FIRST CARBONIFEROUS PROTASPID LARVAE (TRILOBITA)

Journal of Paleontology, Jul 2005 by Lerosey-Aubril, Rudy, Feist, Raimund

It is noteworthy that the hypostome almost fills the ventral opening of the larvae. As mentioned by Fortey (1990) the labral plate of crustaceans must be the homologue of the hypostome. In nauplius larvae, it is also particularly large (Roy and Fahraeus, 1989; Fortey, 1990). Evidently, the hypostome at this stage forms a kind of shield that protects the larvae ventrally. It is only after metamorphosis that the hypostome acquires a function in feeding as it is supposed to have in the adults (Fortey and Owens, 1999). A ligamental connection, possibly with the rostral plate and anterior pits, may have occurred as the hypostome did not seem to be fused to the rostral plate nor directly to the free cheeks (see fig. 8.3, 8.8, 8.12 in Edgecombe et al, 1997). This may have given a certain mobility to the junction.

Post-metamorphic stages.-Comparison between different metaprotaspid types: Metaprotaspid type 1 displays obvious morphological differences comparing to types 2-4. These are: maximum width located at the midlength (tr.) of the protopygidium (leading to a grossly triangular shape), cranidium relatively shorter (Table 1), glabella slightly pointed anteromedially, preglabellar region wide (sag.), α-β less divergent, β-γ more convergent, anterolateral margins of the protopygidium diverging, protopygidial border large, flat and well defined, and furrows generally deeper. As far as it can be observed, the only specimen belonging to type 5 could be considered as the morphologically closest larva for this type. In contrast, types 2-4 are particularly similar. The major difference between these three types is the size (Table 1), especially for type 4. Considering this parameter, type 4 may represent an earlier larval stage of one of the species represented by type 2 and type 3. If we calculate the Dyar's coefficient for maximum width ("a(l)") in the two hypotheses, it seems more probable to regard type 4 as being the stage preceding type 2 (a(l): 1.26) rather than type 3 (a(l): 1.32) according to Prizbram's rule. However, no obvious features support one or the other of these assumptions. For this reason and considering all the discussions about the reliability of Dyar's coefficient (Chatterton and Speyer, 1997, p. 208), we prefer to regard the type 4 as a separate species. Now, if we consider just types 2 and 3, it seems likely that they are not different stages of a single species as they exhibit only slight differences in overall proportions (Table 1). Moreover, type 3 differs from type 2 in having a piriform glabella, a straight anteromedial part of the cranidial margin, a very high degree of effacement, an important enlargement of the cranidium posteriorly, and a less important vault in lateral view. Nonetheless, the morphological similarity of types 2-4 suggests that they all belong to species that are taxonomically very close to each other.

Comparison with previously described proetoid metaprotaspides.-Regarding absence of ornamentation, elongate overall shape, and important size, the metaprotaspides described herein are more similar to previously described proetoid larvae than to larvae from all other trilobites belonging to the Proetida. Hence, the metaprotaspides also separate Proetoids from Aulacopleuroids/Bathyuroids like anaprotaspides and anaprotaspid hypostomes do. Proetoid metaprotaspides have been previously described for four species: Decoroproetus beecheri [previously assigned to Triarthrus eatoni by Beecher (1893) but recently reassigned to D. beecheri by Edgecombe et al. (1997)], Devonoproetus talenti (Chatterton, 1971), Stenoblepharum astinii (Edgecombe et al., 1997), and Dechenella sp. (Chatterton et al., 1999). The comparison of our metaprotaspid types with these previously described larvae is summarized in Table 2.