The multicausal twilight of\xa0South American native mammalian predators (Metatheria, Sparassodonta)

Sergio Daniel Tarquini,Sandrine Ladevèze,Francisco Juan Prevosti

doi:10.1038/s41598-022-05266-z

Abstract

Sparassodonts were the apex mammalian predators of South America throughout most of the Cenozoic, diversifying into a wide array of niches including fox-like and even saber-toothed forms. Their extinction is still controversial, with different authors suggesting competition with other predators (placental carnivorans, terror birds, and carnivorous opossums), extinction of prey, and climate change as causal explanations. Here, we analyse these hypotheses using a novel approach implicating Bayesian analyses. We find that speciation and extinction rates of sparassodonts can be correlated with (i) intrinsic biotic factors such as changes in body mass and diversity of sparassodonts, (ii) extrinsic biotic factors such as potential prey diversity, and iii) extrinsic abiotic factors like the atmospheric CO2, sea level, temperature, and uplift of the Andes. Thus, sparassodonts are a good example of a multilevel mixed model of evolution, where various factors drove the evolutionary history of this clade in a pluralistic way. There is no evidence for competition between Sparassodonta and others predators, and the effect of competition in the face of extinctions of fossil species should be tested and not assumed. Furthermore, we propose a novel approach for evaluating the fossil record when performing macroevolutionary analyses.

Highlights

From the beginning of the Cenozoic until present times, the fauna of South America has undergone considerable changes throughout all trophic levels[1,2,3,4]
We evaluated four datasets of the sparassodont fossil records: (1) a dataset of all occurrences evaluated at the species level dated with precise methods, (2) a dataset of all sparassodont specimens identified at the species level with precise dating, (3) a dataset of occurrences at the species level dated with the boundaries of the South American Age/Stage, and (4) a dataset of occurrences at the genus level with precise dating
The South American fossil record is not unique in this regard, there are marked hiatuses in fossil record mainly during Paleogene, and there are some exceptions, fossil sites are limited to high latitudes[6,7,8,55,56]

Summary

Introduction

From the beginning of the Cenozoic until present times, the fauna of South America has undergone considerable changes throughout all trophic levels[1,2,3,4]. López-Aguirre et al.[37] correlated the diversity of Sparassodonta with different variables and concluded, by means of multiple regression analysis, that the diversity of three specific mammal groups correlated with diversity of sparassodonts: Didelphimorphia (i.e., opossums and relatives), African migrants (i.e., caviomorph rodents and platyrrhine primates), and South American ungulates (i.e., astrapotheres, litopterns, notoungulates, pyrotheres, and xenungulates) All these analyses were based on the numerical count of sparassodont species rather than evolutionary rates. In recent years Silvestro et al.[45] have proposed new Bayesian methods to directly calculate speciation and extinction rates from the fossil record, considering biases in the preservation These rates can be correlated with different variables to test biological interaction hypotheses like competition (e.g., Red Queen model) and the effect of environmental changes (e.g., Court Jester model)[46]. This contribution aims to study the evolution of Sparassodonta using these new Bayesian approaches and to test the different proposed hypotheses about the causes of its extinction

Methods

Results

Conclusion