Liverpool John Moores University, UK

An ecologist enjoys a smelly experiment on a neglected link in the food web.

I have long been fascinated by an idea from the 1970s about rotting food. Daniel Janzen, now at the University of Pennsylvania in Philadelphia, suggested then that many of the noxious chemicals secreted by microbes in decaying food are produced to fend off large animals, allowing the microbes to keep the resource for themselves.

It's an intuitively appealing hypothesis. Our own experience is to be repulsed by putrid food, and several studies have shown that birds prefer fresh over rotted fruit. Most recently, a careful study in the seas off the southeasten United States provided further support for Janzen's idea (D. E. Burkepile et al. Ecology 87, 2821–2831; 2006).

In what must have been a gloriously smelly experiment, the researchers baited crab traps with dead fish, either rotten or fresh. The microbe-laden carrion was four times less likely to be consumed by scavengers than the fresh fish.

This provides clear evidence that microbes compete for food with larger animals, something that has been largely overlooked in the huge ecological literature on food webs and feeding relationships. But it doesn't tell us how the chemicals evolved.

Last year, I published with colleagues a theoretical analysis of the evolutionary implications of Janzen's idea (T. N. Sherratt et al. Ecol. Modell. 192, 618–626; 2006). Our model suggested that the chemicals cannot have evolved solely to protect against large animals, because the temptation for microbes to 'cheat' by free-riding on toxin production by others undermines the system.

The experiments done by Burkepile et al. show that the effect is real, but perhaps these chemicals first evolved for other reasons, such as inter-microbe competition?