Insects are an extremely diverse group accounting for approximately two thirds of described species. Morphology is highly variable across insects; however, preliminary studies have shown a significant difference in the correlation (integration) of linear traits in holometabolous and hemimetabolous groups, with correlations being much higher in holometabolous insects (0.84 vs 0.5). New imaging technologies and analytical approaches allow for a far more detailed study of morphological evolution across a range of invertebrates and for testing fundamental hypotheses on function, development, key innovations, and macroevolutionary patterns across this diverse group of organisms.
In this project, the student will expand a recently-gathered pilot 3D dataset for invertebrates by focusing in on a few clades for more targeted study. Using data from the preliminary CT scans of representatives from the main orders of Insecta we will select specific groups for in-depth analysis of interspecific variation, ecomorphology and intraspecific variation. Analysis of interspecific variation and ecomorphology will use a cutting-edge AI pipeline for rapid segmentation of hundreds of micro-CT scans to be gathered with a robot system developed for this project. We will then apply deep learning and landmark-free methods to conduct surface geometric morphometric analysis, using evolutionary modelling and comparative methods to analyse patterns and rates of shape evolution in different morphological components, how these different regions relate to each other (integration), and how these aspects relate to development, function, and ecology. Intraspecific variation for a few exemplar species will be analysed at the population level, from either the historic museum collections or, where necessary, the collection of contemporary material.