DNA for Identification
What if results of DNA sequencing don't match result of morphology (eg. the organism's appearance in iNaturalist photos)?
The bits of DNA that are sequenced tell you a lot about the relationships between the bits of DNA. That often, but not always, tells about the relationships between the organisms. Why not? Most often, incomplete gene sorting. The population of the common ancestor included both variations in the DNA -- let's call them A and B. In the ancestors of one species, A was lost. In the ancestors of the other, most A was lost and B predominates, but A still exists in low numbers. Thus sampling any single individual can be misleading.
Other possibilities include past hybridization followed by backcrossing, or some kind of horizontal gene transfer. Also, even the best of labs can have occasional incidents of contamination. (Sorry, but that can't be completely excluded from just one run.) Also, though it isn't as much a problem now as in the early days of PCR, the PCR reaction can become highly biased if the first cycle copies some untypical DNA sequence, perhaps from an organism's surface or from its last dinner, and that gets amplified. Also, though I hate to say it, we have probably all mislabeled a photo from time to time.
People forget that DNA sequences are genetic markers but structures of teeth and genitalia and other structures, even many behaviors, are also genetic markers. So is color pattern (though it may be a highly variable genetic marker in some species). Relying on just one marker is a kind of shorthand, good enough in most cases but sometimes misleading.
