Species concepts

The question of what constitutes a species is one of the oldest and most contentious in biology. Despite more than two centuries of debate, there is no single, universally accepted definition of species, and the ongoing disagreement — known as the species problem — reflects both the genuine complexity of the natural world and the different goals of biologists working in systematics, ecology, genetics, and palaeontology.^{6, 9} Over two dozen species concepts have been proposed, each emphasising different properties of the groups they seek to delineate: reproductive isolation, phylogenetic distinctiveness, ecological distinctiveness, morphological diagnosability, or genetic cohesion. The most influential have been the biological species concept, the phylogenetic species concept, and the morphological species concept, though several others — including the ecological, cohesion, and genealogical species concepts — have substantial followings in particular subdisciplines.^{6, 7}

The biological species concept

The biological species concept (BSC) was formalised by Ernst Mayr in his 1942 book Systematics and the Origin of Species and remains the most widely taught species concept in biology. Under the BSC, species are defined as "groups of actually or potentially interbreeding natural populations, which are reproductively isolated from other such groups."¹ The emphasis is on reproductive isolation as the criterion that delineates species boundaries: two populations belong to the same species if they can exchange genes through interbreeding, and to different species if they cannot. Mayr argued that reproductive isolation is the key property that allows species to maintain their identity in sympatry, because without barriers to gene exchange, two populations that come into contact would eventually merge into one.^{1, 2}

The BSC has substantial strengths. It provides a clear mechanism for species cohesion (gene flow) and species divergence (the evolution of reproductive isolation), and it connects the species concept directly to the process of speciation. It has been successfully applied to a vast number of sexually reproducing animal species and has guided much of the modern research on speciation genetics.⁹ However, the BSC has well-known limitations. It cannot be applied to asexual organisms, which do not interbreed at all, or to fossils, for which reproductive compatibility cannot be tested. It also struggles with cases in which species hybridise extensively in nature yet maintain distinct identities, as occurs in many plant genera, some birds, and recently diverged mammalian lineages.^{10, 11} The qualifier "potentially interbreeding" introduces further ambiguity, since allopatric populations cannot be tested for reproductive compatibility in nature, and laboratory crosses may succeed between populations that would never meet in the wild.^{9, 16}

The phylogenetic species concept

The phylogenetic species concept (PSC) emerged in the 1980s from the cladistic revolution in systematics and defines a species as the smallest diagnosable group of organisms within which there is a parental pattern of ancestry and descent.^{4, 5} Under the PSC, any population that can be distinguished from all other populations by at least one fixed character — molecular, morphological, or otherwise — qualifies as a separate species. The PSC does not require reproductive isolation; it requires only diagnosability and monophyly (or at minimum, exclusivity in the sense that all members of the group are more closely related to each other than to any outsider).⁵

The PSC has the advantage of being applicable to all organisms, including asexual lineages and fossils, and it provides an operational criterion (diagnosability) that can be tested directly using morphological or molecular data. However, the PSC tends to recognise substantially more species than the BSC when applied to the same set of organisms, because geographically isolated populations that differ in even a single fixed character qualify as distinct species under the PSC but might be considered conspecific subspecies under the BSC.^{6, 9} This "taxonomic inflation" has practical consequences for conservation, where species counts influence funding allocations and legal protections. It also raises the question of where to draw the line between species-level and population-level variation: with sufficiently fine-grained molecular data, virtually every geographically distinct population might be diagnosable, collapsing the distinction between species and populations.^{8, 15}

The morphological species concept and alternatives

The morphological species concept (MSC), also called the typological or phenetic species concept, predates both the BSC and the PSC and defines species by morphological distinctness. Under the MSC, species are groups of organisms that share a common morphological pattern and can be distinguished from other such groups by consistent differences in form. This was the dominant approach to species delimitation from Linnaeus through the early twentieth century and remains the only practical approach for many palaeontological taxa, for which neither reproductive compatibility nor genetic data are available.¹⁶ The MSC has the advantage of simplicity and broad applicability, but it is undermined by cryptic species (morphologically indistinguishable populations that are genetically and reproductively distinct) and by the subjective judgement involved in deciding which morphological differences are sufficient to warrant species status.⁹

Several additional species concepts address shortcomings of the major three. Templeton's cohesion species concept defines a species as the most inclusive population of individuals having the potential for phenotypic cohesion through intrinsic cohesion mechanisms, including both gene flow and developmental constraints.¹² The ecological species concept defines a species as a lineage that occupies a distinct ecological niche, emphasising adaptive differentiation rather than reproductive isolation or diagnosability. The genealogical species concept uses the criterion of reciprocal monophyly at multiple independent gene loci, requiring that all copies of a gene in one species coalesce more recently with each other than with any copy in another species.¹⁵ Each concept captures a genuine aspect of the speciation process, but none captures all aspects simultaneously.

The unified species concept

Kevin de Queiroz has proposed a reconciliation that he calls the general lineage concept or unified species concept. Under this framework, all species concepts agree on one fundamental point: species are separately evolving metapopulation lineages. What they disagree on are the criteria used to recognise such lineages — reproductive isolation, diagnosability, monophyly, ecological distinctness — and these criteria represent different lines of evidence for the same underlying phenomenon, not competing definitions of what species are.^{6, 7, 8}

Under the unified species concept, reproductive isolation, morphological diagnosability, reciprocal monophyly, and ecological distinctness are all expected consequences of lineage divergence, but they do not arise simultaneously. A newly diverging lineage may initially be distinguishable only by allele frequency differences at a few loci, later becoming diagnosable by fixed molecular differences, later still developing pre-mating reproductive barriers, and eventually evolving morphological distinctness. Different species concepts effectively draw the species boundary at different stages along this continuum.^{7, 8} The unified concept treats these properties as lines of evidence rather than defining criteria, suggesting that the more of them a lineage satisfies, the stronger the evidence that it constitutes a separate species. This framework does not eliminate disagreement over borderline cases — populations that have diverged sufficiently to satisfy some criteria but not others — but it reconceptualises the disagreement as a question about the strength of evidence rather than a question about the definition of species.⁸

Problem cases and the limits of species concepts

Ring species have long been cited as a challenge to the biological species concept. In a ring species, a chain of populations encircles a geographical barrier, with each adjacent pair capable of interbreeding but the terminal populations, where the ring closes, unable to interbreed despite being connected by a continuous chain of gene flow. The classic putative example is the herring gull-lesser black-backed gull complex circling the Arctic, though molecular analyses by Liebers and colleagues have shown that its phylogeography is more complex than the simple ring model suggests.¹³ The greenish warbler (Phylloscopus trochiloides) around the Tibetan Plateau provides a better-supported case, with gradual divergence in song, morphology, and genetics along the ring and reproductive isolation where the terminal forms meet in sympatry in central Siberia.¹⁴

Hybridisation presents perhaps the most pervasive challenge. Among plants, hybridisation is widespread, with an estimated 25% of plant species known to hybridise with at least one other species, and among animals, recent genomic studies have revealed far more introgression than previously suspected. The genomes of modern humans contain approximately 1 to 4% Neanderthal DNA, indicating substantial gene flow between lineages that had diverged hundreds of thousands of years earlier.^{10, 11} Such cases blur the sharp boundaries that the BSC assumes and raise questions about whether species boundaries are best conceived as semipermeable membranes rather than absolute barriers. In the genomic era, discordance among gene trees — where different regions of the genome tell different stories about species relationships due to incomplete lineage sorting and introgression — has become the norm rather than the exception, further complicating species delimitation.^{10, 15}

The species problem is unlikely to be fully resolved because species are products of an ongoing, continuous process — speciation — and any attempt to impose discrete boundaries on a continuum will inevitably encounter borderline cases. Populations in the early stages of divergence may satisfy some species criteria but not others, and different criteria may be met at different times and in different sequences depending on the particulars of geography, ecology, and genetics. The multiplicity of species concepts reflects this reality, and the most productive approach may be to acknowledge that species are real entities in nature but that their boundaries are inherently fuzzy at the margins, precisely because evolution is a process that operates continuously in time and space.^{6, 8, 9}