This has so many edge cases that it's really more of a vague rule of thumb than a hard definition: ring species (as another poster pointed out), infertile indiv

This has so many edge cases that it's really more of a vague rule of thumb than a hard definition: ring species (as another poster pointed out), infertile individuals (does every infertile individual constitute a species by itself?), drift over time (like ring species, but over time: an individual might be able to (in theory) produce fertile offspring with a far ancestor or with a far descendant, but the far ancestor may not be able to mate with the far descendant), ...