The Flowering Plants of Hawaii
Part 26 Malvaceae
The 11th largest contributor of species to the Hawaiian flora is the cotton family, Malvaceae. Malvaceae are a relatively large family with about 100 genera encompassing 2,000 species, according to the Manual. To demonstrate the changes that are underway in the business of reassessing relationships at higher taxonomic levels (largely based on DNA data), it is interesting to note that the family is seen as much larger in recent treatments. Mabberley (pp. 518-519), for example, gives 113 and 5,000 as the number of genera and species, respectively. Several previously recognized families have been absorbed into Malvaceae one of which, the Sterculiaceae, we will meet below.
The family has many members that would be familiar to most readers including cotton (Gossypium spp.), okra (Abelmoschus), hollyhock (Alcea rosea), lavatera (Lavatera spp.), and hibiscus (Hibiscus spp.). The contribution made to the Hawaiian flora by Malvaceae involves at least 49 species representing 24 genera, including some of the common ones just mentioned. Many of these genera make significant contributions to floral variety in the islands; several also serve as excellent examples of the tenuous state of certain elements within the flora. We can begin our introduction to the family with one of the most well known members, Gossypium, the genus to which the cotton of commerce belongs. Currently, 49 species of Gossypium are recognized worldwide, but only a few produce the lints from which useful threads can be produced. The endemic Hawaiian cotton, G. tomentosum, ma`o or huluhulu, (see image), belongs to the group of species that do not produce useful lints (see image). Ma`o occurs in dry habitats, often on comparatively poor soils, on all except the Big Island. The photograph of a community of ma`o (see image) was taken in an arid area on the northern slope of Läna`i, in January. The two naturalized cottons on the islands, both native to the New World, are G. hirsutum, upland cotton; and G. barbadense, sea island cotton, or pulupulu haole in Hawaiian. Cultivation of cotton was atempted in the islands in the 19th century but never developed into a major industry.
In order to discuss the origin of the Hawaiian endemic species of cotton it is necessary to give a little background, although the entire story of cotton’s history is matter for a book in its own right. Cotton species fall into two groups, diploids (two sets of chromosomes), which comprise the bulk of species and are native to the Old World; and tetraploids (four sets of chromosomes) of which there are six species. Four tetraploids occur in the Americas: G. barbadense (sea island cotton), whose range is centered in northern South America and the Caribbean; G. hirsutum (upland cotton), which is centered in Central America and the Caribbean but also extends into northern South America; G. mustelinum, a native of northwestern Brazil; G. lanceolatum, a native of southwestern Mexico; G. darwinii, a native of the Galapagos Islands; and G. tomentosum (once called G. sandwicense), the Hawaiian endemic of interest to us here. The question, then, is which of the continental tetraploid species is the most likely ancestor. The use of morphological comparisons did not resolve the problem because each set of comparisons—floral and fruit characters, leaf structures, pigment chemistry—led to a different conclusion. Breeding biology did not identify the most likely ancestor since the Hawaiian endemic species crossed with all of the continental species, but did serve to indicate that comparatively little genetic differentiation has occurred. DNA sequence data, however, pointed to G. hirsutum as the most likely ancestor. Since G. hirsutum occurs over a fairly wide area a search for was undertaken for genetic clues that might indicate a particular area from which ancestral plants came. No such clues were found suggesting that the island species diverged during an early phase of diversification of G. hirsutum. DNA sequence data did, however, indicate that development of the tetraploid form of cotton occurred between one and two million years ago, well within the time that Hawaiian Islands existed (DeJoode and Wendel, 1992).
It seems likely that transoceanic dispersal of a seed must have occurred. Cotton capsules are buoyant and seeds remain viable for up to three years as has been determined by studies using artificial sea water.
Equally well known to most readers, I suspect, are members of the genus Hibiscus—collectively known as aloalo in Hawaiian. The genus should be familiar owing to the contribution that many members (ca. 675 species according to Mabberley, only 200 according to the Manual) make to the horticultural industry. Ten species of Hibiscus occur in the islands, five endemic, two indigenous (with some question), and three that have become naturalized. The most common species that visitors will likely see in its native habitat is H. tiliaceus, known in Hawaiian simply as hau (see image). Hau occurs widely in the tropics and subtropics preferring coastal, streamside, or other wet habitats. A dense colony of hau can be seen growing beside the Hanalei River (Kaua`i) (see image). In the Hawaiian Islands hau occurs on Midway, French Frigate Shoals, and on all of the main islands except Ni`ihau and Kaho`olawe. It is not possible to establish whether hau was a Polynesian introduction—it is well known in other parts of Polynesia—or whether it arrived by flotation. The seeds of hau are buoyant and remain viable for several months in sea water. As in the case of other island plants, coconut for example, there is no way to determine which route was taken. Perhaps hau followed both routes.
In contrast to the abundance of H. tiliaceus, most of the endemic species, or one of their forms, are rare and in some cases may consist of only a few individuals in nature. The endemic species H. arnottianus, a native of the mountains of O`ahu and Moloka`i, is thought to have several forms. The one shown in the photograph (see image) is subsp. immaculatus, a rare form native to the steep northern cliffs of Moloka`i. Only a few individuals are known in the wild; this photograph was taken at the Maui Nui Botanical Garden where efforts are underway to help preserve this form. Hawaiian names for this beautiful plant are koki`o ke`oke`o, hau hele, koki`o kea, and pämakani.
The next species, H. kokio, is also variable throughout its range, which includes all of the main islands except Läna`i. Two subspecies are recognized, the Kaua`i endemic. subsp. saintjohnianus (see image above), and the more widely distributed subsp. kokio (see image), which is called koki`o, koki`o `ula (`ula = red) in Hawaiian. Subspecies saintjohnianus was photographed in the Limahuli Botanical Gardens. Hibiscus waimeae (see image above) is also a native of the island of Kaua`i where it can be found, although not easily, growing in Waimea Canyon and valleys on the island’s western and southwestern flanks. This specimen was photographed in the NTBG greenhouse.
Hawaiian names for this species are koki`o ke`oke`o, and koki`o kea. Petals of this species are white, a feature noted in the second Hawaiian name where kea, the Hawaiian word for white, appears as the descriptor. An Image of the endemic species of Hibiscus, H. clayi from eastern Kaua`i (see image) was provided by the Starrs. The last example here is H. brackenridgei (see image), which has been reported from all of the main islands. This latter species is called ma`o hau hele in Hawaiian. One of the meanings of hele involves the idea of ‘everywhere’ suggesting that this yellow hau is widely distributed, which, as the Manual informs us, is the case.
The Hawaiian Islands are home to two endemic genera in this family, Kokia and Hibiscadelphus, both of which are in serious trouble. Kokia consists of four species, one of which, K. lanceolata, is extinct, not having been seen in nature since the early 1900s. Its native range was southeastern O`ahu, an area that has seen overwhelming development where little in the way of natural vegetation remains. Kokia drynarioides, native to dry slopes in the North Kona district of the Big Island, is nearly extinct in the wild but is being maintained in cultivation. Two images of this species are included, one showing the flower structure and leaves (see image), the other showing a young tree (see image) at the NTBG. The other member of this little group that we can look at, Kokia cookei (see image), one of the most beautiful members of the family. The natural range of K. cookei was on the slopes of Maunaloa, the extinct volcano in western Moloka`i, an area that has been totally devastated through extensive pineapple cultivation and related developments. Although extinct in the field, this species is being maintained at the NTBG by grafting onto K. kauaiensis. Kokia kauaiensis, native to valleys on northern Kaua`i, is itself a very rare species. Although originally known from a single individual, its situation is somewhat less dire with the subsequent discovery of several additional trees. The generic name for this set of species is based upon the Hawaiian name for the group, koki`a. Another name for the group is hau hele `ula, which can be translated as being similar to a hau but with red ( `ula) flowers.
Hibiscadelphus is described in the Manual as consisting of six species, but a recent report (Lorence and Wagner, 1995) describes a seventh, H. woodii, recently (1994) discovered in the Kalalau Valley in northwestern Kaua`i; only four plants were found none of which appeared to be reproductive. Three species are extinct: H. bombycina, collected once, before 1868, on Hawai`i; H. crucibracteatus, known from a single tree (now dead) discovered in 1981 on Läna`i, but whose seeds failed to germinate; and H. wilderianus, known from a single tree found in 1910 on the southern slopes of Haleakalä. The three remaining species are in perilous condition, as one can judge from information on each gleaned from the Manual. Hibiscadelphus distans, when discovered in 1972, consisted of a single population of 10 individuals near Waimea Canyon, Kaua`i. Subsequent visits to the site revealed that the population had grown to 16, but current status is unknown owing to possible catastrophic hurricanes that have devastated the area, `Iwa in 1982, and Iniki in 1992. (A visit to Kaua`i in 1994 revealed great swathes of forest that had been leveled as if by a large scythe—the scene was horrendous.)
Hibiscadelphus giffardianus, known from a single tree discovered in 1911 on the eastern slopes of Mauna Loa, is probably extinct in the wild, but is being maintained in cultivation by Hawaii Volcano National Park staff (see image) The situation with H. hualalaiensis is much the same: in 1977 two or three individuals were known to exist on Hualälai Volcano (Big Island); it is likely extinct in nature, but it is being maintained in cultivation.
One might ask if there is some reason that this group of species is uniquely in such trouble? The answer is no. There is no extraordinary phenomenon at work here; it is the confluence of two well known biological situations. The first of these involves population size. For whatever reason, a species may consist of a comparatively small number of individuals, a situation that is common in the islands. Problems can arise when events occur that markedly alter the environment of the species in question, such as a volcanic eruption, a landslide, a brush or forest fire, a hurricane, attack by a pathogenic microorganism, or disappearance of pollinators. In the case of Hibiscadelphus, it is thought that the last explanation, loss of pollinators, led—and indeed may be leading—to extinction of the genus. A few words about pollinators are needed to set the stage.
For our purposes, we can define two groups of pollinators, generalists and specialists, and comment on the types of flowers with which each is associated. Open-faced flowers, daisies for example, present an easy target for potential pollinators: they fly in, collect some pollen and nectar, and fly off to another flower. For lack of a technical term one could call this landing-field pollination. If plants have tubular flowers, however, the scene changes dramatically. Big insects can’t fit into small floral tubes, while small insects may get lost trying to find their way around in large tubes. Through the course of evolution many plants and their pollinators have become closely associated with one another; this is true with both insect- and, as we shall see, bird-pollination systems. These adaptations between floral structure and pollinator lead to maximum efficiency of visits and pollen transfer. This is co-evolution at work. Readers familiar with orchids and their requirement for specific insect pollinators will have seen this phenomenon in action. A similar situation exists with Hibiscadelphus whose curved floral tubes require pollinators with similarly curved beaks enabling them to reach the bottom of the tubes for their nectar reward. This service was provided by the honey-creepers, a group of Hawaiian birds with curved beaks. With the disappearance of the honey-creepers, and nothing to fill their unique niche, Hibiscadelphus lost its pollinators. With no pollinators, no seed was set, and with no seed available to give rise to the next generation, these species were doomed.
Why did the honey-creepers disappear? According to Jared Diamond (1984), a bird biologist with broad international experience, Hibiscadelphus has been the victim of “trophic cascade.” This is a technical term for what we more familiarly call the domino effect. Simply put, change something at one point in the food chain and something further along the path will be affected. The intervening factor in the case of Hibiscadelphus was the death of honey-creepers who had become infected with avian malaria when bitten by mosquitoes. Malaria had killed the birds; dead birds don’t pollinate; un-pollinated flowers do not set seed.
But, mosquitoes in the Hawaiian Islands? Yes, and there are lots of different kinds. The particular animal that interests us here, though, is Culex quinquefasciatus, which had become naturalized in the Hawaiian Islands by 1826. This species is a known vector for avian malaria and avian pox; it traveled by catching a ride through the activities of unsuspecting humans. Within a few years this insect had spread to all of the islands and was well established in low elevation wet forests (up to 1,800 m elev.). The result of this invasion was the death of large numbers of lowland bird species, including the honey-creepers so essential for pollination of Hibiscadelphus, as well as other species (T. Smith et al., 1995). Mosquitoes require standing water in which to breed, an environment provided by humans through construction, as well as by general sloppy upkeep of property, and by feral pigs who provided an abundance of puddles through their diggings. Other mosquito breeding sites include water collected in tree stumps, or in cavities of tree-ferns brought to the ground by pigs looking for the tasty inner soft parts of the plants.
From time to time I am asked whether it wouldn’t have been possible for the malaria-prone bird species to have developed some immunity to the disease? The answer to that question is yes, it is theoretically possible for immunities to develop, over time. The key word here is ‘time,’ but an equally important factor, as noted above, is species, or population, size. If population numbers drop below some critical value—and these numbers are different for different organisms—there will be too few individuals surviving to mate successfully. In the case of the malaria-afflicted birds, the population size was likely too small for there to have been sufficient variation in susceptibility to disease upon which selection could work. Selection for new, adaptive traits takes time, normally thousands of generations.
Complicating matters even more, the elevation at which mosquitoes can survive has been gradually rising over the recent past as a result of global climate change. Birds that were once safe from infection are now at risk owing to the upward expansion of the mosquitoes’ ranges.
Abutilon is a moderately large genus with about 160 species widely distributed in tropical and warm areas (Mabberley, p. 2). The Hawaiian Islands are home to seven species, three of which are endemic and rare: A. eremitopetalum from arid western Läna`i; A. sandwicense from the Wai`anae Mountains on O`ahu; and A. menziesii, ko`oloa `ula (see image), originally from Läna`i, East Maui, and the Big Island. The photograph of ko`oloa `ula was taken at the NTBG. Flower color is quite variable in this species ranging into the red as suggested by `ula in its Hawaiian name.
Abutilon eremitopetalum (image) is a rare species endemic on Läna`i. Abutilon incanum(image), which is known from deserts of southern California and Arizona and adjacent states in Mexico, is marked in the Manual as possibly indigenous on the islands. The most widespread species of Abutilon in the islands is the pantropical weed A. grandifolium (image), commonly known as hairy abutilon or ma`o in Hawaiian. This plant can be found in waste fields and roadsides on all of the islands.
An article in the March 2008 issue of Environment Hawai`i gave a good indication of how sensitive some of the island species are to environmental conditions. A land survey in a scrub zone on western Hawai`i (Kona side of the Big Island) associated with development plans for “Villages at `Aina Le`a“ in 1991 revealed 38 individuals of ko`oloa `ula in one of the gulches that crosses the property. Various legal issues arising from the planned development required that the area be resurveyed in 2000. A visit to the gulch showed no plants of this species. There was speculation that their absence could be accounted for by extreme drought and wind conditions. The possibility that seeds exist in the surrounding soil was mentioned, but I am unaware if any further work on this site has been done. A quick search of the Web, however, did reveal an attractive advertisement with information dated mid-February 2012.
One of the most widespread members of the family that visitors will meet on the islands is the indigenous `ilima, Sida fallax (see image). It is also one of the most variable species with growth forms ranging from prostrate to upright shrubs. The flowers are popular for making lei. The degree of variation in this species prompted authors of the Manual to suggest that serious study could well result in recognition of infraspecific groupings (varieties or subspecies). Hawaiians were very much aware of variation within `ilima. In addition to the clearly obvious color differences among different plants—light yellow (ha lenalena), strong yellow (melemele), and bronze red (`ula`ula)—they recognized several varieties (varieties in the common usage, not in the formal botanical sense): low, spreading beach plant ( `ilima kukahakai); tall form growing on pasture land (`ilima ku kula); small flowered form growing on Ka`u lava (`ilima makana); domesticated form used for lei making, tall spreading bush (`ilima lei or `ilima `apiki); and a domesticated form with bronze-red flowers that grows on O`ahu (`ilima kolikukui). It would be of interest to see if these forms have biological reality, which, of course, requires a comprehensive study of morphological and macromolecular features of the species.
I would be remiss not to mention that a colleague, Douglas Pratt, in his excellent little 1998 book Pocket Guide to Hawai`i’s Trees, pointed out that two taxicab companies in Honolulu have used this plant’s names for their fleets. Thus, it is possible to take either the Latin named cab, “Sida”, or the Hawaiian-named cab, “Ilima”, for traveling about the city.
The next member of Malvaceae that we meet here is Thespesia populnea, known as milo or the Portia tree. Since milo occurs widely in littoral habitats throughout the Pacific Basin it is not possible to say with certainty if it got to the Hawaiian Islands by natural means, or whether it was brought by Polynesian colonists who are likely to have known about the usefulness of its wood. This tree is most readily identified by its characteristic display of fruits as seen in (see image). Its flowers (image) are very similar to those of hau.
Malvaviscus, members of which are commonly called Turk’s cap (see images), sleeping hibiscus, or cardinal flower in English, or aloalo pahüpahü in Hawaiian, has been treated in different ways. At least one authority sees the genus as consisting of one highly variable species, while others recognize several species. For example, the Manual lists the only Hawaiian representative as M. penduliforus, where references to the horticultural literature, for example, will cite it as var. penduliflorus of M. arboreus. Mabberley (p. 519) lists Malvaviscus as a synonym of Hibiscus, a view that emphasizes similarities among the various forms. Regardless of its taxonomic status, Turk’s cap is attractive and widely planted, often as a hedge plant as seen in one of the images.