> when viruses infiltrate a bacterial ecosystem, often only about one bacterium in 10 million will gain a spacer that lets it defend itself. Sounds like, when

> when viruses infiltrate a bacterial ecosystem, often only about one bacterium in 10 million will gain a spacer that lets it defend itself. Sounds like, when under mortal threat, grab some DNA. Even if it mostly causes cellular suicide by grabbing its own DNA, if just one bacterium acquires immunity, the group as a whole benefits. Of course, group selection doesn't seem to work... but maybe the very high probability of death means this low probability last-ditch attempt is worth it for the individual - which then passes on the CRISPA machinery and the specific immunity to its "progeny"... which then have a nicely cleared out environment all to themselves... like eucalyptus after a bushfire. Desparate times call for desparate measures; and high risk/high reward.