Patterns of Complexity and Homoplasy in the Evolution of Tetrapods

Abstract

Systematists and paleobiologists interested in the evolution of diversity routinely tally numbers of species or genera, and plot these numbers through time in order to analyse patterns of origination, extinction and turnover, or compare species numbers across contemporaneous clades. Allied to this, it is now common practise to quantify variation in morphological form or morphological disparity. Empirically, morphological disparity is variously decoupled from diversity, with a tendency for low diversity but relatively high disparity early in the evolution of major clades. The Phanerozoic history of life is characterised as one in which species diversity has irregularly increased (modelled by various functions and interrupted by numerous setbacks) but in which disparity has taken a much more damped path.
A widely-acknowledged trend through the history Life is for an increase in maximum complexity. Complexity usually described as a function of number, heterogeneity and hierarchy of dynamic interactions among parts of biological systems. Complexity has been much less studied than species diversity and disparity, not least because of the difficulties inherent in defining and quantifying it. Here, I devise some possible indices of complexity in the context of a single system at a single level, namely the differentiation of the vertebral column of tetrapods. By analogy with frameworks for indexing disparity, there is no universal system, and estimates of complexity are only relative to those frameworks. I apply both geometric and count-based indices of serial differentiation to two samples of mammal taxa, asking whether there are any consistent trends in these indices throughout phylogeny, and whether there are particular signatures associated with different modes of locomotion and different habitats. Surprisingly, there is a consistent trend for the number of vertebrae to increase (more elements and therefore more degrees of freedom), but for count-based indices of element diversity to decrease. Reductions in the number of vertebrae are also associated with increases in the geometric shape differentiation between those elements. There were remarkably few significant locomotory or habitat signatures, with flapping fliers (bats) and to some extent hopping and bipedal mammals having more geometrically diverse element shapes, and marine mammals having longer, less differentiated columns.
As a necessary pre-requisite for this work, I addressed a perennial concern for evolutionary biologists seeking to make inferences from morphological character sets, namely the generality of the signals within subsamples of data. This issue is usually resolved by compiling the largest, holistically-sampled data sets possible on the principle of total evidence. In morphological phylogenetics using fossil taxa, characters may be restricted to particular regions of the body, or to those with higher preservation potential. This can systematically distort inferred trees. We apply established and new tests for partition homogeneity to the phylogeny of dinosaurs, finding that trees inferred from characters of the skull and dentition are significantly at odds with those from the rest of the skeleton in about half of cases (which is more often than in most other vertebrate groups). The issue is pertinent to our use of the vertebral column as a proxy for anatomical complexity, since we do not have data from other anatomical systems (and few good candidates) that might offer tests of generality. In the absence of this, we demonsrate that phylogeny inferred from the mammalian vertebral column is not significantly different from that inferred from characters of the rest of the body. Moreover, the same is true for subclade disparity inferred from these two partitions.
The generality of these patterns observed in mammals remains to be tested in other clades. Birds and squamate furnish excellent candidate groups.

Date of Award	14 Feb 2022
Original language	English
Awarding Institution	University of Bath
Supervisor	Matthew Wills (Supervisor) & Daniel Field (Supervisor)