A theory attempts to identify the factors that determine a class
of phenomena and to state the permissible relationships among
the factors . . . substituting one theory for many facts. A good
theory points to possible factors and relationships in the real
world that would otherwise remain hidden and thus stimulates
new forms of empirical research. . . . If it can also account for,
say, 85% of the variation in some phenomenon of interest, it will
have served its purpose well.
—MacArthur and Wilson (1967, p. 5)
MACARTHUR AND WILSON’S (1963, 1967) dynamic equilibrium theory of island biogeography has a clear claim to be the most influential body of theory within ecological biogeography. Central to its continuing influence, their model invokes fundamental dynamic processes operating on populations, in order to explain key emergent patterns of system species richness, turnover, and endemism. As they envisaged, their theory has found application (with varying success) to all types of insular system, from microcosms to oceanic islands, and from ponds to habitat islands of woodland in “seas” of human-transformed habitat (Whittaker and Fernández-Palacios 2007).
The aim embodied in the 1967 monograph was to promote a research agenda for island biogeography in which the particularities of historical narratives were set aside in the search for the general mechanisms, laws, and rules and their emergent outcomes, beginning at the population level. Within the better-known opening chapters, it can be considered a largely macroecological approach (sensu Brown 1995), whereas the later chapters develop the accompanying evolutionary theory concerning, for example, species radiation and the taxon cycle. There have been numerous attempts to link evolutionary and ecological dynamics building on the MacArthur-Wilson model (e.g., Wilson 1969, Diamond 1975, Heaney 1986, 2000, Peck 1990, Cowie 1995, Peck et al. 1999, Losos and Schluter 2000, Price 2004, Emerson and Kolm 2005a,b, Heaney et al. 2005), notwithstanding which, the model has been less successful and is arguably less complete when applied to oceanic island systems operating on evolutionary time scales than when applied to “ecological islands” (e.g., Haila 1990, Paulay 1994, Cowie 1995, Stuessy et al. 1998, Borges and Brown 1999, Heaney 2000, Whittaker and Fernández-Palacios 2007, Gillespie and Baldwin, this volume, Losos and Parent, this volume).
Recently, Heaney (2007) has called for the development of a comprehensive new model of oceanic island biogeography, reunifying ecological and evolutionary biogeography. Such a model should be based on the identification of general patterns, describe these patterns quantitatively, and capture the underlying mechanisms (Brown and Lomolino 2000). Here, we sketch out an extension to the MacArthur-Wilson dynamic model that combines their reasoning with a simplified model of the ontogeny of oceanic islands to derive a general dynamic theory for the biogeography of oceanic islands.
The MacArthur-Wilson model recognizes that, for a discrete and isolated biological system, the number of species at any point in time must be a function of the number previously occurring there plus those gained through immigration and/or speciation (specifically via cladogenesis1), minus those having gone locally extinct. Their theory proposes that these three fundamental processes should vary in a predictable fashion in response to time since system initiation, and in relation to two principal controlling geographical/environmental influences: isolation and area. Immigration rate (I, species immigrating to the island per unit time) should decline as a function of isolation (distance), and extinction rate (E, species being lost from the island per unit time) should decline as a function of increasing area (a general surrogate for island carrying capacity, K). Taking the case of a newly formed and barren island, I starts at its highest rate and declines as a hollow exponential curve as the proportion of species propagules arriving on an island that represent new species declines, while E gradually rises as the resource space is occupied. Expressed per unit time, I is shown as forming a concave falling curve, with E forming a convex rising curve (MacArthur and Wilson 1967, figure 20) and, in time, these rates intersect to provide a dynamic equilibrium, a condition at which I and E are in balance, with a continual turnover (T) of species occurring thereafter.
MacArthur and Wilson (1963, 1967) recognized that, on the more remote islands, the pace of immigration is sufficiently slow that increasing proportions of the biota on such islands are the result of in situ evolutionary change, with species gain via speciation (again, in this context they were mostly focused on net expansion of lineages), most pronounced on larger islands towards the outer limits of the distributional reach of a taxon: which they denoted the “radiation zone.” Hence, they argued that species gain through in situ speciation increased with island/archipelago remoteness and with island area.
The simplest classification of types of islands found within seas and oceans divides them into three classes: continental shelf islands (many of which have been joined to continents at Pleistocene sea-level minima, i.e. they are land-bridge islands), continental fragments (ancient continental islands), and oceanic islands (Wallace 1902, Whittaker and Fernández-Palacios 2007). Our focus herein is on the last of these groups, the true oceanic islands. They are formed in varied tectonic circumstances but are largely volcanic in origin, building from the oceanic crust to form land masses isolated from mainland source pools by open stretches of ocean. While those formed in arcs associated with subduction zones can be renewed over extended periods of tectonism, many remote oceanic islands (e.g., in hot-spot archipelagoes, fracture zones, etc.) are formed by volcanic activity of limited duration, and once formed experience subsidence and erosion, resulting in their eventual demise, or persistence in tropical waters only as low-lying atolls, sustained by coral growth. Thus, with some well-known exceptions, remote islands forming volcanically over oceanic crust are typically short-lived. The significance of the island life cycle of these oceanic islands has been recognized by a number of authors (e.g., Paulay 1994, Stuessy et al. 1998, Stuessy 2007), most presciently by Peck (1990, p. 375), who wrote that “A relationship [of numbers of eyeless terrestrial cryptozoans] with island age should be expected, but it would not be a straight line. . . . Rather the relationship should be a curve which rises fast at first, reaches a peak or plateau, and then decreases as erosion destroys the island.”
In two recent papers, we have developed this line of reasoning more fully, suggesting that common elements in the ontogeny of oceanic islands should produce common emergent trends in diversity (Whittaker et al. 2007, 2008). Similar to the simple core model at the heart of the MacArthur-Wilson island theory, which focused principally on species richness (a metric indicative of “ecological dynamics”), we focus first on some simple metrics of “evolutionary dynamics,” in particular on numbers and proportions of species restricted to single islands (i.e., single-island endemics, SIEs). SIE data arguably provide only crude metrics but have been used in a number of recent studies as indicators of evolutionary dynamics (e.g., Peck et al. 1999, Emerson and Kolm 2005a, Triantis et al. 2008).
It should be noted that we use the term “evolutionary dynamics” in a broad sense, to encompass biotic and abiotic processes occurring over evolutionary time scales that determine emergent outcomes of species numbers, endemism, and phylogeography. While there is evidence indicating long-term persistence of many island biotas in the absence of catastrophic disturbance, erosion, and subsidence (i.e., where islands are fairly stable and persistent) (Ricklefs and Bermingham 2002), it is not possible to assume when examining the phylogeny of an island clade that all species that have formed within an island or archipelago have persisted to the present day. Hence, estimates of evolutionary rates available in the literature should be regarded as diversification rates, i.e., meaning rates of speciation minus extinction. This recognizes that within radiating archipelagic lineages some species may have formed and long ago gone extinct, something that must, for example, have happened repeatedly during the 32 Ma history (Price and Clague 2002) of the Hawaiian Chain. So, whereas in the model developed herein we invoke trends in speciation and extinction rates through time, in practice, when it comes to evaluating the phylogenetic evidence, we have to accept that even when looking at apparent evidence of speciation rates on young islands like Hawaii (the Big Island), it is more proper to consider them diversification rates (i.e., S − E). Moreover, when examining numbers of single-island endemics, each of speciation, extinction, and interisland colonization has a role. The limitations of using SIEs as metrics are discussed below.
The general dynamic model (GDM) rests on three key premises as stated in table 4.1. The first two premises derive directly from MacArthur and Wilson (1967), and encapsulate both (1) their immigration/speciation-extinction dynamics, and (2) the argument that speciation and diversification in insular habitats are encouraged through the ecological opportunity signified by the concept of “empty niche space,” intertwined with the geographical opportunity provided by isolation (e.g., Lack 1947, Peck et al. 1999, Heaney 2000, Gillespie, 2004, Levin 2004). The final premise recognizes (3) that oceanic islands have a typical developmental life cycle from youth, to maturity, to old age and eventual loss (e.g., Nunn 1994, Price and Clague, 2002), and, crucially, that this life cycle plays itself out at a temporal scale resonant with and strongly influencing the evolutionary dynamics shaping the biota of oceanic island archipelagoes and basins (Peck 1990, Peck et al. 1999, Price and Clague 2002, Stuessy et al. 2005, Whittaker and Fernández-Palacios 2007).
TABLE 4.1
The Three Premises Underlying the General Dynamic Model of Oceanic Island Biogeography
Premise |
Support for the premise |
Biological processes: |
|
The MacArthur-Wilson model is an essentially correct summation of the key biological processes, i.e., island biotas are a function of rates of immigration, extinction and speciation, which lead toward a biotic equilibrium broadly as they envisaged. |
A large body of literature supports the importance of these processes, but evidence of attainment of equilibrium for distant oceanic archipelagoes remains equivocal as progress toward equilibrium is very slow (e.g., Cowie 1995, Whittaker and Fernández-Palacios 2007). |
Evolutionary response: |
|
Diversification within island lineages is typically greatest on larger islands that are remote (i.e. where interactions with closely-related fellow colonists is least) and where lineage persistence for non-trivial periods of time is permitted. |
1. Island systems near the effective dispersal limits of a higher taxon, where few lineages colonize, typically show the greatest diversification per colonist lineage (the “radiation zone” of MacArthur and Wilson, 1967). 2. Within oceanic island archipelagos, single island endemics (SIEs) have a far larger minimum area threshold and increase disproportionately with increasing area relative to native species of the taxon (Peck et al. 1999, Triantis et al. 2008). |
Geological progression: |
|
Oceanic islands are formed volcanically and typically have short life spans; in the simplest scenarios an island builds relatively speedily to maximum area and altitudinal range in its youth, next becomes increasingly dissected as it erodes, and then gradually subsides/erodes to disappear back into the sea or persist as a low-lying atoll. |
Geological dating of oceanic islands indicates much support for this, especially for the Hawaiian hot-spot chain of islands (Price and Clague 2002), although not all volcanic islands follow such a simple developmental sequence (reviewed in Whittaker and Fernández-Palacios 2007). |
|
|
Source: After Whittaker et al. 2008.
In this section we set out the general properties of the GDM through a series of graphical representations, inspired by MacArthur and Wilson’s (1963, 1967) familiar dynamic model. We begin with Heaney’s (2000) representation of the radiation zone concept, figure 4.1, showing how, for a given taxon, declining frequency of colonization translates into decreasing richness combined with increased absolute and relative importance of in situ cladogenesis.
In figure 4.2 we set out a general representation of the life history of an oceanic island, assuming the simplest of oceanic island histories, from initial appearance as a new volcanic island, building to a high cone-shaped form, of maximal area and height, and then becoming increasingly dissected and eroded. In time, such islands typically both subside (some rapidly and substantially, e.g., Moore and Clague 1992) and erode (aerially and through marine action), resulting in loss of both elevational range and area, until they disappear back into the sea, or persist in tropical seas as atolls—coralline islands of low elevation (Nunn 1994, Stuessy et al. 1998, Price and Clague 2002). Maximum topographic complexity will typically occur some time after the maximal elevation and area have been reached and passed.
In reality, most oceanic islands have rather more complicated histories than depicted, sometimes involving separate islands fusing to become one, and often involving catastrophic episodes of volcanism (tailing off with age) and slope failures (sometimes massive) (Price and Clague 2002, Whelan and Kelletat 2003, Le Friant et al. 2004); while Pleistocene climate change and sea-level fluctuations have also left detectable imprints on their biogeography (e.g., Peck 1990, Price and Elliott-Fisk 2004, Carine 2005). Furthermore, those oceanic islands that have formed within island arcs in association with plate margins can experience yet more complex histories, involving both vertical and lateral displacement (e.g., Buskirk 1985, Keast and Miller 1996) and can be more persistent than assumed herein (Paulay 1994). Hence, the simplified ontogenetic argument presented here is most applicable to hot-spot archipelagoes, and while it should, in principle, apply to other volcanic oceanic island archipelagoes, some modification will be necessary to accommodate alternative and more complex geological scenarios.
Figure 4.1. L. R. Heaney’s (2000) model of the development of species richness on large islands or archipelagos that experience varying rates of colonization due to varying degrees of isolation. According to Heaney, on islands near a species-rich source, high rates of gene flow will inhibit speciation. As the average rate of gene flow drops below approximately the level of one individual per generation, anagenesis will begin to take place and endemic species will develop. These endemic species (between lines 1 and 2) will have their sister-taxon in the source area, not on the island/archipelago. As colonization becomes still less frequent, and as time passes, phylogenesis will produce endemic clades diversified within the island/archipelago (species between lines 2 and 3). Over time, the oldest clades will become progressively more species rich (between lines 3 and 5).
Considering the simplified scenario in figure 4.2, the model implies that (1) the maximum carrying capacity K of an island, in terms of biomass and number of individuals across all species, will be reached roughly coincident with maximum area and elevational range (figure 4.3a), with (2) the maximum heterogeneity of environment, and thus maximum opportunity for within-island allopatry, occurring somewhat later, but still within the “middle age” of the island (figure 4.4).
Figure 4.2. Idealized relationships between the age (x-axis, time) and area (dotted line), elevational range (dashed line), and topographic complexity (solid line), of a hypothetical oceanic island. Island maximum altitude and area both peak before maximum topographic complexity, but all three are expected to show a humped pattern. As the period of growth is typically shorter than the period of decline, time may best be considered a logarithmic function. From Whittaker et al. (2008).
These arguments allow us to extend the MacArthur-Wilson theory to incorporate the implications of both an extended preequilibrium phase and an extended postequilibrium phase where K is declining and E > (I + S). Figures 4.3a and 4.3b combine these arguments to provide a graphical model of the dynamic processes involved in the developmental cycle of an island within an oceanic archipelago. The period from island emergence to maximal carrying capacity is typically far shorter than the period of decline (consider, e.g., Stuessy et al. 1998, Carracedo and Tilling 2003, Le Friant et al. 2004), such that the time axis should be represented as some form of logarithmic or power function.
With regard to evolutionary dynamics, the key propositions in relation to the generalized life cycle of an island are:
1. in youth, initially most species can be attributed directly to immigration, typically from older islands in the archipelago;
2. during immaturity, speciation rates (and rates of cladogenesis) peak relatively early on, when there are enough lineages present to “seed” the process (see Percy et al. 2008), but when there is also plenty of adaptive opportunity in the form of empty niche space;
Figure 4.3. Graphical representation of the key rates and properties of the general dynamic model (GDM) of oceanic island biogeography. Island building being typically much more rapid than decline, time should be considered as some form of log function, as also the case for figures 4.2 and 4.4. A. Showing the postulated relationships between the biological characteristics and the ontogeny of a single island, where, for key rates: I is immigration rate, S is speciation rate, and E is extinction rate (each rate referring to number of species per unit time); and for species number: K is the potential carrying capacity, and R is realized species richness. For islands showing sudden extensive loss of territory due to landslips (as suggested by the kinks in the K and R curves) the extinction rate curve would require modification. B. Modification of I and S curves in relation to distance between islands or mobility of the taxa concerned. The amplitude of the S curve will vary between archipelagoes and major taxa as a function of the size of the available species pool/ease of dispersal. This variation in accessibility is signified by the variation between I1, I2, and I3 curves, corresponding respectively to S1, S2, and S3 curves. Note that a suite of modified R curves should also be shown, to match the variations in the balance of rates of immigration, speciation, and extinction, but have been omitted to reduce clutter. From Whittaker et al. (2008).
3. in maturity, species richness peaks, while speciation continues to add new species, partly due to the increasingly dissected topography, which generates increased opportunities for within-island allopatry;
4. in old age speciation declines to a low relative and absolute rate in tandem with reduced K and increased E (and thus reduced richness) as islands decline in elevation, topographic relief, area, and habitat diversity in old age; and
5. finally, all is lost, the island founders.
It is worth noting that the form of the I, E, and S curves for a series of islands should be expected to vary in relation to not only the usually considered parameters of area and isolation of islands, but also the temporal resolution of the analysis. This is probably of greatest significance when considering the early phase of island emergence and biotic colonization. An illustration of this comes from empirical analyses of the recolonization of the Krakatau Islands following their sterilization by volcanism in 1883, and of the interrupted colonization of the emerging island of Anak Krakatau from the 1930s onward (Bush and Whittaker 1991, Thornton 1996, Whittaker and Fernández-Palacios 2007). These studies found departures from MacArthur and Wilson’s (1967) smoothly falling I and smoothly rising E rates for several taxa, as a result of factors such as (a) initially hostile environmental conditions preventing widespread colonization early on, (b) accelerated phases of colonization as successional thresholds were passed (the formation of the first woodlands, etc.), (c) episodes of extinction linked to (b), and (d) bursts of extinction and immigration linked to further disruptive volcanic activity. Such complexities are evident over time periods of years up to several decades. However, as we are here concerned with systems running over several million years, we can think in terms of a temporal resolution of analysis of hundreds to thousands of years, in which Krakatau-like successional dynamics will be largely undetectable. Hence, our model shows smoothly falling I and rising E rates essentially from time zero, ignoring the likelihood that when analyzed at a very fine temporal resolution we might expect to see a more complex early development pattern.
Figure 4.4. Schematic representation of relative roles of different forcing factors through the life cycle of the island. Considering figures 4.2 and 4.3, we can derive the prediction that the greatest opportunities for adaptive radiation (solid line, first peak) will occur earlier than those for non-adaptive processes linked to within-island isolation (dashed line, second peak). Biotic interactions within and across trophic levels may be expected to become more important in the later stage of the island life cycle (dotted line, third peak), past the point of maximum carrying capacity and where extinction rate is climbing with island erosion/subsidence. Such biotic/competitive mechanisms may produce species involved in tight mutualisms, or fine subdivisions of resources sympatrically, but not at a rate sufficient to prevent the eventual decline in the proportion of SIEs. From Whittaker et al. (2008).
Although true oceanic islands arise in varied geological circumstances, they are frequently clustered together in space, forming distinct archipelagoes within which the timing of formation of each island varies significantly (e.g., Nunn 1994, Carracedo and Tilling 2003). Thus, as each island goes through its own life cycle, an archipelago develops in which a wide array of island ages/stages is available at any single time. Hence, a young island is supplied by colonists from nearby older islands, and in time supplies colonists to the next island(s) to form. Therefore, archipelagoes such as the Canaries or Hawaii can be conceived of as consisting of a series of terrestrial platforms each going through the sequences shown in figures 4.2–4.4, but each at a different point along the time axis.
Considering a single island forming within an existing archipelago, developing to maximum size, and elevational range, then becoming increasingly dissected through erosion, and finally entering a long phase of decline in area, elevation, and environmental complexity, we expect a general hump-shaped trend in potential carrying capacity (K) and similar trends in species richness (R), and in speciation rate (S) (figure 4.3). Extinction of species can occur at any stage, but will be driven by differing processes at different stages of an island’s life cycle. During the building/maturity phase, high-magnitude catastrophes (large volcanic eruptions, mega-landslides) will be more important—if highly unpredictable—while the more gradual erosion and subsidence processes associated with older islands will eventually force the background extinction rate to rise above the combined processes of addition (speciation and immigration), inexorably driving species number toward zero for islands that founder beneath the waves, completing the cycle.
We may also derive a general prediction (table 4.2) for the trend in the proportion of single-island endemic species (pSPIE) during the ontogeny of a particular focal island. Initially, as the island ecosystems are seeded (colonized through successional processes) from the nearby older islands in the archipelago, most species are not SIEs, although they may well include archipelago-level endemics, so the pSIE will be low. However, as the available propagule pool is relatively limited, and ecological space is initially unsaturated, speciation rate picks up, often generating significant radiations within single genera (e.g., Gillespie and Baldwin, this volume), thus increasing the proportions of SIEs and simultaneously generating an increased species-to-genus ratio. As the process continues, some part of this diversification process may be attributable to the arrival of “keystone species” such as Metrosideros in the Hawaiian system, providing stimulus to diversification in interacting animal lineages (Percy et al. 2008, and cf. Emerson and Kolm 2005). However, as the island ages and declines, it follows that a point is reached at which E > (I + S), and so species richness and the number of SIEs (nSIE) will each decline.
A further prediction follows, that the proportion of SIEs on our focal island should also decline, for the following reasons: (1) the area threshold for SIEs is on average larger than for non-SIE native species (Triantis et al. 2008), partly as the latter may persist even as fairly small populations if reinforced by occasional propagule flow from other islands; (2) the loss of habitat diversity (e.g., upland habitats, lava tubes [Borges and Hortal 2009]), and corresponding increase in habitat similarity with the coastal lowlands of other islands in the group, results in the collapse of radiations of neo-endemics (including many habitat specialists) on the focal island, while widespread coastal generalists would be anticipated to persist best; and (3) as the focal island supplies colonists to the next island to form, some of the SIE species of the focal island colonize the new island (in accordance with the progression rule [Funk and Wagner 1995]) and lose their status as SIEs. This last mechanism will apply most strongly in hot-spot archipelagoes involving a clear age progression; it may not be so evident in more complex island arc systems, and would not be anticipated at all in, e.g., poorly dispersing sightless troglodytes.
TABLE 4.2
Predictions Derivable from the General Dynamic Model
1. Island species number and the number of SIEs should be a humped function of island age and, when examining snapshot data across an archipelago, this will be combined with a positive linear relationship with area.
2. The amplitudes of the curves shown in figure 4.3a should vary in relation to the size of the island at maturity, with higher peak richness and SIE numbers on islands that attain greatest size (area and elevation) at maturity.
3. The relative amplitudes of the immigration and speciation rate curves should vary in relation to the effective isolation of islands, i.e., in relation either to distance between islands and their sources or to the mobility of the taxon, as shown in figure 4.3b.
4. Lineage radiation (leading to multiple SIEs on individual islands) should be most prevalent after the initial colonization phase, in the period leading up to island maturity, coinciding with maximal carrying capacity (K) and the development of maximal topographic complexity.
5. Montane representatives on old, declining islands should gradually be lost because of loss of habitat, meaning that surviving montane forms are increasingly likely to be relatively old (i.e., basal) forms in relation to other members of an archipelagic radiation.
6. The proportion of SIE should also be a humped function of island age, as islands that decline to small size and carrying capacity should lose SIEs in accordance with the second premise of the GDM (and see also: prediction 8).
7. SIE per genus should be higher on younger islands; intermediate-aged islands will have more lineages showing speciation than do young or old islands; SIE per genus should decline on older islands so that as islands lose SIE, there is a tendency towards monotypic genera, preserving maximal ecological spacing in the remaining endemics.
8. As islands age, some of their SIE species should colonise a younger island, so that they become multi-island species instead. Hence, the GDM also predicts that the progression rule should be a common/dominant phylogeographical pattern within an archipelago.
9. Using Stuessy et al.’s (1990, 2006) approach to classifying speciation modes, there should be a tendency on old, submerging islands for anagenesis to be an increasingly prominent speciation signal. Note: This assumes that where SIEs are the only member of their genus the explanation is in situ speciation. In practice we expect that on the oldest islands “anagenesis” will often be a misnomer, as there will be a trend towards survival of single relicts from former radiations.
10. Adaptive radiation will be the dominant process on islands where the maximum elevational range occurs, as it generates greatest richness of habitats (major ecosystem types), including novel ones few colonists have experienced. Nonadaptive radiation will become relatively more important on slightly older islands, past their peak elevation, due to increased topographical complexity promoting intra-island allopatry (figure 4.4). Similarly, composite islands (e.g. Tenerife, formed from three precursors), should have provided more opportunity than islands of simpler history for within-island allopatry, producing sister-species that lack clear adaptive separation (e.g., Gruner 2007).
Source: From Whittaker et al. 2008.
The GDM thus allows us to derive several predictions (table 4.2) about the emergent properties of the biota: (a) of a single oceanic island through time; and (b) of the islands of an oceanic archipelago at a single point in time. Given the extended time period (millions of years) over which data would ideally be required to fully explore the generality of the assumptions and predictions, we have to make use principally of predictions about temporal “snapshot” patterns in order to assess support for the GDM. This requires the selection of oceanic archipelagoes in which a meaningful portion of the life cycle shown ultimately by a single island is available for study in the form of separate islands of widely different age/stage. The key problem in doing this is that the islands within an archipelago do not all attain identical properties at maturity, and in particular they may vary significantly in maximum attained area and elevational range: properties of key importance (table 4.1, figure 4.3a). To deal with this analytically we need to include a term for island size, assuming that all islands within a group follow the same general trajectory, but that the amplitude of the curves will vary in relation to the maximum area attained.
The postulated humped trends of particular diversity attributes/metrics in relation to island age (table 4.2) constitute a particularly distinguishing and testable feature of the GDM. Whittaker et al. (2008) therefore began the empirical evaluation of the GDM by using data from five oceanic island archipelagoes (the Canaries, Galápagos, Marquesas, Azores, and the Hawaiian Islands) satisfying two criteria: (1) they provide a good span of island ages (maximum island ages were used in the analyses); and (2) fairly comprehensive survey work and compendia were available for particular taxa. Details of data sets, modeling approaches, and specific aspects of island histories, etc., are provided in Whittaker et al. (2008).
Tests of the GDM factoring in both island age and area take the form Diversity (D) = a + b(Time) + c(Time2) + d(logArea), where the use of a logarithmic function of area follows standard practice, empirically derived in numerous published analyses, and where the expectation is for positive exponents for Area and Time but a negative exponent for Time2 to reflect a humped relationship between diversity and island age. We term these fitted regression models ATT2 (i.e., Area+Time+Time2) models to distinguish them from the theoretical GDM. These models were compared with the semilogarithmic and power models for island area (the most commonly favored in the literature), plus a semilogarithmic island age model and a parabolic age model (i.e., D = b1 + b2 × Age + b3 × Age2) to explore the fits derivable from area or age alone. The diversity metrics used were species richness (SR), number of SIEs (nSIE), proportion of SIEs (pSIE), and a simple diversification index (DI), which is the ratio of nSIE to the number of genera containing SIEs (where nSIE = 0, DI was also set to 0).
The ATT2 models describing species richness were statistically significant for each of the fourteen data sets, with a mean R2 value of 0.85 ± 0.08 (SD) and in each case the relationship with island age was humped in form (table 4.3). Similar findings pertained for each of thirteen tests for each of the three SIE-based metrics, which were again significant in all cases. The island age component was humped except in four cases, namely, nSIE and pSIE for Azorean snails, and pSIE and DI for Galápagos beetles. The ATT2 model (with humped age relationship) provided the best model (based on adjusted R2 values) in between eight and ten cases for each metric (table 4.3). The four alternative models are each simpler compared to the ATT2 models, being two-parameter (T + T2) or one-parameter models. The two conventional area models each provide higher adjusted R2 values than the ATT2 model for between one and four cases (depending on the metric used) but, unlike this model, neither provides significant fits to all data sets, with nonsignificant fits most evident for the three Canarian taxa (i.e., standard species-area models are inadequate in this archipelago; see also Triantis et al. 2008). The time-only models generally performed poorly in comparison to the ATT2 models, with one exception, the Azorean snail data, for which, contrary to the expectations of the GDM, the relationship with time is not humped. This particular result can be accounted for within the GDM reasoning if it is accepted that the maximum geological age for some islands differs substantially from the effective age of the island in biological terms; although some might consider this special pleading (see details in Whittaker et al. [2008] and see analyses for other Azorean groups by Borges and Hortal [2009]). In summary, the analyses demonstrate that the ATT2 model provides a generally good fit with data from a range of plant and invertebrate taxa from five oceanic island archipelagoes, both for numbers of native species (SR) and for metrics more directly indicative of evolutionary dynamics (nSIE, pSIE, DI). It is worth emphasizing that in the majority of the cases studied the relationship between the diversity metrics used and island age, when included in a model with island area, was hump shaped, despite the fact that the modeling approach did not impose such a relationship (see table 4.3).
TABLE 4.3
Summary of Tests of the General Dynamic Model Using Diversity Metrics from Five Archipelagoes
The effectiveness of the ATT2 model in fitting data for particular archipelagoes and taxa is expected to depend on the effective isolation of the archipelago (figure 4.3b) and on the extent to which the archipelago provides a full range of island developmental stages. For example, for archipelagoes providing only young (and/or rejuvenated) islands, it would be consistent with the GDM for a simpler “log(area) + linear time” model to provide a better fit than the full ATT2 model (Borges and Hortal 2009). However, across the data sets evaluated, comparison with the alternative models provides confirmation that the ATT2 model, while not the simplest model (and not necessary in all cases), has greater generality than the traditional diversity-area models, or time-only models.
There are a variety of limitations to these tests: (1) the biological data are undoubtedly incomplete, (2) the islands have had more complex histories of formation than we assume, and (3) Pleistocene sea-level fluctuations have altered island areas and repeatedly joined and divided some islands. In addition, it is important to recognize that species may acquire and lose SIEs in several ways, e.g., (1) some current SIE species may have originated on another island (or land mass), from which they subsequently became extinct; (2) some species that evolved in situ as an SIE may have gone extinct and so are not around to be counted; (3) some former SIE species may have colonized another island(s) to become multi-island endemics (MIEs); (4) some MIEs occur on islands that were formerly connected at times of lowered sea level, indicating that their current disjunct distribution may derive from localized vicariance. Hence, we emphasize that the three metrics based on SIE data should be regarded as evolutionary dynamics metrics rather than either diversification or speciation indices. Nonetheless, we hold that in situ speciation will typically be the main driver of change in each of the three evolutionary metrics (nSIE, pSIE, DI) in the lengthy period leading up to the establishment of a dynamic evolutionary equilibrium (sensu Wilson 1969), whereas within-archipelago migration and within-island extinction become more important influences on numbers and proportions of SIEs during the even longer period of island “senescence.”
As volcanism continually requires the founding of new local
populations, genetic shifts and/or other episodic evolutionary
change would be expected to accelerate during the growth phase
of each successive Hawaiian volcano. These influences, however,
would decline as each volcano completes its active phase and
becomes dormant. . . . We suggest that the youngest island at any
one time has always been Hawaii’s major evolutionary crucible.
—Carson et al. (1990, p. 7057)
It is intrinsically difficult to obtain evidence of changes in rates of the vital processes (i.e., migration/immigration, speciation, and extinction) through time and in relation to other island attributes (spacing, overall archipelago isolation, Quaternary climate change, etc.). This is especially the case for the biotas of remote oceanic islands, many of which can be accounted for by mean colonization rates of one species every few thousand years (e.g., Wagner and Funk 1995, Peck et al. 1999). Similarly, attributing evolutionary outcomes to nonadaptive versus adaptive processes (prediction 10, table 4.2) is challenging (but see Barrett 1996, Cameron et al. 1996, Price and Wagner 2004), suggesting that testing some of the predictions in table 4.2 will be rather difficult to accomplish. Hence, while the indices of evolutionary dynamics evaluated in table 4.3 are crude, we have followed other recent authors (e.g., Peck et al. 1999, Emerson and Kolm 2005a) in adopting the rationale that SIE data are a good starting point and are likely to be indicative of trends and patterns in other metrics of evolutionary dynamics. In support of this, tallies of data for the overall number of Canarian endemic plants across the seven main islands of the archipelago (reproduced in Whittaker and Fernández-Palacios 2007) show that, at least in this case, the pattern for the number of Canarian endemics is strongly correlated with the nSIE and again shows a humped relationship with island age.
Several of the predictions derived from the GDM (specifically 4, 5, 7–10: table 4.2) concern the mode (figure 4.4) and pattern of lineage development. As an aid to visualizing the latter, figure 4.5 shows how a typical lineage might look at different points in time on an island progressing from youth to old age. Although we are not yet able to evaluate these model predictions systematically, we can begin to explore these aspects of the model with reference to existing literature from island systems.
Figure 4.5. A hypothetical island lineage conforming to the general dynamic model, examined at four points in time (T1 to T4). T1, a single colonization event to our focal island early in its life cycle leads to rapid onset of radiation, exploiting the relatively uncontested niche space. T2, during the period leading up to island maturity, a full array of habitats is available, and opportunities for within-island allopatry also gradually increase, with both circumstances encouraging speciation and diversification (many, short branches in the tree). T3, the frequency of formation of new species is expected to slow and to increasingly be balanced by losses as island erosion and subsidence reduce the available habitat space. With the passage of time, secondary colonization events from an older island following the progression rule, or sometimes backwards colonization events are possible. Thus, the clade is becoming more diverse and paraphyletic (ancestral) on our focal island compared to the next youngest island to form in the chain (or to T1 of the focal island). T4, speciation rate declines in tandem with reduced K, and extinction increasingly weeds out the tree, nevertheless, while the number of branches/species may be reduced the genetic diversity may remain high (compared to T1 or T2) due to the possession of older endemic lineages (longer branch lengths in a pruned tree).
Silvertown (2004) notes that large endemic taxa within the Canarian endemic flora are typically monophyletic (e.g., 63 species of Crassulaceae, and 37 species of Echium), i.e., they typically derive from single-colonization events. Silvertown suggests that this may be indicative of the operation of niche preemption by early-colonizing lineages that may have inhibited the success of later-arriving mainland relatives and also have spread out across the archipelago as new islands formed, frequently radiating into new habitats. These interpretations are broadly consistent with the GDM, and particularly the notion of greatest lineage radiation occurring on relatively young islands (e.g., Cowie 1995, Carine et al. 2004, Silvertown et al. 2005).
Turning to Hawaii, Gillespie and Baldwin (this volume) identify three basic categories of Hawaiian taxa in respect to speciation rapidity: (a) groups that diversify based on sexual selection speciate rapidly and in cases attain highest diversity very quickly on the youngest island (e.g., Laupala crickets and some Drosophila); (b) groups that predominantly diversify ecologically (many animal, some plant lineages) may reach their highest diversity after a somewhat longer period of time, on a youthful but not perhaps on the youngest island; and (c) groups that appear to have diversified mostly in allopatry (or in parapatry) (e.g., Orsonwelles spiders, many plant groups) show a progressive increase in species numbers with island age, implying that this mode of speciation tends to be rather slower and that equilibrium may not have been reached within the approximately five million year span provided between Hawaii and Kauai. These findings were based on phylogenetic analyses of a range of taxa that were established on the Hawaiian Islands at the time that Kauai was the youngest island in the chain (if not earlier). They appear to be broadly supportive of a number of the predictions arising from the GDM (see table 4.2, predictions 4, 6, and 7) but at the same time highlight that the GDM is capable of further refinement.
Several other phylogenetic analyses also indicate that younger islands are particularly active arenas for genetic differentiation and speciation (although strictly the evidence is generally for diversification rates; see above) (e.g., Carson et al. 1990, Kaneshiro et al. 1995, Barrier et al. 2001, Levin 2004, Percy et al. 2008). On the Hawaiian Islands, Levin (2004) reports that the estimated “speciation rate” for plants is a negative function of island age, varying from 0.20 species per lineage per million years (Myr) on Kauai (5.7 million years old) to 2.1 species per lineage per million years on Hawaii (0.5 Myr). Studies from the flora of the Juan Fernández Islands also support the idea of high initial rates of radiation, with faster rates evident on the younger island (Levin 2004, and see Crawford et al. 1992).
We find additional support for the likelihood that relatively high speciation rates can account for “explosive early” patterns of lineage diversification in recent simulation modeling by Rabosky and Lovette (2008), in a paper providing a method for distinguishing the signal of speciation from extinction in molecular phylogenies. Further analyses of island radiations using this approach hold promise for the evaluation of the ideas presented herein. However, from the data currently available, it has to be allowed that apparently faster evolutionary rates on younger islands could, at least in cases, be the outcome of the effects of erosion and subsidence on older islands reducing the persistence of neoendemic lineages within the older islands (as in figure 4.5, and see Peck et al. 1999, Stuessy 2007). Such extinctions are always going to be hard to quantify from traditional forms of data as we are highly unlikely to find comprehensive fossil evidence for species lost as a result of island erosion and subsidence. There are, however, numerous cases where island phylogenies point to the past existence and extinction of ancestral species that once occurred on land areas that no longer exist, i.e., former uplands and lost islands (those now submerged) (e.g., Wagner and Funk 1995, Keast and Miller 1996, Price and Clague 2002, Butaud et al. 2005, Emerson and Oromí 2005, Pulvers and Colgan 2007), providing general exemplification of the point that island decline forces extinctions and in time a net reduction in diversity.
Phylogeographic analyses of island lineages provide further evidence of the processes of movement and evolution across archipelagoes. One commonly supported pattern involves taxa showing a pattern of movements from older to younger islands within an archipelago, with speciation occurring on newly colonized islands (see figure 4.5). This progression rule pattern (Funk and Wagner 1995) is particularly evident in archipelagoes showing a clear linear age sequence of islands, consistent with our general theory (table 4.2, prediction 8). Examples drawn from many that provide support for this rule include, from Hawaii, Drosophila, Hesperomannia, Hibiscadelphus, Kokia, Orsonwelles, Remya, Metrosideros, and Tetragnatha; from Macaronesia, Olea, Gallotia, Gonopteryx, Hegeter, Pimellia, and possibly Dysdera; from Galápagos, scarabs and weevils; and from the Austral Islands, Misumenops rapaensis (original references in Whittaker et al. 2008, and see Gillespie and Baldwin, this volume, Percy et al. 2008).
We acknowledge that various other phylogeographical patterns (or no resolvable pattern) have been detected from these and other oceanic archipelagoes. In some cases, e.g., Galápagos birds, evolutionary scenarios involve multiple phases of island hopping and of alternating periods of allopatry and sympatry within a single radiation (Lack 1947, Grant and Grant, 1996). Moreover, data for some lineages are most parsimoniously explained by a sequence of colonization in contradiction to the age sequence (e.g., Kvist et al. 2005, Sanmartín et al. 2008). So it should be understood that the progression rule is not without exceptions (see Funk and Wagner 1995, Gillespie and Roderick 2002). However, based on the GDM, it should be expected to be a dominant pattern, followed by many taxa in archipelagoes showing a strong island age sequence, and especially so in taxa which happen to colonize early in the developmental history of an archipelago, yet which also exhibit sufficient dispersal limitation to speciate within the islands of that archipelago.
We are under no illusions that the general dynamic model described herein provides a complete theory of oceanic island biogeography and evolution, but we do consider that it provides an analytically tractable framework that is largely consistent with the larger body of theoretical ideas we have discussed herein. Modification will be necessary for those classes of island that conform poorly to our ontogenetic model, including many island arc archipelagoes and islands of mixed continental/oceanic origins showing complex histories of horizontal and vertical movement, erosion, and rebuilding (e.g., Buskirk 1985, Keast and Miller 1996). For those oceanic islands that do conform to the simple ontogenetic model, perhaps one of the most important omissions from the framework is the role of Pleistocene climate change and accompanying variation in the configuration of islands (e.g., Nunn 1994, Carine 2005, Whittaker and Fernández-Palacios 2007, Ávila et al. 2008). Global environmental change in the Pleistocene altered not only the number, area, and elevational range of islands in these archipelagoes, but also their relationship with source pools. For instance, Carine (2005) argues that the evolutionary pattern in Macaronesian Convolvulus is suggestive of discrete waves of colonization, which he explains through the “colonization window” hypothesis. This postulates that colonization opportunities have varied through time as a function of both the geotectonic mechanisms discussed herein (island formation, island sterilization/disturbance) and periods of climate change. Thus, low sea-level stands during the Pleistocene saw the emergence of stepping-stone islands, aiding dispersal among the more persistent islands of Macaronesia, and between them and the mainland. Similar arguments have been invoked elsewhere, and the notions that dispersal distances and directionality of dispersal related to major current systems can change through time, provide additional components that require integration into a comprehensive general theory of oceanic island biogeography (Cook and Crisp 2005, Cowie and Holland 2006).
In this paper we have outlined a general dynamic theory for the biogeography of oceanic islands, which explicitly places MacArthur and Wilson’s (1963, 1967) dynamic equilibrium model into the geological and evolutionary context of oceanic archipelagoes. The GDM is a deliberately simplified representation of diversity dynamics on oceanic islands. Our aim was to capture the few major factors that drive diversity patterns on oceanic islands of different sizes and ages, not to produce a precise predictive model. The main intended advantage of the GDM is not the better fit of the ATT2 models (which are directly derived from the GDM), since other higher-order models can have this property too, but that it may offer an improved theoretical framework for describing and understanding the evolutionary biogeography of oceanic islands. We envisage that the GDM is capable of further theoretical and empirical development, for example (1) modification to incorporate alternative repeated geological scenarios, (2) tests of genetic/functional trait variation at subspecies level for multi-island native species/endemics, (3) extension to take account of principles of community assembly on oceanic islands (see Gillespie and Baldwin, this volume), (4) analysis of the fit of the model for non-native species, and (5) translation of the current graphical models into a more precise mathematical format. Thus, although a more complete, formal treatment awaits further development, we hope the GDM can offer the foundation for a newly expanded theory of island biogeography, unifying ecological and evolutionary biogeography.
We are grateful to Henning Adsersen, Paulo Borges, Mark Carine, Brent Emerson, Larry Heaney, Joaquín Hortal, Jonathan Losos, José María Fernández-Palacios, Aris Parmakelis, Carsten Rahbek, Robert Ricklefs, Spyros Sfenthourakis, Tod Stuessy, Kathy Willis, and attendees of the Harvard symposium for discussion and/or comments on this and/or our 2008 Journal of Biogeography paper (on which this chapter is based). RJW is grateful to the organizers for the invitation to participate, and for financial support to attend the meeting. KAT was supported in this work by a Marie Curie Intra-European Fellowship Program (project “SPAR,” No. 041095).
Ávila, S. P., P. Madeira, N. Mendes, A. Rebelo, A. Medeiros, C. Gomes, F. García-Talavera, C.M. da Silva, M. Cachão, C. Hillaire-Marcel, and A. M. de Frias Martins. 2008. Mass extinctions in the Azores during the last glaciation: Fact or myth? Journal of Biogeography 35:1123–29.
Barrett, S.C.H. 1996. The reproductive biology and genetics of island plants. Philosophical Transactions of the Royal Society of London, Series B 351:725–33.
Barrier, M., R. H. Robichaux, and M. D. Purugganan. 2001. Accelerated regulatory gene evolution in an adaptive radiation. Proceedings of the National Academy of Sciences U.S.A. 98:10208–3.
Borges, P.A.V., and V. K. Brown. 1999. Effect of island geological age on the arthropod species richness of Azorean pastures. Biological Journal of the Linnean Society 66:373–410.
Borges, P.A.V., and J. Hortal. 2000. Time, area and isolation: Factors driving the diversification of Azorean arthropods. Journal of Biogeography 36:178–91.
Brown, J. H. 1995. Macroecology. Chicago: University of Chicago Press.
Brown, J. H., and M. V. Lomolino. 2000. Concluding remarks: Historical perspective and the future of island biogeography theory. Global Ecology and Biogeography 9:87–92.
Buskirk, R. E. 1985. Zoogeographic patterns and tectonic history of Jamaica and the northern Caribbean. Journal of Biogeography 12:445–61.
Butaud, J-F., R. Rives, D. Verhaegen, and J-M. Bouvet. 2005. Phylogeography of Eastern Polynesian sandalwood (Santalum insulare), an endangered tree species from the Pacific: a study based on chloroplast microsatellites. Journal of Biogeography 32:1763–74.
Cameron, R.A.D., L. M. Cook, and J. D. Hallows. 1996. Land snails on Porto Santo: adaptive and non-adaptive radiation. Philosophical Transactions of the Royal Society of London, Series B 351:309–27.
Carine, M. A. 2005. Spatio-temporal relationships of the Macaronesian endemic flora: A relictual series or window of opportunity? Taxon 54:895–903.
Carine, M. A., S. J. Russell, A. Santos-Guerra, and J. Francisco-Ortega. 2004. Relationships of the Macaronesian and Mediterranean floras: Molecular evidence for multiple colonizations into Macaronesia and back-colonization of the continent in Convolvulus (Convolvulaceae). American Journal of Botany 91:1070–85.
Carracedo, J. C., and R. I. Tilling. 2003. Geología y Volcanología de Islas Oceánicas. Canarias–Hawai. Santa Cruz de Tenerife, Spain: CajaCanarias–Gobierno de Canarias.
Carson, H. L., J. P. Lockwood, and E. M. Craddock. 1990. Extinction and recolonization of local populations on a growing shield volcano. Proceedings of the National Academy of Sciences U.S.A. 87:7055–57.
Cook, L. G., and M. D. Crisp. 2005. Directional asymmetry of long-distance dispersal and colonization could mislead reconstructions of biogeography. Journal of Biogeography 32:741–54.
Cowie, R. H. 1995. Variation in species diversity and shell shape in Hawaiian land snails: In situ speciation and ecological relationships. Evolution 49:1191–202.
Cowie, R. H., and B. S. Holland. 2006. Dispersal is fundamental to biogeography and the evolution of biodiversity on oceanic islands. Journal of Biogeography 33:193–98.
Crawford, D. J., T. F. Stuessy, D. W. Haines, M. B. Cosner, M. O. Silva, and P. Lopez. 1992. Allozyme diversity within and divergence among four species of Robinsonia (Asteraceae: Senecioneae), a genus endemic to the Juan Fernandez Islands, Chile. American Journal of Botany 79:962–66.
Diamond, J. M. 1975. Assembly of species communities. In Ecology and Evolution of Communities, ed. M. L. Cody and J. M. Diamond, 342–444. Cambridge, MA: Harvard University Press.
Emerson, B. C., and N. Kolm. 2005a. Species diversity can drive speciation. Nature 434:1015–17.
———. 2005b. Emerson and Kolm reply. Nature DOI: 10.1038/nature04309.
Emerson, B. C., and P. Oromí. 2005. Diversification of the forest beetle genus Tarphius in the Canary Islands, and the evolutionary origins of island endemics. Evolution 59:586–98.
Funk, V. A., and W. L. Wagner. 1995. Biogeographic patterns in the Hawaiian Islands. In Hawaiian Biogeography: Evolution on a Hot Spot Archipelago, ed. W. L. Wagner and V. A. Funk, 379–419. Washington, DC: Smithsonian Institution Press.
Gillespie, R. G. 2004. Community assembly through adaptive radiation in Hawaiian spiders. Science 303:356–59.
Gillespie, R. G., and G. K. Roderick. 2002. Arthropods on islands: colonization, speciation, and conservation. Annual Review of Entomology 47:595–632.
Grant, P. R., and B. R. Grant. 1996. Speciation and hybridization in island birds. Philosophical Transactions of the Royal Society of London, Series B 351:765–72.
Gruner, D. S. 2007. Geological age, ecosystem development, and local resource constraints on arthropod community structure in the Hawaiian Islands. Biological Journal of the Linnean Society 90:551–70.
Haila, Y. 1990. Towards an ecological definition of an island: A northwest European perspective. Journal of Biogeography 17:561–68.
Heaney, L. R. 1986. Biogeography of mammals in SE Asia: Estimates of rates of colonization, extinction and speciation. Biological Journal of the Linnean Society 28:127–65.
———. 2000. Dynamic disequilibrium: A long-term, large-scale perspective on the equilibrium model of island biogeography. Global Ecology and Biogeography 9:59–74.
———. 2007. Is a new paradigm emerging for oceanic island biogeography? Journal of Biogeography 34:753–57.
Heaney, L. R., J. S. Walsh, Jr., and A. T. Peterson. 2005. The roles of geological history and colonization abilities in genetic differentiation between mammalian populations in the Philippine archipelago. Journal of Biogeography 32: 229–47.
Kaneshiro, K. Y., R. G. Gillespie, and H. L. Carson. 1995. Chromosomes and male genitalia of Hawaiian Drosophila: Tools for interpreting phylogeny and geography. In Hawaiian Biogeography: Evolution on a Hot Spot Archipelago, ed. W. L. Wagner and V. A. Funk, 55–71. Washington, DC: Smithsonian Institution Press.
Keast, A., and S. E. Miller, editors. 1996. The origin and evolution of Pacific Island Biotas, New Guinea to Eastern Polynesia: Patterns and processes. Amsterdam: SPB Academic.
Kvist, L., J. Broggi, J. C. Illera, and K. Koivula. 2005. Colonisation and diversification of the blue tits (Parus caeruleus teneriffae-group) in the Canary Islands. Molecular Phylogenetics and Evolution 34:501–11.
Lack, D. 1947. Darwin’s Finches: An Essay on the General Biological Theory of Evolution. Cambridge: Cambridge University Press.
Le Friant, A., C. L. Harford, C. Deplus, G. Boudon, R.S.J. Sparks, R. A. Herd, and J. C. Komorowski. 2004. Geomorphological evolution of Montserrat (West Indies): Importance of flank collapse and erosional processes. Journal of the Geological Society 161:147–60.
Levin, D. A. 2004. The ecological transition in speciation. New Phytologist 161:91–96.
Losos, J. B., and D. Schluter. 2000. Analysis of an evolutionary species-area relationship. Nature 408:847–50.
MacArthur, R. H., and E. O. Wilson. 1963. An equilibrium theory of insular zoogeography. Evolution 17:373–87.
———. 1967. The Theory of Island Biogeography. Princeton, NJ: Princeton University Press.
Moore, J. G., and D. A. Clague. 1992. Volcano growth and evolution of the island of Hawaii. Bulletin of the Geological Society of America 104:1471–84.
Nunn, P. D. 1994. Oceanic Islands. Oxford: Blackwell.
Paulay, G. 1994. Biodiversity on oceanic islands: Its origin and extinction. American Zoologist 34:134–44.
Peck, S. B. 1990. Eyeless arthropods of the Galapagos Islands, Ecuador: Composition and origin of the cryptozoic fauna of a young, tropical, oceanic archipelago. Biotropica 22:366–81.
Peck, S. B., P. Wigfull, and G. Nishida. 1999. Physical correlates of insular species diversity: the insects of the Hawaiian Islands. Annals of the Entomological Society of America 92:529–36.
Percy, D. M., A. M. Garver, W. L. Wagner, H. F. James, C. W. Cunningham, S. E. Miller, and R. C. Fleischer. 2008. Proceedings of the Royal Society of London, Series B DOI:10.1098/rspb.2008.0191.
Price, J. P. 2004. Floristic biogeography of the Hawaiian Islands: Influences of area, environment and paleogeography. Journal of Biogeography 31:487–500.
Price, J. P., and D.A. Clague. 2002. How old is the Hawaiian biota? Geology and phylogeny suggest recent divergence. Proceedings of the Royal Society of London, Series B 269:2429–35.
Price, J. P., and D. Elliott-Fisk. 2004. Topographic history of the Maui Nui Complex, Hawai’i, and its implications for Biogeography. Pacific Science 58:27–45.
Price, J. P., and H. L. Wagner. 2004. Speciation in Hawaiian angiosperm lineages: Cause, consequence, and mode. Ecology 58:2185–2200.
Pulvers, J. M., and D. J. Colgan. 2007. Molecular phylogeography of the fruit bat genus Melonycteris in northern Melanesia. Journal of Biogeography 34:713–23.
Rabosky, D. L., and I. L. Lovette. 2008. Explosive evolutionary radiations: Decreasing speciation or increasing extinction through time? Evolution 62:1866–75.
Sanmartín, I., P. van der Mark, and F. Ronquist. 2008. Inferring dispersal: A Bayesian approach to phylogeny-based island biogeography, with special reference to the Canary Islands. Journal of Biogeography 35:428–49.
Silvertown, J. 2004. The ghost of competition past in the phylogeny of island endemic plants. Journal of Ecology 92:168–73.
Silvertown, J., J. Francisco-Ortega, and M. Carine. 2005. The monophyly of island radiations: An evaluation of niche pre-emption and some alternative explanations. Journal of Ecology 93:653–57.
Stuessy, T. F. 2007. Evolution of specific and genetic diversity during ontogeny of island floras: The importance of understanding process for interpreting island biogeographic patterns. In Biogeography in a Changing World, ed. M. C. Ebach and R. S. Tangney, 117–33. Boca Raton, FL: CRC Press.
Stuessy, T. F., D. J. Crawford, and C. Marticorena. 1990. Patterns of phylogeny in the endemic vascular flora of the Juan Fernandez Islands, Chile. Systematic Botany 15:338–46.
Stuessy, T. F., D. J. Crawford, C. Marticorena, and R. Rodríguez. 1998. Island biogeography of angiosperms of the Juan Fernandez archipelago. In Evolution and Speciation of Island Plants, ed. T. F. Stuessy and M. Ono, 121–38. Cambridge: Cambridge University Press.
Stuessy, T., J. Greimler, and T. Dirnböck. 2005. Landscape modification and impact on specific and genetic diversity in oceanic islands. Biologiske Skrifter 55:89–101.
Stuessy, T. F., G. Jakubowsky, R. Salguero Gómez, M. Pfosser, P. M. Schlüter, T. Fer, B.-Y. Sun, and H. Kato. 2006. Anagenetic evolution in island plants. Journal of Biogeography 33:1259–65.
Triantis, K. A., M. Mylonas, and R. J. Whittaker. (2008) Evolutionary species-area curves as revealed by single-island endemics: Insights for the interprovincial species-area relationship. Ecography 31:401–7.
Wagner, W. L., and V. A. Funk, eds. 1995. Hawaiian Biogeography: Evolution on a Hot Spot Archipelago. Washington, DC: Smithsonian Institution Press.
Wallace, A. R. 1902. Island Life, 3rd ed. London: Macmillan.
Whelan, F., and D. Kelletat. 2003. Submarine slides on volcanic islands—a source for mega-tsunamis in the Quaternary. Progress in Physical Geography 27:198–216.
Whittaker, R. J., and J. M. Fernández-Palacios. 2007. Island Biogeography: Ecology, Evolution, and Conservation, 2nd ed. Oxford: Oxford University Press.
Whittaker, R. J., R. J. Ladle, M. B. Araújo, J. M. Fernández-Palacios, J. Delgado, and J. R. Arévalo. 2007. The island immaturity–speciation pulse model of island evolution: An alternative to the “diversity begets diversity” model. Ecography 30:321–27.
Whittaker, R. J., K. Triantis, and R. J. Ladle. 2008. A general dynamic theory of oceanic island biogeography. Journal of Biogeography 35:977–94.
Wilson, E. O. 1969. The species equilibrium. Brookhaven Symposia in Biology 22:38–47.
1 Anagenesis (the evolutionary change from a colonist species to a neo-endemic form) does not lead to an increased number of species on an island (although it does increase endemism). Thus, as they were primarily concerned with understanding variation in species richness, MacArthur and Wilson (1963, 1967) focused on evolutionary change giving rise to increased richness, i.e., cladogenesis (sensu Stuessy et al. 1990) when outlining their dynamic equilibrium model.
