Dr. Rob Dillon, Coordinator





Tuesday, June 18, 2013

Pleurocera canaliculata and the Process of Scientific Discovery


Looking back on my post of two weeks ago [1], I fear I may have left the impression that Rob Dillon is smarter than Calvin Goodrich, and indeed smarter than any other malacologist who has ever waded knee-deep into the rivers of the greater Midwest before or since.  That's probably not an uncommon failing of mine, but nothing could be farther from the truth.  In fact, the revelation that the gastropod populations we have been calling Pleurocera acuta for 189 years are actually the same as what we've been calling Pleurocera canaliculata for 192 only dawned on me after several years of being dope-slapped by the glaringly obvious.  And even then, I was only able to figure out the "Goodrichian" relationship between P. acuta and P. canaliculata backwards, through P. pyrenellum.

So to set the situation aright, I hereby offer an amendment to my essay of 3June13.  And in this retelling I propose to lay bare for my readership the agonizingly slow process by which this particular snail-guy's mind actually works.

The initial revelation that Goodrichian taxon shift might be even more dramatic than Calvin Goodrich himself realized first dope-slapped me in the back of my head eight years ago, as I scored a set of allozyme gels comparing pleurocerid populations from the upper Powell drainage on the Virginia/Tennessee border.  I wrote this the bottom of my data sheet late in the afternoon of October 4, 2005: "Good grief!  Pleurocera unciale and Goniobasis clavaeformis at the mouth of Indian Creek are a single random-breeding population!"  John Robinson [2] and I wrote this initial evidence of Goodrichian taxon shift into a gray literature report in 2007 [3].  I then sampled additional populations of nominal clavaeformis, unciale, acutocarinata, curta, carinifera and vestita across East Tennessee and North Georgia in 2008-09, ran a big mess of additional gels, and published the formal paper subsuming the genera Goniobasis and Elimia under Pleurocera in 2011 [4].

Meanwhile, my general survey of the freshwater gastropod fauna of East Tennessee was moving forward.  I spent at least a week or two working in the rivers and streams between Knoxville and Chattanooga each field season between 2007 and 2010, collecting (what I recorded as) P. clavaeformis (in its various phenotypic forms) in essentially every stream I visited for four years.  And incidentally, I also observed that the big river populations (historically referred to P. unciale or curta) sort-of petered out around Knoxville, to be replaced in the main Tennessee River by Pleurocera canaliculata.

I took a second dope-slap to the back of my noggin at approximately 4:30 PM on August 14, 2010.  Standing knee-deep in Savannah Creek (maybe 20 km N of Chattanooga) on that memorable day, it suddenly struck me that the snails crawling around my feet might be Pleurocera acuta.  At that point in my career, I had collected Pleurocera acuta up in Kentucky, Ohio and Michigan, but would not have expected a P. acuta population anywhere within 500 km of the stream where I was standing.  When I got back to my truck, I wrote this in my field notes: "(star)Wow(star)  Is this Pleurocera acuta??? My eyes are opened!"

So back in Charleston, I pulled four years of old samples off the shelves, dug the old literature out of the file cabinets, and (over the course of several weeks) pieced together the following line of reasoning.

Inference.  The snails I collected in Savannah Creek on 14Aug10 must be what Goodrich would have called "Pleurocera pyrenellum," not Pleurocera acuta.  Goodrich (1940) gave the range of P. pyrenellum as "tributaries of the Tennessee River in Morgan and Lancaster Counties, Alabama, and Walker County, Georgia" [5].  Walker County wasn’t much more than 20-30 km downstream from Savannah Creek.  So "pyrenellum" my snails would seem to be.

Therefore Pleurocera pyrenellum populations inhabit Tennessee tributaries further upstream than Goodrich realized.  Going back through my old collections, I recognized P. pyrenellum at 10-12 additional sites, extending up the Tennessee Valley all the way to the environs of Knoxville [6].  They were usually mixed with populations of P. clavaeformis, and (dazzled as I had been by the phenotypic variety of clavaeformis) I had simply missed them.

Revelation.  Travelling down the main Tennessee valley from Knoxville to Chattanooga, Pleurocera canaliculata begins to appear in the main river (red above) as P. pyrenellum begins to appear in the tributaries (pink).  This was the third major dope-slap I took during my long bewilderment with the East Tennessee Pleuroceridae, although I didn't record the date and time of the whacking.  The relationship between pyrenellum and canaliculata looks exactly like the Goodrichian relationship between unciale/curta and clavaeformis (blue above).

Hypothesis.  But the range of P. canaliculata extends throughout the greater Midwest, up the Ohio River all the way to Pittsburgh, while nominal pyrenellum populations are restricted to drainages of the Tennessee.  So might the small-stream form upstream from all those populations of canaliculata in Ohio, Kentucky, Indiana and Illinois be what everybody calls P. acuta in Yankeeland?

The point of this confession is that my colleagues and I were no more able to see the Goodrichian relationship between acuta and canaliculata standing on the shores of the Wabash River than Calvin Goodrich [7].  I flipped the whole story around when I told it two weeks ago, so that it "made sense," following the conventional practices of science.  I made the process of discovery look logical, even inevitable, when it most certainly was not.  I'm really a bonehead.  Ask my wife.

Allow me to close with another of my philosophical peregrinations.  My regular day-job is teaching majors-level genetics at a very ordinary liberal arts college of regional reputation.  It’s a big chore – there are three or four other members of my department doing the same thing.  All younger than I, of course.

So I organize the subject matter of genetics historically, the way I myself learned it, starting with Mendel, then Morgan, then Beadle then Avery and McLeod and McCarty and Watson and Crick and so forth.  And just when it begins to look like we might get anywhere near the present day, I go back to Hardy and Weinberg and finish with the "Modern Synthesis" of the 1930s and 40s.

Some of my colleagues tell me that undergraduate students "don’t get this."  They prefer to start their majors-level genetics sections with our modern understanding of the molecular basis of inheritance, because given the underlying mechanism, Mendel's 1866 results become a lot easier to explain.  And the kids prefer to focus on today's science, in any case.  They don't care about peas.

That line of argument is so self-evidently horrific to me that in my first draft of this essay, I left it dangling in the wind, without comment.  Like Stede Bonnet on the Charleston battery.

But I shall close with the simple observation that the K-12 teaching of science and the profession of science in higher education are two entirely different things.  In the former we are building palaces, and in the latter we are laying roads.  I trust that my readership has enjoyed this month’s brief foray into road building.  Next month, it’s back to the masonry.


Notes

[1] Pleurocera acuta is Pleurocera canaliculata [3June13]

[2] John D. Robinson was working on his MS in Marine Biology with me at the time.  He went on to earn his PhD at the University of Georgia and has recently moved to the Cornell area.

[3] Goodrichian Taxon Shift [20Feb07]

[4] Dillon, R. T., Jr. (2011)  Robust shell phenotype is a local response to stream size in the genus Pleurocera.  Malacologia 53: 265-277.  [pdf]
  • Goodbye Goniobasis, Farewell Elimia [23Mar11]
[5] Goodrich, C. (1940)  The Pleuroceridae of the Ohio River drainage system.  Occas. Pprs. Mus. Zool. Univ. Mich. 417: 1 -21.

[6] By the 8/2011 debut of the FWGTN website, I had documented 8 populations of P. canaliculata in the big rivers below Knoxville and 24 populations of nominal "P. pyrenellum" in the tributaries.  Since I've decided to save the nomen "pyrenellum" as a subspecies (more about that in a future post) the distribution of both shell forms is still apparent on the pdf  map available from the P. canaliculata page on the FWGNA site.

[7] In fact, the Indiana survey of Pyron et al. (2008) retained both P. canaliculata and P. acuta as quite distinct species.  For a pdf of Pyron’s survey, see:
  • The Freshwater Gastropods of Indiana [23Jan09]

Monday, June 3, 2013

Pleurocera acuta is Pleurocera canaliculata


I am pleased to report that the second paper in what I expect will be a continuing series on Goodrichian taxon shift in the North American Pleuroceridae has recently reached formal publication [1].  In this most recent installment, the team of Dillon, Jacquemin and Pyron uses a combination of genetic and shell morphometric techniques to demonstrate that the specific nomina acuta (Rafinesque 1824) and pyrenellum (Conrad 1839) are junior synonyms of Pleurocera canaliculata (Say 1821).  As “Pleurocerus acutus” was designated the type of the genus in ICZN Opinion #1195 of 1981 [2], this finding will be of more than the usual taxonomic interest.

Our analytical approach was patterned after that used by Dillon (2011) on the Pleurocera clavaeformis populations of East Tennessee [3].  We used allele frequencies at nine polymorphic allozyme loci to show that two populations of nominal P. acuta (from Indiana and Kentucky) and one population of nominal P. pyrenellum (from north Alabama) were each more genetically similar to the P. canaliculata population immediately downstream than any of the six populations was to any nominal conspecific.  We then used landmark-based morphometics to explore one of these “Goodrichian” shifts in greater detail, the historically important acuta-to-canaliculata transition found in the Wabash River of Indiana.

On at least two occasions Calvin Goodrich himself expressed doubts about the distinction between P. acuta and P. canaliculata, specifically mentioning the Wabash populations in 1937 [4].  So we borrowed 18 lots of historically-collected Wabash Pleurocera from the University of Michigan Museum of Zoology [5] and documented a significant correlation between shell robustness (as measured using relative warp analysis) and river size at point of collection, as estimated by catchment area.  The seven museum lots bearing more slender, high-spired shells (collected above Wabash River kilometer RK500) were identified by Goodrich as acuta, and the six lots bearing broader, heavier shells collected downstream from RK400 were identified as canaliculata.  Goodrich recognized a mixture of the two nominal species in the five lots collected in the 100 km immediately downstream from RK518, where the Tippecanoe River joins the Wabash, essentially doubling the catchment area (sites 8 - 12 in the map below).

Situations such as we document in the RK400-500 region probably explain why Calvin Goodrich, the man for whom Goodrichian taxon shift was named, did not synonymize acuta under canaliculata himself back in 1939.  The two nominal “species” sometimes seem to mix and retain a degree of distinctiveness.  The phenomenon appears even more dramatically in places where very small streams communicate directly with large mainstem rivers, such as that marked “12” in the map below.  It is not unusual to find snails bearing very slender, high-spired shells washed down into much broader, more robustly-shelled populations inhabiting the main river at spots such as this, looking for all the world like a pair of bona fide, reproductively-isolated species.

Way back in 2007 I defined "Goodrichian Taxon Shift," as “intraspecific variation in freshwater gastropod shell morphology along an environmental cline of such magnitude as to prompt the erroneous recognition of multiple nominal taxa” [6].  The phenomenon was conceived as narrowly applicable to freshwater snails, as Calvin Goodrich himself focused his own research.

But after some soul-searching, my colleagues and I have decided to broaden the concept and coin a new term for it.  So in our paper just published, we propose the term “cryptic phenotypic plasticity,” which we define as “interpopulation morphological variation so extreme as to prompt an (erroneous) hypothesis of speciation” [7].  We think that the sort of insight Calvin Goodrich brought to the study of pleurocerid snails in the 1930s and 1940s has the potential to make a contribution to the understanding of evolutionary processes beyond our small community of freshwater gastropod cognoscenti, even in the present day.

Goodrich himself was strongly influenced by the research of A. E. Ortmann on unionid mussels.  Ortmann described a correlation between river size and shell robustness way back in 1920, on the basis of which he sank quite a few unionid taxa into synonymy [8].  And Ortmann credited the idea to Wilson & Clark (1914) [9].  So in some sense it might not be fair to continue to call the phenomenon “Goodrichian” anything, as we forward cryptic phenotypic plasticity onward to evolutionary science as a whole.

The extent to which cryptic phenotypic plasticity may occur in the biotas of land and sea remains to be seen, however.  It seems unlikely to me that marine and terrestrial environments manifest themselves as variably to the populations that inhabit them as fresh waters, at least for the molluscan fauna with which I have any professional familiarity.  Were the Nucella (“dog whelk”) populations of the Pacific intertidal an (historic) example of cryptic phenotypic plasticity?  What component of the baroque taxonomy of Cerion populations in the Bahamas might be attributable to ecophenotypic plasticity in their shells [10]?  Our colleagues with expertise in environments such as these are hereby invited to take notice.

Notes

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

[2] Joe Morrison and the Great Pleurocera Controversy [10Nov10]

[3] Dillon, R. T., Jr. (2011)  Robust shell phenotype is a local response to stream size in the genus Pleurocera.  Malacologia 53: 265-277.  [pdf]
  • Mobile Basin III: Pleurocera Puzzles [12Oct09]
  • Goodbye Goniobasis, Farewell Elimia [23Mar11]
[4] Goodrich, C. (1937)  Studies of the gastropod family Pleuroceridae – VI.  Occas. Pprs. Mus. Zool. Univ. Mich. 347: 1 – 12.
Goodrich, C. (1939)  Pleuroceridae of the St. Lawrence River basin.  Occas. Pprs. Mus. Zool. Univ. Mich. 404: 1 – 4.

[5] The UMMZ was Goodrich’s home institution from 1924 to 1944.  All the pleurocerid lots curated during this period would have been identified by him, if not collected by him personally.  See “The Legacy of Calvin Goodrich.” [23June07]

[6]  Goodrichian Taxon Shift [20Feb07]

[7]  In the final publication our word “interpopulation” was emended to “intrapopulation” by the editor, without our notice.  Darn it.  Note to self: Check proofs carefully!

[8] Ortmann actually did this backwards.  In his (1918) “The Nayades of the Upper Tennessee Drainage, with Notes on Synonymy and Distribution” (Proc Am Phil Soc 57: 522) he wrote, “A large number of the ‘species’ described by Lea… and of those listed by Lewis… and subsequently described by various authors…are mostly synonyms.”  And as evidence he cited, “a rule…that one and the same shell assumes different shapes in the large rivers and in small streams and headwaters, a rule the existence of which will be shown elsewhere.”  Then in 1920 Ortmann published his “Correlation of shape and station in freshwater mussels (naiades)” (Proc Am Phil Soc 59: 269-312).

[9]  Wilson, C. B. & H. W. Clark (1914)  The mussels of the Cumberland River and its tributaries.  Bur. of Fisheries 781: 1 – 63.

[10]  I thought Woodruff & Gould pretty much answered this question back in the 1980s (eg, Biol. J. Linn. Soc 14: 389-416).  But Harasewych (Nautilus 126: 119-126) seems not to have been paying attention.