modify SLiM engine to do bacterial species

[petrelharp]

As discussed with @jeanrjc: how to do this is described in Cury et al:

1. use a nonWF model
2. have indviduals be haploid
3. switch out the reproduction callback to do HGT

This might be pretty different to the existing SLiM engine because population regulation happens in a different way; for instance in the paper above it works as follows:

early()
{
 inds = p1.individuals;
 ages = inds.age;

 // kill off parental individuals; biologically they don’t even exist,
 // since they split by mitosis to generate their offspring
 inds[ages > 0].fitnessScaling = 0.0;

 // density-dependent population regulation on juveniles, toward Ne
 juvenileCount = sum(ages == 0);
 inds[ages == 0].fitnessScaling = Ne / juvenileCount;
}

However, probably a bunch of the same machinery from the existing implementation will be useful. I don't know if this should be a new engine, or just additional behavior of the same engine.