Morphological variation of hypaxial musculature in salamanders (Lissamphibia: Caudata)

Despite the acknowledged importance of the locomotory and respiratory functions associated with hypaxial musculature in salamanders, variation in gross morphology of this musculature has not been documented or evaluated within a phylogenetic or ecological context. In this study, we characterize and quantify the morphological variation of lateral hypaxial muscles using phylogenetically and ecologically diverse salamander species from eight families: Ambystomatidae (Ambystoma tigrinum), Amphiumidae (Amphiuma tridactylum), Cryptobranchidae (Cryptobranchus alleganiensis), Dicamptodontidae (Dicamptodon sp.), Plethodontidae (Gyrinophilus porphyriticus), Proteidae (Necturus maculosus), Salamandridae (Pachytriton sp.), and Sirenidae (Siren lacertina). For the lateral hypaxial musculature, we document 1) the presence or absence of muscle layers, 2) the muscle fiber angles of layers at mid‐trunk, and 3) the relative dorsoventral positions and cross‐sectional areas of muscle layers. Combinations of two, three, or four layers are observed. However, all species retain at least two layers with opposing fiber angles. The number of layers and the presence or absence of layers vary within species (Necturus maculosus and Siren lacertina), within genera (e.g., Triturus), and within families. No phylogenetic pattern in the number of layers can be detected with a family‐level phylogeny. Fiber angle variation of hypaxial muscles is considerable: fiber angles of the M. obliquus externus range from 20–80°; M. obliquus internus, 14–34°; M. transversus abdominis, 58–80° (acute angles measured relative to the horizontal septum). Hypaxial musculature comprises 17–37% of total trunk cross‐sectional area. Aquatic salamanders show relatively larger total cross‐sectional hypaxial area than salamanders that are primarily terrestrial. J. Morphol. 241:153–164, 1999. © 1999 Wiley‐Liss, Inc.