= Species delimitation
<Species delimitation> infers the boundaries and number of <species> from evidence. No single arbitrary percentage of <nucleotide sequence> difference is a universal boundary. Divergence within an old, widespread <species> can exceed divergence between recently separated <species>; geography and multiple kinds of evidence matter. <Species Concepts and Species Delimitation by Kevin de Queiroz>[pp. 884–885, “Alternatives to the Traditional Properties”: divergence cutoffs and geographic differentiation]{ref}
An original counterexample to a universal percentage rule: an assay could examine a strongly conserved <gene> in one comparison and a rapidly changing <nucleotide sequence> in another. Equal percentages then answer different historical questions. Even the denominator needs specifying: all sites, aligned sites, or only a filtered region? A threshold is an operational rule for a stated assay and group of <organisms>, whose errors need evaluating. It is not a definition that automatically transfers to every <species>.
\Image[/species-evidence-criteria.png]
{title=Contrasting operational evidence for <species delimitation>}
{description=Original hypothetical comparison of <morphology>, mating and ancestry evidence. Each observation tests a different claim; the combinations are thought experiments, not a dataset or verdicts about named <species>. “Not tested” does not mean absent.}
The diagram's hypothetical pair B has similar <morphology> but reduced mating and distinct ancestry evidence. A <morphological species concept> could miss that distinction. Pair C instead has different <morphology> but compatible mating and ancestry that fails to separate the samples. That combination calls for checking variation in <phenotype> within a <species>. Pair A has only a structural difference observed: the other criteria remain untested. These examples show why missing evidence, disagreeing evidence and agreeing evidence require different interpretations.
Back to article page
