Bioacoustics of Spinomantis microtis

Habits. According to the original notes by Ch . P. Blanc, the holotype MNHN 1080 and one paratype MNHN 1081 were found within vegetation next to a stream, while another paratype MNHN 1082 was found within moss percolated by water. A. P. Raselimanana (pers. comm.) observed the species at Andohahela, between 990 and 1500 m of elevation (apparently most abundant at about 1100 m ). Individuals were observed within and along some streams with large boulders. They preferred boulders along small waterfalls 0.50 to 3.0 m in height. Frogs usually adhered to stones, by flattening their body and with spread legs. These habits were confirmed by, P. Bora, F. Glaw and M. Vences, who found the species at Andohahela, at about 1600–1700 m of altitude. So far, no acoustic behaviour is known. Sexual dimorphism. Adult males smaller than females. The body size in males ranged 34.30–55.80 mm (mean ± SD = 41.30 ± 6.03 mm , n = 17), and 37.20–48.40 mm in females (44.40 ± 5.25 mm , n = 10). There is no evident sexual difference in overall coloration. The only apparent dimorphic feature between some males and females is the morphology of the cloacal opening. In likely immature or reproductively inactive individuals, the cloaca is a simple terminal slit, as in most other Mantidactylus species. In adult females (most likely in reproductive season), the cloaca is tubular and curved downwards. In adult males the cloaca is similarly directed downwards, while the opening is overhung by a skin flap, sometimes bilobed. Considering the habitat where the species lives (fast streams), we suspect that the cloacal morphology is functional in regards to the egg laying in running waters. During egg deposition the male and female cloacae are likely juxtaposed, thus forming a nearly closed structure which allows for more efficient fertilization. A somehow similar structure was observed for Boophis albilabris and B. occidentalis , two species that like M. microtis frequent fast running streams ( Cadle 1995 , Andreone et al. 2002 ). Although Guibé (1974) and Blommers­Schlösser & Blanc (1991) indicated the presence of a vocal sac we were unable to detect it in the analysed M. microtis specimens. We suspect that their note was based upon the analysis of a series housed in Paris (MNHN 1975­2636 / 2649 and 2651 /2655). These animals, coming from “Andohahelo” (a misspelling for Andohahela), and collected by J. Arnoult (in January 1954 ), are badly conserved and were likely not well fixed. For this the skin of these specimens is loose, and especially the throat is so relaxed that may have given the aspect of a vocal sac. In specimens more recently collected (e.g., those from UMMZ and ZMS series) there is no evidence of this structure. Similar species. Mantidactylus microtis differs from M. microtympanum (which has a somehow similar colouration) in having a smaller size and more slender habitus, and for the skin texture, which is more granular, and the belly colour, which is more uniformly whitish or greyish, while in M. microtympanum there are several spots on the throat. The head of M. microtis is more flattened and more roundish seen from above than in M. microtympanum , in which the snout is pointed. Furthermore, the nostrils are more relevant when seen from the profile. Mantidactylus microtis is also similar in overall habits and morphology to M. lugubris , which is smaller at adult size (around 30–45 mm ), has a blackish or black and greenish banded colouration (mostly greyish with darker spots and dots in M. microtis ), relevant femoral glands (absent in M. microtis ), and a pointed snout. Furthermore, Mantidactylus lugubris emits trilling vocalisations, while M. microtis is not yet known to vocalise.