Hadrosaurs. David A. Eberth

hatchet tends to be more curved; and associated with these in some individuals there is also an unusual, somewhat irregular mass of bony material located in the center of the chest and referred to as an “intersternal ossification” (Norman, 1980:fig. 56). The humerus is very robust and nearly straight rather than strongly sigmoid (obscured by crushing), and has a massively thickened deltopectoral crest. The forearm bones are equally massive and parallel, with very little gap between the shafts of the two bones, which supports the contention that this dinosaur used its forelimbs for walking and body-weight support. The carpals and metacarpal I are fused into a large block (Norman, 1980:fig. 59); the metacarpal has a roller-like articular surface for articulation of the large, conical, and slightly curved pollex. Unlike the condition seen in Hastings Group taxa (B. dawsoni and H. fittoni), the pollex does not become fused to the carpometacarpal block in mature specimens and remained freely mobile at its base. A small, flattened phalanx is occasionally seen lodged in the base of the pollex ungual. The central metacarpals are more massive, and in proportion shorter, than those of Mantellisaurus (Fig. 2.15).

      The pelvis (Fig. 2.22) is distinct from that seen in Mantellisaurus (Fig. 2.16). The ilium has a very robust, thick, preacetabular process, which is supported by an enlarged medial ridge (note the shape of its cross section in silhouette; Fig. 2.22). The main part of the iliac blade is vertical, but the upper edge is thick and posteriorly it becomes more so – so that it forms a somewhat everted and curved ledge that overhangs the ischiadic peduncle; there is no abrupt inflection along the upper margin of the postacetabular process that characterizes the ilium of Mantellisaurus. The dorsal margin of the posterior ilium is elongate and pointed in profile and its ventral surface forms a broad, shallowly vaulted brevis fossa (br.f) bounded laterally by a prominent ridge (Fig. 2.22). The ischiadic peduncle does not exhibit the prominent lateral and stepped expansion typical of all other Wealden iguanodontians. The prepubic process forms an elongate anterior blade that is transversely thick (Fig. 2.22, silhouette), but dorsoventrally narrow along much of its length, before expanding distally; this is distinct from the thinner and deeper blade that is typical of Mantellisaurus (Fig. 2.16). The pubic shaft forms a tapering rod that is much shorter than the shaft of the ischium. The ischium has a shaft that is elongate and stout, with a generally rounded, rather than angular, cross section; the shaft is curved along its length (J-shaped) and ends in a prominent, anteriorly expanded “boot.” This structure is distinct from the narrower, angular-sided, and far straighter ischial shaft with a small boot that characterizes Mantellisaurus.

      In the hindlimb the femur is large and very stout compared to that of Mantellisaurus; the fourth trochanter is very elongate and forms a very thick crested blade along the postero-interior edge of the mid-shaft. The shin or lower half of the limb is similar in overall shape in the two taxa, but the difference in robustness is notable; this is especially so in the massive construction of the pes. The first metatarsal of the foot in I. bernissartensis forms a small, oblique, flattened spatula-shaped splint bone that is distinct from the thin, pencil-like, metatarsal that lies parallel to the shaft of metatarsal II in one articulated example of Mantellisaurus from the Isle of Wight.

      Reconstruction of Iguanodon bernissartensis

      The reconstruction of I. bernissartensis presented here is modified from an often-published version created originally by the artist Gregory Paul (e.g., Brett-Surman, 1997; Fig. 2.23). Changes introduced are as follows: the joint between the forelimb and manus has been straightened; mammal-style scapular rotation against the ribcage has been removed (this was physically impossible judged by the anatomy of the pectoral girdle); and a lattice of ossified tendons has been added. The overall impression of an Iguanodon in a hurry depicted in the earlier reconstruction is potentially misleading in an animal that attained a body length of 12 m and probably weighed in excess of 5 metric tons.

       SYSTEMATIC ANALYSIS

      The brief anatomical survey of these Wealden taxa provides an opportunity to align them, systematically, with other reasonably well-known taxa from younger stratigraphic stages, with the intention of exploring their phylogenetic relationships generally, as well as exploring the origin of euhadrosaurs. Numerous phylogenetic hypotheses have been proposed (e.g., Norman 2002, 2004, in press; Weishampel et al., 2003; Horner et al., 2004; Prieto-Márquez, Gaete, et al., 2006; Dalla Vecchia, 2009; McDonald et al., 2010; Prieto-Márquez, 2010; McDonald, 2012b). These analyses have either directly or indirectly addressed the relationships of those animals that are proximate to the clade most commonly referred to as Hadrosauridae (= Euhadrosauria, Weishampel et al. 1993), but have resulted in little general consensus on the relationships of individual taxa.

      2.22. Iguanodon bernissartensis Boulenger. Holotype pelvis partially restored, RBINS R51. Abbreviations: ac, acetabulum; br.f, brevis fossa; obt, obturator process. Cross-sectional views shown as solid black (after Norman, 1980).

       Data and Methodological Framework

      A new systematic analysis is presented here for the explicit purpose of exploring the nature of putative phylogenetic relationships between a range of more derived non-hadrosaurian ornithopods. Twenty-four taxa (Appendix 2.1) were selected for this analysis because they are known from generally well preserved skulls and/or skeletons, and have been reasonably well described. A significant number of additional taxa have been named in very recent years but these are, on the whole, more fragmentary, and their addition to the analyses materially affects the resolution and stability of tree topology. These have been removed a priori. More comprehensive analyses incorporating these additional taxa, as well as more basal ornithopods, are being considered in more detail elsewhere (Norman, in press). Lesothosaurus (based on the descriptions of Thulborn [1970, 1972], and the supplementary information from Sereno [1991]) is used as the outgroup taxon (Butler et al., 2008) in order to polarize character-states. The character-states and their codings are presented in Appendix 2.2, and these comprise a suite of 92 characters that have been generated after reassessment and revision of previously published character lists (e.g., Norman, 2004; McDonald et al., 2010; Prieto-Márquez, 2010; Wu and Godefroit, 2012). Unusually, compared to the general trend in the literature, the number of characters used is fewer rather than greater. The data matrix was constructed in MacClade 4.06 (Maddison and Maddison, 2003), and the analysis was undertaken using PAUP* 4.0b10 (Swofford, 2002). All characters were equally weighted, and were first analyzed as “unordered” using the Branch and Bound search option (and the analysis was run twice using ACCTRAN and DELTRAN optimizations). Because a substantial number of characters used in this analysis are multistate in nature, a second run was undertaken using the “ordered” character option. The matrix was again analysed using Branch and Bound search option and was run under both ACCTRAN and DELTRAN optimization protocols.

      2.23. Iguanodon bernissartensis. Reconstruction of the skeleton, modified from an earlier original drawing by Gregory S. Paul (from Brett-Surman, 1997).

       Results

      Running the data matrix with characters unordered produced a fairly well resolved topology in the strict consensus tree. Twelve most-parsimonious trees