What's the right Dunkleosteus for you?

[Halichoeres]

Dunkleosteus is roughly tied with the thoroughly banal megatooth shark for the most frequently made prehistoric fish. But lots of errors in its reconstruction afflict the well known toy versions. First I'm going to note a few relevant points about the animal's anatomy, and then I'm going to go through some of the more prominent exemplars point by point.

Endoskeleton: we have lots of placoderm braincases (primitively bony but often secondarily cartilaginous) and more spinal columns than you might guess, including a series of cartilaginous but partially mineralized vertebrae from Dunkleosteus, which show that the front half of the body was pretty inflexible. Arthrodires did not have ribs.

Exoskeleton: made of intramembranous bone and a few other mineralized tissues. While there is nothing quite like a placoderm now, dermal bone is widespread in modern fishes. Virtually all bony fishes have some in their face--even you! Your skull roof, maxilla, and dentary are all dermal bone. It's just that in tetrapods it tends to have a lot of soft tissue on it. In fishes, this really varies. The maxilla of an actinopterygian, for example, is usually externally visible, bound to other bones by muscles and membranes, and coated on the outside surface with various substances including keratin or enameloid. The body scales of coelacanths, lungfishes, gars, and bichirs are also just dermal bone with layers of various other mineralized tissues similar to tooth enamel. This is about what placoderm armor would have looked like.

What do the dermal bones look like? You can see some of the seams between them, but not all. As a rule of thumb, if the dermal bones can move with respect to each other, you'll be able to see the separation easily. That includes body scales, opercular elements, jaw elements, that sort of thing. Immobile elements might have invisible sutures; for example, it's usually hard to spot the suture between a fish's nasal and frontal bones. Even if you have it in your hand, you often have to feel for it with a finger. A small amount of soft tissue is necessary to make sensory equipment like lateral line canals and nares work; that, too, can obscure sutures.
For more on histology try Giles et al. 2013: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5176033/

So large, mobile plates such as the suborbital are very likely to be externally visible. Sim and I were recently talking about this in another thread and I believe we were talking past each other. Some artists have reconstructed the plates as entirely buried in soft tissue; there is no reason to believe this. To the contrary, having layers of, say, fat and muscle overlying all the plates would obviate their function. They might have had thin layers of skin, but it wouldn't have been necessary in all areas. What would be very strange is having the unarmored skin be pigmented in a way that highly contrasted with the armor. Modern fishes with heavy dermal armor do not usually offset it visually. For example, you don't see a sharp demarcation in the head and body of a sea robin, or a loricariid catfish, or a bichir.

The fins were very similar to those of sharks, with ceratotrichia (cartilaginous elements arranged in rows) supporting the base of the fin. Like a horn core, the ceratotrichia only gives you the absolute minimum fin area. As in sharks, the extent of the fin would have often been greater. A large triangular dorsal fin on Dunkleosteus is quite likely. The pectoral fins are known and look very sharklike.

What about tail morphology? People are often tempted to use the phylogenetic bracket here, inferring that the few well-preserved arthrodire body outlines, like that of Coccosteus, are a good guide to the morphology of all arthrodires. This is probably a mistake; fin shape and body shape are incredibly labile, meaning that they are very changeable. Developmentally, this makes sense--all you need to do is make minor alterations to the relative amount of time particular cell populations are allowed to grow. Their tails were probably more crescent-shaped, like that of a tuna or pelagic shark. But anyway, what good is a tuna-like tail? Large fishes experience lower effective water viscosity, and therefore greater yaw, than smaller fishes. That becomes untenably costly if you are very large and very sinuous. A stiff trunk and tuna-like tail reduces yaw and also shunts vortices away from the streamline, the region of laminar flow that allows the main body to enjoy low friction. There's a large body of morphometric literature showing that ecology is nearly as strong a predictor of body shape and fin shape as phylogeny, which is why you get crescent shaped tails in lineages as divergent as tunas, billfishes, pikeminnows, lamnid sharks, ichthyosaurs, and mosasaurs. Even the Triassic coelacanth Rebellatrix had a tail that looked a bit like this. Dolphins have it too, but because of the constraint of mammalian vertebral flexion, theirs is horizontal rather than vertical. In tunas, billfishes, pachycormids, etc., there is often also a keeled peduncle, which means that the narrow 'neck' of the tail has a horizontal growth that stabilizes and strengthens the caudal fin at high speeds.

The idea that Dunkleosteus had a mechanically efficient tail isn't new. All the way back in 2001, Kaiyodo--quite reasonably!--gave their Dunk figure the tail of a big pelagic predator, and so did Favorite in 2014. The idea became more popular after a 2017 study that put some numbers on it, finding good support for the tuna-like tail. It remains at least possible that it had some suboptimal tail, but arthrodires had tens of millions of years' worth of chances. To me it seems like special pleading to argue that they're the one big pelagic fish group that had to propel themselves like an eel strung through an oil barrel.


So with all these things in mind, how do the various Dunks hold up? I'm not sure I'm comfortable confidently proclaiming one the most accurate, but I'll try to give the good and bad of each.


CollectA (2018)
The good:
• plates are the same color as the rest of the body, it looks like it has skin on it.
• soft tissue around nares and mouth
• nice wrinkles at the nuchal gap (the joint between the head and thorax)
• the sclerotic ring is inside the eyeball (however, the eyeball probably occupies too much of the socket)
The bad
• the body shape is utter nonsense; it looks like an ice cream cone that bites back. This thing would be a nightmare to move through a fluid medium.
• the fins are too conservative, and the dorsal is probably too posterior
• the osteoderms, evidently based on Gemuendina, an extremely distant relative (think rabbits vs. elephants). Great way to add unnecessary drag and weight. Also, if arthrodires had these, we would have found some by now!
• the skin is more generally overtextured; CollectA isn't immune to the Papo-fication of the hobby, where either fans demand, or companies assume fans demand, detail at the expense of realism
• the tail is just that of Coccosteus, feh.
Verdict: this figure is a mess, inexcusably so given the late date of its release. Their miniature from the previous year had fewer problems, and certainly fewer egregious ones. It's probably not as bad as the Schleich, but I have more ire for it because CollectA should know better.


Favorite (2014)
The good:
• hydrodynamically, the best there is. The tail is a nice efficient shape and even includes a caudal keel.
• smooooooth
• the fins are nicely sculpted and believable, although too conservative in extent
• sclerotic ring inside eyeball
The bad:
• the fins are a little too conservative
• the plates are implausibly visually offset, and also raised relative to the rest of the body. They look more wrapped around the body than integrated into it.
• the plates have too little connective tissue; this is most noticeable at the front of the head where you see every suture in the neighborhood of the nares.
Verdict: really good overall. I think this one is the most thoughtfully executed, even though some errors made it through. I admire both the hydrodynamic features and the resistance to detail-for-detail's-sake.


Kaiyodo (2001) (this photo is not mine, it's by Andre Mursch)
The good:
• sharklike fins
• sharklike tail
• clear indication of gill tissue in the correct location
• sclerotic ring inside eyeball
The bad:
• maybe a bit too much offset for the plates, and they're raised compared to the rest of the body, but it's not as bad as other examples of this sin
• hard to be sure at this scale, but maybe a bit too much in the way of visible seams on the face
Verdict: the OG. I personally prefer my large animals at a larger scale, but this is an excellent figure and probably the first toy of the genus (SRG made a bronze of Dinichthys back when the two genera were considered synonyms)


Mojo (2019)
The good:
• skin, largely a good balance. The ventral thoracic shield should probably be visible, but otherwise this is a pretty believable set of pajamas
• sclerotic rings internal to eyeball, although the eyes are a little too buggy.
• The gills are in the right place, below and behind the inferognathal and suborbital plates
The bad:
• the fins are too conservative, and the dorsal is probably too posterior
• the tail is much too eely
• the body is much too elongate
Verdict: I don't love it, but it isn't bad


Safari (2007)
The good:
• skin, with lips and everything. I also love this color scheme.
• sclerotic rings inside the eyeball.
• the way the suborbital flares as the jaw opens is very good.
The bad:
• the armor is probably too raised relative to the other skin, a hydrodynamic implausibility
• tail is out of date, but the arguments for a more sharklike tail hadn't penetrated the popular consciousness as thoroughly then
• the fins are too conservative and their morphology is a little iffy
Verdict: groundbreaking. I don't think it's the most up-to-date, but it was the first attempt by a major toy company, excepting the 2001 Kaiyodo miniature. Deserves a lot of credit, but maybe also deserves an update.


Schleich (2016)
The good:
• there are some nice sculpt details, like the folds in the body at the curves.
• the tail is halfway there
The bad:
• the tail is only halfway there
• the sclerotic ring is shown as an external structure
• the plates are pretty much disembodied; they do not look like an integrated part of an organism. This is the Night King's pet fish.
• the fins look like bichir dorsal finlets, total trash
• sturgeon scutes. Like the Gemuendina osteoderms on the CollectA version, they make no sense and we'd have evidence for them if they existed
• pebble scales all over. We are not a squamate.
Verdict: meh. I was too kind to it in my review for the blog, because it had been a while since anyone had done a placoderm. But it's since become a ubiquitous image of the genus, to the point that it was featured on a t-shirt for the academic Evolution conference that almost happened in Cleveland last year, and in the news conference declaring D. terrelli the state fossil fish. I think the world is worse for its existence.


ThinkArt (2020)
The good:
• the tail profile is at least trying
• the pectoral and pelvic fins are sort of okay, not great, but okay
• the sclerotic ring is inside the eyeball
The bad:
• the anatomy of the midline fins is completely ridiculous; it looks like an acanthomorph teleost
• the plates are entirely naked and every seam is visible; there is no place for necessary connective or sensory tissues
• it appears to have ribs. Even when fish have ribs (arthrodires didn't, remember), they aren't wraparound ribs like tetrapods need to hold their organs in. There are exceptions, but they tend to be more T-shaped in cross section, serving as muscle attachment points. At any rate, arthrodires didn't have ribs at all. This detail really betrays that the artist was working from paintings rather than from anatomy
• moderately over-textured, although not as bad as the CollectA
Verdict: it has shelf presence, I'll give it that. But it's not a very good representation of the animal.

There are lots of other Dunkleosteus figures by more minor companies, including Bandai, Chap Mei, Cog Ltd, Colorata, DeAgostini, Diramix, Recur, Sonokong, and Takara Tomy. If people want specific critiques of those, let me know. I figured this would cover the ones people are most likely to be deciding between.

Update August 2021: Zaha the PNSO Dunkleosteus:


The good:
• They've done a good job giving it skin. The plates have the same color pattern as the rest of the body, and the small plates at the front of the head have some soft tissue overlying them. They haven't gone too overboard in the texture department, as with their Ophthalmosaurus or CollectA's be-tubercled monster.
• The eyes contain the sclerotic rings, rather than vice versa.
• The tail is biomechanically plausible.
• The fins are generally sharklike, which is good because, like shark fins, placoderm fins were supported by ceratotrichia.

The bad:
• The aspect ratio of the fish is pretty bad. In other words, it's way too sinuous. This is a common malady in PNSO's aquatics. I think they're averse to neutral swimming poses, so in order to make the figure seem more dramatic, they contort it, which requires making some parts of the animal implausibly elongate. The same thing happened with their Eurhinosaurus.
• I don't know the position of Dunk's dorsal fin with certainty (maybe someone does, but if so, I haven't found them saying so). This one seems modeled after Coccosteus, but for a big pelagic animal it would probably have been farther forward. On the plus side, they have given it a plausible height.
• The gill openings are too dorsal. They should be ventro-lateral, behind the lower jaw and perhaps the bottom edge of the suborbital plate. The suborbital probably did not move in the manner depicted.

On the whole, I still think Favorite's version is the most faithful depiction of this animal, but in part that's because overall body shape and biomechanical plausibility are important to me. This one definitely has nicer pajamas, so if you find the harsh offset between the armored and unarmored parts of Favorite's version off-putting, you might prefer the PNSO despite its somewhat eel-y aspect.

[Newt]

Thanks, this is an invaluable roundup and critique.

[Gwangi]

Fantastic writeup, I very much enjoyed it. I feel like it should be on the blog, somewhere where it won't get lost to time. Even though those individual figures have all been covered, and it kind of deviates from what the blog traditionally does.

I personally have the Safari dunk and that's probably sufficient. If I were to get another it would be the Favorite. I do like the Mojo dunk too though, for aesthetic reasons. 

[Bread]

I too enjoyed reading this! I've been looking for a Dunkleosteus recently, just none really have interested me besides Favorites (2014) model, which I can not find anywhere for a reasonable price....

Thank you for the read! Very informative and overall perfect!

[Pachyrhinosaurus]

I have the Paleocraft dunkleosteus, though I haven't finished painting it yet. It's a great sculpt but suffers from many of the same downfalls as those you already mentioned. Had I known about the shark-like fins earlier, I would have made new ones for it.

[stargatedalek]

How does the larger vinyl Favorite one fare?

[Faelrin]

Thanks for doing such a detailed write up on this. If I had to pick just one with all the current options I'd go with the Favorite, but honestly I'm still waiting for another take on this that truly nails it (to the current evidence).

I am curious about how the Takara Tomy one fares since that will likely be the first Dunkleosteus in my collection however, though granted I could already catch issues with it (tail for starters), just from the promo pic.

[Flaffy]

Fantastic write up! Very informative and educational to a fish noob like me

Looking back, I was waaaaaaay too kind on the CollectA Dunkleosteus. I was far too focused on praising the decent execution of the head, that I missed literally the rest of the messed up body.

So far, none of the current mainstream Dunkleosteus appeal to me, though if I had to choose it would definitely be the Favorite Co one. I hope to see a proper updated Dunkleosteus from someone at some point in the future. Given it's status as the "Tyrannosaurus of Devonian fish" (in terms of popularity), I think we'll see a well executed one sooner rather than later. 

[Halichoeres]

Thanks, everyone, I'm glad that this has been useful.

Several people pointed out that the Favorite version is hard to find, which is true in some markets. However, HobbyLink Japan carries Favorite products, and they ship worldwide. While Favorite products always show as out of stock, what that really means is that next time they get a consignment from Favorite, they'll include your request as a special order: https://www.hlj.com/dunkleosteus-soft-model-fav77614

If it wasn't so tiny and weirdly painted the Kaiyodo one.
But for now I am still not interested in any of the current ones as they are.
At most the Favorite one if it was easier to get hold of, I like that it has claspers.

I neglected to mention that, but that is definitely a cool feature! I suppose the absence of claspers on other figures can just be excused as their being female, so I don't use it as a criterion. However, it's a good thing to include if you're trying to be educational/accurate because of the strong evidence that many placoderm-grade fishes had internal fertilization (Materpiscis, e.g.).

Fantastic writeup, I very much enjoyed it. I feel like it should be on the blog, somewhere where it won't get lost to time. Even though those individual figures have all been covered, and it kind of deviates from what the blog traditionally does.

I personally have the Safari dunk and that's probably sufficient. If I were to get another it would be the Favorite. I do like the Mojo dunk too though, for aesthetic reasons. 

Of the ones that are easily available in the States, those are my preferences as well.

I have sometimes thought about pitching other kinds of articles for the blog besides individual figure reviews, like the one where I go through the math of why blind bags are a scam. But of course it takes me long enough to just get to my regular scheduled reviews!

I have the Paleocraft dunkleosteus, though I haven't finished painting it yet. It's a great sculpt but suffers from many of the same downfalls as those you already mentioned. Had I known about the shark-like fins earlier, I would have made new ones for it.

That's a beautiful sculpt, but it's definitely not as accurate as Cooper's dinosaurs. Understandable to some extent; if you mess up a piece of theropod anatomy, you have an immediate avalanche of armchair paleontologists lambasting you for it. Fishes don't really have this kind of peanut gallery.

How does the larger vinyl Favorite one fare?

From what I can see in photos it's more or less identical to the "soft model" apart from size, so all the good and bad points apply with equal force. The plates are perhaps a little less embossed relative to the size of the model, but I'm not 100% sure on that without having it in hand.

I am curious about how the Takara Tomy one fares since that will likely be the first Dunkleosteus in my collection however, though granted I could already catch issues with it (tail for starters), just from the promo pic.

It actually looks quite good apart from the tail. It has more of a nurse shark type tail, if anything, which would be fine if it were more of a benthic species, but that's probably incorrect. The dorsal fin shape is also probably wrong--too long of a base, but also not tall enough.

Fantastic write up! Very informative and educational to a fish noob like me

Looking back, I was waaaaaaay too kind on the CollectA Dunkleosteus. I was far too focused on praising the decent execution of the head, that I missed literally the rest of the messed up body.

So far, none of the current mainstream Dunkleosteus appeal to me, though if I had to choose it would definitely be the Favorite Co one. I hope to see a proper updated Dunkleosteus from someone at some point in the future. Given it's status as the "Tyrannosaurus of Devonian fish" (in terms of popularity), I think we'll see a well executed one sooner rather than later. 

I was also too kind to the CollectA initially, because they really did some good work on the head. It seems like they just forgot to attend to the rest of the body. Like they just rolled out a clay snake and added studs to it.

[Bokisaurus]

Cool overview of the different figures.
Each one of them, as flawed as they are, do represent various take and consensus at that time.
I'm glad they are all different, making each one a unique one, it's one of the reasons I like popular species with lots of different variations on how they are depicted.
Might as well have a variety if one would add more popular species that so overdone in a collection!😃

[SBell]

That was very comprehensive. It gives me post hoc justification for not buying either the CollectA or Mojo...

I am actually curious on your takes for the COG and Colorata models; in the case of the former (given its size and overall appearance) I'm not convinced that it wouldn't stand in better for a different, smaller arthrodire.

[Bread]

@Halichoeres Thank you for that link and information! Going to place an order soon, but first I need to ask: How long is the typical wait for the out of stock figures? I assume shipping may take two weeks or so but how about the warehouses' expected time to have said products?

[BlueKrono]

Aaaall of them!

[Gwangi]

I have sometimes thought about pitching other kinds of articles for the blog besides individual figure reviews, like the one where I go through the math of why blind bags are a scam. But of course it takes me long enough to just get to my regular scheduled reviews!

I think it's a fine idea. So much has already been reviewed on the DTB that it gets harder to find new stuff to write about. In addition I fear the blog has all but been replaced by video reviews. Some new kind of content might give it some more life. But I suppose this isn't the place to discuss it.

[Halichoeres]

I am actually curious on your takes for the COG and Colorata models; in the case of the former (given its size and overall appearance) I'm not convinced that it wouldn't stand in better for a different, smaller arthrodire.

@SBell I don't think the Cog is a good stand-in for anything because, like the small Diramix Dinichthys, the thoracic armor is weirdly truncated. It also has prominent sutures that don't look like anything I've ever seen. Like their "Groenlandaspis," it seems to have been hastily sketched from memory with elements of a variety of taxa.

Edited to add: I reviewed the Colorata one for the blog, but briefly, the paired fins look pretty okay, the plates are a bit too skeletal, the eye has a weird donut shape, the tail is unlikely, and the dorsal is too conservative and squared off. I like that it's all the same color, and I like the ball-and-socket display rod.

Halichoeres Thank you for that link and information! Going to place an order soon, but first I need to ask: How long is the typical wait for the out of stock figures? I assume shipping may take two weeks or so but how about the warehouses' expected time to have said products?

@Bread I'm afraid I don't know the answer to that. In my experience it's been a few weeks, but everything is so topsy-turvy now that I couldn't possibly guess.

Aaaall of them!

Yes, for some collectors, this is the correct answer!

[bmathison1972]

I am content and happy with the Safari version 

[SBell]

I am actually curious on your takes for the COG and Colorata models; in the case of the former (given its size and overall appearance) I'm not convinced that it wouldn't stand in better for a different, smaller arthrodire.

@SBell I don't think the Cog is a good stand-in for anything because, like the small Diramix Dinichthys, the thoracic armor is weirdly truncated. It also has prominent sutures that don't look like anything I've ever seen. Like their "Groenlandaspis," it seems to have been hastily sketched from memory with elements of a variety of taxa.

Edited to add: I reviewed the Colorata one for the blog, but briefly, the paired fins look pretty okay, the plates are a bit too skeletal, the eye has a weird donut shape, the tail is unlikely, and the dorsal is too conservative and squared off. I like that it's all the same color, and I like the ball-and-socket display rod.

Thanks. I've definitely noticed that on the cog placoderms; there's one that would act as a better fit for their supposed groendlandaspis but can't recall now.

On the other hand...at least they made some.

[Bread]

Halichoeres Thank you for that link and information! Going to place an order soon, but first I need to ask: How long is the typical wait for the out of stock figures? I assume shipping may take two weeks or so but how about the warehouses' expected time to have said products?

@Bread I'm afraid I don't know the answer to that. In my experience it's been a few weeks, but everything is so topsy-turvy now that I couldn't possibly guess.

Thank you! I figured it kind of be unknown or uncertain. I won't mind the wait since this Dunkleosteus is worth it, in my opinion.

[Faelrin]

@Halichoeres Thanks for the info about the Takara Tomy one. Good to know it is mostly just the tail and fin that are issues on that one.

[Halichoeres]

I am actually curious on your takes for the COG and Colorata models; in the case of the former (given its size and overall appearance) I'm not convinced that it wouldn't stand in better for a different, smaller arthrodire.

@SBell I don't think the Cog is a good stand-in for anything because, like the small Diramix Dinichthys, the thoracic armor is weirdly truncated. It also has prominent sutures that don't look like anything I've ever seen. Like their "Groenlandaspis," it seems to have been hastily sketched from memory with elements of a variety of taxa.

Edited to add: I reviewed the Colorata one for the blog, but briefly, the paired fins look pretty okay, the plates are a bit too skeletal, the eye has a weird donut shape, the tail is unlikely, and the dorsal is too conservative and squared off. I like that it's all the same color, and I like the ball-and-socket display rod.

Thanks. I've definitely noticed that on the cog placoderms; there's one that would act as a better fit for their supposed groendlandaspis but can't recall now.

On the other hand...at least they made some.

Yeah, that's one's really weird. It's like a petalichthyid head on an antiarch body.

I forgot an important figure in my summary before, and that's the PaleoZoo version. It's not a toy exactly, but a model 3d printed in gypsum.



I actually can't find immediate fault with this model, but of course I'm not going to spot everything. ? @Stolpergeist I remember you said it had shortcomings, do you recall what they are?