Variation in timing of ossification affects inferred heterochrony of cranial bones in Lissamphibia.

ABSTRACT

The evolutionary origin of Lissamphibia likely involved heterochrony, as demonstrated by the biphasic lifestyles of most extant orders, differences between Anura (with tadpole-to-froglet metamorphosis) and Urodela (which lack strongly defined metamorphosis), and the appearance of direct development among separate lineages of frogs. Patterns in the timing of appearance of skeletal elements (i.e., ossification sequence data) represent a possible source of information for understanding the origin of Lissamphibia, and with the advent of analytical methods to directly optimize these data onto known phylogenies, there has been a renewed interest in assessing the role of changes in these developmental events. However, little attention has been given to the potential impact of variation in ossification sequence data--this is particularly surprising given that different criteria for collecting these data have been employed. Herein, new and previously published ossification data are compiled and all pairs of data for same-species comparisons are selected. Analyses are run to assess the impact of using data that were collected by different methodologies (1) wild- versus lab-raised animals; (2) different criteria for recognizing timing of ossification; and (3) randomly selecting ossification sequences for species from which multiple studies have been published, but for which the data were collected by different criteria. Parsimov-based genetic inference is utilized to map ossification sequence data onto an existing phylogeny to reconstruct ancestral sequences of ossification and infer instances of heterochrony. All analyses succeeded in optimizing sequence data on internal nodes and instances of heterochrony were identified. However, among all analyses little congruence was found in reconstructed ancestral sequences or among inferred instances of heterochrony. These results indicate a high degree of variation in timing of ossification, and suggest a cautionary note about use of these data, particularly given that in most instances issues associated with the original sources of data (e.g., wild- vs. lab-raised animals; or criteria for identification of earliest ossification) are not addressed. Potential sources of variation in the original data are discussed and may explain the incongruence observed here.