Bioacoustics of Microhyla malcolmi

Natural history notes. Microhyla malcolmi inhabits various habitats, including primary and secondary forests as well as human-modified rural landscapes in lowland and hilly areas at elevations up to 2200 m a. s. l. In the lowland monsoon forests of southern Vietnam , this species appears to be associated with bamboo tangles ( Vassilieva et al. 2016 ) and is also often recorded along forest streams. Microhyla malcolmi is active both day and night; according to Vassilieva et al. (2016) , its diet consists mainly of ants. Breeding was observed in Vietnam and Cambodia and usually had two peaks of reproductive activity: in May-June and October-December ( Neang and Holden 2008 ; Vassilieva et al. 2016 ). The noisy rasping call of breeding males can be heard at night from beside forest pools ( Neang and Holden 2008 ). Breeding mainly takes place in slowly-flowing forest streams, ponds, rain pools, puddles and other similar still waterbodies; the reproductive activity usually starts after heavy rains. Clutches include hundreds of small ( 1.6–1.7 mm ) pigmented floating eggs. In Thailand , the species prefers waterbodies with dead leaves and branches on silty bottoms for breeding ( Meewattana 2022 ).

Advertisement call. The male advertisement call of M. malcolmi represents a series of noisy rasping sounds that resemble the sound of a ratchet to the human ear. Heyer (1971) provided a brief description of an advertisement call for a M. malcolmi population from Sakaerat Experimental Station, Nakhon Ratchasima Prov . , Thailand (referred to as M. berdmorei in his work). The duration of the call was 0.09– 0.26 s , with each call consisting of 3–9 pulses at a pulse rate of 33–35 pulses per second, and the peak frequency of the call estimated at 1500–1800 Hz. Our recording of M. malcolmi advertisement calls taken in Song Hinh Forest Reserve, Phu Yen Province , Vietnam , included call series, each of them consisting of several calls (4–6; average 5.2 ± 1.4) emitted at a 2.3 ± 1.5 s (0.97– 3.79 s ) interval; a single call lasted for 0.22 ± 0.04 s (0.16–0.31) on average. Each call consisted of 9 ± 1.1 (8–10) pulses. The call’s amplitude increased sharply, reached its peak in the first 0.05– 0.10 s of call duration, and then gradually decreased towards the call’s end; the last half of the call duration had a notably lower amplitude (491 ± 57 Hz) than the first half (Fig. 10 B ). We estimated the call peak frequency to be 1631 ± 93 Hz (1507–1809 Hz), which agrees well with the estimates of Heyer (1971) .