Last month I reviewed the legacy of Frank Collins Baker (1867-1942), the "freshwater Pilsbry" whose deep personal connection to the biology of the organisms he studied remains vivid through his writings now three generations after he passed away. But this month I suggest that, as much as we may admire it, Baker's (1911) monograph of the North American Lymnaeidae is simply obsolete, and cannot continue to be the basis of the classification system we use today.
Baker recognized about 113 species and subspecies of lymnaeids in North America. These he divided into seven genera: Acella, Bulimnea, Lymnaea, Pleurolimnaea (a fossil taxon), Pseudosuccinea, Radix and Galba. The first six genera were monotypic in North America, or nearly so, while "Galba" was immensely complicated, and further subdivided into five subgenera and three "groups" (1). Baker synthesized a tremendous amount of data in support of this classification, and his science was as good as any malacologist working in 1911. But he was innocent of the Modern Synthesis, and his species concept was typological. And his understanding of the Lymnaeidae has today been superseded by something much, much finer.
I don't know much about Bengt Hubendick – he worked at the Riksmuseum in Stockholm from 1950 to 1959, then moved to the Natural History Museum in Goteborg, where he remained active until about 20 years ago. He wrote a number of very important works, including his (1955) Phylogeny of the Planorbidae, his (1964, 1967, 1970) Studies on the Ancylidae, and the (1978) chapter he wrote on the systematics of the Basommatophora for the set of volumes on Pulmonates edited by Fretter and Peake. But for sheer beauty, no monograph ever written on any group of mollusks before or since has ever topped Hubendick's 1951 masterpiece, "Recent Lymnaeidae, Their Variation, Morphology, Taxonomy, Nomenclature and Distribution (2)."
This (223 page, 369 figure) work of genius is absolutely worldwide in scope, which is amazing for any era. Hubendick used thin-sectioning techniques as well as gross dissection to construct detailed diagrams showing longitudinal sections of the male copulatory anatomy. His plates I, II, and III are amazing – entirely comprised of photos of Lymnaea peregra, showing dozens of diverse forms from scores of localities. He pioneered the use of the "mean photograph," superimposing the images of as many as 20 specimens on top of each other for each figure. Intraspecific variance was not a nuisance to Bengt Hubendick - it was the stuff of evolution, and it was to be cherished.
Hubendick concluded that, while the Lymnaeidae as a family demonstrate "great morphological uniformity, there is a wide range of variation within the various species." He recognized about 40 valid species worldwide, which he saw no reason to subdivide, placing all in the typical genus Lymnaea (3). Here in North America Hubendick admitted humilis, cubensis, bulimoides, catascopium (?), emarginata, columella, megasoma, utahensis (?), haldemani, and arctica, plus the holarctic species stagnalis, palustris, and (perhaps) truncatula (4). With Baker's 113 species reduced to about a dozen, the continued recognition of seven genera would seem difficult to justify.
Why this clean, modern, rigorous, elegant classification system did not immediately sweep the world is something of a mystery to me. Burch continued to advocate a modification of Baker's seven-genus system in his (1980, 1982) "North American Freshwater Snails." He wrote,
"The genus Lymnaea has been used variously to include nearly all members of the Lymnaeidae (eg Hubendick 1951) or only Lymnaea stagnalis and several very closely related species (eg F.C. Baker 1928). In this latter system, the family contains a number of species groups (genera) equal in rank to Lymnaea s.s. A third system, more or less a compromise between the previous two, uses Lymnaea as a large inclusive genus, but recognizes various subgeneric groups within it. These subgenera correspond to the genera of the F.C. Baker scheme. As a convenience for species-group separation, the less conservative scheme is used here."Burch went on to recognize 55 species and subspecies of lymnaeids in North America (3), which he divided into the same genera recognized by Baker (1911, 1928): Acella, Bulimnea, Fossaria, Lymnaea, Pseudosuccinea, Radix and Stagnicola.
Hubendick's system was received more warmly in Europe, adopted in Belgium by Adam (1960) and in England by Macan (1977), although not in Germany by Gloer (1994) nor in the Czech Republic by Beran (2002). The greatest authority on the Lymnaeidae active today must be Poland's Maria Jackiewicz (5), who prefers the "compromise" system mentioned by Burch - Lymnaea as a large inclusive genus with subgenera Stagnicola, Radix, Galba, and so forth.
Let's go with the compromise, shall we? Hubendick's marvelous monograph has convinced me both that the number of valid biological species of lymnaeid snails is small, and that as a family they collectively demonstrate broad morphological uniformity. But I hate to lose the indexing functions of the old genus names. Higher taxonomic categories play an important roll in information retrieval, and such terms as "Stagnicola" or the "fossarine" lymnaeids have been around for decades. So for the FWGNA project, I have referred all the lymnaeid species to the genus Lymnaea, with subgenera according to Baker and Burch. I'd invite you all to join me.
And have a Happy New Year!
(1) Baker modified this system slightly for his (1928) "Freshwater Mollusca of Wisconsin." The smaller-bodied half of Galba he elevated to the genus Fossaria, and the larger-bodied half became Stagnicola.
(2) Kungl. Svenska Vetenskapskademiens Handlingar. Fjarde Serien Band 3. No. 1. Stockholm: Almquist & Wiksells Boktryckeri AB. I don't understand any of this - I'm just copying it off the cover page.
(3) Excluding the weirdo limpet-shaped Lanx and its relatives.
(4) Hubendick seems to have missed Lymnaea caperata, which I do think is likely valid. He wasn't a god.
(5) Jackiewicz, M. (1998) European species of the family Lymnaeidae. Genus 9: 1 - 93.