Oceanic island archipelagos are profoundly interesting ecosystems in which to ask questions about evolutionary patterns and processes. Their isolation from other land masses restricts numbers of plausible hypotheses that can be advanced to explain evolutionary events, often resulting in stronger scientific inferences. Island archipelagos may rightly be considered one of the best places on earth to understand origins and elaborations of biological diversity.

In recent years, interest in the evolution of plants on oceanic islands has increased with work appearing on the Hawaiian Islands, the Canary Islands, Bonin Islands (Japan), Juan Fernandez Islands (Chile) and numerous additional archipelagos. It seemed pertinent and timely, therefore, to bring persons together with experience in these different island groups for a new view of island plant biology.

The present papers resulted from a symposium convened by the coeditors at the XV International Botanical Congress in Yokohama, Japan, August 1993. The response from that symposium was very positive and suggested that publication of the presentations (in modified and up-dated form) would be highly desirable. Cambridge University Press agreed to take on the project. In addition to papers from the symposium itself, which focused on Pacific archipelagos, several additional invited manuscripts were solicited to provide an even stronger and more comprehensive review of the status of studies of plant evolution on oceanic islands.

It is clear from the chapters of this book that oceanic archipelagos do offer special opportunities for investigating evolutionary phenomena in vascular plants.

Abstract

The Juan Fernandez Islands are located 667 km west of continental Chile at 33° S latitude. They contain 104 endemic angiosperms. Although numerous isolating mechanisms and modes of speciation are known for plants of continental regions, few have been documented for those on Juan Fernandez. A survey of endemic angiosperms reveals spatial isolation (between islands) within the archipelago as the most important primary factor accounting for 70% of closely related species pairs. Environmental (habitat) isolation occurs within individual islands and is known in 18% of cases. External reproductive isolation is less common, with temporal isolation accounting for 7% of cases, mechanical isolation only 3% and autogamy known in only two situations (3%). There is no indication of internal reproductive isolation between close congeners, although direct experimental evidence is lacking. Almost no changes in chromosome number or reticulation during speciation are known. Isozyme and DNA comparisons show low levels of genetic divergence among endemic congeners, yet conspicuous morphological differences often exist. The eroded environment of the older island, Masatierra, gives fewer clues to geological and ecological components of speciation. The younger island, Masafuera, has stronger ecological zonation and species of some genera (e.g., Erigeron) correlate with particular environments. The overall pattern of speciation in the archipelago, therefore, is divergence at the diploid level, initiated by geographic isolation especially between islands, accelerated by ecological partitioning within islands, accompanied by marked morphological change in response to changing environments, but with relatively minor genetic modification.

There are many oceanic islands scattered in the huge area of the western and southern Pacific Ocean. Their origin is diverse; most of these islands are volcanic, but some of them are coral reefs and/or even of continental origin, having previously been connected to other land masses. The flora of these islands is also diverse, with floristic elements of tropical South East Asia, including Taiwan and the Philippines, of East Asia such as China, the Korean Peninsula and the Japanese Archipelago, and of Polynesia or Australia. Among these areas, remarkable plant speciation and evolution has occurred on the Bonin Islands (Japan), Ullung Island (Republic of Korea) and on some South Pacific Islands such as Fiji, New Caledonia, Samoa, the Society Islands, the Solomon Islands (including Vanuatu) and the Marquesas Islands. These archipelagos provide the focus for this part of the book.

Chapters 6 to 8 discuss aspects of the Bonin (Ogasawara) Islands, located 26°30′ N–27°40′ N and 142°00′ E–142°15′ E. The area occupied by the islands is very small; even the largest island is only c. 24 km2. They are isolated from any other continental or large land mass such as Taiwan, the Philippines or the Japanese mainland. The flora of the Bonin Islands consists of floristic elements of tropical South East Asia including Taiwan and the Ryukyu (Okinawa) Archipelago, of continental East Asia such as China, the Korean Peninsula and the Japanese mainland, and even of the Polynesian Islands and Hawaii (e.g., Santalum and Metrosideros).

Abstract

Crossostylis (Rhizophoraceae) is one of the inland genera of Rhizophoraceae. Its distribution is on the South Pacific Islands, encompassing New Caledonia, the Fiji Islands, Vanuatu, the Solomon Islands, Samoa, the Society Islands and the Marquesas Islands. Species of this genus show morphological diversity especially of the inflorescences, flowers and seed coats. Based on cladistic analyses of morphological characters and restriction site variation of chloroplast DNA (cpDNA), we discuss phylogeny, speciation and evolutionary trends of morphological characters in Crossostylis. Cladistic analyses based on morphological and DNA data yield the same topological cladogram. Phylogenetic analyses suggest that species of Crossostylis are divided into two pronounced monophyletic groups: one comprising six species distributed on the Solomon Islands, Vanuatu and the Fiji Islands, and the other comprising four species distributed on New Caledonia and Polynesia. Monophyly of the former group agrees with general assumptions of phytogeographical affinities, but the monophyletic Polynesia–New Caledonian species group is not coincident with general floristic affinities for the region. All species endemic to the Fiji Islands are very closely related and this suggests speciation from a single ancestor on the islands. Some morphological characters were optimized on the cpDNA tree obtained and examined for evolutionary trends. Many of the floral morphological features support macromolecular clades, but all seed coat characters are homoplasious. Perhaps this reflects high adaptive significance, such as is involved with seed dispersal.

Crossostylis J. R. Forst. et G. Forst, one of the inland genera of Rhizophoraceae, is distributed on the major islands of the South Pacific.

Abstract

Speciation in angiosperms is often accompanied by change in chromosome number via euploidy or aneuploidy. Evolution involving autopolyploidy and to a much greater extent allopolyploidy has been documented on numerous occasions. In fact, it is estimated that more than 60% of angiosperms exist at the polyploid level, having resulted from many different allopolyploid reticulate evolutionary events. Evolution via aneuploidy is also common, with both ascending and descending modes known (the latter more common, however). Stimuli for rapid chromosomal evolution include intertaxon hybridization and rapid environmental change. Floras of oceanic islands contain numerous endemic taxa which have evolved from continental ancestors and often have further speciated within the archipelago. Chromosomal surveys of the Hawaiian, Juan Fernandez, Bonin and Galapagos Islands reveal very little change in chromosome number during the evolution of endemic taxa, even though speciation is often accompanied by marked morphological divergence. More continental islands, such as the Queen Charlotte Islands, and older oceanic islands closer to mainland sources, for example the Canary Islands, likewise reveal patterns of chromosomal variation. Explanations for absence of change in chromosome number on oceanic islands include low levels of hybridization due to habitat exclusion of endemic taxa, short periods of geological time and selection against aneuploid cytotypes which might disrupt the adaptive complex of traits that led to successful establishment, colonization and radiation.

Speciation in angiosperms is often accompanied by change in chromosome number via euploidy or aneuploidy.

Abstract

The biology and phylogeny of Dendroseris and Robinsonia (Asteraceae), two genera endemic to the Juan Fernandez Islands, are discussed and contrasted. Morphological and molecular data were used to test the monophyly and to generate phylogenetic hypotheses for the two genera. Restriction site mutations in the chloroplast DNA (cpDNA) and intergenic spacer (IGS) region of the nuclear ribosomal DNA (rDNA) were used to produce a phylogeny of each genus, and sequences from the internal transcribed spacer (ITS) region of the nuclear rDNA were also employed to generate phylogenies. All molecular data sets strongly support the monophyly of each genus despite the morphological and ecological diversity found in them. For Dendroseris, all molecular information, in concordance with morphology, shows that the two subgenera Dendroseris and Phoenicoseris are monophyletic, but this is not true for subg. Rea. The restriction site data for Robinsonia do not provide any phylogenetic resolution whereas the ITS sequences produce a completely resolved tree that is concordant with relationships inferred from morphology. Allozyme diversity is higher in species of Robinsonia than in those of Dendroseris and divergence between species is likewise higher in Robinsonia. Differences in diversity may reflect factors such as amount of genetic variation carried to the island by the original colonizers, population sizes and breeding systems. Higher allozyme divergence between species of Robinsonia could result from sorting of alleles from polymorphic ancestors during speciation. Also, Robinsonia may have been on the Juan Fernandez Islands longer than Dendroseris, thereby allowing time for the accumulation of unique alleles via mutation subsequent to speciation.

Abstract

In this chapter we report the results of studies of genetic diversity of plants on the Bonin Islands, which are oceanic islands in the western Pacific, and compare these results with previous studies in other oceanic archipelagoes.

Oceanic island endemics have lowered genetic diversity in comparison with continental taxa. In populations of endemic plants on the Bonin Islands, low genetic diversity is also observed, and population size and heterozygosity are highly correlated.

In plants of the Bonin Islands, some genera show adaptive radiation. We have studied electrophoretically genetic divergence among these taxa. High genetic identities were observed in Pittosporum and in some species pairs of both Crepidiastrum and Symplocos. This result indicates low genetic differentiation among species in spite of their morphological divergence.

All the endemic plants now found on the Bonin Islands are descendants of progenitors that were immigrants in the past and these lineages have been isolated from original ancestral populations. Average genetic identities between the endemics and mainland species are comparable to those between congeneric species (0.63), as estimated by Crawford (1983). We have estimated divergence time of these endemic lineages from ancestral source populations to be 2.25 million years for Pittosporum, 3.15 million years for Crepidiastrum and 2.10 million years for Symplocos. On the other hand, preliminary data of widespread plants show relatively high genetic identity between insular and mainland populations.

Several studies have measured the genetic diversity of endemic plants on oceanic islands, such as on the Hawaiian Islands (Lowrey & Crawford, 1985; Helenurm & Ganders, 1985; Witter & Carr, 1988), the Juan Fernandez Islands (Crawford et al., 1987a, 1992) and the Galapagos Islands (Wendel & Percival, 1990).

Few young students in the western world have failed to read the famous novel, Robinson Crusoe, by Daniel Defoe (1719). Although set in the Caribbean region, the story does relate to the real Robinson Crusoe Islands in the Pacific Ocean 667 km off the coast of continental Chile. These islands, also known as the Juan Fernandez Islands after the Spanish navigator who first discovered them in 1574 (Medina, 1974), harboured a Robinson Crusoe-type sailor for five years during 1704–9. After having suffered passage around Cape Horn, Alexander Selkirk had a dispute with Captain Thomas Stradling of the ship, Cinque Ports, and demanded to be set off at the next available land, which to his misfortune happened to be the Juan Fernandez Islands. In this solitude he endured a quiet and eventually rewarding experience that created genuine interest in journalists of the day (including Defoe) when he finally returned to England in 1711.

The Juan Fernandez Islands have been of even greater interest for economic and strategic reasons to European countries for over four centuries. The two greatest considerations were as a location for refitting boats and recuperating crews after the long and arduous trip around the tip of South America, and as a place to mount raids against ships and coastal towns throughout the colonial Spanish empire. Ships coming from Europe had an initial long journey across the Atlantic Ocean and down the eastern coast of South America, generally touching Brazil and Argentina.

Abstract

MacArthur & Wilson's (1967) equilibrium theory of island biogeography predicts that numbers of species on oceanic islands are dependent upon the size of islands, distance from major source area, and rates of immigration and extinction. Despite healthy criticisms of this theory, it has been enormously stimulating for understanding species diversity on islands as well as other island-like habitats. More recently, other workers have attempted more complex models to find better ways of predicting species diversity in oceanic archipelagoes. The Juan Fernandez Islands are well suited for specific model assessment due to the small size of the endemic and native flora (156 angiosperm species), the few major islands (Masatierra and Masafuera), their small size (50 km2 each) and proximity to the major source area (southern South America). Factors taken into consideration include size (both ancient and modern), distance from source area, extinction, speciation within and between islands, and differing modes of dispersal. A model for predicting species diversity is first developed to explain numbers of species on Masatierra, the island closest to the continent (667 km in the Pacific Ocean). This approach is then applied to explain species diversity on Masafuera, with encouraging results. The most important point is that numerous geological, ecological and historical factors impinge on determining the total number of species that can be supported on a particular oceanic island. Changing island size through geological time is especially important to consider.

Few islands of the world have received as much attention for evolutionary patterns and processes as the Hawaiian Islands. Reading through Sherwin Carlquist's stimulating Hawaii: A Natural History (1970) always elevates our interest. We have also read about fascinating evolutionary phenomena in picture-wing Drosophila, studied so successfully by Hampton Carson and colleagues (e.g., Carson & Kaneshiro, 1976; Kaneshiro, Gillespie & Carson, 1995; DeSalle, 1995). It is fitting, therefore, that the initial two chapters of this book deal with the Hawaiian Islands. Recent studies have greatly increased our understanding of patterns and processes in the endemic vascular plant flora of the Hawaiian Islands. A monumental achievement was the publication of the two-volume Manual of the Flowering Plants of Hawaii (Wagner, Herbst & Sohmer, 1990) that established for the first time a consistent species concept for the entire archipelago. In the past, some taxa had been split into numerous microspecies and others had been treated broadly, depending upon the perspective of the particular taxonomist. These disparate treatments of plant diversity in the archipelago made it very difficult to approach questions of speciation and biogeography. In fact, publication of the new Manual made it possible to conceive and execute a very meaningful project on biogeography in the archipelago, involving both plants and animals (Wagner & Funk, 1995). This would have been impossible without the consistent foundation of species concepts provided by the flora project.

The basic shape, form and structure of an organism (whether fungal, plant or animal) does not arise all at once. Rather, the shape and form emerge as a result of a sequence of developmental adjustments. Each of these is usually irreversible within its morphogenetic sequence although often reversible by some gross disturbance; for example, differentiated cells being put into tissue culture, nuclear and cell transplants, regeneration after injury, etc. The whole process in which the final organisation and pattern of the organism is established is termed ‘morphogenesis’.

The most extensive research on the topic has been done with animals and from this a vocabulary has been established which describes morphogenetic events without pre-judging the mechanisms which may be involved (Slack, 1991). It is evident that as the embryonic organism develops towards adulthood, each intermediate state represents a reduction in developmental potential compared with the previous state. Each adjustment (or developmental ‘decision’) is made by cells already specified by earlier adjustments to belong to a particular developmental pathway. Consequently, developmental decisions are made from among progressively smaller numbers of alternatives until the particular structure to which the cell will contribute is finally determined. Classic embryological transplantation experiments revealed these states. Where the explant differentiated to a state representative of its old position then it was said to have been determined prior to transplantation.

The basic genetic architecture of fungi is fairly typically eukaryotic, and all the major principles of genetics apply – Mendelian segregations, recombination, chromosomal structure, gene structure, etc. (Clutterbuck, 1995a). Nevertheless, there are some differences between most fungi and most of the rest of the eukaryotes. These will be summarised here; detailed information can be obtained from Carlile and Watkinson (1994), Elliott (1994) and Chiu (1996).

Fungi have a generally smaller genome size than other eukaryotes (Clutterbuck, 1995b). The yeast, Saccharomyces cerevisiae, for example, has a haploid genome of about 15 × 106 base pairs, which is less than four times the size of the genome of the bacterium Escherichia coli (Table 5.1). Fungal nuclei are consequently difficult to study by conventional cytological procedures because both nuclei and chromosomes are small, variable in shape and indistinct by conventional microscopy. Progress in understanding nuclear changes during the fungal life cycle has been slow, though in recent years the study has benefited greatly from application of electron microscopy and molecular techniques for analysing the karyotype.

Fungal mitotic divisions are intranuclear, unlike most animals and plants, which means that their division spindle is formed within an intact nuclear membrane. This makes the progress of the division even more difficult to see and study, but does not appear to affect the biological consequences of the mitotic division.

In this chapter I will describe and illustrate the formation of fruit bodies and related structures from their earliest stages (fruit body ‘initials’) through to maturation. This will include discussion of the cell types concerned and their patterns of distribution, the development of form and the way in which tissue domains are defined, and aspects of fruit body construction that relate to, or are determined by, the mechanics and physical structure of the object. Development is a dynamic process. In a conventional book, conveying the full dimensionality of morphogenesis is extremely difficult. Illustrations are in two dimensions, but morphogenesis occurs in four – the three dimensions of space and the fourth dimension of time. Please remember that as you now continue.

If the activities of any organism can be described as having a purpose, then the purpose of the activities described so far in this book is to provide the fungi concerned with reproductive potential. Resources that the invasive, exploratory mycelium has won are not squandered on vegetative growth. Rather, as we have seen, from a very early stage the mycelium puts aside reserves for later use in asexual and/or sexual reproduction. The network of metabolic regulatory mechanisms provides some sort of partitioning system that diverts part of the absorbed nutrient into reserve materials. With the exception of perennating structures, like dormant spores, overwintering sclerotia, or some perennial (bracket) fruit bodies, fungi store these reserve materials for short times.