Evolving Thoughts

Evolution, culture, philosophy and chocolate! John Wilkins' continuing struggle to come to terms with impermanence... "Humanus sum, nihil humanum a me alienum puto" - Terence

Thursday, December 29, 2005

Speciation conclusion

If we accept the distinction between mode of speciation (geographical relation between populations) and rate of gene exchange, the mechanisms of speciation that fail to be distinguished that way - mostly chromosomal rearrangements and selection-based - add extra axes or dimensions to the notion of reproductive isolation. Depending on how many actual mechanisms there are we can have a number of different ways of attaining species-hood. Add to this the ways of being asexual species, and we get a rich field, one that is more complex than the usual ways of framing the issue.

The traditional RI conception of species known as the "biological" species concept was invented by Theodosius Dobzhansky in his 1935 paper, and revised in the text that effectively kicked off the modern synthesis, Genetics and the Origin of Species. According to Dobzhansky,
... a species is a group of individuals fully fertile inter se, but barred from interbreeding with other similar groups by its physiological properties (producing either incompatibility of parents, or sterility of the hybrids, or both). (Dobzhansky 1935, 353; cf. also Dobzhansky 1941, 312)
Mayr's version is better known:
A species consists of a group of population which replace each other geographically or ecologically and of which the neighboring ones intergrade or interbreed wherever they are in contact or which are potentially capable of doing so (with one or more of the populations) in those cases where contact is prevented by geographical or ecological barriers.

Or shorter: Species are groups of actually or potentially interbreeding natural populations, which are reproductively isolated from other such groups. (Mayr 1942, p120; cf. also 1940)
Mayr contrasted this version with the "typological" conception he claimed, wrongly*, was the default view before he "corrected" Dobzhansky's view. Dobzhansky's was a description of the process but not a definition of the concept of species, he said. I think he was wrong. Mayr tended to overstate his own originality and to denigrate others' contributions. Dobzhansky's definition sure looks to me like a definition. Mayr's adds little - an undefined term "population" and some passing acknowledgement of ecological and geographical aspects. In terms of our chart, Dobzhansky treated the gene exchange as primary. Mayr added geography.

Why does it matter that two populations are in a geographical relation in speciation? Surely this is because Mayr's own version relied upon geography as the proximal mechanism for independent evolution that gave speciation as a by product. It is special pleading. Geography matters to Mayr because geography is the primary mechanism of sexual isolation in his preferred model. But if isolation is achieved in many ways, and it surely is, even in Mayr's own account, then we have an interesting situation. Speciation is something that occurs in multiple modes, using many mechanisms ranging from gamete surface molecules, to behavioural differences, to ecological adaptations, to genomic structure and so on.

So there is a high dimensionality to speciation. For every aspect of RI, for karyotypic change, for ecological isolation and reinforcement, as well as geography, speciation exhibits a plurality of causes. And in the case of non-RI speciation, such as quasispecies/asexuals, the causes can be purely ecological.

So let us work out the hierarchy of speciation processes:

First we have the distinction between speciation that involves gene flow and speciation that doesn't. Quasispecies do not, usually, exchange genes between populations (viruses can superinfect a cell and crossover their genes; and some gene flow can occur due to retroviral insertion). Speciation that involves the rate of gene flow between populations ranges from pure allopatry (0% gene flow) to pure sympatry (50% gene flow), with most being a mixture (parapatry). Cases of near-sympatry must require isolating mechanisms such as selection.

Second we have the distinction between intrinsic and extrinsic selection. Intrinsic selection involves selection against the background of the rest of the population. It includes mate selection, lowered hybrid fitness, and developmental incompatibilities. Extrinsic selection involves selection against only the extra-specific environment.

Third, we have the distinction between determinate speciation and random or Markov chain speciation. In the latter case, the "location" of the population in genome space is a matter of stochastic sampling, with no constraints imposed on viability or fitness. We can illustrate this with a diagram:

Modes of speciation (examples only)

Each of these three axes are independent. A nascent species/population may be maintained at its genomic location by extrinsic selection and yet be able to randomly traverse the equal fitness ridges in the fitness landscape. Each axis is a notional, or abstract, variable. The concrete aspect for a given case will be something physical, like a pairing problem for chromosomes or a haplotype block, or cell surface identifiers on gametes, or a developmental abnormality that precludes viable zygotes, and so forth. Nothing is required and many things may be sufficient to achieve specieshood.

Does this mean that being a species is an abstract property? In philosophical terms this is often held to be equivalent to asking whether or not species are real. Pluralism of modes of being species is held, for example by Marc Ereshefsky, to mean that there is nothing in (biological) reality that corresponds to species. Species are just concepts. There are concrete realities, but they are not necessary to being a species, and in effect "species" is a trashcan categorial, a heterogeneous notion that is dissolved into real things or eliminated. I reject this option for the following reasons (Wilkins 2003).

Some general notions, such as "individual" or "object", are multiply realisable. So long as the category covers something common to all and only those entities, the class concept can be held to denote "real" things. The multiple causal origins of species does not in itself force the conclusion that species are not real. I will appeal to another domain and a solution offered by Jaegwon Kim (1993) from the philosophy of mind.

Thoughts are an output of brains - that much is pretty clear. But it is likely that "having a thought T" is physically different in every mind that has it. For physicalists - those who think that all that exists is physical - such abstract notions as "thinking T" are problematic. Either all instances of T are physically identical, or there is something non-physical that relates all T-instances together under the general abstract notion of "T". Kim's solution is to call these T-instances "supervenient", which can be briefly summarised as "being the same if physically identical, but being multiply realisable". Sober (1984, 2000), for example, has argued that fitness is a supervenient property. Any two organisms in the same environment that are physically identical (in the relevant respects) have the same fitness value, but not all organisms with that fitness value are physically identical. Being a species is, I believe, a supervenient property, and speciation is a superveniently-realising process. I alluded to this with the notion of "reproductive reach". There is a physical fact of the matter whether two organisms with particular genomic and developmental configurations (the physical properties) will have an average number of progeny, but having that number of progeny is something that occurs in many different physical ways. What counts is the persistence of the genomic and developmental configurations C by reproduction (in a particular class of environments E).

Such supervenience explains why we observe species when we can. The phenomenality of species is a physical outcome of the relation between the observer's assays or discriminative capacities and the progenitive properties of the organisms' C in E. From a heuristic perspective, species are phenomenal objects (iff we have the right assay). Speciation is the process of attaining this phenomenal quality. In cryptic species pairs we may not be able to discriminate the differences. Or we may over-discriminate when the physical proerties do not interrupt reproductive reach. But overall, species are real when there is an interruption in the number of progeny that maintain the relevant causal processes of reproduction.

In summary, a species is some object of the relevant kind (a lineage of a population) that is maintained by causal physical processes through selection and gene flow. All species are lineages, as de Queiroz has argued (1998), and as Richard Boyd has argued, they are homeostatically maintained kinds (1999), but neither homeostasis nor lineage is sufficient to identify species. We must in the end make recourse to reproduction rate and gene-exchange in the case of sexual species, and selection in all cases, to distinguish species from subspecific and supraspecific cases. When in sympatry the rate of reproduction between putative species-populations is close to the maximum (50%), we do not have species. When it is closer to the minimum (0%), we do. When asexuals maintain their ecological niche in sympatry, we have species. When they facultatively exchange niches, we don't. And so forth.

It is my opinion that we must take selection into account in every case of putative speciation and species-status. On such a meagre return for so long a discussion, I leave this topic for now.

[Late note: I promised to discuss Templeton's concept. Later.]

* In my book on species which may, or may not, get published soon, I cover the history of the species concept and conclude that the only sense in which people were "typological", or "essentialist", which Mayr added later, before Darwin was exactly the sense in which all later systematists were typological or essentialist - in terms of diagnosis or identification. People didn't suddenly become smart after Darwin, nor were they mindlessly Aristotelian and bad observers before him.


Boyd, R. 1999. Homeostasis, species, and higher taxa. In Species, New interdisciplinary essays, edited by R. Wilson. Cambridge, MA: Bradford/MIT Press.

de Queiroz, Kevin. 1998. The general lineage concept of species, species criteria, and the process of speciation. In Endless forms: species and speciation, edited by D. J. Howard and S. H. Berlocher. New York: Oxford University Press.

Kim, Jaegwon. 1993. Supervenience and mind: selected philosophical essays, Cambridge studies in philosophy. New York, NY, USA: Cambridge University Press.

Mayr, Ernst. 1940. Speciation phenomena in birds. American Naturalist 74:249-278.

———. 1942. Systematics and the origin of species from the viewpoint of a zoologist. New York: Columbia University Press.

Sober, Elliott. 1984. The nature of selection: evolutionary theory in philosophical focus. Cambridge, Mass.: MIT Press.

———. 2000. Philosophy of biology. 2nd ed, Dimensions of philosophy series. Boulder, Colo.: Westview Press.

Wilkins, John S. 2003. How to be a chaste species pluralist-realist: The origins of species modes and the Synapomorphic Species Concept. Biology and Philosophy 18:621-638.