Crato Formation fossils and the new tapejarids

i-feb003a2464a70b7fc12925bea5308ef-Crato book cover 18-1-2008.jpg

Those of you with particularly good memories might recall the little references I've been making here and there to a 'big, personally-relevant publication', and those asides to new papers about pleurodires and enantiornithines. Following horrific delays (caused by amphibians, dinosaur growth rates, ichthyosaurs and conferences) I am, finally, pleased to announce that Cambridge University Press' huge new book, The Crato Fossil Beds of Brazil: Window into an Ancient World (Martill et al. 2007), was published in December 2007 and is now available. If you're interested in Cretaceous anurans, squamates, turtles, birds, or - most importantly - pterosaurs, you will want to see a copy...

To begin with, for those who don't know, I'll tell you that the Crato Formation is a Lower Cretaceous sequence of clastic and carbonate sediments of the Brazilian Araripe Basin. The exact name of the unit has differed among authors (you might have seen it as the Crato Member of the Santana Formation), but the idea that the Crato deserves recognition as a formation looks pretty good and is the system I follow. The precise age of the Crato Formation has been difficult to pin down but various lines of data make a late Aptian date most likely.

From the palaeontological point of view, the Crato Formation is globally significant in being a lagerstätte (a place of exceptional preservation), and it's famous for its exquisitely preserved, spectacularly complete arthropods, fishes and others fossils. Why the fossils are so exquisite and complete is a good question, but one that I'm not going to answer as I want to get straight to the tetrapods. And the Crato Formation is pretty good on the tetrapods, all of which are reviewed in this book. It has anurans* (Leal et al. 2007), araripemydid pleurodire turtles (Naish 2007), squamates (Martill 2007), susisuchid and araripesuchid crocodyliforms (Frey & Salisbury 2007), birds (Naish et al. 2007) and pterosaurs (Unwin & Martill 2007). I'd love to talk about all of those - particularly the araripemydids and birds (and eventually I will) - but right now we need to address one group in particular, the pterosaurs.

* Both pipoids and leptodactyliform hyloid neobatrachians. Now you're pleased with yourself for reading all those anuran articles right?

Dave Unwin and Dave Martill's pterosaur chapter - all 49 pages of it - makes this volume a must-have if you're seriously interested in pterosaurs. Crato Formation pterosaurs include ornithocheiroids and azhdarchoids, and Unwin & Martill (2007) review all the named taxa, as well as describing and figuring lots of unnamed or unnameable fragments.

i-61ff88152a554e512f0c816178108f15-Ludodactylus its eerie wails keeping us awake at night.jpg

Ornithocheiroids are represented in the Crato Formation by some awesome specimens as well as some scrappy ones. Easily the most impressive is the holotype of Ludodactylus sibbicki Frey et al., 2003: one of two ornithocheirid ornithocheiroids that combines a Pteranodon-like cranial crest with dentition (the other is the Lower Cretaceous Isle of Wight taxon Caulkicephalus trimicrodon). Named on the basis of an essentially complete skull (shown at left), the Ludodactylus holotype died with a sharp-tipped leaf stuck in the soft tissues of its lower jaw, and the frayed tip of the leaf indicates that the poor animal tried to remove the offending obstacle by rubbing it against the ground. That's all guesswork of course, but if you can find a better way of explaining the fossil I'd like to hear it. Brasileodactylus, originally described from the Santana Formation, seems to be present in the Crato Formation according to Unwin & Martill (2007), and finally there's Arthurdactylus, a mid-sized ornithocheiroid known from a well-preserved headless skeleton.

i-90e836b3091db36492649ad46b33eec9-SMNK 3830 PAL today it's a sinopterid.jpg

And then there are the azhdarchoids. As yet unassigned to a genus or species is one of my personal favourites, SMNK 3830 PAL (or SMNK Pal 3830, depending on which source you consult: SMNK is the Staatliches Museum für Naturkunde in Karlsruhe, Germany). Consisting of a left forelimb and both hindlimbs, this specimen is a big deal in preserving an extensive portion of cheiropatagium (the membrane that extends between the arm and the body and leg): I blogged about it on ver 1 here and, despite comments otherwise, it preserves clear evidence that the cheiropatagium graded smoothly into the ankle (see also Frey et al. 2003). Unwin & Martill (2007) now identify this specimen as a tapejarid, closest in proportions to Sinopterus dongi.

SMNK PAL 2342 (an articulated wing metacarpal and wing finger) is another interesting specimen: it was first described as a small azhdarchid (Martill & Frey 1999) because some of its phalanges are T-shaped in cross-section. If this identification is correct, the specimen would be one of the smallest azhdarchids, and also one of the oldest... in which case it's fun to speculate that it might be a basal form suggesting a Southern Hemisphere origin for Azhdarchidae. Alas, Unwin & Martill (2007) argue that the proportionally short wing metacarpal of the specimen in fact suggests that the specimen is a tapejarid, and furthermore a T-shaped cross-sectional shape to the wing phalanges is not unique to azhdarchids anyway (Kellner 2004). For the record, 'tapejarid' is used by Unwin & Martill (2007) in the restrictive sense: i.e., for the Sinopterus-Tapejara clade, with Tupuxuara being elsewhere.

A new name for an azhdarchoid: Ingridia

i-372f6e1c2be5ce7cc992561d71c51bc8-Ingridia Munich photo 18-1-2008.jpg

While the pterosaur chapter is mostly a review, it also names a new taxon, and here we run into a bit of a mess. Two azhdarchoid taxa, both based on well-preserved skulls, have been named from the Crato Formation: Tapejara imperator Campos & Kellner, 1997 and Tap. navigans Frey et al., 2003. Both are remarkable sail-crested forms, similar overall yet different in the details (a fantastic life restoration of Tap. imperator was previously shown on Tet Zoo here, and the type specimen of Tap. navigans is shown in the adjacent image). When first named, both species were referred to Tapejara, a genus erected for Tap. wellnhoferi Kellner, 1989 from the Santana Formation (Kellner 1989). Tap. wellnhoferi isn't sail-crested however and also differs from the Crato forms in being shorter-faced, in having a far shorter nasoantorbital fenestra, and in several other details. So when all three species were included in a phylogenetic analysis it wasn't entirely surprising that they failed to form a special little clade (Martill & Naish 2006). Martill & Naish (2006) implied in their study that Tap. navigans and Tap. imperator might not really be members of Tapejara, but might instead deserve their own generic names.

Well, that suggestion is here taken to its logical extreme: Unwin & Martill (2007) regard Tap. navigans and Tap. imperator as generically distinct from Tap. wellnhoferi, and furthermore as each other's closest relatives (a possibility not recovered in any of the more parsimonious cladograms produced by Martill & Naish (2006), incidentally). The new navigans + imperator clade is coined Ingridia, a name that commemorates Ingrid Wellnhofer, wife of leading pterosaur expert Peter Wellnhofer. The hypertrophied premaxillary crest with its spine-like supra-premaxillary ossification, and the relatively elongate preorbital region are regarded as diagnostic for Ingridia within Tapejaridae, and the type species is given as imperator.

Tupandactylus vs Ingridia

i-a62726a1f992edaa428b7327740c7497-Mike Hanson tapejarids.jpg

Now, an interesting thing happened earlier this year: Kellner & Campos (2007) published a paper on tapejarids (in fact, their paper actually came out only a few weeks prior to the publication of the Crato book). Therein they came up with a new generic name for Tap. imperator, and that new name is Tupandactylus. This taxon is diagnosed on the basis of its tall cranial crest which is only ossified in its basal half, on the dorsally oriented spine-like structure that grows from the premaxillary crest, and the posteriorly projecting crest that extends beyond the occiput [adjacent picture is Mike Hanson's scaled reconstructions of all three species referred to Tapejara. For the full-size version go here].

It might seem at first sight that Tupandactylus makes Ingridia redundant, or in other words that the former is an objective senior synonym of the latter. Given that Ingridia was named in honour of Peter Wellnhofer's late wife, itself an act of affection and gratitude for the both of them, it would be unfortunate if this name had to be scrapped, especially given that Peter had been told about the new name at the 2007 Munich pterosaur meeting (held in his honour). A poster displayed at the Munich meeting included the name Ingridia, and we know from the acceptance date of Kellner & Campos (2007) that they were aware of this name while their paper was in press. However, authors do have the right to coin new generic names for their species where this is appropriate. Furthermore, as mentioned above, Martill & Naish (2006) had already argued that imperator and navigans needed new names, so maybe it was only a matter of time before someone acted on this.

Where do we go from here? If imperator and navigans form a clade as Unwin & Martill (2007) state (they don't include a cladistic analysis), then the name Tupandactylus is the one that has priority and the one that should be used for it. So, bye-bye Ingridia. However, Kellner & Campos' diagnosis for Tupandactylus explicitly applies only to imperator and, as mentioned above, Martill & Naish (2006) never found imperator and navigans to form a clade. If anything, we always found imperator to group closest to Tap. wellnhoferi, sometimes grouping as its sister-taxon (Martill & Naish 2006, Fig. 9A: this is the cladogram shown immediately below). This latter relationship would suggest that imperator should arguably be retained in Tapejara. Rather, it was navigans that seemed the most disparate: in some runs of our data we found it to be the most basal member of Azhdarchoidea, and thus well removed from Tap. wellnhoferi and imperator.

i-59d648bfe7155449b196e10be3831056-Martill & Naish cladogram fig 9A resize.jpg

In this cladogram (Fig. 9A from Martill & Naish 2006), all the Sinopterus and 'Tapejara' species form a clade (Huaxiapterus would be in there too had we included it) for which a new, restricted use of the name Tapejaridae was used. Tap. wellnhoferi and imperator formed a clade; navigans was the most basal tapejarid. Tapejaridae, in the node-based form used by Kellner (2003, p. 125), would be synonymous with Azhdarchoidea in this cladogram.

i-e2eaa049908e7765339e5908fa801c93-Tupan imperator Witton 18-1-2008.jpg

Kellner & Campos (2007) were rather uncertain about the position of navigans: they said that, if it really does lack the posteriorly-projecting occipital process present in imperator, then it possibly does represent a distinct taxon (the implication from Kellner & Campos' paper seems to be that navigans can only be distinguished from imperator on the basis of its occipital region, but this is incorrect and the two are quite different: navigans has a more vertical premaxillary crest, lacks a posteriorly-projecting occipital process, differs in the dorsal extent of the striated bone lamina that supports the cranial crest, and has a much broader jugal). However, Kellner & Campos also said that navigans is so similar to imperator that it 'possibly represents the same taxon' (p. 1). Based on current evidence, I think that the two are distinct, as was found in the analyses of Martill & Naish (2006) [adjacent image is a cropped version of Mark Witton's life restoration of imperator. This, and all of Mark's images used in the cladograms here, are from his world-famous, Science-endorsed flickr site].

If the two are distinct, and if they are phylogenetically disparate, then we have a solution to the Tupandactylus vs Ingridia problem: Tupandactylus applies only to imperator, and Ingridia only to navigans. Some workers have argued that taxa should only be named once a cladistic analysis has shown what their relationships are, in order that preconceptions about 'distinctness' are avoided (Bonde 1999), and one could argue that what has happened with imperator and navigans is (more or less) appropriate.

i-63f809a54da0f0b6919bb2b5c42ceedc-Martill & Naish cladogram fig 9B resize.jpg

In this cladogram (Fig. 9B from Martill & Naish 2006), tapejarids are paraphyletic with respect to neoazhdarchians, navigans is the most basal azhdarchoid, and Tap. wellnhoferi doesn't group together with either navigans or imperator. Again, Tapejaridae, in the node-based form used by Kellner (2003, p. 125), would be synonymous with Azhdarchoidea in this cladogram.

However, the big complication here is that Unwin & Martill (2007) chose imperator as the type species for Ingridia. I'm not 100% sure on this (help out if you can), but according to Article 67 of the ICZN, use of the same type species for both of these proposed genera renders them synonyms, I think. Ingridia may indeed remain available for navigans, but someone has to establish this by acting as reviser, and for now the name Ingridia really is synonymous with Tupandactylus.

So, all in all, I think there is a way out of this: anyway, we still have a way to go before we can be more confident about the relationships among the species concerned and it's not like Martill & Naish (2006) is the last word on which topologies are best supported. Our data run was small and preliminary and I hope we can expand on it in future.

Having indulged in this long discussion of species referred to Tapejara, I should point out that the main point of Kellner & Campos' 2007 paper wasn't to address this issue, but rather to (1) assert the generic distinction of Thalassodromeus, a taxon that Martill & Naish (2006) regarded as synonymous with Tupuxuara, and to (2) assert the specific distinction of Tupuxuara leonardii, a species that Martill & Naish (2006) regarded as synonymous with Tup. longicristatus. It turns out that Thalassodromeus really is quite a different beast from Tupuxuara, and Dave Martill and I are happy to admit that we were wrong in sinking it. But I still think there are problems with Tup. leonardii, and indeed with Tupuxuara as a whole (Dave and Mark Witton are currently working on this topic). More on that later - this article is already way too long (but one more thing: Kellner & Campos (2007) formally name Thalassodrominae [sic - it should be Thalassodromeinae] for the clade that includes Thalassodromeus and Tupuxuara. In other papers - including Unwin & Martill (2007) - this group has informally been called 'the tupuxuarids').

i-57ea5f416ff4d88dc04ad1cf77fb573c-Dave doesn't call me anymore.jpg

While on the subject of pterosaurs, the new tiny anurognathid specimen being worked on by Chris Bennett (alluded to in one of the Munich pterosaur meeting articles: here) has now been published (Bennett 2007), as has the new Huaxiapterus species that I also mentioned: it's called H. benxiensis (Lü et al. 2007) [in adjacent photo, Dave Unwin and I ponder the Pterosauria. From the BSPG Mesozoic Vertebrate Research Group pterosaur meeting page, © Oliver Rauhut. Look at the John Conway cartoon on the whiteboard].

Anyway... that's enough about pterosaurs. The big deal is that Martill et al.'s Crato Formation book is out, and in fact has been out since December 2007. Amazon are advertising it here - it's certainly not cheap but, like I said, those interested in Cretaceous tetrapods will definitely want to see it.

Refs - -

Bennett, S. C. 2007. A second specimen of the pterosaur Anurognathus ammoni. Paläontologische Zeitschrift 81, 376-398.

Bonde, N. 1999. Systematic descriptions, diagnoses and cladistics: alternative style, with a new mosasaur as an example. In Hoch, E. & Brantsen, A. K. (eds) Secondary Adaptation to Life in Water. University of Copenhagen, p. 9.

Frey, E. & Salisbury, S. W. 2007. Crocodilians of the Crato Formation: evidence for enigmatic species. In Martill, D. M., Bechly, G. & Loveridge, R. F. (eds) The Crato Fossil Beds of Brazil: Window into an Ancient World. Cambridge University Press (Cambridge), pp. 463-474.

- ., Tischlinger, H., Buchy, M.-C. & Martill, D. M. 2003. New specimens of Pterosauria (Reptilia) with soft parts with implications for pterosaurian anatomy and locomotion. In Buffetaut, E. & Mazin, J.-M. (eds) Evolution and Palaeobiology of Pterosaurs. Geological Society Special Publication 217. The Geological Society of London, pp. 233-266.

Kellner, A. W. A. 1989. A new edentate pterosaur of the Lower Cretaceous from the Araripe Basin, northeast Brazil. Anais da Academia Brasileira de Ciências 61, 439-446.

- . 2003. Pterosaur phylogeny and comments on the evolutionary history of the group. In Buffetaut, E. & Mazin, J.-M. (eds) Evolution and Palaeobiology of Pterosaurs. Geological Society Special Publication 217. The Geological Society of London, pp. 105-137.

- . 2004. New information on the Tapejaridae (Pterosauria, Pterodactyloidea) and discussion of the relationships of this clade. Ameghiniana 41, 521-534.

- . & Campos, D. de A. 2007. Short note on the ingroup relationships of the Tapejaridae (Pterosauria, Pterodactyloidea). Boletim do Museu Nacional, Nova Séroe, Rio de Janeiro - Brasil. Geologia 75, 1-14.

Leal, M. E. C, Martill, D. M. & Brito, P. M. 2007. Anurans of the Crato Formation. In Martill, D. M., Bechly, G. & Loveridge, R. F. (eds) The Crato Fossil Beds of Brazil: Window into an Ancient World. Cambridge University Press (Cambridge), pp. 444-451.

Lü, J., Gao, Y., Xing, L., Li, Z. & Ji, Q. 2007. A new species of Huaxiapterus (Pterosauria: Tapejaridae) from the Early Cretaceous of western Liaoning, China. Acta Geologica Sinica 81, 683-687.

Martill, D. M. 2007. Lizards of the Crato Formation. In Martill, D. M., Bechly, G. & Loveridge, R. F. (eds) The Crato Fossil Beds of Brazil: Window into an Ancient World. Cambridge University Press (Cambridge), pp. 458-462.

- ., Bechly, G. & Loveridge, R. F. 2007. The Crato Fossil Beds of Brazil: Window into an Ancient World. Cambridge University Press, Cambridge.

- . & Naish, D. 2006. Cranial crest development in the azhdarchoid pterosaur Tupuxuara, with a review of the genus and tapejarid monophyly. Palaeontology 49, 925-941.

- . & Frey, E. 1999. A possible azhdarchid pterosaur from the Crato Formation (Early Cretaceous, Aptian) of northeast Brazil. Geologie en Mijnbouw 78, 315-318.

Naish, D. 2007. Turtles of the Crato Formation. In Martill, D. M., Bechly, G. & Loveridge, R. F. (eds) The Crato Fossil Beds of Brazil: Window into an Ancient World. Cambridge University Press (Cambridge), pp. 452-457.

- ., Martill, D. M. & Merrick, I. 2007. Birds of the Crato Formation. In Martill, D. M., Bechly, G. & Loveridge, R. F. (eds) The Crato Fossil Beds of Brazil: Window into an Ancient World. Cambridge University Press (Cambridge), pp. 525-533.

Unwin, D. M. & Martill, D. M. 2007. Pterosaurs of the Crato Formation. In Martill, D. M., Bechly, G. & Loveridge, R. F. (eds) The Crato Fossil Beds of Brazil: Window into an Ancient World. Cambridge University Press (Cambridge), pp. 475-524.


More like this

On to more of my thoughts about the outstanding Wellnhofer pterosaur meeting, held at the Bayerische Staatssammlung für Paläontologie und Geologie (Bavarian State Palaeontological Collection - BSPG) in Munich last week. For part I go here, and for initial thoughts on what a neat conference it was…
In the previous article I discussed the outside section of the Royal Society Summer Science Exhibition's pterosaur display (hosted at Royal Festival Hall on London's South Bank). The exhibition (which finished on July 4th, sorry) incorporated three flying, life-sized azhdarchids - suspended from…
You blog readers are so fickle. Write several thousand words on a spectacular group of carnivorous mammals that have never previously been the subject of any sort of semi-popular review, and get bugger all attention. Write about 50 words and post two pictures, and - hey - the whole world goes nuts…
We know all too little about the biology and behaviour of the pterosaurs, the amazing, often bizarre flying reptiles of the Mesozoic Era. Most of our ideas - about feeding behaviour, locomotion, physiology and social and sexual behaviour - are inferences based on bones, or inferences based on…

Well, unless the new Anurognathus paper is free, I can't get to it. Does anybody have it, and would be willing to send it my way? Thanks in advance. Hahaha...I don't have to buy that book now, because Darren just gave an excellent summary of the pterosaur bits! That Ludodactylus specimen is almost too good to pass up, though. Gorgeous, gorgeous fossil.
Darren, have you heard anything more about that Triassic horned, theropod-skulled, parasagittal-limbed, heterodont pterosaur who you hinted at in the 3rd pterosaur meeting post?

use of the same type species for both of these proposed genera renders them synonyms

Yes. Objective synonyms. Ingridia is toast for all time -- unless some technical mistake can be found in the Tupandactylus paper, in which case Tupandactylus is toast for all time.


On the poor Ludodactylus... are you sure this isn't a postmortem association? The position of the hyoid is impossible when the soft tissues are halfway intact, isn't it?

By David Marjanović (not verified) on 18 Jan 2008 #permalink

Yes, yes I have....

Then I must quote: "Truly you are a bastard[,] Darren."

By David Marjanović (not verified) on 18 Jan 2008 #permalink

Well, it's not published yet, if that's what you mean. Wait until Dave Hone shows up in the comments, he may say more...

It is really Cretacous moth with wing pattern?

Navigans and imperator indeed look like female and male or maybe immature and adult.

What was the paleoenvironment? Was Tapejara a toucan or waterfowl?

And anybody did structural studies on skull => what it ate?

Dave Hone has shown up in the comments!

I have spoken to Rico Stecher this week (discoveror, preparator and describor [it suits the theme]) of the new and amazing Triassic Swiss pterosaur. His paper has been delayed slightly and I have been doing some minor edits to the English for him this last week.

It should not be pretty much sorted and obviously still needs to go though the proof reading before publication. Saldy the journal it is with is quarterly so now he has missed Jan, do not expect it to come out before April. Hopefully an in press advanced release will be avilable before then.

That's about it for now. I will of course flag it massively on Diniobase when it appears and I know Darren will want to blog about it ferrociously. It is, frankly, amazing.

By Dave Hone (not verified) on 18 Jan 2008 #permalink

Sorry, Darren, but as David said, if two genera have the same type species, then the later name is automatically sunk. End of story.

Of course, the really cool thing about Ludodactylus is the name - from the Latin word for "to play", and named because with its mouth full of teeth Ludo looks more like many a cheap toy "Pteranodon" than the actual Pteranodon did.

Jerzy: Preservation of pigmentation pattern (but not colour, only light vs. dark) in insect wings is not uncommon, and examples date back pretty much as far as winged insects have been around. Similar preservation is not uncommon for feathers - Caudipteryx, for instance, had clearly banded feathers.

Just a quick note on feeding in Tapejara and others ther eis a long ongoing debate about them being fish feeders (or similar) vs fruit and seedeaters. Both arguemnts have some merit and I am working on a devious secret project that hopefully can point (not solve by any means) to the right answer. More to come.

In a few months when we have the reults. And then written the paper. And got it published....

By Dave Hone (not verified) on 18 Jan 2008 #permalink

The answer seems to be - really quite well. At the risk of turning into Darren - there is a fairly detailed paper about to come on this shortly (as part of the pterosaur volume from the Munich meeting) and I don't want to jump the gun on my own research (I am one of the authors) but at least in terms of flight mechanics, they don't seem to have much effect.

Weight is another factor entirely, though most were actually pretty thin and therefore quite light.

By Dave Hone (not verified) on 20 Jan 2008 #permalink

Oh, I forgot to add abotu Ludodactylus wrt the position of the leaf. Three things mitigate against a post morten association and a disarticulated hyoid.

First off the leaf is trapped a long way back in the jaw and between the two rami of the jaws. To get right in there (and all frayed up too) would be very difficult from a current, or somethign similar.

Secondly this fact is apparently quite common in pelicans (scooping up something which then gets traspped in a gular pouch) so there is a matching behavioural pattern in extant taxa.

Thirdly, hyoids do regulalry move like this after death, and if Ludo. had a large gular pouch (which matches other pterosaurs and its infereed diet) then it seems reasonable that the leaf was jammed deep into tissue and the damamge to the leaf was done inside the jaw. After death, the rest of the flesh rotted away leaving the leaf in place and the hyoids to move as they would actually be less solidly anchored in the jaw than the leaf at this point.

By Dave Hone (not verified) on 20 Jan 2008 #permalink

I see.

By David Marjanović (not verified) on 21 Jan 2008 #permalink

Man, some people just have the coolest jobs in the world..... [big sigh of envy]

By Luna_the_cat (not verified) on 21 Jan 2008 #permalink

Ludodactylus and the leaf... like you said, poor animal...

...fascinating forensically, though...

I'm intrigued, wondering whether it incurred that wound by grabbing at a food item it spotted among foliage of a spiky-leaved tree, where the leaf was in situ?

... and if so, whether it made its grab from a perched pose, or in flight?

By Graham King (not verified) on 31 Dec 2008 #permalink

hy, i'm a brazilian guy. i study biology im my college and am a curious in paleoart. is so hard find confiable papers in internet to use how database.
i want ask if you can send me some links were i can find confiable informations about pterosaurs, mainly about Tapejara, it´s my favorite!



"What was the paleoenvironment? Was Tapejara a toucan or waterfowl?"

And Dave Hone:

"Just a quick note on feeding in Tapejara and others ther eis a long ongoing debate about them being fish feeders (or similar) vs fruit and seedeaters. Both arguemnts have some merit and I am working on a devious secret project that hopefully can point (not solve by any means) to the right answer. More to come."

According to Mark Witton, they seem to have some features (such as strongly clawed fingers, plus the migration of the shoulder joint to a lower position than in other pterosaurs, thus reducing the wing muscles), which suggest that they were adapted to climb and to walk. The shape of the rostrum is also not akin to that of piscivore pterosaurs, and its now clear they weren't skimmers, so there's no way they could actively hunt fish. So this offers some unorthodox explanations:

1-They were toucan analogues (that means they were adapted to an arboreal lifestyle)

2-They were galliforme analogues (spending most time on the ground but climbing trees in case of predator attacks)

In both cases, they would be omnivores, feeding on whatever could fit on their bills