Kinyongia asheorum, Necas, Sindaco, Koreny, Kopecna, Malonza & Modry, 2009

Necas, Petr, Sindaco, Roberto, Koŕený, Luděk, Kopečná, Jana, Malonza, Patrick Kenyatta & Modrý, David, 2009, Kinyongia asheorum sp. n., a new montane chameleon from the Nyiro Range, northern Kenya (Squamata: Chamaeleonidae), Zootaxa 2028 (1), pp. 41-50 : 43-47

Kinyongia asheorum sp. n.

Holotype. MCC 1 R 1212 , adult male collected 14.II.2004 (estimated elevation 2.000 -2.400 meters) on the eastern slope of Mt. Nyiro , between Korante and the top (Kosi Kosi), Northern Frontier Division of Kenya. The animal was donated to Roberto Sindaco and Luca Borghesio by a local tribesman.

Paratypes. Paratype 2 1 - NMK L/2699 , female collected 06.IX.2003 at Korante clearing, approx. 8 km west of South Horr on the upper part of Mount Nyiro , Northern Frontier Division of Kenya, N 02°07´57.4´´ N, 036°51´51.8´´ E, 2.450 m by David Modrý and Jana Kopečná . Paratype 2 - MCC R 1203 / NMK L/2595 , subadult male collected 06.II.2004 on the western slope of Mt. Nyiro , above Tuum (N 02°08’56’’, E 36°48’50’’) at 2.270 m by Roberto Sindaco and Luca Borghesio . Paratype 3 – ZFMK 3 84821 was collected by local tribesman in montane forest on the eastern slope of Nyiro range .

Diagnosis. A medium-sized, slender chameleon reaching a maximum total length of around 220 mm, with tail length clearly exceeding the head-and-body length. The body scalation is subhomogeneous, with slightly enlarged elongate lenticular scales on the flanks. The gular region wears paramesially enlarged semiovoid tubercular scales, submentally, there is a group of several high appendicular conical scales (the highest are more than 10 times as high as wide at their base), not ordered in a medial row. The ventral crest is indistinct. The head wears a distinct casque, clearly exceeding the dorsal crest on the neck. The canthus parietalis and canthi laterales are formed by enlarged tubercular, sometimes keeled and/or pointed scales. The temporal crest is completely absent. Canthus supraorbitalis is composed of two parallel rows of clearly enlarged scales. Canthi rostrales are composed of strongly enlarged pointed scales; they terminate rostrally on both sides in a long, stump horn (diverging slightly from each other). Kinyongia asheorum n. sp. differs from all known chameleons by the unique scalation of the gular region consisting of paramesial semiovoid scales and a field of several high appendicular pointed scales.

Description of the holotype. The male type specimen of K. asheorum n. sp. is a slender chameleon, with a total length of 219 mm and head-and-body length 92 mm. The tail is very long (58% of the total lenght). The scalation is subhomogeneous, consisting of small round scales and slightly enlarged lenticular, mostly elongate scales, disseminated especially on the cranial half of the flanks. The belly, legs and tail are scaled almost homogeneously. The tail scalation differs significantly on the upper side and underside. While the scales on the upper side of the tail are analogous to the body scalation, the scales on the underside are quadrangular, ordered in transversal rows. The fingers terminate in a fine, single, white claw. There are no tarsal spurs on hind-feet; the soles of the extremities are smooth.

The dorsal crest is developed only cranially, consisting of eight solitary stump conical scales decreasing successively in size caudally. The gular crest wears a field consisting of seven high and thin pointed appendicular scales (the highest are more than ten times as high as wide at base), not ordered in a medial row but situated both medially and paramesially. The throat has narrow longitudinal grooves between the rows of scales, that are either granular or, paramesially, heavily enlarged, prominating and elongately oval or semiovoid. The ventral crest is indistinct; not even indicated by a white midventral line.

The head is relatively long, being more than twice as long as wide (ratio = 2.13). It wears a distinct casque, covered by enlarged flat scales. The level of the casque exceeds the level of the dorsal crest on neck by 5 mm. The head length (from the tip of mouth to the posterior margin of the casque) is 32 mm; the casque height (from the mouth angle to the top of the casque) is 20 mm, slightly less than the length of the mandible (22 mm). There are no occipital flaps.

The distinct parietal crest consists of six heavily enlarged, tubercular, keeled and/or pointed scales; the

1. Museo Civico di Storia Naturale, Carmagnola, Italy

2. National Museum of Kenya, Nairobi, Kenya

3. Zoologisches Forschungsmuseum A. Koenig, Bonn, Germany caudal (last) one is heavily enlarged and pointed. The supraorbital crest is composed of a double row of enlarged tubercular scales, giving it a swollen and serrate appearance. The proper lateral crest consists of heavily enlarged, pointed, tubercular scales. Canthus rostralis consists of three extremely enlarged tubercular scales becoming successively larger rostrally. The rostral scale is elongated, stump and conical, forming a horn, pointing upwards and slightly sidewards ( Figs. 1 and 2 ). The temporal crest is completely absent. The nares are situated at about 1/2 of the distance between the mouth tip and the rostral rim of the orbit. The snout is pointed.

Coloration and pattern. When alive, the coloration of the type was as follows. The body scales are uniformly dark green, except of two white spots, one at the mouth angle, the second in the centre of the flank. The interstitial skin is white. The combination of light interstitial skin and darker scales gives the entire body, head and extremities a reticulated pattern. The dorsal crest is green, with brownish shades. The pointed gular scales are white. The eyelids were pinkish brown, the upper side of the head, especially the cranial crests, were reddish brown ( Fig. 4 ). The extremities are the same color like the body outside, white inside.

Color in alcohol. Uniform pale grey, the insides and soles of legs are cream-white. The whitish spots on the flanks and above the mouth tip are still visible. The head is lighter with a brownish shade.

Etymology. The Mount-Nyiro Bearded Chameleon is named in honor of James and Sanda Ashe, in respect to their life-long contribution to African herpetology and, also, as an expression of our deep mourning for James Ashe, who passed away recently. The name itself ( asheorum ) is formed as a plural genitive case of the family name (Ashe).

Variation of the paratypes. In the female specimen of K. asheorum n. sp. (paratype 1), the dorsal crest is developed only in the cranial fourth of the back, the proper lateral crest is developed weakly supratemporally, but becomes well visible towards the top of the casque, consisting there of enlarged, pointed, tubercular scales, there are two large scales on the top of the casque in comparison to the type, which has only one ( Figs. 3 and 5 ). Canthus rostralis consists of three extremely enlarged tubercular scales becoming successively larger rostrally. The rostral scales are conical, pointing upwards. Paratype 2 is an immature male, therefore, the enlarged scales are not fully developed yet. The appendicular gular scales are underdeveloped, the rostral horns are smaller than in the type and are not divergent but parallel. Paratype 3 is an adult male. It is in full accordance with the type. Being about 20% smaller, it is worth to remark, that the casque is relatively bigger and the horns relatively longer than in the type. The horns are parallel.

When alive, the coloration of paratype 2 was the same as in the holotype. The female was colored slightly differently: the body scales were uniformly light green, except of two white spots, one at the mouth angle, the second in the centre of the flank. The interstitial skin was white, except of a longitudinal stripe on each flank and eyelid, where it was colored with yellow to orange. The dorsal crest was green, the gular crest was white. There was a Y-shape marking on the underside of the tail and undersides of femurs. Preserved specimens are uniformly pale grey, the insides and soles of the legs are cream.

Distribution and habitat. Mt. Nyiro is a massif of mainly Precambrian gneisses ( Dodson 1963) rising steeply up to 2800 m from a flat plateau lying at about 800-1000 m of altitude. According to Herlocker (1979), three successive belts of vegetation are stratified on Mt. Nyiro.

K. asheorum n. sp. is known only from Mt. Nyiro montane forest. The only four specimens available were caught at an altitude of 2'000-2'450 meters, either near the upper edge of the semi-deciduous bushland area (above Tum), or in a clearing in a montane forest. The Afro-montane evergreen forest that occupies most of the upper plateau of Nyiro range is dominated by the Conifers Podocarpus latifolius and Juniperus procera , being mixed with broadleaved trees (e.g. Olea africana , O. capensis , Prunus africana , Faurea saligna ). The highest trees can be up to 35 m high, and canopy is usually 20 m high or more. A few large glades in the forest are clearly anthropogenic, and are used as seasonal pasture by the Samburu pastoralists. Here the vegetation consists of a Dichanthium / Themeda grassland, bordered by dense shrub stands ( Rhamnus prinoides , Gnidia glauca , Vernonia spp. etc.). Patches of bamboo ( Arundinaria alpina ) are found above 2000 m, especially around the highest peaks of the mountain.

Samburu inhabitants of Mt. Nyiro reported the association of K. asheorum with tall trees within the montane forest; however, two specimens were collected on Dodonaea bushes, up to 2 m above the ground.

Conservation status. The forest zone of Mt. Nyiro is declared as a forest reserve and, as such, managed by local forestry authorities. This regulation gives the montane forest at least some level of protection. However, unauthorized timber wood collecting (mainly Juniperus procera ), frequent fires and extensive grazing by herds of Samburu pastoralists during the dry season, all together form a strong anthropogenous pressure leading to forest degradation. K. asheorum n. sp. is probably endemic to the relict forest at Mt. Nyiro, but its current distribution awaits extensive field studies. The range of this species is probably limited to a rather restricted area of high canopy montane forest and high altitude bushlands, representing thus one of the smallest ranges if compared to other African continental chameleons.

Molecular phylogeny. Maximum-likelihood ( Fig. 6 ) and Bayesian (not shown) tree topologies are analogous. The only difference is related to the placement of two different sequences of K. tenuis as sister taxon in ML analysis. This branch was not revealed by Bayesian method. The bootstrap support of this branch is low in ML. When the sequence of K. tenuis ( EF014318 ) was excluded from the dataset, the statistical supports of most of the branches of the phylogenetic trees (both the bootstrap values in ML and the posterior probabilities in Bayesian analysis) significantly increased. K. asheorum is placed as a sister taxon to the clade comprising K. excubitor , K. adolfifriderici , K. xenorhina and K. carpenteri in the ND2 gene tree.

Figures

FIGURE 2. Head of the holotype of Kinyongia asheorum n. sp. (male) dorsally.

FIGURE 3. Head of the paratype 1 of Kinyongia asheorum n. sp. (female) laterally.

FIGURE 4. Holotype of Kinyongia asheorum n. sp. (male), in its natural habitat at Korante, Nyiro Range, Kenya.

FIGURE 5. Paratype 1 of Kinyongia asheorum n. sp. (female) photographed in its natural habitat immediately after its discovery.

FIGURE 6. Phylogenetic tree of the genus Kinyongia as inferred from ND2 nucleotide sequences using Maximum likelihood method. Numbers above branches are bootstrap percentage; numbers under branches indicate bayesian posterior probabilities. The newly sampled specimens have the corresponding museum collection number attached to their names. The NCBI accession numbers of the sequences are displayed under the species names. The accession numbers in the brackets represent the sequences that are identical to the present sequences and were therefore excluded from the dataset. The presence of K. tenuis* did not affect the ML topology, but lowered the statistical support of many of the branches. It was therefore excluded from the dataset before the bootstrap values and posterior probabilities were calculated.