Crisp & Chandler, Par'aphyletic species 
833 
created. Like Linnaeus, most taxonomists want to put organisms in pigeon holes. 
Complementary (paraphyletic) groups at all ranks have always been treated as taxa 
and continue to be spoken of, even by cladists, e.g. gymnosperms, dicots, algae, 
reptiles, dinosaurs, amphibians, fish. 
Returning to species, we conclude that there is not a problem with the recognition 
of paraphyletic and metaphyletic species. Problems arise only because some 
systematists want an all-inclusive concept of taxa that will allow all organisms to be 
assigned to a species using the same criteria as for higher taxa. However both theory 
and the discovered pattern of relationships among organisms show that no such 
universal criterion is possible. Species are different from higher taxa because they 
are basal, so a special criterion is justified. Paraphyletic and metaphyletic species are 
theoretically unavoidable, unless the autapomorphic species concept is adopted, in 
which case the problem is merely swept under the carpet. As we have concluded 
above, the appropriate species concept in a phylogenetic system is either the 
phylogenetic species concept or the related composite species concept. Both predict 
that some species will be monophyletic and others paraphyletic or metaphyletic. 
Implications for historical applications 
One of the arguments against recognising paraphyletic taxa is that they may be used 
in other biological applications with the assumption that they are monophyletic, e.g. 
Cracraft (1989). If the assumption of monophyly is invalid, does this invalidate the 
application? Let us consider cladistic biogeography. 
Cladistic biogeogeography attempts to discover historical patterns of areas manifest by 
congruence among the phytogenies of different taxa occurring in the same areas. Because 
of confounding factors such as failure of some taxa to speciate when areas differentiate, 
extinction, incomplete sampling and dispersal, some methodological rules are necessary. 
Component analysis (Nelson & Platnick 1981; Humphries & Parenti 1986; Page 1990) 
uses three 'assumptions' incorporating different rules (Nelson & Platnick 1981; Zandee 
& Roos 1987). For example. Assumption 0 treats widespread species (Fig. 7a) as if they 
(and the areas in which they occur) were monophyletic. If the species in question (and 
the history of the combined areas in which it occurred) were actually paraphyletic 
(Fig. 7a), this assumption would fail to reconstruct the area-phylogeny (Fig. 7c). 
Assumption 0 is inappropriate if a paraphyletic taxon occupies multiple areas because 
its ancestor failed to differentiate when the areas differentiated (Page 1989: 169). The 
more relaxed Assumption 1 treats widespread areas as either monophyletic or 
paraphyletic. Congruence with a well-corroborated pattern shown by another group 
with a monophyletic taxon in each area would then favour the well-corroborated solution 
(Fig. 7e), rather than a misleading one (Figs. 7c-d). Assumption 2, which considers 
possible polyphyly of areas, would also support the well-corroborated pattern. Tlierefore 
in this example at least, a paraphyletic species would not seriously mislead cladistic 
biogeography, at least under Assumptions 1 and 2. Rage & Jaeger (1995) argue that 
paraphyletic taxa can be used meaningfully in biogeography, the only deficiency being 
the information that is missing because part of the taxon has been excluded taxonomically. 
In simulations of macroevolution, Sepkoski & Kendrick (1993) found that paraphyletic 
taxa treated as if monophyletic did not seriously mislead estimation of speciation 
and extinction rates, except when sampling was poor. Smith (1994b: 88-91) states 
that paraphyletic taxa can be used validly in estimating patterns of taxon origination 
and standing species-level ['phenon-level'l diversity, but not extinction patterns. 
The latter restriction applies because paraphyletic taxa may disappear from the 
fossil record due to pseudoextinction, when a member of the paraphyletic group 
gives rise to an autapomorphic descendant taxon. 
