Haplocanthosaurus, Hatcher, 1903

Whitlock, John A., 2011, A phylogenetic analysis of Diplodocoidea (Saurischia: Sauropoda), Zoological Journal of the Linnean Society 161 (4), pp. 872-915 : 884-886

HAPLOCANTHOSAURUS

Haplocanthosaurus is recovered here as the basalmost diplodocoid, a placement supported by three synapomorphies: dorsal neural spines with prespinal lamina formed by conjoined spinoprezygapophyseal laminae (character 107); cervical ribs shorter than centrum (152); posterolaterally facing fibular facet of the astragalus (186). This is congruent with the findings of Calvo & Salgado (1995; N.B. Haplocanthosaurus was polyphyletic in their analysis), Wilson (2002), Salgado et al. (2004), Remes (2006), Salgado et al. (2006), and Lovelace et al. (2008). In contrast, Upchurch (1998), Rauhut et al. (2005), and Harris (2006c) recovered Haplocanthosaurus outside Neosauropoda. A re-analysis of Upchurch (1998) by Upchurch & Martin (2002) was unable to resolve a single position for Haplocanthosaurus relative to Neosauropoda, recovering it outside the Neosauropoda, outside Macronaria, and outside Titanosauriformes. Re-scoring the matrix of Rauhut et al. (2005) for Amazonsaurus (see ‘ Amazonsaurus ’, below) reduces much of Eusauropoda to a polytomy, including Haplocanthosaurus , Rebbachisauridae , and Flagellicaudata. In the analysis of Harris (2006c), Haplocanthosaurus is excluded from Neosauropoda on the basis of one character, morphology of the lateral pneumatic cavity in cervical vertebrae; all other synapomorphies of Neosauropoda in that analysis were unscorable in Haplocanthosaurus . This character, with three derived states, has a low CI (0.30) and the most number of steps (10) found in that author’s analysis, and Haplocanthosaurus itself is scored as having multiple states (0/1), including the state that appears to be primitive for Neosauropoda (1).

Upchurch (1995), Wilson & Sereno (1998), Casanovas, Santafé & Sanz (2001), and Gallina & Apesteguía (2005) recovered Haplocanthosaurus as a macronarian. Both Upchurch (1995) and Wilson & Sereno

MDD, more derived diplodocoids.

(1998) were superseded by more recent analyses ( Upchurch, 1998; Wilson, 2002; Upchurch et al., 2004) that are discussed elsewhere in this section. Neither Upchurch (1995) nor Wilson & Sereno (1998) were able to reject the alternative hypothesis that Haplocanthosaurus was a basal diplodocoid (P> 0.10). In the analysis of Casanovas et al. (2001), the macronarian position of Haplocanthosaurus was supported by two synapomorphies: pleurocoels in dorsal vertebrae deep; and pleurocoels present in sacral vertebrae. Both derived states are also present in Diplodocus , however. The small sample of diplodocids included in that analysis – only Barosaurus , Dicraeosaurus , and Diplodocus – may have contributed to that result. Adding Apatosaurus to their matrix produces two equally parsimonious trees. Strict, semi-strict, and 50% consensus trees reduce Neosauropoda to a polytomy amongst Flagellicaudata, Brachiosaurus , Camarasaurus , Haplocanthosaurus , and Losillasaurus ; Apatosaurus is recovered in a polytomy with Dicraeosaurus and the clade Barosaurus + Diplodocus .

Gallina & Apesteguía (2005) recovered a polyphyletic Haplocanthosaurus , with H. priscus and H. delfsi members of distinct lineages, as in the analysis of Calvo & Salgado (1995). Unlike the results of Calvo & Salgado (1995), however, Haplocanthosaurus was recovered within Macronaria. Haplocanthosaurus delfsi was recovered as the more derived of the two, with H. priscus basal to the clade H. delfsi + Camarasaurus . The placement of both Haplocanthosaurus species in Macronaria is supported by two synapomorphies in Gallina & Apesteguía’s (2005) analysis: posteriorly curved anterior caudal neural spines; and anterodorsally orientated prezygapophyses in mid-caudal vertebrae. Constraining both Haplocanthosaurus taxa to a position equivalent with this analysis requires five additional steps; a topology with Haplocanthosaurus as a diplodocoid is rejected by a Templeton test (P <0.01) for Gallina & Apesteguía’s (2005) dataset.

Wilson (2002) and Upchurch et al. (2004), the two most recent comprehensive studies of sauropod relationships, disagree on the position of Haplocanthosaurus . Wilson (2002) recovered Haplocanthosaurus as a basal diplodocoid, a position supported only by the presence of short cervical ribs. Character support for this node in Wilson’s analysis is also low (decay index = 1). Conversely, Upchurch et al. (2004) recovered Haplocanthosaurus as a macronarian, sister taxon to the clade Brachiosaurus + MDT. Constraining Haplocanthosaurus to be a member of Diplodocoidea in that analysis requires five additional steps, although a Templeton test reveals that the data cannot reject this alternative hypothesis (P> 0.10). The analyses of Wilson (2002) and Upchurch et al. (2004) also disagree in other content of Diplodocoidea. Upchurch et al. (2004) recovered Nemegtosaurus and Quaesitosaurus as basal diplodocoids; these taxa are considered titanosaurs by Wilson (2002, 2005b) and other authors ( Curry Rogers & Forster, 2001; Apesteguía, 2004; Curry Rogers, 2005).

Constraining the position of Haplocanthosaurus outside Neosauropoda, to a node either above or below Jobaria , in the present analysis requires only a single additional step; all other relationships outside Neosauropoda require two or more additional steps. Placing Haplocanthosaurus at the base of Macronaria also requires only a single additional step, although moving it to a position more derived than Camarasaurus requires at least seven additional steps. This low resistance to re-arrangements may be the result of critical anatomical data that are absent in Haplocanthosaurus , such as cranial characters. Neosauropoda is supported by five synapomorphies, only a single one of which can be identified in Haplocanthosaurus : anterior chevrons without ‘crus’ (character 155). Further confounding the issue with that character is the reversal of that state within Flagellicaudata and the absence of information regarding its state in Amphicoelias and basal rebbachisaurids. The elements from which Haplocanthosaurus is known are also quite plesiomorphic, making it difficult to assign to a particular group because it necessarily shares so few synapomorphies with any other clade. It is not improbable that the description of more complete Haplocanthosaurus remains ( Bilbey, Hall & Hall, 2000) could result in that taxon moving to any number of positions around the base of Neosauropoda.

Removing Haplocanthosaurus from this analysis results in three equally parsimonious trees of length 264. The topology is unaltered from that recovered from the unpruned matrix.

Excluding the referred hindlimb elements (CM 2043 and USNM 4275) from the scored material does not change the topology or the number of recovered trees, but it does negatively impact the robustness of the node Diplodocoidea. One of the three recovered synapomorphies of the group, posterolaterally facing tibial articular facet of the astragalus (186), is recovered instead as a synapomorphy of the more restrictive group Flagellicaudata + Rebbachisauridae when this material is excluded.