There are a number of ways to eat other bacteria, and probably the simplest is phagocytosis, shown to the right (image from the free dictionary). It's a relatively easy system, involving nothing more than the ability to warp the cell membrane in response to binding, and release degrading enzymes once the prey has been captured. However, it does rely on the prey being smaller than you...and not forming complex multicellular structures such as biofilms, or fruiting bodies.
A second method is to parasitise, to crawl into the cell wall of your prey and destroy it from the inside out. This is the method used by the small bacterium Bdellovibrio bacteriovous, which attacks a large number of other bacteria (only Gram negatives though) and grows inside them, eventually destroying them. Its life cycle is shown below:This image is taken from the Nunez Group homepage, which contains a lot more information (and some beautiful pictures) about this method of predation.
The third option for bacteria is to use chemical warfare, release a large number of cell-destroying enzymes and then eat up the debris. This is the strategy of my Streptomyces, which excrete antibiotics capable of destroying a whole range of different bacteria. It is still not totally certain whether they do this for food, or simply to remove predators or potential rivals for space, but either way it's an effective method of killing bacteria which has been exploited by the pharmaceutical industry since Flemming first marketed the idea.
The final major strategy is to hunt. Bacteria do not all exist as solitary blobs in isolation, many species are able to form semi-multicellular structures that can move together, grow together, form spore-producing fruiting bodies and, in the case of Myxococcus, hunt together. Myxococcus xanthus is the model organism for this, capable of forming large swarms of bacteria that can swarm towards prey and then destroy it.
The image on the left (taken from the first reference below) shows a single M. xanthus bacteria (the long thin bacteria highlighted with an arrow) approaching and killing a round coccus bacteria. As soon as the xanthus touches its prey, it releases hydrolytic enzymes which destroy it, producing nutrients for the xanthus to then consume. Unlike Streptomyces (which can't move) these bacteria can form large colonies, which move forward together into an area colonised by prey, forming rippling shapes which can be seen on plates. Although xanthus are perfectly capable of hunting on their own, the large rippling group allows them to disrupt structures such as biofilms, giving more access to prey.
Gathering together in large groups also allows differentiation and division of labour within the group. While consuming prey, the M. xanthus tend to form two distinct subpopulations; bacteria near to the prey will be feeders, forming the characteristic rippling pattern. Behind them, bacteria in the less-nutrient rich area will begin to aggregate and form fruiting bodies which, if conditions suddenly turn bad (i.e all the food runs out) can sporulate, ensuring survival of the population.
It's a fascinating little world; hunters, predators, parasites, and a whole world of physical challenges to get through (swimming through water for bacteria is similar to moving through treacle for a human). The pressure and challenge of surviving in such a world has produced a whole mass of different shapes, sizes and strategies, from single celled packets of explosive chemicals to larger and more complex multicellular assemblies.
However good people get at killing bacteria, other bacteria will always be able to do it better. And that is why why study them.
Berleman JE, & Kirby JR (2009). Deciphering the hunting strategy of a bacterial wolfpack. FEMS microbiology reviews, 33 (5), 942-57 PMID: 19519767