Lepidoptera: butterflies are thought to comprise
about 17 500 full species, but the number of cur-
rently recognized subspecies approaches 100 000
(Groombridge 1992). These problems are general to
taxonomy and systematics, but have particular
force in relation to islands on which there is often
no adequate way of testing the potential reproduc-
tive isolation of geographically isolated ‘species’,
‘subspecies’, or ‘varieties’ on different islands
within an archipelago or ocean basin. Even in well-
studied lineages the integrity of the species unit can
be difficult to resolve. For example, although recent
phylogenies based on mtDNA sequences (Sato et al.
1999) recognize 14 species of Galápagos finches
(13 on the Galápagos plus the Cocos Island finch,
which is descended from a Galápagos ancestor),
Zink (2002) argues that the species-level taxonomy
within the genera Geospiza (ground finches,
6 species) and Camarhynchus(tree finches, 4 species)
should be considered unresolved. Zink raises the
possibility that GeospizaandCamarhynchusare in
effect each functioning as a single, highly variable
species, which would mean reducing the number of
recognized species from 14 to 6. This taxonomic
uncertainty is a product of the degree of hybridiza-
tion going on in this relatively youthful radiation
(Grant and Grant 1996a). Elsewhere in this volume
we follow general practice in recognizing 14 species.
In a review of the species concept, Mallet (1995)
notes the prevalence, even at textbook level, of as
many as seven or eight different notions of how to
define species. This may be bewildering to many,
and for the present purposes a way out is needed.
This is not the place for a lengthy review of the
problems of taxonomy and systematics; the intent
has instead been to ensure that the reader appreci-
ates that difficulties exist with the species concept
and its application. The solution Mallet recommends
is to return to Darwin’s pragmatic usage of ‘species’
and ‘varieties’, but to make use of new knowledge
from genetics as well as morphology in determining
when the species label is appropriate. Darwin’s posi-
tion is made clear by the quote given above, and in
the following passage from On the origin of species:
...‘varieties have the same general characters as species,
for they cannot be distinguished from species, except,
firstly, by the discovery of intermediate linking forms...;
and except, secondly, by a certain amount of difference for
two forms, if differing very little, are generally ranked as
varieties...‘We now recognize that there is more to it than
this, but even with modern genetic techniques it
can be difficult to agree on what constitutes a
‘good’ species.
If species are in essence arbitrary units, or ‘well-
marked varieties’, then it will be apparent that
determining the point at which speciation has
occurred within a radiating lineage is also problem-
atic. ‘That is like asking exactly when a child
becomes an adult. I am content to know that ini-
tially there was one lineage and now there are
more.’ (Rosenzweig 1995, p. 87). The debates on
this crucial matter will continue, and although
some may not like it, the view taken here is a prag-
matic one, following Darwin and Wallace: the pre-
cise level at which species are defined is arbitrary.
As with other currencies, exchange rates can
change as a function of the methods and assump-
tions used in the calculation.
At higher levels, species are grouped into genera,
then families, and so forth. In flowering plants such
groupings of species are based principally on the
evolutionary affinities of floral structure and so it is
possible to find many different growth forms and
ecologies within a single family. Thus the
Asteraceae (Compositae—the daisy family) all have
a composite flower structure, but some are herbs
and others are trees. In general, there tends to be
less variation in functional characteristics within a
genus than between genera (within the same fam-
ily) and there is presumed to be a closer evolution-
ary relationship, with all members of a genus
ultimately descended from a common ancestor,
different from that of other genera within the
same family. Some species are so distinct that they
may be the only members of a genus or family,
whereas others belong to extremely species-rich
genera, e.g. figs (Ficus; Moraceae). Insular examples
include, at the one extreme, Lactoris fernandeziana,
the monotypic representative of the endemic family
Lactoridaceae from the Juan Fernández islands
(Daviset al. 1995); and at the other extreme, on theTHE SPECIES CONCEPT AND ITS PLACE IN PHYLOGENY 197