Use of landmark and outline morphometrics to investigate thecal form variation in crushed gogiid echinoderms

Abstract

Gogia is a primitive genus of pelmatozoan echinoderm the investigation of whose systematics, taxonomy, and palaeoecology has been invigorated recently by the recovery and description of articulated specimens and related genera of blastozoan species from Cambrian strata in Utah and China. To date investigation of gogiid systematics has been undertaken almost exclusively using qualitative methods to assess morphological variation owing, at least in part, to the crushed character of articulated fossils. This practice reflects the widespread belief among palaeontologists that quantitative approaches to morphological analysis can only be employed successfully on well-preserved specimens. Accordingly, no geometric morphometric analysis of gogiid morphology has been published to date. In order to investigate whether morphometric approaches to gogiid character analysis might be used to recover biological signals the thecae of 27 gogiid specimens collected from a single bedding plane were subjected to geometric analysis using landmarks, boundary outline equi-spaced semilandmarks, boundary outline sliding semilandmarks, and combined landmark+outline sampling protocols. Results indicate that, despite the crushed and deformed character of these specimens, robust and unquestionable biological signals can be extracted from the patterns of variation presented by these specimens by all the aforementioned methods. Based on these morphometric results it is concluded that (1) the sample exhibits several complex patterns and shape variation some of which may be inconsistent with the assumption that all specimens included in the sample should be assigned to the same species, (2) the primary pattern of thecal shape change appears to be highly correlated with an aspect of specimen size (thecal length) suggesting that various life-cycle stages are present within the sample, (3) assessments of right-left symmetry can be used to distinguish biological from post-mortem deformational aspects of shape variation, (4) the sliding semilandmark procedure failed to improve semilandmark correspondences across the sample and, indeed, appears to have degraded rather than improved the assessment of the biological shape variation, and (5) morphometric sampling strategies that combine the information recorded by landmarks and equi-spaced boundary outline semilandmarks appear to recover the greatest proportion of biologically interesting shape variation.