80
T.J. Givnish
of the last species pair might simply reflect their recent divergence; in this regard,
it should be noted that 3 of 12 Cyanea taxa on the youngest island of Hawai'i (ca.
0.4 M years old) are considered conspecific with taxa endemic to Maui (Lammers
1990). Three ecologically similar, palm-like species (c. aculeatiflora, C. hamatiflora,
C. macrostegia) cooccur in wet forests near Waikamoi on East Maui, but diverge
somewhat in mean flower length and/or elevation (Fig. 5). C. aculeatiflora and C.
macrostegia are closely related, based on morphology, crossability (Lammers 1990)
and molecular data (Givnish et al. 1995), but C. hamatiflora is quite distant from
both, a member of the purple-fruited clade (Givnish et al. 1995).
These exceptions aside, species of Cyanea appear to have undergone largely
parallel adaptive radiations in flower length and elevational distribution on Kaua'i,
Q'ahu, Maui, and Hawai'i - an evolutionary version of the movie "Groundhog
Day", in which the genus evolved in much the same way (albeit with a few quirks)
on each new island to which it dispersed. This similarity among insular radiations
appears to extend to the interspecific range in maximum plant height and the number of species found on each major island as well, with the notable exception that
Hawai'i, the youngest island, has fewer species than the older islands, in spite of
being much larger in area (Table 2). This shortfall (12 vs. 15-21 spp) probably
reflects the short time (0.4 M years) plants and other organisms have had to evolve
on Hawai'i, compared with that available on Maui (1.9 M years), Q'ahu (3.7 M
years), and Kaua'i (5.1 M years). In accord with this hypothesis, a higher fraction
of species on Hawai'i (27%) are conspecific with populations on other islands than
are species on Maui (24%), Q'ahu (14%), or Kaua'i (0%) (tabulated from data in
Lammers 1990).
2.5 Extinction Rates in Relation to Adaptive Radiation
and Geographic Speciation
Limited dispersal may not only have accelerated speciation and fostered parallel
radiations on individual islands in Cyanea, but - paradoxically - have increased
the likelihood of extinction for individual species as well (Givnish et al. 1995).
Reduced dispersal and high rates of speciation appear to have resulted in narrow
geographic and elevational ranges: 91 % of Cyanea species are restricted to single
islands, vs. only 58% for species in the sister genus Clermontia, while the average
elevational range of individual species of Cyanea is 438 ± 294 m, vs. 840 ± 367 m
for species of Clermontia (P < 0.001, 2-tailed t-test, 59 d.f.). The narrow geographic and elevational ranges of individual species of Cyanea, their limited dispersal, and dependence (at least in some instances) on highly specialized pollinators may have made them more susceptible to extinction (Givnish et al. 1995).
Habitat destruction, decimation of pollinators and seed dispersers, and introduction
of mammalian herbivores has led to 20% of Cyanea species now being thought
extinct, with an additional 29% considered endangered (updated from Givnish et
T.J. Givnish
of the last species pair might simply reflect their recent divergence; in this regard,
it should be noted that 3 of 12 Cyanea taxa on the youngest island of Hawai'i (ca.
0.4 M years old) are considered conspecific with taxa endemic to Maui (Lammers
1990). Three ecologically similar, palm-like species (c. aculeatiflora, C. hamatiflora,
C. macrostegia) cooccur in wet forests near Waikamoi on East Maui, but diverge
somewhat in mean flower length and/or elevation (Fig. 5). C. aculeatiflora and C.
macrostegia are closely related, based on morphology, crossability (Lammers 1990)
and molecular data (Givnish et al. 1995), but C. hamatiflora is quite distant from
both, a member of the purple-fruited clade (Givnish et al. 1995).
These exceptions aside, species of Cyanea appear to have undergone largely
parallel adaptive radiations in flower length and elevational distribution on Kaua'i,
Q'ahu, Maui, and Hawai'i - an evolutionary version of the movie "Groundhog
Day", in which the genus evolved in much the same way (albeit with a few quirks)
on each new island to which it dispersed. This similarity among insular radiations
appears to extend to the interspecific range in maximum plant height and the number of species found on each major island as well, with the notable exception that
Hawai'i, the youngest island, has fewer species than the older islands, in spite of
being much larger in area (Table 2). This shortfall (12 vs. 15-21 spp) probably
reflects the short time (0.4 M years) plants and other organisms have had to evolve
on Hawai'i, compared with that available on Maui (1.9 M years), Q'ahu (3.7 M
years), and Kaua'i (5.1 M years). In accord with this hypothesis, a higher fraction
of species on Hawai'i (27%) are conspecific with populations on other islands than
are species on Maui (24%), Q'ahu (14%), or Kaua'i (0%) (tabulated from data in
Lammers 1990).
2.5 Extinction Rates in Relation to Adaptive Radiation
and Geographic Speciation
Limited dispersal may not only have accelerated speciation and fostered parallel
radiations on individual islands in Cyanea, but - paradoxically - have increased
the likelihood of extinction for individual species as well (Givnish et al. 1995).
Reduced dispersal and high rates of speciation appear to have resulted in narrow
geographic and elevational ranges: 91 % of Cyanea species are restricted to single
islands, vs. only 58% for species in the sister genus Clermontia, while the average
elevational range of individual species of Cyanea is 438 ± 294 m, vs. 840 ± 367 m
for species of Clermontia (P < 0.001, 2-tailed t-test, 59 d.f.). The narrow geographic and elevational ranges of individual species of Cyanea, their limited dispersal, and dependence (at least in some instances) on highly specialized pollinators may have made them more susceptible to extinction (Givnish et al. 1995).
Habitat destruction, decimation of pollinators and seed dispersers, and introduction
of mammalian herbivores has led to 20% of Cyanea species now being thought
extinct, with an additional 29% considered endangered (updated from Givnish et