Crustacean disparity through the Phanerozoic: comparing morphological and stratigraphic data

Abstract

Crustaceans have been an important component of marine diversity and biomass since the earliest Phanerozoic. With a relatively well-documented fossil record, they provide an excellent subject for a continuous study of disparity (approxiamtely bodyplan variety) from the Cambrian to the Recent. A data base of 135 morphological characters forms the basis for cladistic and morphospace studies at the ordinal and sub-ordinal level. Gross cladistic topology is: (Eumalacostraca+Hoplocarida vs Maxillopoda) vs Phyllopoda (paraphyletic). Each of these groups is of approximately equal disparity, and occupies a distinct region of the morphospace plot. A few problematical fossils (e.g.WaptiaandOdaraia) fall close to the base of the tree. Comparison of the cladogram with stratigraphic range data indicates the location of probable ghost lineages, and randomization procedures provide a statistical test of the goodness of fit of a given set of stratigraphic ranges to a given tree topology. Disparity indices are calculated at series and stage intervals. Observed range data indicate that Cambrian disparity was approximately one third its present level. The Earliest Ordovician saw a marked decrease, with an increase and subsequent plateau through rest of the period. Increases through the Silurian and Devonian corresponded to the radiation of branchiopods, cephalocarids, and latterly the Eumalacostraca and Hoplocarida. By the end of the Carboniferous, observed disparity had reached over four fifths of Recent levels, and the remaining history of the group saw a gradual but slightly irregular increase up until the end of the Tertiary. Indices of disparity incorporating ghost lineages exhibit less marked peaks and troughs, with fewer perturbations overall. Cladistically-implied disparity in the Lower Cambrian is estimated at three quarters of that in the Recent. Rarefaction is used to compare actual levels of disparity at each time interval with the mean for a similar number of taxa selected randomly from the list of all realized bodyplans. Most intervals preserved a range of forms more disparate than the mean of random samples drawn from the pool of all the taxa considered. From the Triassic to the Recent this difference was intermittently significant. Once occupied, extremes of morphospace tend not to fall vacant again.