Abstract

Phenotypic integration and the associated concept of modularity have long been of interest to those seeking to understand the mechanics of evolvability. Since the publication of the pioneering work by Olson and Miller (1958), the study of patterns of trait covariances has itself evolved into a dynamic field that has its roots in ideas that were beginning to take shape as early as the nineteenth century. Cuvier’s “principle of correlated parts” was such an observation, despite his flawed claim that the seeming perfection in the fit of each component to form a functional whole opposed evolutionary thought as change in one part would cause the entire system to fail (Mayr, 1982). The idea that organisms are composed of correlated sets of parts lends itself well to methods that utilize morphometric data. Because of this, the study of integration and modularity can be extended into the fossil record, where oftentimes the only data available is that relating to morphology. The fossil record also allows for an investigation of these patterns in deep time, which is an opportunity that is not afforded by the study of extant organisms alone. Of particular interest is the idea that integration serves as a constraint on clade disparification. The term “constraint” is intuitively thought of as a limiting factor in terms of evolvability, but it is perhaps better thought of as a factor that can channel evolutionary variation in certain directions (e.g. Gould 1989, 2002). In order to investigate how phenotypic integration affects patterns of morphological diversity, one must first determine if it does, in fact, serve as a long-term constraint. The following studies aim to address that question within the framework of a fossil study system. Chapter 1 presents the phylogenetic framework for the clade of interest, the Agnostina. These small, blind arthropods have been notoriously problematic in terms of systematic classification given their conserved morphologies and their lack of reliable diagnostic characters. Using Bayesian inference and parismony methods, trees were recovered that support much of the existing rank-based systematics of the Agnostina, Chapter 2 explores the structure of integration in a middle Cambrian agnostine, Pentagnostus brighamensis (Resser, 1939) using geometric morphometric methods. The aim of this study was achieved by determining that the integration structure of this agnostine is driven by direct interactions among developmental pathways, thus potentially serving as a long-term constraint. The final chapter, Chapter 3, compares the structure and level of integration in the sister taxon of P. brighamensis, Pentagnostus segmenta (Robison, 1964), as determined by the phylogenetic analyses in Chapter 1. A phylomorphospace was constructed, highlighting Peronopsidae Westergård, 1936 as a morphologically conserved clade relative to other agnostines. It was shown that while Peronopsidae is conserved, that conservation is not due to phenotypic integration acting as a constraint on the clade as the degree of integration in both species is weak and the structure of integration is not conserved among sister taxa. These studies are among the first to investigate the role of integration as a constraint on morphological evolution in fossil arthropods, and they are the first to do so focusing on agnostine arthropods. These are still the first steps into a rich field of study that remains disproportionately represented by research into extant organisms. The methods used herein are part of a toolkit allowing one to interrogate the fossil record and take advantage of its opportunities to pursue questions that can only be answered by observing patterns through time.