Semantics. The light couldn't have got there without the person putting it there and turning it on in order to observe
Yes it can. For example, we may be comparing the behavior of bacteria in a sunlit puddle with the behavior of bacteria in a dark puddle at night; or we could be comparing the behavior of bacteria in a microscope experiment (under strong illumination) with the behavior of bacteria in a chemosensing experiment (where you detect the bacteria's presence via the chemical evidence left behind, and thus don't use light).
In both cases, the presence of light - irrespective of whether or not it was induced by an experimenter - affects the behavior of the bacteria.
To claim the researcher doesn't affect the observed little critters, because the light does is like saying a chef doesn't actually bake a pie, the oven does. Or, that a soldier didn't actually shoot the enemy, the gun did.
Yes, the oven bakes the pie - or would it get baked if the chef focused his psionic brain waves on it instead?
Yes, the gun shoots the enemy - or would the enemy have fell down screaming if the soldier had focused his psionic brain waves on him instead?
When you hold a hammer everything looks like a nail; and when you're dead bent on imagining that the free action of human being can somehow short-circuit our usual ideas of causation (because you can't think of any other way to discredit the idea of natural selection), you'll imagine that human free action determines everything. And yet it is still true that a chef without an oven cannot bake a pie, a soldier without a gun cannot shoot an enemy, and a scientist without a strong source of illumination cannot agitate phototaxic bacteria. The physical, not the mental, is always the proximate cause.
Of course, your example are example of human agency
but only because there are no natural analogues of the physical causes which these humans create. It's fairly easy to illuminate a colony of bacteria; harder to create a consistent temperature of 210 degrees C for several minutes; and it's just about impossible for you to expect a cylinder of metal to barrel right through your heart at a few meters per second. In those cases where peculiar physical causes correlate only with human intent, and never with non-intentional situations, we properly associate the causation with human intent.
Even so, I could clearly come up with counter-examples that illustrate how absurd your comparison is:
To claim the researcher doesn't affect the observed little critters, because the light does is like saying a chef doesn't actually make me full, a pie does. Or, that the pharmaceutical company didn't actually help my headache get better, the Panadol did.
The question, of course, is: which of these is natural selection more like? Until you show me how pigeon breeding is
physically substantially different from natural selection - other than the presence of human intent, which I have already shown is by itself irrelevant - you haven't a leg to stand on.