Dr. Rob Dillon, Coordinator

Tuesday, September 7, 2021

Intrapopulation gene flow: King Arthur's lesson

A frequent visitor to the Malacology Department at Academy of Natural Sciences, during the sweet gauzy years of my graduate education in Philadelphia, was a charming scholar of aristocratic bearing named Prof. Arthur J. Cain [1].  Prof. Cain made his reputation on the study of color polymorphism in the European land snail Cepaea, publishing several very influential papers on the subject in the 1950s and early 1960s [2].  He was a wellspring of stories about E.B. Ford and the good old days of “Ecological Genetics,” stomping about the English countryside, collecting data on the frequencies of the myriad color morphs of Cepaea and their correlation with temperature, ground cover, predation, and so forth [3]

Arthur J. Cain (1921 - 1999)
King Arthur once remarked, “Since the term was coined by the founding fathers of the modern synthesis – I don’t know by whom, and it does not matter – nobody has ever imagined that the concept of panmixia might apply to a natural population of gastropods.”

So I had arrived in graduate school knowing that I wanted to focus my research on the phenomenon of speciation in freshwater snails.  And it was crystal clear to me that in order to understand speciation, I had to understand population divergence.  And in order to understand population divergence, I had to understand genetic variation within populations.  And in order to understand genetic variation within populations, I had to understand all those things that Arthur Cain loved to talk about – barriers to dispersal, isolation-by-distance, gene flow and the lack thereof.  In short, why freshwater gastropod populations are not panmictic.  That was the first thing.

Like a good student, I started in the library.  The literature on intrapopulation gastropod dispersal was mind-bendingly huge, even when I first dug into it in the late 1970s.  So I focused on the movement of freshwater gastropods in lotic environments, which was the most directly relevant to my research interest in pleurocerid populations in the Southern Appalachians.  And the first impression I got was the universal tendency for freshwater snails to actively disperse upstream, into currents.  And the second impression I got was the equally-universal likelihood of passive dispersal downstream, by wash-down.  Both of these phenomena are functions of current speed.

The oldest paper in the yellowing folder I still have filed under “Intrapopulation Dispersal” in the cabinet to the right of my desk also turned out to be among the most relevant to my subsequent dissertation research with Pleurocera (“Goniobasis”) proxima – a 1952 study of the Campeloma population inhabiting a small creek in Michigan.  Bovbjerg [4] observed striking aggregations of his large-bodied, burrowing study organism downstream from rocky riffles, which they seemed to have difficulty traversing.  His mark-release study, conducted over six days in a long, sandy region without such obstructions, returned a (surprisingly high) mean upstream dispersal rate of 350 cm per day OTSIWATFA [5], with no downstream dispersal whatsoever.

So the increased population densities Bovbjerg observed downstream from riffles would seem to be consistent with the opposite of dispersal… a barrier, right?  Exactly analogous to the grassy fields and gravel roads that Arthur Cain and the ecological geneticists focused so much of their attention upon in the English countryside, yes?  Back in 1978, I marked Bovbjerg’s paper with a little yellow sticky-tab.
On your mark!  Get set...

Marked with an orange sticky-tab in my intrapopulation dispersal file was a section from the 1978 dissertation of Mancini, involving a Pleurocera (“Goniobasis”) semicarinata population inhabiting a small Indiana stream [6].  Between September 1974 and January 1976, Mancini performed 10 release experiments, each two months in duration, involving 100 marked snails.  The fourth and fifth columns of the author’s Table 7 (reproduced below) show net migration, combining both upstream and downstream movement.  The mean over all five summer values was a modest 4.0 cm per day net upstream movement OTSIWATFA, and the winter mean a very modest +1.9 cm per day.

At the top of my intrapopulation dispersal stack, marked with a red sticky-tab, was the 1966 study of Crutchfield [8] on the Pleurocera proxima population of a North Carolina stream – the same study organism I had targeted for my own dissertation research, in exactly the same environment.  His was a single-release study of 15 week’s duration, between late December and early April of 1958.  Of the 53 snails Crutchfield marked in December, only 5 could be relocated in April, a result the author correlated to rains and high water.  All five snails had moved upstream, at a median distance of 55 feet (16.0 cm/d OTSIWATFA), ranging from 20 feet (5.8 cm/d) to “>100” feet (>29 cm/d)

Perhaps unsurprisingly, however, the best studies of freshwater gastropod dispersal published (as of my years in graduate school) focused on the medically-important planorbid Biomphalaria [9].  Adult Biomphalaria are adapted for lentic waters, not lotic, bearing clunky, bulbous shells typically enfolding an air bubble.  But the mark-release experiments of Paulini (for example), conducted in a ditch with a gentle current of 10 cm/sec, yielded a mean upstream dispersal rate of 130 cm/day OTSIWATFA.  After 6 days, the population mode was 15 meters upstream from Paulini’s release point.

In higher currents, the likelihood of dislodgement seems to become quite significant in Biomphalaria.  Pimentel’s mark-release experiment in a rapid Puerto Rican stream returned a net downstream transport that initially averaged 55 cm/day OTSIWATFA, as against only 36 cm/day upstream, for his first week of observation.  The same downstream-bias continued until observations ended at day 42, although at about half the initial rate.

The most refined study of freshwater gastropod dislodgement of which I am aware is Dussart’s [10] comparison of Biomphalaria with six other pulmonates, conducted in a transparent pipe.  He introduced each snail into a gentle current, allowing them to crawl directly on the PVC walls, and (of course) they all oriented upstream.  Gradually increasing the flow rate, he recorded the time at which each snail became dislodged and calculated a peculiar statistic (cubic centimeter minutes) measuring the total flow to which each snail had been exposed.  The most important variable predicting mean detachment flow (in cc.m) turned out to be profile area of the shell, rather than overall mass or foot size.

From Mancini [6]

One of the more memorable field experiences of my undergraduate education at Virginia Tech took place over the 24 hours I spent with a graduate student named Jim Kennedy sampling macroinvertebrate drift from the New River bridge in Fries, Virginia.  Jim had rigged a battery of plankton nets to suspend in the water column under the bridge, and we checked them every hour through one very cold winter afternoon in 1977, the night, and morning that followed. 

Freshwater gastropods are recovered from such nets at a surprisingly high frequency.  The most spectacular case of which I am aware was documented by P. C. Marsh [11], working in a small Minnesota Creek draining 17.6 square kilometers of extensively-ditched agricultural lands.  Marsh reported collecting at least a few Physa [12] in his drift nets at each of 10 sampling periods over the course of 20 months.  Heavy rains and high discharge seem to have been responsible for a peak mean of 326 snails/net/day on one sample Marsh collected in early April. 

But in late May a second peak in snail count was observed, this unaccompanied by rain.  Marsh’s average catch of 1,398 snails/net/day translated to an eye-popping 533,000 snails/cubic meter discharge/second, averaged over the entire day of observation.  This figure represented at that time (and may still be) the highest macroinvertebrate biomass ever documented from stream drift sampling.  Marsh speculated that this drift even may have been triggered by crowding and competition in the rapidly-growing Physa population upstream.

Marsh’s literature review found three studies of macroinvertebrate drift published between 1944 – 1980 containing quantitative data on gastropods [13].  Sitting in my carrel at the University of Pennsylvania Library, I was able to add a 1981 paper by Waters [14] and an excellent study by McKillop and Harrison [15] on the Caribbean Island of St Lucia, published in 1982.

New River Bridge at Fries, VA [16]

McKillop and Harrison positioned standard dip nets at the outfalls of several small, enclosed dasheen (or taro) marshes, recording total captures at 06:00 and 18:00 hours for 11 days during the month of April [18].  Summing all 11 samples taken at 18:00 hours, the authors recorded 29 individual Biomphalaria glabrata and 40 individuals of the little hydrobioid Pyrgophorus parvulus.  Totaled across all samples drawn at 06:00, McKillop and Harrison tallied 59 B. glabrata and 114 P. parvulus.  The authors compared the size frequency distributions of drifting Biomphalaria and Pyrgophorus to those of the resident populations and found that drift contained significantly increased frequencies of the smallest size classes.

The data of McKillop and Harrison demonstrate one of the best-documented characteristics of macroinvertebrate drift, diel periodicity.  At night, especially on the new moon, stream organisms actively enter the water column at an increased frequency.  So as was the case with the Physa population in the Minnesota ditches, the data of McKillop & Harrison suggest that at least some of the Biomphalaria and Pyrgophorus inhabiting St. Lucia marshes intentionally release into the water currents.

For eleven years, ecology students at the University of Cape Town, South Africa [19] sampled five sites along the 11 km Liesbeek River nearby.  The first two sites were near the mountainous source of the river, showing average March current velocities of 61 cm/sec and 47 cm/sec.  Sites 3, 4, and 5 were approaching the coast, with average March current speeds of 41, 30, and 13 cm/s, respectively.  The students first collected invasive Physa acuta at site 5 in 1979.  By 1980 the population had spread 3.1 km upstream to site 4, and in 1981 it first appeared at site 3, another 1.8 km upstream.  Thus, over two years, P. acuta displayed a minimum net upstream movement of 4.9 km, or 671 cm/d.  Active dispersal cannot account for such a pace.

The Physa population became established at sites 3, 4, and 5 over the next seven years, in some places reaching densities in the hundreds per square meter, but as of 1988 had not spread upstream to site 2.  These observations are all consistent with an hypothesis that the Liesbeek River Physa population has relied on avian transport, bait bucket hitchhiking, or some similarly passive agency for upstream dispersal.  My readership is referred back to my 2016-17 series of essays on the arial dispersal of freshwater gastropods [20] for further development of this interesting theme.

The bottom line for this month is that the gastropod populations inhabiting lotic environments move.  Their movement upstream is slow but measurable, while their movement downstream is rapid and episodic.  What are the evolutionary consequences?  Might these two processes be sufficient to maintain panmixia, contrary to King Arthur’s lesson?  Tune in next time.


[1] For more about this “polymath, and one of Britain’s leading evolutionary biologists,” see:

  • Clarke, B.C. (2008) Arthur James Cain, 25 July 1921 – 20 August 1999.  Biographical Memoirs of Fellows of the Royal Society 54: 47 – 57.

 [2] My favorites from Arthur Cain’s bibliography:

  • Cain, A.J. and P.M. Sheppard (1954) Natural selection in Cepaea.  Genetics 39: 89 – 116.
  • Cain, A.J. and P.M. Sheppard (1954) The theory of adaptive polymorphism.  American Naturalist 88: 321 – 326.
  • Cain, A.J. and P.M. Sheppard (1956) Adaptive and selective value.  American Naturalist 90: 202-203.
  • Cain, A.J. and J.D. Currey (1963) Area effects in Cepaea.  Philosophical Transactions of the Royal Society Series B 246: 1 – 81.

[3] The masterful review of J.S. Jones and colleagues might fairly be said to sum up the entire research corpus of the British school of ecological genetics:

  • Jones, J.S., B.H. Leith, and P. Rawlings (1977) Polymorphism in Cepaea: A problem with too many solutions?  Annual Review of Ecology and Systematics 8: 109 – 143.

[4] Bovbjerg, R.V. (1952)  Ecological aspects of dispersal of the snail Campeloma decisum.  Ecology 33: 169 – 176.

[5] OTSIWATFA = Of The Snails I Was Able To Find Again.

[6] Eugene Mancini’s 1978 dissertation was an old-school gem packed with great observations on pleurocerid biology.  I first heard about it from Steve Chambers [7], and in March of 1981 wrote a long letter to Dr. Mancini, then working at Woodward-Clyde Associates in California, to request a copy.  He cordially complied in April.  I was so impressed by the 93 page work that I showed it to my mentor George Davis, at that time editor of Malacologia.  George agreed with me that it should be published, and so I wrote a second letter to Mancini, thanking him for his kindness, complimenting him on his work, and offering a prescription of modest edits by which we felt it could be published in Malacologia.  I never heard from him again.

  • Mancini, E.R. (1978)  The biology of Goniobasis semicarinata (Say) in the Mosquito Creek drainage system, southern Indiana.  Ph.D. Dissertation, University of Louisville.  93 pp.

[7] Steve was an important influence on my young career, and a good friend.   See

  • Fred Thompson, Steve Chambers, and the pleurocerids of Florida [15Feb17]

[8] Crutchfield, P.J. (1966)  Positive rheotaxis in Goniobasis proxima.  Nautilus 79:80 -86.

[9] My favorite references on intrapopulation dispersal in Biomphalaria:

  • Pimentel, D., P.C. White & V. Idelfonso (1957) Vagility of Australorbis glabratus intermediate host of Schistosoma mansoni in Puerto Rico.  Am J Trop Med Hyg 12: 191 – 196.
  • Scorza, J.V., J. Silva, L. Gonzalez, & R. Machado (1961) Stream velocity as a gradient in Australorbis glabratus (Say, 1818).  Zeitschrift fur Tropenmedizin und Parasitologie 12: 191-196.
  • Paulini, E. (1963) Field observation on the upstream migration of Australorbis glabratus.  Bull WHO 29: 838 – 841.
  • Etges, F.J. & L.P. Frick (1966) An experimental field study of chemoreception and response in Australorbis glabratus (Say) under rheotactic conditions.  Am J Trop Med Hyg 15(3):434-438.

[10] Dussart. G.B.J. (1987) Effects of water flow on the detachment of some aquatic pulmonate gastropods.  American Malacological Bulletin 5: 65 – 72.

[11] Marsh. P.C. (1980) An occurrence of high behavioral drift for a stream gastropod.  American Midland Naturalist 104: 410 – 411.

[12] Marsh referred to his study organism as “Physa gyrina,” but I’m sure my readership will agree with me that Physa acuta is a much more likely identification.

[13] Papers documenting freshwater gastropod drift, from Marsh [11]:

  • Dendy, J.S. (1944) The fate of stream animals in stream drift when carried into lakes.  Ecological Monographs 14: 333-357.
  • Logan, S.M. (1963)  Winter observations on bottom organisms and trout in Bridger Creek, Montana.  Trans Am Fish Soc 92: 140 -145.
  • Clifford, H.F. (1972) Drift of invertebrates in an intermittent stream draining marshy terrain of west-central Alberta.  Can J Zool 50: 985 – 991.

[14] Waters, T.F. (1981) Drift of stream invertebrates below a cave source.  Hydrobiologia 78: 169 – 175.

[15] McKillop. W. & A. Harrison (1982)  Hydrobiological studies of eastern Lesser Antillean Islands.  VII. St. Luca: Behavioral drift and other movements of freshwater marsh mollusks.  Archiv fur Hydrobiologie 94: 53 – 69.

[16] You are looking at the only straight patch of road in all of Grayson County, Virginia.  Jim and I spent 24 hours living in a van, parked down by the river at left.  About 3-4:00 PM a good old boy, driving a jacked-up Plymouth, stopped by for a chin-wag.  As he departed, he boasted that he could “get rubber in four gears” across that bridge, in the roughly 300 yards from the store at left to the mountain wall at right.  I’m not sure how he knew he could do it [17], but he did it.

[17] But I’ve got a pretty good hunch.

[18] McKillop & Harrison sampled other months, but the diel periodicity is not as dramatic, primarily because of the confounding effects of rain.  Refer to their paper directly for the entire data set.

[19] Appleton, C.C. & G.M. Branch (1989)  Upstream migration by the invasive snail, Physa acuta, in Cape Town, South Africa.  South African Journal of Science 85: 189 – 190.

[20] The phenomenon of aerial dispersal in freshwater gastropods has been reviewed on four occasions in the long history of this blog:

  • Freshwater gastropods take to the air, 1991 [15Dec16]
  • A previously unrecognized symbiosis? [11Jan17]
  • Accelerating the snail’s pace, 2012 [24Apr17]
  • Freshwater snails and passerine birds [26May17]


  1. Neat stuff!! I have seen large numbers of snails (Juga, Fluminicola, Lanx, Vorticifex, Physa) release from the substrate and drift away when crayfish emerge and start to hunt in the evening. I have also witnessed it when I have walked through riffles. I expect any snail predator or large mammal walking across a creek bottom could cause a mass drift. I look forward to the next exciting episode . . .

  2. Gee, that's a remarkable observation. I can't say I've ever noticed any such thing. Thanks for sharing!

  3. Adding to what D.C. said, watching live Physa and other snails in tanks I've noted that some species will regularly detach from all solid substrate with no provocation needed, with no current present or any possibility of forced dislodgement.

    This can involve leaving solid emergent objects to "crawl" on the water surface itself while inverted, or sometimes detaching from the bottom and gently ascending through the water column directly.

    Physa especially do this. Other snails, like Lymanaea will do it also, but some, like Pleurocera, I have never seen to attempt it, even after long periods of observation.

    From this I assume it is mainly a behavior of "air-breathing" freshwater snails (although perhaps not universal there), and not a habit of the gilled snails. This would suggest different dispersal patterns could hold for these different physiological groups, even when their shell profile geometry was similar.

  4. Bryan - yes, I think that's a very common observation, sometimes ascribed to "grazing on the surface film," and other times correlated with the simple act of breathing in pulmonate snails. And yes I agree, such behavior should indeed lead to differing dispersal patterns for pulmonates and prosobranchs.

    Pulmonate surface-feeding behavior is entirely limited to lentic waters, in my experience. But it's not hard to imagine a situation where a snail feeding on the surface of a calm backwater might find itself cycled into a current and carried downstream. So the bottom-line prediction would be that Pulmonates, as a rule, disperse better.

  5. Having a "lung" with an air-bubble seems to give pulmonates a different kind of buoyancy than most prosobranch snails--close to neutral for pulmonates, and definitely negative for prosobranches (with exceptions-that-prove-the-rule for non-pulmonates with "lungs", like apple snails).

    We all know that many freshwater "lunged" snails are weedy supertramps while freshwater gills-only snails are comparatively very conservative at dispersal, but I never put together the biomechanical "why?" until you laid it all out in your essay here.

  6. Yes, that and their adaptation to rich (albeit unpredictable) habitats, rapid growth, huge reproductive output, short generation time, hermaphroditism, ability to self-fertilize, etc, etc!

  7. Exactly! No point in having neutral buoyancy if you can't reproduce by yourself after you've been carried who-knows-where.

    It is like Darwin's observation on the frequent evolution of flightlessness in insects on tiny oceanic islands.

    If getting 'blown away' is likely to mean certain death (or no reproduction, no ability to establish a sustainable population), then you're better off not getting blown away.

    Better to have no wings (or negative buoyancy), unless you also have the ability to survive aerial (or in-current) travel.

    But if you do have the ability to survive such travel, then there's a big advantage in it, and you'll tend to keep your wings (or neutral buoyancy).

    Flies and snails -- so much in common.