Genetic diversity in Asteraceae endemic to oceanic islands: Baker’s Law and polyploidy

Abstract

When considering the two factors ploidy and breeding system of ancestral colonizers together from the fi ve archipelagos, two things are most evident. First, the Canary Islands are exceptional because of the frequency of diploid colonizers compared to the other island groups. The hypothesis advanced to explain this is the much closer proximity of the Canaries to a continental source area than the other island systems; thus, multiple early introductions may have reduced the selective advantage of the greater genetic diversity aff orded the single polyploid propagules dispersed to more remote archipelagos. The other notable observation is the greater frequencyof SC ancestors for Hawaii relative to the other island systems. It is tempting to suggest that the higher frequency of SC colonizers in Hawaii is due to the greater distance to source areas; however, the most spectacular radiation in Hawaii, the silversword alliance, originated from SI or PSF ancestors. An equally large radiation in Hawaii, but perhaps not as spectacular in terms of ecological and morphological diversity as the silverswords, is Bidens where the thirty species originated from a SC ancestor. Bidens in Hawaii is a high polyploid, so a single ancestor colonist could have carried extensive diversity to the island. Although species of Bidens are SC, there has been evolution of sexual expression in the lineage, including factors that promote outcrossing, with the net result that species exhibit a mixed mating system. One of the lineages that seems to “break the rules” for Hawaii, and indeed for the other four archipelagos, is Tetramolopium. The colonizing ancestor of this ecologically and morphologically diverse lineage was both SC and diploid . The progenitor of Hawaiian Tetramolopium is from New Guinea and possible dispersal mechanisms are birds and wind; the pappus and glandular trichomes on the fruits could facilitate adherence to the feathers and feet of birds. Though highly speculative, it is suggested that there may have been multiple fruits in a single dispersal event, and Tetramolopium evolved from more than a single colonizer. Tetramolopium is an illustration of the limitation of attempting to interpret the genetic diversity of colonizers with only two variables of ploidy and breeding system. While the dispersal agent of Asteraceae isnormally the individual fruit, this does not mean that single dispersal events involve only one fruit, as was admitted by Baker (1967). However, an important point with regard to Tetramolopium, and other lineages originating from SC colonizers, is that if a single dispersalevent included more than one fruit from the same selfi ng population, then the multiple fruits may not collectively contain signifi cantly more genetic diversity than a single fruit. In contrast to multiple propagules from selfi ng source populations, dispersal of multiple fruits in a single event from an outcrossing source population would enhance genetic diversity in a founding population relative to a single propagule. This may explain why the colonizing ancestors of so many island lineages were SI (or more likely PSF) regardless of whether there was one or multiple propagules in a single dispersal event. Despite the limitations of considering only the two variables of ploidy and breedingsystem, observations for the fi ve archipelagos show that both polyploidy and SI accompanied by PSF are common attributes of successful lineages. It seems reasonable to hypothesize that these two factors, both alone or in concert, provide colonizers with the requisite diversity necessaryfor radiation and speciation in an archipelago. There are plausible hypotheses to explain the two notable exceptions to the above generalizations, the paucity of polyploidy in the Canary Islands and SC colonizers in Hawaii.