830 
Telopea Vol. 6(4): 1996 
Asymmetric speciation 
Commonly hypothesized modes of speciation are allopatric, sympatric and peripheral 
isolation (Mayr 1982; Lynch 1989; Rieseberg et al. 1991; Theriot 1992). Allopatric 
speciation, in which a widespread ancestral species is split into vicariant 
(geographically isolated) populations which then differentiate and speciate, seems 
more likely to be symmetric, resulting in a pair of autapomorphic species. This is the 
speciation model usually assumed as a basis for cladistic vicariance biogeography 
(Nelson & Platnick 1981; Humphries & Parenti 1986; Cracraft 1989). However, 
sympatric speciation and peripheral isolation seem likely to result in the two daughter 
species that have populations very unequal in size. Lynch (1989) uses the admittedly 
arbitrary rule that if one species has a distributional area no more than 5% of its 
sister, this represents peripheral isolation rather than vicariance (allopatry). Similarly, 
local (sympatric) ecological differentiation and speciation of a small population under 
strong selection (Andersson 1990; Rohwer & Kubitzki 1993; Linder 1995) should also 
result in asymmetry. 
Thus some modes of speciation involve isolation of a small population. This may 
differentiate rapidly and evolve fixed differences from the parental population 
(autapomorphies) through processes such as founder effect, bottlenecks, genetic drift, 
selection and lineage sorting (Eldredge & Cracraft 1980; Wiley 1981; Rieseberg & 
Brouillet 1994). The much larger parental populations would be buffered against these 
effects, and so may persist for a long period without diverging from their ancestral 
condition. If the ancestral species were widespread, and had already fragmented into 
isolated, partly divergent populations, then the peripherally speciating population 
may be historically more closely related to an adjacent population than to other, more 
distant populations. In other words, the parental species would be paraphyletic. 
But is allopatric speciation necessarily symmetrical? If two relatively large populations 
become isolated from one another, divergence is likely to be gradual, and fixation of 
novel characters probably would occur some time after isolation. Thus some authors 
(Nixon & Wheeler 1992; Patton & Smith 1994; Rieseberg & Brouillet 1994) have 
suggested that species start out as a metaphyletic or paraphyletic group of populations 
(or even polyphyletic, due to reticulation) and only gradually become monophyletic 
(cf. Frost & Kluge 1995: fig. 3). Therefore, through any mode of speciation, whether 
symmetrical or not, either one or both daughter species is likely to appear at least 
temporarily non-monophyletic. 
Species concepts: a solution to the conundrum 
If species are no different from higher taxa (except in rank), and if taxa are forbidden 
to be paraphyletic or metaphyletic, then there is a paradox, because we have shown 
above that many species are unavoidably paraphyletic or metaphyletic. One solution 
would be to adopt the monophyletic species concept, and treat only demonstrably 
monophyletic species as taxa. Thus Donoghue (1985) and de Queiroz & Donoghue 
(1988) suggested marking metaspecies with an asterisk to identify them as different 
in kind from monophyletic species. This is tantamount to excluding paraphyletic 
and metaphyletic populations from species. Instead they would sit as unassigned 
residual populations at the base of higher (monophyletic) taxa. This treatment seems 
unacceptable to most systematists. It is clear that the great majority prefer to treat 
species as taxa, being part of the system to which all higher taxa belong. Therefore 
the only reasonable alternative is to drop the monophyly requirement for species. In 
other words, species may be considered taxa but with the special provision that they 
may be paraphyletic or metaphyletic. Is there a logical basis for defining species as 
special taxa? There would be if species had a property that made them different 
