Dichotomous Dichotomous Keys

Last month [1] we reviewed the Gastropoda chapter contributed by Christopher Rogers to the new Fourth Edition of Thorp & Covich’s Keys to Nearctic Fauna [2].  The bottom line was, “Buy this book.”  And among the several lines of support I offered for this recommendation was a somewhat enigmatic observation to the effect that the new fourth-edition dichotomous key complements, but is in many cases strikingly different from, the old third-edition key.  What did I mean by that?

In two words, Christopher’s new key is evolutionary, whereas the old one was ecological.  Christopher has divided the North American freshwater gastropods phylogenetically, designing his dichotomous key to branch as a phylogenetic tree might branch.  The older key divided the fauna functionally, according to morphological adaptation.

Broad/flat vs. Narrow/filiform, if you're curious...

 So, couplet #1 of the old Third Edition was a choice between “shell coiled” and “shell an uncoiled cone.”  The former choice took the user onward into the body of the key, while the latter choice immediately took the user aside to seven limpet genera.  Those seven limpet genera are assorted into three families: the Ancylidae with four, the Acroloxidae with one, and the Lymnaeidae with two.  This is an ecological distinction – the limpet shape demonstrating superior strength against predation, and superior performance against hydrodynamic drag, given a solid substrate upon which its bearer can graze.

That patelliform shape has, of course, evolved many, many times independently in many, many different gastropod lineages, both freshwater and marine [3].  And it has apparently evolved (at least) three times separately in the freshwater gastropods – in the Acroloxoidea, in the Lymnoidea, and in the ancylid taxa of the Planorboidea.

So Christopher elected to open his new Fourth Edition gastropod key with operculum present/absent as his couplet #1. This is the easiest character by which to distinguish pulmonates from prosobranchs, the primary phylogenetic division in the freshwater gastropod fauna.  All the freshwater limpets are pulmonates, missing an operculum, hence all go together to his couplet #7.  Then at Christopher’s couplet #7, the user finds: 

7(1) Shell not patelliform, or if patelliform, then spire sinistral (apex centered or to right of midline) and blunt, with adult patelliform shell larger than 7 mm … go to #8.

7’ Shell patelliform with spire dextral (apex to left of midline), acute; adult shell less than 7 mm in length … Acroloxidae.

So I understand what Christopher is trying to do here, and I appreciate his effort.  He is beginning to sort the pulmonate families using evolutionary distinctions.  But I’m just not sure it works.

First, as a practical matter, users cannot determine if the spire is sinistral or dextral for 99.999% of all the limpets in the creek – they’re blunt and smooth.  So, users must read over that text to “apex centered or to the right of midline” vs. “apex to left of midline.”  And second, being finely evolutionary, both Acroloxus and the lymnaeid limpets (Lanx and Fisherola) demonstrate shell apexes to the left of midline, at least when young [4].

Now I understand if Christopher does not want to engineer his dichotomous key to the entire freshwater gastropod fauna of North America around an accommodation for juvenile-Lanx-collectors.  So I’ll let it go.  Let’s suppose we have under our scope some non-patelliform pulmonate, like Physa or Helisoma.  And we have dodged through couplet #7 to arrive at couplet #8, undiscouraged.  Here we read: 

8(7) Tentacles narrow, filiform …. 9

8’ Tentacles broad, flat, triangular; haemoglobin absent; coiled shell always dextral, patelliform shell with apex central or sinestral [sic]; never planospiral … Lymnaeidae

For heaven sake!  Now Christopher seems to expect us to have a living animal under our scope, or at least a very well-preserved one, to distinguish narrow tentacles from broad tentacles?  Narrow compared to what?  There’s no tentacle figure.  Haemoglobin, are you serious?  Must I medevac this little speck of coiled brown nothing to England and set up an IV?  Typed and cross-matched, stat?

In the Third Edition key, after I had observed that my Physa was coiled and been directed to couplet #8, I was next asked if the shell was planospiral or “with raised spire.”  If planospiral go to Planorbidae, otherwise go on to couplet #20.  Simple.

I am not going to criticize every couplet in Christopher’s entire 20-page key.  It’s a tremendous effort, and I don’t want to diminish his contribution.  I will simply observe that the evolutionary approach taken in the Fourth Edition is not as user-friendly as the ecological approach taken in the third.

So, here’s the bottom line for this month’s essay.  Don’t throw away your old Third Edition of Thorp & Covich.  Open it up on the lab bench next to your new Fourth Edition and use both simultaneously.  The two works side by side are a dichotomous perspective on the wonderful diversity that marks the North American freshwater gastropod fauna.

Notes

[1] REVIEW: Thorp & Covich Fourth Edition [12Apr18]

[2] Thorp, J. H., and D. C. Rogers (2016) Keys to Nearctic Fauna.  Thorp and Covich’s Freshwater Invertebrates, Fourth Edition.  Volume II.

[3] Actually, the “hypothetical ancestral mollusk,” from which all sevenish of the molluscan classes diverged back in the Precambrian, is generally modelled with a limpet-like (or plate-like) shell similar to that borne by the present-day Monoplacophora.

[4] Basch, P. (1963) A review of the recent freshwater limpet snails of North America. Bull. Mus. Comp. Zool. Harvard 129: 399-461.