78
T.J. Givnish
et al. 1997), Laupala crickets (Shaw 1995), and Tetragnatha spiders (Gillespie and
Croom 1995) based on molecular data, and the plant genera Hesperomania, Remya,
Hibiscadelphus, and Kokia (Funk and Wagner 1995) based on morphology.
Cyanea species appear to partition each island via differences in elevational
distribution within moist/wet forests. For example, within the Hardyi clade of purplefruited Cyanea, there appears to have been an evolutionary progression from low to
high elevations on Kaua'i, involving C. coriacea (150-220 m), c. hardyi (425700 m), and C. spathulata (700-1200 m) (Givnish et al. 1995). More broadly,
species of Cyanea appear to have invaded the full elevational range of appropriate
habitats on each major island. Average species elevation (calculated as the mean of
the upper and lower limits in Lammers 1990, 1992, 1996, as well as unpubl. pers.
obs.) ranges from 185 to 1215 m on Kaua'i, from 320 to 950 m on Q'ahu, from 455
to 2130 m on Maui, and from 465 to 1850 m on Hawai'i (Table 2). The much
greater elevations achieved on the latter two, young islands partly reflects the greater
maximum elevations of Hawai'i (4205 m) and Maui (3055 m) compared with the
two older islands of Q'ahu (1225 m) and Kaua'i (1598 m), as well as the absence of
moist or wet forests above roughly 2200 m and the dominance of bogs and wet
shrub lands near the crests of Q'ahu and Kaua'i (Gagne and Cuddihy 1990).
When species on each island are plotted by their mean corolla length and elevation, there is a roughly even distribution of species along these combined gradients,
suggesting an even and strikingly similar partitioning of pollinators and habitats
on each island (Fig. 5). Exceptions to this rule are few: (1) C'yanea humboldtiana
and C. superba on Q'ahu; (2) C. copelandii and C. mceldowneyi in wet forests near
Waikamoi on East Maui; and (3) C. degeneriana and C. pilosa on Hawai'i, smallflowered species with pilose foliage which were originally described as subspecies
of the same taxon by Rock (1919).
The close similarity of the first pair of species largely disappears if geographic
variation in elevational distribution is taken into account. Cyanea (formerly
Rollandia [Lammers et al. 1994]) humboldtiana is endemic to the Ko'olau MounTable 2. Range, mean, and standard deviation of average corolla tube length, plant height,
and elevation for Cyanea taxa occurring on each of the four rain-forested islands of the
Hawaiian archipelago (see text). Data compiled from Lammers (1990, 1992, 1996), Rock
(1919), and personal observations. Some or all data are missing for C. glabra, C. linearifolia,
C. longissima, C. pinnatifida, C. remyi, and C. pycnocarpa
Island No. of
Corolla tube length
Plant height
Elevation
species Range
Mean ± s.d.
Range Mean ± s.d.
Range
Mean ± s.d.
Kaua'i
17
17-65 36.7 ± 14.0 1.0-14
4.1 ± 3.6
185-1215 631 ± 328
Q'ahu
15
24-75 53.5±17.3 0.5-6.0 2.7 ± 1.6
320-950 562 ± 147
Maui
21
18-75 49.3 ± 17.0 1.5-8.0 4.4 ± 2.3
455-2130 1127 ± 465
Hawai'i 12
22-75 51.1 ± 20.4 1.2-10
4.0 ± 2.8
465-1850 1064 ± 451
T.J. Givnish
et al. 1997), Laupala crickets (Shaw 1995), and Tetragnatha spiders (Gillespie and
Croom 1995) based on molecular data, and the plant genera Hesperomania, Remya,
Hibiscadelphus, and Kokia (Funk and Wagner 1995) based on morphology.
Cyanea species appear to partition each island via differences in elevational
distribution within moist/wet forests. For example, within the Hardyi clade of purplefruited Cyanea, there appears to have been an evolutionary progression from low to
high elevations on Kaua'i, involving C. coriacea (150-220 m), c. hardyi (425700 m), and C. spathulata (700-1200 m) (Givnish et al. 1995). More broadly,
species of Cyanea appear to have invaded the full elevational range of appropriate
habitats on each major island. Average species elevation (calculated as the mean of
the upper and lower limits in Lammers 1990, 1992, 1996, as well as unpubl. pers.
obs.) ranges from 185 to 1215 m on Kaua'i, from 320 to 950 m on Q'ahu, from 455
to 2130 m on Maui, and from 465 to 1850 m on Hawai'i (Table 2). The much
greater elevations achieved on the latter two, young islands partly reflects the greater
maximum elevations of Hawai'i (4205 m) and Maui (3055 m) compared with the
two older islands of Q'ahu (1225 m) and Kaua'i (1598 m), as well as the absence of
moist or wet forests above roughly 2200 m and the dominance of bogs and wet
shrub lands near the crests of Q'ahu and Kaua'i (Gagne and Cuddihy 1990).
When species on each island are plotted by their mean corolla length and elevation, there is a roughly even distribution of species along these combined gradients,
suggesting an even and strikingly similar partitioning of pollinators and habitats
on each island (Fig. 5). Exceptions to this rule are few: (1) C'yanea humboldtiana
and C. superba on Q'ahu; (2) C. copelandii and C. mceldowneyi in wet forests near
Waikamoi on East Maui; and (3) C. degeneriana and C. pilosa on Hawai'i, smallflowered species with pilose foliage which were originally described as subspecies
of the same taxon by Rock (1919).
The close similarity of the first pair of species largely disappears if geographic
variation in elevational distribution is taken into account. Cyanea (formerly
Rollandia [Lammers et al. 1994]) humboldtiana is endemic to the Ko'olau MounTable 2. Range, mean, and standard deviation of average corolla tube length, plant height,
and elevation for Cyanea taxa occurring on each of the four rain-forested islands of the
Hawaiian archipelago (see text). Data compiled from Lammers (1990, 1992, 1996), Rock
(1919), and personal observations. Some or all data are missing for C. glabra, C. linearifolia,
C. longissima, C. pinnatifida, C. remyi, and C. pycnocarpa
Island No. of
Corolla tube length
Plant height
Elevation
species Range
Mean ± s.d.
Range Mean ± s.d.
Range
Mean ± s.d.
Kaua'i
17
17-65 36.7 ± 14.0 1.0-14
4.1 ± 3.6
185-1215 631 ± 328
Q'ahu
15
24-75 53.5±17.3 0.5-6.0 2.7 ± 1.6
320-950 562 ± 147
Maui
21
18-75 49.3 ± 17.0 1.5-8.0 4.4 ± 2.3
455-2130 1127 ± 465
Hawai'i 12
22-75 51.1 ± 20.4 1.2-10
4.0 ± 2.8
465-1850 1064 ± 451