Placoderm double penis: not the primitive trait they think it is

Summary, for those in a hurry:
Contra the title of this Long et al. 2014 paper, the presence of bony claspers (shark-like dual lateral penises) does not signal the origin of gnathostome internal fertilization on this tiny, jawless placoderm, Microbrachius (Fig. 1). Contra Long et al., internal fertilization never gave rise to external fertilization in gnathostomes. Long et al. added Microbrachius to a cladogram invalidated by the exclusion of extant taxa (Fig. 2).

Massive taxon exclusion
in Long et al. 2014 recovered an incomplete and invalid cladogram (Fig. 2) that mistakenly (yet traditionally) nested placoderms with osteostracans and spiny sharks with sharks by excluding all extant taxa and several pertinent extinct taxa.

By contrast
the large reptile tree (LRT, 1744+ taxa, subset Fig. 3), minimizes taxon exclusion and separates all four clades. In both cladograms (Figs. 2–4) claspers and internal fertilization are phylogenetically separated and thus arose at least twice by convergence in sharks and placoderms.

From the Long et al. 2014 abstract
“Reproduction in jawed vertebrates (gnathostomes) involves either external or internal fertilization¹. It is commonly argued that internal fertilization can evolve from external, but not the reverse. Male copulatory claspers are present in certain placoderms, fossil jawed vertebrates retrieved as a paraphyletic segment of the gnathostome stem group in recent studies.”

This is incorrect, according to the LRT (subset Fig. 3). A valid phylogenetic context is missing in Long et al. 2014 (Fig. 2) largely due to excluding extant taxa. In Long et al. 2014 Microbrachius is a basal placoderm, basal to sharks and bony fish. By contrast, in the LRT Microbrachius is a derived taxon, a sister to Bothriolepis, leaving no descendants.

The Long et al. 2014 abstract continues:
“This suggests that internal fertilization could be primitive for gnathostomes, but such a conclusion depends on demonstrating that copulation was not just a specialized feature of certain placoderm subgroups.”

This is incorrect. According to the LRT sturgeons (Pseudoscaphirhynchus) are pre-gnathostomes and they produce millions of eggs through external fertilization. Paddlefish (Polyodon) are basal gnathostomes and they produce hundreds of thousands of eggs through external fertilization. Between them the Chondrosteus clade includes whale sharks (Rhincodon) and manta rays (Manta) and those two produce only a few young with live birth after internal fertilization. In the same clade, Loganellia (Early Silurian) is the earliest known fossil gnathostome and its reproductive strategy remains unknown at present.

The Long et al. 2014 abstract continues:
“The reproductive biology of antiarchs, consistently identified as the least crownward placoderms and thus of great interest in this context, has until now remained unknown.”

The addition of fish to the LRT postdates the publication of Long et al. 2015, so the authors did not realize their cladogram became erroneous by excluding extant taxa. After adding taxa (in the LRT) antiarchs are bottom-dwelling placoderms that reduce or lack mandibles as a derived, not a primitive condition.

The Long et al. 2014 abstract continues:
“Here we show that certain antiarchs [placoderms] possessed dermal claspers in the males, while females bore paired dermal plates inferred to have facilitated copulation. These structures are not associated with pelvic fins. The clasper morphology resembles that of ptyctodonts, a more crownward placoderm group, suggesting that all placoderm claspers are homologous and that internal fertilization characterized all placoderms.”

In the LRT ptyctodonts (like Campbelllodus) are closer to catfish (like Clarias) rather than the rest of the traditional placoderms.

In their efforts to simply illustrate their hypothesis,
the cladogram diagram in Long et al. 2014 (Fig. 4) cherry-picked taxa and “Pulled a Larry Martin” by focusing on copulatory organs to the exclusion of a suite of other traits. Excluding all extant taxa is the main problem with this paper. The LRT includes extant taxa and falsifies the cladograms of Long et al. 2014 (Figs. 2, 4) restricted to a select list of extinct taxa, that was the traditional list until the LRT expanded to include fish.

The Long et al. 2014 abstract continues:
“This implies that external fertilization and spawning, which characterize most extant aquatic gnathostomes, must be derived from internal fertilization, even though this transformation has been thought implausible.”

No. That conclusion arises from a phylogeny invalidated by taxon exclusion. In the LRT, by minimizing taxon exclusion the transformation from internal to external fertilization is not documented. External fertilization is always more primitive. Internal fertilization occurred by convergence several times in vertebrate phylogeny.

Just so you know…

• Sea lampreys produce 30 thousand to 100 thousand eggs
• Sturgeons produce 100 thousand to 3 million eggs
• Manta rays give birth to 1.5m babies
• Whale sharks give birth to .5m babies
• Paddlefish produce 70 thousand to 300 thousand eggs
• Moray eels produce 10 thousand eggs
• Catfish produce 4 thousand to 100 thousand eggs
• Coelacanths give birth to 5 babies
• Basal tetrapods, like frogs and salamanders, produce dozens of eggs

References
Long J Mark-Kurik E, Johanson Z et al. 2014. Copulation in antiarch placoderms and the origin of gnathostome internal fertilization. Nature 517, 196–199 (2015). https://doi.org/10.1038/nature13825

wiki/Incisoscutum
wiki/Microbrachius