Download Diversification in the Tropical Pacific: Comparisons Between Marine

INTEGR. COMP. BIOL., 42:922–934 (2002)
Diversification in the Tropical Pacific: Comparisons Between Marine and Terrestrial Systems
and the Importance of Founder Speciation1
GUSTAV PAULAY2
AND
CHRIS MEYER
Florida Museum of Natural History, University of Florida, Gainesville, Florida 32611-7800
INTRODUCTION
Patterns of species distribution can differ strikingly
between land and sea, or between islands and continents. For example, terrestrial or insular organisms
tend to have more limited geographic ranges than marine or continental taxa. The processes underlying
these distributional patterns, such as the relative importance of dispersal or vicariance, are often considered to also differ markedly among biotas. Dissimilar
distributional patterns and implied differences in process were quite evident between the terrestrial and marine studies featured in this symposium. However differences in pattern can arise not only through differences in process, but also through differences in the
spatial or temporal scale across which processes operate.
The islands of Oceania offer an excellent setting to
explore the contrast between terrestrial and marine distributional patterns and processes, because the geographic setting for terrestrial and shallow water benthic
marine organisms is the same: the myriad islands
strewn across the vast Pacific. The size of distributional ranges and the geographic distribution of endemism are two biogeographic attributes that are
thought to differ markedly between the terrestrial and
marine biotas of Pacific islands (see below). While terrestrial species are frequently confined to single islands
or archipelagoes throughout Oceania, marine species
tend to have wide to very wide distributions, and are
rarely restricted to single island groups except on the
most isolated archipelagoes. These differences have
led to the perception that different processes govern
distributional dynamics and diversification in terrestrial and marine settings. For example, founder speciation is thought to be the primary, and perhaps only,
process in the differentiation of the Pacific island land
biotas (e.g., Mayr and Diamond, 2001; Carson, 1983;
Wagner and Funk, 1995), while vicariant processes
tend to be preferentially considered for the origin of
marine species of the area (e.g., Springer, 1982; Lessios et al., 2001; Planes and Fauvelot, 2002; Bellwood
and Wainwright, 2002).
We first review distributional patterns in Oceania,
then consider how gene flow establishes and connects
insular populations, and how it varies in space and
time. We show that: 1) founder speciation and vicariance are ends of a continuum rather than dichotomies,
2) founder speciation is more important in the sea than
previously thought, 3) the same processes control gene
flow in marine and terrestrial systems, although their
efficacy varies both within and between these systems,
and 4) there is considerable overlap in both patterns
of distribution and diversification processes on land
and sea.
DISTRIBUTIONAL PATTERNS
IN
OCEANIA
Extending from New Guinea to Hawaii and Easter
Island, Oceania encompasses most of the oceanic islands in the Pacific basin. Oceania forms the eastern
part of the Indo-West Pacific (IWP), an area that extends further westward to the Red Sea and East Africa,
making it the largest marine biogeographic region on
Earth. The IWP is characterized in part by the numerous species that range throughout its vast extent (Ekman, 1953). Although marine species with more re-
1 From the Symposium Integrated Approaches to Biogeography:
Patterns and Processes on Land and in the Sea presented at the
Annual Meeting of the Society for Integrative and Comparative Biology, 2–6 January 2002, at Anaheim, California.
2 E-mail: [email protected]
922
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
SYNOPSIS.
Patterns of distribution and processes of differentiation have often been contrasted between
terrestrial and marine biotas. The islands of Oceania offer an excellent setting to explore this contrast,
because the geographic setting for terrestrial and shallow-water, benthic, marine organisms are the same:
the myriad islands strewn across the vast Pacific. The size of species ranges and the geographic distribution
of endemism are two biogeographic attributes that are thought to differ markedly between terrestrial and
marine biotas in the Pacific. While terrestrial species are frequently confined to single islands or archipelagoes throughout Oceania, marine species tend to have wide to very wide distributions, and are rarely
restricted to single island groups except for the most isolated archipelagoes. We explore the conditions under
which species can reach an island by dispersal and differentiate. Genetic differentiation can occur either
through founder speciation or vicariance; these processes are requisite ends of a continuum. We show that
founder speciation is most likely when few propagules enter the dispersal medium and survive well while
they travel far. We argue that conditions favorable to founder speciation are common in marine as well as
terrestrial systems, and that terrestrial-type, archipelagic-level endemism is likely common in marine taxa.
We give examples of marine groups that show archipelagic level endemism on most Pacific island groups
as well as of terrestrial species that are widespread. Thus both the patterns and processes of insular diversification are variable, and overlap more between land and sea than previously considered.
DIVERSIFICATION
TABLE 1.
Endemism in insular gastropods.
Mariana
Land snails
Marine snails
IN THE
90% (N 5 96)
1% (N 5 843)
Hawaii
97% (N 5 781)
21% (N 5 572)
% endemism and number of species given; data from Eldredge
and Miller, 1995; Bauman, 1996; and Smith, 2002.
923
if they do, why? In the final section of the paper we
show that there is considerably more overlap in distributions and endemism than currently perceived. Do
the differences in pattern imply differences in process,
as suggested by some, or only differences in the scale
across which processes operate? By considering temporal and spatial variation in gene flow we show that
the same processes operate, but scaling is fundamentally important. Below we first explore spatial and temporal variation in gene flow, then consider how various
factors affect gene flow and thus influence the likelihood of speciation, and conclude by reexamining patterns of speciation and endemism in marine vs. terrestrial settings.
PROCESSES AFFECTING DISTRIBUTION
AND DIFFERENTIATION
Oceanic islands acquire their biota strictly through
dispersal, and dispersal frequency (gene flow) is a major determinant of whether populations can become
established and whether they differentiate from each
other. If dispersal is too rare to lead to successful colonization, then an organism cannot become established. The absence of large suites of taxa, i.e., a ‘‘disharmonic’’ biota, is one of the most striking feature of
islands (Carlquist, 1974). Differentiation is prevented
by too frequent dispersal, as demonstrated by numerous studies on typical, marine species among IWP islands (Lessios et al., 1998, 1999; Bowen et al., 2001).
At intermediate levels of dispersal, an organism can
colonize in the long term, yet is not prevented from
differentiating by gene flow in the short term. Alternatively if dispersal frequency decreases through time,
then an established population can become genetically
isolated and differentiate. Finally, selection can lead to
differentiation even in the face of gene flow, a process
that we do not further consider in this paper.
Thus spatial and temporal variation in dispersal determines whether a population can both establish and
differentiate on an island. Genetic differentiation can
occur if: 1) frequency of dispersal is very low (K1
migrant/generation time), the process of founder speciation, or 2) dispersal probability changes through
time so that gene flow approaches zero, the process of
vicariant speciation. The need for a species to become
widespread through dispersal before it can be sundered
by vicariance, and the implications of this necessity,
are often ignored. These processes can be depicted
graphically as follows. The density of potential recruits
decreases away from an island, so that up to a certain
distance around an island gene flow is too high to allow differentiation (zone F: Fig. 1). There is a zone
beyond this high gene flow zone that colonists occasionally reach and can establish in, but where gene
flow is too limited to prevent divergence. This second
band, which we call the speciation zone (S: Fig. 1), is
where founder speciation can occur. MacArthur and
Wilson’s (1967; p. 175) ‘‘radiation zone’’ is analogous
to the speciation zone here depicted: ‘‘In equilibrial
biotas . . . the following prediction is possible: adap-
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
stricted ranges are more common than pan-IWP taxa,
most still range across several archipelagoes and over
thousands of kilometers (Kay, 1984). That most marine taxa range widely has often received support under genetic scrutiny. Most published phylogeographic
studies have shown genetic cohesion in IWP marine
species over basinal to subbasinal ranges, even when
showing isolation by distance (Benzie and Williams,
1997; McMillan et al., 1999; Planes, 2001). Although
cryptic species have often been discovered, they are
often themselves wide-ranging (Colborn et al., 2001;
Lessios et al., 2001), and frequently sympatric with
congeners (Borsa and Benzie, 1993; Palumbi et al.,
1997; Williams, 2000). Archipelago-level endemism
among marine organisms is perceived to be substantial
only on the most isolated, peripheral island groups in
Oceania. Randall (1998) documents high fish endemism in the Hawaiian (23%), Easter (22%), Marquesas
(10%), Lord Howe-Norfolk (7.2%) and Rapa (5.5%)
islands, while endemism in more centrally located archipelagoes is ,2%. Similar patterns of endemism
characterize other marine groups surveyed (Kay,
1984).
In contrast to the wide ranges and rarity of nonperipheral, archipelagic endemism in most marine
taxa, substantial endemism characterizes many terrestrial groups throughout the Pacific, irrespective of location. Although there is variation among taxa (often
relating to differences in dispersal ability), endemism
in many groups is high at the archipelagic and even
single-island level on volcanic islands (atolls have little endemism because their terrestrial habitats are
ephemeral). Endemism in terrestrial taxa can lead to
in situ diversification through intra-archipelagic and
even intra-island speciation (Paulay, 1985; Wagner and
Funk, 1995). The resulting evolutionary radiations are
one of the most striking features of insular biotas. This
contrast between marine and terrestrial endemism is
well illustrated by the gastropods of the Mariana and
Hawaiian Islands. The land snail faunas of both areas
are highly endemic, somewhat more so on isolated Hawaii than in the west Pacific Marianas. However while
marine snails in Hawaii show notable endemism, very
few are restricted to the Marianas (Table 1).
Thus a major perceived contrast between land and
sea biotas in Oceania is the geographic scale and location of endemism and speciation: marine organisms
are generally widespread, differentiating on an archipelagic scale only on the most isolated islands, while
many terrestrial organisms differentiate on any high
island. Does this contrast hold up to scrutiny: do land
and sea biotas of Oceania differ broadly like this and
TROPICAL PACIFIC
924
G. PAULAY
AND
C. MEYER
tive radiation will increase with distance from the major source region and, after corrections for area and
climate, reach a maximum on archipelagos and large
islands located in a circular zone close to the outermost
dispersal range of the taxon.’’
Both zones can change in diameter, because factors
affecting dispersal probability vary through time; such
variation can lead to vicariance when appropriately
scaled. Three scales are relevant in determining whether founder or vicariant speciation is more likely: 1) the
location and width of the speciation zone, 2) the spatial
scale of temporal fluctuations in the boundaries of the
gene flow and speciation zones, and 3) the temporal
dynamics of those fluctuations. Variations in the first
two are depicted in Figure 1. As we show below, variation can occur over a wide range of time scales,
which taken together with variability in the first two
factors implies that founder and vicariant speciation
are extremes of a complex continuum.
Temporal fluctuations in gene flow can originate either from variation in inter-island distance, variation in
the velocity of transport medium (air or water), or
from variation in the number of propagules entering
the dispersal medium (this last is treated below). Variation in inter-island distance can result from 1) islands
arising or disappearing, 2) habitats on stepping stone
islands becoming favorable or unfavorable, and 3) islands moving relative to one another. Islands can arise
through volcanic action, tectonic uplift, or sea-level
fall, and can be lost through erosion, subsidence, or
sea-level rise. New islands become stepping stones between older islands; one manifestation of this process
has been described as island integration (Rotondo et
al., 1981). Suitable habitats on islands can arise or be
lost in a variety of ways, perhaps most strikingly as a
result of sea-level changes (Paulay, 1991). Thus rising
sea level can drown islands, resulting in the loss of
terrestrial and potential loss of shallow-water habitats,
but can also create sheltered lagoons where none existed before. In contrast, falling sea level can change
atolls, with marginal terrestrial habitats, into high,
limestone islands, well suited as stepping stones (Ladd,
1958; Leopold, 1969). Falling sea levels can strand
lagoons and expose seamounts, or bring them into the
depth range of littoral species. Finally while the spacing of islands on the same tectonic plate are relatively
invariant, islands on different plates can move toward
or apart from each other. The tectonic and eustatic processes that affect inter-island distance occur over 10
Ka to Ma time frames.
The other physical factor that affects dispersal probability is variation in velocity of transport medium.
Wind speed is considerably more variable than current
velocity. Wind speeds vary over 0–100 m·sec21, although speeds .45 m·sec21 are restricted to within cyclonic systems. Conversely, ocean current speeds
range over 0–3 m·sec21. Thus wind transport can be
much more effective from a velocity perspective, but
its efficacy is limited by the relatively greater density
of the propagules than of the transport medium, so
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
FIG. 1. Graphical model of vicariant and founder speciation. Dark zone corresponds to zone of high gene flow (F), outer zone corresponds
to zone of speciation (S); source and recipient islands marked. Vicariance is likely when speciation zone is relatively narrow but width of
zone of high gene flow changes through time (A). Founder speciation is likely when speciation zone is relatively wide (B). Arrow denotes
time.
DIVERSIFICATION
IN THE
925
ciation events geographically cluster in a handful of
locations characterized by major dispersal barriers:
wide stretches of open ocean or otherwise inhospitable
habitats. Speciation across such dispersal barriers
could occur either by vicariance (if an identifiable
event can be linked to the barrier) or by founder speciation. We found that speciation events across all barriers were spread out over time, implying that they
have been continuously important in speciation. This
pattern is consistent with founder speciation or recurrent vicariance (events recurring on frequent time
scales as discussed above), but not with classical vicariance.
The likelihood of vicariant speciation depends on
the temporal scale of physical fluctuations, which need
to be determined empirically. As discussed above,
classic vicariance may be uncommon within ocean basins. The likelihood of founder speciation depends on
how the probability of dispersal varies spatially across
the terrain (see Fig. 3 below), and on the efficacy of
colonists at establishing populations. The spatial distribution of dispersal probability depends on the number of propagules entering the dispersal medium, their
speed, the rate at which they leave the dispersal medium (through death, settling out, etc.), and their dilution with distance due to physical processes (diffusion, transport speed and direction, etc.). How these
factors affect dispersal probability with distance is examined in a simple model below, but first we review
how these factors vary among organisms with different
dispersal strategies.
DISPERSAL MECHANISMS
There is a voluminous literature documenting the
diversity of dispersal mechanisms and evaluating their
importance and efficacy (Carlquist, 1981). Terrestrial
organisms disperse between islands through air or at
the air/water interface. Active flight or turbulence can
keep propagules in the air stream. Dispersal by active
flight is limited by metabolic resources and psychological barriers (Diamond et al., 1976). Hitching on
active flyers is further limited by the propensity of the
propagule to become and remain attached to the dispersal vector, in addition to inherent limitations of the
vector. Attachment probability depends on encounter
rate and efficacy of attachment mechanism, either external (barbed or sticky propagules, entrapment in mud
attached to bird) or internal (by ingestion or inoculation). Passive wind transport can be a typical part of
the life cycle of organisms adapted for wind dispersion
(e.g., fern spores), be a likely chance event for organisms that enter the air column on their own but are too
weak to fly against stronger winds (e.g., minute flying
insects), or be a rare event for organisms that are only
buoyed and transported in extreme wind events (e.g.,
small mammals transported modest distances by extreme storms). The likelihood of dispersal at the air/
water interface also varies from a frequent process for
organisms with propagules that float and survive in
salt water and live near ocean or river shores (e.g.,
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
settling out of suspension becomes a limitation to
transport. Variation in wind and current velocities
range over time scales from diurnal, through years/
decades (e.g., ENSO), through 104–105 years (e.g.,
Milankovitch climatic cycles), and over millions of
years (circulation and wind patterns driven by tectonic
changes). For example, the North Equatorial Counter
Current varies seasonally from absent to a mean flow
of 0.39 m·sec21, but individual buoy tracks show flow
rates as high as 0.9 m·sec21 (Richmond, 1988). Such
extreme events are probably more important to gene
flow than mean velocity. Peak flow during the 1982–
83 ENSO reached 1.4 m·sec21, about 50% higher than
in typical years.
The variability of dispersal in time and space implies that vicariance and dispersal, or vicariant and
founder speciation, represent end members of a continuum. ‘‘Classic’’ vicariant speciation occurs when
gene flow that has kept populations in genetic continuity is suddenly severed, i.e., when there is long-term
temporal variation in dispersal frequency. ‘‘Classic’’
founder speciation occurs when dispersal is so unlikely
at all times that a population resulting from a successful colonization event experiences insufficient gene
flow to prevent differentiation, i.e., when there is
large-scale spatial variation in dispersal frequency. As
dispersal probability varies over temporal and spatial
scales, these distinctions are blurred. For example inter-island distance and velocity of transport medium
vary over multiple time scales, so that ‘‘classic’’ vicariance gives way to recurrent vicariance when variation is over Milankovitch (10 Ka–100 Ka) time
scales, and to ecological-scale fluctuations in dispersal
when occurring on daily to millennial time scales (cf.,
Bennett, 1990). Veron (1995) has highlighted variation
in currents over Milankovitch time scales as ‘‘circulation vicariance.’’ However we emphasize that currents, as well other factors impacting dispersal probability, exhibit complex variations over all time scales.
This implies in part that ‘‘classic’’ vicariance is less
likely across the ocean than often considered, as it requires that the amplitude of variations in current velocities on a time-scale relevant to speciation substantially exceed the amplitude of variations in the short
to medium (Milankovitch) term.
While the model of founder speciation is routinely
applied to explain the origin of terrestrial endemics on
islands (e.g., Zimmerman, 1948; Carson, 1983; Wagner and Funk, 1995), it is frequently not considered
for the origin of marine endemics (e.g., Springer, 1982;
Lessios et al., 2001; Planes and Fauvelot, 2002; Bellwood and Wainwright, 2002). One of the points we
wish to emphasize is that founder speciation is likely
also important in marine systems. We came to appreciate the potential importance of this process in a
large-scale phylogeographic analysis of the gastropod
family Cypraeidae (Meyer and Paulay, submitted). We
found that cowrie sister taxa frequently are allopatric,
retaining the geographic signature of their origin
through allopatric speciation. We also found that spe-
TROPICAL PACIFIC
926
G. PAULAY
AND
C. MEYER
FIG. 2. Proportion of the indigenous vascular flora considered to have been derived from propagules that arrived by various modes of dispersal
on Pacific atolls (A) and high islands (B). Modified from Carlquist, 1974, figure 2.7.
pelagic life history stages (planktonic larvae, medusae), active swimming (larger vertebrates), or rafting.
Planktonic life history stages are widespread among
marine organisms, and in many groups species that
lack planktic development rarely or never reach oceanic islands (Bouchet and Poppe, 1988; Kohn and Perron, 1994). There is often (Kohn and Perron, 1994;
Meyer, 2002), but not always (Thresher and Brothers,
1985; Wellington and Victor, 1989), a correlation between length of pelagic period and geographic range
of species. Rafting on algae, pumice, and other floating
materials is an effective dispersal method for many
taxa, especially sessile and small-bodied mobile species (Highsmith, 1985; Jokiel, 1989, 1990; Hobday,
2000), as demonstrated by the wide distribution of
many clonal, sessile metazoans with very short
(,days) or no planktonic larval stages (Jackson,
1986). The propensity to enter the dispersal medium
varies substantially among rafters.
As these examples illustrate, variation in propensity
to enter and remain in the transport medium is widespread, with important consequences for the efficacy
of dispersal. Carlquist (1974) emphasized this variation by dividing drift into ‘‘frequent’’ and ‘‘rare’’ categories, corresponding to seed floatation and rafting
respectively. Below we explore the implications of this
variation.
DETERMINANTS
OF
COLONIZATION PROBABILITY
General considerations
In light of the variety of ways to disperse in both
land and sea, what are the main determinants of colonization probability and gene flow and thus of founder vs. vicariant speciation? Successful colonization depends on dispersal plus the ability to establish a population after dispersal. Dispersal probability depends
on 1) inter-island distance, 2) velocity of transport, 3)
rate of entry into the transport medium, and 4) rate of
leaving the transport medium. The first two factors are
external, governed by tectonic and climatic factors.
Temporal variation in these can lead to either founder
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
most strand plants), to rare events for species that can
float only on rafts and live away from the coast.
Taxon-specific studies have documented the relative
importance of different dispersal modes in Oceania.
Carlquist (1974, 1981) showed that bird-assisted dispersal is generally most important, while aerial dispersal is least important for seed plants. Atolls differ
from high islands however in the greater importance
of drift dispersal (Fig. 2), and absence of unassisted
aerial dispersal. This reflects the maritime conditions
of atolls and low diversity of their flora. Terrestrial
arthropods can disperse by passive wind transport, active flight, bird-assisted transport, and rafting, especially for deadwood associates (Zimmerman, 1942,
1948). Aerial plankton traps deployed on ships and
airplanes have documented a diversity of small to minute terrestrial insects and arachnids in the air stream
over the open Pacific (Yoshimoto and Gressitt, 1960;
Holzapfel, 1978; Holzapfel et al., 1978). Yoshimoto
and Gressitt (1960) noted the importance of storms for
transport: ‘‘nearly all specimens taken more than 800
km from land were taken during storm periods.’’ Zimmerman (1948, 1976) also emphasized the importance
of hurricane-force winds as a necessity for long-distance arthropod dispersal. The small body size of Pacific land snails has been considered indicative of the
importance of aerial transport, with drift transport secondary (Vagvolgyi, 1975). There is ample anecdotal
evidence for bird-assisted transport of land snails, and
the possibility of independent aerial transport is suggested by results of aerial plankton surveys (Vagvolgyi, 1975). The efficacy of aerial transport for small
propagules can be tremendous. Many terrestrial and
freshwater protist and microscopic metazoan morphospecies have attained cosmopolitan distributions as a
consequence (Finlay, 2002).
The dispersal of marine organisms is by water, although the potential cross-continental aerial dispersal
of a fungal pathogen capable of terrestrial and marine
existence (Shinn et al., 2000) shows there are exceptions even to this generalization. Shallow-water, benthic marine organisms can disperse through specialized
DIVERSIFICATION
IN THE
TROPICAL PACIFIC
927
speciation if changes envelop new islands within Zone
S (Fig. 1) or vicariance if changes exclude islands
from Zone F. The third and fourth factors are at least
partly intrinsic to the organism. The influence of these
four factors in the simplest case can be expressed as
follows:
N 5 Q e2lt f(D) 5 Q e2Dl/v f(D)
Eqn. 1
Where N is the number of propagules that reach the
target island, Q is the number that started dispersing
(factor 3), l is the mortality rate (factor 4), t is time,
v is velocity of transport (factor 2), and f(D) is a function of D (distance between source area and target island [factor 1]) that determines what proportion of the
surviving propagules are in the path of the target island
(cf., Jackson and Strathmann, 1981). f(D) varies with
type of dispersal. In the simplest case, if the probability of transport is equal and steady in all directions
(such as an active flyer leaving the source island in
random, but straight direction in calm weather), it is a
fraction of D (f(D) 5 tan21((I/D)/2p), where I is the
diameter of the target island’s signature seen by the
dispersing agent). It is usually more complex however,
thus for water transported propagules it includes advection and diffusion terms (see e.g., Cowen et al.,
2000). Although the model is simplistic in that it assumes no variation or time dependence in any of these
parameters, it illustrates the impact of these four factors well.
It is instructive to examine how N varies with D,
and how this depends on Q and l/v (Fig. 3). l/v has
a greater influence on the relative width of the gene
flow vs. speciation zones, whereas Q strongly influences how these zones scale with D. Increasing l/v
leads to a decrease in the relative width of the speciation zone, while increasing Q leads to an increase in
the width of both zones. Thus high l/v with low Q
leads to little dispersal and thus narrow zones of gene
flow as well as speciation—such taxa are not expected
to reach most oceanic islands. High l/v with high Q
leads to a wider zone of gene flow and slightly wider
speciation zone—such taxa are expected to reach nearby but not remote islands. Low l/v with high Q leads
to a very wide zone of gene flow followed by a wide
speciation zone—such taxa are expected to be panmictic over wide areas, although they could also undergo founder speciation if sufficiently remote islands
exist. Low l/v with low Q leads to narrow zone of
gene flow followed by a wide speciation zone—such
taxa are expected to reach islands and undergo founder
speciation there. Thus species that place few propagules into the dispersal medium, but have high survivorship over the distance dispersed are the most likely
to undergo founder speciation.
How does l/v vary among organisms with different
dispersal strategies on land and sea? Velocity of transport is greater in air than ocean, although many terrestrial organisms disperse by sea. There is substantial
variation in l in both marine and aerial dispersal. l
includes mortality as well as leaving the transport medium in other ways. Mortality rates of small larvae in
the plankton are generally high (.0.1 d21) but are lower for large marine larvae (Strathmann, 1985), potentially lower for sessile taxa attached to flotsam (Helmuth et al., 1994; Hobday, 2000), and likely quite low
for large, floating seeds. Among propagules passively
transported by air, l is in large part dependent on the
rate of settling out of suspension, which is related to
turbulence and the surface area/mass ratio of the propagule. Thus spores and minute, winged insects stay
suspended longer than seeds and large, flightless insects. Active flyers and organisms hitchhiking on them
have low l, if mortality is low in flight. Thus low
values of l/v, which allow propagules to disperse far
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
FIG. 3. Colonization vs. distance curves based on eq. 1, in the simplest case with no diffusion or advection, when f(D) 5 tan21((I/D)/2p).
Parameters set at l 5 0.1 day21, I 5 1 km, v 5 1 m·sec21 (A, B) vs. 0.05 m·sec21 (C), Q 5 10,000 (A) vs. 100 (B, C). Thus curves B vs. C
contrast in l/v, while curves A vs. B contrast in Q. Founder speciation is most likely with low Q and low l/v (curve B). This can be visualized
by arbitrarily defining the speciation zone to fall at 0.1 . N . 0.01, or in the lowest section of the graph.
928
G. PAULAY
and thus open the possibility of founder speciation, are
possible in both terrestrial and marine systems: e.g.,
rapid transport of seeds on birds (Carlquist, 1974) or
of small sessile biota rafting on algae (Helmuth et al.,
1994; Hobday, 2000). An important aspect of mortality
not addressed in the simple model is time dependence
of l. Thus planktonic mortality rates are not time-independent in many taxa; most importantly l reaches
unity at the end of the competent period for larvae
with finite life spans. The implications of this dependence are not further explored here.
C. MEYER
may remain in the vicinity of their source island as a
result of varied hydrographic and behavioral (both of
spawning adult and of larvae) factors (Johannes, 1978;
Phillips, 1981; Lobel and Robinson, 1988; Swearer et
al., 1999).
Entry into the transport medium for terrestrial organisms includes flying, becoming suspended in air,
attaching to a bird, or ending up floating in the sea,
free or on a raft. There is considerable variation in the
probability of each of these. Thus many plants have
propagules (fern and bryophyte spores, minute seeds,
etc.) well suited for aerial dispersal, while others (large
seeds and fruits) can only become airborne briefly in
violent storms. The efficacy of bird transport depends
in part on the likelihood of the propagule encountering
a bird, the facility of attachment, and the probability
of the bird flying to another island. As sea and shore
birds travel among islands routinely while land birds
rarely leave, habitat has a great influence on which of
these vectors an organism encounters, and thus on the
likelihood of transport (see Carlquist, 1974 for a review of bird-assisted dispersal). Probability of a propagule becoming water-born depends on the proximity
of the organism to shore or streams.
The probability of entering the air stream off the
island also varies greatly. For passive floaters this depends largely on exposure to wind. Thus spores have
a much greater probability of entering the air stream
from ferns living on wind-swept, open terrain than
from ferns living as groundcover in dense forest.
Among volant insects, behavior and time spent in
flight varies greatly, both influencing probability of being swept off by the air stream. For example Zimmerman (1942) noted the habit of the damselfly Ischnura
aurora to fly straight up and out of sight immediately
after hatching as a likely reason for the wide distribution of the species across Oceania. Active flyers like
birds, although frequently in the transport medium,
have highly variable inclination to cross water (Diamond et al., 1976). Floating and rafting propagules
have similar entrainment problems as the marine organisms discussed above.
Thus rate of entry into the transport medium varies
in both marine and terrestrial systems. High rates of
entry occur in organisms like fecund marine invertebrates with small, planktotrophic larvae and ferns with
abundant spores or protists that can form resistant
cysts. Low rates of entry typify marine direct developers that do not readily associate with algae or pumice, and the majority of terrestrial organisms.
Ability to found population after dispersal
The final factor that determines colonization probability is the ability of dispersed propagules to found
a population. Abundant anecdotal literature on the occasional arrival of species on islands that fail to establish populations underscores the importance of this
factor (e.g., Zimmerman, 1948; Emerson, 1991). Colonization success depends in part on availability of
appropriate habitat, which varies greatly among islands
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
Entry into the transport medium
When l/v is not so high that it effectively prevents
organisms from reaching an island by dispersal, then
entry into the transport medium becomes the major
determinant of whether an organism can undergo
founder speciation. Again there is considerable variation within, and wide overlap between, marine and terrestrial systems. Rate of entry into the transport medium depends not only on a propagule’s probability of
entry, but also on the number of propagules available
to begin with, thus the size and fecundity of the source
population. Entry into the transport medium includes
not only becoming suspended/floating in air/water, but
also entering the airstream or current that carries the
organism away from the source area. These two processes are examined in turn for marine and terrestrial
taxa.
Entry into the transport medium may or may not be
a regular part of the life cycle of an organism. In the
sea many species produce specialized dispersal stages
that invariably enter the water column, while others,
with direct development, will only enter the transport
medium at a much lower frequency, on rafts. Although
both distributional data on clonal, sessile species with
abbreviated development (Jackson, 1986), and observation of organisms attached to flotsam (Jokiel, 1989,
1990; Hobday, 2000) indicate that rafting is an important mode of dispersal, the efficacy of rafting
among taxa is highly variable. For example Jokiel
(1990) found that a few coral groups made up the bulk
of the colonies rafting on pumice in the South Pacific.
Thus 70% of colonies belonged to the Pocilloporidae,
which includes about 5% of the species diversity of
the area, while the majority of genera were not encountered. Although this does not imply that other corals never raft, it indicates that rafting can be a rare
event even among taxa thought to be especially well
suited to it. Thus there is a great range of variation
among marine species in probability of entering the
dispersal medium, from species with larvae adapted to
a planktonic existence, to groups that raft commonly
(e.g., Sargassum-associated biota), to species that rarely if ever get so transported.
Even if a propagule enters the water column, it may
nevertheless not enter a transporting current. Our understanding of entrainment vs. transport of planktonic
propagules around islands is still developing, but it is
clear that a large proportion of even long-lived larvae
AND
DIVERSIFICATION
IN THE
PATTERNS OF SPECIATION
The foregoing leads to the expectation that endemism through founder speciation is most likely for organisms that rarely enter the transport medium but survive well in it. Before considering this hypothesis it is
important to establish that organisms that rarely enter
the air/water indeed are capable of reaching distant
islands. This needs little emphasis for terrestrial organisms, as most lack ready means of regular dispersal, yet have colonized remote islands, giving rise to
largely-endemic biotas (Zimmerman, 1948; Carlquist,
1974). Among some marine taxa, however, there has
been a perception that groups that lack pelagic dispersal stages are largely absent from oceanic islands.
Thus Kohn and Perron (1994) showed that direct-developing cones (Conus) are virtually absent from oceanic islands, while Bouchet and Poppe (1988) and
Paulay (1997) have commented on the almost complete restriction of the direct-developing gastropod
family Volutidae to continental shores. The same applies to directly developing members of the Cypraei-
929
dae, that although have evolved on several occasions
(Meyer, 2002) are all restricted to continental areas.
Although direct development is rare among large mollusks on islands, it is not uncommon among small species, thus among taxa most capable of rafting dispersal
(Soot-Ryen, 1960; Johannesson, 1988; Sleurs and
Preece, 1994; Paulay and Preece, in preparation). Similarly, no clear relation between larval duration and
distribution is evident in many clonal animals, where
rafting dispersal is thought to be especially common
(Jackson, 1986).
The hypothesis that organisms that rarely enter the
dispersal medium are more prone to founder speciation
than those that enter it commonly is supported by
Wagner’s (1991) analyses of the Hawaiian and Marquesan floras. As discussed above, entry into the dispersal medium is less likely on/in a bird than as airborne spores or floating seeds. Wagner showed that
within-archipelago diversification is most common
among bird-dispersed plants, and less common among
plants that rely on drift or aerial dispersal. The driftdispersed strand flora (Whistler, 1992) and ferns,
whose minute spores are readily carried by the wind,
tend to have especially wide geographic ranges. Endemism among ferns is considerably lower than
among seed plants both in Hawaii (69% vs. 89%) and
the Marquesas (16% vs. 55%) (Wagner, 1991). The
only genetic test we are aware of that has assessed
differentiation across the Pacific of a readily dispersed
plant is Wright et al.’s (2000) work on Metrosideros,
a myrtaceous tree with minute seeds that only require
5–19 km/hr winds to be lofted. Metrosideros shows
almost no genetic differentiation across Oceania from
Hawaii to New Zealand. Although Wright et al. (2000)
interpreted these results as evidence of a recent spread
of these plants across Polynesia, an alternative hypothesis is that Metrosideros colonized Oceania further back in time, but remains genetically undifferentiated because of high levels of gene flow. The observation that Metrosideros harbors a diverse biota of
specialist herbivores throughout Polynesia (Gagné,
1979; Stein, 1983; Paulay, 1985) supports the latter
hypothesis.
Marine organisms that rarely enter the dispersal medium are also more prone to speciation on islands.
Thus endemism is higher among marine organisms
with direct development than among those with pelagic dispersal stages. For example, while 66% of marine
amphipods (obligate brooders) are endemic to Hawaii,
18–34% of other crustacean taxa (with exclusively/
predominantly pelagic development) are endemic (Kay
and Palumbi, 1987). Direct developing mollusks described from Oceania are often highly endemic (Reid
and Geller, 1997; Sleurs and Preece, 1994; Paulay and
Preece, in preparation). Several lineages of minute bivalves that live in caves and reef interstices and have
direct (typically brooded) development (Hayami and
Kase, 1993, 1996) are both widespread across the Pacific and differentiated on an archipelagic scale across
Oceania (Hayami and Kase, 1996; Paulay and Preece,
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
and species, but not systematically between land and
sea. It also depends on the ability of the organism to
complete its life cycle by finding a mate (if sexual)
and retaining offspring in sufficient density to allow
the population to persist. Thus asexual species, selfing
hermaphrodites and species capable of sperm storage
have advantages as colonizers. Hermaphroditism is
also advantageous over gonochorism, as it increases
the effective population size and probability that both
sexes are present in the incipient population. Potential
mates need to encounter each other; thus mobility is
also advantageous. Free-spawning organisms of lowmobility are limited by fertilization success at low
population densities (Levitan, 1995). Similarly, the absence of suitable pollinators can limit the establishment of plants, and pollinators are often of low diversity on remote islands (Zimmerman, 1948; Carlquist,
1974).
Highly dispersive, pelagic stages in the life cycle of
many marine organisms facilitate dispersal to remote
islands, but hinder successful establishment, because
first generation larvae may become too spread out to
meet potential mates after settlement. Thus outcrossing
organisms with long-lived planktonic stages may need
a substantial number of colonists for successful establishment. This makes founder speciation less likely, as
the necessity of a substantial founder population
makes subsequent gene flow more likely. The actual
number of colonists need not be very high though, as
introductions of marine taxa with such life cycles have
shown. For example Trochus niloticus, a gonochoric
marine gastropod with a short-lived (days) planktonic
larva, has been repeatedly introduced to Pacific islands
from as few as tens of colonists (Eldredge, 1994).
However, these few initial colonists were likely placed
in close proximity on the reef, increasing the likelihood of finding potential mates, and establishing a viable population.
TROPICAL PACIFIC
930
G. PAULAY
C. MEYER
pear spread out over time (Meyer and Paulay, 2002;
Geller et al., in preparation). As discussed above
founder speciation is expected for direct developers
like brooding bivalves, but is more surprising for cypraeid, patellid and turbinid gastropods, groups with
planktonic larvae, and thus relatively high Q. However
members of all three groups differentiating in this
manner have relatively short-lived larvae: the erroneines, the cowrie group most prone to such fine scale
differentiation, have protoconchs suggestive of abbreviated planktonic development (Meyer, 2002), while
patellids and turbinids have short-lived, non-feeding
larvae. Hydrographic entrainment and retention of larvae around islands (Lobel and Robinson, 1988; Swearer et al., 1999) may be relatively much more important
for short-lived than for more long-lived propagules,
leading to a relatively low effective Q for these taxa.
Entrainment and retention also make the establishment
of a population from second generation larvae more
facile.
As this paper reviews, new studies are filling in the
gaps between terrestrial and marine distributional patterns indicating taxa in both systems span the gamut
of fine spatial structure to interconnectedness (Fig. 5).
While highly-structured marine clades across Oceania
were not known in the past, many are being discovered, particularly in poorly dispersing taxa, filling in
that end of the marine spectrum. Also, widespread terrestrial taxa are known that show little structure
throughout the basin, completing the opposite end in
the terrestrial spectrum. Future studies are likely to
flesh out the range of differentiation across Oceania
even more. However, true gaps, such as the lack of
marine intra-island or intra-archipelagic endemics,
point to unique differences between these two systems
as one moves to finer spatial scales. Perhaps such finescale differentiation is constrained by selective forces
acting on dispersal potential in the marine realm. This
paper enforces the continuity of these patterns for both
terrestrial and marine biotas under a single model—
variation in connectivity in time and space. The fact
that the same changing dynamics under our simplistic
model can create both vicariant and founder speciation
events depending on the specific geographic setting
reinforces the continuous nature of the two seemingly
dichotomous schools: vicariance vs. founder speciation. Marine biogeographers have been quick to jump
on the vicariance bandwagon in the past as the primary
explanation for differentiation. Given our model of
dispersal and connectivity, we suggest that founder
speciation should be given at least an equal chance,
albeit more difficult to test, when it comes to explaining diversification in the sea.
CONCLUSIONS
As the examples above illustrate, there is considerable overlap in the spatial scale of differentiation and
of diversification between the terrestrial and marine
biota of Oceania. Although marine organisms on average are much more widespread than terrestrial ones
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
in preparation). The only known potential example of
intra-archipelagic diversification in a marine organism,
that of Conus in the Cape Verde Islands, is of a directdeveloping species group as well (Kohn and Perron,
1994).
Marine organisms that have pelagic dispersal stages,
and thus commonly enter the dispersal medium, can
nevertheless differentiate on surprisingly fine spatial
scales. That such groups can run the gamut in levels
of geographic differentiation is becoming especially
clear under molecular genetic scrutiny. Gastropods
provide an excellent illustration of the diverse scales
of differentiation possible. At one extreme, the morphologically established observation of pan-Indo-Pacific species can be substantiated genetically. Among
cypraeids, members of Lyncina and related genera tend
to show this pattern; such species among cowries appear also to have relatively long-lived pelagic larvae
(Meyer, 2002).
Other cowries are widespread and genetically well
mixed across much of the Indo-West Pacific, but show
peripatric differentiation in the most isolated regions,
such as Hawaii, or SE Polynesia. Among cowries examples include the Monetaria caputserpentis complex
(Fig. 4a). A pattern of peripheral differentiation is also
evident on a smaller scale around the periphery of the
Indo-Malayan area, where taxa widespread within this
area have given rise to peripatric species differentiating in near-Oceania. Several examples illustrate this
type of differentiation in cowries, especially among the
erroneines. Williams et al. (2002) provide examples in
the same area among littorines. Both of these types of
peripatric speciation fit the geographic pattern of classic founder speciation in terrestrial systems (cf., Mayr,
1963).
More surprising are taxa that show a mosaic pattern
of differentiation across the Indo-West Pacific, subdividing the region into a mosaic of reciprocally monophyletic taxa on a variety of spatial scales. The scale
of differentiation ranges from subbasinal (e.g., the Leporicypraea mappa-complex among cowries—Fig.
4b) to multi-archipelagic (e.g., the cowrie genus Cribrarula) to archipelagic and finer. Archipelagic-level
differentiation is shown by, for example the limpet
Scutellastra flexuosa (Fig. 4c) or the turbinid Astralium rhodostoma-complex (Geller et al., in preparation).
Such fine-scale mosaic differentiation in marine organisms is reminiscent of terrestrial endemism in Oceania, although the high single-island endemism and
great within-archipelago diversification seen in some
terrestrial taxa has no known parallels in the sea. High
marine endemism implies that such taxa can get out
into far Oceania, yet can differentiate at surprisingly
fine (for a marine organism) spatial scales. Such an
observation is consistent with founder speciation (low
Q with low l/v). Recurrent vicariance may also explain these patterns, but would require a complex scenario of shifting oceanographic and tectonic factors to
account for the mosaic distributions whose origins ap-
AND
DIVERSIFICATION
IN THE
TROPICAL PACIFIC
931
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
FIG. 4. Range of phylogeographic variation in marine gastropods. All topologies are derived from the mtCOI gene (614 bp). Bootstrap values
are indicated below branches; numbers in parentheses indicate number of individuals sequenced. A. Monetaria caputserpentis complex exhibits
a widespread IWP taxon, M. caputserpentis caputserpentis, with two peripheral Pacific endemic lineages, M. c. caputophidii (Hawaii) and M.
caputdraconis (Easter Is.). B. Leporicypraea mappa complex exhibits basinal level variation with 5 lineages partitioning the IWP. C. Scutellastra
flexuosa exhibits fine-scale archipelagic differentiation, even between close archipelagoes within French Polynesia.
of comparable body size, they also show surprisingly
high levels of differentiation in some groups. Morphological and limited genetic data indicate that some terrestrial organisms, like marine taxa, can be truly widespread and remain in genetic contact across much of
the Pacific basin. Archipelagic and island-level endemism, the hallmark of the land biota of Oceania, is
also turning up in the marine realm in several groups,
however. High levels of differentiation can be seen in
groups with direct development that appear to rely on
rare rafting events for dispersal, as well as taxa with
relatively short-lived, pelagic larvae.
The likelihood of colonization and differentiation in
both terrestrial and marine systems depend on the
same parameters: interisland distance, rate of entry
into, and leaving of, the dispersal medium, velocity of
transport, and colonization success once dispersal is
successful. There is a great deal of variation in these
parameters among taxa and systems, and this underlies
some of the differences perceived between terrestrial
932
G. PAULAY
AND
C. MEYER
and marine taxa. Variation in rate of entry into the
dispersal medium is especially important, as this parameter has a strong influence over the likelihood for
founder speciation to occur. Organisms that rarely enter the dispersal medium but survive well there are the
most likely to undergo founder speciation. The high
levels of endemicity observed in rafted, direct developing marine mollusks, and bird- and raft-dispersed
terrestrial organisms support this hypothesis.
ACKNOWLEDGMENTS
We thank Rachel Collin and Marta deMaintenon for
organizing the symposium and symposium participants
for discussion, and David Steadman and two anonymous reviewers for comments. Support by NSF (DEB9807316/0196049) and the University of Florida Research Opportunity Fund are gratefully acknowledged.
REFERENCES
Bauman, S. 1996. Diversity and decline of land snails on Rota:
Mariana Islands. Master’s Thesis, University of Guam.
Bellwood, D. R. and P. C. Wainwright. 2002. The history and biogeography of fishes on coral reefs. In P. F. Sale (ed.), Coral reef
fishes. Dynamics and diversity in a complex ecosystem, pp. 5–
32. Academic Press, San Diego.
Bennett, K. D. 1990. Milankovitch cycles and their effects on species in ecological and evolutionary time. Paleobiology 16:11–
21.
Benzie, J. A. H. and S. T. Williams. 1997. Genetic structure of giant
clam (Tridacna maxima) populations in the West Pacific is not
consistent with dispersal by present-day ocean currents. Evolution 51:768–83.
Borsa, P. and J. A. H. Benzie. 1993. Genetic relationships among
the topshell Trochus and Tectus (Prosobranchia: Trochidae). J.
Moll. Stud. 59:275–84.
Bouchet, P. and G. Poppe. 1988. Deep water volutes from the New
Caledonian region, with a discussion on biogeography. Venus
47:15–32.
Bowen, B. W., A. L. Bass, L. A. Rocha, W. S. Grant, and D. R.
Robertson. 2001. Phylogeography of the trumpetfishes (Aulos-
tomus): Ring species complex on a global scale. Evolution 55:
1029–39.
Carlquist, S. 1974. Island biology. Columbia University Press, New
York.
Carlquist, S. 1981. Chance dispersal. Amer. Sci. 69:509–16.
Carson, H. L. 1983. Chromosomal sequences and interisland colonizations in Hawaiian Drosophila. Genetics 103:465–82.
Colborn, J., R. E. Crabtree, J. B. Shaklee, E. Pfeiler, and B. W.
Bowen. 2001. The evolutionary enigma of bonefishes (Albula
spp.): Cryptic species and ancient separations in a globally distributed shorefish. Evolution 55:807–20.
Cowen, R. K., K. M. M. Lwiza, S. Sponaugle, C. B. Paris, and D.
B. Olson. 2000. Connectivity of marine populations: Open or
closed? Science 287:857–859.
Diamond, J. M., M. E. Gilpin, and E. Mayr. 1976. Species-distance
relation for birds of the Solomon Archipelago, and the paradox
of the great speciators. Proc. Nat. Acad. Sci. U.S.A. 73:2160–
2164.
Ekman, S. 1953. Zoogeography of the sea. Sidgwick and Jackson,
London.
Eldredge, L. G. 1994. Perspectives in aquatic exotic species management in the Pacific islands. Vol. 1. Introductions of commercially significant aquatic organisms to the Pacific islands. S.
Pacific Comm. Inshore Fish. Res. Proj. Tech. Doc. 7:1–127.
Eldredge, L. G. and S. E. Miller. 1995. How many species are there
in Hawaii? Bishop Mus. Occ. Pap. 41:1–18.
Emerson, W. K. 1991. First records for Cymatium mundum (Gould)
in the Eastern Pacific Ocean, with comments on the zoogeography of the tropical trans-Pacific tonnacean and non-tonnacean
prosobranch gastropods with Indo-Pacific faunal affinities in
West American Waters. Nautilus 105:62–80.
Finlay, B. J. 2002. Global dispersal of free-living microbial eukaryote species. Science 296:1061–1063.
Gagné, W. C. 1979. Canopy-associated arthropods in Acacia koa
and Metrosideros tree communities along an altitudinal transect
on Hawaii Island. Pacific Insects 21:56–82.
Hayami, I. and T. Kase. 1996. Characteristics of submarine cave
bivalves in the Northwest Pacific. Amer. Malac. Bull. 12:59–
65.
Hayami, I. and T. Kase. 1993. Submarine cave Bivalvia from the
Ryukyu Islands: Systematics and evolutionary significance.
Bull. Univ. Mus. Univ. Tokyo, 35:1–133.
Helmuth, B., R. R. Veit, and R. Holbton. 1994. Long-distance dis-
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
FIG. 5. Ranges of differentiation across Oceania in terrestrial and marine systems. As levels of interconnectedness decrease from right to left,
they lead to differentiation on increasingly finer spatial scales. (A) Intra-island endemics complete speciation within an island. (B) Intraarchipelagic endemics differentiate among islands in an archipelago. (C) Archipelagic endemics differentiate among archipelagoes. (D) Basinal
mosaics differentiate on the scale of multiple archipelagoes per lineage. (E) Marginal endemics are peripatric isolates that evolve at the
periphery of more widepspread species. (F) Extreme peripheral endemics are like marginal endemics, but differentiate only on the most remote
archipelagoes, such as Hawaii. (G) Panmictics are those that range across IWP. (H) Cosmopolitan species are those that occur outside the IWP.
Note the absence of examples of marine intra-archipelagic and intra-island differentiation.
DIVERSIFICATION
IN THE
933
Palumbi, S. R., G. Grabowsky, T. Duda, L. Geyer, and N. Tachino.
1997. Speciation and population genetic structure in tropical
Pacific sea urchins. Evolution 51:1506–17.
Paulay, G. 1985. Adaptive radiation on an isolated oceanic island:
The Cryptorhynchinae of Rapa revisited. Biol. J. Linn. Soc. 26:
95–187.
Paulay, G. 1991. Late Cenozoic sea level fluctuations and the diversity and species composition of insular shallow water marine
faunas. In E. C. Dudley (ed.), The unity of evolutionary biology.
The Proceedings of the Fourth International Congress of Systematic and Evolutionary Biology, pp. 184–93. Dioscorides,
Portland, Oregon.
Paulay, G. 1997. Diversity and distribution of reef organisms. In C.
E. Birkeland (ed.), Life and death of coral reefs, pp. 298–353.
Chapman & Hall, New York.
Paulay, G. and R. C. Preece. in prep. Sea level history, speciation,
and species turnover: Evidence from fossil bivalves on Henderson Island.
Phillips, B. F. 1981. The circulation of the southeastern Indian Ocean
and the planktonic life of the western rock lobster. Ocean. Mar.
Biol. Ann. Rev. 19:11–39.
Planes, S. 2002. Biogeography and larval dispersal inferred from
population genetic analysis. In P. F. Sale (ed.), Coral reef fishes.
Dynamics and diversity in a complex ecosystem, pp. 201–220.
Academic Press, San Diego.
Planes, S. and C. Fauvelot. 2002. Isolation by distance and vicariance drive genetic structure of a coral reef fish in the Pacific
Ocean. Evolution 56:378–99.
Randall, J. E. 1998. Zoogeography of shore fishes of the Indo-Pacific
region. Zool. Stud. 37:227–68.
Reid, D. G. and J. B. Geller. 1997. A new ovoviviparous species of
Tectarius (Gastropoda: Littorinidae) from the tropical Pacific,
with a molecular phylogeny of the genus. J. Moll. Stud. 63:
207–33.
Richmond, R. H. 1989. The effects of the El Nino/Southern Oscillation on the dispersal of corals and other marine organisms.
Elsevier Oceanogr. Ser. 52:127–140.
Rotondo, G. M., V. G. Springer, G. A. J. Scott, and S. O. Schlanger.
1981. Plate movement and island integration, a possible mechanism in the formation of endemic biotas, with special reference
to the Hawaiian Islands. Syst. Zool. 30:12–21.
Shinn, E. A., G. W. Smith, J. M. Prospero, P. Betzer, M. L. Hayes,
V. Garrison, and R. T. Barber. 2000. African dust and the demise
of Caribbean coral reefs. Geophys. Res. Lett. 27:3029–32.
Sleurs, W. J. M. and R. C. Preece. 1994. The Rissoininae (Gastropoda: Rissoidae) of the Pitcairn Islands, with the description of
two new species. J. Conch. 35:67–82.
Smith, B. D. 2003. Prosobranch gastropods of Guam. Micronesica
35. (In press)
Soot-Ryen, T. 1960. Pelecypods from Tristan da Cunha. Res. Norw.
Sci. Exp. Tristan Da Cunha 49:1–47.
Springer, V. G. 1982. Pacific plate biogeography, with special reference to shorefishes. Smiths. Contr. Zool. 367.
Stein, J. D. 1983. Insects infesting Acacia koa (Legumosae) and
Metrosideros polymorpha (Myrtaceae) in Hawaii annotated list.
Proc. Haw. Entom. Soc. 24:305–16.
Strathmann, R. R. 1985. Feeding and nonfeeding larval development
and life-history evolution in marine invertebrates. Ann. Rev.
Ecol. Syst. 16:339–61.
Swearer, S. E., J. E. Cadelle, D. W. Lea, and R. R. Warner. 1999.
Larval retention and recruitment in an island population of a
coral-reef fish. Nature 402:799–802.
Thresher, R. E. and E. B. Brothers. 1985. Reproductive ecology and
biogeography of Indo-West Pacific angelfishes (Pisces: Pomacanthidae). Evolution 39:878–87.
Vagvolgyi, J. 1975. Body size, aerial dispersal, and origin of the
Pacific land snail fauna. Syst. Zool. 24:465–88.
Veron, J. E. N. 1995. Corals in space and time. The biogeography
and evolution of the Scleractinia. Univ. New South Wales Press,
Sydney.
Wagner, W. L. 1991. Evolution of waif floras: A comparison of the
Hawaiian and Marquesan archipelagoes. In E. C. Dudley (ed.),
The unity of evolutionary biology. The proceedings of the
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013
persal of a subantarctic brooding bivalve (Gaimardia trapesina)
by kelp-rafting. Mar. Biol. 120:421–426.
Highsmith, R. C. 1985. Floating and algal rafting as potential dispersal mechanisms in brooding invertebrates. Mar. Ecol. Progr.
Ser. 25:169–79.
Hobday, A. J. 2000. Persistence and transport of fauna on drifting
kelp (Macrocystis pyrifera (L.) C. Agardh) rafts in the Southern
California Bight. J. Exp. Mar. Biol. Ecol. 253:75–96.
Holzapfel, E. P. 1978. Transoceanic airplane sampling for organisms
and particles. Pacific Insects 18:169–89.
Holzapfel, E. P., H. B. Clag, and M. L. Goff. 1978. Trapping of airborne insects on ships on the Pacific, Part 9. Pacific Insects 19:
65–90.
Jackson, G. A. and R. R. Strathmann. 1981. Larval mortality from
offshore mixing as a link between precompetent and competent
periods of development. Am. Nat. 1:16–26.
Jackson, J. B. C. 1986. Modes of dispersal of clonal benthic invertebrates: Consequences for species’ distributions and genetic
structure of local populations. Bull. Mar. Sci. 39:588–606.
Johannes, R. E. 1978. Reproductive strategies of coastal marine fishes in the tropics. Envir. Biol. Fishes 3:65–84.
Johannesson, K. 1988. The paradox of Rockall: Why is a brooding
gastropod (Littorina saxatilis) more widespread than one having
a planktonic larval dispersal stage (L. littorea). Mar. Biol. 99:
507–513.
Jokiel, P. L. 1989. Rafting reef corals and other organisms at Kwajalein Atoll. Mar. Biol. 101:483–93.
Jokiel, P. L. 1990. Long-distance dispersal by rafting: Reemergence
of an old hypothesis. Endeavour 14:66–73.
Kay, E. A. 1984. Patterns of speciation in the Indo-West Pacific. In
F. J. Radovsky, P. H. Raven, and S. H. Sohmer (eds.), Biogeography of the tropical Pacific, pp. 15–31. Bishop Mus. Spec.
Pub. No. 72, Honolulu.
Kay, E. A. and S. R. Palumbi. 1987. Endemism and evolution in
Hawaiian marine invertebrates. Trends Ecol. Evol. 2:183–86.
Kohn, A. J. and F. E. Perron. 1994. Life history and biogeography
patterns in Conus. Clarendon Press, Oxford.
Ladd, H. S. 1958. Fossil land shells from Western Pacific atolls. J.
Paleont. 32:183–98.
Leopold, E. B. 1969. Miocene pollen and spore flora of Eniwetok
Atoll, Marshall Islands. U.S. Geol. Surv. Prof. Pap. 260-II.
Lessios, H. A., B. D. Kessing, and J. S. Pearse. 2001. Population
structure and speciation in tropical seas: Global phylogeography
of the sea urchin Diadema. Evolution 55:955–75.
Lessios, H. A., B. D. Kessing, and D. R. Robertson. 1998. Massive
gene flow across the world’s most potent marine biogeographic
barrier. Proc. Roy. Soc. London 265:583–88.
Lessios, H. A., B. D. Kessing, D. R. Robertson, and G. Paulay. 1999.
Phylogeography of the pantropical sea urchin Eucidaris in relation to land barriers and ocean currents. Evolution 53:806–17.
Levitan, D. R. 1995. The ecology of fertilization in free-spawning
invertebrates. In L. R. McEdward (ed.), Ecology of marine invertebrate larvae, pp. 123–56. CRC Press, Boca Raton, FL.
Lobel, P. S. and A. R. Robinson. 1988. Larval fishes and zooplankton in a cyclonic eddy in Hawaiian waters. J. Plankton Res. 10:
1209–23.
MacArthur, R. H. and E. O. Wilson. 1967. The theory of island
biogeography. Princeton University Press, Princeton, NJ.
Mayr, E. 1963. Animal species and evolution. Belknap Press, Harvard, Cambridge, Massachusetts.
Mayr, E. and J. Diamond. 2001. The birds of Northern Melanesia:
Speciation, ecology, and biogeography. Oxford University
Press, Oxford.
McMillan, W. O., L. A. Weigt, and S. R. Palumbi. 1999. Color
pattern evolution, assortative mating, and genetic differentiation
in brightly colored butterflyfishes (Chaetodontidae). Evolution
53:247–60.
Meyer, C. P. 2002. Molecular systematics of cowries (Gastropoda,
Cypraeidae) and diversification patterns in the tropics. Biol. J.
Linn. Soc. (In press).
Meyer, C. P. and G. Paulay. 2002. Recurrent speciation at biogeographic boundaries drives diversification of reef fauna. Proc.
Nat. Acad. Sci. submitted.
TROPICAL PACIFIC
934
G. PAULAY
Fourth International Congress of Systematic and Evolutionary
Biology, pp. 267–84. Dioscorides, Portland, Oregon.
Wagner, W. L. and V. J. Funk. (eds.) 1995. Hawaiian biogeography.
Evolution on a hot spot archipelago. Smithsonian, Washington,
D.C.
Wellington, G. M. and B. C. Victor. 1989. Planktonic larval duration
of one hundred species of Pacific and Atlantic damselfishes (Pomacentridae). Mar. Biol. 101:557–67.
Whistler, W. A. 1992. Flowers of the Pacific island seashore. Isle
Botanica, Honolulu, Hawaii.
Williams, S. T. 2000. Species boundaries in the starfish genus Linckia. Mar. Biol. 136:137–48.
Williams, S. T., J. Jara, E. Gomez, and N. Knowlton. 2002. The
marini Indo-West Pacific break: Contrasting the resolving power
of mitochondrial and nuclear genes. Integr. Comp. Biol. 42:
941–952.
AND
C. MEYER
Wright, S. D., C. G. Yong, J. W. Dawson, D. J. Whittaker, and R.
C. Gardner. 2000. Riding the ice age El Nino? Pacific biogeography and evolution of Metrosideros subgenus Metrosideros
(Myrtacea) inferred from nrDNA. Proc. Nat. Acad. Sci. U.S.A.
97:4118–23.
Yoshimoto, C. M. and J. L. Gressitt. 1960. Trapping of air-borne
insects on ships on the Pacific (Part 3). Pacific Insects 2:239–
43.
Zimmerman, E. C. 1942. Distribution and origin of some eastern
oceanic insects. Am. Nat. 76:280–307.
Zimmerman, E. C. 1948. Insect of Hawaii. Vol. 1. Introduction.
Univ. Hawaii Press, Honolulu.
Zimmerman, E. C. 1976. Curculionidae captured on ships and aircraft on or over the Indian and Pacific Oceans. Pacific Insects
16:380–382.
Downloaded from http://icb.oxfordjournals.org/ at Tunghai University on June 30, 2013