The reality of scientific species

Are species real? Some people, even more skeptical than I, argue that the very idea of a species is no more than a set of artificial categories, more a product of the human need to classify the richness of the natural world than any genuine reality; this is the essence of the "species problem". I think that, though scientists may argue about the larger categories — whether such-and-such a group is a Sub-Order or a Super-Family — the biological concept of species seems pretty secure. A good definition of species — the kids are sure to ask: "a species is a population whose members are able to interbreed freely under natural conditions." This definition is explained and expanded in E.O. Wilson's book, The Diversity of Life, 1992 (page 38ff), a book to read and re-read, part of the deep background that enriches the experience of keeping aquaria.
 
Traditionally species descriptions have been based on a single holotype preserved in a university museum or a national natural history collection. You may be amused to know that the official holotype of Homo sapiens is Dr Edward Drinker Cope, lately of Philadelphia. The definitive dead white male. Conversely, you may be profoundly irritated: Why Philadelphia? Since the selection of which fish first swam into a biologist's net is also partly random, biologists have turned away from treating these holotypes, and their supportive paratypes, as embodying a kind of Platonic ideal, the archetypes against which all individuals are to be measured. Modern thinking in population dynamics suggests that species may show an unexpected range of individual variation. Among the Aphyosemion killifishes of West Africa, where isolated pools are liable to drying up in drought seasons, specific populations, rather than merely the species, are tracked by careful fishkeepers, who keep a record of the specific location from which their strain originated.
 
Populations of Aphyosemion species, or the desert pupfishes of the Western US, are constantly on the brink of extinction, in the isolated waterholes or springs where chance has stranded them. In the natural world, populations of individuals are equally strongly kept separate by various "reproductive isolating mechanisms." Some of these are genetic, based on incompatabilities of DNA or of the physical structures that DNA encodes. Other isolating mechanisms are behavioral. Populations may spawn at different seasons, for example, or just at different times of day.
 
Female choice is a mechanism that drives some speciation; the females are the ones deciding that an Apistogramma male's elongated dorsal fin spine is a hot look. These reproductive isolating mechanisms are seldom absolute, and the edges of species may be naturally blurred. Evolution works through its actions on populations ("gene pools"), rather than on individuals or on species as a whole. Admittedly, this "biological species concept" works best in a single locality and over a short period of time. Scientists begin to get into arguments whenever they start tracking a species as it's transformed through time, but, hey! that's not going to be a problem for us! May you live so long! And there is sometimes controversy spawned by the geographical shadings of populations that will partially interbreed where they naturally overlap, say among Apistogramma or Aphyosemion or other wide-ranging and flexible genera. So we amateur fishkeepers might become all the more concerned to know and remember where our individual fish were originally found, in other words, which natural population they came from.
 
Hybrids. "Reproductive isolation between breeding populations is the point of no return in the creation of biological diversity," Wilson reminds us. Nevertheless, closely-related species will sometimes hybridize in aquarium conditions, a measure of how far from natural our best fishkeeping efforts remain. Up to a point we can overcome the isolating mechanisms and get fish to spawn, but where genuine species exist, the hybrid offspring will suffer from reduced fertility, and the progeny of the cross will die out after a generation or two. The casual hybridizer of fishes won't mention this drawback when he's selling to you, if he's even aware of it.
 
Cryptic sibling species are another affair entirely. They are so similar in shape and color we are unable to tell them apart. But the fish know the difference. Mating experiments where a female is given a choice between males will often separate them out, and DNA analysis will confirm their existence. Why do some genetic variations lead to new species and others don't? James Mallet, University College London, and co-workers are suggesting that species don't need to be separated by dramatic geological events: exploding volcanos, splitting continents, new mountain ranges. Species splitting can also be a result of ecological speciation: "small-scale accomodations to everyday life" in which a population discovers a new food or a new way to eat it, adapts to shallower water for depositing eggs, or to lower oxygen levels or slightly cooler water. The concept of ecological speciation helps account for the amazingly rich variety of freshwater fishes and marine reef fishes. James Mallet argues for continuums of populations and widespread hybridization in  Philosophical Transactions of the Royal Society B: Biological Sciences, 2008. A googlesearch "Mallet ecological speciation" will uncover lots of documents inspired by this new concept.