>Dogs are ostensibly trained to distinguish between a lingering odor (or "residue") and actual material.

Have you got any citation for this? I am wondering how is it even possible.