found for continental floras; for islands there was
much more scatter in the relationship. Rosenzweig
does not distinguish the nature of the islands in his
analysis, i.e. whether they were predominantly
oceanic or continental in origin, and this may be a
confounding factor in his analysis. It is necessary to
look a little further into the data to establish
whether polyploidy has provided a significant con-
tribution to island evolution and whether it has
developedin situ. The following studies are cited in
order of increasing proportion of polyploidy.
Stuessyet al. (1990) examined about one third of
the 148 endemic plants of the Juan Fernández
islands (33S), which are volcanic, oceanic islands
off the coast of Chile. Remarkably, none of the
species examined was found to be polyploid.
Borgen (1979) reports 25.5% polyploidy among
360 Macaronesian endemic plant species from the
Canaries (28–29N). This is a low proportion in
comparison to most known floras, and is signifi-
cantly lower than the 36.4% polyploids amongst the
151 non-endemic Canarian species for which data
were available (Borgen 1979). It appears also that
several of the polyploid events were fairly ancient
ones, subsequent to which gradual speciation
within certain polyploid lineages has occurred (an
example being the genus Isoplexis in the
Scrophulariaceae), but to a lesser degree than
within diploid lineages (Humphries 1979). It may
be concluded that polyploidy has not been a major
evolutionary pathway in Macaronesian lineages.
The New Zealand (c.35–47S) flora consists of
approximately 1977 species, 40% of which are
known cytologically according to Hair (1966).
Polyploidy is absent in the gymnosperms, but char-
acterizes 417 (63%) of the 661 angiosperms investi-
gated, a figure approaching the 70–81% polyploidy
suggested for Oceanic–Subarctic and Arctic areas of
high northern latitudes. New Zealand is an ancient
archipelago, and its flora has been claimed to have
been primarily of continental origin, relictual of
Gondwanan break-up. However, this has been dis-
puted by Pole (1994), who argues against the large-
scale continuity of lineages in the New Zealand
flora, claiming that palynological data support
derivation of the entire forest flora from long-
distance dispersal (predominantly from Australia).
This remains a hotly contested issue (see e.g. Heads
2004, Cook and Crisp 2005, McGlone 2005) making
it difficult to put a time frame on the interpretation
of the New Zealand data.
Few firm conclusions can be drawn from the data
currently available. Polyploidy is more important in
some island floras than others. The latitude and age
of the islands in question may each be of relevance,
although often, as shown, age and origin of the flora
are imperfectly known. If the Macaronesian data
can be taken as a good guide, it appears that the
more spectacular radiations of island plant lineages
do not tend to rely upon polyploidy.8.4 Lineage structure
A number of different models and facets of island
evolution have been introduced in this chapter. Too
many frameworks may serve to confuse, but there
is one more that appears particularly helpful in
making sense of island speciation patterns, and
although the terms may not be instantly memo-
rable, the ideas seem useful. In some cases, evolu-
tionary change takes the form essentially of the
continuation of a single lineage, whereas in other
cases the splitting of lineages is involved. Stuessy
et al. (1990) provide three terms that codify what
they see as the chief outcomes (Fig. 8.2).●Anagenesisis when the progenitor species/form
becomes extinct.
●Anacladogenesisis when the progenitor sur-
vives with little change alongside the derived
species. What is meant by alongside is open to dif-
ferent interpretations—it could be within the same
island, or less restrictively, within the same archi-
pelago.
●Finally, cladogenesisis where the progenitor is
partitioned into two lines and becomes extinct in its
original form, a model suggestive of the classic
examples of adaptive radiation. This framework
sets aside cases where lineages converge or cross,
via hybridization.206 SPECIATION AND THE ISLAND CONDITION