Comparative morphology of the gastrointestinal tract in relation to diet in from a locality in south

Indraneil Das

Animal Ecology Research Group, Department of Zoology, University of Oxford, South Parks Road, Oxford OXI 3PS, UK Present address: Centre for Herpetology,Madras Crocodile Bank Trust, Post Bag 4, Mamallapuram 603 104, Tamil Nadu, India

A close coupling between diet and dentition has been demonstrated for numerous taxa; the relationship between diet and morphology of the gastrointestinal tract but seldom. The association between gut dimensions and diet in the various post- metamorphic stages of the Anura appears never to have been examined, although studies have addressed the question in the larval stages (summarised by Noble, 1931; see also Altig and Kelly, 1974). The comparative morphology and ecology of the gastroinestinal tracts of eight species of anuran were studied in Chengai-MGR (formely Chengleput) District, Tamil Nadu State, south India, between March-August, 1989 and November-Decem- ber, 1990. The specific question asked was: do the gastroinestinal tracts of these frogs reflect diet? Eight species of frogs were found in sympatry at the study site. These were the Indian green (Rana hexadactyla), Jerdon's bullfrog (R. crassa), the skipping frog (R. cya- nophlyctis), the south Indian burrowing frog (Tomopterna rolandae), the marbled balloon frog ( systoma), the red narrow-mouthed frog ( rubra), the ornate nar- row-mouthed frog (M. ornata) and the common Indian treefrog (Polypedatesmaculatus). M. ornata has been excluded from the following analyses because of the small sample size. Frogs were caught using a variety of techniques, including dip-netting, by hand, or with baited hooks; they were brought to the laboratory generally around two hours of capture and sacrificed. The entire gastrointestinal tract, including the stomach and the large and small intestines, of each frog was dissected out and their contents removed. Measurements (to the nearest 0.1 mm) taken of the visceral anatomy were the greatest and least diameters 290 of the emptied and stretched stomachs and the lengths of the entire gastrointestinal tract, comprising the stomach, and the small and large intestines. The scaling of organs between species was expressed in terms of change as a function of snout-vent length (SVL). Stomach volumes (STVOL) were estimated using the formula for an oblate spheroid (4/3 Ka2 b, were a and b represent the major and minor axes, respectively). Several workers have calculated volumes of vertebrate stomachs using the formula for prolate spheroids (e.g., Huey et al., 1974; for lizards) or even spheres (e.g., Chivers and Hladik, 1980; for mammals), but it was thought that an oblate spheroid better describes the shape of stomachs. The terms "gastrointestinal tract length", "GL" and "gut" have been used interchangeably. Adjusted stomach volumes in the seven study species are show in table 1. The largest stomach was found in the folivorous Rana hexadactyla, in which the mean STVOL of 213.51 cm3 is more than 3.4 times greater than that of the species with the next most voluminous stomach (R. crassa, mean STVOL = 62.36 cm3). Within species, the rela- tionship between SVL and STVOL was strong (r = 0.828; log-transformed data), the stomach scaling isometrically, with the slopes (b) of the regressions not significantly different from 3 (t-tests, P>0.05) in those species in which the fit of regression is good enough to enable a valid conclusion to be drawn. To examine the relative size of the stomach among species, while adjusting for body size, STVOL was divided by the cube of SVL (STOL/SVL3); the results are also presented in table 1. A between-species comparison of STVOL indicates that the stomach of R. hexadactyla was significantly larger than those of its sympatrics (t-test, P<0.05); this is generally true even after adjusting for body size, except for comparisons with the - and -specialist, U. systoma (P>0.05). The largest mean ratio (6.93) was found in R. hexadactyla, also the largest member of the assemblage (SVL up to 132.2 mm), while the next highest figure was for the small U. systoma (SVL up to 54.7 mm), with a mean STVOL3 ratio of 3.76. Between species, relationship for log-mean SVL to log-mean STVOL/SVL3 of each species showed a positive (r = 0.702) and significant (P<0.05) correlation. The largest mean gastrointestinal tract length within the study species was found in R. hexadactyla (table 1). An interspecific comparison of this species versus its sympatrics (t- tests) revealed that the gastrointestinal tract is significantly longer at the 0.001 level, except U. systoma (significant at the 0.05 level), which appears to be equipped to handle relatively large amounts of a food type that is available only seasonally. When adjusted for body size, by dividing the gastrointestinal tract length by SVL (GL/SVL), gut length in the largely folivorous adults of R. hexadactyla was significantly (t-tests, P<0.05) longer than those of its sympatrics, including R. crassa, T rolandae and P. maculatus. The highest mean GL/SVL ratio (hence the longest gastrointestinal tract) was found in R. hexadactyla; the lowest (or the shortest gut) in P. maculatus. The relationship of the length of the gastrointestinal tract to that of the body (SVL) showed positive correlations