Cyclicity and frequency of reproduction, fecundity variation, and correlations between body size and fecundity in the parthenogenetic blind snake Ramphotyphlops braminus were studied on the basis of captive observations of 49 individuals and dissection of 98 fresh dead specimens from the Ryukyu Archipelago. Ovulation commenced in mid May and oviposition occurred from mid June to mid July. At all adult snout-vent length's (SVL), the mean number of enlarged ovarian follicles was greater than the mean number of eggs laid. Some follicles may be retained in the ovaries or become atretic during ovulation. Clutch size inferred from the number of enlarged follicles therefore could be overestimated. Relative clutch mass (RCM) was somewhat greater than the mean RCM of henophidian and caenophidian species. In R. braminus, clutch size was significantly correlated with SVL, whereas RCM was not, as in most other snakes. This might suggest the presence of particular factors of selection operating on RCM of R. braminus as well as other snakes. Snake reproduction has been studied by many of general reproductive patterns in this group researchers, and several reviews have attemptof reptiles from ecological and evolutionary ed to analyze available data for the elucidation view points (see Seigel and Ford, 1987 for review). However, most original works on snake reproduction have dealt with henophidian and 1To whom all correspondences should be adcaenophidian species; very little attention has dressed. been paid to the reproductive biology of the 9 This content downloaded from 157.55.39.10 on Mon, 08 Aug 2016 04:54:49 UTC All use subject to http://about.jstor.org/terms M. KAMOSAWA AND H. OTA infraorder Scolecophidia (i.e., Anomalepididae, Leptotyphlopidae, and Typhlopidae), a group of snakes characterized by small body size, cylindrical body shape, and fossorial habits, despite the high species diversity of this group (Erasmus and Branch, 1983; Shine and Webb, 1990). The paucity of quantitative reproductive data for this large group of snakes has seriously limited discussion of the general ecological patterns in snakes (e.g., Seigel and Ford, 1987; Dunham et al., 1988). In this paper, we provide quantitative reproductive data for the northernmost populations of Ramphotyphlops braminus, a triploid, all-female, parthenogenetic typhlopid species widely distributed in tropical and subtropical regions around the world (McDowell, 1974; Ota et al., 1991). MATERIALS AND METHODS Forty-nine specimens were collected in the Ryukyu Archipelago, Japan (30 from Takarajima, 16 from Okinawajima, two from Sesokojima, and one from Kumejima Islands) between 2 May and 28 August 1993. These animals were housed individually in plastic cups (100 x 80 x 45 mm in top diameter x bottom diameter x height) filled with damp sphagnum. Ant eggs, larvae, and pupae, as well as adult termites, were provided as food. Considering arguments raised by Cuellar (1984) and Seigel et al. (1986) in taking clutch mass data, each cup was checked every 24 to 72 h, and most eggs were measured (to the nearest 0.1 mm with dial-calipers) and weighed [to the nearest 0.01 g with electronic balance (Shimazu EB-330H)] on their first appearance. However, because all eggs subjected to this procedure developed mould and died, four clutches of eggs obtained in the latter period of these observations were not handled for measurements. Oviposited females were measured for snout-vent length (SVL) and weighed, and were fed and maintained as above to the end of December to see if a second clutch was laid. Relative clutch mass was calculated for each clutch after Shine (1980) [clutch mass/body mass of oviposited female: referred as RCM1], and Seigel and Fitch (1984) [clutch mass/(clutch mass + body mass of oviposited female): referred as RCM2]. Mean egg mass [clutch mass/clutch size: referred as MEM] was also calculated for each individual. Ninety-eight individuals (27 from Haterumajima, 20 from Miyakojima, 13 from Iriomotejima, 12 from Yonagunijima, 16 from Okinawajima, four from Ishigakijima, three from Kohamajima, and one from Iheyajima, Kuroshima and Minami-Daitoujima) obtained in March (N = 59), April (N = 29), early May (N = 5), October (N = 2), and December (N = 3) (most freshly dead from sugar-cane cultivation), were also measured for SVLs while fresh, and were subsequently dissected for gonadal inspection. For each specimen, the number of enlarged ovarian follicles (Nussbaum, 1980; Ota et al., 1991) were counted, and presence or absence of oviductal eggs were investigated. Due to the insufficient sample size for each island, data for all samples were pooled for analyses under an a priori assumption for the absence of significant geographic variation in reproductive traits within the Ryukyus.