1 School of Biological Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand
2 Department of Plant Sciences, University of Oxford, South Parks Road, Oxford OX1 3RB, UK
Correspondence
Tim Cooper
(t.cooper{at}auckland.ac.nz)
In a recent report, Boles et al. (2004) compared the effect of an environmental stress on high-diversity (wild-type) and low-diversity (recA mutant) biofilm communities of the bacterium Pseudomonas aeruginosa. They found increased resistance to an environmental perturbation in the diversified community, which they interpreted as support for the insurance hypothesis' an ecological model that predicts that a more diverse community will be better able to resist an external stress. The report is important in that it represents one of very few attempts to examine biofilm communities in the light of ecological theory. Nevertheless we question the authors' conclusion for two reasons, one general and one specific. We also question the authors' interpretation that diversification in a biofilm is somehow a programmed response to that lifestyle.
In the ecological literature insurance effects are defined as ... any long-term effects of biodiversity that contribute to maintain or enhance ecosystem function in the face of environmental fluctuation (Yachi & Loreau, 1999). The mechanism underlying these effects is functional redundancy between community members: a more diverse community is more likely to have members that respond differently to an environmental stress. If there is some amount of redundancy between these members then overall ecosystem functioning can be maintained even if some members fail. This intuitive idea has been around for many years, but was first formally stated in 1999 by Yachi & Loreau
. They modelled communities of differing levels of diversity and examined how this diversity affected community productivity in the face of varying degrees of fluctuating environmental stress. They found that increasing community diversity could increase average community function; however, this outcome depended on their being an asynchronous response of the members of the community such that members responded differently to different environmental stresses. For example, member A is more successful than member B in environment 1, but member B is more successful than member A in environment 2 (i.e. a trade-off in response between environments). A key finding of their study was that this trade-off is a necessary aspect of the insurance hypothesis. If member A was more successful in both environments 1 and 2, there would be no difference in community productivity between a monoculture of A and a mixture of A and B.
|
Our more specific concern with the experiment of Boles et al. (2004) involves the choice of the environmental stress (peroxide) applied to the high- and low-diversity communities. The stress had already been shown to differentially affect the (pre-diversified) ancestral type and the derived wrinkly type. Thus the finding that the diversified community was more resistant to this stress is not surprising the stress was chosen on the basis of differentially affecting high and low diversity communities. An unbiased test of the relationship between community diversity and resistance should determine the effect of a range of stresses chosen without prior knowledge of their effects on the members of the diversified community.
Finally, we wish to highlight the authors' discussion of how (and why) diversity originates in biofilms. They consider two explanations for their observation that diversity arises in biofilm but not planktonic communities: that variation is created equally in both communities, but that selection for variant forms is much stronger in biofilms and that there exists some specific ...programmed response to the biofilm state... (Boles et al., 2004, p. 16635) that generates the observed diversity. Boles et al. (2004)
favour the second of these possibilities for two reasons. (i) Communities derived from recA mutants are less diverse than those derived from the wild-type genotype, even though RecA is not thought to be involved in generating growth-dependent mutations. (ii) Auxotrophic mutants reproducibly arise within the biofilm communities in the absence of any obvious selective advantage. Both results can be explained without the need to invoke the involvement of a specific genetic programme. First, recA mutants typically display growth defects (Capaldo et al., 1974
; Sciochetti et al., 2001
), therefore the low diversity in communities derived from the recA mutant may have little to do with impairment of a biofilm-specific diversity generating mechanism, but rather, be a consequence of the fact that the recA-derived communities have gone through fewer generations compared to the control populations: selection thus has less opportunity to act. Second, as has been shown elsewhere, bacteria adapting to one environment often lose the ability to grow in alternative environments as a pleiotropic consequence of adaptation (Cooper & Lenski, 2000
; Cooper et al., 2001
; Maclean et al., 2004
). In this case, auxotrophy arises as a side-effect of adaptation without having to be selected directly. For example, in an experiment in which 12 independent populations of Escherichia coli were evolved in a minimal glucose medium, all 12 populations lost the ability to grow on ribose. This loss was subsequently shown to confer a fitness benefit of
1·5 % during growth in glucose. A test of this auxotrophy as a side-effect of adaptation hypothesis would be to compare the nature of auxotrophic mutations across independent experiments; parallel loss of function would strongly support there being an underlying adaptive cause (Harvey & Pagel 1991
). In the absence of any direct evidence for a programmed response the most parsimonious explanation of biofilm specific diversity is one coupling random variation and strong selection.
The diversity and interactions that can arise in biofilm communities represent unique opportunities for testing ecological and evolutionary theories on a real-time timescale. Boles et al. (2004) present interesting observations to this end, but don't extend these observations to rigorously examine competing explanations. It is precisely these extensions that need to be made if biofilm biology is to incorporate the ecological and evolutionary dimension that is currently lacking.
REFERENCES
Boles, B. R., Thoendel, M. & Singh, P. K. (2004). Self-generated diversity produces insurance effects. Proc Natl Acad Sci U S A 101, 1663016635.
Capaldo, F. N., Ramsey, G. & Barbour, S. D. (1974). Analysis of the growth of recombination-deficient strains of Escherichia coli K-12. J Bacteriol 118, 242249.[Medline]
Cooper, V. S. & Lenski, R. E. (2000). The population genetics of ecological specialization in evolving Escherichia coli populations. Nature 407, 736739.[CrossRef][Medline]
Cooper, V. S., Schneider, D. & Lenski, R. E. (2001). Mechanisms causing rapid and parallel losses of ribose catabolism in evolving populations of Escherichia coli B. J Bacteriol 183, 28342841.
Harvey, P. H. & Pagel, M. (1991). The Comparative Method in Evolutionary Biology. Oxford: Oxford University Press.
Maclean, R. C., Bell, G. & Rainey, P. B. (2004). The evolution of a pleiotropic fitness tradeoff in Pseudomonas fluorescens. Proc Natl Acad Sci U S A 101, 80728077.
Sciochetti, S. A., Blakely, G. W. & Piggoti, P. J. (2001). Growth phase variation in cell and nucleoid morphology in a Bacillus subtilis recA mutant. J Bacteriol 183, 29632968.
Yachi, S. & Loreau, M. (1999). Biodiversity and ecosystem productivity in a fluctuating environment: the insurance hypothesis. Proc Natl Acad Sci U S A 96, 14631468.
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
J MED MICROBIOL | ALL SGM JOURNALS |