Kenichthys is supposed to be a basal tetrapodomorph

Zhu and Ahlberg 2004 presented a hypothesis of relationships that Kenichthys campbelli (Chang and Zhu 1993, Fig. 1) represented a transitional taxon between two purported ‘sarcopterygians’, Youngolepis and Eusthenopteron.

From their abstract:
“The choana, a unique ‘internal nostril’ opening from the nasal sac into the roof of the mouth, is a key part of the tetrapod (land vertebrate) respiratory system. There is no consensus on the origin of the choana despite decades of heated debate. Here we present new material of Kenichthys, a 395-million-year-old fossil fish from China, that provides direct evidence for the origin of the choana and establishes its homology: it is indeed a displaced posterior external nostril that, during a brief transitional stage illustrated by Kenichthys, separated the maxilla from the premaxilla.”

From their caption:
“Figure 1. Nostril positions on the heads of sarcopterygian fishes.”

Unfortunately,
in the Zhu and Ahlberg cladogram (Fig. 2) Kenichthys nests many nodes apart from Eusthenopteron. Compare to figure 3. Taxa at the bases of clades are more like outgroup taxa than others are. So care must be taken that all candidate sisters are included to test relationships.

Unfortunately
in the large reptile tree (LRT, 1482 taxa, subset Fig. 3) Kenichthys does not nest with either Youngolepis or Eusthenopteron. But Kenichthys does nest at the base of all taxa with an internal choana, including lungfish.

Note most of the ray-fin fish
are derived from one branch of lobefins (Fig. 3), and they reverted to a morphology in which both the in and out nares were on the side of the face, beginning with the lobefins Onychodus and Strunius.

Even so,
there are several air-breathing ray-fin taxa in this cladogram (Fig. 3) without choana.

Now the LRT needs to be confirmed
or refuted with a similar taxon list.

Apparently taxon exclusion
stung the Zhu and Ahlberg report. They assumed Kenichthys was a sarcopterygian and they did not test a long list of other taxa to make sure they were not confirming their own biases. They excluded the basal placoderm, Coccosteus (Fig. 4), which has a similar lateral nasal bone that extends nearly to the jaw line. Was there a tentative external naris present? A personal examination of the fossil with this possibility in mind is needed. If not present, then Kenichthys ‘invented’ the choana and thereafter it was free to move inside or outside the mouth. Contra the original hypothesis, according to the LRT Kenichthys does not represent a taxon transitional between an inner and outer choana.

In the Zhu and Ahlberg illustration
(Fig. 1) note the lack of vomer fangs and a distinctly different set of skull sutures in Kenichthys, which does not nest with Eusthenopteron in the LRT, but at a much more basal node, closer to coelacanths. The outgroup basal placoderm, Coccosteus, shares the trait of a large orbit (within an orbit’s length of the rostrum tip), distinct from competing sisters and outgroups, plus a long list of other traits.

Entelognathus is a more primitive placoderm
(Fig. 4) without the derived infolding of the maxilla and enlargement of the cranial shield seen in derived Coccosteus (Fig. 3).

Kenicthys campbelli (Chang and Zhu 1993; Zhu and Ahlberg 2004; Early Devonian, 395 mya) was originally considered a basal sarcopterygian demonstrating the migration of the exit nostril to the jaw rim on its way to the palate to become the choana. Here Kenichthys nests basal to the coelocanth clade. Yesterday we looked at a sister taxon, Stensioella.

References
Chang M and Zhu M 1993. A new Middle Devonian osteolepidid from Qujing, Yunnan. Mem. Assoc. Australas. Palaeontol. 15 183-198.
Zhu M and Ahlberg P 2004. The origin of the internal nostril of tetrapods. Nature 432:94-97.

wiki/Kenichthys