178 Short Notes

Sumida, M., Allison, A., Nishioka, M. (1998): Genetic relationships and phylogeny of Papua New Guinean Hylid frogs elucidated by allozyme analysis. Japan. J. Herpetol. 17: 164-174. Trenerry, M., Dennis, A., Werren, G. (1995). Frog calls of north-east Queensland. Audio Cassette produced by J.M. Hero. Townsville, Queensland, Australia. Tyler, M.J. (1968): Papuan hylid frogs of the genus Hyla. Zool. Verhand. 96: 1-203. Tyler, M.J., Davis, M. (1978): Species groups within the Australopapuan hylid frog genus Litoria Tschudi. Austr. J. Zool., Suppl. Ser. 63: 1-47. Van Kampen, P.N. (1923): Amphibia of the Indo-Australian Archipelago. Leiden. Zweifel, R.G., Tayler, M.J. (1982): Amphibia of New Guinea. In: Biogeography and Ecology of New Guinea, p. 759-801. Gressit, J.L., Ed., Monographiae Biologicae Vol. 42, The Hague, Junk Publishers.

Received: January 8, 2003. Accepted: April 30, 2003.

Diet of the large water snake Hydrodynastes gigas (Colubridae) from northeast Argentina

María Soledad López, Alejandro R. Giraudo

National Institute of Limnology, CONICET, José Maciá 1933, (3016) Santo Tomé, Santa Fe, Argentina

Hydrodynastes gigas (Dúmeril, Bibron and Dúmeril, 1854), is one of the largest snakes of South America reaching a maximum total length of 3 m (Bernarde and Moura-Leite, 1999). It is a semiaquatic species (Strüssmann and Sazima, 1990, 1993; Cadle and Greene, 1993; Giraudo, 2001) that inhabits several types of wetland in subtropical and tropical regions in Peru, central and south-eastern Brazil, eastern of Bolivia, Paraguay to northern and east Argentina (Abalos and Mischis, 1975; Cei, 1993; Williams and Scrocchi, 1994; Moura-Leite et al., 1996; Bernarde and Moura-Leite, 1999; Giraudo, 2001). The diet of H. gigas had been mentioned by few authors, and most of the data are imprecise, lacking quantitative and qualitative analyses (Serié, 1919; Amaral, 1977; Cei, 1993; Strüssmann and Sazima, 1993; Williams and Scrocchi, 1994; Achaval and Olmos, 1997). These authors showed that it is a general predator eating fishes, frogs, mammals, birds and carrion. H. gigas is one of the most frequently found snakes in the lowland of Paraná and Paraguay basin (Strüssmann and Sazima, 1993; López, 2000). Here we report the diet of H. gigas based on stomach contents and regurgitated items from several wild specimens from this region. We examined 40 specimens of Hydrodynastes gigas (see appendix 1) from field sampling and museum collections. Most of the field observations are from north-eastern Argentina (in the provinces of Formosa, Chaco, Corrientes and Santa Fe), with some from Paraguay (in the department of Ñeembucú). This region is a lowland area (50 m average altitude) with alluvial flood plains of the Paraná and Paraguay rivers. The vegetation is formed

© Koninklijke Brill NV, Leiden, 2004 Amphibia-Reptilia 25: 178-184 Also available online - www.brill.nl Short Notes 179 by a mosaic of physiognomies ranging from wet savannas and grasslands to subtropical dry forests, gallery forests, shrublands and a wide variety of wetland (rivers, streams, marshes, swamps) (Cabrera, 1976; Prado, 1993; Dinerstein et al., 1995). The annual precipitation varies from 1,000 to 1,500 mm (Prado, 1993). The snakes were obtained from the field through time-constrained search and road sampling. We moved through the habitats turning logs, inspecting retreats and watching for surface-active snakes, totaling 400 man- hours. Road sampling was performed driving a vehicle at 40-50 km/h along secondary roads, totaling 4,034 km (Campbell and Christman, 1982). Some road-killed specimens in fairly good condition were collected and preserved. We checked for stomach contents by making a midventral incision, and also obtained some items by inducing regurgitation in some specimens caught in the wild. Whenever possible, we recorded the precise place were the specimen was found and its snout-vent length (SVL, in mm). Prey size corresponds to total length (TL, in mm) and was measured with callipers. Intact or slightly digested prey were measured directly; in some cases, anatomic parts of digested prey were used as an inference of their total length by comparison with intact specimens. The amphibians were measured including their extended hindlimbs. The correlation analysis between log transformed (base 10) prey size (TL) and predator size (SVL) was done using Pearson correlation. The normality distribution of variables was tested by Kolmogorov-Smirnov method. Nineteen Hydrodynastes gigas (SVL of females x¯ = 1,498 mm, range = 2,360- 1,010 mm, n = 9; SVL of males x¯ = 1,240 mm, range = 1,585-415 mm, n = 10), 48% of the specimens examined, contained 32 prey items (table 1). Four types of vertebrate prey were recorded, amphibians (30%), colubrid snakes (26%), fish (22%) and mammals (22%) (table 1), without significant differences between their frequencies (χ 2 = 0.4783, df = 3, P>0.9). One snake contained fish eggs (table 1). We consider that small invertebrates found in one H. gigas (SVL = 1,010 mm) are secondary prey first ingested by one amphibian (Leptodactylus sp.) recorded in the same stomach. The total length of 15 measured prey ranged from 7 to 39% of the snout-vent length of the snakes that ingested them. As in most snakes, large individuals of H. gigas tend to ingest larger prey, as indicated by the positive correlation with slope significantly different from zero (r = 0.5616, P = 0.0294, F = 5.99, n = 15) between the prey size (TL) and snake size (SVL) (see Arnold, 1993 for revision). The prey size (total length) did not vary significantly among fish, amphibians and snakes (Kruskal-Wallis test, KW = 4.500, P = 0.1054). The mammals were not compared for lack of data (table 1). We could not analyze the prey weight, an important variable to consider in snake diet (Arnold, 1993), due to the advanced digestion degree of most prey items. We observed that H. gigas has significant sexual dimorphism in some scalation charac- ters (ventrals and subcaudals number), but not in size variables (total length, tail length, snout-vent lenght, head lenght) (table 2). Probably, this is related with a not significant dif- ference between prey size eaten by males and females (t = 1.751, df = 13, P = 0.1035), but our sample sizes for sexes were, possibly, insufficient to obtain robust conclusions on this question. Representatives of two different fish families were identified, one with nude skin (Synbranchidae) and another with bony plaques (Callichthyidae) (Britski et al., 1999). Two adults of H. gigas regurgitated four swamp eels (Synbranchus marmoratus): one of them regurgitated three items, the other regurgitated one item (fig. 1). The shape and behaviour of swamp eels are similar to those of aquatic snakes (fig. 1). Swamp eels breath