Dr. Rob Dillon, Coordinator

Friday, May 7, 2021

Fun with Campeloma!

Editor’s Notes: The essay that follows is the third in a three-part series on the systematics and taxonomy of viviparid gastropods.  I’d recommend that you back up and read my essays of [9Mar21] and [5Apr21] before going forward, if you haven’t already.

It was subsequently published as: Dillon, R.T., Jr. (2023c)  Fun with Campeloma!  Pp 97 – 110 in The Freshwater Gastropods of North America Volume 7, Collected in Turn One, and Other EssaysFWGNA Project, Charleston, SC.

Hey boys and girls!  Want to match wits with an international team of 16 professional scientists?  Click the picture below and download our keen quiz #1!

The figure shows representative shells from the six species of North American Campeloma sequenced by Dr. Björn Stelbrink and his colleagues for their worldwide viviparid gene tree [1].  If you click it, you can download a pdf circular that also reprints the Burch/Vail dichotomous key [2] that Dr. Björn and his team used to identify the snails that bore those six shells.  How many can you get right?  The answers were hidden in our March essay, posted [9Mar21].

Keen Campeloma Quiz #1

So now kids, do you think you’ve got what it takes for an exciting career in the fast-growing field of freshwater gastropod Malacology?  Are you looking for a role model?  Look no further than the example set by our colleague Dr. Steven G. Johnson, currently Dean of the College of Science at the University of New Orleans [3].  The 13 papers he published on the evolutionary biology of Campeloma 1992 – 2007 are as good as any body of work on any malacological topic anywhere, ever.  I’ve listed my favorites at footnote [4] below.

Over that 16-year period, Dr. Steve brought a variety of techniques – allozyme electrophoresis, flow cytometry, mtDNA sequencing – to bear on the evolutionary relationships among dozens of populations of Campeloma, initially sampled throughout the eastern US, then subsequently focused on a sweeping arc of Atlantic and Gulf drainages from South Carolina to Louisiana.  He documented, quite thoroughly and beautifully, multiple origins of parthenogenesis in these populations – sometimes spontaneously by autodiploidy and sometimes by allotriploidy, through hybridization and backcrossing.

The grayscale background of the figure below shows the study area for a 1999 paper that Dr. Steve published with Eric Bragg [4].  He assigned these particular 31 Campeloma populations [5] to five specific nomina using the same Burch/Vail dichotomous key that you used for Keen Quiz #1, with an eye toward type localities.  So populations inhabiting the Wekiva River drainage must be C. floridense by definition, and populations inhabiting the Ochlockonee must be C. parthenum Campeloma limum was described from South Carolina, so that name would be appropriate for Atlantic drainages, and C. geniculum was described from the Flint River, so appropriate for Gulf drainages [6].  The assignment of Say’s specific nomen Campeloma decisum to populations west of the Mobile Basin was a vanilla call, by shell only.

It is interesting to notice that at no point in no body of water ever sampled by Dr. Steve Johnson during his entire 16 year career did more than a single nominal species of Campeloma occur sympatrically.  Where (sexual) C. limum populations and (sexual) C. geniculum populations have apparently come into contact, he found parthenogenic populations which he identified by a third name, C. parthenum.

Johnson & Bragg Fig 1, Stelbrink mapped in red

So, Johnson & Bragg sequenced the mitochondrial cytB gene for 1 – 5 snails sampled from each of these 31 populations.  Their mtDNA gene tree (maximum parsimony) showed six distinct C. parthenum sequences (and one for C. floridense) scattered haphazardly inside a large C. limum cluster, with C. geniculum on one outside branch and C. decisum on another.  Nominal decisum and nominal geniculum are apparently separated by a big, beautiful discontinuity of some sort at the Mobile Basin, but Dr. Steve’s data did not address whether that barrier is reproductive or merely geographical.

Which brings us back to Dr. Björn Stelbrink’s study of worldwide viviparid phylogeny with which we kicked off this essay [1].  Because Dr. Björn’s little sample of six nominal Campeloma species, marked in red on the map above, was centered on the Mobile Basin and North Alabama, right where Dr. Steve Johnson mapped his mysterious discontinuity.  And although Dr. Björn’s group only analyzed single snails from single populations for each of those six species in the old-school U1S2NMT3 gene tree we reviewed in March, reference back to his team’s supplementary table S1 reveals that sometimes they actually sequenced two.

So the map above shows samples from two populations of C. decisum, two populations of C. decampi, and two populations of C. regulare, as well as singletons for parthenum, geniculum and limum.  All nine of these individual snails were identified using the same Burch/Vail key everybody has used for 40 years and sequenced for three genes [7].  And this suggests a test.

If the morphological characters upon which the Burch/Vail key was based fairly reflect the evolutionary history of the Campeloma populations which bear those shells on their backs, one would expect the two snails that Dr. Björn identified as decisum1 and decisum2 to be the most similar to each other genetically, and ditto for the two snails identified as decampi1 and decampi2, and ditto for regulare1 and regulare2.

The figure below is a simple neighbor-joining network generated by the software inside GenBank, connecting the nine CO1 sequences deposited by Dr. Björn Stelbrink for his six nominal species of Campeloma, with the nearest neighbor for each of the six test snails identified by a red arrow.

Of the six nearest-neighbors, only one matched correctly.  The nearest neighbor of decisum1 was in fact decisum2 (98.5% sequence identity).  But oddly, the nearest neighbor of decisum2 was not decisum1, but rather decampi2, with which it demonstrated 99.1% identity.  And further, the nearest neighbor of decampi2 was not decampi1 (by a long shot), but mutually with decisum2.  The nearest neighbor of decampi1 was decisum1, as was the nearest neighbor of regulare1 as was the nearest neighbor of regulare2.  In fact, of the six CO1 sequences obtained for nominal species pairs, only the one sequence, that of decisum1, correctly matched its sister.

Neighbor-joining network from Stelbrink [1] COI sequences

Turning now to the two nuclear genes sequenced by Dr. Björn’s team, very little variance was apparent among the nine sequences for either, none correlating with the taxonomy.  In the histone H3 data set (328 bp), the pair of decisum sequences matched each other and the singleton parthenum sequence.  That group of three differed by a single A/G transition at position 262 from a group that contained both decampi sequences and both regulare sequences.  The limum sequence and the geniculum sequence were both a bit more distinctive, differing from each other (and from the decisum group) by four transitions.

For 28S data set (1,058 bp) the two decisum sequences were again identical to each other, to the parthenum, and to one of the regulare.  The other regulare differed by a single A/G transition at residue 535 and the geniculum differed by a single C/T transition at residue 833.  The two decampi were also again identical to each other and (this time) to the single C. limum sequence.  That set of three differed from the decisum cluster by a two-nucleotide indel at positions 835-6, which is surprising, especially given the geography, and I think significant.  There most certainly is some genetic structure to this set of 3x9 DNA sequences, but the 19th-century taxonomy is not capturing it.

Hey kids, are you ready for some more fun?  Now’s the time to try Keen Quiz #2, shown on page 3 of that same pdf circular you downloaded at the beginning of this essay.

Can you find the three secret-decoder Campeloma hidden in the weeds?  Now can you identify those three shells using the Burch/Vail key?  Circle each shell and write what you think it is in the spaces provided.  Answers below, no peeking!

Not uncommonly, when I am casting about for larger analogies to apply to the messy evolutionary biology of freshwater gastropods, I find myself looking toward the botanical, rather than to the zoological [8].  And in the case of North American Campeloma, I have found inspiration in the dandelions.

Keen Campeloma Quiz #2
The reproductive biology of dandelions is as diverse as one could possibly imagine – outcrossing, selfing, parthenogenetic cloning, sexual/asexual cycling, everything.  An elaborate taxonomy built up in nineteenth-century Europe to describe all the morphs, forms, subspecies and sections of dandelions, but today, most botanists refer the entire yellow-blooming, blowball-sprouting mess to Taraxacum officinale, because after a couple hundred years of squinting at them and screwing around with them, all dandelions pretty much look alike.

The widespread incidence of parthenogenesis in North American populations of the viviparid genus Campeloma voids the biological species concept and necessitates a retreat to the morphological.  And since (judging from the work of Johnson and Stelbrink) there are apparently no consistent phenotypic characters, shell morphological or otherwise, by which Campeloma populations can be distinguished, the FWGNA Project will refer all populations of Campeloma to the oldest available specific nomen, Campeloma decisum (Say 1817).

But how about inconsistent characters?  Nothing in the paragraph above should be interpreted as foreclosing the recognition of C. decisum subspecies – geographically separate and morphologically different forms – even if they intergrade, even if there is no genetic basis for the distinction [9].  The shell morphology of Campeloma populations most certainly does vary regionally.  Indeed, big rivers of the American interior are often inhabited by populations of Campeloma bearing distinctly heavy shells, which have traditionally been identified as C. crassulum Rafinesque 1819, almost certainly a case of cryptic phenotypic plasticity [10].  The FWGNA Project recognizes these populations at the subspecific level, as C. decisum crassulum, just as we recognize heavily-shelled Pleurocera canaliculata canaliculata in those same big rivers, and more gracile P. canaliculata acuta in the little streams that feed into them.

And certainly Call’s (1886) floridense could be saved as a subspecies name for C. decisum with brown apertures, right?  And Binney’s (1865) decampi also seems a likely candidate for retention at the subspecies level.  Although I have no personal observations to contribute, the figures and photos I have seen (e.g., shell #1 way up above) seem to suggest that Campeloma populations in North Alabama may indeed bear shells that are atypically slender and higher-spired, likely for ecophenotypic reasons we do not understand.  And the nomen “decampi” has appeared in recent literature connected to some conservation concerns [12].

As we have often pointed out when faced with analogous situations in the North American pleurocerids, the Latin nomina assigned to biological populations may bear an important indexing function, as well as evolutionary significance.  So, in addition to crassulum and floridense and (perhaps) decampi, it would be a shame to see the names that Dr. Steve Johnson used for his important works – geniculum (Conrad 1834), limum (Anthony 1860), and parthenum Vail 1979 – forgotten by the google machine.  Those names have been useful to index Campeloma populations across the southeastern United States for quite a few years now.

But here’s a big problem, kids.  And maybe you can help!  Let’s look at how you did on Quiz #2.  Those secret-decoder shells didn’t have brown apertures, did they, so they are not floridense.  Did you think that their whorls have angled shoulders, or are their shoulders rounded, or do they have no shoulders at all?  Angled?  So are those shells broadly ovate or narrowly ovate?  Are they Campeloma limumCampeloma geniculum?  Did any of you guess Campeloma decisum for any of the secret-decoders?

Surprise!  All three of the shells hidden in the dandelion patch above are topotypic Campeloma decisum, collected from a pond in Philadelphia’s Fairmount Park, along the banks of the Schuylkill River, where Thomas Say almost certainly explored in 1817.  The snails that bore those three shells were C. decisum, by definition.

That makes them what scientists call a “control,” and what kids like us call super-duper secret-decoders.  Every other Campeloma population, bearing every other shell shape, form, or size, that anybody has ever seen, are (scientifically-speaking) unknown.  That includes all nine of those populations identified by Dr. Björn Stelbrink’s team we had fun with in Keen Quiz #1, and all 31 of those populations studied by Dr. Steve Johnson.  If Dr. Björn’s shells and Dr. Steve’s shells look like the secret-decoders, their correct identification is C. decisum, by definition.  Only if they look different, can they be identified as anything else.

Did Jack Burch and Virginia Vail look at secret-decoders first, when they were making their 1982 dichotomous key?  I don’t know.  The University of Michigan Museum of Zoology does (indeed) hold a couple lots of Campeloma decisum collected from the Schuylkill River in Philadelphia, so it is possible.  But I don’t know about you, kids, I cannot make the Burch/Vail key work at all.  I get messed up early, way up at couplet 14, because I really think that at least some of our secret-decoder shells demonstrate “angled” shoulders, which sends me down the wrong path.

So it would be great to save Dr. Steve Johnson’s limum, geniculum, and parthenum as subspecies names, it really would.  But even setting aside the Burch/Vail key, I cannot distinguish limum collected from its type locality (“South Carolina”) and control decisum from Philadelphia.  Could I distinguish geniculum or parthenum from decisum/limum?  I don’t know that, either [13].  I don’t have control collections from the Flint River or Lake Talquin, or really anywhere in the Gulf drainages of Georgia and Florida.

Boys and girls, I’m going to leave this challenge to you all, the next generation of malacologists. Here’s what I want you to do.  Mail your completed quiz sheets to “Fun With Campeloma,” P.O. Box 31532, Charleston, SC 29417.  Don’t forget to include your name and address and $1.25 for postage and handling.  No, I won’t grade them – I don’t know the right answers myself.  But all entries will be entered into a random drawing.  And one lucky winner will get a genuine FWGNA deputy’s badge and a license to classify the subspecies of Campeloma decisum throughout North America any way he or she wants.

But one last warning, kids, before we sign off this month.  You may hear some grown-ups asking why we have to fall back to an 18th century morphological species concept, as I so boldly asserted ten paragraphs above.  Accepting that the biological species concept is void for Campeloma, those grown-ups may be suggesting that we move forward to the phylogenetic species concept, or one of those other recent concepts based on gene trees.  Don’t all the genetic data we have reviewed this month suggest some sort of evolutionary structure to these populations?  Surely all the Campeloma populations spread across half of North America aren’t equally related to each other, are they?  DNA sequences and fancy tree-generating algorithms are a more scientific way to name snail populations than some kid’s subjective judgement calls on fat brown shells, right?

Be careful around grown-ups like that, boys and girls!  They’re not malacologists, they’re phylogenetic systematists!  We’re going to learn a lot more about phylogenetic systematics in the next few months, and why ideas like that don’t hold water.  So stay tuned!


[1] Stelbrink, B., R. Richter, F. Köhler, F. Riedel, E. Strong, B. Van Bocxlaer, C. Albrecht, T. Hauffe, T. Page, D. Aldridge, A. Bogan, L-N. Du, M. Manuel-Santos, R. Marwoto, A Shirokaya, and T. Von Rintelen (2020)  Global diversification dynamics since the Jurassic: Low dispersal and habitat-dependent evolution explain hotspots of diversity and shell disparity in river snails (Viviparidae).  Systematic Biology 69: 944 – 961.

[2] This is a difficult work to cite.  J. B. Burch's North American Freshwater Snails was published in three different ways.  It was initially commissioned as an identification manual by the US EPA and published by the agency in 1982.  It was also serially published in the journal Walkerana (1980, 1982, 1988) and finally as stand-alone volume in 1989 (Malacological Publications, Hamburg, MI).

[3] Rule #1 is, stay out of museums! Those musty cabinets of dusty shells will kill your scientific career faster than a confession of the Apostle’s Creed.  And Rule #2 is, never return emails from anybody who ever has.

[4] Johnson, S. G. 1992. Spontaneous and hybrid origins of parthenogenesis in Campeloma decisum (freshwater prosobranch snail). Heredity 68:253-261.  Johnson, S.G., R. Hopkins, and K. Goddard. 1999. Constraints on elevated ploidy in hybrid and non-hybrid parthenogenetic snails. Journal of Heredity 90: 659-662.  Johnson, S.G. and W. Leefe. 1999. Clonal diversity and polyphyletic origins of hybrid and spontaneous parthenogenetic Campeloma (Gastropoda: Viviparidae) from the southeastern United States. Journal of Evolutionary Biology 12:1056-1068.  Johnson, S.G. and E. Bragg. 1999. Age and polyphyletic origins of hybrid and spontaneous parthenogenetic Campeloma (Gastropoda: Viviparidae) from the southeastern United States. Evolution 53:1769-1781.  Johnson, S.G. 2000. Population structure, parasitism and survivorship of sexual and parthenogenetic Campeloma limum (Gastropoda: Viviparidae). Evolution 54:167-175.  Johnson, S. G., 2005. Mode of origin differentially influences the fitness of parthenogenetic freshwater snails. Proc. Roy. Soc. Lond. B. 272: 2149-2153.  Johnson, S. G. 2006. Geographic ranges, population structure and ages of sexual and parthenogenetic snail lineages. Evolution 60:1417-1426.

[5] The 1999 paper by Johnson with W. R. Leefe surveyed 55 populations.  The Johnson & Bragg paper from which the figure was extracted retained the Johnson & Leefe numbering system but focused on a slightly-less-cluttered 31-population subset.

[6] Johnson never sampled the Flint River proper, up in Georgia, where it comes in close contact with the Ocmulgee R., which drains east to the Atlantic.  This would certainly have been intellectually fascinating, and might well have clarified Campeloma taxonomy substantially, and the omission drives me nuts, absolutely nuts.

[7] The Stelbrink team sequenced mitochondrial CO1 and nuclear 28S and H3.  They did not (alas) sequence mitochondrial cytB, so their data cannot be integrated with that of Steve Johnson.

[8] The most obvious example is my “USR” model for life history evolution in the freshwater mollusks, which I patterned after J. P. Grimes” CSR model for the plants: Competitors, Stress-tolerators, and Ruderals.  That idea was the unifying theme of my (2000) book for Cambridge University Press.  I thought it would make me rich and famous.  I wonder why not.

[9] To refresh your memory on the definition of the word, “subspecies,” see:

  • What Is a subspecies? [4Feb14]
  • What Subspecies Are Not [5Mar14]

[10]  Quoting Dillon and colleagues [11], “phenotypic plasticity may be considered cryptic when intrapopulation morphological variance is so extreme as to prompt an (erroneous) hypothesis of speciation.”  To refresh your memory:

  • Pleurocera acuta is Pleurocera canaliculata [3June13]
  • Pleurocera canaliculata and the process of scientific discovery [18June13]
  • Elimia livescens and Lithasia obovata are Pleurocera semicarinata [11July14]

[11] Dillon, R. T., S. J. Jacquemin & M. Pyron (2013) Cryptic phenotypic plasticity in populations of the freshwater prosobranch snail, Pleurocera canaliculata.  Hydrobiologia 709: 117-127. [PDF]

[12] Haggerty, T.M. and J.T. Garner (2008)  Distribution of the armored snail (Marstonia pachyta) and slender Campeloma (Campeloma decampi) in Limestone, Piney, and Round Island Creeks, Alabama.  Southeastern Naturalist 7: 729-736.  Haggerty, T.M., J.T. Garner and L. Gilbert (2014)  Density, demography, and microhabitat of Campeloma decampi (Gastropoda: Viviparidae). Walkerana 17: 1 – 7.

[13] But reading all the way down through the Burch/Vail key to couplet #19, one gets the strong impression that even Virginia Vail could not distinguish her Campeloma parthenum from Campeloma decisum.  The basis for the decisum/parthenum distinction is geographical only.  If you find it in the Ochlockonee River, it’s parthenum.  Otherwise, it’s decisum.  So parthenum is an allotriploid hybrid of nominal limum and geniculum, which Burch/Vail distinguished from decisum/parthenum by that problematic shoulder-shape couplet #14?   So two nominal species with angled shoulders hybridize to form populations with rounded shoulders?  Hmmm.  That’s as far down this rabbit hole as I care to go, even in a discursive footnote.