Pablo J. Venegas; Omar Torres-Carvajal; Vilma Duran; Kevin de Queiroz

“ Seven of the eight individuals of Enyalioides azulae reported in this paper were collected at night sleeping on low vertical stems of bushes 15–80 cm above the ground. One adult male (the holotype) was collected during the day on a narrow trail after a rain; when approached, it fled and hid under a fallen log. This species is found in sympatry and possibly syntopy with Enyalioides binzayedi sp. n. (see below) and Enyalioides laticeps. The smallest individuals (CORBIDI 08790–08791, SVL = 61 and 62 mm, respectively) were collected in January. ”

“ Enyalioides azulae, Enyalioides cofanorum and Enyalioides microlepis differ further from Enyalioides binzayedi (character states in parentheses) in having more than 33 dorsal scales in a transverse line between the dorsolateral crests at midbody (31 or fewer) , a low vertebral crest (high, with vertebrals on neck more than four times the size of vertebrals between hind limbs in both sexes) , and a black gular patch (absent) . ”

“ Meristic and morphometric characters of Enyalioides binzayedi are summarized in Table 1. The holotype is the only adult male specimen available; it differs from female and subadult male paratypes in having projecting scales on each side of the vertebral crest on the neck. Additionally, female paratypes CORBIDI 08789 and 09216 are unique in having a double vertebral crest from midbody to pelvic region. ”

“ Additionally, Enyalioides azulae has a marked sexual dichromatism, with males having greenish and females brownish background coloration (Fig. 2) , whereas the other two species have brownish background coloration in both sexes. Enyalioides azulae further differs from Enyalioides cofanorum in lacking scattered enlarged scales on the dorsum, well-developed dorsolateral crests between the hind limbs, and a dark gular patch in females.
Description of holotype.
Male (Fig. 1) ; SVL = 96 mm; TL = 140 mm; maximum head width = 21.28 mm; head length = 26.35 mm; head height = 17.95 mm ”

“ To reduce the chance of converging on a local optimum, four runs were performed. Each consisted of five million generations and four Markov chains with default heating values. Trees were sampled every 1000 generations resulting in 5000 saved trees per analysis. ”