www.mampam.comFrogs of Coorg





Study Area








All fieldwork was carried out in August/September 1998. Pilot surveys identified six
study sites with a varying sizes of amphibian assemblage and heterogeneity of
microhabitats; each site was mapped to scale using squared paper and a 30m measuring
tape. Microhabitat patches were described by randomly placing 0.5m2 quadrats, the
number of which was proportional to the patch size. This determined the % vegetation
cover; vegetation type (shrub, grass, dead wood, leaf litter) and average vegetation height
(cm) of microhabitats.

The attributes of each pond were calculated using a number of methods. Approximate
pond dimensions were recorded using a 30m tape measure. The average depths of the
larger ponds were obtained from estate records. The depths of smaller ponds were
measured using a marked 2m rod at several points in the water body. The volume of each
pond was calculated by multiplying the area by the average depth. Pond pH was
measured using a hand held pH meter. Readings were taken at different locations in the
water and an average was taken. Whether or not a pond was temporary or permanent and
whether it was natural or man-made was deduced from estate records. The ratio of
arboreal to terrestrial microhabitat was calculated by dividing the total area of arboreal
habitat by the total area of terrestrial habitat in the same study area. Vegetation above 1m
was regarded as arboreal for this calculation.

Quadrat sampling
The amphibian community was sampled with 1m2 quadrats. For terrestrial microhabitats
quadrats were made from rigid plastic sheeting. This had 0.6m high walls, which
prevented the frogs from escaping before they were counted, and was supported by a
vertical wooden strut at each corner. The ‘walled’ quadrat could not be effectively placed
in marsh, brash and hedge microhabitats; alternatively a straight stick, 1 metre in length,
was used to determine the sample area within these microhabitats. Marsh, brash and
hedge were open areas in which frogs were easily visible and could be counted and identified without handling, this reduced the probability of frogs escaping before they were counted.
Fieldwork took place between 8.30pm and 11pm when amphibian species are most
active. Sampling protocol required 3 field personnel. Two people haphazardly placed
the quadrat or metre stick; they captured the frogs within the quadrat by hand and
identified species using recently published keys, Daniels 1997. Identified frogs were
placed outside of that quadrat, on the side opposite the direction of observer movement to
avoid recapture. This was the only way in which frogs were handled; individuals were
not removed from the study site. A third person carried a clipboard and recorded the
biological and environmental data: number of individuals of each species; number
calling; number of pairs in amplexus (copulatory position); substrate moisture
(dry/damp/wet/water-logged); rain (nil/light/moderate/heavy); cloud cover
(clear/scant/moderate/complete); temperature (C, read 3 times, at intervals in the
sampling period) and moon (% visible). These data were recorded separately for each

To minimise disturbance to unsampled patches, sampling started at one end of the site
and continued towards the opposite end, a distance of 70m. Fieldwork began at alternate
ends each night to avoid bias caused by surveying at different times of night. A standard
number of quadrats were placed in each patch every sampling night, in proportion to the
patch area. Forty-three 1m2 quadrats were placed each night for a total of 17 nights at
Pond 1, and 5 nights each at Ponds 2-6.

Transect Sampling.
In order to determine the spatial distribution of aquatic species at Pond 1, a 70m long, 1m
wide transect line with distance markers every 2.5m was placed along the water edge.
The transect was walked once before and once after quadrat sampling. The position and
species of each individual was recorded.

Pitfall Traps
Eight pitfall lines were set at Luckunda Estate between the evening of 24 July and the morning of
29 August 1998. Positions are given in Map 2. Traps were 24 liter plastic buckets with depth of
35cm and top diameter of 35cm. Drift fencing was 70cm high corrugated plastic sheeting. Six
buckets were set per line, spaced 2m apart with a total of 8m of fencing, in the manner described by
Bennett (1999). The original intention was to check traps early morning, mid afternoon and late
evening but logistics and safety considerations (proximity of elephants) meant that most evening
checks were conducted before 1900hrs. A voucher specimen of each species caught was taken and
deposited at Zoological Survey of India, Calcutta. Otherwise, frogs caught were identified, marked
by toe clipping and released on paths close to the pitfall line. Reptiles caught in pitfall traps were
identified to species wherever possible. Other animals were identified in an arbitary manner.

Advertisment Calls
Of the 21 species of amphibian recorded at Lackunda, 18 were heard making
advertisment calls, which were recorded with a Sennheiser 240 microphone onto
cassette tape via a Sony WMD6C recorder. Air or water temperature at the calling
point were noted and the recordings deposited at the British Library of Wildlife
Sounds. For most species encountered this appears to be the first attempt at recording
advertisement calls. Calls of some species suggested the presence of additional,
cryptic species, similar to Microhyla rubra and Limonectes limnocharis. Because all
the species are easily recognised by ear, detailed analyses of the sounds have not been
performed. Rather key elements of calls are described, all of which can be heard on
the tape/CD accompanying this volume.
Microhyla ornata
Rhacophorus lateralis

Images and text Daniel Bennett and Katie Hampson 2000

Sounds British Library of Wildlife Sounds 1998