I followed ion and atom trapping a long time ago, when it was still fairly new. (i.e., a couple decades ago) That was a valid question from the git-go, and there were a number of techniques. One way is that the trap is less stable when there are multiple atoms, so the extra atoms will eventually evaporate away, and you can measure the diminishing intensity of the scattered laser light. A paired atom will also interact with the light differently, for instance having a different spectral signature.