Is a new paradigm emerging for oceanic island biogeography?


  • Lawrence R. Heaney

    Corresponding author
      Lawrence R. Heaney, Field Museum of Natural History, 1400 S Lake Shore Drive, Chicago, IL 60605, USA.
    Search for more papers by this author

Lawrence R. Heaney, Field Museum of Natural History, 1400 S Lake Shore Drive, Chicago, IL 60605, USA.


Following several decades during which two dissimilar and incompatible models (equilibrium and vicariance) dominated island biogeography, recent publications have documented patterns that point the way towards a new paradigm that includes elements of both models, as well as some novel aspects. Many of these seminal contributions have been made possible by the recent development of robust, temporally calibrated phylogenies used in concert with increasingly precise and reliable geological reconstructions of oceanic regions. Although a new general model of oceanic island biogeography has not yet been proposed, in this brief overview I present six hypotheses that summarize aspects of the emerging paradigm. These hypotheses deal with: the frequency of dispersal over oceanic water barriers by terrestrial organisms; the existence of substantial variation in the amount of dispersal (and gene flow) within a given set of related species within a given archipelago; the frequency, extent and impact on species richness of diversification within archipelagos; the frequent correlation of island age and the age of the species that live on the island; the long-term persistence of species on oceanic islands; and the occasional recolonization of continents by species from clades that diversified on islands. Identifying, testing, and seeking means of synthesizing these and other emerging hypotheses may allow a new conceptual paradigm to emerge.

Conceptual models in biogeography guide our thought processes, leading us to interpret data in certain ways, and, at a more fundamental level, to ask (and therefore answer) only certain questions, while leaving other questions unasked. In spite of this process of ‘normalization’, data sometimes accumulate that do not readily fit into existing models. Resemblances to results from other studies are noted, new patterns are seen, and a new paradigm begins to emerge.

Island biogeography is almost certainly currently undergoing a paradigm shift, due to just such an accumulation of new data. In this case, it is not one paradigm that is being set aside, but two, simultaneously, that have existed side-by-side for about 40 years. One of these, the equilibrium model of island biogeography (MacArthur & Wilson, 1963, 1967) emphasized island biotas as being typified by rapid turnover through on-going colonization and extinction, and implied that rates of gene flow would often be high due to frequent dispersal. The other, vicariance biogeography (e.g. Rosen, 1978; Nelson & Platnick, 1981) implicitly requires great stability of island biotas and emphasizes the process of diversification, with persistence of species over millions of years with little or no colonization/dispersal and static biotic composition. It has been clear for many years that both models could not both be correct and complete; but for years, many studies utilized only one model or the other, despite calls for integration and synthesis (e.g. Brown, 1986, 2004; Heaney, 1986, 2000; Whittaker, 1998, 2004; Lomolino, 2000; Wiens & Donoghue, 2004).

I believe that recent publications, including many in the Journal of Biogeography, show that a new paradigm is beginning to emerge. The changes in perspective are often incremental, and most of the publications focus on a single group of organisms in a single region, but they present conclusions that are controversial because they do not fit within the perspectives of either of the formerly prevailing paradigms or an associated set of hypotheses and assumptions. One such case is presented by Dávalos (2007), whose data ‘suggest a complex history starkly at odds with previous hypotheses about Caribbean biogeography in general, and short-faced bats in particular’. Rather than representing a group that reached the islands recently by dispersal from one of the surrounding continents during Pleistocene periods of low sea level, or alternatively during the Oligocene (c. 30 Ma) by land-bridge, Dávalos shows that these bats mostly likely reached the islands by over-water dispersal during the Miocene, 10–20 Ma. Rather than being a group widespread within but confined to the islands (traditionally, this pattern would be ascribed to continental competitors; e.g. MacArthur, 1972; Mayr & Diamond, 2001), one lineage most likely re-invaded the continent several million years ago and then expanded its range greatly across a large portion of the continental Neotropics in the course of diversifying into four genera. As Dávalos points out, this hypothesis could be tested readily: ‘a single ancient short-faced bat found on the continent would overturn’ her hypothesis, but as it stands, ‘the fossil record is compatible with the Caribbean radiation hypothesis’. A significant change in the hypothesized phylogeny could overturn her interpretation as well.

In this case, as in many others, several factors allow the author to reach her unconventional conclusions. The development of cladistic phylogenetic methods that produce robust estimates of phylogeny; the availability of inexpensively generated DNA sequence data that produce some of those robust phylogenies; the ability to make rough but reasonable estimates of the timing of phylogenetic events using DNA sequence data (e.g. Arbogast et al., 2002); the availability of some well-dated fossil material; and the existence of increasingly detailed and reliable geological reconstructions for oceanic regions (e.g. Hall, 1998) are all crucial empirical and methodological components. It is the combination of these factors that allows many previously plausible hypotheses to be refuted, and for novel hypotheses to emerge.

In spite of the novelty of the conclusions reached by Dávalos (2007) for these organisms (bats) in this place (the Caribbean basin), similar conclusions are being reached elsewhere. While an explicitly stated and robustly tested general model has yet to be developed, I believe that the shape of a new paradigm for oceanic island biogeography is emerging from these studies. As a means of advancing discussion and research, I present the following six assertions based on recent studies, each of which is an essential component of the general model that I believe is emerging.

1. Dispersal over oceanic water barriers by terrestrial organisms is common over ‘geological time’

Many recent studies have presented strong empirical evidence of long-distance dispersal from continents to other continents, from continents to oceanic islands, and among oceanic islands (e.g. Gillespie & Roderick, 2002; Filardi & Moyle, 2005; Glor et al., 2005; de Queiroz, 2005; Cowie & Holland, 2006; Jansa et al., 2006; Rocha et al., 2006). This is not to suggest that geological vicariance is not a genuine and relevant process also (it clearly is quite real), but rather that both dispersal and vicariance processes are present and contribute to the patterns that we see on islands in oceanic regions of the world.

2. Within any given set of neighbouring islands, the amount of dispersal/gene flow between island populations varies markedly: it is common in some species and absent in others, even when the organisms seem superficially similar

Within a given closely-related set of organisms in a single region, different species have very different patterns of dispersal and gene flow, the consequent function of dispersal. For example, bats within the Caribbean (Carstens et al., 2004; Dávalos, 2007), Indonesia (Hisheh et al., 1998, 2004) and the Philippines (Heaney et al., 2005; Roberts, 2006a,b) vary greatly in the amount of gene flow, such that some genera show little geographic variation within an archipelago (because they have high gene flow), while others are broken into substantially divergent monophyletic units on each island (or set of nearby islands), which demonstrates the absence of gene flow. This is true even for bats of similar size and appearance; it seems that the ecology of the bats, especially their habitat selection (e.g. closed forest vs. disturbed or open habitat), plays a substantial role in determining the frequency of long-distance dispersal. Thus, it appears that the degree to which the geological history of a region influences the patterns of diversification is substantially influenced by the ecology of the species being studied.

3. Diversification within a lineage on a set of oceanic islands is frequent, often leading to high species richness and endemism in archipelagoes

Although neither the equilibrium model nor vicariance biogeography emphasizes adaptive radiation, this was classically the source of much attention (Darwin, 1859; Wallace, 1880; Lack, 1947), and recent studies are again focusing on this process (Schluter, 2000). Monophyletic groups within oceanic archipelagoes often show high levels of allopatric species richness, and much of the diversity within an archipelago may be generated in this way, e.g. anoles in the Caribbean (Losos et al., 1998; Glor et al., 2005), damselflies in Hawaii (Jordan et al., 2003), spiders in Hawaii and the Austral Islands (Gillespie, 2004; Garb & Gillespie, 2006), birds in the southwest Pacific (Filardi & Moyle, 2005; Lecroy & Barker, 2006), and murid rodents in the Philippines (Heaney & Rickart, 1990; Steppan et al., 2003). Further, in some cases, much sympatric species richness may lie within an endemic clade (e.g. Grant, 1986; Wagner & Funk, 1995; Burns et al., 2002; Jansa et al., 2006).

It should be noted that although the equilibrium model sensu stricto does not include diversification, MacArthur and Wilson's broader model of island biogeography was explicit in doing so (MacArthur & Wilson, 1963, 1967, pp. 173–175). Ironically, textbook treatments of their seminal publications usually have not included this aspect, and therefore did not point the way towards the reconciliation of ‘ecological’ and ‘evolutionary’ biogeography (Brown, 1986; Whittaker, 1998; Heaney, 2000) that I believe is now under way.

4. Geologically younger islands often have younger species than nearby, geologically older islands

Because volcanic islands can be dated readily using potassium-argon dating, it is possible to determine the relationship between the age of an oceanic island and the age of the organisms that live on them. Studies in Hawaii (Wagner & Funk, 1995; Baldwin & Sanderson, 1998) have noted the ‘progression rule’ that younger species (i.e. phylogenetically derived species that diverged from their closest relative at some proportionately recent time) usually occur on younger islands, and the age of the divergence events of the species is correlated with the appearance of the new islands above the sea, e.g. weevils in the Galapagos islands (Sequeira et al., 2000), endemic biotic clades in the Canary Islands (Juan et al., 2000), forest mice in the Philippines (Steppan et al., 2003), and spiders in Hawaii and the Austral Islands (Gillespie, 2004; Garb & Gillespie, 2006). This suggests that both long-distance dispersal and divergence/diversification may often be highly regular phenomena, and implies that much of the pattern of phylogeny and geography may be predicted in any place where the geological history of an oceanic archipelago is known.

5. Species on oceanic islands often persist for very long periods; persistence typifies island biotas as much as does extinction

Many of the examples cited above document a phenomenon that has profound implications for how we view the dynamics of species richness in oceanic islands: although some species and clades show evidence of rapid turn-over of the type envisioned by MacArthur & Wilson (1963, 1967), a great many species show clear evidence of having occupied a given island or set of islands for long periods, often millions of years. Indeed, what has been called the ‘late Pleistocene model of speciation’ that dominated studies of birds and mammals for many decades recently has been largely refuted in favour of a paradigm in which divergence of living taxa from their closest living relatives occurred in the late Miocene to early Pleistocene, with a probable peak in the Pliocene (e.g. Klicka & Zink, 1997; Mercer & Roth, 2003; Steppan et al., 2003; Glor et al., 2005; Lovette, 2005; Jansa et al., 2006). This in turn leads to the conclusion that many island endemic species and clades have been in existence on the islands they now occupy for millions of years. Although we have no means at hand to measure the rates of extinction for most of these taxa, the extant species have unambiguously demonstrated their ability to withstand massive climatic and sea-level changes (during the Pleistocene), typhoons (among some tropical species), and volcanic eruptions (since many oceanic islands are volcanic in origin). Persistence, not extinction, typifies these insular organisms under natural circumstances. The recent surge in extinctions of island species most often is associated with some form of human disturbance (e.g. Sax et al., 2005; Borges et al., 2006; Steadman, 2006). The long-term persistence of so many oceanic species calls into question any assumption that natural (undisturbed) island communities are, in any sense, more intrinsically vulnerable to extinction and high turnover than communities on continents.

6. It is unusual, but not rare, for island lineages to colonize continents

Until recently, it was widely assumed that once a lineage became established on an island or set of islands, it then virtually never recolonized a continental area (e.g. Mayr & Diamond, 2001). The study by Dávalos (2007) is only one of several recently documented cases: the others involving small birds in the southwestern Pacific Ocean (Filardi & Moyle, 2005; LeCroy & Barker, 2006), anoles in the Caribbean Basin (Glor et al., 2005), and scincid lizards (Rocha et al., 2006) and chamaeleons in Madagascar and Africa (Raxworthy et al., 2002). Given how few studies have been done that use the procedures that could result in such a finding, these cases suggest that recolonization of continents from islands may be an unusual but widespread phenomenon, and establish the need for more broad-scale studies of phylogenetic biogeography of insular clades.


Clearly, a new comprehensive model of biogeography is needed. Until such a model emerges, we can make essential progress by continuing to conduct thorough and inquisitive investigations of individual taxa or sets of taxa, and identify the conceptual implications of the results. We can also propose, test and debate more restricted, specific hypotheses, such as the six listed above (and others that have been or will be proposed), as a means of synthesizing the patterns as they emerge. It is an exciting time to conduct research in island biogeography; the scope for new, insightful conceptual development has rarely, if ever, been greater.


Lawrence Heaney is Curator and Head of the Division of Mammals at the Field Museum of Natural History. His research is focused on the evolutionary origin, ecological maintenance and conservation of mammalian diversity on islands, especially in Southeast Asia.

Editor: Dov Sax