Developmental palaeobiology of trilobite eyes and its evolutionary significance

Understanding of the calcified composite eyes of trilobites, the oldest preserved visual system, has advanced greatly in recent decades. Three types of trilobite eye occur, the more derived abathochroal and schizochroal types having evolved neotenically from holochroal eyes. Comparative morphology and phylogenetic considerations suggest that all three eye-types were underlain by common developmental systems. So far, understanding of these systems has been based entirely on morphological data from fossils, particularly the way the visual surface grew and the patterning of lens emplacement. Lenses characteristically form a hexagonal array comprising horizontal rows and, conspicuously in schizochroal eyes, dorso-ventral files. Because individual trilobites sometimes have eyes with different numbers of files, file number must reflect the operation of a developmental programme rather than being under immediate genetic control. An empirical developmental model has been devised to describe trilobite eye development, with separate rules dealing with the initiation of lens emplacement, growth and differentiation of the visual surface, and the termination of lens emplacement. Rarely, trilobites may have visual surfaces of normal size, but which lack lenses. This confirms that visual surface growth must have been regulated separately from lens emplacement, and is a feature that cannot be accounted for by the existing developmental model. Such a developmental separation is one of a number of similarities shared with Drosophila, the modem arthropod in which eye development is best understood. Many aspects of eye development are conserved in the Euarthropoda, and in bilaterian metazoans in general. A revised model for trilobite eye development is proposed using extant phylogenetic bracketing, interpreting morphological data from the fossils in the context of the hierarchy of developmental controls now becoming known from living animals. This new model suggests that overall eye shape and size did not require differential growth of the generative zone, as previously thought, and that no separate instruction was needed to specify the termination of lens emplacement. Instead, these features were regulated directly, by controlling the proliferation of cells making up the nascent visual surface. A process documented from Drosophila, which involves the selective inhibition of cells in front of a wave-like front of differentiation, and that is regulated by widely conserved genes, can be used to explain how the trilobite visual surface became differentiated. The model implies also that changes in hormonally regulated developmental pathways known from recent arthropods may have been responsible for the development of abathochroal and schizoehroal eyes, and for heterochronic secondary eye reduction and blindness in trilobites. (c) 2005 Elsevier B.V. All rights reserved.