In his letter, Dieter Haas reviews some interesting data on Pseudomonas aeruginosa that support our argument that it is necessary to consider the natural history of species to understand the evolution of quorum sensing. In our review1, we argued that both the production of signal molecules and the response to these molecules can incur costs. Therefore, both the signalling and response strategies should depend on the costs and benefits of signalling and responding. In the case of P. aeruginosa, Haas correctly points out that the cost of response (that is, the production of extracellular enzymes that degrade macromolecules) to N-acyl-homoserine lactones (AHLs) might be higher than the cost of signal production. Accordingly, one would predict that strains should more readily become quorum-sensing-deficient than signalling deficient, and this is indeed apparently the case, as Haas points out2.

Haas also makes the interesting point that roughly 20% of all environmental and clinical isolates are quorum-sensing-deficient. This again indicates that it is important to consider the ecology of species. As correctly pointed out by Haas, quorum-sensing-proficient P. aeruginosa have a selective metabolic advantage over quorum-sensing-deficient cells when casein or adenosine is the carbon source but a disadvantage in a nutrient-rich, alkaline environment. One would predict that many other ecological factors should influence the benefits of expressing extracellular enzymes that degrade macromolecules. A challenge for microbiologists is therefore to identify what these factors are in natural environments. Moreover, the evolution and spread of signalling- or response-defective cells should also depend on the kin structure of communities and whether several strains compete with each other in the same environment1.

In conclusion, Haas's comments are welcome and demonstrate the need for a better understanding of P. aeruginosa growth and evolution in natural settings.