A "Magical" Experience Identifying Fossils

Old Lodge Skins (after lying down and waiting to die, is aroused by rain):  Am I still in this world?

Jack Crabb:  Yes, Grandfather.

Old Lodge Skins:  I was afraid of that.  Well, sometimes the magic works.  Sometimes it doesn't.

From the movie Little Big Man (1970)

I recently worked to identify the genus and species of two fossilized shells in my collection.  I thought that keeping close track of the steps I followed in this process would make it more directed, less random than it usually is for me.  I was wrong.  This post is an account of the path I followed.

I believe the two macro attributes of any fossil probably most important for identification are (1) shape and (2) geological provenience (that is, where the fossil was found).  Both of these attributes may feed into each other repeatedly during the identification process.

Here are the two fossils I sought to identify.  The dorsal views of these shells (first picture) are labelled 1a and 2a; the ventral views (second picture showing the apertures) are 1b and 2b.  These specimens are referred to below as #1 and #2.

Pictured below is shell #2 labelled with some of the general conchology terminology used in the literature.  For this illustration of terms, I have relied on Percy A. Morris' A Field Guide to Shells of the Atlantic and Gulf Coasts and the West Indies (third edition, original copyright of 1947, renewed 1975, endpapers).

I knew from the outset these two specimens were gastropod or snail shells because of their shape and form.  To deconstruct this recognition, which occurred instantaneously, I would say that my decision tree branched from animal (not plant) to shell (not tooth or bone) to gastropod (not a mollusk bivalve or other type of mollusk).

To take the decision tree down to the level of genus and species of the gastropods at hand, I had to address a second main attribute - geological provenience.  These specimens came to me as part of a small clutch of fossil shells collected by a friend on the western shore of the Chesapeake Bay in Maryland.  This is a site where, I believe, fossils from the Windmill Point member of the St. Marys Formation can be found.  These particular fossil shells, therefore, are likely to have come from a geological strata which is dated from roughly 11.6 to 7.2 million years ago in the Late/Upper Miocene epoch.  (See Paleobiology Database entry for Busycotypus.)  A study of the stratigraphy of the site posited that back in the Miocene, this was "a warm-temperate marine environment" with, some "subtropical influences."  (Lauck W. Ward and George W. Andres,  Stratigraphy of the Calvert, Choptank, and St. Marys Formations (Miocene) in the Chesapeake Bay Area, Maryland and Virginia, Virginia Museum of Natural History Memoir Number 9, 2008, p. 51.)

At this stage in the identification process, I turned to the literature on the fossils of the St. Marys Formation, including the Windmill Point member.  Paleontologists have identified many members of the molluscan assemblage that populated that location during the Miocene.  The literature I consulted included:  

• Timothy Abbott Conrad, Monograph of the Genus Fulgur:  Notes on Shells, Proceedings of the Academy of Natural Sciences of Philadelphia, Vol. 6 (1852 - 1853), 1853
• George C. Martin, gastropoda discussion in Miocene, Maryland Geological Survey, 1904, and Miocene Plates, Maryland Geological Survey, 1904
• Harold E. Vokes, et al., Miocene Fossils of Maryland, Bulletin 20, Maryland Geological Survey, 1999
• Lauck W. Ward, Molluscan Biostratigraphy of the Miocene, Middle Atlantic Coastal Plain of North America, Virginia Museum of Natural History Memoir Number 2, 1992

Based, primarily on perusal of the illustrations in these sources, I found two likely candidate species:  Busycotypus coronatum and B. rugosum.  These names are those provided by Ward in Molluscan Biostratigraphy.  There have been changes in these names over time (e.g., Busycotypus coronatum began its taxonomic life as Fulgur coronatus).  Except in direct quotations from these sources, I usually refer below to each of these species as B. coronatum or B. rugosum, regardless of the name actually used by the author.

I must stress that I certainly did not exhaust the available taxonomic literature on these species.  So, any conclusion reached in this post, should be weighed against the possibility that a full sweep of the literature would have led me elsewhere in the identification process.

The species' descriptions in the literature I did consult proved singularly challenging with inconsistencies, widely varying levels of detail, and conflicting assertions.  Further, the illustrations accompanying these descriptions often failed to capture, at least for me, the differences purportedly distinguishing the two species.

The original identifier of many of the mollusks found in the Windmill Point member, including the gastropod genera and species listed above, was paleontologist Timothy Abbott Conrad (1803-1877), a prolific discoverer and namer of species.  (See Ward and Andres, p. 50-51.)  So, I thought it would most appropriate to consider first what he wrote about these two species in the report listed above.  (In addition, the text written by paleontologist George Martin (also cited above) includes excerpts from several of the many early to mid-19th century descriptions Conrad penned.)

According to Conrad (1853), the shell attributed to B. coronatum is shaped liked a spindle ("fusiform") and swollen ("ventricose"), in contrast B. rugosum is pear-shaped ("pyriform").  B. coronatum has fine lines circling the shell while B. rugosum's shell is marked by coarse lines.  The spire in the former is "short," while that of the latter is "prominent."  The nodes on B. coronatum are "elevated, compressed," while those of B. rugosum are "obtuse."  Their sizes differ markedly.  Conrad stated that B. coronatum's maximum length (the conchology literature often identifies this measurement as "height") is six inches while the length of B. rugosum is just three inches.  (p. 317)

Did Conrad's descriptions of these two species help me?  Less than I hoped.  Some of his observations suggested to me that #1 might be B. coronatum and #2 B. rugosum (e.g., #1 appears to be more spindle-like and #2 perhaps more pear-like; the spire of #1 appears shorter than that of #2).  But the size measurements he gave pointed in the opposite direction.  Of my two specimens, #2 clearly has a greater length than #1, consistent with an identification of B. coronatum.  To further complicate the identification, the differences in some of the attributes he highlighted aren't clear from language he used, particularly regarding the nodes.  And, in their present condition, elements such as the lines circling the body of the shells aren't clear enough to help in the identification process.  Fossil wear is a factor that can thwart an effort to identify the genus and species of a fossil.

Consideration of another paleontologist's comparison of these two species also highlighted the difficulties I faced.  Here is an excerpt from the publication by Harold Vokes (1999) in which he briefly contrasted these species.  He used the genus name Busycon for both and, significantly, considered B. rugosum to be a variety of B. coronatum (that is, not a separate species and at a taxonomic level lower than the species).

The large short-spired shells of Busycon coronatum (Conrad) ... are common in the St. Marys fauna associated with the variety rugosum (Conrad) . . . which differs in having coarser spiral ornamentation and more numerous and heavier elongated nodes on the shoulder of the whorl.  (p. 25)

The everyday adjectives in this description are a bit unclear to me.  Consider "large short-spired."  Is there a comma missing?  I assumed he was characterizing the shell as "large" and the spire as "short."  But that didn't get me very far.  Is B. coronatum bigger than B. rugosum?  Or is it simply, in the scheme of things, "large," though B. rugosum might be bigger still or possibly smaller?  Unfortunately, Vokes provided no measurements for these species.  I concluded that he thought the spire of B. coronatum to be shorter than that of B. rugosum.  Conrad would agree with this characterization of the B. coronatum spire.

To help address my confusion regarding Vokes' treatment of overall shell size and tallness of spire, I turned to his illustrations.  I discovered that his illustrations of these two species are actually from the earlier Maryland Geological Survey publication by George Martin, specifically, plate XLVI (Miocene Plates).  From that plate, Vokes chose to use just the illustration of the dorsal side of B. coronatum and only that of the ventral side of B. rugosum.  Why?  I don't know.  I thought the full plate from Martin (see below), showing dorsal and ventral views of both shells, would be potentially more informative.  The two views at the top of the plate (1a and 1b) are of B. coronatum and those at the bottom of the plate (2a and 2b) are B. rugosum (which, as noted, Vokes considered a variety of B. coronatum).

Looking at the shells depicted in Plate XLVI and assuming they are drawn to scale (a significant assumption), it seemed obvious to me that B. coronatum (1a and 1b), as illustrated, is larger than B. rugosum (2a and 2b), certainly consistent with Conrad's description given above and presumably Voke's as well.  Yet, according to Martin's own text describing these species, that understanding would be incorrect.  In his description of B. coronatum, he provided these measurements:  length, 130 mm (5.1 inches); diameter, 75 mm (3.0 inches).  (Presumably the distance across the widest part of the last or body whorl is the diameter.) For B. rugosum, he cited these:  length, 170 mm (6.7 inches); diameter, 90 mm (3.5 inches).  So, despite the illustrations, B. coronatum, according to the text, is smaller than B. rugosum.  I noted also how different Martin's measurements are from those given by Conrad (see above).

So, at this juncture, I wondered whether I assume that Martin got the measurements wrong, inadvertently flipping them between species.  Ultimately, though, I concluded that size, with respect to my specimens, might be a red herring.  Martin included a quotation from an 1843 report by Conrad that reminded me of a serious issue that should be addressed in the identification process.  Conrad noted that the shell of B. rugosum is shorter and more swollen than that of B. coronatum "when adult."  (As quoted in Martin, p. 182.)  This is an important consideration in the identification process:  is the fossil in question from an adult organism or some younger version of that species?  Size differences between specimens may reflect different maturation stages of the same species.  Frankly, given that neither of my specimens comes very close to the measurements Martin provided, I concluded that neither is from an adult and, therefore, size in this case probably wasn't dispositive.

There were still other aspects of Vokes' comparison between these species that I needed to consider.  What about the height of the spires on the specimens illustrated in Plate XLVI?  I thought the spire of B. rugosum as illustrated appeared to be shorter than that of B. coronatum, contradicting Vokes.  As already noted, I thought my specimen #1 had a shorter spire, suggesting that #1 might be B. coronatum and #2 B. rugosum.  Still hardly persuasive as to identification.

Vokes was as confounding as Conrad on the nodes of these shells.  Vokes stated that B. coronatum had more nodes than B. rugosum, and that its were "heavier."  I found the adjective "heavier" unclear.  So, I turned to considering the number of nodes, a feature that seemed likely to be fairly easy to assess.  Hmmm, not likely.  Examination of Plate XLVI suggested to me that B. rugosum as depicted had more nodes, which didn't fit with Vokes' description.  When I examined my fossils (#1 and #2 in the first photos in this post), I found it nearly impossible to count those on one of these (#1) because the specimen is worn.  As already noted, wear and tear on a fossil constitutes an important aspect of fossils that may confound their identification.

Finally, one of the discussions of these two species raised an additional and rather fundamental factor that should be considered in the identification process.  Similar fossils might not be different species at all, but, instead, reflections of naturally occurring variation, or a fossil might be a transitional stage in the evolution from one species to another.  These possibilities may make drawing sharp distinctions among fossils a fool's errand.  Indeed, B. coronatum and B. rugosum are, as Lauck Ward observed, "related."  He wrote:

Busycotypus coronatum first is differentiated from the parent stock (B. rugosum) in the Little Cove Point beds of the St. Marys Formation.  (p. 132)

I understood Ward (with the phrase "parent stock") to be positing that B. coronatum evolved from B. rugosum, which is consistent with the former appearing after the latter in the fossil record.  But that perplexed me because, as already discussed, Martin and Vokes thought rugosum to be a variety of B. coronatum.  Doesn't that suggest B. coronatum was the predecessor?  Maybe I've misunderstood "variety" all these years.

At that stage, I considered whether, from the outset, my effort to distinguish between the worn fossils of apparently immature and evolutionarily related animals that lived at the same time had been doomed to failure.  To salvage something from this taxonomic journey, I took refuge in what might be considered a cop-out, though I thought it a safe harbor.  I concluded that both specimens are from the genus Busycotypus, but couldn't go any further than that.  I applied the following taxonomic label to both:  Busycotypus sp.

Not a very happy conclusion.  Though it's one that paleontologist George Martin presumably would have understood.  Considering B. rugosum to be a variety of B. coronatum, he noted that it is "sometimes difficult" to separate the two, and, curiously, he added, "it is not essential that it should be done."  (p. 181)

Amen to that.