Ontogeny and systematics of Toernquistiidae (Trilobita, Proetida) from the Ordovician of the Argentine Precordillera

Journal of Paleontology, Mar 1998 by Chatterton, Brian D E, Edgecombe, Gregory D, Waisfeld, Beatriz G, Vaccari, Norberto E

The third larval, and second metaprotaspid, stage of the Argentinian toernquistd species is larger, with more distinct furrows, and a more convex posterior margin (between subtransverse and distinctly convex backward). The posterior marginal furrow of the cranidium and a preglabellar field are usually distinguishable. Where the free cheeks have been found in place, the genal spines may or may not project outward as independent spines (see Chatterton, 1994, figs. 6.20, 6.21, Dimeropyge).

The ontogeny of Lasarchopyge benedettoi includes larger metaprotaspid larvae than the other two species (Figure 3). These protaspides probably represent a third instar, which differs in having a more developed (more distinct somites), relatively longer protopygidium.

PHYLOGENETIC ANALYSIS

In order to classify our new species from Argentina, it is useful to formulate a hypothesis of their phylogenetic relationships. To this end, we have conducted a cladistic analysis of the Argentinian species, the better known species of Toernquistdae and Dimeropygidae (including type species), and some taxa within the Hystricuridae, a likely paraphyletic group which was probably ancestral to these two families (see Table 2 for list of species included). Some other toernquistiid and dimeropygid species were also considered for analysis, but were not included in the analyses because they are too incompletely known. We generated a set of 55 characters (see list below) that were scored for the included taxa. The relationships, inferred from our cladistic analyses, were used in part to justify assigning our species to pre-existing or new genera, erecting the new genera Lasarchopyge and Paratoernquistia, and separating Toernquistdae from Dimeropygidae, despite the fact that several workers have regarded these family level taxa as synonyms (e.g., Whittington in Moore, 1959). We note, however, a major problem in the amount of data available for some hystricurid and many toernquistiid species (often based only upon cranidia).

The characters and character states included in our analysis are all from cranidia, free cheeks, rostral plates (where present), and pygidia (with the exception of a single thoracic character). Hypostomes and thoraxes are known for few of the species considered. Characters of the larval stages are not used in the analysis (known for too few of the species), but are considered in discussing relationships, and in assessing polarity. Some characters were considered and then discarded because their states were found too difficult to code consistently, they are available for too few of the species considered, or the derived states are autapomorphic. These include: number of thoracic segments; and presence/absence of large axial spine on the sixth thoracic segment.

Hystricurids were chosen as an outgroup to root the trees because a number of authors have suggested that the Hystricuridae includes the ancestors of a number of proetide groups, including the Dimeropygidae and, by inference, the Toernquistdae (Chatterton, 1971, 1994; Fortey and Owens, 1975; Fortey and Peel, 1989).