ArticlePDF Available

Abstract and Figures

With more than 250 species, the Mantellidae is the most species-rich family of frogs in Madagascar. These frogs are highly diversified in morphology, ecology and natural history. Based on a molecular phylogeny of 248 mantellids, we here examine the distribution of three characters reflecting the diversity of eye colouration and two characters of head colouration along the mantellid tree, and their correlation with the general ecology and habitat use of these frogs. We use Bayesian stochastic character mapping, character association tests and concentrated changes tests of correlated evolu-tion of these variables. We confirm previously formulated hypotheses of eye colour pattern being significantly correlated with ecology and habits, with three main character associations: many tree frogs of the genus Boophis have a bright col-oured iris, often with annular elements and a blue-coloured iris periphery (sclera); terrestrial leaf-litter dwellers have an iris horizontally divided into an upper light and lower dark part; and diurnal, terrestrial and aposematic Mantella frogs have a uniformly black iris. Furthermore, the presence of a frenal streak and a dark tympanic patch were associated with each other, with horizontally divided iris colour, and with terrestrial habits. Our study is restricted to the mantellid radia-tion, and the performed tests detect the simultaneous distribution of independent character states in a tree, rather than providing a measure for phylogenetic independent correlation of these character states. The concentrated changes tests suggest that the evolutionary origin of a bright iris might indeed be correlated to arboreal habits. Yet, rather than testing hypotheses of adaptive evolution of eye colour in anurans, our study serves to formulate hypotheses of convergence more precisely and thus to open perspectives for their further testing in a comparative framework along the anuran tree of life. For instance, a brightly coloured iris and sclera might serve mate recognition or as aposematic defensive strategy especially in tree frogs, and a horizontally divided iris colour might constitute a disruptive defensive strategy in frogs inhabiting the leaf litter stratum.
Content may be subject to copyright.
7
Correlates of eye colour and pattern in mantellid frogs
All articles available online at http://www.salamandra-journal.com
© 2013 Deutsche Gesellscha für Herpetologie und Terrarienkunde e.V. (DGHT), Mannheim, Germany
SALAMANDRA 49(1) 7–17 30 April 2013 ISSN 0036–3375
Correlates of eye colour and pattern in mantellid frogs
F A
1
, KC. W
2,3
M V
4
1)
Àrea d‘Herpetologia, Museu de Granollers-Ciències Naturals, Francesc Macià 51, 08400 Granollers, Catalonia, Spain
2)
Department of Biology, School of Science, Engineering and Mathematics, Bethune-Cookman University,
640 Dr. Mary McLeod Bethune Blvd., Daytona Beach, FL 32114, USA
3)
Department of Biogeography, Trier University, Universitätsring 15, 54286 Trier, Germany
4)
Zoological Institute, Division of Evolutionary Biology, Technical University of Braunschweig, Spielmannstr. 8,
38106 Braunschweig, Germany
Corresponding author: M V, e-mail: m.vences@tu-bs.de
Manuscript received: 18 March 2013
Abstract. With more than  species, the Mantellidae is the most species-rich family of frogs in Madagascar. ese frogs
are highly diversied in morphology, ecology and natural history. Based on a molecular phylogeny of  mantellids, we
here examine the distribution of three characters reecting the diversity of eye colouration and two characters of head
colouration along the mantellid tree, and their correlation with the general ecology and habitat use of these frogs. We use
Bayesian stochastic character mapping, character association tests and concentrated changes tests of correlated evolu-
tion of these variables. We conrm previously formulated hypotheses of eye colour pattern being signicantly correlated
with ecology and habits, with three main character associations: many tree frogs of the genus Boophis have a bright col-
oured iris, oen with annular elements and a blue-coloured iris periphery (sclera); terrestrial leaf-litter dwellers have an
iris horizontally divided into an upper light and lower dark part; and diurnal, terrestrial and aposematic Mantella frogs
have a uniformly black iris. Furthermore, the presence of a frenal streak and a dark tympanic patch were associated with
each other, with horizontally divided iris colour, and with terrestrial habits. Our study is restricted to the mantellid radia-
tion, and the performed tests detect the simultaneous distribution of independent character states in a tree, rather than
providing a measure for phylogenetic independent correlation of these character states. e concentrated changes tests
suggest that the evolutionary origin of a bright iris might indeed be correlated to arboreal habits. Yet, rather than testing
hypotheses of adaptive evolution of eye colour in anurans, our study serves to formulate hypotheses of convergence more
precisely and thus to open perspectives for their further testing in a comparative framework along the anuran tree of life.
For instance, a brightly coloured iris and sclera might serve mate recognition or as aposematic defensive strategy especially
in tree frogs, and a horizontally divided iris colour might constitute a disruptive defensive strategy in frogs inhabiting the
leaf litter stratum.
Key words. Amphibia, Anura, Mantellidae, Madagascar, correlated evolution, iris colour, sclera, frenal streak, tympanic
patch.
Introduction
e evolution of amphibians is known to be strongly char-
acterized by homoplasy, and convergence has been demon-
strated in the evolution of characters of external morpho-
lo gy (e.g., O D ) and skeleton (K-
 V , M et al. , F
E ). Convergence has putatively shaped spe-
cies in multiple unrelated lineages (E ) to such
similar general external appearances that distantly relat-
ed species can look virtually identical (e.g., miniaturized
frogs, tree frogs, or leaf litter frogs) and their evolutionary
relationships could only be reliably deciphered with the aid
of molecular phylogenetic tools (e.g., B M-
 , V  M et al. , M 
et al. ). Similar to the situation in some lizards (W-
 , L T), the putative associa-
tion between general habits and morphology has led biolo-
gists to hypothesize whether so-called ecomorphscould
also be dened in frogs, referring to forms with similar
ecology-morphology relationships that appear repeatedly
in the evolution of insular frog radiations (B ).
ese similarities among largely unrelated frogs also ex-
tend to patterns of colouration, a eld that has only recent-
ly received increased attention (e.g., S , H-
 F , S et al. , W
et al. , , R et al. , W  S ,
W M , B et al. , S-
 et al. ,  Let al. , S C-
 , O’N et al. , B  Z ).
8
F Aet al.
Frogs show a remarkable diversity in colouration, includ-
ing instances of colour polymorphism (H
B , B Z ) and the co-occur-
rence of skin toxins and bright colours that oen qualify
as aposematic (S et al. ). Convergent evolu-
tion of amphibian colour is obvious from the occurrence of
very similar patterns in multiple independent clades, e.g.,
midvertebral stripes or dorsolateral bands that are known
to be, in at least some species, inherited by simple Men-
delian genetics (reviewed in H B ).
However, such convergences have so far only been explic-
itly analysed with respect to aposematic colour patterns
of poison frogs (e.g., S et al. , S ,
V et al. , C et al. ). Even less studied is
the colouration of anuran eyes. Many species of frogs have
a remarkably colourful iris that oen is in stark contrast to
the colour of the body (e.g., G  V ).
Madagascars fauna has been agged as an excellent
model to study evolutionary questions (V et al. )
as it contains several species-rich endemic radiations. By
far the largest endemic frog radiation in Madagascar is the
family Mantellidae, which currently contains nominal
species in eight genera (G V , ) plus a
large number of undescribed but already identied candi-
date species (V et al. ), making up a total of over
 species. Mantellids contain a striking diversity not only
of species but also of morphological diversity and adapta-
tions to dierent habitats, ranging from large (>mm)
semiaquatic frogs living in streams to minute (<  mm)
species that mainly inhabit the leaf litter of rainforests. In
the course of their diversication, mantellids have con-
quered such diverse habitats as high mountains (>m
altitude) and the xeric areas of Madagascars south-west,
but their centre of diversity is in the rainforest where
they include arboreal, terrestrial and semiaquatic species,
breeding in ponds, streams, or fully independently from
open water (GV , ). Mantellids are
also very diverse in their colouration, including iris colour
(G  V ). Among the most remarkably col-
oured mantellids are Malagasy poison frogs in the genus
Mantella, which sequester dietary alkaloids in their skin
and have bright, aposematic dorsal colourations, ranging
from bright orange to black-yellow-orange or blue (e.g.,
D et al. ). Many other mantellids have a brownish,
cryptic colour, with or without vertebral or dorso lateral
lines. Some mantellid tree frogs in the genus Boophis have
a bright green dorsal colour with a translucent shade, sim-
ilar to the unrelated Neotropical tree frogs of the family
Centrolenidae. e sister clade to the Mantellidae is the
mainly Asian family Rhacophoridae (V  M et
al. , B et al. , F et al. ), which
mainly comprises arboreal species that are morphological-
ly convergent with many mantellids (B M-
 ).
e diversity of eye colouration in the Mantellidae trig-
gered the informal analysis of G  V () who
observed that bright iris colouration was found mainly in
the largely arboreal genus Boophis, and black eyes mainly
in the aposematic Mantella and proposed some further hy-
potheses of the possibly causal association of certain an-
uran colour patterns with their ecology. At the time, how-
ever, no reliable phylogeny for the Mantellidae was avail-
able, the taxonomy of these frogs was only incompletely
known, and no quantitative analysis of character associa-
tion was carried out. In the meantime the number of new
frog species from Madagascar, mostly mantellids, has sky-
rocketed (K et al. , V et al. ), their ge-
nus-level classication has been revised (GV
), and a comprehensive molecular phylogeny has been
published (W et al. ). In the light of this ad-
vanced state of knowledge, we here aim to rene and more
precisely formulate the hypotheses of character association
of GV () based on explicit Bayesian re-
construction of character evolution along a densely sam-
pled phylogenetic tree of mantellid species.
Materials and methods
Molecular phylogeny
A phylogeny of  of the  described species in the Man-
tellidae plus  undescribed conrmed candidate spe-
cies (V et al. ) was reconstructed based on 
basepairs of mitochondrial DNA ( bp of S rRNA,
bp of cytochrome b, and  bp of cytochrome oxidase
sub unit I). e backbone of the phylogeny (i.e., relation-
ships among subfamilies and genera) was constrained on
the basis of a combined analysis of  species represent-
ing all major mantellid lineages, for a total of  base-
pairs (bp) of fragments of the mitochondrial genes S
rRNA, (bp), S rRNA (two fragments of  bp and
 bp), cytochrome b ( bp), cytochrome oxidase sub-
unit I (bp), and of the nuclear genes rhodopsin exon
(bp) and regulation-activating gene ( bp). For
details of the analyses performed see Wet al.
(). From the resulting time-calibrated Bayesian infer-
ence tree we pruned those taxa for which no or incomplete
ecological and colour data were available, and used this -
nal tree with a total of  species for comparative analyses.
Ecological and colour character coding
Data on the natural history, habitat and habits of mantel-
lid frog species, as well as on their head and eye colour
patterns, was compiled from a large collection of original
live photographs that are to the largest extent reproduced
in GV () (see Supplementary Table S).
Translating ecological and morphological traits into cat-
egorical character states always requires making uncom-
fortable decisions because the complexity of nature rarely
ts perfectly into human-made categories. is particular-
ly refers to the interpretation of the habits and general ecol-
ogy of a given species, where these decisions were some-
9
Correlates of eye colour and pattern in mantellid frogs
times dicult to make. However, we are convinced that in
general our categorization reects true ecological dier-
ences between species. Another source of uncertainty, the
individual variation of colour patterns within species, was
of lesser importance. As previously discussed (e.g., G
 V ), eye colouration appears to be remarkably
constant within species, even between geographically dis-
tant conspecic populations. e same is true for the fre-
nal streak (e.g., V  G ) and to a somewhat
lesser degree also for the dark tympanic patch. Coding of
ecological and morphological characters was as follows:
General ecology and habits: () arboreal, () terrestri-
al, () semiarboreal, () saxicolous, () rheophilous (semi-
aquatic), () terrestrial to rheophilous (riparian).
Detailed iris pattern: () uniform black (also used for
taxa with a small amount of light pigment in the upper half
of the iris), () more or less uniform, any other colour but
black, sometimes with reticulation; () densely reticulated
(dark reticle on a light iris), () annular, divided in an outer
and inner iris colour (not counting a colourful area around
the iris margin, called iris periphery by GV
 and probably constituting the sclera), () horizontally
striped with central dark stripes on both sides lateral to the
pupil, () horizontally divided into two halves of dierent
colour, usually an upper light and a lower dark half.
Iris contrast: () iris of similar tone as body, () iris dark-
er than body, () iris distinctly more colourful.
Colour of iris periphery (sclera): () indistinct, () blue,
() bluish (including light blue to turquoise), () green,
()yellow, () red, () white.
Dark tympanic patch on both sides of the head or broad
dark longitudinal line behind eyes: () present, () absent.
Frenal streak (light streak running from below tympa-
num along upper lip): () present, () absent.
General dorsal colouration: () presumed cryptic, non-
aposematic, () presumed aposematic.
Reconstruction of character evolution
and correlation
We chose Bayesian stochastic character mapping
(H et al. ) to reconstruct character evo-
lution instead of parsimony or maximum likelihood for
several reasons. First, parsimony is an unrealistic method
for fast-changing characters, but also performs poorly on
conservative characters evolving across long time trajecto-
ries, underestimating the number of changes (R
). is latter situation is the case in mantellid frogs,
which separated from other frog lineages circa  mya
(R et al. ). Second, stochastic mapping allows
transitions among states of characters along the branch-
es in phylogenetic trees and evaluates character histories
based on their posterior probabilities (R ).
All analyses were performed using SIMMAP .. so-
ware (B ). Because of the large size of the
group examined, stochastic mapping on a subset of trees
obtained by Bayesian inference was computationally not
feasible. erefore, we conducted the analyses using our
preferred tree (the majority-rule consensus tree from
Bayesian analysis). Correlated evolution among morpho-
logical characters, and between morphological and eco-
logical characters was examined using the pairwise asso-
ciation value d
ij
and overall character correlation D
ij
(see
Supplementary Materials for detailed values obtained) as
described by H et al. () and B
(). Predictive P-values (P) for determining signi-
cance of character state associations and posterior prob-
abilities of relevant nodes in the phylogeny were calculated
by averaging  realizations and  simulations of the
null hypothesis as the probability of observing a value larg-
er than expected by the null model of character independ-
ence. e same scheme was used to obtain the null distri-
butions for d
ij
statistics of state association. To be certain to
only consider highly signicant rather than spurious char-
acter state associations, we applied sequential Bonferroni
correction, taking into account their total number in all 
tests performed (considering both the association tests of
general ecology vs. colour patterns, and among colour pat-
terns, thus following the most conservative approach pos-
sible).
We tested for an evolutionary correlation among ecolo-
gy and colour pattern with the concentrated-changes test
of M () to assess the association of changes
in these two binary characters (see L  E, ).
is test, as implemented in MacClade . (M
M ), uses only binary characters and we
therefore performed it on simplied characters of arbore-
al vs. non-arboreal frogs (counting semi-arboreal species
as non-arboreal), and iris contrast as bright vs. dull (spe-
cies with an iris darker than body counted as dull), as well
as the already binary characters “frenal streakand dark
tympanic patch. is test determines the probability that
various numbers of gains and losses of the dependent char-
acter state (colour pattern) would occur in certain distin-
guished areas of the clade selected (dened by ecology
arboreal vs. non-arboreal), given that a certain number of
gains and losses occur in the whole clade, and given the
null model that changes are randomly distributed among
the branches of the clade.
Results
Evolution of ecology, habitat, eye and head colours
in the Mantellidae
Bayesian analysis of character evolution indicates a com-
plex history of ecological diversication in mantellid frogs.
e ancestral general ecology at the mantellid root could
not be reliably resolved; both a saxicolous state (posteri-
or probability of this character state: PP = .; Supple-
mentary Table S; Fig. A) as observed in some deep man-
tellid clades of low species diversity (Tsingymantis and
Boehmantis), and an arboreal origin (PP = .) received
10
F Aet al.
11
Correlates of eye colour and pattern in mantellid frogs
comparable posterior probabilities. In general terms, most
mantellid genera are relatively uniform and well-dened
regarding their general ecology and habits (Fig. A). Sev-
eral independent evolutionary transitions leading to arbo-
real, semiarboreal, terrestrial and saxicolous habits are re-
constructed within the family (Fig. A). For example, ar-
boreality might have evolved directly from saxicolous an-
cestors in Boophis, from terrestrial forms in Spinomantis,
or through intermediate semiarboreal stages as in Guibe-
mantis. Semiarboreal habits evolved from terrestrial ances-
tors in Blommersia, or from terrestrial-rheophilous ones in
Gephyromantis and Mantidactylus argenteus, in this latter
case clearly supported by the nested position of the species
within its genus. Remarkably, progressive adaptation to
streams has occurred from terrestrial-rheophilous general-
ists to rheophilous specialists four times. Saxicolous habits
appear derived from arboreal and terrestrial-rheophilous
ancestors in three cases.
All mantellids have horizontal pupils, but iris colour
and pattern has been modied multiple times across the
mantellid tree (Fig. B; Supplementary Table S). Uni-
formly black eyes evolved exclusively (and most prob-
ably twice convergently) in the genus Mantella but this
state is not derived from the densely reticulated iris that
is found in some Mantidactylus and Blommersia. Bright
iris colour arranged in an annular pattern evolved con-
vergently in various clades of the genus Boophis, exclud-
ing the pond-breeding subgenus Sahona. A contrasted
iris colour also originated in two other clades of arbore-
al and semi arboreal frogs (Mantidactylus argenteus and
Guibe mantis), although in these species, the iris is clearly
less bright than in most Boophis and has no annular pat-
tern (Fig. C). All Boophis have a brightly coloured iris
periphery (sclera) (Fig. D), and this character state also
evolved in several other, mainly arboreal or saxicolous
clades, e.g., Spino mantis or Guibemantis frogs. Generally,
clades with a brightly coloured iris typically contain spe-
cies with dierent iris periphery colours (blue, bluish, yel-
low, and white). Green sclera, however, were only present
in one clade of Boophis. Of the head colour characters an-
alysed, our reconstruction indicates that a dark tympanic
patch and a frenal streak (which might however be weak-
ly expressed) are ancestral in mantellids, and both these
characters experienced many independent secondary
losses (Fig. and Supplementary Table S). Aposematic
colouration is exclusive to the genus Mantella and is re-
constructed as having been present in the ancestor of this
clade, with two reversals within the clade (Supplementary
Table S).
Correlation between ecology, eye and body colours
in the Mantellidae
As summarized in Table , our analysis supported with
statistical signicance an association of several characters
of eye and head colours with the general habits and ecol-
ogy of mantellid frogs. Test statistics are detailed in Sup-
plementary Materials. e largest dierences in eye colour
were found between arboreal and non-arboreal frogs. A
signicant tendency of iris colour being brighter than the
body was only observed in arboreal frogs whereas in most
other ecological clusters, this association was signicantly
negative. An annular iris pattern was associated with arbo-
real frogs (only occurring in those of the genus Boophis)
and negatively associated with terrestrial-rheophilous and
semiarboreal habits. In contrast, semiarboreal and rheo-
philous frogs shared a prevalence of horizontally divided
eyes. Arboreality was the only ecological state negatively
associated with an indistinctly coloured iris periphery, but
it was positively associated with blue iris periphery colour,
which in turn was negatively associated with all other hab-
its except the saxicolous one. Dark tympanic patches and
a frenal streak were positively associated with terrestriality,
but negatively with stream-bank dwellers (terrestrial-rheo-
philous), and arboreal frogs.
We furthermore identied several instances of associa-
tion of eye colour and head colour patterns (Supplemen-
tary Table S). Aposematic body colouration was associ-
ated with black eyes (only occurring in the Mantella clade).
Frogs with indistinctly coloured sclera had most oen hor-
izontally divided eyes. Blue sclera occurred in frogs whose
iris was brighter than the body, while an iris of similar
brightness as the body occurred in frogs that had indis-
tinctly coloured sclera. Frenal streak and dark tympanic
patch were strongly correlated with each other, and both
were more frequently found in frogs with a horizontally di-
vided or striped iris than expected by chance.
To obtain some rst indications whether these charac-
ter associations would also point to a correlated evolution-
ary origin of the respective character states, we performed
concentrated changes tests of character correlation as im-
plemented in MacClade (using , simulations), which
compare the real data against the null hypothesis that gains
and losses of a character are randomly distributed across
the phylogeny. In this test, only binary characters can be
used and tracing is carried out using parsimony criteria.
Both ecology and iris contrast were therefore simplied
for analysis (see Materials and Methods). Counting only
strictly arboreal frogs in the “Arboreal” category, using the
Le page. Figure 1. One-character evolutionary histories reconstructed through Bayesian stochastic character mapping on a phylogeny
of mantellid frogs (from W et al. 2011). Inset photos show exemplary species for the various character states. (A) General
ecology and habits in mantellid frogs; (B) Iris pattern in mantellid frogs; (C) Iris contrast; (D) Iris periphery colour (area usually
hidden under eyelid, here visible to the right of the iris, probably corresponding to the sclera). Inset photos from top to bottom:
(A)Boophis jaegeri, Gephyromantis cornutus, G. silvanus, Mantidactylus guttulatus, M. femoralis, Aglyptodactylus securifer; (B) Boophis
luteus, Mantella aurantiaca, Guibemantis liber, G. kathrinae, Mantidactylus madecassus, Boophis majori; (C) Mantidactylus betsileanus,
Mantella crocea, Boophis miniatus; (D) Boophis viridis, B. picturatus, Aglyptodactylus securifer, Boophis praedictus, B. madagascariensis,
B. axelmeyeri, Gephyromantis cornutus.
12
F Aet al.
Figure 2. One-character evolutionary histories reconstructed through Bayesian stochastic character mapping on a phylogeny of man-
tellid frogs (from W et al. 2011). Inset photos show exemplary species for the various character states. (A) presence/
absence of frenal streak (as indicated by arrows in the upper three inset photos); (B) presence/absence of dark tympanic patch (as
indicated by arrows in the upper three inset photos). Inset photos from top to bottom: (A) Mantella betsileo, Gephyromantis granulatus,
Boophis rhodoscelis, Mantidactylus melanopleura, Mantella laevigata, Gephyromantis zavona, Boophis arcanus, Mantidactylus aerum-
nalis; (B)Aglyptodactylus madagascariensis, Mantidactylus sp. a. aerumnalis, Boophis doulioti, Laliostoma labrosum, Mantidactylus
sp. a. biporus, Boophis majori.
MINSTATE simulation, the probability of observing, out
of  gains and  loss, of the character state “bright iris, the
observed  and  (dened as more than  and fewer than
), respectively, on branches distinguished by the character
state climbing”, was P < .. e evolutionary origin of
frenal streak and dark tympanic patch was not signicantly
correlated with terrestrial habits:  gains and losses of
the frenal streak were reconstructed of which  and  oc-
curred in terrestrial frogs (P = .) and  gains and  loss-
es of the dark tympanic patch of which  and  occurred in
terrestrial frogs (P = .).
Discussion
Among the more than  species and candidate species
of mantellid frogs (V et al. , W et al.
), most have arboreal habits. According to the character
reconstructions performed herein, this state evolved vari-
ous times and comprises some variability, such as phyto-
telmic breeders adapted to particular plants only or tree
dwellers that reproduce in streams or ponds. In contrast,
strictly rheophilous frogs evolved in only one mantellid
clade (Mantidactylus). Moist rocks and slopes are inhab-
ited by only a few clades, which typically are species-poor,
and they share morphological adaptations with arboreal
species such as terminal toe pads (M et al. ).
eir rather isolated phylogenetic position suggests that
these saxicolous frogs could be relicts of older lineages sur-
viving through competitive exclusion from other frogs in
this rare and marginal habitat (Tsingymantis and Boehman-
tis), or small, regional radiations in areas where such hab-
itats are more common (some Gephyromantis, K-
 et al. ). is pattern in mantellids agrees with
that in other tropical frogs, where arboreal habits are gen-
erally more common than truly rheophilous or saxicolous
habits (I C , D , D ,
P ).
Our study provides evidence for multiple courses of par-
allel evolution of eye and head colour patterns in mantellid
13
Correlates of eye colour and pattern in mantellid frogs
frogs. ese characters are furthermore statistically associ-
ated with the general ecology of mantellid frogs, corrobo-
rating the informal analysis by G  V (). Al-
though we could also show the simultaneous distribution
of morphological and ecological states in the phylogeny,
it must be emphasized that this analysis is purely correla-
tive and can neither prove any causal relationship between
ecology and morphology, nor the convergent evolution of
certain eye and head morphological characters with cer-
tain ecological states.
e concentrated changes test instead provided evidence
that the evolution (not just the occurrence) of bright iris-
es occurred signicantly more oen, given a background
of arboreal vs. non-arboreal habits. is, however, strongly
depends on the binary character coding of the test: semi-ar-
boreal frogs such as Gephyromantis and Blommersia (with-
out bright irises) were scored as terrestrial in this analysis
and bright irises arose almost exclusively, yet several times,
within a single clade of arboreal frogs (Boophis). Although
the separate origins of bright iris colouration within dis-
tinct clades of Boophis, strictly speaking, are phylogeneti-
cally independent events, it still is questionable whether
they should be really counted as such given that, for in-
stance, a genetic basis for colourful irises might have arisen
only once (in the ancestor of Boophis) and could then have
reversed and re-evolved several times within the genus.
Despite the restrictions of these analyses to (i) mantellid
frogs only and (ii) character association rather than con-
vergent evolution, they allow us to formulate and rene hy-
potheses for further testing (P, E ).
In particular the idea that similar processes of natural or
sexual selection might have convergently shaped these
characters, and that the selective pressures on eye and body
colours dier among frogs adapted to dierent lifestyles,
appears attractive for further studies.
Indications that our results might have wider implica-
tions derive from the observation that the observed cor-
relations probably also apply beyond the Mantellidae, as
was in part already discussed by GV ().
()Bright iris colours are likewise found in arboreal frogs
of numerous unrelated clades, as exemplied by prominent
species such as the Neotropical Agalychnis callidryas (Hyli-
dae), or African Leptopelis (Arthroleptidae) with bright red
eyes. An annular iris pattern is found in phylogenetically
unrelated tree frogs such as Rhacophorus baluensis (Rhaco-
phoridae) or Litoria chloris (Hylidae). () e presence of
uniform black eyes in aposematic species is also repeated
in the Neotropical poison frogs (family Dendrobatidae),
which exhibit a striking convergence with Malagasy Man-
tella in diet, ecology and colouration (C et al. ).
() Similar to many mantellids, other unrelated frogs living
in leaf litter have a horizontally contrasted iris pattern and
a large dark tympanic patch (TH ),
as is the case with the Asian Hylarana luctuosa (Ranidae)
or Australian Mixophyes (Myobatrachidae). Together with
obvious similarities in external body proportions among
many of these frogs, this suggests that future attempts to
objectively dene anuran ecomorphological guilds (as in
tadpoles by A J ) or ecomorphs (as
in Anolis lizards; W , L T ,
L ) might lead to novel insights into the evolu-
tion and macroecology of amphibians.
GV () proposed three alternative hy-
potheses for eye colour function in anurans, in addition to
the null hypothesis of no function: () a physiological role
in that it would inuence vision, () a function as preda-
tor deterrent, or () a function as mate recognition signal.
Furthermore, they mention () a possible function of iris
colour supporting a generally cryptic colouration, for in-
stance when a horizontal pattern of the iris is associated
Table 1. Summary of signicant character state associations between general ecology and habits of mantellid species and colour and
pattern of eye and head. For each association, the table shows the character states that have a statistically signicant association (P =
positive or N = negative) with the respective general ecology state aer passing a Bonferroni correction over all tests. “Positiveindi-
cates that the two states occur together at a higher frequency than expected by chance, while “negativeindicates they occur together
at a lower frequency than expected by chance. For detailed d
ij
pairwise statistics see Supplementary Materials Table S3.
Arboreal Terrestrial Saxicolous Semiarboreal Rheophilous Terrestrial rheophilous
Iris contrast P: Brighter than
body
N: Similar to body
N: Brighter than
body
N: Brighter than
body
P: Similar to body
N: Brighter than body
Detailed iris pattern P: Annular.
N: Horizontally
divided
P: Horizontally
divided
N: Annular
P: Horizontally
divided
N: Annular,
Horizontally divided
Iris periphery colour P: Blue
N: Indistinct
N: Blue P: Indistinct
N: Blue
N: Blue P: Indistinct
N: Blue
Dark tympanic patch P: Absent
N: Present
P: Present
N: Absent
P: Absent
N: Present
Frenal streak P: Absent
N: Present
P: Present
N: Absent
P: Absent
N: Present
14
F Aet al.
with a ank-dorsum colour dierence. In the absence of
experimental results it is not possible to reliably discrimi-
nate among these alternatives, although some appear to be
more probable than others at present.
Sexual selection has been hypothesized to inuence eye
colouration in humans (F ) and birds (S
), but these hypotheses have remained little investi-
gated (e.g., H  MG). A functional associa-
tion has been presumed with image sharpness: light irises
may allow more light to reach the retina, which may re-
duce the sharpness of images relative to dark irises (H 
MG). We consider it likely that bright iris col-
ours in tree frogs indeed have a function in mate recog-
nition and that they evolve under the inuence of sexual
selection, which is also supported by the fact that in tree
frogs with bright-coloured irises, the eye colour is usually
more stable and species-specic than their body coloura-
tion (G V ). C  H () did
not nd a correlation of bird iris colouration with social
behaviour, but these authors emphasized that critical in-
formation was missing for many species in their analysis.
In arboreal frogs with a bright eye colour, several colours
are usually present in dierent species. Visual signalling is
mainly known from frogs living along noisy streams, but
also from some tree frogs, and might play a role in individ-
ual recognition (H  A ). e conspicu-
ousness of iris colour is magnied by a division into circu-
lar areas with dierent pigmentations in annular-patterned
eyes. is might favour species recognition in species-rich
frog communities, considering that in Madagascar, around
 species of Boophis are known to co-occur in some areas
(V et al. ). Assuming a sensory-driven role of
iris colour in mantellid speciation as known in cichlid sh-
es (M et al. , S et al. ) appears to be
far-fetched. However, such a mechanism, i.e., a divergent
evolution of the visual system associated with dierences
in colour (in this case of the iris) and colour preference
biasing mate choice, should not be a priori disregarded
when designing future studies on anuran eye colouration.
A better understanding of the intraspecic variation of iris
colour is also necessary. While in birds, some groups such
as Ploceus weavers with sexually dimorphic iris colour are
known (C H ), no such pattern has thus
far been observed in mantellids or other frogs (G
V ). On the contrary, ontogenetic changes that
are common in birds (C H, N
et al. ) are probable (although not studied in detail) in
anurans.
T  H () assign an adaptive function
as disruptive colouration to the horizontally divided iris
pattern and assume that it contributes to breaking up the
frog’s shape. is hypothesis seems straightforward, also
because such a pattern, in Madagascar and elsewhere, is
oen found in partly diurnal leaf litter frogs, which oen
are light brown dorsally but dark on the anks, and in these
cases, the upper light iris colour prevents that dark eyes
make a frog silhouette stand out in dorsal view. Similar mi-
metic functions might also be hypothesized for the dark
tympanic patch of many terrestrial frogs, while the frenal
streak, which we found being statistically associated to this
patch, might play a role in intraspecic communication as
in some mantellids, it has a divergent expression in closely
related sympatric species such as Mantidactylus melano-
pleura and M. opiparis (V  G ).
To better understand selective pressures and the func-
tion of eye colour in frogs and other vertebrates, it will
also be crucial to understand its genomic and genetic ba-
sis. Amphibian pigment cells are located in the epidermis,
but are well known not to be ectodermal but derived from
the neural crest. In the vertebrate eye, mesenchymal cells
(head mesoderm and neural crest cells) form the iris stro-
ma as well as other structures such as the corneal ento-
derm, structures at the iridocorneal angle, and ciliary body
stroma (S L ). us, the pigment cells both
in the iris and the body skin are derived from the neural
crest. However, either the genetic regulation of cell dier-
entiation seems to be dierent between eye and skin, or the
migration of their precursor cells during embryonic devel-
opment might be dierently aected. is is obvious from
semialbinistic frogs that oen lack pigments in the body
but have normally pigmented eyes (G  V ).
Given that iris colour in humans and other primates is in-
uenced by numerous loci (e.g., B et al. , L
et al. ), it is probable that the colours of iris and scle-
ra are determined by a complex interplay of various genes
and gene regulatory mechanisms also in amphibians.
Conclusion and outlook
e evolutionary history of ecological diversication in
mantellids is complex as indicated by character mapping
on a well-resolved phylogeny of these frogs. Also iris con-
trast and pattern, colour of sclera, and several conspicu-
ous colour patterns on the head have evolved and reversed
multiple times within the Mantellidae, and most of these
characters have certain states signicantly associated both
with each other and the general ecology of the frogs. In
particular, a brightly coloured iris and iris periphery with
annular pattern was mainly observed in Boophis tree frogs
whereas a frenal streak and dark tympanic patch were as-
sociated with terrestrial habits. Experimental behaviour-
al studies are needed to understand the function of these
traits. In particular, the role of bright eye colour in either
intraspecic communication or predator deterrence is
worth further analysis. Similar analyses of character asso-
ciation in other major clades of frogs could clarify whether
not only the same character states but also the same char-
acter state associations evolved convergently in independ-
ent anuran radiations. Additionally, a better understanding
of the underlying developmental genetics might allow dis-
entangling the evolutionary processes that inuence pig-
mentation in tree frogs. As a testable hypothesis, natural
selection might act more strongly on body colour while
sexual selection, at least in some groups, might act more
strongly on eye colour.
15
Correlates of eye colour and pattern in mantellid frogs
Acknowledgements
We are grateful to the Département de Biologie Animale of the
University of Antananarivo for its continuous support and col-
laboration, and to the Madagascar National Parks board as well
as the Malagasy Ministère des Eaux et Fôrets for issuing permits
over the past years that made data collection for this study pos-
sible.
References
A, R. G. G. J (): Guilds of anuran larvae: re-
lationships among developmental modes, morphologies, and
habitats. − Herpetological Monographs, : –.
B, R. C. K. R. Z (): Sexual dichromatism in
frogs: natural selection, sexual selection and unexpected di-
versity. − Proceedings of the Royal Society B, : –.
B, J. P. (): SIMMAP: Stochastic character mapping
of discrete traits on phylogenies. − BMC Bioinformatics, : ar-
ticle .
B, F. M. C. M (): Convergent adap-
tive radiations in Madagascan and Asian ranid frogs reveal co-
variation between larval and adult traits. Proceedings of the
National Academy of Sciences of the U.S.A., : –.
B, F., R. M. B, D. M. H, D. C. C
M. C. M (): Phylogeny and biogeography of
a cosmopolitan frog radiation: Late Cretaceous diversication
resulted in continent-scale endemism in the family Ranidae.
Systematic Biology, : –.
B, B. J., A. P  N. I. M (): Blue eyes in
lemurs and humans: Same phenotype, dierent genetic mech-
anism. American Journal of Physical Anthropology, :
–.
B, R. M. (): Frogs. pp. – in: Gillespie, R. G. &
D. A. Clague (eds.): Encyclopedia of Islands. Berkeley and
Los Angeles, California. University of California Press.
B, J. L., M. M, M. C K. S ():
Selection on color in Panamanian poison frogs: a coalescent
approach. − Journal of Biogeography, : –.
C, Y., M. V, D. R. V, F. R, P.
B, O. R R  A. M
(): New evidence for parallel evolution of colour patterns
in Malagasy poison frogs (Mantella). − Molecular Ecology, :
–.
C, V. C., C. J. R, V. R, P. S
B. L. F (): Convergent evolution of chemical de-
fense in poison frogs and arthropod prey between Madagascar
and the Neotropics. Proceedings of the National Academy of
Sciences of the U.S.A., : –.
C, A.J. F. K. P. E. H (): Iris colour in passerine
birds: why be bright-eyed? − South African Journal of Science,
: –.
D, J. W., N. R. A, M. A
C. W. M (): Madagascan poison frogs (Mantella)
and their skin alkaloids. − American Museum Novitates, :
–.
D, I. (): Resource use and foraging tactics in a south Indian
amphibian community. − Journal of South Asian Natural His-
tory, : –.
D L, G., W. H A. A (): Colour, size, and
movement: the role of visual stimuli in species recognition by
males of the frog Allobates femoralis. − Animal Behaviour, :
–.
D, W. E. (): e biology of an equatorian herpeto-
fauna in Amazonian Ecuador. − Miscellaneous publications of
the Natural History Museum of the University of Kansas, :
–.
E, S. B. (): Convergence and morphological con-
straint in frogs: variation in postcranial morphology. − Fieldi-
ana, Zoology, : –.
E, S. B. (): Vertebrate secondary sexual characteris-
tics: physiological mechanisms and evolutionary patterns.
e American Naturalist, : –.
F, M. S. E (): Parallelism and convergence
in anuran fangs. − Journal of Zoology, : –.
F, D. R., T. G, J. F, R. H. B, A. H, C.
F. B. H, R. D S, A. C, M. W, S. C.
D, C. J. R, J. A. C, B. L. B-
, P. M, R. C. D, R. A. N, J. D. L, D.
M. G W. C. W (): e amphibian tree of
life. Bulletin of the American Museum of Natural History,
: –.
F, P. (): European hair and eye color: A case of frequen-
cy-dependent sexual selection? Evolution and Human Be-
haviour, : –.
G, F. M. V (): A review of anuran eye coloura-
tion: denitions, taxonomic implications and possible func-
tions. pp. – in: Böhme W., Bischo W. & T. Ziegler
(eds.): Herpetologica Bonnensis. Societas Herpetologica Eu-
ropaea, Bonn, Germany.
G, F.  M. V (): Phylogeny and genus-level classi-
cation of mantellid frogs. − Organisms Diversity and Evolu-
tion, : –.
G, F. M. V (): A Field Guide to the Amphibi-
ans and Reptiles of Madagascar ird Edition. − Vences, M. &
Glaw, F. Verlag, Köln,  pp.
H, M.  A. F (): Correlated evolution of con-
spicuous coloration and body size in poison frogs (Dendro-
batidae). − Evolution, : –.
H, G. E. K. J. MG (): Bird coloration: function and
evolution. − Harvard University Press, Harvard.  pp.
H, E. A.  M. S. B (). A review of colour and
pattern polymorphism in anurans. Biological Journal of Lin-
nean Society, : –.
H, D. L.  A. A (): Visual signaling in an-
uran amphibians. pp. – in R, M. J. (ed.): Anuran
communication. Washington, DC. Smithsonian Institution
Press.
H, J. P., R. N  J. P. B (): Stochas-
tic mapping of morphological characters. − Systematic Biolo-
gy, : –.
I, R. F.  R. K. C (): Organization of contiguous
communities of amphibians and reptiles in ailand. Eco-
logical Monographs, : –.
K, N., K. C. W, J. K, F. G, D.
R. V  M. V (): Molecular phylogeny and bio-
geography of Malagasy frogs of the genus Gephyromantis.
Molecular Phylogenetics and Evolution, : –.
16
F Aet al.
K, M.  M. V (). Terminal phalanges in ra-
noid frogs: morphological diversity and evolutionary correla-
tion with climbing habits. − Alytes, : –.
K, J., D. R. V. R. M. B, F. H G, F.
G, D. S  M. V (): New amphibians and
global conservation: a boost in species discoveries in a highly
endangered vertebrate group. – Bioscience, : −.
L, P. D. J.M. E (): Power of the concentrated
changes test for correlated evolution. − Systematic Biology, :
−.
L, J. B. (): Lizards in an evolutionary tree. Ecology and
adaptive radiation of anoles. University of California Press.
Berkeley, Los Angeles & London.  pp.
L, J. B. R. S. T(): Evolutionary diversication
of Caribbean Anolis lizards. pp. – in: Dieckmann U.,
M. J.Doebeli, A. J. Metz & D. Tautz (eds.): Adaptive speciation.
– Cambridge University Press. Cambridge.
L, F., K.  D, K. V, J. R. H, A. U-
, A. G. J A. C. J. W. M. K (): Eye
color and the prediction of complex phenotypes from geno-
types. − Current Biology, : R–R.
M, W. (): A method for testing the correlated evolu-
tion of two binary characters: Are gains or losses concentrated
on certain branches of a phylogenetic tree? Evolution, :
−.
M, W. P. D. R. M (): MacClade: analysis
of phylogeny and character evolution, version .. − Sunder-
land, Massachusetts, Sinauer.
M, M. E., K. D. H, J. J. M.  A O. S
(): Sensory drive in cichlid speciation. e American
Naturalist, : –.
M, A., M. F M. V (): Intercalary el-
ements, tree frogs, and the early dierentiation of a complex
system in the Neobatrachia. e Anatomical Record, :
–.
M, R. L., J. R. M, M. J, D. B. W J. L. B
(): Morphological homoplasy, life history evolution, and
historical biogeography of plethodontid salamanders inferred
from complete mitochondrial genomes. Proceedings of the
National Academy of Sciences of the U.S.A., : –.
N, D. M. M. A. S. A (): Iris colour as an in-
dicator of age feature in female Brazilian tanagers (Passeri-
formes: Emberizidae) conrmed by a molecular sexing tech-
nique. − Revista de Biologia Tropical, : –
O, A. A. D (): Démonstration de l’origine in-
dépendante des ventouses digitales dans deux lignées phylo-
génétiques de Ranidae (Amphibiens, Anoures). Comptes
Rendus de lAcademie des Sciences du Paris, : –.
O’N, E. M., K. H. B M. E. P (): Cast
adri on an island: introduced populations experience an al-
tered balance between selection and dri. Biology Letters,
: –.
P, J. R. (): Trophic ecology of a tropical anuran as-
semblage. − Scientic Papers of the Natural History Museum
of the University of Kansas, : –.
P, R. O. (): Phylogenetic tests of alternative intersexual
selection mechanisms: trait macroevolution in a polygynous
clade (Aves: Pipridae). − American Naturalist, : –.
R, C., J. S, M. S, C. S W. H
(): Turning blue and ultraviolet: sex-specic colour
change during the mating season in the Balkan moor frog.
Journal of Zoology, : –.
R, K., D. J. G, M. W, S. P. L., S. D.
B, K. G, L. M  F. B (): Global
patterns of diversication in the history of modern amphibi-
ans. Proceedings of the National Academy of Sciences of the
U.S.A., : –.
R, F. (): Bayesian inference of character evolution. −
Trends in Ecology and Evolution, : –.
S, J. C.  D. C. C (): Phenotypic integration
emerges from aposematism and scale in poison frogs. − Pro-
ceedings of the National Academy of Sciences of the U.S.A.,
: –.
S, R. A., R. Z, M. R, G. G. G M.
A. D (): Experimental evidence for aposema-
tism in the dendrobatid poison frog Oophaga pumilio. Co-
peia, : –.
S, O., Y. T, I. S. M, K. L. C, M.
D. J. M, R. M, I.   S, M. V. S,
M. E. M, H. T, H. I  N. O (): Spe-
ciation through sensory drive in cichlid sh. Nature, :
–.
S, A., T. W. C, E. L, R. V K. S
(): Interspecic and intraspecic views of color signals in
the strawberry poison frog Dendrobates pumilio. − Journal of
Experimental Biology, : –.
S, N. G. (): Visual isolation in gulls. − Scientic Ameri-
can, , –.
S, K. A.  B. A. L (): Morphogenesis of the anterior
segment in the zebrash eye. − BMC Developmental Biology,
: article .
S, K. (): Convergent evolution of bright coloration
and toxicity in frogs. − Proceedings of the National Academy
of Sciences of the U.S.A., : –.
S, R., R. S  K. S (): Molecular phylo-
genetic evidence for a mimetic radiation in Peruvian poison
frogs supports a Müllerian mimicry hypothesis. Proceed-
ings of the Royal Society of London, Series B, : –.
S, M., C. S, A. B, C. R W. H
(): Chin up – are the bright throats of male common frogs
a condition-independent visual cue? − Animal Behaviour, :
–.
T, L. F. C. F. B. H (): Colors and some mor-
phological traits as defensive mechanisms in anurans. In-
ternational Journal of Zoology, : article ID , –.
doi:.//.
V  M, A., M. V, S. H  A. M ():
A previously unrecognized radiation of ranid frogs in south-
ern Africa revealed by nuclear and mitochondrial DNA se-
quences. Molecular Phylogenetic and Evolution, : –
.
V, M. F. G(): Revision of the subgenus Chono-
mantis (Anura: Mantellidae: Mantidactylus) from Madagas-
car, with description of two new species. − Journal of Natural
History, : –.
17
Correlates of eye colour and pattern in mantellid frogs
V, M., J. K, R. B, C. F. B. H, E. L
M, S. L M. V (): Convergent evolu-
tion of aposematic coloration in Neotropical poison frogs: a
molecular phylogenetic perspective. − Organisms Diversity
and Evolution,: –.
V, M., K. C. W, D. R. V D. C. L
(): Madagascar as a model region of species diversica-
tion. Trends in Ecology and Evolution, : .
V, D. R., K. C. W, F. A, J. K, F.
GM. V (): Vast underestimation of Mada-
gascar’s biodiversity evidenced by an integrative amphibian
inventory. − Proceedings of the National Academy of Sciences
of the U.S.A., : –.
W, I. J.  H. B. S (): Rapid color evolution in an
aposematic species: a phylogenetic analysis of color variation
in the strikingly polymorphic strawberry poison-dart frog.
Evolution, : –.
W, K. D. (): e ecology and behavior of amphibians.
University of Chicago Press. Chicago.  pp.
W, E. E. (): Ecomorphs, faunas, island size, and di-
verse end points in island radiations of Anolis. pp. –
in: Huey R. B., E. R. Pianka & T. W. Schoener (eds.): Lizard
ecology. Harvard University Press. Cambridge, MA.
W, K. C., F. G, A. M M. V ():
Molecular phylogeny of Malagasy reed frogs, Heterixalus, and
the relative performance of bioacoustics and color-patterns
for resolving their systematics. − Molecular Phylogenetics and
Evolution, : –.
W, K. C., D. R. V, F. G  M. V ():
Speciation in little: the role of range and body size in the diver-
sication of Malagasy mantellid frogs. BMC Evolutionary
Biology, , article .
W, K. C., S. L, C. M-F M. V
(): Disentangling composite colour patterns in a poison
frog species. Biological Journal of the Linnean Society, :
–.
W, K. C. G. J. M (): Why colour in sub-
terranean vertebrates? Exploring the evolution of colour pat-
terns in caecilian amphibians. Journal of Evolutionary Bio-
lo gy, : –.
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Online Supplementary Material
F A, KC. W M V (2013): Correlates of eye colour and pattern in
mantellid frogs. - Salamandra, 49(1): 7–17.
 Supplementary tables
 Supplementary gures
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Figure S1. One-character history reconstructed through stochastic character mapping of presumed adaptive functions
of colouration: aposematic and cryptic. Reconstruction of cryptic ancestor, PP=0.999.
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Table S1. Matrix of species, ecology, habitat, eye morphology variables and body colour characters in the mantellid
frogs used in this study: ECO, general ecology and habits; DIP, detailed iris pattern; ICB, iris colour brightness; CIP, iris periphery
(sclera) colour; FS, frenal streak; TPS, dark tympanic patch or streak; COL, general dorsal colouration. State number follows material
and methods denitions. Names of species in second column refer to those used in the supplementary gures S2-S7.
Species name
(W et al. 2011)
Species name
(GV 2007)
ECO DIP ICB CIP FS TPS COL
Aglyptodactylus laticeps Aglyptodactylus laticeps 1 3 1 6 1 1 0
Aglyptodactylus madagascariensis Aglyptodactylus madagascariensis 1 3 0 6 1 1 0
Aglyptodactylus securifer Aglyptodactylus securifer 1 3 0 6 1 1 0
Aglyptodactylus sp 2 Aglyptodactylus sp. a. madagascariensis east 1 3 0 6 0 1 0
Aglyptodactylus sp 3 Aglyptodactylus sp. a. madagascariensis Ranomafana 1 3 0 6 0 1 0
Blommersia blommersae Blommersia blommersae 3 2 0 3 1 1 0
Blommersia domerguei Blommersia domerguei 3 3 0 3 1 1 0
Blommersia grandisonae Blommersia grandisonae 3 2 0 3 1 1 0
Blommersia kely Blommersia kely 3 4 0 3 1 1 0
Blommersia sarotra Blommersia sarotra 3 3 0 3 1 1 0
Blommersia sp. 2 Blommersia sp. a. blommersae Maroantsetra 3 5 0 3 1 1 0
Blommersia sp. 1 Blommersia sp. a. blommersae Nosy Boraha 3 5 0 3 1 1 0
Blommersia sp. 3 Blommersia sp. a. blommersae Toamasina 3 5 0 3 1 1 0
Blommersia sp. 5 Blommersia sp. a. wittei west 3 3 0 3 1 1 0
Blommersia sp. 4 Blommersia sp. Comoros 3 3 0 3 1 1 0
Blommersia wittei Blommersia wittei 3 3 0 3 1 1 0
Boehmantis microtympanum Boehmantis microtympanum 2 5 0 6 0 0 0
Boophis albilabris Boophis albilabris 0 5 0 5 1 0 0
Boophis albipunctatus Boophis albipunctatus 0 5 0 0 0 0 0
Boophis andohahela Boophis andohahela 0 3 1 0 0 0 0
Boophis andreonei Boophis andreonei 0 0 2 6 0 0 0
Boophis anjanaharibeensis Boophis anjanaharibeensis 0 0 2 0 0 0 0
Boophis ankaratra Boophis ankaratra 0 0 0 0 0 0 0
Boophis axelmeyeri Boophis axelmeyeri 0 3 0 5 0 0 0
Boophis blommersae Boophis blommersae 0 5 0 6 0 0 0
Boophis boehmei Boophis boehmei 0 0 2 0 0 0 0
Boophis bottae Boophis bottae 0 0 2 0 0 0 0
Boophis brachychir Boophis brachychir 0 5 0 6 0 1 0
Boophis burgeri Boophis burgeri 0 0 0 0 0 0 0
Boophis doulioti Boophis doulioti 0 4 2 6 1 1 0
Boophis elenae Boophis elenae 0 5 2 0 0 0 0
Boophis englaenderi Boophis englaenderi 0 0 2 0 0 0 0
Boophis erythrodactylus Boophis erythrodactylus 0 0 2 0 0 0 0
Boophis feonnyala Boophis feonnyala 0 5 0 0 0 0 0
Boophis goudoti Boophis goudoti 0 5 0 0 0 0 0
Boophis guibei Boophis guibei 0 3 2 0 0 0 0
Boophis haematopus Boophis haematopus 0 5 0 0 0 0 0
Boophis idae Boophis idae 0 5 0 6 0 0 0
Boophis jaegeri Boophis jaegeri 0 0 2 0 0 0 0
Boophis laurenti Boophis laurenti 0 5 0 2 0 0 0
Boophis liami Boophis liami 0 5 2 0 0 0 0
Boophis lichenoides Boophis lichenoides 0 5 0 0 0 0 0
Boophis luteus Boophis luteus 0 0 2 0 0 0 0
Boophis madagascariensis Boophis madagascariensis 0 0 0 6 0 0 0
Boophis majori Boophis majori 0 5 2 0 0 0 0
Boophis mandraka Boophis mandraka 0 5 2 0 0 0 0
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Species name
(W et al. 2011)
Species name
(GV 2007)
ECO DIP ICB CIP FS TPS COL
Boophis marojezensis Boophis marojezensis 0 5 2 0 0 0 0
Boophis microtympanum Boophis microtympanum 0 5 0 2 0 1 0
Boophis miniatus Boophis miniatus 0 5 2 0 0 0 0
Boophis occidentalis Boophis occidentalis 0 0 0 0 0 0 0
Boophis opisthodon Boophis opisthodon 0 5 0 6 1 1 0
Boophis pauliani Boophis pauliani 0 5 0 1 0 0 0
Boophis periegetes Boophis periegetes 0 5 0 6 0 0 0
Boophis picturatus Boophis picturatus 0 0 0 0 0 0 0
Boophis pyrrhus Boophis pyrrhus 0 5 2 5 0 0 0
Boophis rappiodes Boophis rappiodes 0 0 2 0 0 0 0
Boophis reticulatus Boophis reticulatus 0 0 0 6 0 1 0
Boophis rhodoscelis Boophis rhodoscelis 0 5 0 0 0 1 0
Boophis ruoculis Boophis ruoculis 0 0 2 6 0 0 0
Boophis sambirano Boophis sambirano 0 5 2 0 0 0 0
Boophis schuboeae Boophis schuboeae 0 0 2 0 0 0 0
Boophis septentrionalis Boophis septentrionalis 0 3 2 0 0 0 0
Boophis sibilans Boophis sibilans 0 5 2 0 0 0 0
Boophis solomaso Boophis solomaso 0 5 2 0 0 0 0
Boophis sp. 5 Boophis sp. a. albilabris red eyes 0 5 2 4 1 0 0
Boophis sp. 20 Boophis sp. a. ankaratra Andohahela fast 0 0 0 0 0 0 0
Boophis sp. 19 Boophis sp. a. Ankaratra Andohahela slow 0 0 2 0 0 0 0
Boophis sp. 16 Boophis sp. a. boehmei Ranomafana 0 5 0 0 0 0 0
Boophis sp. 11 Boophis sp. a. brachychir 2 0 5 0 0 0 1 0
Boophis sp. 22 Boophis sp. a. elenae vigoi 0 5 2 0 0 0 0
Boophis sp. 15 Boophis sp. a. lichenoides Ambatolahy 0 5 0 6 0 0 0
Boophis sp. 12 Boophis sp. a. madagascariensis north 0 0 0 5 0 0 0
Boophis sp 35 Boophis sp. a. majori Ranomafana long call 0 5 0 0 0 0 0
Boophis sp. 28 Boophis sp. a. mandraka Marojejy 0 0 2 0 0 0 0
Boophis sp. 33 Boophis sp. a. microtympanum low altitude 0 5 0 2 0 1 0
Boophis sp. 29 Boophis sp. a. miniatus Mahakajy 0 5 0 0 0 0 0
Boophis sp. 4 Boophis sp. a. occidentalis Berara 0 0 0 0 1 0 0
Boophis sp. pauliani Boophis sp. a. pauliani Tolagnaro 0 5 0 6 0 0 0
Boophis sp. 13 Boophis sp. a. periegetes Ranomafana 0 3 0 0 0 0 0
Boophis sp. 31 Boophis sp. a. rappiodes Ambre 0 5 2 0 0 0 0
Boophis sp. 32 Boophis sp. a. rappiodes lilianae 0 5 2 0 0 0 0
Boophis sp. 31 Boophis sp. a. rappiodes northeast 0 4 2 6 0 0 0
Boophis sp. 34 Boophis sp. a. rhodoscelis Ambohitantely 0 5 0 0 0 1 0
Boophis sp. rhodoscelis Boophis sp. a. rhodoscelis Ranomafana 0 5 0 0 0 1 0
Boophis sp. 8 Boophis sp. a. ruoculis Ranomafana 0 0 0 0 0 0 0
Boophis sp. 17 Boophis sp. a. sibilans trill call 0 5 0 0 0 0 0
Boophis sp. 4 Boophis sp. Comoros 0 5 2 6 0 0 0
Boophis sp. 27 Boophis sp. n. a. mandraka Andreone 1 0 5 0 0 0 0 0
Boophis sp. 2 Boophis sp. sarotra 0 5 0 0 0 0 0
Boophis tampoka Boophis tampoka 0 3 2 0 0 0 0
Boophis tasymena Boophis tasymena 0 0 2 0 0 0 0
Boophis tephraeomystax Boophis tephraeomystax 0 4 0 6 1 1 0
Boophis viridis Boophis viridis 0 0 2 0 0 0 0
Boophis vittatus Boophis vittatus 0 5 0 0 0 0 0
Boophis williamsi Boophis williamsi 0 5 0 0 0 0 0
Boophis xerophilus Boophis xerophilus 0 5 0 0 0 0 0
Gephyromantis ambohitra Gephyromantis ambohitra 3 5 0 3 0 0 0
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Species name
(W et al. 2011)
Species name
(GV 2007)
ECO DIP ICB CIP FS TPS COL
Gephyromantis asper Gephyromantis asper 3 5 0 3 0 0 0
Gephyromantis azzurae Gephyromantis azzurae Andreone Isalo Andrianamero 3 5 0 3 0 0 0
Gephyromantis blanci Gephyromantis blanci 1 5 0 3 1 1 0
Gephyromantis boulengeri Gephyromantis boulengeri 1 5 0 3 0 0 0
Gephyromantis cornutus Gephyromantis cornutus 3 3 0 3 0 0 0
Gephyromantis corvus Gephyromantis corvus 3 5 0 3 0 0 0
Gephyromantis decaryi Gephyromantis decaryi 1 5 0 3 1 1 0
Gephyromantis eiselti Gephyromantis eiselti 1 5 0 3 1 1 0
Gephyromantis enki Gephyromantis enki 1 5 0 3 1 1 0
Gephyromantis granulatus Gephyromantis granulatus 3 3 0 3 1 1 0
Gephyromantis horridus Gephyromantis horridus 3 5 0 3 0 0 0
Gephyromantis klemmeri Gephyromantis klemmeri 1 5 0 3 1 1 0
Gephyromantis leucocephalus
Ste Luce
Gephyromantis leucocephalus Ste Luce 1 5 0 3 1 1 0
Gephyromantis leucomaculatus Gephyromantis leucomaculatus 3 3 0 3 0 0 0
Gephyromantis luteus Gephyromantis luteus 3 3 1 3 0 0 0
Gephyromantis malagasius Gephyromantis malagasius 3 5 0 3 0 0 0
Gephyromantis moseri Gephyromantis moseri 3 3 0 3 0 0 0
Gephyromantis plicifer Gephyromantis plicifer 3 3 1 3 0 0 0
Gephyromantis pseudoasper Gephyromantis pseudoasper 3 5 0 3 0 0 0
Gephyromantis redimitus Gephyromantis redimitus 3 5 0 3 0 0 0
Gephyromantis rivicola Gephyromantis rivicola 2 5 0 3 1 1 0
Gephyromantis runewsweeki Gephyromantis runewsweeki 1 5 0 3 1 1 0
Gephyromantis salegy Gephyromantis salegy 3 3 0 3 0 0 0
Gephyromantis schil Gephyromantis schil 3 3 0 3 1 1 0
Gephyromantis sculpturatus Gephyromantis sculpturatus 3 3 1 3 0 0 0
Gephyromantis silvanus Gephyromantis silvanus 2 5 0 3 1 1 0
Gephyromantis sp. 1 Gephyromantis sp. a. Ambohitra Marojejy 3 5 0 3 0 0 0
Gephyromantis sp. 5 Gephyromantis sp. a. blanci Andohahela 1 5 0 3 1 1 0
Gephyromantis sp. 10 Gephyromantis sp. a. corvus Bemaraha 3 5 0 3 0 0 0
Gephyromantis sp. 11 Gephyromantis sp. a. horridus Marojejy 3 5 0 3 0 0 0
Gephyromantis sp. 17 Gephyromantis sp. a. leucomaculatus Marojejy 3 3 0 3 0 1 0
Gephyromantis sp. 13 Gephyromantis sp. a. malagasius highlands 3 5 0 3 0 0 0
Gephyromantis spinifer Gephyromantis spinifer 3 5 0 3 0 0 0
Gephyromantis striatus Gephyromantis striatus 3 5 0 3 0 0 0
Gephyromantis tandroka Gephyromantis tandroka 3 3 0 3 0 0 0
Gephyromantis thelenae Gephyromantis thelenae 1 5 0 3 1 1 0
Gephyromantis tschenki Gephyromantis tschenki 3 3 0 3 0 0 0
Gephyromantis ventrimaculatus Gephyromantis ventrimaculatus 3 5 0 3 0 0 0
Gephyromantis webbi Gephyromantis webbi 2 5 0 3 1 1 0
Gephyromantis zavona Gephyromantis zavona 3 3 0 3 0 1 0
Guibemantis albolineatus Guibemantis albolineatus 0 4 0 3 0 1 0
Guibemantis bicalcaratus Guibemantis bicalcaratus 0 4 0 3 0 1 0
Guibemantis depressiceps Guibemantis depressiceps 0 3 0 0 0 1 0
Guibemantis avobrunneus Guibemantis avobrunneus 0 4 0 3 0 1 0
Guibemantis kathrinae Guibemantis kathrinae 0 3 0 0 0 1 0
Guibemantis liber Guibemantis liber 0 4 0 3 0 1 0
Guibemantis pulcher Guibemantis pulcher 0 5 2 3 0 1 0
Guibemantis punctatus Guibemantis punctatus 0 4 2 3 0 1 0
Guibemantis sp. 3 Guibemantis sp. a. albolineatus Andasibe 0 4 0 3 0 1 0
Guibemantis sp. 8 Guibemantis sp. a. bicalcaratus Besariaka 0 4 0 3 0 1 0
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Species name
(W et al. 2011)
Species name
(GV 2007)
ECO DIP ICB CIP FS TPS COL
Guibemantis sp. 14 Guibemantis sp. a .bicalcaratus Fierenana 0 4 0 3 0 1 0
Guibemantis sp. 12 Guibemantis sp. a. bicalcaratus Manongarivo 0 4 0 3 0 1 0
Guibemantis sp. 9 Guibemantis sp. a. bicalcaratus Nosy boraha 2 0 4 0 3 0 1 0
Guibemantis sp. 7 Guibemantis sp. a. bicalcaratus Tolagnaro 0 4 0 3 0 1 0
Guibemantis sp. 19 Guibemantis sp. a. depressiceps Andohahela 0 4 0 3 0 1 0
Guibemantis sp. 10 Guibemantis sp. a. avobrunneus Manombo 0 4 0 3 0 1 0
Guibemantis sp. 20 Guibemantis sp. a. liber giant 0 4 0 3 0 0 0
Guibemantis sp. 5 Guibemantis sp. a. liber Vevembe 0 4 0 3 0 1 0
Guibemantis sp. 6 Guibemantis sp. a. punctatus south 0 4 0 3 0 1 0
Guibemantis timidus Guibemantis timidus 0 5 0 6 1 1 0
Guibemantis tornieri Guibemantis tornieri 0 3 0 0 0 1 0
Laliostoma labrosum Laliostoma labrosum 1 5 0 5 0 0 0
Mantella aurantiaca Mantella aurantiaca 1 1 1 3 0 0 1
Mantella baroni Mantella baroni 1 1 0 3 0 1 1
Mantella bernhardi Mantella bernhardi 1 3 0 3 0 1 1
Mantella betsileo Mantella betsileo 1 3 0 3 1 1 0
Mantella cowani Mantella cowani 1 1 0 3 0 1 1
Mantella crocea Mantella crocea 1 1 0 3 0 1 0
Mantella ebenaui Mantella ebenaui 1 3 0 3 1 1 0
Mantella expectata Mantella expectata 1 3 0 3 1 1 0
Mantella haraldmeieri Mantella haraldmeieri 1 3 0 3 0 1 0
Mantella laevigata Mantella laevigata 1 1 0 3 0 1 0
Mantella madagascariensis Mantella madagascariensis 1 1 0 3 0 1 1
Mantella manery Mantella manery 1 3 0 3 1 1 0
Mantella milotympanum Mantella milotympanum 1 1 0 3 0 0 1
Mantella nigricans Mantella nigricans 1 1 0 3 0 1 1
Mantella pulchra Mantella pulchra 1 1 0 3 0 1 0
Mantella sp. 1 Mantella sp. a. expectata Tranomaro 1 3 0 3 1 1 0
Mantella viridis Mantella viridis 1 3 0 3 1 1 0
Mantidactylus aerumnalis Mantidactylus aerumnalis 4 3 0 3 1 0 0
Mantidactylus albofrenatus Mantidactylus albofrenatus 4 3 0 3 1 1 0
Mantidactylus alutus Mantidactylus alutus 5 5 0 3 0 0 0
Mantidactylus ambreensis Mantidactylus ambreensis 5 5 0 3 1 1 0
Mantidactylus argenteus Mantidactylus argenteus 3 5 2 3 0 0 0
Mantidactylus bellyi Mantidactylus bellyi 5 5 0 3 0 0 0
Mantidactylus betsileanus Mantidactylus betsileanus 5 5 0 3 0 0 0
Mantidactylus biporus Mantidactylus biporus 5 5 0 3 0 0 0
Mantidactylus bourgati Mantidactylus bourgati 5 5 0 3 0 0 0
Mantidactylus brevipalmatus Mantidactylus brevipalmatus 4 3 0 3 1 1 0
Mantidactylus charlotteae Mantidactylus charlotteae 4 3 0 3 1 1 0
Mantidactylus cowanii Mantidactylus cowanii 2 2 0 3 0 0 0
Mantidactylus curtus Antoetra Mantidactylus curtus Antoetra 4 5 0 3 1 0 0
Mantidactylus delormei Mantidactylus delormei 4 3 0 3 1 1 0
Mantidactylus femoralis Mantidactylus femoralis 5 4 0 3 1 1 0
Mantidactylus guttulatus east Mantidactylus guttulatus east 4 5 0 3 0 0 0
Mantidactylus guttulatus north Mantidactylus guttulatus north 4 5 0 3 0 0 0
Mantidactylus lugubris Andasibe Mantidactylus lugubris Andasibe 2 5 0 3 0 0 0
Mantidactylus madecassus Mantidactylus madecassus 4 2 0 3 0 0 0
Mantidactylus majori Mantidactylus majori 5 5 0 3 1 1 0
Mantidactylus melanopleura Mantidactylus melanopleura 4 3 0 3 1 1 0
Mantidactylus mocquardi Andasibe Mantidactylus mocquardi Andasibe 5 5 0 3 0 0 0
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Species name
(W et al. 2011)
Species name
(GV 2007)
ECO DIP ICB CIP FS TPS COL
Mantidactylus noralottae
Andreone Isalo Andrianamero
Mantidactylus noralottae
Andreone Isalo Andrianamero
5 5 0 3 0 0 0
Mantidactylus opiparis Mantidactylus opiparis 4 3 0 3 1 1 0
Mantidactylus pauliani Mantidactylus pauliani 4 2 0 3 0 0 0
Mantidactylus sp. 26 Mantidactylus sp. a. betsileanus Andranofotsy 5 5 0 3 0 0 0
Mantidactylus sp. 27 Mantidactylus sp. a. betsileanus Nosy Boraha 5 5 0 3 0 0 0
Mantidactylus sp. 28 Mantidactylus sp. a. betsileanus slow calls 5 5 0 3 0 0 0
Mantidactylus sp. 36 Mantidactylus sp. a. betsileanus Toamasina 5 5 0 3 0 0 0
Mantidactylus sp. 29 Mantidactylus sp. a. betsileanus Tolagnaro 5 5 0 3 0 0 0
Mantidactylus sp. 17 Mantidactylus sp. a. biporus Ambohitantely 5 5 0 3 0 0 0
Mantidactylus sp. 23 Mantidactylus sp. a. biporus Andasibe 5 5 0 3 0 0 0
Mantidactylus sp. 26 Mantidactylus sp. a. biporus Andranofotsy 5 5 0 3 0 0 0
Mantidactylus sp. 24 Mantidactylus sp. a. biporus Ranomafana 5 5 0 3 0 0 0
Mantidactylus sp. 32 Mantidactylus sp. a. biporus Tsaratanana camp 0 5 5 0 3 0 0 0
Mantidactylus sp. 33 Mantidactylus sp. a. biporus Tsaratanana camp 1 5 5 0 3 0 0 0
Mantidactylus sp. 13 Mantidactylus sp. a. charlotteae Ranomafana 4 3 0 3 1 1 0
Mantidactylus sp. 48 Mantidactylus sp. a. cowanii small 2 2 0 3 0 0 0
Mantidactylus sp. 18 Mantidactylus sp. a. curtus Ambohitantely 5 5 0 3 1 0 0
Mantidactylus sp. 30 Mantidactylus sp. a. curtus Andohahela short snout 5 5 0 3 0 0 0
Mantidactylus sp. 19 Mantidactylus sp. a. curtus Ankaratra 5 5 0 3 1 0 0
Mantidactylus sp. 44 Mantidactylus sp. a. femoralis Ambohitsara 5 5 0 3 0 0 0
Mantidactylus sp. 42 Mantidactylus sp. a. femoralis Ambre 5 3 0 3 1 1 0
Mantidactylus sp. 40 Mantidactylus sp. a. femoralis Tsaratanana 5 5 0 3 0 1 0
Mantidactylus sp. 57 Mantidactylus sp. a. grandidieri north 4 5 0 3 0 0 0
Mantidactylus sp. 52 Mantidactylus sp. a. lugubris Marojejy 2 5 0 3 0 0 0
Mantidactylus sp. 49 Mantidactylus sp. a. lugubris south 2 5 0 3 0 0 0
Mantidactylus sp. 41 Mantidactylus sp. a. majori andapa 5 5 0 3 1 1 0
Mantidactylus sp. 47 Mantidactylus sp. a. mocquardi Ambatolahy 5 5 0 3 0 0 0
Mantidactylus sp. 46 Mantidactylus sp. a. mocquardi Marojejy 5 5 0 3 1 1 0
Mantidactylus sp. 45 Mantidactylus sp. a. mocquardi Tsaratanana 5 4 0 3 1 1 0
Mantidactylus sp. 20 Mantidactylus sp. a. pauliani Itremo 4 5 0 3 0 0 0
Mantidactylus tricinctus Mantidactylus sp. a. tricinctus parvus 5 5 0 3 0 0 0
Mantidactylus sp. 14 Mantidactylus sp. a. ulcerosus Isalo 5 5 0 3 0 0 0
Mantidactylus sp. 7 Mantidactylus tricinctus Manantantely 5 5 0 3 0 0 0
Mantidactylus tricinctus Manombo Mantidactylus tricinctus Manombo 5 5 0 3 0 0 0
Mantidactylus ulcerosus Mantidactylus ulcerosus 5 5 0 3 0 0 0
Mantidactylus zipperi Mantidactylus zipperi 4 3 0 3 1 1 0
Mantidactylus zolitschka Mantidactylus zolitschka 5 5 0 3 1 1 0
Spinomantis aglavei Spinomantis aglavei 0 5 0 0 0 0 0
Spinomantis bertini Spinomantis bertini 1 5 0 3 1 1 0
Spinomantis elegans Spinomantis elegans 2 3 0 3 1 1 0
Spinomantis mbriatus Spinomantis mbriatus 0 5 0 0 0 0 0
Spinomantis guibei Spinomantis guibei 1 3 0 3 1 1 0
Spinomantis massorum Spinomantis massorum 0 5 0 5 0 0 0
Spinomantis microtis Spinomantis microtis 2 5 0 3 0 0 0
Spinomantis peraccae Spinomantis peraccae 0 5 0 6 0 0 0
Spinomantis phantasticus Spinomantis phantasticus 0 5 0 5 0 0 0
Spinomantis sp. 6 Spinomantis sp. a. bertini Andohahela low altitude 1 5 0 3 1 1 0
Spinomantis sp. 8 Spinomantis sp. a .bicalcaratus Maharira 0 5 0 3 0 0 0
Tsingymantis antitra Tsingymantis antitra 2 4 0 0 0 0 0
Wakea madinika Wakea madinika 1 5 0 3 1 1 0
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Table S2. Posterior probabilities of the character states (in bold) for the eight characters analysed as reconstructed
for the following relevant nodes: Aglyptodactylus, Blommersia, Gephyromantis, Guibemantis, Mantella, Mantidactylus and Spinomantis,
nodes representing the most recent common ancestor (MRCA) of all species of each genus; Boophinae, Mantellinae, Laliostominae,
nodes of the MRCAs of the mantellid subfamilies according to GV (2006). ECO, general ecology and habits; DIP, de-
tailed iris pattern; ICB, iris colour brightness; CIP, iris periphery (sclera) colour; FS, frenal streak; TPS, dark tympanic patch or streak;
COL,general dorsal colouration. Character state values are detailed in Materials and Methods.
Node ECO DIP ICB CIP FS TPS COL
Aglyptodactylus (1)1.000 (4)1.000 (0)0.999 (6)1.000 (1)0.999 (1)0.999 (0)0.999
Blommersia (3)1.000 (4)0.984 (1)1.000 (4)1.000 (1)1.000 (1)1.000 (0)1.000
Boophis (0)0.998 (6)1.000 (1)1.000 (1)0.892 (1)0.885 (1)0.890 (0)0.999
Gephyromantis (3)0.729 (6)1.000 (1)1.000 (4)1.000 (1)0.933 (1)0.938 (0)1.000
Guibemantis (0)1.000 (5)1.000 (1)1.000 (4)1.000 (1)1.000 (1)1.000 (0)1.000
Mantella (1)1.000 (4)0.522 (1)1.000 (4)1.000 (1)1.000 (1)1.000 (1)0.879
Mantidactylus (2)0.400 (6)1.000 (1)1.000 (4)1.000 (1)0.932 (1)0.937 (0)1.000
Spinomantis (1)0.884 (6)1.000 (1)1.000 (4)1.000 (1)0.932 (1)0.937 (0)1.000
Mantellinae (2)0.999 (6)0.999 (1)1.000 (4)0.985 (1)0.932 (1)0.937 (0)1.000
Laliostominae (1)1.000 (6)0.978 (1)1.000 (6)0.869 (1)0.928 (1)0.933 (0)0.999
Boophinae (0)0.998 (6)1.000 (1)1.000 (1)0.892 (1)0.885 (1)0.890 (0)0.999
Mantellidae (3)0.431 (5)1.000 (0)1.000 (0)0.847 (1)0.999 (1)0.897 (0)0.999
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Table S3. Matrix of correlations among eye and body morphology and general ecology and habits in Malagasy
mantellid frogs. e upper number gives the statistics of character association d
ij
, the lower number is predictive of the p-values
below. Under each sub-table, D
ij
values (overall character correlation) and the associated signicances are given. Pairwise association
values (d
ij
) that remained signicant (P≤0.05) aer sequential Bonferroni correction (over the total number of independent tests) are
highlighted in bold. Positive values indicate that the two states occur together at a higher frequency than expected by chance, while
negative values indicate they occur together at a lower frequency than expected by chance. Negative values indicate a negative cor-
relation among the character states.
Detailed iris pattern / General ecology and habits
Arboreal Terrestrial Semiarboreal Saxicolous Rheophilous Terrestrial rheophilous
Black eyes -0.004 0.010 -0.001 -0.001 -0.001 -0.003
0.124 0.019 0.296 0.550 0.439 0.273
Uniform -0.015 -0.010 -0.008 0.003 -0.011 0.041
0.002 0.006 0.013 0.171 0.001 < 0.001
Densely reticulated -0.003 -0.001 0.001 0.003 0.001 -0.002
0.200 0.495 0.307 0.173 0.402 0.259
Annular 0.040 -0.008 -0.013 -0.002 -0.003 -0.011
< 0.0001 0.025 < 0.001 0.201 0.213 0.001
Horizontally divided -0.040 0.014 0.030 -0.004 0.018 -0.017
< 0.001 0.004 < 0.001 0.151 0.001 < 0.001
Horizontally striped 0.023 -0.005 -0.007 -0.001 -0.002 -0.007
< 0.001 0.058 0.036 0.509 0.178 0.034
D=0.307 P<0.001
Iris contrast / General ecology and habits
Arboreal Terrestrial Semiarboreal Saxicolous Rheophilous Terrestrial rheophilous
Similar to body -0.049 0.009 0.012 0.001 0.004 0.021
< 0.001 0.010 0.004 0.324 0.049 < 0.001
Darker than body 0.001 0.001 0.003 0.001 -0.001 -0.005
0.311 0.316 0.168 0.290 0.348 0.040
Brighter than body 0.048 -0.010 -0.015 -0.001 -0.004 -0.015
< 0.001 0.001 0.001 0.207 0.115 < 0.001
D=0.307 P<0.001
General ecology and habits / Colour of iris periphery (sclera)
Indistinct Blue Bluish Green Yellow Red White
Arboreal -0.143 0.120 0.002 0.003 0.009 0.021 0.005
< 0.001 < 0.001 0.207 0.121 0.025 0.293 0.085
Terrestrial 0.014 -0.024 0.001 -0.001 0.010 -0.001 0.001
0.002 < 0.001 0.511 0.450 0.013 0.520 0.513
Semiarboreal 0.059 -0.042 -0.001 -0.001 -0.011 -0.001 -0.003
< 0.001 < 0.001 0.280 0.421 0.003 0.361 0.243
Saxicolous 0.007 -0.001 0.001 0.001 0.001 0.001 0.001
0.056 0.005 0.384 0.491 0.504 0.327 0.417
Rheophilous 0.015 -0.011 -0.001 -0.001 -0.002 -0.001 -0.001
0.005 0.001 0.524 0.448 0.254 0.541 0.422
Terrestrial rheophilous 0.046 -0.032 -0.001 -0.001 -0.007 -0.001 -0.002
< 0.001 < 0.001 0.465 0.355 0.018 0.352 0.284
D=0.627 P<0.001
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
General ecology and habits / Head with dark tympanic patch or streak
Presence Absence
Arboreal -0.013 0.013
< 0.001 < 0.001
Terrestrial 0.023 -0.023
< 0.001 < 0.001
Semiarboreal -0.001 0.001
0.263 0.240
Saxicolous 0.001 -0.001
0.152 0.162
Rheophilous 0.003 -0.003
0.025 0.025
Terrestrial rheophilous -0.013 0.013
< 0.001 < 0.001
D=0.141 P<0.001
General ecology and habits / Head with frenal streak
Presence Absence
Arboreal -0.014 0.014
< 0.001 < 0.001
Terrestrial 0.022 -0.022
< 0.001 < 0.001
Semiarboreal 0.002 -0.002
0.041 0.042
Saxicolous -0.002 0.002
0.111 0.109
Rheophilous 0.003 -0.003
0.021 0.021
Terrestrial rheophilous -0.013 0.013
< 0.001 < 0.001
D=0.139 P<0.001
General dorsal colouration / Detailed iris pattern
Black eyes Uniform Densely
reticulated
Annular Horizontally
divided
Horizontally
striped
Cryptic -0.006 0.002 -0.001 -0.001 0.002 0.004
0.006 0.062 0.190 0.141 0.102 0.013
Aposematic 0.006 -0.002 0.001 0.001 -0.002 -0.004
0.004 0.062 0.180 0.138 0.103 0.014
D=0.074 P<0.104
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Detailed iris pattern / Colour of iris periphery (sclera)
Indistinct Blue Bluish Green Yellow Red White
Black eyes 0.001 -0.002 0.001 0.001 0.001 0.001 -0.001
0.325 0.291 0.481 0.461 0.423 0.402 0.464
Uniform 0.009 -0.001 0.001 0.001 -0.012 -0.001 0.001
0.027 0.432 0.352 0.233 < 0.001 0.467 0.304
Densely reticulated 0.001 -0.002 0.001 -0.001 0.001 0.001 0.001
0.402 0.233 0381 0.501 0.444 0.362 0.397
Annular -0.045 0.042 0.001 0.001 0.001 0.001 0.001
< 0.001 < 0.001 0.486 0.418 0.258 0.473 0.385
Horizontally divided 0.023 -0.025 -0.001 -0.001 0.007 -0.001 -0.002
< 0.001 < 0.001 0.368 0.373 0.039 0.390 0.316
Horizontally striped 0.010 -0.012 -0.001 -0.001 0.002 -0.005 -0.005
0.022 0.002 0.538 0.404 0.234 0.552 0.449
D=0.247 P<0.060
Iris contrast / Colour of iris periphery (sclera)
Indistinct Blue Bluish Green Yellow Red White
General tones 0.071 -0.055 -0.002 -0.001 -0.008 -0.002 -0.001
< 0.001 < 0.001 0.201 0.411 0.009 0.204 0.274
Darker than body -0.012 0.003 0.001 0.001 0.005 0.001 0.001
0.003 0.189 0.367 0.361 0.049 0.346 0.247
More colourful -0.059 0.052 0.001 -0.001 0.003 0.001 0.001
< 0.001 < 0.001 0.258 0.459 0.176 0.283 0.434
D=0.311 P<0.001
Detailed iris pattern / Head with dark tympanic patch or streak
Presence Absence
Black eyes 0.003 -0.003
0.028 0.028
Uniform -0.026 0.026
< 0.001 < 0.001
Densely reticulated -0.001 0.001
0.343 0.318
Annular -0.012 0.012
< 0.001 < 0.001
Horizontally divided 0.019 -0.019
< 0.001 < 0.001
Horizontally striped 0.016 -0.016
< 0.001 < 0.001
D=0.166 P<0.001
Frenal streak / Head with dark tympanic patch or streak
Presence Absence
Presence 0.028 -0.028
< 0.001 < 0.001
Absence -0.028 0.028
< 0.001 < 0.001
D=0.122 P<0.001
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Figure S2 (next 3 pages). One-character history reconstructed through stochastic character mapping of general ecology and habits in mantellid frogs (as in Figure 1 of
main paper but including taxon names as in second column of Table S1).
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Figure S3 (next 3 pages). One character history reconstructed through stochastic character mapping of dierent character states of iris pattern in mantellid frogs (as in
Figure 1 of main paper but including taxon names as in second column of Table S1).
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Figure S4 (next 3 pages). One character history reconstructed through stochastic character mapping of dierent character states of iris contrast in mantellid frogs (as in
Figure 1 of main paper but including taxon names as in second column of Table S1).
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Figure S5 (next 3 pages). One character history reconstructed through stochastic character mapping of dierent character states of iris periphery colour in mantellid frogs
(as in Figure 1 of main paper but including taxon names as in second column of Table S1).
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Figure S6 (next 3 pages). One character history reconstructed through stochastic character mapping of presence vs. absence of the frenal streak in mantellid frogs (as in
Figure 1 of main paper but including taxon names as in second column of Table S1).
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Figure S7 (next 3 pages). One character history reconstructed through stochastic character mapping of presence vs. absence of the dark tympanic patch in mantellid frogs
(as in Figure 1 of main paper but including taxon names as in second column of Table S1).
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Figure S8 (next 3 pages). One character history reconstructed through stochastic character mapping of function of body colouration in mantellid frogs (as in Figure 1 of
main paper but including taxon names as in second column of Table S1).
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
Supplementary Material to Aet al. (2013) Salamandra 49(1): 7–17
... Boophis are notorious for lacking distinct morphological differences between closely related species and their intraspecific variability in body colour and pattern can be substantial (e.g. in B. picturatus, see Glaw et al. 2001). In contrast, the colouration of the eyes turned out to be species-specific for numerous Boophis species and therefore is a crucial character for their taxonomy (Glaw & Vences 1997, Amat et al. 2013. Similarly, the colour of the webbing between toes and fingers can be a reliable character to distinguish closely related species (e.g. ...
... In contrast, the body colouration of B. albilabris, B. praedic tus and B. tsilomaro is known to vary substantially within and among populations. These observations confirm previous studies, which found species-specific eye colouration in frogs and especially in treefrogs (Glaw & Vences 1997, Amat et al. 2013. Boophis and other treefrog species appear to be largely nocturnal, especially when breeding, making it unlikely that the colourful eyes can be used as a prezygotic isolation mechanism. ...
Full-text available
Article
Eye and webbing colouration as predictors of specific distinctness: a genetically isolated new treefrog species of the Boophis albilabris group from the Masoala peninsula, northeastern Madagascar Abstract. We describe a large and distinctive new treefrog species with blue webbing from the west coast of the Masoala peninsula in northeastern Madagascar. Boophis masoala sp. n. is morphologically similar to the other species of the Boophis albilabris group but can be distinguished from them easily by several chromatic characters of the eyes. Despite its similar morphology, it is genetically highly differentiated (10.5-13.3% pairwise p-distance in a segment of the 16S rRNA gene) from all other species in the B. albilabris group including the morphologically most similar Boophis praedictus. Both species share the blue webbing between toes and are distributed on the Masoala peninsula, but so far were not found in close sympatry. Although we recorded the new species only from the unprotected areas near the coast, we are confident that it also occurs within the adjacent Masoala National Park. We discuss the importance of eye colouration as a predictor of specific distinctness in the genus Boophis and that of webbing colouration as taxonomic characters of large treefrogs. Based on a micro-CT scan we provide a comprehensive description of the osteology of the new species, which is the first for any Boophis species, and furthermore describe its distress call which consists of three distinct sections corresponding to (1) the starting phase with closed mouth, (2) the opening of the mouth and (3) the final section with an open mouth.
... This latter example confirms that colour anomalies can affect the eyes and body separately. Given that both the pigment cells of iris and body skin are derived from the neural crest, this indicates that either the neural crest cells of the two tissues undergo separate cell regulation processes, or the mutations underlying colour anomalies may affect their migration in different ways (Amat et al., 2013). ...
... Phenotypic eye color has been suggested as an indicator of genetic predisposition toward certain behaviors, where dark-eyed subjects would tend to display behaviors requiring sensitivity, speed, and reactive responses, while with ones with light-colored eyes, behaviors requiring hesitation, inhibition, and self-paced responses, both in humans and in animals (Elias et al., 2008). Furthermore, it has been proposed that eye coloration in various species may be related to social ranks, aggression, mate recognition, and sexual selection (e.g., Volpato et al., 2003, Amat et al., 2013. Chicken eye color is largely determined by genetics, but age, diet, and disease can affect it as well (Nelson, 1947). ...
Full-text available
Article
Despite the intensive genetic selection in modern poultry, variability of domestic fowl phenotypes has remained, especially in breeds adapted to local conditions. The relevance of this variability to the chicken outdoor ranging activities remains unknown. The aim of this study was to investigate if external features were associated with the ranging frequency of the 48 female chickens from each of the two breeds: Sasso and Green-legged Partridge. In each of six single-breed pens, eight hens and two roosters were housed under conditions of EU requirements for organic meat chicken production, including free access to an outdoor range, from weeks 5 to 10 of age. The birds were video-recorded during the experiment to obtain frequencies of individual birds' use of the ranges. Comb size (length and height) was measured using a digital ruler, while the sizes of the dark area of neck plumage and beak were processed and analyzed using ImageJ software. The same traits were scored using direct visual assessment by a trained observer on a scale of 1-3. In addition, the eye color of the bird was recorded. Statistical analysis was conducted independently for each breed using regression models, ANOVAs and Spearman correlations. Significant positive associations between neck plumage (P<0.01), beak darkness (P=0.03) measurements, comb length (P<0.01) and comb height (P<0.01) and frequency of range use were identified for Sasso. Sasso hens scored with darkest neck plumage (P=0.03) and biggest comb size (P=0.04) ranged the most, while their external features were significantly and positively correlated between each other, except beak darkness and comb length. No significant associations between ranging and external features were found in Green-legged Partridge birds, except that their comb height was significantly and positively correlated with neck plumage and beak darkness (r=0.39 and 0.33, respectively). In some genetic strains, better understanding of the associations between chickens’ external features with ranging behavior could contribute to improve selection programs and bird welfare, assuring production of breeding stock suitable for outdoor conditions.
... Results of our quantitative morphometric analysis indicated that the three known species in the genus Alexteroon exhibit highly conserved general body plans and external morphology, rendering a quantitative distinction between species based on preserved material alone difficult, particularly when comparing the two superficially similar patterned yet genetically well separated A. obstetricans and A. jynx. Meanwhile, phenotypic features that can only be observed in live specimens, such as the gular coloration observed in reproductively active males of A. hypsiphonus and A. obstetricans or marked differences in coloration between A. hypsiphonus on one hand and A. obstetricans and A. jynx on the other hand appear to be more useful for ad hoc species delimitation (compare Amat et al. 2013;Glaw et al. 2018 for similar findings in mantellid frogs). These observations call for detailed morphological studies, as well as for photo-documentations of live specimens that should be a standard feature of anuran species descriptions. ...
Article
The African reedfrog taxon Alexteroon consists of only three described species with rather restricted geographical ranges. Although the assignment to a distinct genus is supported by multiple evidence, its position within the larger African hyperoliid radiation has been disputed. Available molecular data are scarce and the geographic records are few and scattered. The partially formalin fixed type series were previously not accessible to molecular analyses. This changed only very recently with the advancement of Next Generation Sequencing and ancient DNA techniques. Here we provide a reassessment of the current distribution and identity of all known species in this taxon based on (a) historical and new records, (b) morphological reanalyses of the type material and newly collected specimens from Angola and Gabon, and (c) newly established, nearly complete mitochondrial genome data from historical type and modern non-type material. We also present a molecular phylogeny (five mitochondrial loci 12S, 16S, ND1, ND2, COI, Cytb) for 78 sequences from 75 different species of Hyperolius retrieved from GenBank and 14 newly established Alexteroon sequences. We demonstrate that Alexteroon is more widely distributed than previously thought with records from northern Angola representing major southern range extensions. Results of the quantitative morphometric analyses show that the group has a rather conserved general body plan. Therefore qualitative phenotypic features observable in live specimens appear to be more useful for ad hoc species delimitation. We found Alexteroon to be nested within Hyperolius, corroborating previous findings. However, the combination of molecular data and consistent differences observed in morphology and ecology provide strong support for the distinctiveness of this evolutionary lineage within Hyperolius sensu lato. We therefore treat Alexteroon as a subgenus of Hyperolius and argue that the large and diverse genus Hyperolius is in need of revision that may result in new generic arrangements.
... Several anuran studies have emphasized on the usefulness of eye colour and pattern as a character for species level identification (e.g., Duellman, 1970;Glaw & Vences, 1997;Amat, Wollenberg & Vences, 2013;Glaw et al., 2018) or study of ontogenetic colour changes (e.g., Hoffman & Blouin, 2000;Biju et al., 2013); however, the application of this trait for field identification of frogs is seldom attempted (Glaw & Vences, 1997;Stuebing & Wong, 2000). Among the~230 known frog species of the Western Ghats, genus Raorchestes is the most remarkably diverse in terms of skin colouration as well as eye colours and patterns. ...
Full-text available
Article
The genus Raorchestes is a large radiation of Old World tree frogs for which the Western Ghats in Peninsular India is the major center for origin and diversification. Extensive studies on this group during the past two decades have resolved long-standing taxonomic confusions and uncovered several new species, resulting in a four-fold increase in the number of known Raorchestes frogs from this region. Our ongoing research has revealed another five new species in the genus, formally described as Raorchestes drutaahu sp. nov., Raorchestes kakkayamensis sp. nov., Raorchestes keirasabinae sp. nov., Raorchestes sanjappai sp. nov., and Raorchestes vellikkannan sp. nov., all from the State of Kerala in southern Western Ghats. Based on new collections, we also provide insights on the taxonomic identity of three previously known taxa. Furthermore, since attempts for an up-to-date comprehensive study of this taxonomically challenging genus using multiple integrative taxonomic approaches have been lacking, here we review the systematic affinities of all known Raorchestes species and define 16 species groups based on evidence from multi-gene (2,327 bp) phylogenetic analyses, several morphological characters (including eye colouration and pattern), and acoustic parameters (temporal and spectral properties, as well as calling height). The results of our study present novel insights to facilitate a better working taxonomy for this rather speciose and morphologically conserved radiation of shrub frogs. This will further enable proper field identification, provide momentum for multi-disciplinary studies, as well as assist conservation of one of the most colourful and acoustically diverse frog groups of the Western Ghats biodiversity hotspot.
... A less well-studied source of ornamental diversity is the eye. Eye color arises from the deposition of pigments in the pigmented epithelium of the iris [4], and is thought to play many of the same signaling functions of integumentary ornaments, both in intra-specific and inter-specific communication [5][6][7][8][9]. Dermal chromatophores and the iris pigmented epithelium share a common developmental origin from neural crest cells [10][11], and similarities in ultrastructure and pigment type composition have been described [3,12]. ...
Full-text available
Article
Birds exhibit striking variation in eye color that arises from interactions between specialized pigment cells named chromatophores. The types of chromatophores present in the avian iris are lacking from the integument of birds or mammals, but are remarkably similar to those found in the skin of ectothermic vertebrates. To investigate molecular mechanisms associated with eye coloration in birds, we took advantage of a Mendelian mutation found in domestic pigeons that alters the deposition of yellow pterin pigments in the iris. Using a combination of genome-wide association analysis and linkage information in pedigrees, we mapped variation in eye coloration in pigeons to a small genomic region of ~8.5kb. This interval contained a single gene, SLC2A11B , which has been previously implicated in skin pigmentation and chromatophore differentiation in fish. Loss of yellow pigmentation is likely caused by a point mutation that introduces a premature STOP codon and leads to lower expression of SLC2A11B through nonsense-mediated mRNA decay. There were no substantial changes in overall gene expression profiles between both iris types as well as in genes directly associated with pterin metabolism and/or chromatophore differentiation. Our findings demonstrate that SLC2A11B is required for the expression of pterin-based pigmentation in the avian iris. They further highlight common molecular mechanisms underlying the production of coloration in the iris of birds and skin of ectothermic vertebrates.
... Due to the inherent conspicuousness of the vertebrate eye [2], color patterns that seem to mimic eyes are often incorporated into signals and traits designed to be maximally salient [7]. However, the function of conspicuous real eyes, found in multiple species across all vertebrate classes (e.g., [16][17][18]), is largely unknown. Indeed, many fish are capable of greatly increasing their eye salience via rapid color change of chromatophores distributed across their irises [14,15,19], making them an ideal system for studying the adaptive function of eye coloration. ...
Article
Understanding the adaptive function of conspicuous coloration has been a major focus of evolutionary biology for much of the last century. Although considerable progress has been made in explaining how conspicuous coloration can be used in functions as diverse as sexual and social signaling, startling predators, and advertising toxicity [1], there remain a multitude of species that display conspicuous coloration that cannot be explained by existing theory. Here we detail a new “matador-like” divertive antipredator strategy based on conspicuous coloration in Trinidadian guppies (Poecilia reticulata). Guppies encountering predatory fish rapidly enhance the conspicuousness of their eyes by blackening their irises. By pitting biomimetic robotic guppies against real predatory fish, we show this conspicuous eye coloration diverts attacks away from the guppies’ center of mass to their head. To determine the function of this seemingly counterintuitive behavior, we developed a method for simulating escape probabilities when live prey interact with ballistic attacking predators, and find this diversion effect significantly benefits black-eyed guppies because they evade capture by rapidly pivoting away from the predator once it has committed to its attack. Remarkably, this antipredator strategy reverses the commonly observed negative scaling relationship between prey size and evasive ability, with larger fish benefiting most from diverting predators. Taken together, our results introduce a new antipredator divertive strategy that may be widely used by conspicuously colored prey that rely on agility to escape their predators.
... It is often assumed, for example, that vertically elongated pupils are associated with an arboreal lifestyle. However, while vertical pupils are common to all phyllomedusine frogs (Tyler and Davies, 1978), a genus of Central and South American tree frogs, among mantellid frogs both arboreal and terrestrial species have horizontally elongated pupils (Amat et al., 2013). Similarly, a genus of African hyperoliid tree frog has a horizontally elongate pupil, while those of close relatives are vertical (Roedel et al., 2009). ...
Article
The timecourse and extent of changes in pupil area in response to light are reviewed in all classes of vertebrate and cephalopods. Although the speed and extent of these responses vary, most species, except the majority of teleost fish, show extensive changes in pupil area related to light exposure. The neuromuscular pathways underlying light-evoked pupil constriction are described and found to be relatively conserved, although the precise autonomic mechanisms differ somewhat between species. In mammals, illumination of only one eye is known to cause constriction in the unilluminated pupil. Such consensual responses occur widely in other animals too, and their function and relation to decussation of the visual pathway is considered. Intrinsic photosensitivity of the iris muscles has long been known in amphibia, but is in fact widespread in other animals. The functions of changes in pupil area are considered. In the majority of species, changes in pupil area serve to balance the conflicting demands of high spatial acuity and increased sensitivity in different light levels. In the few teleosts in which pupil movements occur they do not serve a visual function but play a role in camouflaging the eye of bottom-dwelling species. The occurrence and functions of the light-independent changes in pupil size displayed by many animals are also considered. Finally, the significance of the variations in pupil shape, ranging from circular to various orientations of slits, ovals, and other shapes, is discussed.
... All four monophyletic clades were reliably distinguished based on a combination of upper and lower iris colour while alive. Iris colouration in frogs has been demonstrated to have a clear association with ecology [34]. A horizontally contrasted iris pattern, differing between the upper and lower iris, and a large dark tympanic patch are a typical eye structure for frogs inhabiting leaf litter [35]. ...
Full-text available
Article
An important aspect of evaluating biodiversity in a region, starting a monitoring program or informing conservation management decisions is having a good understanding of the taxonomy of local species. However, identification to the species-level can be challenging. A combination of DNA-based and phenotypic character analysis can provide a preliminary species list and help identify diagnostic features for taxonomically difficult groups such as the Neotropical leaf litter frogs (Craugastor spp.). We used 16S and COI marker sequences to assess the number of phylogenetic Craugastor species present in Cusuco National Park, Honduras. We conducted a linear discriminant analysis to determine if phenotypic characteristics could validate identified monophyletic species. Subsequently, we evaluated the efficacy of the Automatic Barcode Gap Discovery (ABGD) algorithm, a DNA sequence similarity-based tool, for species delineation within Neotropical amphibians. Phylogenetic analyses conducted on sequences derived from 194 specimens produced concordant results between both loci, with reciprocal monophyly of mitochondrial DNA haplotypes for all clades, revealing the presence of four Craugastor species: C. rostralis, C. chac, C. laticeps and C. c.f. charadra. Iris colouration was discovered to be a diagnostic character and the ABGD algorithm accurately identified all four monophyletic species within the phylogenetic and phenotypic analyses. A further three species have been reported from Cusuco National Park including C. milesi, C. laevissimus and C. coffeus. These species are more readily identifiable than the cryptic species we examined, but they have yet to be confirmed using molecular analyses. We demonstrate that the use of molecular tools, in conjunction with the post hoc evaluation of phenotypic variation, can aid with the delineation of cryptic biological diversity and with the discovery of key diagnostic features for accurate species recognition in the field.
Full-text available
Article
The iris of the eye shows striking color variation across vertebrate species, and may play important roles in crypsis and communication. The domestic pigeon (Columba livia) has three common iris colors, orange, pearl (white), and bull (dark brown), segregating in a single species, thereby providing a unique opportunity to identify the genetic basis of iris coloration. We used comparative genomics and genetic mapping in laboratory crosses to identify two candidate genes that control variation in iris color in domestic pigeons. We identified a nonsense mutation in the solute carrier SLC2A11B that is shared among all pigeons with pearl eye color, and a locus associated with bull eye color that includes EDNRB2, a gene involved in neural crest migration and pigment development. However, bull eye is likely controlled by a heterogeneous collection of alleles across pigeon breeds. We also found that the EDNRB2 region is associated with regionalized plumage depigmentation (piebalding). Our study identifies two candidate genes for eye colors variation, and establishes a genetic link between iris and plumage color, two traits that vary widely in the evolution of birds and other vertebrates.