Morphological disparity and its correlates across clades

: (Alternative Format Thesis)

Abstract

Morphological disparity refers to the degree of difference in physical form or structure among a group of different, but related species (clade) or other operational taxonomic units (e.g., genus, family, and order) and can be used as an index of the evolutionary diversity within these groups. It contrasts with species richness or diversity, which is usually construed as a tally of the number of species present within a group. Morphological disparity can be quantified using numerous indices, including the number of unique morphological features present within a group and the range or variance in the size or shape of organs or body parts. High morphological disparity is often regarded as a correlate of evolutionary innovation and adaptability, as it reflects the ability of a group of organisms to evolve a wide range of forms and functions in response to different selective pressures.

Morphological disparity is typically quantified in studies of deep time and is used as a proxy for amounts of evolution represented by species across clades or time bins. In general, morphological disparity and diversity are believed to be decoupled, since modest numbers of species can encompass high levels of morphological disparity, while highly speciose groups can display great morphological conservatism. Multidimensional ordinations are often used to visualise the distributions of taxa within abstracted “morphospaces”, which are effectively maps of morphological space in which the most similar species are plotted close together and the most dissimilar far apart. Where the underlying data are continuous variables or geometric morphometric data, it is often possible to interpret changes along axes geometrically and in verbal semantic terms. Where the underlying data are discrete characters, such interpretations are typically not possible.

In this thesis I investigate patterns of morphological disparity across a sample of vertebrate and invertebrate clades. My initial focus is the Paleopheidae, an enigmatic family of extinct marine snakes where I sought to clarify their poorly understood evolutionary origins, as well as their post K-Pg radiation. The morphological disparity of Paleopheidae peaks early in their history, a pattern of “early high disparity” that has been widely demonstrated in other groups. Within this dataset, I also noted a positive correlation between the number of variable discrete characters within subclades, and the disparity inferred for those same subclades. This implies that there may be artefacts related to the intensity of character sampling and morphological disparity.

Therefore, I conceived of several possible indices of molecular variation, or quantifiers of the “amount” of evolution as analogues of – and adjuncts to – indices of morphological disparity. In contrast to codifying morphological character data, the acquisition of molecular sequence data is a much more objective process. Principally, I devised two approaches to quantifying molecular disparity. The first approach used data on the presence/absence of gene families and was applied to primates at the familial level. This relied upon comprehensive genome annotation across our sample, which is often unavailable for many species. The second approach used nucleotide sequence data, coupled with variously sophisticated models of nucleotide evolution. Such models can be applied to single loci, and inter-taxon distance matrices between single loci were all found to be strongly correlated. It can also be applied to large, concatenated matrices of many hundreds of genes. This latter method allows for the repurposing of data from total-evidence analysis. I applied this latter method across tetrapods, plants and invertebrate groups and contrasted and compared this index to morphological disparity.

In general, molecular, and morphological disparity were found to be decoupled. As with morphological disparity, molecular disparity also appears to be decoupled from species richness. In Aves, I found that the molecular disparity of subclades correlates positively with geographic range and latitude (but not with annual temperature). I also found a correlation between morphological disparity and indices of evolutionary distinctiveness (ED) which is utilised to assess conservation priorities. In this study I used an adjunct to ED: “morphological eccentricity” (ME), first described in (Wills, Briggs and Fortey, 1994). ED is given by the down-weighted (by the number of species subtended on each branch length) to each species and is combined with the globally endangered (GE) score to give the EDGE score used to assess conservation priorities. ME is tallied as the per-species generalised Euclidian distance from the morphological centroid. By combining the ME score with the globally endangered score I envisage a morphological index of conservation given by the “MEGE” score, combining ME and GE. I utilised the same morphological dataset to assess the degree of homoplasy across different partitions of avian morphology, where I found that within the osteological characters, the skull and leg bones are the most phylogenetically informative. Within the largest morphological character set I found the soft characters to be more informative than the hard character partition, however, this was reversed in two further analyses, albeit of smaller partitions. Phylogenetic partition homogeneity tests revealed significant incongruence between the majority of these partitions. In the largest morphological dataset soft characters to be more phylogenetically informative than osteological characters. Within the osteological partitions, the cranium and leg bones revealed the least homoplasy.

Morphological disparity is a measure of the physical differences in form and structure among species within a group or clade and is often used as an index of evolutionary diversity. It can be quantified using various indices, such as the number of unique morphological features or the range in size or shape of organs/body parts. Morphological disparity is typically studied in deep time and can be used as a proxy for amounts of evolution represented by species in subclades or time bins. Multidimensional ordinations are commonly used to visualize the distribution of taxa within abstracted "morphospaces", which are maps of morphological space.

Within the field of evolutionary analysis, both molecular and morphological data offer valuable insights. Phylogeny, a fundamental aspect of evolutionary studies, can be inferred from both molecular and morphological character data. While likelihood and Bayesian tools were initially developed for analysing molecular sequences, they are now increasingly applied to morphological character data as well. In turn, clock and rate studies, originally focused on molecular data are now being extended to include morphological data. I explored the applicability of molecular disparity indices, what such an index might entail and how it may be applied as an evolutionary index. Work on molecular indices is relatively limited compared to the extensive research on morphological disparity. Though I present some use cases for molecular disparity as an evolutionary index, more research on the topic is required to assess the robustness of the various indices. Furthermore, studying homoplasy patterns across large datasets and exploring the ontogenetic and architectural depth of characters can provide valuable insights into evolutionary processes. Future research efforts focusing on these areas will contribute to a more comprehensive understanding of evolutionary patterns and relationships, bridging the gap between molecular and morphological approaches.

Date of Award	13 Sept 2023
Original language	English
Awarding Institution	University of Bath
Supervisor	Matthew Wills (Supervisor), Mark Puttick (Supervisor), Araxi Urrutia (Supervisor) & Daniel J. Field (Supervisor)