Throughout the tropical savanna of Africa, and abundantly present on suitable habitats there is always found a characteristic, conspicuous Hyperolius which, although highly variable, shows a number of common characters, most of them unique within the genus:
- high population density and exposed calling site
- strictly savanna-living
- a characteristic melodic voice, like small xylophones
- a brief snout
- very large shagreened gular flap and large dilatable vocal sac in males
- transverse gular fold in females
- extensive webbing
- a large clutch of eggs, deposited in water
- absence of external metatarsal tubercle
- horizontal pupils
In the colour pattern of phase F, however, there is extreme variation in these savanna-living forms, between populations and sometimes within populations, a variation which is probably greater that in any other vertebrate animal. In general this variation can be characterised thus:
- variation in some populations is small, while others are highly variable.
- in some populations variation results in distinct morphs (Mo-1, Mo-2 and Mo-3), in others there is an intergrading variation.
This form is always present on the savanna and its beautiful voice is as characteristic of the African night as that of Kassina senegalensis. It is also a characteristic of the complex that one form, and only one such form, occurs at any given locality with very few exceptions.
The variation in colour-pattern within this group has been analysed by Schiøtz
(1971). The conclusions are that single characters (i.e., dorsum green or brown, or subdermal dark lateral band present or absent) are generally widely distributed, beyond the distribution of the recognised subspecies. Furthermore, the different characters vary independently of each other. Finally, when sufficient samples are available to demonstrate it, most recognised subspecies show a gradual transition in pattern to the neighbouring forms. This clinal trend has not been observed by Laurent in Central Africa; instead, he recognizes hybrids between subspecies, which may be saying the same.
A peculiar character used in the taxonomy of this group is the subdermal dark lateral streak, a black pigmentation of the musculus obliquus externus in some forms which reveals itself as a dark lateral band, dark bluish in life, and sometimes difficult to see if the sides are heavily pigmented. This character tends to disappear after some time in alcohol. There is a difference in the distribution of this character in the two sexes. It is widespread in the parallelus subgroup but is also found in several viridiflavus forms in eastern Africa.
It is generally recognized that more than one species is involved since some cases of sympatry between different forms has been reported.
There is, however, no agreement as to the species structure. The forms from southern Africa have traditionally been regarded as subspecies of H. marmoratus. Schiøtz has argued that a number of forms from western-southern Central Africa have so many common characters that they should be regarded as belonging to one species H. parallelus. Several authors have regarded some of the forms as full species.
A recent paper (Wieczorek et al. 2001) has split up part of the forms in this superspecies into 10 full species. Also, some other authors have regarded some of these forms as separate species. Considering the taxonomic uncertainty, such forms are treated both under the heading Hyperolius viridiflavus, with the distinct forms treated as vernacular (or subspecies) names, as well as under their respective names. Further DNA studies may resolve these issues. Taxa are listed here roughly in geographical order, starting in West Africa.
The viridiflavus subgroup:
Most often small forms, Phase F not falling into distinct morphs as in the parallelus-marginatus subgroup. Some populations show very limited variation, others very much.
Phase F light grey with small black dots. Perhaps part of the
Nitidulus-cline, although Emms et al. (2005) disagree.
Westernmost West Africa.
incl. AUR of Schiøtz
Phase F dorsum brown to grey, with or without darker spots and marbling. Flanks spotted, marbled, or with a dark band.
There is a west-east cline in this form which may also include Spatzi: The western form (Sierra Leone) has an unmarked dorsum and few dark spots on the flanks. Further east the flanks become more spotted, marbled or with a lateral band, until in the Cameroun population (Aureus) we have a conspicuous dark lateral band and often a spotted dorsum. The population from Sierra Leone has phase J with an hour-glass pattern, rare in this part of Africa.
From Sierra Leone to Cameroun.
incl. BAN, ANE
Dorsum uniform greyish or with minute black dots, lateral area with small black dots.
These forms are described from the drier savanna of Cameroun to Rep. Centrafricaine, and may also occur in the dry, northern savanna of Nigeria.
Dorsum brown with small round yellow spots, often with a black centre.
Northern R. D. Congo, southern Sudan, south-western Ethiopia, probably northern Uganda.
Dorsum green with small rounded yellow to red spots, often with a black centre. Ethiopia, north-western Kenya.
(1998) has suggested that although samples from Kenya show the typical characters for this form they may be isolated from the Ethiopian populations (incl. the type locality), by a broad expanse of arid lowland, possibly unsuitable for this species.
Dorsum green to bluish green, unmarked or with small dark blue to black dots and freckles; posterior face of thigh iridescent pale blue to purple with blackish speckling, ventrum yellowish, sometimes speckled with grey or black.
Largen (1998) brought attention to the existence of this distinct form.
Low to moderate altitude on floor and eastern wall of Rift Valley, Ethiopia.
incl. BAY, SCH, KWI
A most variable form. Dorsum finely spotted to coarsely marbled in green, blue and black. No small red dorsal spots.
A form called Bayoni (BAY of Schiøtz
(1971), H. v. bayoni Schiøtz
(1975), is distributed over a greater part of Uganda and Parc Virunga, R. D. Congo, and therefore well represented in collections. It combines the characters of the northern Pachydermus and Viridiflavus (small round yellow or red dorsal spots) with those of the southern Variabilis (very variable pattern) and populations furthermore show a gradual transition from one to the other. It is a matter of personal choice whether one recognises Bayoni as a distinct subspecies or simply as a range of gradual transition forms in one of the few areas in Africa sufficiently well collected to demonstrate such transition. Schubotzi seems very similar to Bayoni and Kwidjwiensis to Variabilis.
At Sango Bay Variabilis is sympatric with Argentovittis, the two forms showing conspicuous differences in voice and size, no transition in pattern and very few signs of hybridisation. This is the only case of sympatry in the Hyperolius viridiflavus complex I have met - although several other instances have been reported in the literature - and is clear evidence that the group includes more than one species. The sympatry could be explained by recent colonisation from the north (Variabilis) and south (Argentovittis) in Uganda west of Lake Victoria, an area where the savanna, the habitat for this group, has replaced the forest in recent times.
Rwanda, Burundi, Southwest Uganda, Northwest Tanzania.
Many specimens have longitudinal stripes.
A montane form from Mokoto Lakes (1700-2000 m), R. D. Congo.
A large form from montane grasslands. Dorsum uniform brown. Flanks in life bluish (presumably what is in this book termed subdermal dark lateral streak) and red ventrum. Mt. Karissimbi, Rwanda, above 2000 m.
incl. PIT, XAN.
A number of rather large montane forms, Karissimbiensis, Pitmanni, Xanthogrammus and Françoisi from eastern R.D. Congo and adjacent areas are so similar in appearance that it is questionable whether they should be maintained as separate forms. While Karissimbiensis has a uniform dorsum, Pitmanni often has small yellow spots on the brown dorsum, Xanthogrammus normally so, sometimes arranged in rows, while Françoisi has a yellow network on the dorsum. These forms are found isolated from each other in the highlands of eastern R. D. Congo, south-western Uganda, Rwanda and Burundi, surrounded by lowland forms such as Variabilis and Bayoni. They are often found in open swamps in forested areas.
Ssp from Marsabit
There is an undescribed form from Marsabit, Kenya, possibly isolated from other forms by semi-desert. Large, with the dorsum uniform grey or brown. Subdermal dark lateral streak present.
A large form, phase J with hourglass pattern. Phase F dorsum dark with diffuse light spots. No subdermal dark lateral streak.
Highlands north of Nairobi, Kenya.
Brown to grey, uniform or with minute black spots.
There may be a gradual transition towards the similar Pantherinus so that it may be unnecessary to maintain two names.
Uplands of southern Kenya, including the Nairobi area.
A small form from the dry savanna of southern Kenya, dorsum brown with darker brown longitudinal lines.
This form was described as a full species, since it differs rather much from the adjacent forms in this complex.
Dry savanna of southern Kenya.
A small form. Dorsum dark with lighter vermiculation. Coastal Kenya, possibly southern Somalia.
Dorsum dark, almost black with white spots and streaks.
Near Teita Hills, Kenya.
Dorsum dark, almost black with small white rings. Apparently no gradual transition to the similar Glandicolor. Between Kilimanjaro and Mt. Meru, Tanzania.
Dorsum and limbs translucent green, sometimes with bluish green spots. Ventrum white or green. Throat of males yellow or green.
A very distinct form with a unique coloration, not met in any other member of the superspecies.
Collected in swamps near Lake Victoria in the vicinity of Mwanza. None were heard in the Papyrus on the shore itself. One would expect frequent gene exchange across Lake Victoria with its many floating Papyrus islands and, nowadays, its semi-solid layer of Eichhornia, so this distinct form with restricted distribution is a surprise.
Dorsum grey to brown, sometimes spotted or marbled. Flanks marbled. Found over a large area in upland Tanzania, above the escarpment formed by the Eastern Arc Mountains.
Dorsum coarsely marbled in light and dark brown. The pattern seems constant. A very few Goetzei show the same pattern.
Ngorongoro Crater, Tanzania.
Very similar to Mariae. The main difference is that the latter has a black band along the canthus rostralis, continuing over the upper eyelid. Schiøtz's sample from Amani, Usambaras, has specimens with and without black markings on the side of the head, so the variation along the Tanzanian should be studied.
Distribution follows Pickersgill's coastal regions of Tanzania and Unguja. Pickersgill's distribution for H. mariae is restricted to eastern and north-eastern Tanzania, avoiding the coastal strip.
A small form with a uniform yellow to brown dorsum and a conspicuous subdermal dark lateral streak. A black canthal stripe is often present, sometimes continuing behind the eye.
The pattern is related to Goetzei and Reesi, and intergrades to Goetzei seem to occur.
The voice is somewhat different from that of other members of the group, a fast series of high-pitched clicks.
Eastern Tanzanian lowlands, Pemba, Zanzibar.
Dorsum light yellow to light green. Flanks with a broad subdermal dark lateral streak overlaid by 2-3 vertical yellow stripes.
This strikingly coloured form undoubtedly belongs to the viridiflavus complex, vicariating for other forms in the Kilombero Floodplains, Tanzania. If one disregards the conspicuous bright yellow stripes, the pattern is very similar to that of Mariae to the north-east and to Bitaeniatus and Rhodoscelis to the Southwest.
Dorsum uniform yellowish or with black spots. A subdermal dark lateral streak present. South-western Tanzania and north-eastern Zambia.
Can be regarded as a transition from Rhodoscelis to Goetzei. South-western Tanzania.
A large form. Dorsum uniform yellow.
A dark subdermal lateral streak with a thin bright yellow longitudinal line.
Eastern Katanga (R. D. Congo) and north-western Zambia. The range seems to overlap that of Melanoleucus, an argument for the two being specifically distinct. Sympatry between the two has been reported by Laurent (1943) from Kasenge.
Dorsum yellow, uniform or with dark spots, sometimes arranged in longitudinal rows.
Northern shores of Lake Malawi.
Dorsum varies from jet black to silvery white with a dusting of black. Belly light pink and hidden parts of limbs vivid pink. No subdermal dark band. Maximum length 27 mm.
Only known from Beira, Mozambique, where Pickersgill also found many intergrades of H. v. taeniatus.
Dorsum dark with five straight light bands which may sometimes be broken up into spots.
Intergrades with Marmoratus, but apparently not with Albofasciatus.
KwaZulu-Natal to Zimbabwe and Malawi.
Dorsum marbled or mottled, very variable.
Central coastal South Africa.
Spotted dorsum. Intergrades to marmoratus.
Eastern Cape province, South Africa.
The parallelus-marginatus subgroup:
The following forms from southern and south-western Africa are large with well defined female morphs (1, 2 and 3), not all of which may be present in each population. Two of these morphs (1 and 3) vary very little from form to form throughout the vast range of the subgroup:
- Mo. 1. Dark dorsum with light middorsal stripe and more or less well-defined light lateral stripes.
- Mo. 3. Dark dorsum with fine light vermiculation or spots
- Mo. 2 is the term used for a number of variable patterns, in some populations forming transitions between Mo. 1 and 3, while in other forms it varies gradually, without the existence of well-defined morphs 1 and 3.
The frequency of these morphs varies from 0-100% in the different populations, so that this variation in itself is a systematical character.
In this group (unlike the viridiflavus-group) there are two tibial patterns: in some populations tibia is always finely spotted regardless of the dorsal pattern, while in others it is marked as the dorsum, that is finely spotted if the specimen is morph 3, coarsely marbled if it is morph 1 or 2.
There seem to be gradual clines between the populations with the different forms merging into each other.
Nearly all phase F are morph-1: Dorsum dark with narrow white dorsal and lateral lines, the latter well separated from the light ventrum. Tibia light with small dark and white spots. Subdermal dark streak present in females, absent in males. M-3, only recorded from R. Congo, is pale fawn, minutely freckled with darker grey-brown pigment. Synonymous with H. bocagei (Schiøtz and Van Daele 2003).
Coastal R. Congo, western R. D. Congo, north-western Angola.
Morph 1 with dark dorsum with very wide white dorsal stripe. Tibia coarsely marbled. Subdermal dark lines present in females. The form seems ill-defined, and there are intergrades to Parallelus and Huillensis/Angolensis. May be synonymous with H. bicolor.
incl. HUI, QUA
Almost all specimens in the type series from north-eastern Angola are morph 3 with a dark dorsum with light vermiculations. Subdermal lateral streak present in both sexes. Further south, in Botswana and Zambia, much more variable and could be regarded as a separate subspecies.
A form called Huillensis from the south-western part of Angola is described as having round dotson the dorsum rather than vermiculation but the known specimens are so few and the variation in pattern so great that its distinctness from Angolensis seems doubtful. Quarrei from southern R. D. Congo also seems ill-defined.
Known from Angola, northernmost Namibia, north-western Botswana and western Zambia.
Light ground colour with spots or vermiculation in red-brown. Subdermal dark streak present in both sexes. Mo 1 rare. Seems ill-defined in an area with many very variable but similar forms (Angolensis, Alborufus, Pyrrhodictyon, Aposematicus).
Distributed where Angola, Zambia and R. D. Congo meet.
Morphs 1, 2 and 3 present. Tibia finely spotted (northern populations) or as dorsum (southern populations.). Subdermal dark streak present in both sexes. Light lateral line not well defined. Morph 1 pattern often irregular. No red pigmentation after preservation.v
We lack collections from R. D. Congo south of the forest that may show a gradual transition to the very similar Parallelus. There seems to be a gradual transition to Angolensis.
At Sango Bay, Uganda, Argentovittis is sympatric with Variabilis and at Gita, Rwanda with the so-called Schubotzi (Variabilis).
Southern Uganda, eastern R. D. Congo, Rwanda, Burundi, western Tanzania.
An isolated highland form. Dorsum uniform light, diffuse lateral spots. Subdermal dark streak present.
Parc Upemba, R. D. Congo. Possibly a member of the viridiflavus subgroup, related to Rhodoscelis.
Dorsum uniform yellowish to green or almost black, delimited laterally by black. Flanks light with red spots, sometimes with black centre. In Schiøtz (1971) I argued for full specific rank for this form, but Laurent (1976) is probably correct in regarding it as a member of this subgroup. Sympatric with Nyassae (a viridiflavus-form) in Malawi (Fonesca & Jorque 1979).
Central and western Zimbabwe, eastern Mozambique, eastern central Zambia, northern Malawi, south-eastern Tanzania.
The distribution is disjunct since the Tanzanian populations are isolated from the more western populations by several forms of the viridiflavus subgroup, perhaps a reflection of the two subgroups being specifically distinct.
The large type series from south-eastern R. D. Congo (240 phase F) are all morph 1, with a very regular pattern. Several separate oblique light lateral lines. Red lines centrally in light lines, especially on flanks. Further south, in central Zambia, this form is very variable both within and between populations and the separation into several subspecies is justifiable.
Further south there is a transition to Pyrrhodictyon.
South-eastern R. D. Congo to central Zambia.
Dorsum uniform brownish, often with rather diffuse lighter (green in life) spots, sometimes with a black centre. Ventrum with red spots or vermiculation.
The populations of the parallelus-marginatus subgroup in southern Zambia show extreme variation and the present nomenclatorical separation from south-eastern populations of Melanoleucus, from Rhodesianus and Aposematicus is very subjective.
Kafue River and uplands to the south, Zambia.
Pattern very variable both within and between populations. Spotted to marbled. No red pigmentation.
Ventrum with strong red vermiculation which disappears after preservation. Dorsum yellow with green spots. Perhaps not distinguishable from Pyrrhodichtyon.
Only morph 1 present. Very little variation in pattern. Red lines in the white pattern disappear after preservation. Light dorsolateral lines well-defined. No red ventral pigmentation. Tibia as dorsum. It has been discussed whether there are populations showing transitions to Taeniatus, a very variable form belonging to the viridiflavus sub-group. Recent studies (Channing et al.) seem to show that the two are specifically distinct. Transitions to Marginatus are known.
Very variable and often asymmetrical dark dorsal marbling on light ground colour. No subdermal dark lateral streak.
South-eastern highlands of Zimbabwe.
Only morph F-1 with very regular pattern present, tibia as dorsum. No light dorsolateral lines but black or red spots on flanks. No subdermal dark lateral streak.
Very similar to other forms such as Broadleyi and Melanoleucus, but geographically separated from these by other, very different forms (Marginatus and Taeniatus). Apparently no transition to these. Highlands of southern Malawi.
Distribution and Habitat
Country distribution from AmphibiaWeb's database: Ethiopia, Kenya, Tanzania, United Republic of
This species shows developmental changes in patterning, with two phases, J (juveniles and many mature males) and F (mature females and some mature males). All newly metamorphosed individuals are phase J, which is normally brownish to green with paired light dorsolateral lines, or an hourglass pattern. All females, and some males, develop into phase F before the first breeding season. Phase F is often colorful and variable, showing the diagnostic color characteristics for the species or subspecies. Either well-defined morphs may be present, or graded variation.
This account was taken from "Treefrogs of Africa" by Arne Schiøtz with kind permission from Edition Chimaira publishers, Frankfurt am Main.
Emms, C., Jambang, M. D. K., Bah, O., Mankali, B., Rödel, M. O., and Barnett, L. (2005). ''The amphibian fauna of the Gambia, West Africa.'' Herpetological Bulletin, 94, 6-16.
Largen, M.J. (1998). ''The status of the genus Hyperolius (Amphibia Anura Hyperoliidae) in Ethiopia.'' Tropical Zoology, 11(1), 61-82.
Laurent, R.F. (1976). ''Noveaux commentaires sur la superespèce Hyperolius viridiflavus.'' Annales du Musée Royal de l'Afrique Centrale, 213, 70-114.
Schiøtz, A. (1971). ''The superspecies Hyperolius viridiflavus (Anura).'' Vedenskabelige Meddelelser fra Dansk Naturhistorisk Forening, 134, 21-76.
Schiøtz, A. (1975). The Treefrogs of Eastern Africa. Steenstrupia, Copenhagen.
Schiøtz, A. (1999). Treefrogs of Africa. Edition Chimaira, Frankfurt am Main.
Schiøtz, A. and Von Daele, P. (2003). ''Notes on the treefrogs of the North-Western Province, Zambia.'' Alytes, 20(3-4), 137-149.
Wieczorek, A. M., Drewes, R. C., and Channing, A. (2001). ''Phylogenetic relationships within the Hyperolius viridiflavus complex (Anura: Hyperoliidae), and comments on taxonomic status.'' Amphibia-Reptilia, 22, 155-166.
Written by Arne Schiøtz (arne AT schiotz.dk), *
First submitted 2001-01-29
Edited by Kellie Whittaker, updated by Michelle S. Koo (2013-02-06)
Feedback or comments about this page.
Citation: AmphibiaWeb: Information on
amphibian biology and conservation. [web application]. 2016. Berkeley, California:
(Accessed: Jul 23, 2016).
AmphibiaWeb's policy on data use.