2004 Denver Annual Meeting (November 7–10, 2004)

Paper No. 12
Presentation Time: 10:45 AM


ANDERSON, Laurie C.1, ROOPNARINE, Peter D.2, ARONOWSKY, Audrey3 and CALLIER, Viviane2, (1)Geology and Geophysics, Louisiana State Univ, Baton Rouge, LA 70803, (2)Dept. of Invertebrate Zoology and Geology, California Academy of Sciences, 875 Howard St, San Francisco, CA 94103, (3)Geology and Geophysics, Louisiana State Univ, E235 Howe-Russell Geoscience Complex, Baton Rouge, LA 70803, laurie@geol.lsu.edu

Morphological and phylogenetic character data can be difficult to obtain and frustrating to interpret in morphologically conservative groups, particularly at low taxonomic levels. For such groups, a new morphometric network technique developed by one of us (Roopnarine) is particularly useful in quantifying patterns of morphologic variation. Morphometric network analysis uses geometric landmark data and calculates patterns of integration using partial correlations among landmarks. These patterns may be visualized directly on configurations of landmarks, and used in conjunction with geometric morphometric analysis and standard multivariate techniques to define character states and recognize morphospecies. In addition, patterns of integration can be treated as mathematical networks, allowing modularity to be detected and used to test hypotheses of constraint and evolvability. Further, these integration data may be appropriate for use as characters in phylogenetic analyses.

We tested the utility of morphometric network analysis using several morphologically conservative molluscan clades: the corbulid bivalves Bothrocorbula and Caryocorbula, the venerid Lirophora, and the gastropod family Naticidae. For some of these taxa, previous attempts to use morphometric data to detect morphospecies has met with limited success. In fact, Lirophora has resisted traditional morphological description, with single species names being applied to temporally and geographically distant taxa. For naticids, the new source of characters provided by geometric morphometrics and integration are invaluable to the phylogenetic placement of problematic fossil taxa. In the two corbulid genera, species-level phylogenies are hampered by few discrete characters. In these clades, species differ primarily in valve outline shape, which is difficult to deconstruct into character states appropriate for phylogenetic analyses. In addition, strong interspecific allometry in Caryocorbula indicates that constraint channeled its morphologic evolution. We found that for most taxa the use of morphometric network analyses greatly increased our ability to distinguish taxa, recognize modularity in shape variation, and parse out overall shape differences into character states of use in phylogenetic analyses.