98 



Current Herpetol. 20(2) 2001 



interactions between coexisting species 

 (Inger and Greenberg, 1966; Stewart and 

 Sandison, 1972; Miller, 1978; Licht, 1986; 

 McAlpine and Dilworth, 1989; Werner 

 et al., 1995). In the mainislands of 

 Japan, food habits have been studied 

 for R. nigromaculata, R. porosa brevipoda, 

 R. rugosa, and Hyla japonica inhabiting 

 rice fields (Hirai and Matsui, 1999; 2000a, 

 b, 2001a, b). With respect to food habits 

 of Japanese populations of R. limnocharis, 

 however, a list of prey items recovered 

 from four individuals (Tomida, 1976) is the 

 only available information to the present. 



In this paper, we present diet composi- 

 tion of R. limnocharis inhabiting rice fields 

 of central Japan, and compare it with 

 published data for conspecific populations 

 from outside of Japan. 



Methods 



We collected frogs from rice paddy fields 

 in Nara City, Nara Prefecture, central 

 Japan (34°42'N, 135°51'E). Collections 

 were conducted at night between 2100 h 

 and 2300 h on 20 June, 14 July, and 6 

 August 1997. Immediately after we cap- 

 tured frogs, we extracted their stomach 

 contents by forced regurgitation with 

 forceps, and preserved the contents in lO'Vo 

 buffered formalin for later identification 

 and analysis. For each frog, we measured 

 snout-vent length (SVL), mouth width and 

 body mass, and examined the presence or 

 absence of male secondary sexual characters 

 (nuptial pads, vocal sacs, and an M-shaped 

 dark marking on the throat). Individuals 

 were divided into three sex-age classes: 

 juvenile, adult male, and adult female. 

 Adult males are easily differentiated from 

 the others in possessing the male secondary 

 sexual characters, but adult females and 

 juveniles, both lacking those characters, 

 are not distinguishable from each other 

 with certainty unless the gonads are 

 inspected by dissection. Nevertheless, 

 because females are known to reach sexual 



maturity at larger body size than males 

 (Maeda and Matsui, 1999), for we practi- 

 cal purposes, we regarded individuals 

 smaller than the smallest adult male 

 (actually 32.2 mm in SVL) as juveniles, 

 and larger individuals as adult females, 

 including sub-adults. In addition, we 

 clipped toes in unique combinations for 

 individual identification. After these 

 procedures, we released frogs where they 

 were captured. 



In the laboratory, we identified stomach 

 contents to the lowest practical taxonomic 

 level, and measured maximum length (L) 

 and width (W) of each prey item to the 

 nearest 0.1 mm using either a caliper or 

 a calibrated ocular micrometer fitted to 

 a dissecting microscope. For partially 

 digested prey items, we estimated lengths 

 by measuring width and then using prede- 

 termined length-width regressions from 

 intact prey (see Hirai and Matsui [2001c] 

 for more details). Volume (V) of each prey 

 item was calculated using the formula for 

 an ellipsoid: 



V=4/3jr(L/2)(W/2)2 



Results 



Diet composition 



Forty frogs (17 adult females, 15 adult 

 males, eight juveniles) were captured, of 

 which six (five adult females, one adult 

 male) had empty stomachs. The bulk of 

 the diet was represented by three inverte- 

 brate phyla (Arthropoda, Mollusca, and 

 Annelida), with Arthropoda including 

 four classes (Insecta, Araneae, Crustacea, 

 and Diplopoda), making up 94.8% in 

 number and 52.9% in volume of total 

 prey items (Table 1). Insecta contained 

 seven orders, and comprised 62.3% in 

 number and 31.0% in volume. Ants (For- 

 micidae) comprised the largest proportion 

 in number (26.9%), followed by spiders 

 (Araneae; 20.3%) and dipterans (18.4%). 

 In volume, however, earthworms (Oli- 

 gochaeta; 36.2%) predominated in the 



