Perini, Casali and Flynn 2020
try to resurrect an invalid clade in the LRT (subset Fig. 1), ‘Notoungulata‘. Either that or they never thought to test it for validity.
From their abstract:
“During most of the Cenozoic, South America (SA) was an isolated continent, leading to evolution of an unique and endemic fauna. Among these were the Notoungulata, an extinct group of ungulate-like mammals that included a wide array of species assigned to 14 families and more than 150 genera, occupying many distinct herbivorous niches and showing convergent morphologies with distantly related mammalian herbivore groups.”
That ‘wide array of species’ is a clue that this traditional clade is not monophyletic. The large reptile tree (LRT, 1751+ taxa, see Fig. 1) nests various traditional notogunulates in various disparate clades, some placental, others marsupial. This makes the Notoungulata invalid (in 2016, Fig. 1) due to polyphyly.
“Despite this diversity, few studies have addressed the evolution of morphological disparity among notoungulates.”
In counterpoint, the LRT addressed the evolution of morphological disparity in this traditional clade, and found its diversity spelled its doom.
“In this study, we applied modern comparative methods to investigate macroevolutionary patterns within the clade.”
...and that’s there main mistake. They assumed they had a monophyletic clade. Don’t assume! Test by adding taxa. Don’t cherry pick taxa, especially if those taxa are on traditional lists. Add taxa and add more taxa until all the taxa in your study nests in monophyletic clades, as shown online in the LRT. When this is done the traditional members of the Notoungulata fail to remain monophyletic (Fig. 1).
“We used a comprehensive morphological matrix to perform a Bayesian phylogenetic analysis, obtaining a fully resolved phylogeny and divergence time estimates for the major groups of notoungulates.”
Cherry-picking taxa makes this possible. Cherry-picking = cheating or laziness or reliance on others, none of which makes good science.
Perini, Casali and Flynn 2020 continue:
“This phylogeny supports the division of Notoungulata into Toxodontia and Typotheria, including many traditionally recognized families, as well as indicates paraphyly of some groups such as “Notohippidae”, “Notopithecidae”, and “Isotemnidae”. The diversity of Notoungulata rose steadily from the beginning of the Paleocene, reaching its apex by the end of the Oligocene, with marked decreases in the beginning, middle, and end of the Miocene, until final extinction in the Pleistocene. Toxodontia and Typotheria show no overlap, but occupy increasingly more distinct areas of the morphospace over time. Etc.”
Some traditional notoungulates are marsupials. Others are placentals in the LRT. You can weed out convergence by adding taxa. I’ll keep saying that until it becomes rote and the standard operating principle.
Nothing should proceed in detail
until a valid phylogenetic context is present. Otherwise you are wasting your time. Once you have your wide gamut cladogram, the kind that minimizes taxon exclusion, like the LRT, you will have a tool that you can use over and over again with little additional effort. It’s like a Periodic Table of Elements you can refer to. If you never have such a tool you will always risk taxon exclusion and the specter of convergence.
Perini FA, Casali DD and Flynn J 2020. Notoungulata, an endemic radiation of extinct South American herbivorous mammals. SVP abstracts 2020.