Metoposaurus gets one more finger

Konietko-Meier et al. 2020 discover digit 5
where they did not expect to find one, on Metoposaurus (Figs. 1-3).

We’ve long wondered, how many fingers did the first tetrapod have? 
If more than five, when did four or five come to be?
If five, when did four or more than five come to be?
If four, when did five or more than five come to be?

From the Konietko-Meier et al. 2020 introduction:
“In contrast to crown tetrapods that rarely have more than five digits, basal tetrapod groups possessed more digits, such as Acanthostega gunnari Jarvik, 1952 which had eight in the forelimb (Coates and Clack, 1990) and Ichthyostega Säve-Söderbergh, 1932 with seven digits in the hindlimb (Säve-Söderbergh, 1932; Jarvik,1996). This fact indicates that polydactyly is the plesiomorphic condition for the tetrapod autopodium (Laurin et al., 2000).”

No. That’s a myth. The large reptile tree (LRT, 1713+ taxa; subset Fig. 4) recovered four fingers in basal tetrapods. Five fingers are derived in several convergent clades. More than five fingers occurs only in Acanthostega and kin, a derived clade without descendants. (Maybe Ichthyostega, too, but we have no hands for it).

From the Konietko-Meier et al. 2020 abstract:
“Temnospondyli are commonly believed to have possessed four digits in the manus
and five in the pes. However, actual finds of articulated autopodia are extremely rare. The most important observation is the presence of five metacarpals in this specimen. This allows reconstructing the manus as pentadactyl.”

From the Konietko-Meier et al. introduction:
“The first known record of a pentadactyl hand belongs to the Early Carboniferous stegocephalian Casineria kiddi (Paton et al., 1999).”

Chronology does not always mirror phylogeny. Casineria nests as an archosauromorph reptile, off the bottom of the chart (Fig. 4). Many more primitive taxa had only five digits. The Late Devonian reptilomorph, Tulerpeton, had only five fingers, as we learned earlier.

Among temnospondyls in the LRT
(Fig. 4) the derived taxa leaving no descendants, Parotosuchus and Paracyclotosaurus, were illustrated with five fingers. Trematosaurus is known from skull material only. These fifth fingers are appearing de novo, not as reversals.

Proterogyrinus
developed five fingers. Fingers are not preserved in related taxa, none of which left descendants.

Dissorophids
developed five fingers without leaving descendants.

Reptilomorpha,
starting with Utegenia + Seymouriamorpha, developed five fingers and we are their descendants.

The Konietko-Meier et al. chart
(their Fig. 4) indicates the outgroup taxon, Greererpeton (Fig. 5; Godfrey 1986, 1989 had five fingers.

This is an error. Only the PhD thesis illustrates fingers and only four are illustrated (Fig. 5). Maybe the five-digit pes was accidentally added to the manus database?

From the Konietko-Meier et al. introduction:
“Reconstruction of the evolution of digit reduction of the most basal and post-Devonian stegocephalians is not possible because of the lack of informative fossils. It is known that reductions in the number of digits have occurred frequently during tetrapod evolution, but it is still not known exactly when or even how many times the number of digits was reduced to five or less (Laurin et al., 2000).”

The LRT clarifies this problem. Reductions in the number of digits occurred less frequently than envisioned by Konietko-Meier et al. since ‘four fingers’ is the primitive and plesiomorphic condition, even in Greererpeton.

References
Godfrey SJ 1989. The postcranial skeletal anatomy of the Carboniferous tetrapod Greererpeton burkemorani Romer, 1969. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 323(1213), 75–133.
Konietzko-Meier D, Teschner EM, Bodzioch A and Sander PM 2020. Pentadactyl manus of the Metoposaurus krasiejowensis from the Late Triassic of Poland, the first record of pentadactyly among Temnospondyli. Journal of Anatomy 00:1–11. DOI: 10.1111/joa.13276