Functional morphology of the feeding mechanism in aquatic ambystomatid salamanders

This study addresses four questions in vertebrate functional morphology through a study of aquatic prey capture in ambystomatid salamanders: How does the feeding mechanism of aquatic salamanders function as a biomechanical system? How similar are the biomechanics of suction feeding in aquatic salamanders and ray-finned fishes? What quantitative relationship does information extracted from electromyograms of striated muscles bear to kinematic patterns and animal performance? and What are the major structural and functional patterns in the evolution of the lower vertebrate skull? During prey capture, larval ambystomatid salamanders display a kinematic pattern similar to that of other lower vertebrates, with peak gape occurring prior to both peak hyoid depression and peak cranial elevation. The depressor mandibulae, rectus cervicis, epaxialis, hypaxialis, and branchiohyoideus muscles are all active for 40-60 msec during the strike and overlap considerably in activity. The two divisions of the adductor mandibulae are active in a continuous burst for 110-130 msec, and the intermandibularis posterior and coracomandibularis are active in a double burst pattern. The antagonistic depressor mandibulae and adductor mandibulae internus become active within 0.2 msec of each other, but the two muscles show very different spike and amplitude patterns during their respective activity periods. Coefficients of variation for kinematic and most electromyographic recordings reach a minimum within a 10 msec time period, just after the mouth starts to open. Pressure within the buccal cavity during the strike reaches a minimum of -25 mmHg, and minimum pressure occurs synchronously with maximum gill bar adduction. The gill bars (bearing gill rakers that interlock with rakers of adjacent arches) clearly function as a resistance within the oral cavity and restrict posterior water influx during mouth opening, creating a unidirectional flow during feeding. Durations of electromyographic activity alone are poor predictors of kinematic patterns. Analyses of spike amplitude explain an additional fraction of the variance in jaw kinematics, whereas the product of spike number and amplitude is the best statistical predictor of kinematic response variables. Larval ambystomatid salamanders retain the two primitive biomechanical systems for opening and closing the mouth present in nontetrapod vertebrates: elevation of the head by the epaxialis and depression of the mandible by the hyoid apparatus.