The paradigm of the microbial food web, exemplified by the famous figure in Pomeroy [10], has been extremely productive during the past 20 years and is now well established. According to the workings of science proposed by T.S. Kuhn [7], now is the time to check its predictions in as wide a range of conditions as possible, in order to look for the paradoxes that may eventually lead toward a new paradigm. Many of its statements are based on aggregating all bacteria together and ignoring differences among them. This is obviously an oversimplification and illustrates the problem I will discuss: in order to measure boxes and arrows in any model of the food web we need to aggregate organisms into arbitrarily chosen groups. If we go too far in pooling organisms together, we will lose all the natural history that ultimately drives evolution of both organisms and communities. If we stop short and retain many groups, we will be confronted with a complexity that will escape analysis. Therefore, we need to recognize explicitly the criteria we use to aggregate organisms. This is a problem common to all ecologists. In microbial ecology, however, this aggregation has been absolutely empirical and the problem has been explicitly addressed in only a very limited way. [...]

3 pages