Taxic and morphological diversification during the early radiation of Clupeomorpha (Actinopterygii, Teleostei)

Evolutionary radiation is a problematic concept whose definition and classification have recently changed. Radiations can be defined as the pattern of abrupt increase in diversity of a lineage. It is relevant to evaluate the presence and interaction of different types of radiation in extant and fossil organisms to adequately delimitate the radiation types and to know the diversity in the context of Earth´s history. Here, we employed superorder Clupeomorpha in the Early-Late Cretaceous boundary as a study case to recognize the radiation types and their interactions, using both taxic and morphologic approaches. Clupeomorpha is a diverse and ancient teleostean superorder with a wide geographic and ecological distribution, that has been extensively studied. We proposed a model to calculate rates of origination to analyze the taxic diversification and employed geometric morphometrics to analyze the morphological diversification that occurred at the time boundary. The results suggest the absence of taxic radiation due to the constant increase in the taxa origination. However, the expansion of the phylomorphospace occupation and the disparity increase suggest the presence of a climatic-geographical-disparification and broad diversification-like, according to the current classifications. This exhibits the incompatibility of the current radiation classifications with the case study. 

Methods
Data are part of a geometric morphometric analysis of the body shape diversity of the clupeomorph fishes during the Cretaceous. We digitalized twice 15 landmarks and 35 semilandmarks on TPS software (Clupeomorpha_2_14ene22.tps; Clupeomorpha_3_14ene22.tps). The code used can be consulted on the Paper Supporting Information and next to these files.  

As more phylogenetic-related taxa tend to resemble morphologically each other and lose data independence, phylogeny was considered in the analysis. Clupeomorph phylogeny was obtained from the character matrix of Than-Marchese & Alvarado-Ortega (2022). A Bayesian inference analysis of 56 taxa and 71 characters was performed with 20 million generations, 2 runs, 4 strings, and 0.2 temperature. Bayesian analysis was performed with MrBayes software and the unrooted consensus tree was cut with Mesquite software. The phylogeny obtained was attached to this repository (Matriz_11ene22_Bayes_JAO_tax_fil.con.tre)