Comprehensive DescriptionRead full entry
Diagnosis: Larvae are reddish or pinkish with dense black marksor reticulations on the dorsal surface and an unmarked venter. Well-developed gills with each middle arch having about 20 gill rakers. Pterygoid teeth 10-10. Vomerine teeth about 30-30. Maxillary-premaxillary teeth about 30-30. Splenial teeth about 50. Webbing involves only the metatarsals and metacarpals (Taylor 1941).
Description: A robust, medium-sized ambystomatid. Terrestrial adults reach a maximum SVL of 76.6 mm in males and 84.8 mm in females, a maximum tail length of 75.3 mm in males and 71.6 mm in females, and a maximum total length of 164.4 mm in males vs. 163.5 mm in females. Males have longer tails and shorter bodies than females. Costal grooves number 11-12 (average 11.0). Vomerine teeth usually in two series (range 42-70, mean 50.4) with a midline diastema. Glandular ridge on proximal third to half of the tail. Parotoid gland is present. Neither the ridge nor the parotoid gland are as well-defined as in A. gracile (Anderson 1961; Anderson 1978).
Hatchlings are dark-brown to black, while older larvae are patterned with black and yellow (Anderson and Webb 1978; Webb and Baker 1984). The typical larval coloration consists of black and yellow mottling arranged in longitudinal rows (Taylor 1941; Shannon 1951), although specimens from Yépomera can have more of a silver-gray to brown background color (Shaffer 1983). Metamorphs are uniform olive-brown, gradually changing to brownish-black with light yellow spots at maturity in Chihuahua specimens (Van Devender 1973) or being relatively uniformly colored with only occasional spots in Durango specimens (Anderson 1961). Tanner (1989) also reported that larvae transited from the typical mottled larval pattern to a darker, more uniform color, before the yellow-spotted adult coloration began to emerge. Terrestrial adults generally have large distinct yellow spots on a dark background that may coalesce into veins or streaks; some individuals have little to no spotting (Raffaëlli 2007). Paedomorphs are blackish with paler marks (Raffaëlli 2007).
Similar species: Larvae of this species can be distinguished from sympatric Ambystoma tigrinum stebbinsi, the Sonora tiger salamander, by coloration and gill raker count (USFWS 2002); A. rosaceum larvae are pinkish in color and have dark patterning on the dorsum and sides (Taylor 1941), plus fewer gill rakers (9-15) than tiger salamanders in Arizona and Mexico (15-24) (Collins 1979). A. tigrinum stebbinsi larvae are gray dorsally and lack dark patterning (USFWS 2002).
It may be a species complex, with northern and southern populations representing distinct lineages (Shaffer 1983). Four subspecies were recognized by Shannon (1951), mainly on the basis of differing larval coloration in various populations, but Anderson (1961) did not find these subspecies to be valid. However, Shaffer (1983) subsequently recognized two subspecies (A. r. rosaceum, in the northern part of the range, and A. r. nigrum, in the south) on the basis of allozyme differences. Highton (2000) thought that the data in Shaffer and McKnight (1996) indicated A. rosaceum might actually be three species.
Although two papers (Van Devender 1973 and Van Devender and Lowe 1977) reported that A. rosaceum and A. tigrinum were sympatric in streams and springs of Yépomera, Chihuahua, Tanner (1989) pointed out that the larval specimens were always identified as A. rosaceum and the adults were always identified as A. tigrinum, and concluded from patterning that all of these specimens were in fact A. rosaceum (supported by Shaffer's (1983) allozyme analysis of other Yépomera specimens). Anderson (1961) thought that the two species described in Taylor's (1941) paper were both likely to be A. rosaceum, with the second species simply in the process of metamorphosis.
Although Taylor (1941) is regarded as the species authority, Webb and Baker (1984) noted that this species was actually first described by Owen (1844) as Axolotes maculata. According to Webb and Baker (1984), Bishop (1945) shared this view and noted that Owen's name appeared to be a senior synonym of Taylor's A. rosaceum, but that the name maculata was already in use for Ambystoma maculata (then referred to as Lacerta maculata). Webb and Baker (1984) note that Owen's type locality was "in fluviis Sierrae Madre, Chihuahua, lat. 26°6'N, long. 106°50'W", or about 13 km east of Guadalupe y Calvo; they reported finding A. rosaceum larvae in most permanent streams in the area, lending credence to Owen's report.