*Gansus*

[David Marjanovic <david.marjanovic@gmx.at>]

What chance is there that *Gansus* actually belongs into Hesperornithes?

As explained in their supplementary information, You et al. (2006) found 24 
MPTs, in which *G.* was either outside of (Hesperornithes + Neornithes) or 
closer to Neornithes than *Ichthyornis* or (as in the majority-rule 
consensus tree in the paper itself) closer to (Neornithes + *Ichthyornis*) 
than to Hesperornithes.

Neither *Enaliornis* nor *Pasquiaornis* (nor, unsurprisingly, *Potamornis*) 
were included; the only representatives of Hesperornithes are *Baptornis* 
and *Hesperornis*. The supp. inf. states that *E.* was excluded because it 
shows hesperornithean autapomorphies and would therefore predictably have 
come out as the sister-group to *Baptornis* + *Hesperornis* -- but I can't 
see why that's an argument. It would after all have broken up the very long 
branch that leads to (*B.* + *H.*); I wonder if we are seeing long-branch 
attraction of (*B.* + *H.*). away from *G.*. The occurrence of MPTs where 
*G.* is outside (Hesperornithes + Neornithes) shows that the phylogenetic 
position of (*B.* + *H.*) is unstable.

According to the supp. inf., *Gansus* shares the following character states 
with *Ichthyornis* and Neornithes but not (*B.* + *H.*):

- "three sacral vertebrae with dorsally-directed costolateral eminences" 
(...sacral ribs + transverse processes?)
- "lateral process of coracoid present"
- "medial surface of coracoid in vicinity of N. supracoracoideus foramen 
flat to convex"
- "ischium straight, with no dorsal process"

I wonder: could the first and the last character states be related to 
foot-propelled diving, and could the other two be flight-related and 
therefore reversed in (*B.* + *H.*)?

Fig. S1C is intriguing. That cnemial crest is enormous.