Publication Type: | Journal Article |
Year of Publication: | 2008 |
Authors: | A. K. Roberts, Smith, D. M., Guralnick, R. P., Cushing, P. E., Krieger, J. |
Journal: | Geological Society of America Special Paper |
Volume: | 435 |
Pagination: | 105-116 |
Date Published: | 2008 |
ISBN Number: | 0072-1077 |
Keywords: | ] [Colorado / Florissant Formation / ] [Eocene / Upper Eocene / ]., Araneae (Arachnida): [Fossil]., Araneae [Identification techniques / Outline morphometric approach to, Biometrics, Caenozoic, Geological time periods, identification / ] [Palaeontological techniques / / ] [Biometrics / /, Land zones, Nearctic region, North America, Palaeogene, Techniques, Tertiary, USA |
Abstract: | Using available shape characters we conducted an outline morphometric analysis to make family-level identifications of fossil spiders from the Florissant Formation in Colorado. In this analysis we used carapace shape because it is a character that can be observed on most fossil spiders, and we also used linear leg characters. All measurements were first made on 202 modern spiders from eight families found in localities similar to the fossil lake environment. A multiple discriminant analysis (MDA) of the eigenshape axes was used to predict family placement among the modern data set to test the accuracy of the predictions. The modern spider families that were predicted correctly most often were the Salticidae (91.2%), Linyphiidae (80%), Dictynidae (76.5%), Tetragnathidae (68.2%), Clubionidae (66.7%), and Araneidae (65.5%). Families that produced less successful results were the Agelenidae (46.7%) and the Lycosidae (39.1%). Forty-three fossil spiders from Florissant were then added to the model to determine their family placement. All fossils were placed into modern families with varying degrees of accuracy. Only 42% of our identifications agree with those made by previous authors, but it is likely that these specimens were originally misidentified. With the addition of more taxa and characters, we believe that an outline morphometric approach shows great promise for helping to identify fossil taxa that are lacking traditional taxonomic characters. |
URL: | <Go to ISI>://ZOOREC:ZOOR14609064889 |