Pleistocene extinction and geographic singularity explain differences in global felid ensemble structure

Abstract

Extant felids are hyper-carnivorous predators that originated in Asia c. 11 Mya and diversified in 8 distinct lineages, with 41 species surviving to the Recent. These species occupy almost every terrestrial habitat available in the four continental land masses they occupy and exhibit morphological and behavioral specializations to various locomotor styles and hunting modes. Today, distinct felid ensembles inhabit each continent and major biogeographic region. How the differential structuring of these ensembles was generated, and which evolutionary processes shaped these differences across ensembles, are key emerging questions. Using multivariate statistics, we analyzed a large dataset of 31 cranial and 92 postcranial linear variables describing shape and functional proxies of the entire skeleton of extant felids. We statistically demonstrate the existence of nine felid morphotypes at the global scale, whose occurrence is characteristic of different continental or biogeographic ensembles. Phylogenetically explicit analyses show that morphotypes from different felid lineages converged in different continents, but still ensembles remain distinct due to the fact that various morphotypes are missing in several of those ensembles. However, fossil evidence suggests that most of these missing morphotypes were represented by species from those territories that went extinct during the Quaternary. Furthermore, reconstructing the hypothetical felid ensembles before Pleistocene extinctions rendered the continental felid faunas remarkably more similar to each other than they presently are, leaving their remaining, relatively minor differences to outstanding geographic singularities of each continental land mass.

The Cranial Morphospace of Extant Marsupials

“…In turn, Australasian marsupials underwent a major ex�nc�on of debated (human versus clima�c) cause(s), especially megafauna (see Black et al 2012). Effects of Quaternary ex�nc�ons on morphospace structuring have been documented in placental clades (e.g., Felidae; Morales and Giannini 2021).…”

Using linear measurements to diagnose the ecological habitat of Spinosaurus

“…To account for isometric scaling due to body size, each measurement was divided by the geometric mean of the skull ( Sakamoto & Ruta, 2012 ). The resultant dimensionless Mosimann shape variables have been demonstrated to out-perform residuals as size-adjusted shape variables, and have the additional benefit of only requiring information from a single specimen ( Mosimann, 1970 ; Sakamoto & Ruta, 2012 ; Morales & Giannini, 2021 ). These values were log-transformed ( Glazier, 2013 ), centred on 0 and scaled to unit variance to conform to the assumptions of cluster analysis and to a lesser extent, principal component analysis (PCA).…”

“…Despite a fragmentary cranial fossil record, sufficient material of multiple spinosaurids (both those considered semi-aquatic and terrestrial) exists to compare these features and their relation to ecological realm occupancy, both within Spinosauridae and to other taxa where ecological realm occupancy is undisputed. To this end, cranial linear morphometric analyses ( Mosimann, 1970 ; Sakamoto & Ruta, 2012 ; Morales & Giannini, 2021 ) have previously been successful in revealing the taxonomic affinities of unidentified specimens ( Blake et al, 2014 ; Naish et al, 2014 ). Linear morphometric analysis could thus be a suitable method for categorising the ecological realm occupancy of Spinosaurus .…”