W. & T.
ZlEGLER (eds.): HERPETOLOGIA BONNENSIS,
definitions, variation, taxonomic implications
and possible functions
FRANK GLAW & MIGUEL VENCES
Different characters related
rize anurans. Pupil shape, which
distinguish between closely related species,
seldom used taxonomically,
often been emphasi-
fact that taxono-
mists traditionally worked mainly with preserved specimens which
lost their characteri-
stic colour patterns. Just
last decades colour photographs
many anuran species
ve been published, potentially allowing
the use of
In this paper
species from Madagascar
to published data referring
often similar patterns
considered here). Since
rhacophorine genus Boophis
variable between species
A large collection
Madagascan anurans, made between 1987
more than 200 different species,
and 2) to
closely related species.
advertisement calls were used
possibility that specimens with similar morphology
only colour morphs
single species. Most photographs were made
colour slide films.
consulted colour photographs
from different parts
world, mainly those published
South Africa, BLOMMERS-SCHLOSSER
Australia, MENZIES (1974)
Hong Kong, ZHAO
North America, RODRIGUEZ
South America. Colour pictures
most Madagascan species mentioned
the present text have been published
VENCES (1994). Refer
frog collecting localities
generally also apply
and urodeles. However,
excluded that some
correspond with morphological nomenclature.
we did not
study whether 'iris
another tissue type.
Frog eyes - definitions and terminology
Boophis and Mentidactylus}
In several species
In many Boophis
(of special shape
the Malagasy hyperdiid
In some Boophis
peripheral area dorsally
to the right they continue
following pigment cell types:
Melanophores contain black
brown melanine, xantophores
contain real pigment,
physical effects (WINTER 1988).
some anurans xanthophores
also present (DUELLMAN
TRUEB 1986). According
DUBOIS (1976) black eyes
phores, a reduction of xanthophore numbers, and a dominance of melanophores.
The iris of anurans can be dilated or contracted to control the size of the pupil and thus
the amount of light that strikes the retina. Therefore a comparison of iris colouration is only
possible in frogs with similar degree of dilatation. A specimen pictured in darkness will
have a large (blackish) pupil, with the pigmented iris as only a small ring around. Another
problem is that eyes are not always fully opened, especially during the day. In these cases
the colouration of the outer iris area is not or only partly visible (e.g. in Boophis viridis).
According to the summary of DUELLMAN & TRUEB (1986), the retina of anurans and uro-
deles contains four types of receptor cells. These are two types of cones and two types of
and their relative numbers vary in relation to the amount of light to which the species
normally is exposed. Rods are responsible for colour vision in mammals, and the presence
of a second type of rods (green rods) is unique to anurans and urodeles among vertebrates.
PENZLIN (1980), probably referring to relatively ancient literature, reported interspecific dif-
ferences in amphibian colour recognition: no such ability was detected in Alytes, whereas
Rana temporaries Triturus alpestris and T. vulgaris could differentiate between red and
Few experimental data about colour vision in amphibians are available, but these
indicate at least a limited distinction of different wavelengths of light (DUELLMAN & TRUEB
Contrasting this opinion, ZWEIFEL (1992) states that no indication exists that anurans
can distinguish colours.
Pattern and variation of eye colouration
Common eye colour combinations: Several main types of colour
combinations in the iris can be distinguished in Boophis and other anuran genera (see also
I Iris with a blue ring
a) blue or turquoise outer iris area, inner iris area brown: Boophis albilabris occidentalis,
viridis und B. sp.n. 3, and the microhylid Platypelis pollicaris. This colour pattern is also
present in Rhacophorus angulirostris (see MALKMUS 1995 and a picture made by P. HOFF-
MANN in Borneo). In this latter species the bluish colour is present on the dorsal part of the
to a lesser extent on the ventral part of the iris, and not recognizable laterally.
b) blue outer iris area, inner iris area silvery: Boophis erythrodactylus, B. rappiodes, and
the microhylid Plethodontohyla sp.n.
II Iris with a red or yellow ring
a) red or orange outer iris area, inner iris area beige or brown: Boophis luteus, B. guibei,
boehmei, B. miniatus, B. sp.n. 4 and B. sp.n. 7, Litoria chloris from Australia.
b) yellow or whitish outer iris area, inner iris area brown: Boophis madagascariensis, B. cf.
HI Iris uniformly coloured
a) iris greenish: Boophis microtympanum, B. laurenti, many Bufo viridis and B. calamita.
b) iris reddish brown: Boophis goudoti, B. pauliani, Stumpffia tridactyla and many non-Ma-
dagascan species (e.g. Bufo bufo).
a sp.n. 7 (cf.reticulatus
B. cf. burgeri
B. cf. rhodoscelis
Andasibe externally cream white, inside brown
Andasibe externally yellow, inside brown
Ankeniheny externally yellowish, inside brown
An'Ala externally yellow, inside brown
Marojezy externally yellowish, inside brown
golden to light brown
An'Ala externally golden-yellow, inside silvery-grey
Ranomafana externally golden-yellow, inside brown
An'Ala externally red inside, golden-brown
Andasibe externally yellow, inside red-brown
Marojezy above dark brown, externally silvery, inside brown
Andasibe externally red, inside beige
Ankeniheny externally red, inside beige
Ranomafana externally orange, insideTxown
Andringitra externally red, inside beige
Ambohitantely externally red, inside brown
Andasibe golden to red-brown ;
B. microtympanum (normal) Ankaratra green
(yellow mutant) Ankaratra grey-green
a cf. microtympanum Andringitra green
B. laurenti Andringitra green
golden •• j?
externally red, inside silvery-grey J3
golden or brown ••>
B. sp.n. 10
genus Boophis. Species
B. a. albilabris
B. a. occidentals
B. I. luteus
localityIris colourationiris periphery
B. sp.n. 2
B. sp.n.3 (cf. miniarus)
B. sp.n. 4
B. sp.n. 5
B. sp.n. 6
Benayony externally copper, inside brown
Marojezy externally golden, inside brown (copper)
Anjanaharibe externally copper, inside brown
Andasibe externally copper, inside brown
An'Ala above silvery, rest dark grey
Ambatolahy uniformly orange
Isalo externally blue, inside brown
8. /. septentrionalis
B. cf. andreonei
B. cf. sibilans
B. sp.n. 1
B. cf. mandraka
externally red, inside beige
externally red, inside beige
externally red, inside beige
externally yellow, inside silvery
externally white, inside pink-red
externally white, inside red to brown
externally golden, inside red-brown
externally silver-grey, inside brown
externally yellow, inside silver-grey
externally golden, inside silvery
externally silvery or golden, inside red-brown
externally silvery, insjde red-brown
externally yellow, inside red-brown- pattern
externally yellow, inside red-brown pattern
externally yellow, inside silvery with red-brown
externally golden-yellow, inside brown pattern
externally golden-yellow, inside copper
externally blue, inside silvery
externally blue, inside silvery
externally blue, inside silvery
white with brown reticulation
uniformly silvery/externally silvery inside brown
externally black, blue in the middle, centrally brown
Andasibe externally black/blue, inside silvery
Tolagnaro externally red, inside silvery
Andasibe externally trtackAurquoise,cream(middle),inside brown
An'Ala above yellow or orange, inside silvery
An'Ala above and below yellow, inside silvery
An'Ala above and below orange, inside silvery
Mandraka externally orange, inside silvery-beige
M. d'Ambre silvery-golden
Marojezy above yellow, below silvery, inside brown
Maroiezy above orange, inside silvery
Andasibe above blue, inside silvery with reticulations
no periphery, upper margin green
no periphery, upper margin green
no periphery, upper margin green
c) iris red: not known from Boophis, but present in several neotropical hylids: Agalychnis
callidryas and other species of this genus; most species of Duellmanohyla (CAMPBELL &
SMITH 1992); Ptychohyla legleri (see picture in WEIMER et al. 1993a, but compare with
CAMPBELL & SMITH 1992); one colour morph of Eleutherodactylus caryophyllaceus (see
picture in WEIMER et al. 1993b); also visible in a specimen of Phyllomedusa tarsius from
Tarapoto, Peru (SCHLUTER 1987, cover photograph).
d) iris silvery or golden: Boophis cf. rhodoscelis, B. idae, B. opisthodon, B. sp.n. 9, B. te-
phraeomystax, Mantidactylus argenteus, many Heterixalus and many non-Madagascan spe-
IV Iris with reddish or brownish ornamentations
a) irregular reticulations: Boophis mandraka.
b) symmetrical or regular markings: Boophis albipunctatus, B. sibilans, B. sp.n. 1, B. maro-
jezensis, B. elenae.
V Iris dorsally light, ventrally darker: see section iris colour and body colour.
If inner and outer iris area are differently coloured, the inner area is generally less colour-
ful (in most cases silvery, beige or brown). Red iris colour in the genus Boophis is present
in most species groups: B. luteus-gtaap: B. luteus; B. difficilis-group: B. miniatus, B. sp.n.
Boophis goudoti-group: B. boehmei, B. sp.n. 7; B. tephraeomystax-group: B. guibei.
Beside the colour of the iris
the colour of the iris periphery, especially its posterior
part, is an important feature (see tab. 1). Species of the Boophis luteus- and B. rappiodes-
group always have a blue or grey iris periphery, whereas the B. tephraeomystax-group has
an indistinct and inconspicuously coloured, never blue, iris periphery.
Intraspecific geographic variation of eye colouration:
In most Boophis species we did not observe important colour differences in the iris or iris
periphery of specimens from different localities (see tab. 1). However, there are minor dif-
ferences in the iris colouration of Boophis luteus from Tolagnaro (southeastern Mad.) ver-
sus Andasibe and the Ranomafana area (central eastern Mad.). Specimens from both latter
localities are characterized by additional reddish pigment on the iris. Comparable patterns
are also present in the eye of Boophis boehmei from Andringitra and lacking in specimens
from Andasibe, Ranomafana and Ambohitantely.
Geographic variation of iris colouration seems to occur also in neotropical frogs, but in
it can not always be excluded that differences actually refer to (still unrecognized) different
species or are due to misidentifications.
et al. (1993b) figured two colour morphs
of Eleutherodactylus caryophyllaceus from Costa Rica. Both morphs show extremely diffe-
rent iris colouration. A specimen of Phyllomedusa tarsius figured in RODRIGUEZ & DUELL-
MAN (1994) shows a distinctly reticulated iris, whereas a specimen from Tarapoto (Peru)
has a dark red iris (see SCHLUTER 1987: cover photograph).
Altitudinal variation of iris colour was found in Hyla lancasteri ^TRUEB 1968: 293). Few
examples (B. luteus: Tolagnaro, near sea level, versus Andasibe, ca. 900 m) indicate that
such variation could be present also in Boophis. The differences in B. boehmei (see above)
can clearly not be explained by altitudinal variation, since the localities Andringitra and
Ambohitantely are at similar altitude.
Eye colour variation within a population: In several anurans
body colouration differs substantially within a population whereas eye colouration shows re-
latively low variability. This is especially remarkable in an undescribed Boophis (B. sp.n.
3) in which several different body colour morphs occur. Eye colour variation within a po-
pulation mainly concerns colour intensity but not general pattern. The outer iris area and
iris periphery of B. madagascariensis from Andasibe vary from nearly white to orange-yel-
low. In an undescribed Boophis species from An'Ala the outer iris area varies from yellow
to orange, in B. elenae from Andasibe pigment in the inner iris area is red or red-brownish.
Intrapopulational iris colour variation is often correlated to dorsal body colouration (see
Iris colour and body colour: Obvious relationships between iris colour
and dorsal body colour exist between individuals of a population as well as between diffe-
rent species. Some specimens of Boophis cf. mandraka (from Ranomena) have a uniformly
pale green back and a uniformly silvery iris, whereas other specimens with brown pigments
on the back have also brown pigments in the iris. At Andasibe there are specimens of Boo-
phis idae with a silvery back and a golden iris beside specimens with a brownish back and
a golden-brownish iris. Boophis microtympanum has a predominantly green back and a
green iris. A flavistic mutant of B. microtympanum from Ankaratra (with a yellow instead
of green back) had a grey-green instead of a metallic green iris as found in normal speci-
Another example can be found on the photographs in PASSMORE & CARRUTHERS
The iris of a greenish morph oiHyperolius argus (from Richards Bay) has a gree-
nish iris, a brown-backed morph from the same locality has a brownish iris. Two morphs
of Hyla leucophyllata occur at Iquitos (Peru): A dark morph with largely dark iris and a
lighter reticulated morph with a light iris (RODRIGUEZ & DUELLMAN 1994). Another corre-
lation will be described in the section "Obligatorily black-eyed species".: Black eyes are
often associated with aposematic body colouration.
In several anuran families a correlation between colourations of dorsal body and iris can
be observed when related species are compared. The common toad (Bufo bufo) has a brown
or reddish-brown back and a copper iris, whereas Bufo calamita and Bufo viridis, with a
partly greenish back, have also a greenish iris. Males of Rana lessonae are predominantly
yellow with a golden iris during the breeding period, whereas Rana ridibunda has more
dark colour on dorsum and iris. Other examples are obvious in frogs of South Africa (see
photographs in PASSMORE & CARRUTHERS 1995): A brown-backed Kassina maculata
shows a brown iris, that of a more yellow-backed Kassina senegalensis a more golden iris.
The iris of the five Heleophryne species largely reflect their dorsal colouration. The same
is true for the pictured specimens oiHyperolius horstoc/d and H. tuberilinguis. The iris of
the tree bark-like coloured Boophis sp.n. 10 from An'Ala has a colour very similar to the
neighbouring skin. The same is true tor Hyla marmorata from Peru (see photograph in
DRIGUEZ & DUELLMAN 1994).
In other groups such relationships between dorsum and iris colouration are completely
lacking. Among the green species of the Boophis luteus-gmup none has a greenish iris. One
adult flavistic specimen of this group (probably B. sibilans) with a yellow instead of green
back had a normally coloured iris.
In many terrestrial (and some scansorial) species, the eye is concealed in a blackish head-
side or in a completely dark lateral colouration. In such species the dorsal part of the iris
light pigment whereas the ventral part is darker. This pattern, already mentioned
(1986), can be found in many Madagascar! species as
A. sp.n. 1, A. sp.n. 2, several Mantella, in different subgenera of
Mantidactylus (e.g. M.
M. luteus, M. aerumnalis). As far as can be judged
from photographs it occurs also in South African and Australian species as
i, Mixophyes sp.,
(see PASSMORE &
An exceptional correlation between body and eye colouration exists in specimens with re-
or absent pigmentation, classified as "albino" or "semi-albino".
a table with a lot of published cases of albinism in anurans. Comparing the pic-
(1995), we can distinguish between albinos with translucent reddish eyes, semi-albinos with
pigmentless body and normally pigmented eyes, and light yellowish, 'flavistic', specimens
with normally pigmented eyes (often the iris in these specimens is somewhat lighter). Alt-
the reddish-eye-forms have been quoted to have a pigmentless iris, in several pictu-
res rests of the iris pigmentation are still recognizable
1988). The trans-
reddish colour is most probably caused by the blood
in the retina.
Eye colour differences between males and females: There
still many Madagascan frog species in which females or their
colouration are un-
known. In those cases in which photographs of both
are available we found no signi-
ficant differences in the colour of iris or iris periphery between males and females. In the
males in the breeding season have a more yellowish dorsal body
and a golden iris, whereas females have more dark pigment on body and iris.
Since it is currently nearly impossible to determine juveniles of Madagascan frogs, we can
treat intraspecific iris colour differences between juveniles and adults, which probably
exist especially in
Eye colouration and habits: Colourful eyes are mainly found in arboreal-
frogs of the tropics, which often share a "typical" treefrog habitus. Arboreal anu-
have generally relatively larger eyes
aquatic or fossorial species
1986), and large eyes are especially evident in treefrogs. Two treefrog groups, the
family Hylidae and the rhacophorine frogs, show the greatest diversity in eye colouration.
species with colourful eyes occur primarily or exclusively in rainforest habi-
The same seems to be true for treefrog groups in other tropical regions.
Black eyes as rare mutations : Eyes with a blackish iris have been described as rare mu-
in different anuran families (also known from urodeles): Discoglossidae:
et al. 1990); Bufonidae: Bufo
1976); Hylidae: Hyla
1976); Ranidae: Rana clamitans
& NACE 1983), Rana cyanophlyctis (DUBOIS 1976), Rana esculenta (BOULENGER 1897,
DUBOIS 1968, 1979), Rana graeca (BOUI^NGER 1898, after DUBOIS & VACHARD 1971),
Rana lessonae (DUBOIS 1979; DANOVA et al. 1995), Rana nigromaculata (RICHARDS et al.
Ranapipiens (RICHARDS & NACE 1983), Rana ridibunda (see DUBOIS 1979), Rana
sylvatica (RICHARDS & NACE 1983), Rana temporaria (ROSTAND 1953, after DUBOIS &
VACHARD 1973), Rana tsushimensis (RICHARDS et al. 1969).
Some of these variants were kept in captivity and not found to be more fragile or less
healthy than those with normal eye colouration (DUBOIS 1979). Own unpublished observa-
tions on a partially albinotic "black eyed" specimen of Alytes obstetricans confirm this
view. This specimen which was described and illustrated in GALAN et al. (1990) was held
in captivity by us for several years. During that period it did not show obvious differences
in vitality to normal-eyed specimens held in the same terrarium. These observations suggest
that black eye mutations can survive in wild populations under certain conditions.
- Obligatorily black-eyed species : There are several species of anurans, urodeles and sna-
kes with obligatorily blackish eyes which may have their origin in mutations of normal
eyed ancestors. We classify these species into three groups:
Black eyes in not aposematically coloured species are seldom observed. They are com-
mon in the genus Nyctimystes which occurs mainly in New Guinea (MENZIES 1974, COG-
GER 1992, MARTENS 1992). The single Australian species (N. dayi) also has black eyes
(COGGER 1992). A black iris is also typical for Phyllomedusa boliviano which shows a
special behaviour that can be interpreted as deathfeigning (KOHLER et al. 1995). Both gene-
ra Phyllomedusa and Nyctimystes are nocturnal rainforest treefrogs of the family Hylidae
with vertical pupils and a reticulated translucent eyelid, whereas the Black-eyed Litter Frog
(Leptobrachium nigrops) from Singapore (LlM & LlM 1992) and Vibrissaphora boringii
from China (ZHAO & ADLER 1993) are terrestrial.
Aposematically coloured terrestrial species in which the body colouration is largely or
completely uniform yellow, orange or red with a very contrasting largely black iris. Among
these are: Mantella aurantiaca from Madagascar, the neotropical dendrobatids Dendrobates
pumilio and Phyllobates terribilis, the brachycephalid Brachycephalus ephippium from
South America, males of the bufonid Bufo periglenes (however, the latter has not a comple-
tely black iris). Exceptions are found in the bufonid Atelopus zeteld and the Madagascan
microhylid Dyscophus antongili. These species have also a yellow or orange-red body co-
louration but the iris is light.
Aposematically coloured and black-eyed species in which the body is yellow, orange or
red with black. Among anurans this is known from the Madagascan Mantella baroni, M.
M. laevigata, several neotropical dendrobatids (e.g. Dendrobates leucomelas, D.
ventrimaculatus), some specimens of the African microhylid Phrynomantis bifasciatus and
the Australian myobatrachid Pseudophryne corroboree (some light pigment is still visible
in the less aposematically coloured P. dendyi, see COGGER 1992). On the other hand, other
aposematically coloured amphibians (e.g. several Atelopus) do not have uniformly black
Eye colouration and taxonomy
Sibling species with different eye colour pattern: Within
Boophis several species are morphologically very similar to each other and can mainly be
distinguished by advertisement calls. These species can be considered as sibling species.
The majority of these siblings in Boophis can clearly be distinguished by iris colouration:
B. a. occidentalis; B. madagascariensis
B. brachychir; B. elenae
L luteus - B. I. septentrioncdis; B. miniatus - B. sp.n. 3; B. reticulatus - B. sp.n. 7; B.
B. sp.n. 2; B. jaegeri
B. andreonei; B. majori
In contrast, only a few sibling species of the speciose mantelline genus Mantidactylus
(M depressiceps -M. tornieri; M.
fimbriatus) can be clearly distinguished by
iris colouration. Only single examples are available for other Madagascan genera: The
mantelline Mantella baroni has distinctly black eyes, whereas the similar M. pulchra has
light iris pigments; the arboreal microhylid Plethodontohyla sp.n. has a bluish outer iris area
which is brownish in the sibling species P. notosticta. No species of the hyperoliid treefrog
genus Heterixalus can be identified by eye colouration.
This clearly demonstrates that the availability of eye colouration as taxonomic character
strongly depends on the group under consideration. However, single examples of diagnostic
eye colouration between closely related species seem to be widespread among anurans.
Taxonomic implications: Evidence from Boophis supports the assumption
that larger differences in eye colouration occur mainly between taxonomically different
Within frog populations we observed relatively low variability of this character, and
few examples of geographic variation do exist. Judging from our current knowledge, eye
colour patterns can therefore be a rather reliable taxonomic character, especially in groups
with large interspecific variability of eye colour pattern. Larger differences in eye coloura-
tion between different "morphs" of a species (e.g. in Eleutherodactylus caryophyllaceus, see
WEIMER et al. 1993b) may indicate that different taxa are involved. They should be used
as stimulation to search for additional (e.g. bioacoustic) differences, but descriptions of new
taxa should not be based exclusively on eye colouration. Considering the low number of di-
agnostic morphometric characters in anurans, more interest should be focused on the com-
parative investigation of life colouration.
Possible functions of eye colouration
Except for those examples in which iris pigments are integrated into a cryptic body colour
(e.g. the terrestrial species with a light dorsal streak on the iris, or examples like Hyla mar-
any statement on the function of eye colouration remains speculative. In the follo-
wing we will discuss pros and contras of four alternative hypotheses.
Alternative 1 - Eye colouration without any significance : Correlations between
body and iris colouration (see section "Iris colour and body colour") demonstrate that iris
colouration can not be always explained by structures and functions of the iris
can be interpreted as an effect of pleiotropy, indicating a partly common pigment control-
ling system for eye and body. They could also be used as argument to deny a separate
functional significance of iris colour. However, at least two strong arguments stand against
such a 'neutralist' approach. If colourful eye pigments had no adaptive importance,
- differences in pigmentation would be expected to increase with phylogenetic distance.
However, strong differences in iris pigmentation are especially found in closely related syn-
topic sibling species.
- a random distribution of eye colours through ecological frog types would be expected.
In contrast to this expectation colourful eye patterns evolved convergently in different anu-
ran groups with similar habits (arboreal and nocturnal).
Alternative 2 - Eye colouration with physiological significance : It can be argu-
ed that the iris may be the place of some specialized physio-chemical reactions (and the co-
lour a by-product of these processes). The iris could also be considered as a kind of 'waste
pit, for chemical agents which accumulate during certain metabolic activities.
It can not be excluded that light not only reaches the retina through the pupil, but also
through the iris. This iris-filtered light may provide some special (unknown) informations,
especially during the day when the pupil is largely closed. Such a hypothetic physiological
function of iris pigments is in agreement with the fact that some combinations of iris colour
pattern have evolved convergently in different anuran groups. However, this fact can also
be explained by other hypotheses. We are not aware of any other data which could confirm
a physiological or waste pit function of eye colouration, which, in contrast, applies with a
certain probability to several other animal pigmentations such as the regular ornaments of
mollusc shells; in many species these are covered by the periostracum in life, and therefore
can not have any function in communication.
Alternative 3 - Eye colouration with function in prey/predator relationships :
Prey of small and medium sized arboreal frogs are mainly rather small arthropods, and it
seems extremely unlikely that eye pigmentation could play any role in attracting such prey
or distracting it while the frog approaches. A possible function as antipredator adaptation
remains as the only conceivable function of colourful eye pattern in interspecific communi-
Before discussing this subject we first will have a look to the so-called "eyespots" which
are common on the wings of butterflies and can also be found in the inguinal region of se-
veral terrestrial frogs. Leptodactylids of the genera Pleurodema and Physalaemus have large
inguinal glands and assume a defence posture lowering the head and elevating the pelvic re-
gion, thereby presenting the glands to the predator (DUELLMAN & TRUEB 1986). Similar
spots (without gland-like elevation) occur in the Madagascan microhylid Plethodontohyla
ocellata. Such ocelli-like markings have been interpreted as "eyespots" with the suggestion
that the broad pelvic region with elevated "eyes" gives the image of a much larger orga-
nism (DUELLMAN & TRUEB 1986). The interpretation of eyespots as an antipredator mecha-
nism seems convincing at least in some species as Pleurodema cinereum which show exact
imitations of eyes with a dark pupil. On the other hand the eyespots of Physalaemus natte-
reri (figured in DUELLMAN & TRUEB 1986) are uniformely dark and could be interpreted
as imitations of black eyes.
If the ocelli-like markings in the inguinal region of these frogs are really imitations of
it would be easy to imagine that real eyes can also have an impact on predators. This
seems probable for the black eyes of the non-aposematically coloured species of Nyctimy-
Phyllomedusa and Leptobrachium as well as for the strongly contrasting black eyes in
light aposematically coloured species (e.g. Mantella aurantiaca). It is also possible (but not
very probable) for certain types of colourful eyes.
In poisonous species Mullerian Mimicry should be expected and would predict a strong
convergence of eye colour patterns in syntopic species. This convergence is clearly not
found in Boophis. A possible explanation is that Boophis species are probably not particu-
larly poisonous. Thus, predators could potentially learn that they are edible; different eye
colours of related species would be interpreted as a mechanism to prevent habituation of the
predator (apostatic selection). However, the same argument would also predict high in-
traspecific variability of eye colour which is not found.
In species with an aposematic body colouration which includes black parts, the (black)
eyes are concealed in these black parts; thus they do not break up the contrast, enhancing
the effect of the general pattern. Many of these species have been demonstrated to be poi-
sonous (e.g. the frog genera Dendrobates and Mantella, the urodele Salamandra salaman-
dra, the snake genus Micrurus), and it can be assumed that the distinctly contrasting pattern
constitute a strong signal to predators.
In cryptically coloured species in which iris colour is similar to body colour, and in terre-
strial species with darker head side or flanks (see section "Iris colour and body colour") eye
colouration most probably functions enhancing the concealing effect of body colouration.
Alternative 4 - Eye colouration as visual reproductive isolating mechanism : It
can be hypothesized that eye colour is an important optical signal to recognize conspecific
mates and/or rivals. In territorial species such a mechanism would save time and energy by
avoiding superfluous battles with non-conspecifics. The obvious advantage of improved ma-
te recognition is to avoid hybridization which often results in unviable offspring. Gull
species from the arctic region mainly recognize their mates by the colouration of the area
around the eye (SMITH 1967). A striking example of iris colour differences as probable re-
productive character displacement is also found in Asian turtles (MOLL et al. 1981). Frogs
normally use their advertisement calls as primary premating isolation mechanism, but addi-
tional visual mate recognition can be assumed for species with colourful iris pigmentation.
The fact that closely related and syntopic sibling species often strongly differ by eye colour
supports such a function.
The largest problem for this hypothesis is that most frogs (as well as geckos and snakes)
with colourful eyes are nocturnal, often living in the deep darkness of dense forest. The low
amount of available light probably does not allow colour-vision at night. Additionally the
pupil of nocturnal frogs is enlarged at night and the colourful iris is reduced to a narrow
margin along the border of the pupil, thus probably not recognizable. However, during or
after heavy rain several Boophis species display diurnal activity and start calling long befo-
re sunset. The same is true for Nicaraguan^lga/yc/i/iw callidryas (pers. observations). Diur-
nal calling activity from arboreal positions is generally extremely rare in frogs, probably be-
cause of predation by birds. During the day visual recognition of conspecifics could therefo-
re partly replace the acoustic communication. To test this hypothesis it should be studied
to which extent male arboreal rainforest frogs conquer and defend calling territories before
sunset, and how many females approach their breeding water bodies and are clasped by ma-
les during the day.
BEHLER, J.L. & F.W. KING (1979): The Audubon Society Field Guide to North American Reptiles
& Amphibians. - Alfred. A.
Inc., New York, 744 pp. •
BLOMMERS-SCHLOSSER, R.M.A. & C.P. BLANC (1993): Amphibiens (deuxieme partie). - Faune de
Madagascar, 75(2): 385-530 + 20 planches. '"
BOULENGER, G.A. (1897): The tailless Batrachians of Europe. I. - London.
— (1898): The tailless Batrachians of Europe. II. - London.
CAIN, B.W. & S.R. UTESCH (1976): An unusual color pattern of the green tree frog, Hyla cinerea. -
Southwest. Nat., 21: 235-236.
CAMPBELL, J. A. & E.N. SMITH (1992): A new frog of the genus Ptychohyla (Hylidae) from the Sierra
de Santa Cruz, Guatemala, and description of a new genus of Middle American strean-breeding
treefrogs. - Herpetologica, 48(2): 153-167.
COGGER, H.G. (1992): Reptiles & Amphibians of Australia. -Fifth Edition, Reed books, Chatswood,
DANOVA, R., KOTUK, P. & V.
(1995): Ein Fall von Albinismus beim Kleinen Wasser-
frosch. - Salamandra, 31(1): 57-60.
DUBOIS, A. (1968): Sur deux anomalies de la grenouille verte (Rana esculenta). - Bull. Soc. linn.
Lyon, 37(7): 316-320.
— (1969): Sur un crapaud commun aux yeux noirs. - Bull. Soc. linn. Lyon, 38(4): 105-106.
— (1976): Deux Rana cyanophlyctis du Nepal aux yeux noirs (Amphibiens, Anoures). - Bull. Soc.
linn. Lyon, 45: 303-307.
— (1979): Anomalies and mutations in natural populations of the Rana "esculenta" complex (Am-
phibia, Anura). - Mitt. Zool. Mus. Berlin, 55: 59-87.
DUBOIS, A. & D.
(1971): Sur une anomalie pigmentaire de la grenouille verte (Rana escu-
lenta) et de quelques autres amphibiens anoures et urodeles. - Bull. Soc. linn. Lyon, 40: 40-52.
DUEIXMAN, W.E. (1970): The hylid frogs of Middle America. - Monog. Mus. Nat. Hist. Univ. Kan-
DUELLMAN, W.E. & L. TRUEB (1986): Biology of Amphibians. - New York (McGraw-Hill), 670 pp.
ENGELMANN, W.-E. & FJ. OBST (1976): Partielle Pigmentlosigkeit bei Bufo viridis viridis (Amphi-
bia, Anura, Bufonidae). - Zool. Abh. Mus. Tierk., Dresden, 34:
GABRIEL, M. (1987): Ein albinotischer Seefrosch Rana ridibunda PALLAS, 1771 aus Sibenik, Jugosla-
wien (Anura: Ranidae). - Salamandra, 23(4):
GALAN, P., VENCES, M., GLAW, F., FERNANDEZ ARIAS, G. & M.
gen zur Biologie von Alytes obstetricans in Nordwestiberien. - herpetofauna, 12(65): 17-24. <,,
GLAW, F. & M. VENCES (1994): A Fieldguide to the Amphibians and Reptiles of Madagascar. 2nd |
edition, including mammals and freshwater fish. - K61n (Vences & Glaw Verlag), 480 pp. + 48 !
colour pages. '?
KARBE, B. & D. KARBE (1988): Fund eines adulten Albino-Grasfrosches Rana temporaria LINNAEUS, |
1758 in Overath, Oberbergisches Land (Anura: Ranidae). - Salamandra, 24(4): 316-317.
KARSEN, S.J., LAU, M.W. & A. BOGADEK (1986): Hong Kong Amphibians and Reptiles. - Urban
Council Hong Kong, 136 pp.
KLEMZ, C. & K.-D. KOHNEL (1986): Ein Beitrag zum Albinismus beim Moorfrosch Rana arvalis
NILSSON, 1842. - Salamandra, 22(2/3): 190-195.
KOHLER, J., IBISCH, P.L., DIRKSEN, L. & W. BOHME (1995): Zur Herpetofauna der semihumiden Sa-
maipata-Region, Bolivien. 1. Amphibien. -herpetofauna, 17(98): 13-24.
LIM, K.K.P. & F.L.K. LIM (1992): A Guide to the Amphibians and Reptiles of Singapore. - Singa-
pore Science center, 160 pp.
MALKMUS, R. (1993): Ein Albino des Grasfrosches Rana temporaria LINNEAUS, 1758 aus dem Spes-
sart, Nordwestbayern. - herpetofauna, 15(85): 17-18.
— (1995): Kopfseitenflecken bei beiden Geschlechtern von Rhacophorus mgulirostris AHL, 1927
(Amphibia: Rhaoophoridae). - Salamandra, 31(4): 245-246.
MARTENS, H. (1992): Froschlurche aus Papua-Neuguinea. - herpetofauna, 14(79): 18-24.
MENZIES, J.I. (1974): Handbook of common New Guinea Frogs. - Wau Ecology Institute Handbook
1, Wau, Papua New Guinea, 75 pp.
E.O., MATSON, K.E. & E.B. KREHBIEL(1981): Sexual and seasonal dichromatism in the Asian
River Turtle Callagur borneoensis. - Herpetologica, 37(4): 181-194.
NOLLERT, A. & C. NOLLERT (1992): Die Amphibien Europas. - Stuttgart (Franck-Kosmos Verlag),
N.I. & V.C. CARRUTHERS (1995): South African Frogs. A Complete Guide. - Revised
edition, Southern book publishers, Witwatersrand University Press, 322 pp.
PENZLIN, H. (1980): Lehrbuch der Tierphysiologie (3. Auflage). - Stuttgart (Gustav Fischer Verlag),
RICHARDS, CM. & G.W. NACE (1983): Dark Pigment Variants in Anurans: Classification, New De-
scriptions, Color Changes and Inheritance. - Copeia, 1983: 979-990.
RICHARDS, CM., TARTOF, D.T. & G.W. NACE (1969): A melanoid variant in Rana pipiens. - Co-
peia, 1969: 850-852.
RODRIGUEZ, L.O. & W.E. DUELLMAN (1994): Guide to the Frogs of the Iquitos Region, Amazonian
Peru. - The University of Kansas Natural History Museum, Special Publication No. 22, 80 pp.
+ 12 plates.
ROSTAND, J. (1953): Sur l'anomalie "iris brun" chez Rana esculenta L. et chez Rana temporaria L.
- C. r. Acad. Sci., 237: 762-764.
SCHLOTER, A. (1987): Die Froschlurche an einem StillgewSsser im tropischen Regenwald von Peru.
- herpetofauna, 9(50): 17-24 (title photograph).
SMITH, N.G. (1967): Visual isolation in gulls. - Scientific American, 94-102.
L. (1968): Variation in the Tree Frog Hyla lancasteri. - Copeia, 1968(2): 285-299.
WBMER, R., FEICHTINGER, W., BOLANOS, F. & M. SCHMID (1993a): Die Amphibien von Costa Rica.
Herpetologische Eindriicke einer Forschungsreise. Teil 1: Einleitung, Hylidae (1). - Sauria, 15(2):
— (1993b): Die Amphibien von Costa Rica. Herpetologische Eindriicke einer Forschungsreise. Teil
Leptodactylidae (1). - Sauria, 15(4): 19-24.
WINTER, H.G. (1988): Farbungund Zeichnung. -In: KLEWEN, R.: Die Landsalamander Europas, Teil
- Die Neue Brehm Bucherei, Wittenberg Lutherstadt (A. Ziemsen Verlag), pp. 54-55.
ZHAO, E. & K. ADLER (1993): Herpetology of China. - Society for the Study of Amphibians and
Reptiles, Oxford (Ohio), 522 pp., 48 plates, 1 map. Contr. Herpetol. 10.
ZWHFEL, R.G. (1992): Frosche und Kr6ten. - In: COGGER, H.G. & R.G. ZWEIFEL (eds.): Reptilien
und Amphibien, Hamburg (Jahr-Verlag), pp. 76-105.
Authors' addresses: FRANK GLAW and MIGUEL VENCES, Zoologisches Forschungsinstitut und Muse-
um Alexander Koenig, Adenauerallee 160, D-53113 Bonn, Germany.