Studies of insects have played a major role in the general understanding of the biota of islands, touching on all areas of biogeography, ecology, evolution, and conservation. The notable writings of Darwin and Wallace were influenced heavily by the biological diversity that each witnessed on islands and the processes inferred to underlie that diversity. Rather more recently, studies exploiting the discrete nature of islands have given rise to pervasive organizing theories of community ecology, particularly involving island biogeography. With the advent of accessible molecular genetic tools, research on islands has allowed unique insights into the processes that generate biotic diversity, especially the mechanisms of speciation. Unfortunately, islands are also prime targets for biological invasions, mediated largely by anthropogenic disturbance. The severity of such impacts on island biotas may result from their evolution in isolation, but is certainly compounded by their characteristically small population sizes. Yet, for many islands, extinction among arthropods is largely unknown, although this may be attributable more to lack of knowledge than any innate security that arthropods might possess. Indeed, it is likely that many island arthropods will go extinct before they have been collected and described.
THEORY OF ISLAND BIOGEOGRAPHY
Larger islands contain more species. MacArthur and Wilson formalized this idea in the 1960s with the development of the Equilibrium Theory of Island Biogeography (ETIB). This theory relates species and area by the formula, S = cAz, where S is species number, A is area, c is a constant measuring overall species richness, and z measures the extent to which increases in area have diminishing returns in terms of species number. Values of z tend to vary between 0.18 and 0.35; that is, doubling the species number requires increasing the area by a factor of between 7 and 100. The premise of the theory is that the rate of immigration decreases with increasing distance from the source, whereas the rate of extinction decreases with increasing island size. The balance of these processes results in an equilibrium number of species on any one island. As the number of resident species on an island increases, the chance of an unrepresented species arriving on that island decreases and the likelihood of extinction of any one resident species increases. The predictions of the model are that: (1) the number of species on an island should change little once the equilibrium is reached; (2) there should be continual turnover of species, with some becoming extinct and others immigrating; (3) small islands should support fewer species than large islands; and (4) species richness should decline with remoteness of the island because islands farther from the source will have lower rates of immigration.
Rigorous tests of the ETIB have been surprisingly few, and they have supported some aspects of the theory but not others. For example, Simberloff used insecticides to defaunate mangrove islands and found that species of insects and spiders accumulated to an equilibrium number. However, contrary to expectation, turnover of species was not randomly distributed among species—particular types of species were likely to colonize or go extinct. In some cases, species numbers have been found to be affected unpredictably by both area and isolation, whereas others have shown that an equilibrium does not exist, or that parameters other than area per se may dictate species richness. Such factors include habitat diversity, climatic conditions, island age, and even the status of knowledge concerning the presence of resident species. However, the theory has proven to be remarkably useful and, although it was developed for islands, it has had relevance for the study of many kinds of ecological communities.
ADAPTIVE RADIATION
The ETIB assumes that islands are within the geographic distance that a species is likely to disperse, thus maintaining genetic populations between source and island populations. On islands that are beyond the range within which populations can maintain genetic contact with source populations, one might predict (based on the theory) that few species should be present. But this tends not to happen. Isolated islands that are formed initially without life are often found to have large numbers of closely related species. This phenomenon, where single colonists, isolated genetically from their source population, give rise to a series of species that have diversified ecologically, is termed adaptive radiation, and usually occurs beyond the so-called “radiation zone,” or normal range of dispersal of a given organism. Species that form through adaptive radiation are typically neo-endemics, formed i n situ and found nowhere else. Among arthropods, the Hawaiian Islands hold the record in having the largest number of neo-endemics, an extraordinary 98% of the fauna. Important factors involved in adaptive radiation can include founder effects, genetic drift, behavioral isolation, ecological isolation, host-associated isolation, and the presence or absence of competitors, predators, and symbionts, the relative importance of these varying across biological systems. For insects, particularly noted examples of adaptive radiation include: Drosophila flies, which are well known for their diversity of mating behaviors; lineages of crickets that have diversified explosively in song repertoire; sap-feeding planthoppers and psyllids that have proliferated by tracking and switching between plant hosts; and beetles that have formed new species on different substrates. Diversification may follow a predictable pattern, at least in some groups; for example, among Tetragnatha spiders similar ecological sets of species have evolved over and over again on each of the different Hawaiian Islands.
Compared to their hypothetical colonizing ancestors, terrestrial arthropod species on remote oceanic islands are often characterized by a reduction in dispersal ability, including loss or reduction of wings or a reduced tendency to fly or balloon. This effect is most pronounced in species that would have colonized islands passively by wind, including a number of spiders and likely planthoppers and many other insects. Other arthropod species on remote islands, such as beetles in the genus Rhyncogonus, likely were never good dispers-ers and colonized remote islands by hitching rides on vectors such as birds, or on floating mats of vegetation. Regardless of their mode of transport, the species that colonize remote islands are a small sample of the species inhabiting their various continental sources. Island biological communities are therefore unrepresentative of the biotic diversity on continents, a phenomenon accentuated by the frequent proliferation of successful colonists.
DIFFERENT KINDS OF ISLANDS: NEO-ENDEMICS AND PALEO-ENDEMICS
Neo-endemics typically form on isolated islands that have been created de novo, and have abundant empty ecological space into which those few colonists can diversify. Besides Hawaii, other volcanic archipelagoes, including the Marquesas, Societies, and Galapagos in the Pacific and the Canaries in the Atlantic, have provided ideal conditions for the formation of neo-endemics. However, species can also form on fragment islands, formed as a mass of land has broken away from a larger continental region. Examples of such islands include some of the Caribbean islands, and the islands of New Zealand and Madagascar. These islands are formed upon losing connection with a continental source; as they become more isolated, gene flow between island and continental populations is reduced and may become insufficient to overcome genetic divergence. Unlike volcanic islands that form in isolation, starting without any species and accumulating species through time, fragment islands are usually ecologically saturated at the time of separation and tend to lose species through ecological time. Over evolutionary time, the species on these islands may change through a process termed relictualiza-tion, with the formation of paleo-endemics, usually without adaptive radiation.
HABITAT ISLANDS
Many of the ideas originally developed for islands in the sea have been extended also to so-called “habitat” islands of a particular habitat type within a matrix of unlike terrain. Most such islands are fragments of habitats that were historically connected, such as remnant trees and forest patches. For habitat islands, as for islands in the sea, ecological and evolutionary processes are governed largely by isolation, time, and the nature of the matrix relative to the dispersal abilities of the organisms in question. Habitat islands, because of their discrete nature and ease of manipulation compared to islands in the sea, have been exploited in the development of many ecological principles including those related to meta-population dynamics and physical design of nature reserves.
CONSERVATION
The biota of islands is often unique—for example, the islands of the Pacific have been designated a biodiversity hotspot. Assessing this diversity, particularly for arthropods, is problematic. The major impediment is a lack of taxonomic understanding of arthropods on many islands, particularly those that are more remote. New species are being collected at a remarkable rate in areas such as French Polynesia, Madagascar, and even the relatively well-studied Canary Islands, New Zealand, Hawaii, and the Galapagos, yet the training of arthropod systematists has lagged behind.
Anthropogenic disturbance has also had its impact, not only in present times but historically, such as witnessed through the colonization of the Pacific by Polynesians several thousands of years ago. In the islands of the Pacific and Indian oceans, new paleoarcheo-logical studies are now documenting what arthropods were present before the arrival of humans. A number of characteristics of arthropod populations on islands, including high local endemism, limited dispersal abilities, and small population sizes, make them particularly vulnerable to both demographic accidents and environmental change. In addition, islands have been impacted heavily by invasive species, many of which are also arthropods. These invasives tend to disperse well and therefore can often recolonize disturbed habitats before indigenous species. The impact has been both direct, such as through the extirpation of species by invasive predatory ants, and indirect, such as through diseases of vertebrates vectored by mosquitoes.
Although islands have long proven themselves as extraordinary laboratories for studying processes associated with the generation of diversity, they are now contributing to our understanding of processes leading to the loss of diversity. For example, studies of invasive species on islands have shown the importance of environmental factors as well as species-specific attributes that facilitate biological invasions and its negative effects. New tools are urgently needed: rapid biodiversity assessment techniques that bypass traditional taxonomic identification will be important in recognizing areas of high conservation priority, as will genetic and ecological approaches that can distinguish native species from those introduced in more recent history.
