Hyperolius acuticeps Ahl, 1931

Lötters, S., Schick, S., Scheelke, K., Teege, P., Kosuch, J., Rotich, D. & Veith, M., 2004, Bio-sketches and partitioning of sympatric reed frogs, genus Hyperolius (Amphibia; Hyperoliidae), in two humid tropical African forest regions, Journal of Natural History 38 (15), pp. 1969-1997 : 1983-1985

Hyperolius acuticeps Ahl

(figure 6)

We found small, sharp-nosed reed frogs belonging to the species complex H. nasutus Günther. The latter undoubtedly comprises several sibling taxa, scattered through Africa (cf. Schiøtz, 1999). Channing et al. (2002), based on bioacoustics, were able to distinguish three largely sympatric species, found in both savannah and forest. The authors did not consider Kenyan material. Comparing their call data with ours suggests that our population from western Kenya is referable to H. acuticeps (as is another population found in the central Kenyan highlands). We examined the types of H. acuticeps , which coincide with Kenyan specimens except that they are somewhat smaller (lectotype 17.2 mm and paralectotype 19.1 mm). However, external characters do not seem to be useful in distinguishing taxa of the H. nasutus complex (cf. Schiøtz, 1999; Channing et al., 2002).

Moreover, Channing et al. (2002) discussed available names referable to the H. nasutus complex. For six of them they were unable to determine the taxonomic status. These include H. granulatus (Boulenger) , H. oxyrhynchus (Boulenger) and H. sagitta Laurent , from the Democratic Republic of Congo, and H. papyri Werner from Uganda. They all originate from sites in relative proximity to western Kenya, and consequently we suggest detailed analyses before applying with certainty the name H. acuticeps to Kenyan material (the type locality of H. acuticeps is in Tanzania). We can state at least that the male holotype of H. sagitta ( SVL 21.5 mm) compares closely with our Kenyan material.

Diagnosis. (1) SVL of two males 20.5 and 20.9 mm, SVL of females unknown (non-collected females did not differ in size as typical for H. nasutus sensu lato; cf. Schiøtz, 1999); (2) TIBL / SVL 0.40–0.52 (n ~2), HW/ SVL 0.24– 0.26 (n ~2); (3) dorsal surface smooth; (4) snout shape dorsally sub-elliptical to pointed and laterally acute, nares visible from above; (5) E-N/EYE 0.85–0.96 (n ~2), canthus rostralis convex from tip of snout to nostril and straight from nostril to eye; (6) tympanum covered by thick skin; (7) FOOT / TIBL 0.74–0.80 (n ~2); (8) foot webbing formula: 1(1), 2i(1) 2e(K), 3i(1–1K) 3e(K), 4i/e(1), 5(K–1); (9) PhJ and PhF are dorsally and ventrally translucent green, dorsal surface occasionally with irregular tiny reddish brown spots and/or white dorsolateral line (and sometimes up to three faint dorsal lines), mid-belly white, toe and finger tips yellowish; gular flap of males greenish yellow and white; the iris is golden (information provided is based on collected and several noncollected specimens of both sexes); (10) for sequence of 560 bp fragment of mitochondrial DNA of the 16S ribosomal gene see GenBank under AY323926 (~ NMK A/3922/1); (11) LTRF 1/3.

Hyperolius nasutus sensu stricto and H. viridis Schiøtz, 1975 seem indistinguishable from H. acuticeps (or material here assigned to this species) on the basis of external features ( Channing et al., 2002).

Life history. The nocturnal, arboreal H. acuticeps was found only between May and June in the Kakamega Forest. Both males and females were observed by night on vegetation 0.3–2 m above ground in swampy areas, and at ponds in disturbed primary forest. Males emitted calls frequently and pairs in amplexus were observed. As typical for the H. nasutus group (e.g. Schiøtz, 1999), egg deposition was in water; several clutches were attached to submerged vegetation with relatively little jelly. The ca 180 eggs obtained from one female were halfpigmented. Hyperolius acuticeps seems to be intermediate between explosive and prolonged breeding strategies. Larvae are lentic and omnivorous.

The advertisement call is a relatively long, high-pitched ‘scream’, actually a composition of several pulsed notes, the first of which is always longer and exhibits more pulses than the following ones. Table 1 lists numerical vocalization parameters. Channing et al. (2002) described the same advertisement call from elsewhere in Africa (‘call type A’). However, the authors refer to one call consisting of several pulses (instead of several notes) that slow down during the last half of the call.