Controlling for the species-area effect supports constrained long-term Mesozoic terrestrial vertebrate diversification

Abstract

Variation in the geographic spread of fossil localities strongly biases inferences about the evolution of biodiversity, due to the ubiquitous scaling of species richness with area. This obscures answers to key questions, such as how tetrapods attained their tremendous extant diversity. Here, we address this problem by applying sampling standardization methods to spatial regions of equal size, within a global Mesozoic-early Palaeogene data set of non-flying terrestrial tetrapods. We recover no significant increase in species richness between the Late Triassic and the Cretaceous/Palaeogene (K/Pg) boundary, strongly supporting bounded diversification in Mesozoic tetrapods. An abrupt tripling of richness in the earliest Palaeogene suggests that this diversity equilibrium was reset following the K/Pg extinction. Spatial heterogeneity in sampling is among the most important biases of fossil data, but has often been overlooked. Our results indicate that controlling for variance in geographic spread in the fossil record significantly impacts inferred patterns of diversity through time.

The apparent exponential radiation of Phanerozoic land vertebrates is an artefact of spatial sampling biases

“…Understanding how terrestrial diversity was assembled through deep time is crucial for settling fundamental debates about the diversification process, such as whether it is constrained by ecological limits [3,4]. However, there is no consensus about the long-term trajectory of terrestrial diversity-in particular, whether or not exponential increases occured through the Phanerozoic, leading to diversity being higher today at local, regional and global scales than at any point in the geological past [3,[5][6][7][8][9][10][11].…”

“…Most problematically of all, 'global' palaeodiversity curves based on the worldwide fossil record are not truly global, because the spatial extent of the fossil record varies substantially among intervals of geological time [10,11]. In reality, the 'global' fossil record comprises a heterogeneous set of regional assemblages, with palaeogeographical regions that vary markedly in number, identity and extent (both within and between continental regions) through the intervals of geological time.…”

“…Patterns inconsistent with expansionism are recovered by analyses applying rigorous sampling standardization to estimate regional diversity of more restricted groups of tetrapods [6,[21][22][23] or over shorter intervals of time (the Mesozoic-early Palaeogene; [10,11]). Analyses of Phanerozoic tetrapod diversity at the local community scale [24] also contradict the expansionist model of diversification.…”

Improving global paleogeography since the late Paleozoic using paleobiology

“…The PBDB is a widely used resource (e.g., Wright et al, 2013;Finnegan et al, 2015;Heim et al, 2015;Mannion et al, 2015;Nicolson et al, 2015;Fischer et al, 2016;Tennant et al, 2016;Close et al, 2017;Zaffos et al, 2017), yet, the spatial coverage of data is still highly heterogeneous, with relatively few data points across large areas of the globe for some time periods. Hence, it is important to combine it with other geological data, such as stratigraphic data from StratDB Database (http://sil.usask.ca) and Macrostrat Database (https://macrostrat.org/) and other sources of paleoenvironment and paleo-lithofacies data, to further constrain the paleogeographic reconstructions.…”

Biotic and abiotic factors driving the diversification dynamics of Crocodylia

“…Sepkoski ; Holland ; Close et al . ). The Bayesian method here employed (PyRate) explicitly models the preservation process (comprising fossilization, sampling and taxonomic identification; Silvestro & Schnitzler ) and several tests performed on a range of simulated datasets reveal their accuracy and robustness against various degrees of data incompleteness (Silvestro et al .…”