Sunday, August 20, 2017

The joys and agonies of other peoples' dissertations

In the hierarchy of information sources, dissertations and theses get a mixed reception. They rank higher than conference abstracts, which isn't saying much; after a few years, even a good abstract is kind of like a tooth that shows you duckbills or tyrannosaurs were present in such-and-such formation, but not much else. They get a bad rap for a number of reasons. Except in rare cases, their authors don't have vast experience writing scientific works, and the proofreeding and editorial processes are not generally as stringent. Speaking for myself, I worked on my thesis like a rat gnawing through concrete, and I wouldn't take back the accomplishment, but at the same time I could do a much, much better job today. I'd bet a lot of other people feel the same way about their graduate work. The quality of scientific review also varies, and is not considered to be of the same quality expected for a peer-reviewed paper, so questionable ideas can sneak through that would be weeded out in other publication venues. Finally, they can be difficult to nigh-impossible to find: the number of publicly available hard copies of a given dissertation or thesis is often one (1), so if the one you're looking for has not been scanned, cannot be purchased, and/or only exists in a university library hundreds or thousands of miles away, it might as well have been lost in the Library of Alexandria for all the good it's doing you. It's not hard to understand why published versions of graduate work are preferred.

If you can get access, though, a dissertation or thesis can be uniquely useful, provided you recognize the faults. They are great for detail. A peer-reviewed publication in a journal is usually on the order of twenty pages or fewer. A dissertation can be much, much longer, and contain the seeds of multiple papers plus stuff that was left out. For example, in my work on Saint Croix National Scenic Riverway geology, I've had access to many relevant graduate works through the University of Minnesota libraries, including dissertations by Clem Nelson (1949) and Bob Berg (1951) that basically reset the paleontology and stratigraphy for the St. Croix Valley. Although the two of them published several papers that incorporated parts of their graduate work (Nelson 1951, 1956; Berg 1953, 1956; Berg et al. 1956; etc.), these publications did not use everything. In particular, if you're just going by post-graduate publications, you only get some of Nelson's and Berg's stratigraphic sections.

Another thing that dissertations and theses are uniquely good for is documenting work from people who either ended up leaving the field or did not publish their results. Not publishing doesn't mean you did a terrible job; it can mean as little as your project's results didn't really have a venue. There are plenty of "Geology of the [insert name here] Quadrangle" projects that haven't been published in journals but are still valuable geologic records. There are a clutch of these at the University of Minnesota for the St. Croix Valley and nearby areas of Washington County for the late 1940s through the 1960s, and they've found their true homes as base data for Minnesota Geological Survey maps. One interesting example is Quaschnick (1959), a thesis on the geology of the Marine [on St. Croix] Quadrangle and the Falls Creek area. Although the Paleozoic rocks of Minnesota are not noted for their structural complexity, there is this one weird area in Scandia where faulting has preserved a thin sliver of St. Peter through Decorah between the expected Upper Cambrian rocks. You can visit part of this area in the Falls Creek Scientific and Natural Area, but don't expect to see much of the geology, thanks to vigorous plant growth. It was published on back in the 1920s (Peterson 1927, itself based on her 1924 thesis), but was still strange enough to warrant a revision. Unfortunately, Quaschnick's results were not formally published, so you have to go to the thesis. Another odd one: a few years ago, I noticed in geological maps of the St. Paul Park quadrangle and Washington County that an outcrop of Platteville Formation was plotted along 70th Street in Cottage Grove, but a personal inspection turned up nothing. This year, while going through Matsch (1962) on the St. Paul Park quadrangle, I found the outcrop described there; whatever it was in 1962, it's simply no longer visible today.

In the end, you still need to keep an eye open when using graduate work. Even apart from errors, often the author will make revisions if they do publish, sometimes substantial revisions. For example, when Nelson published his trilobites (1951), he named a number of new species. If you compare this publication to his dissertation, though, three of the new species have different names and seven new species mentioned in the dissertation were not named in the journal publication, instead being subsumed into other species. There are also some inconsistencies with specimen numbers. (Of course, dissertations and theses don't count for the purposes of formally naming species.)

These two specimens, UMPC (University of Minnesota Paleontological Collections) T6558b and T6558a left to right, were the paratype and holotype specimens of "Stigmacephalus marinensis" in Nelson (1949), but two years later he had second thoughts and identified them as the "paratype" and "holotype" of Stigmacephalus oweni var. B (Nelson 1951).


Berg, R. R. 1951. The Franconia Formation of Minnesota and Wisconsin. Dissertation. University of Minnesota, Minneapolis, Minnesota.

Berg, R. R. 1953. Franconian trilobites from Minnesota and Wisconsin. Journal of Paleontology 27(4):553-568.

Berg, R. R. 1954. Franconia formation of Minnesota and Wisconsin. Geological Society of America Bulletin 65(9):857-881.

Berg, R. R., C. A. Nelson, and W. C. Bell. 1956. Upper Cambrian rocks in southeast Minnesota. Pages 1–23 in R. Sloan and G. M. Schwartz, editors. Lower Paleozoic geology of the Upper Mississippi Valley. Geological Society of America, New York, New York. Guidebook Series Field Trip 2.

Matsch, C. L. 1962. Pleistocene geology of the St. Paul Park and Prescott quadrangles (Minn.). Thesis. University of Minnesota, Minneapolis, Minnesota.

Nelson, C. A. 1949. Cambrian stratigraphy of the St. Croix Valley. Dissertation. University of Minnesota, Minneapolis, Minnesota.

Nelson, C. A. 1951. Cambrian trilobites from the St. Croix Valley. Journal of Paleontology 25(6):765–784.

Nelson, C. A. 1956. Upper Croixan stratigraphy, Upper Mississippi Valley. Geological Society of America Bulletin 67(2):165-183.

Peterson, E. 1924. The Cambrian geology of the lower St. Croix valley: Osceola to Stillwater. Thesis. University of Minnesota, Minneapolis, Minnesota.

Peterson, E. 1927. Block-faulting in the St. Croix Valley. Journal of Geology 35(4):368–374.

Quaschnick, R. K. 1959. The geology of the Marine Quadrangle and the Falls Creek area, Minnesota. Thesis. University of Minnesota, Minneapolis, Minnesota.

Sunday, August 13, 2017

On the treatment of venerable names

The two major dinosaur news stories since Borealopelta have been the giant titanosaur Patagotitan and a shake-up in troodontid dinosaurs that, among other things, puts another stake into Troodon formosus. For some relevant thoughts on Patagotitan, I direct you to "On this occasion of receiving a new giant dinosaur" from a few years back. I'd instead like to spend some time on the dark and subtle art of taxonomy and how it relates to old names based on poor material.

Sunday, August 6, 2017


At last, the "Suncor ankylosaur" takes its place among the described! With the description of Borealopelta, I can cross off one more from "Coming Attractions". (That's four and a half down, with half-credit on Daspletosaurus, and eighteen and a half to go.) The first thing I recommend that you do is to download the description and the supplemental information from here.

Borealopelta, photo borrowed from Wikimedia Commons.

Sunday, July 30, 2017

The limitations of the layer cake

To be perfectly honest, we use simplifications for practically everything. Atoms don't really look like bunches of colorful spheres surrounded by smaller spheres orbiting them. The Earth is an oblate spheroid, which is close to but not quite a sphere. The planets of the Solar System don't have nice circular orbits centered on the center of the sun, lying in a flat plane. The need to simplify complex topics is obvious, both on the grounds of providing what someone needs to know to do something, and tailoring material to what someone can understand. There's a simple version for grade school kids, a more complex version for undergrads, and so on, until you're working professionally, where you've got very detailed models which are still abstractions, only closer (hopefully) to reality. One of these simplifications in geology is "geological formations as layer cakes", where formations maintain their thicknesses and are easily distinguished. The layer cake abstraction is most useful at a local level, in settings where deposition wasn't switching back and forth between different processes and sediment sources. For example, the Ordovician rocks of the Twin Cities fit pretty well. However, the cake starts getting funny-looking as you head into southeastern Minnesota. The photo below is of the Sogn roadcut, where some familiar rocks are exposed.

No, I don't know how to pronounce "Sogn".

At Sogn, though, what we would know as the lower half of the Platteville Formation is absent. Instead, the deposition of the Glenwood Formation persisted much longer (Sloan et al. 1987). Similarly, the Decorah Shale is at its thickest at the Brickyard in Lilydale, but going southeast, the upper part is replaced by the Cummingsville Formation. We can get these shifts in deposition from a number of causes. Sometimes you're looking at the boundary between two different modes of deposition shifting over time (such as a shoreline prograding or regressing). Sometimes there is a tectonic component, such as a basin subsiding. Sometimes the source of sediment changes or runs out.

Here's a more advanced example: the interval long known as the Franconia Formation and now known as the Tunnel City Group is divided into four parts in the St. Croix Valley. These are the Mazomanie Formation and three members of the Lone Rock Formation, from oldest to youngest the Birkmose, Tomah, and Reno members. We've met the Mazomanie Formation before; it's a quartz-rich light-colored very-fine- to medium-grained sandstone with abundant burrows and various forms of cross-bedding. (This of course is also a simplification, boiling down the essence of a rock unit that was deposited across some hundreds of thousands of years over parts of two states.) The Lone Rock Formation is a finer-grained, darker, wormier unit. The Birkmose Member is a greenish-gray very-fine to fine-grained sandstone, with a lot of feldspar and glauconite grains (glauconite being a green mineral that likes to form on marine bottoms with little sedimentation). The Tomah Member is a brownish-gray feldspar-rich siltstone and very-fine-grained sandstone with thin interbeds of gray-green shale. Finally, the Reno Member is similar to the Birkmose Member, but somewhat finer-grained and with better defined sedimentary structures. The Mazomanie Formation is a lateral equivalent to most of the Lone Rock Formation. While the Lone Rock Formation was deposited in an offshore setting centered in Minnesota, the Mazomanie was deposited under shallower conditions, and its sediment came from topographic highs to the north and east in central Wisconsin. The two formations intertongue over a wide geographic and vertical range. If you trace the zone of intertonguing, you're seeing deposition fluctuating over time, as pulses of uplift and erosion on the Wisconsin highs sent sand to the south and west. It doesn't look much like a layer cake, at least not a competent example. There are at least three major Mazomanie tongues, plus who-knows-what going on between Franconia and Marine-on-St. Croix. The Tomah seems to go quietly, but the Reno is engaged in some kind of geological close-quarters combat with the Mazomanie.

Schematic colorized version of St. Croix diagram in Berg (1951, 1954), with information from Quaschnick (1959) taken into account for northern Tomah extent. Thick black lines are reasonably certain stratigraphic contacts, thin black lines are inferred, red line are biostratigraphic boundaries, and brown vertical lines show the extent of the measured sections (with the locations identified below the lines). With Berg's Woodhill Member removed (Ironton Sandstone Member of the Wonewoc Sandstone), the Tunnel City Group here is around 52 m (170 ft thick). The rocks continue for a long way south of Afton, but there aren't any good outcrops for a while.

The concept of a simple planar formational contact is in itself a simplification. Sometime you get a nice flat contact between two units. Sometimes you get a contact with vertical relief, because the underlying formation was eroded into hills and valleys before the overlying unit was deposited. Sometimes the contact is arbitrary, because the lower rock type grades into the upper rock type. Sometimes the contact is arbitrary because the two units meet over a zone of alternating beds, due to the two types of deposition switching from time to time. This last kind is what we're seeing here between the Mazomanie and the Lone Rock formations, and if we could see through the ground to get a full picture of what is going on from Taylors Falls from Afton, the contacts would probably look "fuzzy" due to smaller and smaller-scale interbedding.

Finally, I've mentioned a few times how the Franconia Formation was problematic because of mixing rocks with biostratigraphy. Back in the day, people tried to define subunits based on trilobites. Berg (1951, 1954) pointed out that the zones don't actually follow the rocks. When your biostratigraphic formations don't correspond to rock types, it makes it a real pain to try to map. In addition, you have to have both a paleontologist who can identify the relevant species, and well-preserved examples of those species in the rocks you are studying. (Of course, it gets even worse if some significant number of the species you are dealing with are actually minute variations on a single species, but who would ever do that to you?) The red lines in the diagram show that the trilobite zones skew upward going north in the St. Croix Valley. This is not entirely surprising, when you get down to it: the Lone Rock Formation is notable for its glauconite content, which as mentioned is a sign of low sedimentation rate. The Mazomanie Formation lacks glauconite. I'm going to guess that the Mazomanie had a greater rate of sedimentation than the Lone Rock, which would naturally cause the zone boundaries to skew higher where there is more Mazomanie deposition.


Berg, R. R. 1951. The Franconia Formation of Minnesota and Wisconsin. Dissertation. University of Minnesota, Minneapolis, Minnesota.

Berg, R. R. 1954. Franconia Formation of Minnesota and Wisconsin. Geological Society of America Bulletin 65(9):857–881.

Quaschnick, R. K. 1959. The geology of the Marine quadrangle and the Falls Creek area. Thesis. University of Minnesota, Minneapolis, Minnesota.

Sloan, R. E., D. R. Kolata, B. J. Witzke, and G. A. Ludvigson. 1987. Description of major outcrops in Minnesota and Iowa. Pages 197–231 in R. E. Sloan, editor. Middle and Late Ordovician lithostratigraphy and biostratigraphy of the Upper Mississippi Valley. Minnesota Geological Survey, St. Paul, Minnesota. Report of Investigations 35.

Sunday, July 23, 2017

Graptolites of Afton

"Saw Clinton R. Stauffer, with a big rock in his hands
Says he found the graptolite site again
Gonna celebrate at Selma's Ice Cream Parlour
Send 'em off to Rudolf Ruedemann

[imitation of the sound of a graptolite]
Graptolites of Afton..."

(I apologize for nothing!)

University of Minnesota Paleontological Collection (UMPC) 4093, a particularly photogenic paratype of Callograptus staufferi, also depicted as Figure 5, Plate 55 in Ruedemann (1933).

Sunday, July 16, 2017

Follow-up: Pipestone National Monument, Scenella, Cylindrocoelia

Here's a little more information on a few enigmas from previous posts, with some additional photos from the University of Minnesota paleontological collections. First up is Pipestone National Monument's "Lingula calumet", then Scenella, and finally Cylindrocoelia minnesotensis.

Sunday, July 2, 2017

National Park Service dinosaurs

Here we are, three and a half years into this thing, and I haven't done a summary of National Park Service dinosaurs (non-avian variety)? No better time than the Fourth of July!

Time for another giant caption! This map shows the National Park Service units where non-avian dinosaur bones or tracks have been found in situ, or are historically associated with a park. 1. Yellowstone National Park; 2. Bighorn Canyon National Recreation Area; 3. Dinosaur National Monument; 4. Capitol Reef NP; 5. Arches NP; 6. Canyonlands NP; 7. Bryce Canyon NP; 8. Zion NP; 9. Glen Canyon NRA; 10. Rainbow Bridge NM; 11. Pipe Spring NM; 12. Petrified Forest NP; 13. Colorado NM; 14. Curecanti NRA; 15. Mesa Verde NP; 16. Chaco Culture National Historical Park; 17. Yukon-Charley Rivers National Preserve; 18. Denali NP & Preserve; 19. Wrangell-St. Elias NP & Preserve; 20. Katmai NP & Preserve (possibly); 21. Aniakchak NM & Preserve; 22. Big Bend NP; 23. Lewis and Clark National Historic Trail; 24. Springfield Armory National Historic Site.