All members of the chordate phylum are united by a shared body plan – a stereotypical composition and topology of tissues that emerges in the wake of gastrulation and defines the major anatomical patterns emerging in later development. For this reason, although there is remarkable anatomical diversity between adult chordates, they are all united by a common set of design principles. The epicentre of the body plan is the notochord, located at the axial midline, which is flanked dorsally by a neural tube, bilaterally by a metameric pattern of somites, and ventrally by a primitive gut tube. If we are to understand how chordates emerged in evolution, and have subsequently diverged, it is therefore pertinent to ask how these traits are assembled in the embryo. However, here lies a paradox. The conservation of form emerging in the wake of gastrulation is matched by diversity in the morphogenetic processes that put it together. Therefore, if we are to understand chordate evolution, we must take a comparative approach to body plan morphogenesis. In this case, we can infer ancestral principles of development by comparing vertebrates with the most basally-branching member of the chordate phylum – the cephalochordate, amphioxus.

In this PhD thesis, I present a decomposition of body plan morphogenesis in the amphioxus embryo at three primary scales of observation. First, I identify the major changes in tissue shape and cellular architecture involved in body plan assembly, using a fusion of classical embryological approaches and three-dimensional morphometrics. This exposes the amphioxus embryo as a largely growth-free system, in which axial development primarily depends on tissue-specific programmes of convergent extension behaviour. Within this, volumetrically-reductive cell division acts to regulate tissue complexity via focal regulation of cell size and number, and is required in axial progenitor cells of the tailbud for full elongation of the body axis. Second, I reconstruct patterns of cell shape change underpinning notochord development using single-cell morphometrics and morphospatial embedding. Despite its superficial simplicity, I show the amphioxus notochord to be remarkably complex, with evidence of region-specific behaviours, and an interplay between growth and cell rearrangement that generates axial length. Finally, I study the diversity of axial progenitor cell types in amphioxus using quantitative in situ imaging of gene expression, and a bespoke in silico pipeline for cell state classification and spatial mapping. Here, aided also by in vivo signalling perturbations, I locate a population of vertebrate-like neuromesodermal progenitors that generate the posterior spinal cord.

The results I present in this thesis reveal remarkable complexity in amphioxus axial development, and a developmental programme replete with morphogenetic principles shared with vertebrates. For this reason, we can infer a basic repertoire of developmental innovations required for body plan formation in the first chordates, as relatively simple structural perturbations of the gastrula. From this foundation, I use my findings to propose that the radiation of chordates has depended more on tweaks in the magnitude of processes already present in the ancestral chordate, than the innovation of new processes de novo. This thesis therefore contributes new understanding in chordate development and evolution. It also offers a suite of techniques suitable for studying and integrating morphogenesis a diversity of research organisms.