(i) Preamble: why study islands? 

"The Zoology of Archipelagos will be well worth examination" (C. Darwin, 1835)

Islands can be broadly defined as areas of suitable habitat that are to some degree isolated from much larger surrounding areas of unsuitable habitat. In the classsical sense, islands are terrestrial habitats that are isolated from continental habitats by freshwater or marine areas that represent some degree of barrier to dispersal between the island and the continent. Other examples of "islands" include freshwater lakes for aquatic life, montane habitats surrounded by lower elevation habitats, and "nunataks" (mountain peaks isolated by large areas of snowfields).

Islands have always held a special place for biogeographers and evolutionary biologists because they are characterized by isolation, and isolation is a principal factor driving evolutionary change. In addition, islands often represent "simpler" ecosystems that are usually more tractable in terms of the processes that influence ecological and evolutionary processes operating within species and communities. The work of scientists like Darwin (Galapagos finches), David Lack (Galapagos and West Indies birds), Ernst Mayr (East Indies bird faunas) and G.E. Hutchinson (community ecology of freshwater lakes) that have contributed so much to evolution, ecology, and biogeography focused much of their work on island biotas.

Island biogeography usually dealt with three major aspects of island biotas:

(1) Descriptions of the diversity and composition of island biotas and how they differed from continental flora and fauna, and the nature of adaptations that influence dispersal to and colonization of islands,

(2) Identification and quantification of factors that influence rate of dispersal to islands, rates of extinction on islands, and the numbers and kinds of species islands can support,

(3) Evolution of communities in "novel" environments following colonization of islands.

Classic examples of evolutionary and ecological studies of island faunas and how they differ from continental faunas from which they are presumably derived include, of course, Darwin’s finches, but there are other spectacular examples:

The Hawaiian Islands have many examples of adaptive radiations from a single or few ancestral species. One example shown below are the honeycreepers that are thought to have descended from a single Asian ancestral species. These birds have diversified to fill several different feeding niches.

The cichlid fishes of the central African Great Lakes provide another example. In lakes such as Lake Victoria, Lake Malawi, and Lake Tanganyika, literally hundreds of species of cichilds are thought to have evolved within each lake basin from one or a few species of ancestral species that colonized these "islands" from inflowing rivers. Again, within each lake the ancestral species has diversified into a huge array of species characterized by very different (and some very unusual) trophic (feeding) niches (see figure below). It’s the fanatastic diversity that one often finds on islands that has fascinated biologists for centuries.

(ii) Foundations of island biogeography theory

(1) the species – area relationship.

The species-area relationship is perhaps the only true "law" in ecology and biogeography. Across countless taxa and geographic areas, there is a strong tendency for species (or any other taxon) number to increase as island area increases.

This relationship is so pervasive that it was the single largest factor that stimulated the development of a quantitative theory of island biogeography (the effect was most obvious and easily observed on islands). The increase in species diversity tended to follow a power function and level off as island area increased. The general form of the relationship can be expressed as:

Where S = species diversity, A = island area, and c and z are fitted constants. This expression was adapted in the 1920 from the well known allometric equation used by physiologists to describe the changes in metabolic properties with changes in body size.

The relationship is linearized by taking the logarithms of species diversity and area. Hence:

where c now represents the intercept and z represents the slope of the relationship between S and A.

An example is shown below: the upper panel represent the distribution of islands in the south Pacific where diversity of genera in conifers and flowering plants have been recorded. The islands range from very large (New Guinea, New Zealand) to very small ones (Norfolk Island ("N"), Easter Island ("E"). The first number shown with each island is the total number of genera on the island, the second number represents the number of genera that are endemic to that island.

Below is plotted the species-area relationship for the different islands. Note the log-log scales. Clearly, generic diversity increases as island area increases. The line through the points represents the least-squares line of best fit or central tendency line describing the relationships between S and A. Note that there is considerable scatter about that line – variation in island area explains much of the variation in generic diversity, but not all of it.

Two other examples follow below from other kinds of "islands":

Birds and mammals from isolated mountain ranges in the North American Great Basin

Freshwater fish from North American temperate lakes and African lakes. Note the greater slope in the African fish relationship which are much older (> 1 million years) than the North American lakes (postglacially formed < 10,000 years ago.).

What drives the species area relationship? In part, the relationship is probably driven by sampling effects along. In any given area, only a few species tend to predominate and most species are only moderately abundant or are rare and these species will tend not to be observed in small sample areas. As the area sampled increases, more individual species will be present and one will tend to have a greater chance of encountering rare species to increase the species count. In addition, larger areas tend to have a greater diversity of habitats that can accommodate species specialized to those habitats with an attendant increase in total species number. The figure below shows the increase in species diversity as habitat diversity increased at 55 sites in SW Australia. Habitat diversity was measured as variability in foliage height along the x-axis.

(2) The species – isolation relationship

In addition to island size, the distance of an island from a continental "source" fauna has a strong influence on species diversity on islands. This "isolation" effect can be seen in the figures below. The left panels in each figure show the decline in species richness in isolated patches of montane forests in the southwest USA and in parts of South America.

Also, examine the plot of generic diversity and island area in the previous figure for plants. Note that there is "scatter" about the central trend line. Some islands (e.g. Easter Island and New Zealand) fall below the line – they have fewer species than their areas alone would predict. Others (e.g. West Carolines, "WC") have more species than their small area would predict. New Zealand and Easter Island are relatively distant from rich faunal sources (New Guinea, Australia) whereas the West Carolines are part of an archipelago and are adjacent to New Guinea as well as the rich fauna of Southeast Asia. Note, however, that the WC, even though they have rich diversity in terms of the numbers of genera, they have no endemics, yet New Zealand has 39 endemic genera (but relatively low total diversity). What might explain the low number of endemic genera on the WC????

(3) Species turnover and extinctions

The third component of the theory of island biogeography involved the observations of rapid dispersal and colonization of new habitats which hinted that island ecology was not static, but rather very dynamic. The most impressive example of this phenomenon is the observations made on the island of Krakatau in the Indonesian Archipelago. This island, located between Sumatra and Java, essentially "blew-up" owing to a massive volcanic eruption in 1883. It was split into three smaller islands each devoid of life. As such, it represented a "natural experiment" in the magnitudes and patterns of dispersal and colonization of new habitat from faunal sources. Faunal and floral surveys of the islands, in particular for the largest island called Rakata, indicated rapid dispersal to the island presumably from Java and Sumatra.

For instance, from a total lack of a bird fauna in 1883 (just after the eruption), 34 species were recorded on Rakata by 1935. The figures below show the pattern of increase of species diversity for various taxa between 1883 and 1983. Note the relatively rapid increase over the first few decades, followed by an apparent "leveling off" of species diversity. Also note that these patterns vary between taxa.