North Central Washington

Radiation supercedes speciation following relaxed environmental constraints

Or which came first, the chicken or the egg?

Or, better yet: Paleontologists Need to Give a Fig about Plant Evolution

Review by George Wooten

Plants and flowers of Hawai‘i, by S. Sohmer and R. Gustafson, Univ. of Hawaii Press, 1987.

The flora of the Hawaiian archipalego is one of the best known and most thoroughly studied in the world. A book by Sohmer and Gustafson (1987) squeezes in a wealth of information about the origin and development of the Hawaiian flora. The book outlines the process that a plant species undergoes in adapting to its new environment (adaptive radiation). The details of this transformation are different from that of the classic, Darwinian model of natural selection. The classic model holds that a species adapts to new, unexploited niches by acquiring beneficial mutations that give greater advantage to each succeeding generation. In contrast, the Hawaiian plants appear to adapt faster than the rate of speciation.

The Hawaiian flora has many examples of plants that acquired novel physical forms (phenotypes) faster than the process of genetic divergence from the parent stock. This implies that the classic model that couples each genetic mutation with a new survival trait may be incomplete. The process of phenotypic variation appears to occur faster, and possibly independently from, genetic variation. What is the evidence for this, and could this be due to something more than Mendelian inheritance? Is this just a plant thing or is it more universal?

The provenance of most of Hawaii's 1,000 native species is known (Fosberg 1948), including approximately 30 species brought by early Polynesians (Abbott 1992). Based on phylogenetic relationships, Fosberg postulated that the native flora would have required about 272 successful introductions to exist as it did prior to the human introductions. The table below shows the provenance of the Hawaiian flora according to Fosberg's analysis:

Provenance Percent of native flora
Indo-Pacific 40%
Southern (Austral) 17%
American 18%
Northern 3%
Cosmopolitan 12%

The geology of the Hawaiian archipalego is understood well enough to ascribe a geologic age of about 5 million years to the present set of 132 islands. The Hawaiian Islands are the summits of the seamounts in the Hawaiian-Emperor chain. This line of 107 volcanoes stretches over 1,000 miles. It is formed by northwestward migration of the Pacific Plate at 3.4 in/year over a stationary magmatic hot spot in the earth. As the plate moves over the hot spot, new volcanoes erupt from the seafloor, eventually to become islands, and ultimately, to erode down and sink below sea level once again over millions of years. The islands are progressively younger toward the big island in the southeast, which has the most active volcanism. The age of the entire chain including the undersea seamounts is somewhere between 30 and 70 million years old.

Taking the figure of 70 million years for the age of the islands, the rate of successful colonization by Fosberg's 272 species is on the order of 1 every 260,000 years, or for the younger age estimate, about twice as frequent. What sort of process occurs as this rare event of becoming successful unfolds?

A disharmonic flora

In comparing the characterists of plants in Fosberg's list of provenance with the floras of their origins, it is quickly apparent that the Hawaiian flora is not representative of the sources. The lineages and lifeforms are disharmonic, that is they are very different in appearance and family membership from that of the original floras.

For instance, the Hawaiian flora has the world's higher proportion of dioecious species (15%; Saxai and others 1995), in which pollen and ovules are produced on separate plants. Outside of Hawaii, dioecy is uncommon. Monoecious flowers, or hermaphrodites, have both types of flowers on a single plant. Saxai's group also found that dimorphic plants that are not strictly dioecious account for 21% of Hawaii's flora.

But this is entirely at odds with what one would expect the Hawaiian flora to look like! A dioecious plant requires two individuals to reproduce. The odds are stacked against a seed, having travelled thousands of kilometers, arriving and then germinating and then flowering at the same time and place as another plant of the same species that it can cross with. There is even a rule called "Baker's (1967) law", holding that self-compatible individuals are more successful colonists because they do not require a mate. But these odds are somewhat mollified by Carlquist's (1974) determination that 40% of the colonists were carried internally in birds, making it plausible that two seeds could germinate and flower simultaneously.

It is risky to draw any conclusions from the prevalence of dioecy because it is also correlated with tropical plants, perennial habit, fleshy fruits, wind pollination and climbing growth (Sakai and others 1995). While Baker (1967) suggests that the dioecious plants arrived as monoecious plants and later evolved to become dioecious, Carlquist (1974) ascribes this primarily to preferential establishment of dioecious colonists. The analysis of Sakai determined that while 10% of the colonists were dimorphic, 21% of current species are dimorphic. At least one third of the dimorphs evolved from monomorphic colonists, while half of the dimorphs are in lineages arising from dimorphic colonists.

Hermaphrodites can self-fertilize, which confers an advantage to the plant during hard times. But such cloning has the disadvantage that inbreeding depression evenutally leads to a genetic dead end. Dioecious plants have an advantage in their ability to outcross and recombine with twice as much available DNA. The offspring of dioecious plants draw their genes from a pool twice as large as those of monecious plants and the result is greater variability in the offspring. Is this what is observed in the Hawaiian flora?

Yes. It turns out that amongst related species in the Hawaiian flora, there is a disproportionately greater amount of phenotypic variation than amongst the same groups elsewhere. Extreme examples of phenotypic variation are legion in allies of the Hawaiian silversword. The genera Dubautia and Argyroxiphium are very phenologically different from each other. The differences are great even within the genera, with Dubautia species containing both shrubs of dry lava beds as well as vines and obligate wetland species. Yet even though these species barely resemble each other, their recent divergence from a common ancestor is evident because they all frequently hybridize in nature.

link to silversword image
Haleakala Silversword in flower - Haleakala National Park, Hawaii - October 13, 2005.

Photo courtesy of Forest and Kim Starr - U.S. Geological Survey

The ability to form natural hybrids is another feature of the Hawaiian flora that is uncommon elsewhere. Hybridization maximizes variability, and it makes sense that this would be of benefit for a colonist presented with a variety of new habitats to colonize.

The features of floral disharmony described above, dioecy, hybridization and phenological variability, do fit reasonably well within Darwin's mechanism of natural selection and survival of the fittest genotype. But there is another twist that requires a different paradigm.

As might be expected, the ancestors of the Hawaiian floral colonists are dominated by plants having seeds or fruits that are small, lightweight or otherwise easily moved (Carlquist 1974).

But something truly remarkable happened to those small-seeded species once they arrived on the islands--they evolved into large-seeded species. Nearly every group of Hawaiian plants possesses the largest seeds of any other in their genus or species throughout the world. Carlquist (1970) gave the example of the Hawaiian species of Bidens, which in mainland ecosystems have long awns or barbs that allow them to hitch-hike rides on birds and animals. These awns are completely lacking in the Hawaiian species of Bidens, and the fruits are larger, rendering them unable to perform the hitch-hiking feat that got them to the islands in the first place. A likely explanation is given by Sohmer and Gustaphson (p. 30):

Another feature that probably bears upon the loss of dispersibility as species evolve from open coastal habitats into closed wet forest ones is the loss of the dispersal agent. Species that are dispersed by particular agents (wind, water or wing) will, as they evolve into new habitats, often leave those agents behind. Evolution, as a general rule, is usually very efficient. When there is no longer a "need" for something it is more efficient not to produce that something and eventually it is lost.

This bears repeating. The key point is that it is not selection pressure, but rather the loss of suppression pressure, that enables the change. It is unlikely that the gain of a new gene is occurring as would be the case if natural selection were the primary determinant. There are at least two genetic functions required for speciation to occur via natural selection (incompatibility, phenotypic variation) and yet these plants remain compatible. A likely explanation for the change of appearance is that there is a loss of gene suppression, or the removal of some of the gene products that provide negative biofeedback to the products of gene transcription and translation. Michael King (2011) enumerated eleven different mechanisms of gene control in eukaryotes that could account for this.

Compatibility (the ability to hybridize) is a good threshold to use as a surrogate for speciation. It is at least partly dependent on background mutation rates that underly natural selection. The tenets of natural selection, speciation and genetic drift are all fundamentally tied to compatibility. A single well-placed mutation of any number of genes would be sufficient to disable compatibility. Thus, by their very existence, compatible species have not had enough time to speciate. Gradual loss of compatibility is documented in livestock, suggesting it may occur at a fairly constant rate over longer time periods. On the other hand, phenotypic expression requires acquisition of complexes of genes with multiple functions (operons), suggesting changes should generally occur slower than loss of compatibility. But operons can be stored for a long time in a suppressed form (King 2011).

For compatibility to be maintained in a "species" with gross anatomical differences, the question of which came first is answered by the chicken, not the egg.

Some of you may worry that this is courting Lamarkism--the theory that reproductive traits are acquired during life. No, this is different, not that Lamarkism is all bad, either. Traits acquired through natural selection still take long times to evolve. But examples of Lamarkism have now been recognized. A stellar example is the transfer of genetic material between bacteria via an F(+) pilus that sometimes breaks off the marriage due to an environmental disturbance, leaving the gene transfer half completed and leaving the F(-) recipient with only half of a new set of genes. This well-known process is so much richer in possibilities over what can be drawn by natural selection alone, that we are past due for an evolution revolution.

And there is no reason to dictate that the mechanisms of radiation should be wholly genetic. King (2011) described a number of repressor mechanisms, both genetic and non-genetic (epigenetic). Phenotypic expression may be controlled by regulatory sequences, intergenic segments, chromosomal structures, developmental ontogeny, prions, the order of gene transcription, and a number of other means. There are segments of DNA whose purpose is to stop and start replication, and these segments are themselves regulated by higher order processes and environmental feedback. Some forms of cancer appear to be an uncontrolled return to natal physiology.

All of this bears consideration in regard to the theory of evolution. It indicates that sometimes, perhaps most of the time, random mutation may be insignificant when compared with the effects of relaxing controls on constrained operons. Before a mutation occurs, a species is already exerting a large amount of energy to maintain its genetic configuration. Radiation into new forms can occur rapidly when these constraints vanish, freeing the energy for other pursuits. In the Hawaiian examples, this radiation happens faster than the rate of speciation, so that we have radically different life forms that are still capable of cross fertilization. This supports the idea that even tightly conserved genes that are necessary for survival may not need to wait eons for natural selection to confer a new gene. The mechanism for change of form is already in place--all that is needed is to switch off the controls against it.

Our current understanding of evolution would benefit from a greater emphasis on plant genetics. Experimental studies of plant genetics can control for variables that are confounding theories based on animal studies, e.g., rate of environmental change, rate of drift, background mutation rate, data gaps in the record, etc. When combined with theories of island biogeography, plant genetics can reveal many factors of evolution that are important to all scientists and interesting to all of us.

Analogously to turning off the gene repressors, our appreciation for evolution would benefit by turning off the neo-Darwinian rhetoric that has cast us in a war over monkeys.

Botanists are renegades from the -ologists. They must be, to sit by and watch their paleontological colleagues get splattered across the pages of National Enquirer, while failing to bring up Bidens.

I suspect that botanists are not so smug about the concept of species either. We are used to seeing plants behave aberrrantly, cloning themselves, and even immortalizing themselves through apomixis. Are botanists afraid of speaking out for fear of castigation from the inner circles of the neo-Darwinists and the funding sources?

Maybe, but more likely their work is just not appreciated beyond the fen. Fosberg published his work 60 years ago, Carlquist had his time in the spotlight 40 years ago and Sohmer and Gustafson spelled it out again as clear as day over 20 years ago. Paleontologists need to give a fig about plant evolution. Some prominent scientists are so busy buttressing natural selection that they have failed to note its further decline in steamy scientific speakeasies, the fern journals, where the evolution revolution lies quietly underappreciated.

But times change. New theories are circulating amongst the flock. A theoretical framework for developmental evolution by Snell-Rood and others now considers a broad array of species, that aren't just primates or mollusks. (see this paper by Snell-Rood and others (2010). And every once in a while, some scientist will trigger a tipping point that allows a collective "ah-ha" moment to occur.


  • Abbott, Isabella (1992). La‘au Hawai‘i: Traditional Hawaiian Uses of Plants, Bishop Museum Press, Honolulu.
  • Baker, H.G. (1967). Support for Baker's law - as a rule. Evolution, 21: 853-6.
  • Carlquist, S. (1974). Island biology. Columbia Univ. Press, New York, 660 pp.
  • Carlquist, S. (1970). Hawaii, a Natural History. Natural History Press, New York, 463 pp.
  • Fosberg, F.R. (1948). Derivation of the flora of the Hawaiian Islands: in Zimmerman, E.C., Insects of Hawaii. Vol. 1. Univ. Press of Hawaii, Honolulu, pp. 107-119.
  • King, Michael, IU School of Medicine (2011). Gene Control in Eukaryotes
  • Sakai, Ann K.l Warren L. Wagner, Diane M. Ferguson, Derral R. Herbst (1995). Biogeographical and ecological correlates of dioecy in the Hawaiian flora. Ecology. 23 Feb, 2011.
  • Sohmer, S.H.; R. Gustafson. 1987. Plants and flowers of Hawai‘i, Univ. of Hawaii Press, Honolulu.

Written and researched by:
George F. Wooten
226 West Second Avenue
Twisp, WA 98856