Caecilia annulata: Mikan, 1820: pl. II.

Siphonops annulatus: Goeldi, 1899: 170; Spengel, 1915: 220; Dunn, 1942: 480; Taylor, 1968: 555; Hoogmoed, 1979: 273; Frost, 1985: 631; Lescure & Marty, 2000: 297; Lavilla et al., 2004; Frost, 2008.

Diagnosis. Maximum known TL 450 mm. Annuli 78–98 (Taylor, 1973). Annular grooves completely encircling body, except for three or four near vent.

Description. TL 17–26 times BW (Taylor, 1968). Head slightly longer than wide, narrower than body. Snout projecting distinctly beyond mouth. Eyes visible in open socket slightly elevated above surface of head. Nuchal grooves distinct dorsally and ventrally except second collar partially fused below with first primary fold; dorsal transverse groove on each collar, shorter and less distinct on first. Ventral transverse groove on first collar present or absent. Body cylindrical, slightly wider than deep. Width along body may vary slightly. Primary annular grooves completely encircling body, except for anterior primary groove that is ventrally incomplete as well as some in region near vent (e.g. four in MPEG 9193). Large unsegmented terminal shield extending from anterior of vent to posterior tip of body. Opening of cloaca (vent) subcircular, T-shaped, or I-shaped (Taylor, 1968). AD 9–10; generally about as many on anterior as on posterior edge of vent (i.e. five anterior, four posterior in MPEG 9193). PMT monocuspid, maximally 43 (Taylor, 1968) with little variation in size, posterior maxillary teeth slightly smaller, extending posteriorly of choanae. Maximally 47 monocuspid PPT (Taylor, 1968) with no apparent variation in size; teeth smaller than PMT. DT monocuspid, at most 32 (Taylor, 1968), slightly larger than PMT. “Fetal” teeth in hatchlings in three rows on lower jaw (Wilkinson & Nussbaum, 1998; Wilkinson et al., 2008).

Color. Blackish or bluish with white grooves. Wilkinson et al. (2008) reported that females brooding young are often much paler than other adults.

Distribution. Cis-Andean South America (Argentina, Bolivia, Brazil, Colombia, Ecuador, French Guiana, Guyana, Paraguay, Peru, Suriname, Venezuela) (Fig. 18), except in Uruguay (Dunn, 1942; Taylor, 1968; Lynch, 1999; Lavilla et al., 2004; Frost, 2008).

Natural history. Goeldi (1899) mentioned a female with a clutch of eggs that was found under a tree trunk moved during ploughing of a pasture, and thus was the first to report oviparity in Siphonops annulatus. Taylor (1968) linked the wide distribution of S. annulatus to the fact that its moisture requirements are less than those of many other caecilians. Lynch (2006) also mentioned the tolerance of S. annulatus to dry conditions and reported finding specimens “under decomposing trunks of African oil palms, even in grazed pastures. Wilkinson et al. (2008) described maternal dermatophagy by young.

Remarks. The caecilian with the largest reported distribution (Wilkinson et al., 2008). Surprisingly, within its large distribution S. annulatus presents low variation, at least in external morphological characters (Dunn, 1942; Taylor, 1968; Lynch, 1999).

Although the species was reported from Suriname by Dunn (1942), Nussbaum & Hoogmoed (1979) stated that “Its occurrence in Surinam could not be confirmed by recent collections”. Despite continued collecting this situation has not changed since (Hoogmoed, in press).

Taylor (1973) discussed geographic variation in number of annuli, and noted that specimens from Ecuador, Peru and Bolivia had higher counts (90–98) than specimens from the eastern part of the distribution of the species (80–90). Lynch (1999), however, found a range of 81–97 annuli for specimens from Colombia, with lower and higher limits close to the known extremes of variation. Our limited data on three Brazilian specimens show a range of 84–94. Two specimens from Guajará-Mirim and Espigão D’Oeste, State of Rondônia, have respectively 84 and 94 annuli and the other one from Juruá, State of Amazonas, has 84 annuli. Thus, the difference between eastern and western populations outlined by Taylor (1973) is not real, and probably was due to lack of material.

In Brazil this species is recorded from Caatinga (Freitas & Silva, 2007), Atlantic forest (Taylor, 1968; M.S. Hoogmoed, pers. obs.), Cerrado, and from Amazonia (our data). Although it is a species considered common in scientific collections (Taylor, 1968) the number of specimens from Brazilian Amazonia we could locate is very low. We did not analyze specimens from outside Amazonia.