The Arctic Cretaceous revisited

One of the earliest posts here, way back in March 2014, was about an assemblage of high-Arctic Late Cretaceous coprolites that had been part of my graduate work. We did quite a bit with them (Chin et al. 2008), but there's always more that can be discovered. One of the things Chin et al. (2008) noticed was the rarity of body fossils for large potential coprolite producers. They broached the idea that the producers were only around part of the time, living elsewhere during the polar winter and only turning up to feed during the long sunny summer days. Duffy et al. (2026) takes this idea and runs with it.

You may recall that back in 2014 we discussed two major categories of coprolites, those with a dominantly greensand composition and those with a dominantly phosphatic composition. The greensand coprolites had various kinds of inclusions, including crustacean carapaces, bivalves, and squid hard bits* and such (interestingly without evidence of processing by teeth), suggesting consumers that were bottom-feeders, whereas the phosphatic coprolites were loaded with planktonic microfossils, suggesting filter feeding (or feeding on soft-bodied filter feeders) (Chin et al. 2008). The division is nuanced a bit more this time around, with an intermediate group with greensand embedded in phosphate. At the time, we could easily see that the phosphatic coprolites had spiral structures, as in shark excrement. As it turns out, at least some greensand coprolites also have internal tubular structures, although usually not as easy to spot (Duffy et al. 2026). What that means is that most of the coprolite producers therefore had spiral intestinal valves like sharks.

*I had the hardest time figuring out those squid-pen bits when I was working with the fossils. I thought they might be horseshoe crab tails.

There is one group that fits quite well for producing the whole range of coprolites while also not producing abundant body fossils: sturgeons. Sturgeons have spiral valves. The adults don't have teeth to shed, so they aren't leaving hundreds of potential fossils, and they aren't crushing or cutting prey. Rather, they are bottom-feeders that suck in prey items (and sediment). They are also big enough to produce the size range of greensand coprolites. Meanwhile, young sturgeons have small teeth (but they lose them), and they feed on zooplankton. Finally, there are sturgeons today that migrate from marine water to estuaries or shallow marine settings to feed on seasonal food blooms before spawning in freshwater (Duffy et al. 2026). They're about as perfect of "poopetrators"** as we could ask for.

**I still regret nothing! 

It's the migration part that particularly interested Duffy et al. Most behaviors are pretty darn difficult to fossilize clearly; or, if you want to get philosophical about it, all behaviors are reflected somewhere in anatomy, but most of them produce effects that are too subtle to pick out or are swamped by the effects of other behaviors. We can look for evidence for migration in certain large land animals such as mammoths by studying stable isotopes in bones. These animals were big enough to travel long distances (and thus drink water in places with different isotopic signatures, for example) and had nice big bones that allow sampling for time sequences. The Devon Island situation is not quite so convenient, but a few lines of evidence are suggestive (Duffy et al. 2026):

1. The microfossils in the coprolites suggest plankton blooms during long polar summer days, which is a pretty common high-latitude pattern; make hay when the sun shines, after all. Blooms of one kind of organism attracts populations of other organisms to consume them. This is a pretty simple way to establish a migratory pattern. (The flip side is that everything would clear out during the polar winter. Think of an Old West boom town, except for having annual booms and busts.) This is the strongest line of evidence in my mind.
2. There are lots of coprolites and not much skeletal evidence for what made them. This is interpreted as evidence for producers that only lived there part of the time. I think this is a bit weaker, as there are always going to be more turds than bodies, but it's worth noting.
3. Sturgeons, as likely producers for some large percentage of the coprolites, are known to be migratory today, and anatomically haven't changed much since the Late Cretaceous.
4. Finally, many of the vertebrates inhabiting the Western Interior Seaway and neighboring areas have distributions that are pretty darn cosmopolitan in this region, consistent with migratory patterns. Like the second point, I don't think this is as strong as the first, but again it's worth noting.

All in all, I like it, and I think it's great to learn new tricks from old turds. You never know what will turn up once you start looking at these humble fossils!

References

Chin, K., J. Bloch, A. Sweet, J.Tweet, J. Eberle, S. Cumbaa, J. Witkowski, and D. Harwood. 2008. Life in a temperate Polar sea: a unique taphonomic window on the structure of a Late Cretaceous Arctic marine ecosystem. Proceedings of the Royal Society B 275(1652): 2675–2685. doi: 10.1098/rspb.2008.0801.

Duffy, F., K. Chin, S. Cumbaa, and L. Wilson. 2026. Coprolite evidence for marine vertebrate migration in the warm Cretaceous Arctic. Historical Biology. doi: 10.1080/08912963.2026.2670771.

Your Former Friends The Ex-Titanosaurs

If you frequent The Compact Thescelosaurus, you may have noticed that several sauropods formerly placed in Titanosauria have been reclassified. (Apropos of nothing, I often wonder what the people who are browsing the sheets think when they see me active. Do they get excited to see what I'm working on? Or is it an inconvenience to whatever searching or sorting they're doing? Sometimes more appear while I'm working. I know it's just a coincidence, but it amuses me to think there's some kind of alert I don't know about that is issued when I show up.) This is not the first time this has happened. Back in 2019 Mannion et al. (2019) led me to move Baotianmansaurus henanensis and Dongyangosaurus sinensis to Titanosauria? (the question mark, the second-to-last refuge of a coward) and Jiangshanosaurus henanensis and Yongjinglong datangi out of Titanosauria altogether. After Beeston et al. (2024), the diamantinasaurs were also put at Titanosauria?. There has now been another purge of Early Cretaceous forms following Mannion and de Souza Carvalho (2026).

(Wait a second... Mannion et al. 2019, Mannion and de Souza Carvalho 2026, Mannion as third author on Beeston et al. 2024... Philip Mannion, stop taking my titanosaurs!)

In this case, the affected species were Hamititan xinjiangensis, Ninjatitan zapatai, and Volgatitan simbirskiensis. Although from different continents and formations, all three share one key characteristic: Supposed Early Titanosaur. SET is almost a curse. As soon as someone starts thinking a particular sauropod represents an Early Titanosaur, it is liable to transform, as if by perverse magic, into something else, and I don't recall that any have actually gotten back to being classified as titanosaurs. The main culprit seems to be that we just don't really have a good grasp on what somphospondyls were up to in their early years. (Well, that and the inevitability that the closer you get to the base of any lineage, the more generalized the taxa. And, perhaps, sometimes people might get too enthusiastic hoping for an Early Titanosaur and read a bit more into specimens than is warranted.) Somewhere in there is the lineage that led to titanosaurs, but until they established their monopoly, it's difficult to distinguish that thread from various also-rans, plus other sauropod groups that may be confused with them when you only have a couple of bones.

Backing up for a moment, Mannion and de Souza Carvalho (2026) is not primarily about reclassifying three disparate Early Cretaceous titanosaur-like sauropods. It's actually a redescription of Triunfosaurus leonardii, another victim of SET. In a minor upset for how these descriptions usually go, the type material doesn't turn out to be chimeric (the "middle-posterior caudals" are more likely anterior, but that's about as close as it gets to major anatomical reinterpretation). This then turned into an opportunity to look at the relationships of five SETs: T. leonardii, the three mentioned above, and Tengrisaurus starkovi. Running equal weighting (EQW) and extended implied weighting (EIW) against their data, they found the following placements:

• Hamititan was a turiasaurian under EQW and deeply nested in Titanosauria as a saltasauroid under EIW, which is a good trick. The authors in passing noted issues with its diagnosis and suggested it is not diagnostic at the genus level. For our purposes, I took the lowest common denominator and reassigned it to Eusauropoda.
• Ninjatitan was a diplodocid under EQW and a non-titanosaurian somphospondyl under EIW, hanging out with Chubutisaurus insignis. It isn't entirely comfortable in either position, and diplodocoids are known from the same formation, so the scrappy type could be chimeric and include both (Mannion and de Souza Carvalho 2026). For our purposes, I reassigned it to Neosauropoda.
• Tengrisaurus, boringly, was a clean titanosaur either way, although of course its exact placement in Titanosauria varied. No change was needed, and at the moment it is our oldest named titanosaur by default. Congratulations.
• Triunfosaurus was a non-titanosaurian somphospondyl under EQW and a basal titanosaur under EIW. Essentially it was either just inside or just outside the velvet rope, so hopefully that means there's a pretty good handle on it. No change was needed, as I already had it at Somphospondyli.
• Finally, Volgatitan was quite consistent... consistently a mamenchisaurid, which the authors found somewhat puzzling and not supported by the most robust of characters. A type specimen consisting of seven partial caudals also did not inspire great confidence in the results. Nevertheless, I moved it to Eusauropoda (I'm not using Mamenchisauridae until someone determines what Mamenchisaurus is and isn't).

Of these five, the two with the best cases to be recognized as true Early Titanosaurs are Tengrisaurus, which always ended up within it, and Triunfosaurus, on the doorstep. Although it's tempting to take one and plant a flag for the origin of the group, the situation is too messy for anything that neat, as noted by Mannion and de Souza Carvalho (2026). There's almost no record of somphospondyls in the Late Jurassic and earliest Cretaceous, when by definition they must have been around (because their sister group Brachiosauridae was around), and when we do start seeing them, they're all over the place (Mannion and de Souza Carvalho 2026).

References

Beeston, S. L., S. F. Poropat, P. D. Mannion, A. H. Pentland, M. J. Enchelmaier, T. Sloan, and D. A. Elliott. 2024. Reappraisal of sauropod dinosaur diversity in the Upper Cretaceous Winton Formation of Queensland, Australia, through 3D digitisation and description of new specimens. PeerJ 12:e17180. doi: 10.7717/peerj.17180.

Mannion, P. D., and I. de Souza Carvalho. 2026. Re-evaluation of the Early Cretaceous titanosauriform sauropod dinosaur Triunfosaurus leonardii from the Triunfo Basin, Brazil: implications for the initial radiations of Somphospondyli and Titanosauria. Zoological Journal of the Linnean Society 207(1): zlag073. doi: 10.1093/zoolinnean/zlag073.

Mannion, P. D., P. Upchurch, X. Jin, and W. Zheng. 2019. New information on the Cretaceous sauropods of Zhejiang Province, China: impact on Laurasian titanosauriform phylogeny and biogeography. Royal Society Open Science 6(8):191057. doi: 10.1098/rsos.191057.

Waukartus

There are so many outstanding fossil sites and productive formations (this link is just a sample) that it's not really feasible to be conversant with all of them and still have time for normal human interactions and responsibilities. I'd love to have that knowledge, but realistically I'd be doing pretty darn good if I only knew North America. A few of them on the linked list are not all that far from the Late Ordovician of the Twin Cities in time and space. We saw the Winneshiek Shale briefly when looking at the Decorah impact crater. Another example is the Waukesha Biota in southeastern Wisconsin, dating to the early Silurian. This assemblage came to mind because of the publication this month of Waukartus muscularis, a cousin to modern millipedes (Briggs et al. 2026).

The Waukesha Biota is found in basal dark shale of the Brandon Bridge Formation, otherwise composed of reddish dolomite. The productive beds are quite limited in distribution, described as extending about 350 m (about 1,150 ft, or not much more than a fifth of a mile) (Briggs et al. 2026). The strata were deposited at the toe of an erosional scarp at the beginning of a marine transgression (Briggs et al. 2026); think of them as akin to sedimentary filler. One of the things that's easy to forget when dealing with Paleozoic marine assemblages that are packed with shells and other hard parts, like our old friend the Decorah Shale, is that there were also a lot of things that just didn't fossilize well, particularly "worms" and arthropods that did not have the convenient durable exoskeletons of trilobites. You can find evidence of them through burrows and microfossils, but it's just not the same thing (it's hard to establish the taxonomic diversity, for one thing!). The Waukesha Biota is a Konservat-Lagerstätte, meaning the preservation is exceptional, and so we get to see those soft-bodied organisms. In fact, the Waukesha Biota is kind of Bizarro World as far as the Paleozoic is concerned, with uncommon brachiopods, crinoids, and mollusks, but abundant and diverse arthropods and "worms" (Wendruff et al. 2020). Preservation seems to have been greatly enhanced by microbial mats (Wendruff et al. 2020).

One of these otherwise unlikely fossils is the present subject, Waukartus muscularis. The genus name refers to Waukesha and limbs, which are an important part of the story, and the species name refers to the preservation of musculature (Briggs et al. 2026). This animal is not actually something that was just found; reports of this fossil animal go back to the 1980s. It was mentioned in the earliest papers on the Waukesha Biota (Mikulic et al. 1985a, 1985b) as a "myriapod-like animal". (Myriapoda is the group including centipedes and millipedes.) Specimens representing parts and sometimes counterparts of nearly three dozen individuals have been found. They top out at a little less than 3 cm (1.2 inches) long and perhaps 10% of that wide, and would have looked rather like chunky basic millipedes from a human's-eye-view. The body features a head, as many as 11 trunk segments (each segment looking deceptively like they were actually two parts), and a terminal segment. Despite the fairly large sample size, there isn't an especially well-preserved head, but there appears to have been four appendages on the head and eyes, likely on stalks. The terminal section is also poorly preserved but had a pair of blade-like projections on the underside (Briggs et al. 2026).

Some of the 35 individuals of Waukartus muscularis, including the holotype (A–E, part and counterpart) (Figure 1 in Briggs et al. 2026, which see for full caption; 5 mm scale in A–D and F, 2 mm for E, H, and I, 1 mm for G and J; ). CC-BY-4.0.

The limbs are the feature that has drawn the most comment. There is one pair of (rather stocky) limbs per trunk segment, unlike true millipedes, which have two. (Hence the scientific name for the group, "Diplopoda", meaning "double feet".) They are uniramous rather than biramous, the ancestral arthropod condition. A uniramous limb has "one branch", whereas a biramous limb forks into two branches. Many aquatic arthropods have biramous appendages and use one branch for locomotion and the other for respiration (think trilobites). Terrestrial insects, arachnids, and myriapods have uniramous limbs, and this has long been thought to be a specific adaptation to living on land (having feathery gill-like things on your limbs like trilobites did isn't quite as useful in the open air). Waukartus, though, was found in marine shales with nothing thought to be definitively terrestrial, and so is thought to have been marine as well. Therefore, uniramous limbs may not have been a terrestrial adaptation, at least in myriapods, but something that came in handy when the move occurred (exaptation, or pre-adaptation if you're older than you'd like to admit) (Briggs et al. 2026). It may have been respiring through its cuticle (Briggs et al. 2026), which is a neat trick you can get away with when you're a little less than 3 cm long.

Waukartus muscularis out for a stroll (head lower center, terminal segment upper left) (Figure 5 in Briggs et al. 2026, restoration by Leia Francis). CC-BY-4.0.

References

Briggs, D. E. G., J. C. Lamsdell, J. Kluessendorf, and D. G. Mikulic. 2026. A marine stem-myriapod from the Silurian Waukesha Lagerstätte, Wisconsin, USA: terrestrial traits pre-date the transition to land. Proceedings of the Royal Society B: Biological Sciences 293(2070). doi: 10.1098/rspb.2026.0131.

Mikulic, D. G., D. E. G. Briggs, and J. Kluessendorf. 1985a. A Silurian soft-bodied biota. Science 228: 715–717.

Mikulic, D. G., D. E. G. Briggs, and J. Kluessendorf. 1985b A new exceptionally preserved biota from the Lower Silurian of Wisconsin, USA. Philosophical Transactions of the Royal Society of London. B, Biological Sciences 311: 75–85.

Wendruff, A. J., L. E. Babcock, J. Kluessendorf, and D. G. Mikulic. 2020. Paleobiology and taphonomy of exceptionally preserved organisms from the Waukesha Biota (Silurian), Wisconsin, USA. Palaeogeography, Palaeoclimatology, Palaeoecology 546(109631). doi: 10.1016/j.palaeo.2020.109631.

Your Friends The Titanosaurs: Phosphatotitan khouribgaensis

The latest friendly titanosaur to come along hails from Morocco, making it the first named Moroccan titanosaur but not the first record. Other occurrences are mentioned here. For some reason it seems like the new names are always specimens that weren't included. Still lots of titanosaurs out there!

Fish Creek Canyon

You never know where interesting geology will turn up. I've been poking around the Twin Cities for years and I'm still coming across places I'd never dreamed were there. Case in point: today's pictorial topic, Fish Creek Canyon. I know the name sounds like a place you might find in Montana or Idaho, where the water is cold, the fish are biting, and the bears are waiting for you to wander away from your cooler, but this particular Fish Creek Canyon is just about where St. Paul, Maplewood, and Woodbury meet. It's not a huge canyon, and I don't imagine the fish are very large these days, but it's the kind of little hidden backyard gem that makes exploring worthwhile. I visited it with a friend back in late November, not long before our on-off winter hit the "on" switch for the first time.

Down in the valley on a great fall day, with Fish Creek in view.

Fish Creek is spread across two city jurisdictions. Maplewood has Fish Creek Natural Area, which is mostly the heights above 494 on the south and east and Highway 61 on the west. If you scout around, you'll find that this is one of a handful of small bluff-top parks between Battle Creek Regional Park and 494 overlooking 61. Adjacent on the north is the actual canyon of Fish Creek, which is within St. Paul and owned by Ramsey County Parks but apparently not organized at the moment, per se. The area is also largely within Mississippi National River and Recreation Area (so I'm also *officially* curious). The bluff-top lands are well worth walking around in their own right. There are no official access points to the canyon from the bluff-top, but it is possible to reach the canyon via social trails, leading to a steep descent into the narrow valley.

The view from the bluff, with the November afternoon sun on a clear day.

The canyon itself is not vast or deep. The water power that carved the original ravine has long since dwindled to a hoppable creek. And yet, what the creek has cut for itself is such a perfect little feature, complete with a miniature waterfall.

You're walking along, and then the creek disappears.

A miniature waterfall.

There ought to be gnomes.

It widens a bit going down (note that this feature was put in by people, presumably for water management, although it's not bad for aesthetics, either).

The rock that has been cut through here is the St. Peter Sandstone in its case-hardened form, producing the steep-sided, narrow slot. If you've driven Highway 61 near here, you may have noticed how the St. Peter sinks out of view for most of the stretch between Battle Creek and Camels Hump in Cottage Grove. It's still there, it just only shows its face sporadically, and this is the most picturesque place to find it. Interestingly, the location of Frederick Sardeson's "Highwood" collecting locality for St. Peter Sandstone fossils was supposedly a little north of here, about a mile and a half south of Battle Creek Park on 61. This works out to about the area where the ravines now occupied by Highwood Avenue and Springside Drive empty out, so our old friend was successfully trying his luck out here back in the 1890s.

Sometimes the St. Peter feels like crumbling, and sometimes it feels like holding a wall.

Afton graptolites revisited

There are two great lost fossil sites in the Twin Cities area. (The Brickyards don't count; they're not lost, they just aren't open to collection.) One is the Johnson Street Quarry, where workers cut into a bed in the Hidden Falls Member of the Platteville Formation that had unusually abundant echinoderms. As described in Sloan et al. 1987: 200, "Sardeson mined out a spot in this unit in the old Johnson Street Quarry in Minneapolis (now filled with garbage, and covered with Interstate 35) that produced about 20 specimens of the starfish Protopalaeaster narrawayi, several specimens of the crinoid Cremacrinus arctus (Fig. 16.2), edrioasteroids, cystoids, brachiopods, bryozoans, molluscs, and graptolites." This is slightly out of date; instead of a dump, there's now a Quarry Shopping Center with a Cub Foods, Home Depot, and Target, although even with all those options you can't get an edrioasteroid there anymore. Regardless of the exact character of the overburden, it seems unlikely that anyone will be doing any paleontological follow-up there anytime soon. The other locality is the Afton graptolite locality in the St. Lawrence Formation. We already had a post on why this locality was important; what I'm curious about is where exactly it was. A locality, even if "lost", had to have been *somewhere*, and apart from the scientific and historic interest, there very well could be similar fossils in rocks nearby. Indeed, Hughes and Hesselbo (1997) reported graptolites in the lowest strata of the St. Lawrence Formation in their Afton section, where collection may have postdated the road work that destroyed the classic location. For some reason, despite its “classic” nature, nobody ever saw fit to just put a pin on the map. What clues do we have?

New page: Quaternary vertebrate inventories (USA)

There's an electronic stack of topics I've considered for blog posts, and within that a subset earmarked "do before hanging up the keyboard". One of these, almost from the very beginning, was a bibliography of Quaternary vertebrate locality inventories like the kind Oliver Perry Hay pulled together. Being a detail hoarder, I've always loved having this kind of information. Maybe I'm the only one, but it's my blog and if I want to put together a page of old locality lists, I'll do it. One might think "If I have to find this information, all I need is the Paleobiology Database" and call it a day, to which I just smile and nod politely. Anyway, you can find the page here.

Pop quiz! Mammoth or mastodon?