Externally maintained faunas.

Imagine a situation in which individuals are so scarce that they cannot be detected by ordinary sampling methods. In such a scenario, males and females would presumably encounter difficulties in finding each other in order to mate. This is a problem recognised with the sparse faunas of oceanic islands (MacArthur and Wilson, 1967), where it is assumed that island faunas are maintained by immigration from nearby mainland habitats. These mainland habitats have been dubbed source habitats (Wiens and Rotenberry, 1981). Here I consider the case of habitat islands where the faunas may be maintained externally in this way. I refer to ephemeral rain pools in general and to those on rock surfaces in tropical Africa in particular. In a 1988 paper, I discuss the possibility that the seemingly clear adaptations to the tropical rain pool habitat may be illusory. I made this suggestion because larval populations of the relevant species, the chironomid midges Chironomus pulcher and C. imicola, could theoretically be maintained by ovipositing females travelling from permanent lakes and rivers.

In a hypothetical argument, I reasoned that if permanent waters provide ‘source’ faunas, with pools as ‘sink’ habitats, the interesting question is – where does selection operate to shape adaptations in these rain pool species – in source or sink habitats? I have generally assumed that pool faunas are adapted by selection to meet the stringent requirements of life in a transient pool habitat. This seems a reasonable assumption and there is good circumstantial evidence that it is so (McLachlan and Ladle 2001). But, and here lies the problem, if the bulk of the gene pools for these species is to be found in permanent waters, seeking adaptation to rain pools would be a waste of time.

I now wish to overturn the hypothesis that rain pool faunas are maintained externally. I suggest instead that there is a fundamental difference between oceanic island of Wiens and Rotenberry (1981) and MacArthur and Wilson (1967) on the one hand and transient pools on the other. I make this suggestion for two reasons. First, though I have never observed mating, there is ample evidence from counts of egg masses (McLachlan 1988), that breeding occurs in the rain pools. Second, the low densities proposed for source habitats would presumably make breeding unlikely there because of the difficult of finding mates. So, strong selection for parthenogenesis might be expected in the putative source habitats, the lakes and rivers and would therefore be expected to occur in rain pools. Yet I have never been able to demonstrate parthenogenesis among rain pool species. This finding does not eliminate the possibility of parthenogenesis but does suggest that it is not the routine method of reproduction. Indeed, though parthenogenesis is not unknown among chironomid midges, it appears confined to genera other than Chironomus (Williams 1974). I therefore conclude that the low density putative populations in lakes and river is not a a source habitat for rain pool dwelling chironomids.  So the search for adaptations to the extraordinary rain pool habitat is probably a worthwhile enterprise after all.

References

MacArthur, R. H. and Wilson, E. O. (1967). The Theory of Island Biogeography. Princeton University Press.

McLachlan, A. J. (1988). Refugia and habitat partitioning among midges (Diptera: Chironomidae) in rain pools. Ecological Entomology, 13, 185-193.

McLachlan, A. J. and Ladle, R. (2001). Life in the puddle: behavioural and life-cycle adaptations in the Diptera of tropical rain pools. Biological Reviews. 76, 377-388.

Wiens, J. A. and Rotenberry, J. T. (1981). Censoring and the evaluation of avian habitat occupancy. Studies in Avian Biology.6, 522-532.  

Williams, D. N. (1974). An infestation by a parthenogenetic chironomid. Water Treatment and Examination. 23, 215-232.