Mike's Missing Molluscs Redux: More on Methodology

by David Dudgeon

Porcupine! 15 contained an article by Mike Dickman entitled 'Why are there no snails in Shap Sze Heung Stream?'. Since it dealt with a stream that I have visited many times and animals which I have studied, I read the article with some interest, noting while I did so that two of my papers were cited and that the article drew upon my work on the predatory behaviour of freshwater crabs. My fascination grew as I traced the author's train of thought through the piece. In essence, the explanation for the lack of snails in this stream was attributed to crab predation and water chemistry. I believe that this conclusion is wrong, and I think that it may be instructive to examine the evidence for and against Mike's inference to see if we can derive any lessons from this difference of opinion.

After an incorrect reference to Sinotaia quadrata and Melanoides tuberculata as pulmonates (they are proso-branchs), the first paragraph of the article ends with '...there are no records of Mollusca reported from this stream'. There are no written records, true, but many visits were made to this stream during undergraduate field courses in the 1980s and 1990s. I have collected lymnaeids and planorbids (Segmentina sp.) as well as freshwater limpets (Ferrissia baconi) from Shap Sze Heung Stream, and the prosobranchs Clithon retropictus and Steno-melania sp. turned up in the lower course. They weren't numerous, but they were there. This example shows us that one should always ask other biologists if they have any relevant facts about the habitat one is studying. The article continues by postulating that a lack of calcium limits mollusc populations in Shap Sze Heung, and evidence is presented showing that Pokfulam Stream, with a high calcium concentration, has more snails than Shap Sze Heung. This may be so, but the streams differ in other important ways and any number of variables could vary between them (in addition to calcium). Unfortunately, if you find a significant difference between two sites, habitats, or whatever, all you can conclude is that they are intrinsically different. It is unwise to make any decisions about cause and effect. You cannot determine what caused the difference without manipulative experiments or (and this is less powerful) testing for differences in snail density between groups of streams which have different concentrations of calcium (i.e. you need replicates of streams in each 'treatment' group). Another approach is to use regression analysis to look at the relationship between snail abundance and calcium concentrations across a range of streams. However, neither regression analysis nor testing for differences between groups of streams will get around the problem of confounding variables. If, for example, food availability increases as calcium levels rise (perhaps because algal abundance is enhanced) then the response observed might be caused by food - not calcium. To reiterate, manipulative experi-ments are best!

Proceeding through Mike's article we are told that '...instead of snails...' the dominant herbivore in Shap Sze Heung Stream is the crab Cryptopotamon anacoluthon. I will leave aside the question of how one decides whether or not a species dominates a particular functional group (and whether snails are, indeed, dominant herbivores). Instead l note that, although Mike cites Dudgeon & Corlett ( 1994), he doesn't mention that we stated that Cryptopotamon feeds on leaf litter. Indeed, Rob Kennish (a noted local crab biologist) used this characteristic of the species to produce a small demonstration of crab feeding which delighted undergraduates during one field course. He also refers to the diet of Cryptopotamon in his Ph.D. thesis. The crabs that Rob dissected were collected from Shap Sze Heung Stream. The fact that Mike found that, in aquaria, Cryptopotamon eats Biomphalar-ia straminea - an exotic snail which does not occur in Shap Sze Heung Stream - is neither here nor there, and does not make it a molluscivore. Clearly, it is important to check your facts (especially those in the sources that you cite), and this is something that almost everyone is guilty of slipping up on now and again.

Mike's article goes on to suggest, citing two of my papers as supporting evidence, that the absence of snails in Shap Sze Heung might be due to predation by Cryptopotamon. He states that I said that thin-shelled pulmonates are often abundant in habitats with no crabs, and that habitats with crabs tend to have few pulmonates but many prosobranchs with heavy shells which protect them from crabs. These statements were made in the context of experimental studies on another crab entirely: a species of Somanniathelphusa which was investigated by Dudgeon & Cheung (1990) and named by Ng & Dudgeon (1992). Note that Somanniathelphusa is in the family Parathelphusidae; Cryptopotamon is in the Potamidae. They are neither the same nor equivalent. The lesson here is that the devil dwells in the details.

Two of the publications cited by Mike (Dudgeon, 1992; Dudgeon & Corlett, 1994) mention freshwater crabs and discuss research on Somanniathelphusa carried out by Dudgeon & Cheung (1990). Mike's article doesn't mention this paper (although it is quoted and referenced in the publications that he cites), yet it is the original source of our knowledge of freshwater crab predation in Hong Kong. By overlooking this paper, Mike was misled into supposing that Somanniathelphusa and Cryptopotamon were the same. This shows that one should always check the original sources.

At the end of the article Mike concludes that a combination of low calcium (so that thick-shelled snails can't survive) and crab predation (eliminating thin-shelled snails) accounts for the absence of gastropods at Shap Sze Heung. I have already dealt with the 'predation hypothesis'. What about the 'calcium hypothesis'? Given that snails have calcareous shells, we would expect some correlation between the occurrence of gastropods in streams and the calcium content of the water. Unfortunately, as Hutchinson (1993: pp. 146- 154) points out, the chemical ecology of freshwater snails is by no means straightforward. For example, Brotia hainanensis has a thick shell, yet abounds in Tai Po Kau Stream where hardness is < 2 mg l^-1 (Dudgeon, 1992: Table 3) and calcium is scarce. In a recent survey of 34 Hong Kong streams (Chan, 1996), Tony Chan - who was formerly the multivariate--statistician-in-residence in my research lab. - found calcium concentrations ranging from 0 to 4.2 mg l^-1 (mean 0.8 mg l^-1 ). Streams in which Brotia was most numerous had mean calcium levels of 1.2 mg l^-1, but these snails occurred at sites where the calcium levels were much lower than this. In addition, multiple regression models which significantly 'predicted' (R² = 0.57) the abundance of Brotia across the 34 sites did not include calcium among the independent variables (alkalinity, silica and sodium were critical). Although Tony's data weren't available to Mike when he wrote his Porcupine! article, they show that calcium is not the most important (or the only) influential factor determining the occurrence of one snail species in Hong Kong streams. In passing, I wonder why Shap Sze Heung Stream supports so many crabs and shrimps if it is so poor in calcium?

What can we learn from this difference of opinion? Firstly, it's easy to criticize (sorry Mike!). Secondly, it is easy to 'answer' questions on the basis of limited or inadequate data. It is also very tempting. Maybe the controlling factor here is calcium. However, the data at hand do not permit us to draw a robust conclusion. I think crab predation is unimportant in Shap Sze Heung Stream: I could be wrong, but the data at hand do not support Mike's conclusions about predation. The most appropriate means of settling arguments about cause and effect is to perform manipulative experiments. That is what is needed in this case, and (to be controversial) I believe such approaches should be mandated for other studies of the Hong Kong biota.

 

References

Chan, K. T., 1996. Multivariate Analysis of Benthic Macroinvertebrate Communities of Hong Kong Streams. Unpublished M. Phil. Thesis. The University of Hong Kong: 177pp.

Dickman, M., 1996. Why are there no snails in Shap Sze Heung Stream? Porcupine! 15: 7-8.

Dudgeon. D., 1992. Patterns and Processes in Stream Ecology: A Synoptic Review of Hong Kong Running Waters. Schweizer-bart'sche Verlags-buchhandlung. Stuttgart: 147 pp.

Dudgeon, D. & Cheung, C.P.S., 1990. Selection of gastropod prey by a tropical freshwater crab. J. Zool., Lond. 220: 147-155.

Dudgeon, D. & Corlett, R.T., 1994. Hills and Streams: an Ecology of Hong Kong. Hong Kong University Press, Hong Kong: 234 pp.

Hutchinson, G.E., 1993. A Treatise on Limnology. Volume IV. The Zoobenthos. John Wiley & Sons, New York: 944 pp.

Ng, P.K.L. & Dudgeon, D., 1992. The Potamidae and Parathelphusidae (Crustacea: Decapoda: Brachyura) of Hong Kong. Inveriebr. Taxon. 6: 741--768.

P.12-13

Back to Contents
Back to Porcupine Homepage
Go to Departmental Homepage