richness of Madeira and the Canaries and how it is
that these islands come to be regarded as the
authentic core of Macaronesia (e.g. Kunkel 1993),
sharing some 50 Canarian–Madeiran endemic
plant species (Sziemer 2000), as well as 8 endemic
genera (Bystropogon, Heberdenia, Isoplexis,
Marcetella, Normania, Phyllis, Semele and Visnea)
(Santos-Guerra 2001). In contrast, the Azores and
Cape Verde islands were less well served by
former stepping stones, again consistent with
the pattern of affinities (Dias 1996; Brochmann
et al. 1997).
Recent molecular phylogenetic analyses have
greatly advanced our understanding of these rela-
tionships, indicating the most likely colonization
pathways for particular taxa both between the
mainlands and Macaronesia and within the region.
For example, it appears that some plants of
Canarian origin (certainly species of Aeonium,
Echium,Sonchus, and likely LavandulaandLimonium
as well) dispersed across the c.1300 km from the
Canaries to the Cape Verde islands, where endemic
forms developed (Santos-Guerra 1999).
Furthermore, at least six plant taxa appear to have
colonized Madeira from the Canaries: Bystropogon,
Aeonium,Convolvulus,Crambe,Pericallis, and the
woodySonchusalliance (Trusty et al. 2005). Some
lineages (e.g. Pericallis) subsequently colonized the
Azores, most likely via the southernmost Azorean
island, Santa Maria, some 850 km from Madeira. In
contrast, Argyranthemum and Crambe, two
Asteraceae genera that have radiated profusely in
Macaronesia, are believed to have colonized the
Canaries from Madeira (Santos-Guerra 1999).
Amongst birds, it has been suggested that the
chaffinches colonized the Azores from the Iberian
peninsula, and then colonized the western Canaries
via Madeira, whereas the Canarian blue tit appears
to have colonized Tenerife, in the centre of the
Canaries, and then to have spread to the other
islands of the Canaries (Kvist et al. 2005).
In terms of African links, the recent discovery in
the Anti-Atlas valleys of typical Macaronesian
floristic elements, including the Dragon tree
(Dracaena draco) and the laurel (Lauruscf.novoca-
nariensis), highlight the role of northwest Africa in
the colonization of Macaronesia. Another very inter-
esting piece of the jigsaw is the discovery of North
African endemic species of genera considered to be
Canarian in origin. Specific examples include the
following species of Aeonium:A. arboreum,A. kor-
nelius-lemsii(both in Morocco), A. leucoblepharum,
andA. stuessyi(both in Ethiopia). A further example
is a species of Sonchus(S. pinnatifidus), still extant in
the easternmost Canaries, from where it has colo-
nized the African coast, indicating a return trip for
lineages that developed on the Canaries having
originally colonized from Africa some 4 Ma (Santos-
Guerra 2001).
Summing up current understanding based on
recent molecular studies, Carine et al. (2004) report
that in the majority of genera investigated so far,
Macaronesian endemic taxa form a single mono-
phyletic group, in most cases with western
Mediterranean sister groups, but also with affinities
revealed with North Africa and the Iberian penin-
sula, and indeed with more widely separated
regions, such as the New World (Madeiran Sedum),
East Africa (Solanum) and southern Africa (Phyllis).
They also note cases of multiple congeneric colo-
nizations, and plausible cases of back-colonization
from Macaronesia to continental areas (e.g. in Tolpis
andAeonium).
How robust these scenarios will prove to be in
the light of future collecting and phylogeography is
unknowable at the present time. However, we can
draw several conclusions from these studies:●There have been multiple and indeed conflicting
colonization pathways followed by different taxa.
●The Macaronesian flora comprises lineages
diverging from their closest continental relatives at
different times, some ancient and some relatively
recent.
●There may well have been distinct waves of colon-
ization related to the greatly changing geological
and climatic history of the region.
●Following from this, although the islands are
truly oceanic, such that long-distance dispersal is
necessary to the exchange of lineages with the
mainlands, this is not to rule out a role for past vari-
ation in the strength of barriers between these62 THE BIOGEOGRAPHY OF ISLAND LIFE