Considering decoupled phenotypic diversification between ontogenetic phases in macroevolution: An example using Triggerfishes (Balistidae)

Across the Tree of Life, most studies of phenotypic disparity and diversification have been restricted to adult organisms. However, many lineages have distinct ontogenetic phases that differ from their adult forms in morphology and ecology. Focusing disproportionately on the evolution of adult forms unnecessarily hinders our understanding of the pressures shaping evolution over time. Non-adult disparity patterns are particularly important to consider for coastal ray-finned fishes, which often have juvenile phases with distinct phenotypes. These juvenile forms are often associated with sheltered nursery environments, with phenotypic shifts between adults and juvenile stages that are readily apparent in locomotor morphology. Whether this ontogenetic variation in locomotor morphology reflects a decoupling of diversification dynamics between life stages remains unknown. Here we investigate the evolutionary dynamics of locomotor morphology between adult and juvenile triggerfishes. We integrate a time-calibrated phylogenetic framework with geometric morphometric approaches and measurement data of fin aspect ratio and incidence and reveal a mismatch between morphospace occupancy, the evolution of morphological disparity, and the tempo of trait evolution between life stages. Collectively, our results illuminate how the heterogeneity of morpho-functional adaptations can decouple the mode and tempo of morphological diversification between ontogenetic stages.

Methods

Collection of images for morphometric analyses: A total of 724 images of triggerfishes (juvenile and adult fishes) were taken from online repositories and digitized specimens from different museums (North Carolina Museum of Natural Science, Burke Museum, Bishop Museum, Philadelphia Academy of Science, LA County Museum, Yale Peabody Museum, Filed Museum and Museum of Comparative Zoology). 
Phylogenetic hypotheses were produced by using available DNA sequence data. The dataset included DNA sequence data from 11 genes (12S, 16S, RAG1, RAG2, rhod, Tmo-4c4, Bmp4, COI, CytB, Glyt, MYH6, RAG2).

Locations of specimens examined (images) and GenBank accession numbers are provided in Table S1 and Table S3 from the supplementary materials.