
(2009): Madagascar as a model region of species diversification. Trends Ecol. Evol. 24: 456-465.
Via, S. (2001): Sympatric speciation in animals: the ugly
duckling grows up. Trends Ecol. Evol. 16: 381-390.
Via, S. (2009): Natural selection in action during speciation.
Proc. Natl. Acad. Sci. USA 106(Suppl. 1): 9939-9946.
Vieites, D.R., Min, M.S., Wake, D.B. (2007): Rapid diversification and dispersal during periods of global warming by plethodontid salamanders. Proc. Natl. Acad. Sci.
USA 104: 19903-19907.
Vieites, D.R., Wollenberg, K.C., Andreone, F., Köhler, J.,
Glaw, F., Vences, M. (2009): Vast underestimation of
Madagascar’s biodiversity evidenced by an integrative
amphibian inventory. Proc. Natl. Acad. Sci. USA 106:
8267-8272.
Wollenberg, K.C., Vieites, D.R., Glaw, F., Vences, M.
(2011): Speciation in little: the role of range and body
size in the diversification of Malagasy mantellid frogs.
BMC Evol. Biol. 11: article 217.
Submitted: April 7, 2012. Final revision received: October
7, 2012. Accepted: October 11, 2012.
Associated Editor: Sebastian Steinfartz.