General Discussion

And produce fertile offspring.

It's serious problem in cactus taxonomy, for instance, where you get continua of species across the Sonoran desert. At this end, you get one set of genes, flower morphology, spination, etc., etc. At some point in the middle, you get a different set, sufficient that they can't interbred. At the far end of the continuum, you get something that can't interbreed with either the midpoint or the first end point.

That makes it sound like there's just a line. But it's in 2D, with the same problem along any line you draw through the sometimes discontinuous territory.

Yet at every point along the way, there's a lot of interbreeding and gene flow. Even at recent discontinuities there's gene flow.

Taxonomy's a bear because where you define the type for the first species essentially dictates the number of species you come up with in the end and what they look like. Start in the middle, you get one number. Start in the southern extreme, you might get a second. Start in the NW corner, you get a third set of species.

Rules of historical precedence come into play, cladistic analysis, and the old fallback of having a type set by a good, solid description and analysis for the individuals in a well defined given locality.

(I teach HS physics, but was a cactus fancier married into a family of desert botanical garden curator-professors. Currently wondering how my small Ariocarpus and Rebutia collections will fair in Houston's unpredictable winters.)