<p>Modularity is a fundamental concept in biology. Most taxa within the colonial invertebrate phylum Bryozoa have achieved division of labor through the development of specialized modules (polymorphs), and this group is perhaps the most outstanding exemplar of the phenomenon. This thesis addresses several gaps in the literature concerning the morphology, ecology, energetics, and evolvability of bryozoan polymorphism. It has been over 40 years since the last review of bryozoan polymorphism, and here I provide a comprehensive update that describes the diversity, morphology, and function of bryozoan polymorphs and the significance of modularity to their evolutionary success. While the degree of module compartmentalization is important for the evolution of polymorphism in bryozoans, this does not appear to be the case for other colonial invertebrates. To facilitate data collection, I developed a classification system for polymorphism in cheilostome bryozoans. While classification systems exist for bryozoan colony form, the system presented here is the first developed for polymorphism. This system is fully illustrated and non-hierarchical, enabling swift classification and statistical comparisons at many levels of detail. Understanding community assembly is a key goal in community ecology, but previous work on bryozoan communities has focused on colony form rather than polymorphism. Environmental filtering influences community assembly by excluding ill-adapted species, resulting in communities with similar functional traits. An RLQ (a four-way ordination) analysis incorporating spatial data was run on a dataset of 642 species of cheilostomes from 779 New Zealand sites, to investigate environmental filtering of colony form and zooid polymorphism. This revealed environmental filtering of colony form: encrusting-cemented taxa were predominant in shallow environments with hard substrata (200 m). Furthermore, erect taxa found in shallow environments with high current speeds were typically jointed. Surprisingly, polymorphism also followed environmental gradients. External ovicells (brood chambers) were more common in deeper, low oxygen water than immersed and internal ovicells. This may reflect the oxygen needs of the embryo or increased predation intensity in shallow environments. Bryozoans with costae (rib-like spines) tended to be found in deeper water as well, while bryozoans with calcified frontal shields were found in shallow environments with a higher concentration of CaCO₃. Avicularia (defensive grasping structures) were not related to environmental conditions, and changes in pivot bar structure with depth likely represent a phylogenetic signal. Factors influencing community assembly were somewhat partitioned by levels of organization, since colony form responds to environmental conditions, while the effects of evolutionary history, predation, and environmental conditions were not well-separated for zooid-level morphology. Finally, rootlets may have been a key innovation that allowed cementing taxa to escape hard substrata, potentially contributing to the cheilostome radiation. Despite the diversity of life on earth, many morphologies have not been achieved. Morphology can be limited by a variety of constraints (developmental, historical, biomechanical) and comparing the distribution of realized forms in a theoretical form-space (i.e. “morphospace”) can highlight which constraints are at play and potential functions. If traits cluster around biomechanical optima, then morphology may be shaped by strong selective pressures. In contrast, a well-explored (filled) morphospace suggests weak constraints and high morphological evolvability. Here, constraints on morphospace exploration were examined for 125 cheilostome bryozoan species from New Zealand. The mandible morphospaces for avicularia (beak-like polymorphs) were visualized using Coordinate-Point Extended Eigenshape analysis. Mechanical advantage, moment of inertia, drag, peak force, and rotational work required to close the mandible were calculated for theoretical (n=47) and real mandibles (n=224) to identify biomechanical optima. The volume and surface of area of the parcel of water passed through by the closing mandible (referred to as the “domain”) was also calculated. The theoretical morphospace of avicularia is well-explored, suggesting they are highly evolvable and have relaxed developmental constraints. However, there may be constraints within lineages. A well-developed fulcrum (complete pivot bar) may be an evolutionary pre/corequisite to evolving mandibles with extreme moments of inertia such as setose and highly spathulate forms. The most common mandible shape, triangular, represents a trade-off between maximizing domain size, minimizing energetic cost (force and construction material), and minimizing the potential for breakage. This suggests that they are well suited for catching epibionts, representing the first empirical evidence for avicularian function. Tendon length and mechanical advantage are limited by tendon width, which itself is constrained by the base width of the mandible. This explains the low mechanical advantage of setose mandibles and suggests that they are unable to grasp epibionts. The calories required to close the mandible of an avicularium (estimated from rotational work) are quite small (1.24 x 10⁻¹⁶ to 8.82 x 10⁻¹¹ cal). Overall, this thesis highlights the complexity of bryozoan polymorphism and suggests cheilostome avicularia could provide a unique evolutionary system to study due to their apparent lack of strong developmental constraints. Future studies into the ecology of polymorphism should focus on the degree of investment (polymorph abundance within a colony) rather than presence or absence.</p>
Neural network detected in a presumed vestigial trait: ultrastructure of the salmonid adipose fin
