Reconstruction of Ichthyovenator laosensis returning from an afternoon swim in the Grès supérieurs palaeoenvironment Credit: Own work

The true swimming dinosaurs

Compilation of general spinosaurid (A), spinosaurine (B), and spinosaurin (C) anatomical traits linked with piscivorous and semiaquatic lifestyles (not to scale); see list below for key.  Credits = A1 & A2: Etemenanki3 (CC BY-SA 4.0); A3 & A4: Andrew Cuff, Emily Rayfield (CC BY 2.5); A5, B1 & B2: Marcos Sales & Cesar Schultz (CC BY 4.0); A6 & A7: Own work; A8: Serjoscha Evers​, Oliver Rauhut, Angela Milner, Bradley McFeeters & Ronan Allain (CC BY 4.0); A9: Ghedoghedo (CC BY-SA 4.0); A10: Own work (based on Casanovas et al. 1993; B3: Own work; B4: Modified from Marco Auditore (CC BY 4.0); B5: Own work (based on Aureliano et al. 2018)

• A1. Long, low and shallow skulls with spoon-shaped, notched snouts extremely reminiscent to those of fish-eating crocodiles (Hone and Holtz, 2017).
• A2. Eyes placed far back and up on the head to see above the waterline (Hone and Holtz, 2017).
• A3. Retracted nostrils to prevent water entering them whilst submerged or hunting (Hone and Holtz, 2017).
• A4. A crocodile-like sensory system on the snout (Dal Sasso et al., 2009; Foffa et al., 2013).
• A5. Extensive secondary palates to stiffen their skulls from bending or twisting whilst feeding (Rayfield et al., 2007).
• A6. Conical, slightly recurved to straight teeth with reduced or absent serrations, a type of dentition well-adapted for impaling and gripping prey (such as slippery fish) (Hone and Holtz, 2017).
• A7. Similar oxygen isotope ratios of their phosphatic remains (e.g. teeth) to those of modern aquatic animals (Romain et al., 2011).
• A8. Powerful, muscular necks for lashing out at small prey, quickly withdrawing the head from water and shaking it in an up-and-down motion (Sues et al., 2002; Evers et al., 2015).
• A9. & A10. Four functional toes, with large flat claws for walking on soft substrate; and possibly webbed feet (Casanovas et al., 1993; Ibrahim et al., 2014).

• B1. Shrunken antorbital fenestrae for an even stiffer skull (Rayfield et al., 2007).
• B2. Eyes angled upwards to more easily see above the waterline whilst submerged (Arden et al., 2019; though see Comment on Arden et al., 2019 by Hone and Holtz).
• B3. Reduced hip bones and proportionately short legs (Ibrahim et al., 2014; Arden et al., 2019).
• B4. Osteosclerosis (dense bone tissue) of the skeleton for buoyancy control, making the body heavier and thus easier to submerge (Ibrahim et al., 2014; Aureliano et al., 2018).

And in Spinosaurini:

• C1. Long, shallow trunks similar to those of crocodilians (Ibrahim et al., 2014; Arden et al., 2019).

Furthermore, a study into the neuroanatomy of Irritator challengeri showed features of the brain consistent with those in predators that hunt small, agile prey such as fish (Schade et al., 2020). Much of our spinosaurid material has been recovered from geological formations representing wet and fluvial environments, often with a plethora of aquatic taxa providing an extensive source of potential prey for spinosaurs (Aureliano et al., 2018). 

Direct fossil evidence for a piscivorous diet in spinosaurids is also known: examples include partially-digested fish scales discovered in the stomach cavity of a Baryonyx walkeri skeleton (Charig and Milner, 1997), and a sclerorynchid fish denticle found embedded in the fossil snout of a spinosaurine (possibly Spinosaurus) (Dal Sasso et al., 2005). Note however that a more generalistic/opportunistic diet for some taxa has also been proposed, since spinosaurids are also known to have fed on pterosaurs and other dinosaurs (Hone and Holtz, 2017). 

And now that we have some context, we arrive at the main topic of this post. In addition to those already mentioned, there's yet another aquatic adaptation that, though having remainined decently plausible for quite a while, had no strong fossil evidence in favour of it until recently, and that is: tail fins/deep tails for swimming.

What's in a tail?

Restored skeleton of Spinosaurus aegyptiacus based on the 2014 reconstruction Credit: modified from Marco Auditore (CC BY 4.0)

Tail-propelled swimming in spinosaurids was first proposed in 2014 with the announcement of the Spinosaurus neotype* specimen (FSAC-KK 11888), which revealed the animal's low, elongated trunk; reduced hip girdle, short hindlimbs, and pelican-like feet. In this publication, Nizar Ibrahim and colleagues posited that Spinosaurus may have employed tail-assisted swimming similar to that of crocodilians. They based this on the morphology of its tail vertebrae, which had shrunken articulating processes—preventing the caudals from significantly overlapping each other—and short vertebral centra. Both traits made the tail more flexible and thus easier to oscillate from side-to-side for swimming, in contrast to the stiffer tails of most other tetanuran theropods^. The authors also noted that this would explain the proportionately tiny legs, which suggested that Spinosaurus relied more on propelling itself via undulating axial body movements rather than stroking with its limbs (Ibrahim et al., 2014).

*Though this skeleton almost certainly represents Spinosaurus, its assignment as a neotype for S. aegyptiacus has been questioned.
^Note however that stiffer tails don't mean other theropods weren't capable of swimming, they were simply less specialised for it and relied moreso on other methods such as their limbs, as can be seen in fossil track sites (Milner and Lockley, 2006).

Nile crocodile shown swimming via lateral tail and body movements, in a manner similarly hypothesised for Spinosaurus. Credit: MathKnight & Zachi Evenor (CC BY 4.0)

In 2016, this was further explored by Jan Gimsa and colleagues, who hypothesised that Spinosaurus's dorsal sail may have potentially been used to herd fish before stunning, wounding or killing them by striking with the long and flexible tail. Basically, a combination of hunting methods used by modern sailfish and thresher sharks (Gimsa et al., 2016). Though this is of course very speculative and it's difficult to determine if the tail was actually used like this (even if Spinosaurus is proven to have been anatomically and hydrodynamically capable of doing so), it's certainly an intriguing possibility which will hopefully be more thoroughly investigated in future research.

Footage of a pelagic thresher shark (Alopias pelagicus) employing tail whips while hunting sardines 
Credit: Simon Oliver, John Turner, Klemens Gann, Medel Silvosa, Tim Jackson (CC BY 2.5)

Given the extensive indications of a semiaquatic lifestyle for members of this clade, something that had become a growing trend in spinosaurid palaeoart (especially during the late 2010s after the Spinosaurus neotype publication) was depicting these animals with expanded tails, either in the form of fleshy mosasaur-esque tail flukes or crocodile-like osteoderms and scales. The latter had even been speculated in a diagram by Gimsa and colleagues, as a way to deepen the tail for greater water displacement and swimming efficiency (Gimsa et al., 2016).

Spinosaurus illustrated with a hypothetical mosasaur-like tail fluke - note that the quadrupedal gait is now outdated Credit: Fred Wierum (CC BY-NC-ND 3.0)

Though highly speculative at the time, this did seem like a logical "next step" in spinosaurid evolution. For one, it would be more advantageous in swimming than the shallower, more cylindrical tails of conventional theropods, and would also better explain why Spinosaurus's hindlimbs became so reduced. Expanded and/or laterally flattened tails can be observed in numerous living semiaquatic and aquatic tetrapods, such as crocodilians (Rooney, 2018), newts (Raxworthy, 2015), sailfin lizards (Moosmueller and Wilhelm, 2019), Asian water dragons, and iguanas (Bedford and Christian, 1996).

Extant semiaquatic and aquatic reptiles and one amphibian with fin or paddle-like tails. Clockwise from top left: Indian gharial (Gavialis gangeticus), with a deep tail bearing tall osteoderms; central newt (Notophthalmus viridescens louisianensis), with a fleshy blade-like tail paddle; Asian water dragon (Physignathus cocincinus), with a tail heightened by a crest of skin and fat in adult males; Philippine sailfin lizard (Hydrosaurus pustulatus), with elongated neural spines on the tail connected by a skin membrane.  Credits (in same order): Goodfriend19 (CC BY-SA 4.0), Fredlyfish4 (CC BY-SA 4.0), Rushenb (CC BY-SA 3.0), Elliot Brown (CC BY 2.0)

In contrast to most mammals—which mainly use their limbs for locomotion, and thus show strong muscle attachment sites in their limb girdles—reptiles' locomotion is more reliant on the axial body region (including the head, neck, trunk and tail) especially in semiaquatic species (Rooney, 2018). This is due in part to the importance that tail muscles such as the caudofemoralis play in the movement of reptiles (Dikes et al., 2012). In lepidosaurs and crocodilians for example, the tail is usually the primary method of propulsion through water. Thus, many species have them flattened sideways or vertically expanded for greater efficiency (Rooney, 2018).

So the possibility of this anatomical trait occurring in spinosaurids in some shape or form was certainly not unreasonable. Yet it all remained speculative until recently. In fact, we had confirmed evidence for it at least as early as 2012, thanks to an unusual species from Southeast Asia.

The fish hunter of Laos

My old skeletal reconstruction of Ichthyovenator laosensis, showing the original holotype elements in place Credit: Own work

Discovered in the Grès supérieurs Formation of Laos, Ichthyovenator laosensis was a large spinosaurid most notable for the sail on its back, which unlike in other spinosaurs had a sinusoidal shape that dipped down over the hips into two separate sails*. The species was initially known from a mostly intact pelvis, one rib, and a partial spinal column from the 12th dorsal to the 2nd caudal vertebra. Ichthyovenator was initially assigned by Ronan Allain and colleagues to the Baryonychinae subfamily in 2012, as a closer relative to Baryonyx than Spinosaurus (Allain et al., 2012).

*The "split sail" in Ichthyovenator—though definitively present in the holotype individual and not the result of erosion or taphonomic breakage—may represent a pathology, unfortunately however there remains to be published research reviewing this possibility.

My old life reconstruction of Ichthyovenator laosensis, largely based on baryonychines Credit: Own work

This classification would later change in 2014, when in a conference proceeding, Allain announced that additional remains from the original holotype had been excavated in 2012, including teeth, the complete cervical series, more ribs, another dorsal, a pubis, and seven more caudal vertebrae. These as-of-yet undescribed remains resulted in Ichthyovenator being reclassified as a spinosaurine in Allain's abstract, due to its unserrated teeth (a distinguishing trait in spinosaurines) and neck vertebrae reminiscent to those of Sigilmassasaurus (Allain, 2014: 74:78). In 2019, Thomas Arden and colleagues also placed it as a spinosaurine (Arden et al., 2019). Aside from the remarkably complete neck, the new Ichthyovenator fossils are important in that they include additional tail vertebrae with remarkably elongated neural spines. And so you can probably tell where this is going.

Some caudal vertebrae of Ichthyovenator laosensis, the leftmost four are additional material from the holotype, the rightmost two are from the original discovery. Credit: modified from ★Kumiko★ (CC BY-SA 2.0)

Earlier this year, I had begun expanding Ichthyovenator's Wikipedia page for Featured Article candidacy and part of this project involved updating old images, including my baryonychine-esque skeletal and reconstruction. But one other thing that prompted these revisions was a new photograph uploaded to Wikimedia.

Vertebrae casts from the Ichthyovenator laosensis holotype at the MNHN in Paris - mind the outdated baryonychine proportions of the skeletal diagram Credit: Ghedo (CC BY-SA 4.0)

Skeletal diagram of Ichthyovenator laosensis, with the original elements of the holotype (MDS BK10) in white, and new material from the same specimen in red. References & citations for this diagram can be found here. Credit: Own work

Unveiling of the Triton theropod

Life reconstruction of Spinosaurus aegyptiacus based on current knowledge Credit: ДиБгд (CC BY-SA 4.0)

And wouldn't you know it, the very next year after Arden and colleagues' paper, we finally got confirmation of this adaptation in Spinosaurus itself. In April 2020, Ibrahim and colleagues published an almost complete, partially articulated tail belonging to the neotype specimen. This incredible find revealed that Spinosaurus's tail had significantly elongated neural spines and chevrons, deepening the tail even more dramatically than in Ichthyovenator, forming a huge paddle-like structure most akin to that of newt tails. The articulating processes were also found to not only have shrunk towards the back half of the tail, but almost fully disappeared, making the tail extremely flexible (Ibrahim et al., 2020). The unique anatomy and proportions made all the more sense now, and this publication was the proverbial final nail in the coffin for the theory that Spinosaurus was semiaquatic.

In the paper, the swimming efficiency of Spinosaurus's tail was determined by comparing it to those of more terrestrial theropods (Coelophysis bauri and Allosaurus fragillis) and two modern aquatic tetrapods (the Nile crocodile and Danube crested newt), using cutout plastic models waved from side to side by a robotic arm in a simulated water current (Ibrahim et al., 2020).

Four species whose tail propulsion capabilities were compared to Spinosaurus by Ibrahim and colleagues. Clockwise from top left: Allosaurus fragillis, Nile crocodile (Crocodylus niloticus), Danube crested newt (Triturus dobrogicus), and Coelophysis bauri. Credits (in same order): Fred Wierum (CC BY-SA 4.0), Harvey Barrison (CC BY-SA 2.0), Zsoldos Márton (CC BY-SA 3.0), Petrified Forest (CC BY 2.0)

The newt and crocodile tails achieved greater thrust and efficiency (respectively) than Spinosaurus; but, in comparison to other theropods, Spinosaurus's tail shape was found to produce 8 times as much thrust, and be 2.6 times more efficient. The authors also suggested that the deep tail may have also have helped stabilize the animal and prevented it from rolling onto its side whilst floating (Ibrahim et al., 2020), similarly to a proposed use for the dorsal sail a few years prior (Gimsa et al., 2016). This would also answer a problematic aspect with the swimming Spinosaurus model that had been pointed out by Donald Henderson (Henderson et al., 2019).

Spinosaurus aegyptiacus depicted hunting a giant lungfish Credit: Denny Navarra (CC BY-SA 4.0)

Ibrahim and his team therein considered Spinosaurus a pursuit predator that would have used its tail to actively hunt underwater (Ibrahim et al., 2020). This particular hypothesis has received some criticism however, and Mark Witton has pointed out that the tall, heavy sail wasn't exactly hydrodynamic; also lending difficulty to any scenario of it being used for herding fish. This leads us to an important point: there is little doubt that Spinosaurus swam with its tail, but to what extent—i.e. did it use it to pursue prey, as a way to travel from shore to shore, or both?—and exactly how it did so will remain unclear until more extensive and detailed research is done.

The plastic tail model experiment is somewhat preliminary in nature; though it did clear up that Spinosaurus's tail shape was significantly more efficient for swimming than in other known theropods, there's still much that it can't answer. Personally, I'd love to see a future biomechanical study with a more fleshed out, robotic tail reconstruction at a closer scale to that of the real animal, to more accurately determine its capabilities and hydrodynamics.

And of course, just because it used its tail to swim, this does not rule out other possible additional uses. In extant lepidosaurs such as those shown earlier in this post, other functions have also been observed for such tails, including climbing, balance, defense against predators, and as a courtship display organ (FONZ; Raxworthy, 2015). Though I doubt spinosaurids were pulling a Hypuronector and climbing trees, the possibility of additional functions for these tails should not be dismissed. These may include balance, as a weapon against fish/small prey, as a sexual or threat display structure, or even (as Andrea Cau has suggested) as a prop to assume a tripodal posture. All of these deserve more rigorous analyses however, so I'll leave it to actual palaeontologists to do that work.

Preliminary comparison of selected overlapping tail vertebrae in Spinosaurus and Ichthyovenator; most vertebrae are in side view, whilst the first caudal is in top view. Note that there are two additional vertebrae between C16 and C23 in Ichthyovenator in comparison to the MNHN diagram, since the tail seemed to me as overall too short there compared to other known spinosaurids. Credit: Own work

So from what we know about both animals' vertebrae so far, Ichthyovenator seems to have had a shallower and less extensive tail fin than that of Spinosaurus, which would make sense since it lived earlier and was less derived. The articulating processes in Ichthyovenator's caudal vertebrae also appear to have been similarly reduced from what I can tell from images, suggesting a flexible tail, and the neural spines and chevrons are also noticeably more gracile in Spinosaurus than Ichthyovenator.

These fossils show that spinosaurids within at least the Spinosaurinae—the more semiaquatic of the two subfamilies—evolved deeper tails that provided them with an effective way of propelling themselves underwater, by swishing them from side to side as modern crocodilians and newts do (Arden et al., 2019; Ibrahim et al., 2020). It also appears that this deepening of the tail may have been correlated with reduction of the hindlimbs, known in Spinosaurus, and inferred from the reduced pelvic girdle in Ichthyovenator. This would make sense if spinosaurs were transitioning from using mainly appendicular locomotion—by stroking with their legs as in typical theropods—to axial locomotion like that of many living aquatic reptiles (Arden et al., 2019).

Implications for the rest of Spinosauridae

So naturally, this all begs the following questions: did other spinosaurids have similar tails? If so, which ones and to what extent? How did tail morphology vary between the two subfamilies—Baryonychinae and Spinosaurinae—and when did the family evolve tail fins to begin with?

Unfortunately, answering these is easier said than done, and there remains to be extensive and detailed research on the topic. However, based on what is currently known about spinosaur phylogenetics, we may be able to make a few assumptions. Keep in mind a few things however: A, This is very preliminary stuff; B, I am not a palaeontologist; and C. There are multiple ways of interpreting the evidence and detailed conclusions cannot be made until we have more complete remains of the animals I'll be discussing.

So first off, just like how not all spinosaurids had tall sails (e.g. Baryonyx), the presence of paddle-like tails was probably similarly variable. Unfortunately there's little in the way of specimens with well-preserved tail remains for this dinosaur family, so we have to heavily rely on inferences. Thus, I went ahead and scoured the literature in search for spinosaur tail bones and put together a diagram to illustrate roughly what we have so far. Bear in mind however that this is not a comprehensive diagram; some specimens have been excluded due to lack of descriptions and images.

Comparison of tail material from various spinosaurid individuals. The silhouettes of fragmentary specimens are highly speculative, being based mostly on the animals' phylogenetic classification and temporal range. Credit: Own work

Tails of baryonychines: Baryonyx walkeri and Cristatusaurus lapparenti

Starting off from the top, it's probably safe to assume that earlier and more basal spinosaurids like Baryonyx and Cristatusaurus* did not have tail fins. Besides their crocodile-like skulls and interlocking conical teeth—congruent with a diet of aquatic prey and small animals—the rest of their anatomy was not particularly well-adapted towards semiaquatic life, and they lacked the dense osteosclerotic bones of spinosaurines (Arden et al. 2019).

Life restorations of Baryonyx walkeri (top) and Cristatusaurus lapparenti (bottom)  Credits (in same order): Rebecca Slater (CC BY-SA 3.0), Андрей Белов (CC BY 3.0)

Additionally, the tail material we have for them does not seem to indicate such anatomy. In Cristatusaurus the caudal neural spines are very elongated, which, along with its relatively short legs were noted as possible adaptations for swimming by Arden and colleagues (Arden et al. 2019). However, the spines are also very broad and blade-shaped, almost coming in contact with one another; and their articulating processes appear to be of typical prominence for tetanuran theropods (Sereno et al., 1998). These factors would make the tail more rigid than those of Ichthyovenator and Spinosaurus, which both had more slender caudal neural spines and shrunken articulating vertebral processes. Caudal material for Baryonyx is poor, but isolated neural spines probably belonging to the rear sacrum or front of the tail show a similarly broad, blade-shaped morphology (Charig and Milner, 1997).

Overall, it has been reasonably inferred that baryonychines were more terrestrial and had generalist diets and lifestyles, similar to herons or kodiak bears (Arden et al. 2019). This could either represent baryonychines niche partitioning from the more aquatic spinosaurines, or it could mean baryonychines were simply earlier spinosaurs that more closely resembled typical theropods, before the clade transitioned fully to semiaquatic body plans. This might be the case if Baryonychinae is a paraphyletic group as suggested by Marcos Sales and Caesar Schultz in 2017.

*Cristatusaurus lapparenti hails from the same formation as and is almost identical anatomically to Suchomimus tenerensis (Hendrickx et al., 2016), though the former is based on more fragmentary material and so a formalized synonymy has not yet been made pending further examination of the material. In this post, I'll consider Suchomimus tenetensis a junior synonym, given that this is almost certainly the case.

Tails of Camarillasaurus cirugedae, Phuwiang spinosaurid B, and Vallibonavenatrix cani

We have eight caudal vertebrae from Camarillasaurus and ten from Phuwiang spinosaurid B, both of which date to the Barremian (same time as Baryonyx) and most do not preserve complete neural spines (Samathi, 2019). Due to this early date we can assume they are more basal taxa, for this reason and the lack of data I restored these in the diagram as conventional theropod tails. Though their true complete shape is unknown.

The Vallibonavenatrix holotype preserves four caudal vertebrae (one figured) and is placed within the Spinosaurinae. Its skeleton has many similarities with that of Ichthyovenator, more so than it does with its fellow European spinosaurs, especially in regards to the pelvis and an isolated, fan-shaped neural spine (Malafaia et al., 2020). Thus, I've speculatively restored it above with a similar, though less pronounced tail fin than Ichthyovenator.

Tails of Ichthyovenator laosensis and Spinosaurus aegyptiacus

Life reconstruction of Ichthyovenator laosensis Credit: Own work

The only spinosaurids we have preserved tail fins for are Ichthyovenator and Spinosaurus. The former appears to have had a lobe-shaped one and the latter a more newt/crocodilian-like paddle form. It is worth noting however that that neither of the tail remains we have from these taxa are fully intact, so their complete shapes are still slightly speculative and may change upon future redescription or the discovery of other specimens.

Tails of Irritator challengeri and Oxalaia quilombensis

Since Irritator was a late (Aptian) spinosaurine that has been placed between Ichthyovenator and Spinosaurus in phylogenetic analyses and in the geological timeline (Arden et al., 2019), it's probably reasonable to assume that it also possessed a fin or paddle-like tail. The shape in which I have restored Irritator's tail in the diagram above is intended to be an intermediate morphology between those two taxa. 

A series of sacrocaudal vertebrae from the same formation as Irritator may belong to the genus or its likely synonym (Angaturama), though since both of these taxa are known only from skulls, lack of overlapping remains makes this difficult to determine (Bittencourt and Kellner, 2004).

Speculative reconstruction of Oxalaia quilombensis, based largely on Spinosaurus aegyptiacus Credit: Own work

Oxalaia quilombensis was a giant spinosaurine from the Cenomanian of what is now Brazil. Known from only two snout fragments, it had extremely close affinities to Spinosaurus, potentially being synonymous with the genus or species S. aegyptiacus (Smyth et al., 2020). Given this close relationship, I've restored the tail as pretty much identical to the paddle-shape of Spinosaurus. One caudal vertebral centra (UFMA 1.10.229) from the Alcântara Formation was referred to Sigilmassasaurus by Manuel Medeiros and colleagues, a strange assignment given that only neck vertebrae are known from said genus. Mickey Mortimer has pointed out that the fossil is just as likely to belong to Oxalaia, and so I've tentatively placed it in my diagram above under Oxalaia quilombensis?.

Conclusion

Size comparison of six spinosaurid taxa; some of the tail shapes are more speculative than others, but this is probably the most consevative approach we can take to restoring their silhouettes as of now. Credit: Own work

So to sum it all up, we know spinosaurids developed many aquatic adaptations during their evolution, including deep tails in at least two taxa (Ichthyovenator laosensis and Spinosaurus aegyptiacus). Given what we currently know about the family's phylogeny, it's very likely that other spinosaurine species—such as Irritator challengeri and Oxalaia quilombensis—also had fin or paddle-like tails. For earlier/more basal spinosaurids like baryonychines however, this was probably not the case. There's little we can say for spinosaurs with poor tail fossils and uncertain phylogenetic placement; we're just beginning to enter a whole realm of this clade's anatomy we didn't even know existed.

So if you're a palaeoartist and plan on or are illustrating a spinosaurid, I hope this serves as a useful guide for how to tackle reconstructing their tails! The final choice/interpretation is entirely up to you however, since there's so much we don't know as of yet. After all, The spinosaurid fossil record is still very poor, especially in comparison to extensively well-documented theropod groups like tyrannosaurs. 

What we do know for sure is that spinosaurids were clearly even more bizarre and perfectly-suited for a life near and in the watery depths than we previously imagined. I for one cannot wait to see what the future brings for how we view these incredible animals.

References:

• Alyssa Bell & Luis M. Chiappe (2016) A species-level phylogeny of the Cretaceous Hesperornithiformes (Aves: Ornithuromorpha): implications for body size evolution amongst the earliest diving birds, Journal of Systematic Palaeontology, 14:3, 239-251.

• Agnolín, F.L., Egli, F.B., Chatterjee, S. et al. Vegaviidae, a new clade of southern diving birds that survived the K/T boundary. Sci Nat 104, 87 (2017).

• Ksepka, Daniel T., Sara Bertelli and Norberto P. Giannini. “The phylogeny of the living and fossil Sphenisciformes (penguins).” (2006).

• Romain Amiot, Eric Buffetaut, Christophe Lécuyer, Xu Wang, Larbi Boudad, Zhongli Ding, François Fourel, Steven Hutt, François Martineau, Manuel Alfredo Medeiros, Jinyou Mo, Laurent Simon, Varavudh Suteethorn, Steven Sweetman, Haiyan Tong, Fusong Zhang, Zhonghe Zhou; Oxygen isotope evidence for semi-aquatic habits among spinosaurid theropods. Geology ; 38 (2): 139–142.

• Hone, David & Holtz, Thomas. (2017). A Century of Spinosaurs - A Review and Revision of the Spinosauridae with Comments on Their Ecology. Acta Geologica Sinica - English Edition. 91. 1120-1132.

• Aureliano, Tito & Ghilardi, Aline & Buck, Pedro & Fabbri, Matteo & Samathi, Adum & Delcourt, Rafael & Fernandes, Marcelo & Sander, Paul. (2018). Semi-aquatic adaptations in a spinosaur from the Lower Cretaceous of Brazil. Cretaceous Research. 90. 283-295.

• M.S.Arden, Thomas & Klein, Catherine & Zouhri, Samir & Longrich, Nick. (2018). Aquatic adaptation in the skull of carnivorous dinosaurs (Theropoda: Spinosauridae) and the evolution of aquatic habits in spinosaurus. Cretaceous Research. 93. 10.1016/j.cretres.2018.06.013.

• Lee, Y., Ryan, M.J. & Kobayashi, Y. The first ceratopsian dinosaur from South Korea. Naturwissenschaften 98, 39–49 (2011).

• Ji Q., Wu X., Cheng Y., Ten F., Wang X., and Ji Y. 2016. Fish-hunting ankylosaurs (Dinosauria, Ornithischia) from the Cretaceous of China. Journal of Geology, 40(2).

• Cau, A., Beyrand, V., Voeten, D. et al. Synchrotron scanning reveals amphibious ecomorphology in a new clade of bird-like dinosaurs. Nature 552, 395–399 (2017).

• Dal Sasso, Cristiano & Maganuco, Simone & Cioffi, Armando. (2009). A Neurovascular cavity within the snout of the predatory dinosaur Spinosaurus.

• Foffa, Davide & Sassoon, Judyth & Cuff, Andrew & Mavrogordato, Mark & Benton, Michael. (2014). Complex rostral neurovascular system in a giant pliosaur. Naturwissenschaften.

• Emily J. Rayfield, Angela C. Milner, Viet Bui Xuan & Philippe G. Young (2007) Functional morphology of spinosaur ‘crocodile-mimic’ dinosaurs, Journal of Vertebrate Paleontology, 27:4, 892-901

• Hans-Dieter Sues, Eberhard Frey, David M. Martill & Diane M. Scott (2002) Irritator challengeri, a spinosaurid (Dinosauria: Theropoda) from the Lower Cretaceous of Brazil, Journal of Vertebrate Paleontology, 22:3, 535-547.

• Evers SW, Rauhut OW, Milner AC, McFeeters B, Allain R. A reappraisal of the morphology and systematic position of the theropod dinosaur Sigilmassasaurus from the "middle" Cretaceous of Morocco. PeerJ. 2015;3:e1323. Published 2015 Oct 20.

• M. L. Casanovas et al., Zub. Monogr. 5, 135–163 (1993).

• Ibrahim, Nizar; Sereno, Paul C.; Dal Sasso, Cristiano; Maganuco, Simone; Fabri, Matteo; Martill, David M.; Zouhri, Samir; Myhrvold, Nathan; Lurino, Dawid A. (2014). "Semiaquatic adaptations in a giant predatory dinosaur". Science. 345 (6204): 1613–6.

• Hone, David & Holtz, Thomas. (2019). Comment on: Aquatic adaptation in the skull of carnivorous dinosaurs (Theropoda: Spinosauridae) and the evolution of aquatic habits in spinosaurids. 93: 275–284. Cretaceous Research. 10.1016/j.cretres.2019.05.010.

• Schade, M., Rauhut, O.W.M. & Evers, S.W. Neuroanatomy of the spinosaurid Irritator challengeri (Dinosauria: Theropoda) indicates potential adaptations for piscivory. Sci Rep 10, 9259 (2020).

• Charig, A. J.; Milner, A. C. (1997). "Baryonyx walkeri, a fish-eating dinosaur from the Wealden of Surrey". Bulletin of the Natural History Museum of London. 53: 11–70.

• Cristiano Dal Sasso, Simone Maganuco, Eric Buffetaut & Marco A. Mendez (2005) New information on the skull of the enigmatic theropod Spinosaurus, with remarks on its size and affinities, Journal of Vertebrate Paleontology, 25:4, 888-896.

• Milner, Andrew & Lockley, Martin. (2006). A large collection of well-preserved theropod dinosaur swim tracks from the Lower Jurassic Moenave Formation, St. George, Utah. New Mexico Museum of Natural History and Science Bulletin. 37.

• Gimsa, J., Sleigh, R., & Gimsa, U. (2016). The riddle of Spinosaurus aegyptiacus’ dorsal sail. Geological Magazine, 153(3), 544-547.

• Rooney, Laura, "Postcranial Morphology and the Locomotor Adaptations of Extant and Extinct Crocodylomorphs and Lepidosaurs" (2018). Electronic Theses and Dissertations. Paper 3418.

• Raxworthy, Christopher. (2010). Courtship, Fighting and Sexual Dimorphism of the Banded Newt, Triturus vittatus ophryticus. Ethology. 81. 148-170. 

• Kyra Moosmueller & Benjamin Wilhelm (2019) Skeletal Adaptations for an Aquatic Lifestyle in the Tail of Reptiles. Department of Natural Sciences and Mathematics, SUNY Cobleskill

• Bedford, Gavin S. and Keith A Christian. “Tail morphology related to habitat of varanid lizards and some other reptiles.” (1996).

• Dilkes, D.W.; Hutchinson, J.R.; Holliday, C.M.; Witmer, L.M. (2012). "Reconstructing the Musculature of Dinosaurs". In Brett-Surman, M.K.; Holtz, T.R.; Farlow, J.O. (eds.). The Complete Dinosaur (2 ed.). pp. 150–190. ISBN 978-0-253-00849-7.

• Allain, Ronan & Xaisanavong, Tiengkham & Richir, Philippe & Khentavong, Bounsou. (2012). The first definitive Asian spinosaurid (Dinosauria: Theropoda) from the Early Cretaceous of Laos. Die Naturwissenschaften. 99. 369-77.

• Allain, Ronan (2014). "New material of the theropod Ichthyovenator from Ban Kalum type locality (Laos): Implications for the synonymy of Spinosaurus and Sigilmassasaurus and the phylogeny of Spinosauridae". Journal of Vertebrate Paleontology Programs and Abstracts. 74: 78.

• Ibrahim, N., Maganuco, S., Dal Sasso, C. et al. Tail-propelled aquatic locomotion in a theropod dinosaur. Nature 581, 67–70 (2020).

• Henderson DM. 2018. A buoyancy, balance and stability challenge to the hypothesis of a semi-aquatic Spinosaurus Stromer, 1915 (Dinosauria: Theropoda) PeerJ 6:e5409

• "Chinese Water Dragon"World Association of Zoos and Aquariums.

• Sereno, P.C.; Beck, A.L.; Dutheil, D.B.; Gado, B.; Larsson, H.C.E.; Lyon, G.H.; Marcot, J.D.; Rauhut, O.W.M.; Sadleir, R.W.; Sidor, C.A.; Varricchio, D.D.; Wilson, G.P; Wilson, J.A. (1998). "A long-snouted predatory dinosaur from Africa and the evolution of spinosaurids". Science. 282 (5392): 1298–1302. 

• Hendrickx C, Mateus O, Buffetaut E. Morphofunctional Analysis of the Quadrate of Spinosauridae (Dinosauria: Theropoda) and the Presence of Spinosaurus and a Second Spinosaurine Taxon in the Cenomanian of North Africa. PLoS One. 2016;11(1):e0144695. Published 2016 Jan 6.

• Samathi, Adun (2019). "6: "A Spinosaurid from Thailand and the Reassessment of Camarillasaurus from the Early Cretaceous of Spain"". Theropod Dinosaurs From Thailand and Southeast Asia Phylogeny, Evolution, and Paleobiogeography. University of Bonn. pp. 161–181.

• Elisabete Malafaia; José Miguel Gasulla; Fernando Escaso; Iván Narváez; José Luis Sanz; Francisco Ortega (2020). "A new spinosaurid theropod (Dinosauria: Megalosauroidea) from the late Barremian of Vallibona, Spain: Implications for spinosaurid diversity in the Early Cretaceous of the Iberian Peninsula". Cretaceous Research. in press: 104221.

• Bittencourt, Jonathas & Kellner, Alexander. (2004). On a sequence of sacrocaudal theropod dinosaur vertebrae from the Lower Cretaceous Santana Formation, northeastern Brazil. Arq Mus Nac. 62. 309-320.