o
Thi
M a r i N e P o P u l at i o N C o N N e C t i V i t y
r c
s a
o
ll
r
e
t
c
i
c
t
l
i
e h
v
e r
a
e
s b
di
s
e
t
e
ir
n p
b
u
ub
t
i
o
li
n o
s
Population Connectivity and
h
e
f a
d in
n
y p
O
o
c
r
ea
t
i
n
o
o
n o
g
Conservation raf tphhiys a, V
of r otilumcle be 20, Ny ph
Marine Biodiversity ot umbocopey mr 3, a qachuinaret, rerelpy joosturinnga, ol or of The ther mo
B y G e o F F r e y P. J o N e S ,
c
e
e
a
a
n
n
o
s i
M aya S r i N i Va S a N , a N D
g
s p
r
a
p
e
h
r
m
y S
G l e N N r . a l M a N y
i
t
o
t
c
e
i
d o
e
t
y
nl
. C
y w
o
p
i
y
t
r
h t
i
g
h
h
t 2007 b
e a
p
p
r
o
v
y The
a
An understanding of the extent
l o
f The
o
to which marine populations are
c
e
a
o
n
c
o
e
connected by larval dispersal is
g
a
r
n
a
o
p
g
h
r
y S
vital, both to comprehend past
a
p
h
o
y S
c
i
e
impacts and future prospects for
o
t
y
c
.
i
a
e
t
ll r
sustaining biodiversity. Marine popu-
y
. S
i
g
e
h
n
t
d a
s r
lations and their supporting ecosystems
e
ll c
s
e
o
r
are now subject to a multitude of threats, most notably overharvesting, pollution,
v
r
e
r
e
d. P
s
p
and climate change (Hixon et al., 2001; Jackson et al., 2001; Hutchings and Reynolds,
o
e
n
r
m
d
e
i
s
n
si
2004; Kappel, 2005; Lotze et al., 2006). The intensity and scale of anthropogenic
c
o
e t
n i
o: in
s g
impacts in the world's ocean have increased dramatically during the industrial age
r
f
a
o@t
n
t
e
(Jackson et al., 2001; Lotze et al., 2006), and these impacts are combining to acceler-
d t
o
s
.o
o c
r
ate the loss and fragmentation of important coastal marine habitats, including man-
g o
o
p
r Th e
y t
h
groves (Ellison and Farnsworth, 1996; Alongi, 2002), seagrasses (Duarte, 2002; Orth
i
s a
o
r
c
t
i
et al., 2006), kelp forests (Dayton et al., 1998; Steneck et al., 2002), and coral reefs
e
c
a
l
e f
n
o
o
g
r u
(McClanahan, 2002; Gardiner et al., 2003; Hughes, et al., 2003; Aronson and Precht,
r
a
p
s
h
e in t
y S
2006). The increasing risk of extinction in the sea is widely acknowledged (Roberts
o
e
c
a
i
e
c
t
h
and Hawkins, 1999; Powles et al., 2000; Dulvy et al., 2003; Kappel, 2005; Reynolds et
y
in
, P
g a
o
B
n
al., 2005), and the conservation of marine biodiversity has become a high priority for
d r
o
x 1931,
e
s
e
researchers and managers alike.
a
r
c
h
r
.
o
r
c
e
k
p
v
ub
ill
e
li
, MD 20849-1931,
c
a
t
i
o
n
, s
y
s
t
e
mm
a
u
t
i
c r
S
a
e
.
p
r
o
du
100
Oceanogr
O
ap
ceanogr hy
ap
c
Vo
V l.
l. 20, No. 3
o
t
i
o
n
,

---

The increasing diversity, intensity,
recent advances in our understanding of
be of limited benefit where habitat loss
and scale of human impacts on marine
larval retention and connectivity, and to
and fragmentation, pollution, and cli-
systems will likely reduce potential con-
explore their implications for evaluating
mate change are contributing to declines
nectivity among remnant populations,
threats to marine biodiversity as well as
in marine biodiversity (Allison et al.,
due to declining numbers and increas-
different management options for mini-
1998; Jameson et al., 2002; Jones et al.,
ing fragmentation. It follows that the
mizing these threats.
2004; Aronson and Precht, 2006). Also,
resilience of a species to these impacts
if MPAs simply result in a displacement
will depend to a large degree on its dis-
MariNe reSerVeS aND
of fishing effort (Hilborn et al., 2004,
persal capability. Marine larvae exhibit
BioDiVerSity ProteCtioN
2006), potential benefits of biodiversity
extremes of larval dispersal, from those
Marine reserves or no-take marine pro-
protection inside MPAs may be offset by
that travel just a few meters to others
tected areas (MPAs) are now routinely
an increase in the detrimental effects of
with the potential to disperse thousands
established both as fisheries management
fishing outside reserves.
of kilometers (Kinlan and Gaines, 2003;
tools and for biodiversity protection.
In some situations, the design of MPA
Shanks et al., 2003). Species with wide
However, while there is ample evidence
networks for fisheries management and
dispersal capabilities may be less sus-
that MPAs can provide a host of benefits
biodiversity protection may have con-
ceptible to global extinction because of
to exploited populations within their
flicting goals or outcomes. The optimal
their large ranges, multiple populations,
boundaries (Roberts and Polunin, 1991;
sizes of MPAs for biodiversity conser-
and potential for local recovery through
Jones et al., 1993; Halpern and Warner,
vation are likely to be larger than those
larval transport. On the other hand,
2002; Halpern, 2003), the effectiveness of
designed for protecting fish stocks and
invasive species and disease vectors with
MPA networks in biodiversity protection
enhancing recruitment to adjacent fished
high dispersal potential pose greater
has received much less attention. Ideally,
areas (Hastings and Botsford, 2003).
global threats to marine biodiversity. The
marine reserves should encompass rep-
Large MPAs may be ideal for biodiversity
problem we face is that the actual larval
resentative regions/habitats so as to pro-
conservation because they encompass
dispersal distances of highly overfished,
tect as much of the regional biodiversity
more species, but they may limit the
threatened, or invasive species are sel-
dom known. While new technologies are
leading to major discoveries concerning
the scales of dispersal (Thorrold et al.,
...a much greater knowledge of connectivity
2002; Palumbi et al., 2003; Levin, 2006),

is required in order to optimize strategies
as yet, this knowledge is too incomplete
to be comprehensively incorporated into

for conserving marine biodiversity.
management strategies (Sale et al., 2005).
Most approaches to the manage-
ment of marine species and ecosystems
are based on untested assumptions
as possible (e.g., Airame et al., 2003;
exploitation of fish stocks to well below
about typical larval dispersal distances.
Beger et al., 2003; Fernandes et al., 2005).
sustainable levels. Small MPAs may pro-
Understanding connectivity is critical
However, the degree to which reserves
vide a protective umbrella for the bio-
both for the design of marine reserve
achieve the goal of "protecting" species
diversity of sedentary species but are
networks to protect biodiversity and for
is uncertain. Given that the majority of
unlikely to provide an effective refuge for
the development of conservation strate-
marine species are not eaten (at least not
highly mobile exploited species (Hilborn
gies to protect species associated with
yet), closing areas to fishing or collect-
et al., 2004; Nardi et al., 2004). Increases
degrading and fragmenting seascapes.
ing does not necessarily address the pri-
in the abundance and biomass of large
The aims of this essay are to highlight
mary threats to most species. MPAs may
exploited predators or space occupiers
Oceanography September 2007
101

---

in reserves may result in the decline of
al., 2005; Purcell et al., 2006; Gerlach
duration (PLD) of 10 days (Figure 1b).
prey or inferior competitors, and thus
et al., 2007), biophysical and hydrody-
Almany et al. (2007) marked larvae of
an overall decline in biodiversity (Jones
namic models (Cowen et al., 2000, 2006;
another clownfish (A. percula) with
et al., 1993; Micheli et al., 2004). Getting
James et al., 2002), metapopulation
an 11-day PLD and a butterflyfish
the right balance between reserve design
models (Armsworth, 2002; Hastings and
(Chaetodon vagabundus) with a 38-day
for both exploitation and conservation
Botsford, 2006), and fish otolith chem-
PLD at Kimbe Island using maternally
requires a detailed understanding of lar-
istry (Swearer et al., 1999) have all indi-
transmitted barium isotopes (Figure 1c).
val dispersal patterns for the widest pos-
cated both significant local retention and
They found that 60% of the juveniles
sible range of marine species.
(a)
Cape Heussner
the potential for long-distance dispersal.
of both species were locally spawned,
20 km
Many larvae may settle much closer
returning to a reef of only 0.3 km2. In
New DiSCoVerieS oN
to home than once thought possible.
both studies, the other immigrant juKimbe Island
ve-
VariatioN iN DiSPerSal:
For example, larval marking studies at
niles must have traveled 10­20 km from
Kimbe Bay
iMPliCatioNS For
two small islands in Kimbe Bay, Papua
MaNaGeMeNt
New Guinea, (Figure 1a) show that a
GeoFFrey P. JoNeS (geoffrey.jon
Schumann Island es@jcu.
Directly tracking the movements of
large proportion of juveniles of three fish
edu.au) is Professor, School of Marine and
marine organisms through their pelagic
species are locally spawned. Jones et al.
Tropical Biology, and Chief Investigator,
larval stages is seldom possible. However,
(2005) used both direct larval mark-
Australian Research Council Centre of
recent advances in technology are pro-
ing (embryo immersion in tetracycline)
Excellence for Coral Reef Studies, James
viding new insights into the extent of
(a)
and genet
Cape Heussner
(b)
ic parentage analysis to e 20 km
xam-
Cook University, Townsville, Queensland,
marine larval dispersal, indicating more
ine retention of clownfish (Amphiprion
Australia. Maya SriNiVaSaN and
local retention and a greater variation
polymnus) larvae at Sch
Kimbe Island
umann Island.
GleNN r. alMaNy are Postdoctoral
in dispersal distances than once appre-
They showed that a significant propor-
Researchers, Australian Research Council
Kimbe Bay
ciated. Direct larval marking (Jones et
tion (~ 30%) of larvae settling at the
Centre of Excellence for Cora
Schumann l Reef Studies,
Island
al., 1999, 2005; Almany et al., 2007),
island were within a few hundred meters
James Cook University, Townsville,
Schumann Island
advanced genetic techniques (Jones et
of their parents, despite a pelagic larval
Queensland, Australia.
200 m
Amphiprion polymnus
(a)
Cape Heussner
(b)
(c)
20 km
Kimbe Island
Amphiprion percula
Kimbe Bay
Schumann
Kimbe
Island
Island
Schumann Island
200 m
Amphiprion polymnus
200 m
Chaetodon vagabundus
(b)
Figure 1. (a) Kimbe Bay, Papua New Guinea. Direct larval marking studies at two islands (Schumann and Kimbe) s
(c)
how that many coral reef fish larvae settle
much closer to home than previously thought. (b) Schumann island. larval marking via tetracycline immersion of embryos and genetic parentage analysis dem-
onstrated ~ 30% self-recruitment in a population of the panda clownfish (Amphiprion polymnus), which have a 10-day pelagic larval duration. (c) Kimbe island.
larval marking via maternal transmission of stable barium isotopes demonstrated ~ 60% self-recruitment in both the orange clownfish (A. percula; 11-day
Amphiprion percula
pelagic larval duration) and the vagabond butterflyfish (Chaetodon vaganbundus; 38-day pelagic larval duration). Percent of self-recruitment was defined as the
proportion of th
Schumann e recruitment to a population that was a product of that population. Kimbe
Island
Island
102
Oceanography Vol. 20, No. 3
200 m
Amphiprion polymnus
200 m
Chaetodon vagabundus
(c)
Amphiprion percula
Kimbe
Island
200 m
Chaetodon vagabundus

---

the nearest adjacent populations. Hence,
relate closely to larval dispersal potential
of suitable habitat (Figure 2d,e). A wide
reef fishes from semi-isolated subpopu-
(Figure 2a­c). The higher the variance
variance in dispersal kernels, including
lations with a range of PLDs can exhibit
in dispersal, the greater the realized con-
significant levels of self-recruitment, will
a wide variance and multimodal patterns
nectivity within a species' population
result in multimodal patterns of realized
in dispersal distances.
range (Figure 2c). More often than not,
dispersal that will maximize the persis-
Realized dispersal distances may
suitable habitat is discontinuous, and
tence of the metapopulation (Figure 1f)
be explained by a combination of the
marine (meta) populations are made up
(Hastings and Botsford, 2006).
actual dispersal potential (a dispersal
of many subpopulations that are linked
Realized dispersal patterns have
kernel based on how far larvae have the
to an unknown degree and distance by
important strategy implications for bio-
intrinsic ability to disperse) and the dis-
larval dispersal (Figure 2d­f) (Kritzer
diversity conservation, such as optimal
tribution of suitable habitat (Figure 2).
and Sale, 2004). Greater dispersal abili-
reserve size and spacing. As a general
Where suitable habitat is continuous
ties do not necessarily increase realized
rule, an increasing range of effective
and not saturated, the observed connec-
dispersal unless they are sufficient to
dispersal will require larger MPAs to
tivity through much of the range may
bridge the gaps between isolated patches
achieve recruitment benefits within
C O N T I N U O U S H A B I T A T
(a) Low
(b) Medium
(c) High
R1 R2
R1
R2
R1
R2
dispersal potential
realized dispersal
R1 = reserve 1
R2 = reserve 2
r o p o r t i o n
P
F R A G M E N T E D H A B I T A T
(d) Low
(e) Medium
(f) High
R1
R2
R1 R2
R1
R2
r o p o r t i o n
P
Distance
Distance
Distance
Figure 2. Schematic representation of the influence of dispersal potential and habitat fragmentation on realized dispersal and
optimal reserve size and spacing. where habitat (orange horizontal bars) is continuous, realized dispersal reflects dispersal poten-
tial, and optimal reserve (blue vertical bars) size and spacing increase with variance in realized dispersal (a­c). where habitat
is discontinuous, realized dispersal reflects a combination of dispersal potential and habitat fragment size and spacing (d­f).
realized dispersal only occurs if dispersal potential is sufficient to bridge gaps between patches of habitat. optimal reserve size
and spacing will be constrained by the size and spacing of habitat fragments.
Oceanography September 2007
103

---

boundaries and will allow a greater spac-
representation of species inside reserves
persal distances may under-represent
ing of MPAs to achieve recruitment ben-
(e.g., Diamond, 1975; Simberloff, 1988;
the potential spread of larvae from
efits beyond boundaries or connectivity
Margules et al., 1982; Pressey et al.,
source populations.
among MPAs (Figure 2a­c). Where habi-
1993). These approaches are increasingly
If we explicitly incorporate larval
tat is discontinuous, the optimal reserve
being applied to MPA design, particu-
retention and connectivity into the
design of MPA networks for biodiversity
conservation, then optimal outcomes
depend on whether the goal is to maxi-
mize benefits within MPAs, beyond their
Critical questions, such as how population
boundaries, or a balance between these


connectivity will be influenced by the
objectives (Figure 3). Individual, isolated
MPAs cannot function as sanctuaries for
increasing loss and fragmentation of marine
threatened species within their boundar-

habitats, are only beginning to be answered.
ies unless the populations are big enough
and there is sufficient self-recruitment
to ensure persistence. Also, they cannot
function as sources to repopulate locally
size and spacing to ensure retention
larly in relation to site selection (Turpie
extinct populations beyond their bound-
and connectivity may be constrained by
et al., 2000; Roberts et al., 2002, 2003;
aries unless there is sufficient longer-
the size and spacing of habitat patches.
Beger et al., 2003; Fox and Beckley, 2005;
distance dispersal to do so. Given the
Subpopulations that are not connected
Fernandes et al., 2005).
likely variance in dispersal (based on the
because of limited dispersal capabil-
There are relatively few analyses that
information gathered so far), it is highly
ity, geographic isolation, or increasing
explicitly incorporate larval connectiv-
probable that reserve size and spacing
habitat fragmentation (Figure 1d,e), will
ity and population persistence in reserve
can be adjusted to achieve both of these
take high conservation priority because
design, and those that do primarily focus
goals (Halpern and Warner, 2002; Jones
of their reduced ability to recover from
on reserves for fisheries management
et al., 2005; Almany et al., 2007).
local depletion or habitat degradation
(Roberts, 1997; Botsford et al., 2001;
(Roberts et al., 2006).
Lockwood et al., 2002; but see Hastings
1. reserve Size: large Versus Small
and Botsford, 2003). These models set
The pros and cons of small versus large
CoNNeCtiVity aND the
out to predict the optimal size, spacing,
marine reserves have received increased
DeSiGN oF MariNe reSerVe
or cumulative area under different lev-
attention (Halpern, 2003; Roberts et
NetworKS to ProteCt
els of dispersal. However, until realized
al., 2003; Palumbi, 2004; Sale et al.,
BioDiVerSity
dispersal distances are confirmed for a
2005). There is substantial variation in
The design of MPA networks, includ-
wide variety of species, application of
the size of existing no-take MPAs, from
ing the size of individual reserves, the
these models is limited. Typical larval
< 1 km2 to > 1000 km2 (Halpern 2003),
number of reserves, cumulative total
dispersal distances, often inferred from
but which is best? Even the smallest
reserve area, the trade-off between a few
indirect information such as PLD or
reserves monitored appear to have local
large or several small reserves, and the
measures of genetic distance, have been
benefits in terms of increases in fished
spacing and locations of reserves, can be
used to argue for the optimal size and
species (e.g., Roberts and Hawkins, 1997;
varied to achieve different conservation
spacing of marine reserves for particu-
Francour et al., 2001). However, in MPA
goals. Most recommendations for the
lar marine taxa (e.g., Sala et al., 2002;
networks, larger reserves have advan-
design of MPA networks are based on
Shanks et al., 2003; Palumbi, 2004; Mora
tages in terms of protecting significantly
theory and practices that maximize the
et al., 2006). However, "typical" dis-
larger populations (Halpern, 2003) and
104
Oceanography Vol. 20, No. 3

---

enforcing compliance (Kritzer, 2004;
recruitment benefits either inside or
dramatically with reserve size. However,
Little et al., 2005).
beyond their boundaries (Figure 3a).
little benefit may be achieved by increas-
In terms of connectivity, extremely
With significant local self-recruitment,
ing reserve size above the level at which
small reserves will provide minimal
the benefit inside reserves should rise
persistent populations can be achieved.
Lockwood et al. (2002) calculated that
to have a persistent population in an
isolated reserve, reserve size should be
about two times the mean dispersal dis-
(a) Reserve size
(b) Reserve number
tance to ensure that it is substantially
Inside
Inside
Better
self-recruiting. Recruitment subsidies
beyond boundaries will increase in pro-
Outside
Outside
portion to the perimeter of the reserve,
so a bigger reserve will always be better.
Few have attempted to put actual
dimensions on optimal reserve size
orse
W
based on mean larval dispersal estimates.
Small
Large
One
Many
Shanks et al. (2003) argue that marine
reserves should be 4­6 km in diameter
(c) Total reserve area
(d) Single large or several small
to be large enough to contain the larvae
Inside
Inside
of short-distance dispersers. However,
Better
Palumbi (2004) argues that greater
Outside
Outside
variation in reserve sizes (1­100 km in
diameter) is necessary due to the large
variance in dispersal distances across
taxa. Laurel and Bradbury (2006) sug-
orse
W
gest larval dispersal distances in fishes
0%
100%
Several
Single
increase substantially with latitude,
Percent total area
small
large
suggesting marine reserves in temper-
(e) Reserve spacing
(f) Reserve location
ate climes should be larger than those
R e s e r v e N e t w o r k D e s i g n Q u a l i t y
in the tropics.
Better
Inside
Inside
2. reserve Number
Assuming reserve sizes are the same, for
Outside
Outside
conservation purposes, it should also be
better to have as many MPAs as possible
orse
(Figure 3b). However, self-sustaining
W Near
Far
Source
Sink
populations inside reserves may be
Isolated
Not isolated
achieved by a relatively small number
R e s e r v e P a r a m e t e r
of reserves, while recruitment subsi-
Figure 3. Design guidelines for marine protected area networks based on connectivity and achiev-
dies outside reserves and connectivity
ing optimal outcomes for recruitment benefits inside (green line) and outside (blue line) boundar-
ies. (a) individual reserve size. (b) Number of reserves. (c) total reserve area (as percentage of total
among reserves should increase in pro-
habitat). (d) Single large versus several small. (e) reserve spacing. (f) reserve location (source versus
portion to reserve number (Figure 3b)
sink, isolated versus clustered). These guidelines are based on an assumption of a wide variance in
dispersal ranges, including a significant level of localized recruitment.
(Roberts et al., 2006).
Oceanography September 2007
105

---

...total area protected is probably
best adjusted to maximize recruitment
subsidies beyond boundaries.
3. total reserve area
of exchange may be unnecessary for
4. Single large or Several Small:
The conservation benefits of MPAs inside
persistence inside reserves. Hence, total
the SloSS Debate
and outside reserves vary in relation to
area protected is probably best adjusted
The decision over whether a fixed pro-
the proportion of the total area to be pro-
to maximize recruitment subsidies
portion of the total reserve area should
tected (Hastings and Botsford, 2003). At
beyond boundaries.
be divided into either a few large or
the lower extreme, with 0% of the habitat
Few workers have put a figure on
many small reserves is an important
inside MPAs, obviously there can be no
the proportion of the total habitat
practical decision faced by managers.
benefits either within or beyond bound-
area inside reserves that is required for
There has been conflicting opinion over
aries. However, if 100% of the available
biodiversity conservation, although
which strategy maximizes the number
habitat were protected, this would maxi-
recommendations ranging between
of species inside reserves, the so-called
mize the benefits within boundaries but
20 and 50% have been published (e.g.,
SLOSS debate (Single Large or Several
provide no benefits beyond boundaries
Botsford et al., 2001, > 35%; Sala et
Small) (Diamond, 1975; Simberloff,
(Figure 3c). While this could be seen as
al., 2002, 40%; Airame et al., 2003,
1988). Whether or not one of these alter-
a fundamental conflict between design-
30­50%; Gell and Roberts, 2003,
natives is better than the other depends
ing MPAs for biodiversity protection and
10­20%; Halpern et al., 2004, < 50%).
on the degree to which small reserves
for sustainable fishing, increasing evi-
Theoretically, the total area needed will
represent nested subsamples of species
dence for localized recruitment suggests
increase with decreasing connectivity
from larger reserves (Lomolino, 1994;
that a relatively small total area will sup-
(Roberts et al., 2006). Small, unique hab-
Worthen, 1996). The few published
port self-sustaining populations. While
itats or endangered species will require
comparisons of small and large marine
inter-reserve connectivity will increase
protection of 100% of the habitat or the
reserves suggest that it may not make
with the total proportion protected
area supporting the population.
much difference to the number of spe-
(Roberts et al., 2006), such high levels
cies protected (McNeill and Fairweather,
106
Oceanography Vol. 20, No. 3

---

1993; Stockhausen and Lipcius, 2001).
There are a few specific recommen-
outside reserves. Sinks (places that rely
In terms of larval dispersal and bio-
dations about MPA spacing based on
solely on larvae imported from upstream
diversity protection, whether a single
generalizations about dispersal distances.
for their persistence) will receive little
large or several small is better again
Shanks et al. (2003) recommend a spac-
benefit from protection and should be
depends on the goal. A single large MPA
ing of 10­20 km for species with typi-
resilient to "recruitment" overfishing
may maximize population persistence
cal pelagic larval durations to promote
(Roberts, 1997).
through self-recruitment, but several
connectivity among adjacent reserves.
It has been suggested further that
smaller MPAs will maximize recruit-
However, Palumbi (2004) argues for
potential source reefs that are resistant
ment beyond boundaries (Hastings and
more variation in spacing (10­200 km)
to perturbations should have the high-
Botsford, 2003) (Figure 3d). However, if
to reflect the likely variation in the dis-
est MPA value, as they alone can source
there is a high level of local retention, it
persal abilities of fish and invertebrates.
the recovery of damaged habitats (Salm
is unlikely that it would make any differ-
Kaplan and Botsford (2005) show that
et al., 2006). However, while protecting
ence over a broad range of reserve size/
variable spacing is better than fixed spac-
source populations is critical, locating
number combinations. Several medium-
ing when there are several small reserves
them is a difficult matter. Populations
sized reserves, appropriately spaced, are
rather than few large reserves.
are usually classified as sources on the
likely to maximize recruitment subsidies
basis of hydrodynamic models (Roberts,
from MPAs, simply because the mag-
6. reserve location
1997; Bode et al., 2006). For example,
nitude of recruitment subsidies should
Discussions about reserve location in
Bode et al. (2006) predicted that popula-
increase as a function of the sum of the
relation to connectivity center on three
tions on the northern Great Barrier Reef
circumferences of all reserves (Figure 3d).
main issues: (1) protecting source popu-
in Australia are self-sustaining source
lations, (2) protecting isolated popu-
reefs, while those in the south may be
5. reserve Spacing
lations, and (3) protecting spawning
reliant on the north for their persistence.
As long as individual reserves are suf-
aggregation sites. Larval "sources," if they
However, models that incorporate larval
ficiently large to be substantially self-
can be identified, make better reserves
behavior and demography suggest that
sustaining, the spacing of reserves may
have little effect on reserve population
persistence (Figure 3e). Persistence of
populations inside reserves may be
...the increasing evidence for local retention
marginally better if reserves are close
together, because of increased recruit-
of larvae argues that biophysical models must
ment subsidies from other reserves
be able to predict patterns of dispersal
(Roberts et al., 2006). However, spac-

and connectivity at fine spatial scales.
ing will be much more critical to maxi-
mizing recruitment subsidies outside
reserves. Again, there is likely to be an
optimal spacing, because if they are too
close, this effectively restricts the area in
than "sink" populations (Roberts, 1997),
the scale at which marine populations
between, and if they are too far apart,
regardless of whether the priority is bio-
act as sources may be more limited than
they will not receive recruits from other
diversity conservation or fisheries man-
once thought (Cowen et al., 2006).
reserves. As a general rule, the lower the
agement. Sources must be prioritized
Isolated islands or habitats have a high
effective dispersal, the closer MPAs will
because they (a) must be self-recruiting
conservation priority because they often
have to be to provide benefits to unpro-
to persist, and (b) will provide an above-
have unique assemblages and popula-
tected areas (see Figure 2a-c).
average recruitment subsidy to areas
tions that are disconnected from all
Oceanography September 2007
107

---

others (Jones et al., 2002; Roberts et al.,
why MariNe reSerVeS are
within their boundaries without restrict-
2006; Perez-Ruzafa et al., 2006). Sources
iNSuFFiCieNt to CoNSerVe
ing access to an unreasonably large total
of replenishment for populations along
MariNe BioDiVerSity
reserve area. Hence, such species will
coastlines or archipelagos may be greater,
It is increasingly appreciated that marine
always require a safety net of manage-
affording them greater resilience. Given
reserves "are necessary, but not suf-
ment actions outside MPAs (Jones et al.,
that isolated islands rely on self-recruit-
ficient" to manage exploited species or
2002). Overfished species may also need
ment for their persistence, protecting a
protect marine biodiversity (Allison et
to be managed outside protected areas,
relatively high proportion of the repro-
al., 1998; Jameson et al., 2002; Aronson
both to control overall fishing effort
ductive population may be necessary to
and Precht, 2006). A comprehensive
(Hilborn et al., 2004, 2006) and to ensure
avoid potential extinction.
management plan must involve mini-
an adequate source of larvae for all
Many large marine organisms gather
mizing human impacts both inside and
unprotected areas (Almany et al., 2007).
at widely separated, but spatially predict-
outside MPAs. Individual populations
able, spawning aggregation or breed-
and local biodiversity inside MPAs can
CoNCluSioNS
ing sites (Vincent and Sadovy, 1998;
be threatened by the buildup of large
Clearly, a much greater knowledge of
Claydon, 2005). As these areas encom-
predators (Jones et al., 1993; Micheli et
connectivity is required in order to opti-
pass the main sources of larvae for local
al., 2004) and extrinsic sources of envi-
mize strategies for conserving marine
or regional replenishment of popula-
ronmental degradation, such as sedi-
biodiversity. To our knowledge, empiri-
tions, their protection is of paramount
mentation and global climate change
cal estimates of connectivity have never
importance (Roberts et al., 2006).
(Allison et al., 1998; Rogers and Beets,
been incorporated into the design and
However, where and how far larvae go
2001; Jones et al., 2004). Although some
implementation of an MPA network.
from particular aggregation sites remains
exploited populations in degraded MPAs
While improved estimates of connec-
a mystery. On coral reefs, populations
can benefit from protection (Hawkins
tivity may not bring about changes to
returning to well-known aggregation
et al., 2006), the majority of unexploited
existing MPA networks, it will help us
sites are generally in decline, and recov-
species may not. An idealized MPA
to understand how they operate and
to identify any deficiencies. Hopefully,
future MPA designs will explicitly take
into account larval sources, optimal MPA
sizes for animal sanctuaries, and optimal
the increasing risk of extinction in the sea is
spacing to maximize recruitment sub-

widely acknowledged..., and the conservation
sidies in non-MPA areas. While there is
increasing direct evidence that MPAs can
of marine biodiversity has become a high
provide benefits within and beyond their

priority for researchers and managers alike.
boundaries, such evidence is limited.
There is also information to suggest that
extrinsic disturbances from beyond their
boundaries can negate these benefits.
ery at locally extinct aggregation sites is
network based on reliable estimates of
Critical questions, such as how popula-
limited (Sadovy, 1993; Sala et al., 2001).
dispersal may change if ocean warming
tion connectivity will be influenced by
Local management of individual spawn-
and climatic conditions alter patterns
the increasing loss and fragmentation
ing aggregation sites may be critical
of circulation and larval development
of marine habitats, are only beginning
if particular sites are not replenished
(Munday et al., 2007). MPA networks
to be answered.
from other sources.
cannot be designed to encompass all
It is impractical to suggest that we
rare and potentially threatened species
will ever have detailed empirical data on
108
Oceanography Vol. 20, No. 3

---

larval dispersal for all marine species.
national cooperation, especially among
the world's mangrove forests. Environmental
Conservation 29:331­349.
Ultimately, we must rely on biophysical
small countries.
Armsworth, P.R. 2002. Recruitment limitation, popu-
models or other proxies of dispersal that
The large range in dispersal distances
lation regulation, and larval connectivity in reef
fish metapopulations. Ecology 83:1,092­1,104.
can be applied across a range of species
that has been observed, even within spe-
Aronson, R.B., and W.F. Precht. 2006. Conservation,
with similar life history characteristics.
cies, is probably no accident. Populations
precaution, and Caribbean reefs. Coral Reefs
However, the increasing evidence for
that are capable of persisting through
25:441­450.
Beger, M., G.P. Jones, and P.L. Munday. 2003.
local retention of larvae argues that bio-
either mechanisms of self-recruitment
Conservation of coral reef biodiversity: A compari-
physical models must be able to predict
or from larval supply from upstream
son of reserve selection procedures for corals and
fishes. Biological Conservation 111:53­62.
patterns of dispersal and connectivity
populations will be resilient to the widest
Bode, M., L. Bode, and P.R. Armsworth. 2006. Larval
at fine spatial scales. Cross-validation of
range of disturbances. This variance may
dispersal reveals regional sources and sinks in the
different techniques for estimating dis-
be sufficient to ensure populations can
Great Barrier Reef. Marine Ecology Progress Series
308:17­25.
persal and testing predictions using lar-
benefit from a wide range of MPA sizes
Botsford, L.W., A. Hastings, and S.D. Gaines. 2001.
val marking studies will be necessary to
and configurations, and indeed that may
Dependence of sustainability on the configuration
of marine reserves and larval dispersal distance.
increase the reliability of these models.
be the best strategy. While it is unlikely
Ecology Letters 4:144­150.
To a large extent, managing marine
that we can hope to get the manage-
Claydon, J. 2005. Spawning aggregations of coral reef
fishes: Characteristics, hypotheses, threats and
biodiversity on the basis of information
ment of all marine species completely
management. Oceanography and Marine Biology:
on connectivity will be a balancing act,
right, we can take comfort in the fact
An Annual Review 42:265­301.
exploiting the advantages and minimiz-
that it would also be impossible to get
Cowen, R.K., K.M.M. Lwiza, S. Sponaugle, C.B. Paris,
and D.B Olson. 2000. Connectivity of marine pop-
ing the disadvantages of the extremes
it completely wrong.
ulations: Open or closed? Science 287:857­859.
of dispersal. For species or habitats
Cowen, R.K., C.B. Paris, and A. Srinivasan. 2006.
Scaling of connectivity in marine populations.
with low connectivity, local protection
aCKNowleDGeMeNtS
Science 311:522­527.
of populations will be enhanced, small
This manuscript has benefited from
Dayton, P.K., M.J. Tegner, P.B. Edwards, and K.L.
Riser. 1998. Sliding baselines, ghosts, and reduced
reserves will be effective, and there will
discussions with members of the ARC
expectations in kelp forest communities. Ecological
be demonstrable benefits for local fisher-
Centre of Excellence for Coral Reef
Applications 8:309­322.
ies and less need for international coop-
Studies Connectivity Program, and
Diamond, J.M. 1975. The island dilemma: Lessons of
modern biogeographic studies for the design of
eration in management. Low-dispersal
members of the Coral Reef Targeted
nature reserves. Biological Conservation 7:129­146.
regions may also be more impervious to
Research and Capacity Building (CRTR)
Duarte, C.M. 2002. The future of seagrass meadows.
Environmental Conservation 29:192­206.
invasive species and pests. However, with
Program, Connectivity Working Group
Dulvy N.K., Y. Sadovy, and J.D. Reynolds. 2003.
low connectivity comes the greater threat
(CWG). Special thanks to S. Thorrold
Extinction vulnerability in marine populations.
Fish and Fisheries 4:25­64.
of local and global extinction, and low
and an anonymous reviewer for their
Ellison, A.M., and E.J. Farnsworth. 2006.
recruitment subsidies from MPAs or ref-
contributions to the final draft.
Anthropogenic disturbance of Caribbean man-
uge populations. The total reserve area
grove ecosystems: Past impacts, present trends, and
future predictions. Biotropica 28:549­565.
may have to be large and reserves closely
reFereNCeS
Fernandes, L., J. Day, A. Lewis, S. Slegers, B. Kerrigan,
spaced to maximize their benefits.
Airame, S., J.E. Dugan, K.D. Lafferty, H. Leslie, D.A.
D. Breen, D. Cameron, B. Jago, J. Hall, D. Lowe, and
McArdle, and R.R. Warner. 2003. Applying ecologi-
others. 2005. Establishing representative no-take
On the other hand, with high larval
cal criteria to marine reserve design: A case study
areas in the Great Barrier Reef: Large-scale imple-
dispersal and connectivity, subpopula-
from the California Channel Islands. Ecological
mentation of theory on marine protected areas.
tions will be resilient to local extinctions
Applications 13:S170­S184.
Conservation Biology 19:1,733­1,744.
Allison, G.W., J. Lubchenco, and M.H. Carr. 1998.
Fox, N.J., and L.E. Beckley. 2005. Priority areas for
and recruitment subsidies from MPAs,
Marine reserves are necessary but not sufficient
conservation of Western Australian coastal fishes: A
and remnant populations will be greater.
for marine conservation. Ecological Applications
comparison of hotspot, biogeographical and com-
8:S79­S92.
plementarity approaches. Biological Conservation
However, total protection of a popula-
Almany, G.R., M.L. Berumen, S.R. Thorrold, S. Planes,
125:399­410.
tion or species may be impossible, the
and G.P. Jones. 2007. Local replenishment of coral
Francour, P., J.G. Harmelin, D. Pollard, and S.
reef fish populations in a marine reserve. Science
Sartoretto. 2001. A review of marine protected
value of small reserves may be limited,
316:742­744.
areas in the northwestern Mediterranean region:
and there will be a premium on inter-
Alongi, D.M. 2002. Present state and future of
Siting, usage, zonation and management. Aquatic
Oceanography September 2007
109

---

Conservation: Marine and Freshwater Ecosystems
and others. 2001. Historical overfishing and the
2005. Effects of size and fragmentation of marine
11:155­188.
recent collapse of coastal ecosystems. Science
reserves and fisher infringement on the catch
Gardiner, T.A., I.M. Cote, J.A. Gill, A. Grant, and A.R.
293:629­638.
and biomass of coral trout, Plectropomus leopar-
Watkinson. 2003. Long-term region-wide declines
James, M.K., P.R. Armsworth, L.B. Mason, and L. Bode.
dus, on the Great Barrier Reef, Australia. Fisheries
in Caribbean corals. Science 301:958­960.
2002. The structure of reef fish metapopulations:
Management and Ecology 12:177­188.
Gell, F.R., and C.M. Roberts. 2003. Benefits beyond
Modelling larval dispersal and retention patterns.
Lockwood, D.R., A. Hastings, Botsford, L.W. 2002. The
boundaries: The fishery effects of marine reserves.
Proceedings of the Royal Society of London, Series B
effects of dispersal patterns on marine reserves:
Trends in Ecology and Evolution 18:448­455.
269: 2,079­2,086.
Does the tail wag the dog? Theoretical Population
Gerlach, G., J. Atema, M.J. Kingsford, K.P. Black, and V.
Jameson, S.C., M.H. Tupper, and J.M. Ridley. 2002.
Biology 61:297­309.
Miller-Sims. 2007. Smelling home can prevent dis-
The three screen doors: Can marine "protected"
Lomolino, M.V. 1994. An evaluation of alternative
persal of reef fish larvae. Proceedings of the National
areas be effective? Marine Pollution Bulletin
strategies for building networks of nature reserves.
Academy of Sciences of the United States of America
44:1,177­1,183.
Biological Conservation 69:243­249.
104:858­863.
Jones, G.P., R. Cole, and C.N. Battershill. 1993.
Lotze, H.K., H.S. Lenihan, B.J. Bourque, R.H.
Halpern, B.S. 2003. The impact of marine reserves:
Marine reserves: Do they work? Pp. 29­45 in
Bradbury, R.G. Cooke, M.C. Kay, S.M. Kidwell,
Do reserves work and does reserve size matter?
The Ecology of Temperate Reefs: Proceedings of the
M.X. Kirby, C.H. Peterson, and J.B.C. Jackson.
Ecological Applications 13:S117­S137.
Second International Temperate Reef Symposium,
2006. Depletion, degradation, and recovery
Halpern, B.S., S.D. Gaines, and R.R. Warner. 2004.
Auckland. NIWA Publications, Wellington. ISBN
potential of estuaries and coastal seas. Science
Confounding effects of the export of production
0-478-08327-0.
312:1,806­1,809.
and the displacement of fishing effort from marine
Jones, G.P., M.I. McCormick, M. Srinivasan, and J.V.
Margules, C., A.J. Higgs, and R.W. Rafe. 1982. Modern
reserves. Ecological Applications 14:1,248­1,256.
Eagle. 2004. Coral decline threatens fish biodiver-
biogeographic theory: Are there any lessons for
Halpern, B.S., and R.R. Warner. 2002. Marine reserves
sity in marine reserves. Proceedings of the National
nature reserve design? Biological Conservation
have rapid and lasting effects Ecology Letters
Academy of Sciences of the United States of America
24:115­128.
5:361­366.
101:8,251­8,253.
McClanahan, T.R. 2002. The near future of coral reefs.
Hastings, A., and L.W. Botsford. 2003. Comparing
Jones, G.P., M.J. Milicich, M.J. Emslie, and C. Lunow.
Environmental Conservation 29:460­483.
designs of marine reserves for fisheries and for bio-
1999. Self-recruitment in a coral reef fish popula-
McNeill, S.E., and Fairweather, P.G. 1993. Single
diversity. Ecological Applications 13:S65­S70.
tion. Nature 402:802­804.
large or several small marine reserves? An experi-
Hastings, A., and L.W. Botsford. 2006. Persistence of
Jones, G.P., P.L. Munday, and M.J. Caley. 2002. Rarity
mental approach with seagrass fauna. Journal of
spatial populations depends on returning home.
in coral reef fish communities. Pp. 81­101, in Coral
Biogeography 20:429­440.
Proceedings of the National Academy of Sciences of
Reef Fishes: Dynamics and diversity in a complex
Micheli, F., B.S. Halpern, L.W. Botsford, and R.R.
the United States of America 103:6,067­6,072.
ecosystem. P.F. Sale, ed., Academic Press, San Diego.
Warner. 2004. Trajectories and correlates of
Hawkins, J.P., C.M. Roberts, C. Dytham, C. Schelten,
Jones, G.P., S. Planes, and S.R. Thorrold. 2005. Coral
community change in no-take marine reserves.
and M.M. Nugues. 2006. Effects of habitat char-
reef fish larvae settle close to home. Current Biology
Ecological Applications 14:1,709­1,723.
acteristics and sedimentation on performance of
15:1,314­1,318.
Mora, C., S.Andréfouët, M.J. Costello, C. Kranenburg,
marine reserves in St. Lucia. Biological Conservation
Kaplan, D.M., and L.W. Botsford. 2005. Effects of
A. Rollo, J. Veron, K.J. Gaston, and R.A. Myers.
127:487­499.
variability in spacing of coastal marine reserves on
2006. Coral reefs and the global network of marine
Hilborn. R., F. Micheli, and G.A. De Leo. 2006.
fisheries yield and sustainability. Canadian Journal
protected areas. Science 312:1,750­1,751.
Integrating marine protected areas with catch reg-
of Fisheries and Aquatic Sciences 62:905­912.
Munday, P.L., G.P. Jones, M. Sheaves, A.J. Williams,
ulation. Canadian Journal of Fisheries and Aquatic
Kappel, C.V. 2005. Losing pieces of the puzzle: Threats
and G. Goby. 2007. Vulnerability of fishes of the
Sciences 63:642­649.
to marine, estuarine and diadromous species.
Great Barrier Reef to climate change. Pp. 357­391
Hilborn, R., K. Stokes, J.J. Maguire, T. Smith, L.W.
Frontiers in Ecology and Environment 3:275­282.
in Climate Change and the Great Barrier Reef. P.
Botsford, M. Mangel, J. Orensanz, A. Parma, J. Rice,
Kinlan, B.P., and S.D. Gaines. 2003. Propagule dispersal
Marshall and J. Johnson, eds, Great Barrier Reef
J. Bell, and others. 2004. When can marine reserves
in marine and terrestrial environments: A commu-
Marine Park Authority.
improve fisheries management? Ocean & Coastal
nity perspective. Ecology 84:2,007­2,020.
Nardi, K., G.P. Jones, M.J. Moran, and Y.W. Cheng.
Management 47:197­205.
Kritzer, J.P. 2004. Effects of noncompliance on the suc-
2004. Contrasting effects of marine protected areas
Hixon, M.A., P.D. Boersma, M.L. Hunter Jr., F. Micheli,
cess of alternative designs of marine protected area
on the abundances of two exploited reef fishes
E.A. Norse, H.P. Possingham, and P.V.R. Snelgrove.
networks for conservation and fisheries manage-
at the sub-tropical Houtman Abrolhos Islands,
2001. Oceans at risk. Pp. 125­154 in Conservation
ment. Conservation Biology 18:1,021­1,031.
Western Australia. Environmental Conservation
Biology: Research Priorities for the Next Decade. M.
Kritzer, J.P., and P.F. Sale. 2004. Metapopulation ecol-
31:160­169.
Soulé and G. Orians eds, Island Press, Washington,
ogy in the sea: From Levins' model to marine
Orth, R.J., T.J.B. Carruthers, W.C. Dennison, C.M.
DC.
ecology and fisheries science. Fish and Fisheries
Duarte, J.W. Fourqurean, W. James, H.K. Heck Jr,
Hughes, T.P., A.H. Baird, D.R. Bellwood, M. Card,
5:131­140.
A.R. Hughes, G.A. Kendrick, W.J. Kenworthy, and
S.R. Connolly, C. Folke, R. Grosberg, O. Hoegh-
Laurel, B.J., and I.R. Bradbury. 2006. "Big" concerns
others. 2006. A global crisis for seagrass ecosys-
Guldberg, J.B.C. Jackson, J. Kleypas, and others.
with high latitude marine protected areas (MPAs):
tems. Bioscience 56:987­996.
2003. Climate change, human impacts and the
Trends in connectivity and MPA size. Canadian
Palumbi, S.R. 2004. Marine reserves and ocean neigh-
resilience of coral reefs. Science 301:929­933.
Journal of Fisheries and Aquatic Science 63:
borhoods: The spatial scale of marine popula-
Hutchings, J.A., and J.D. Reynolds. 2004. Marine fish
2,603­2,607.
tions and their management. Annual Review of
population collapses: Consequences for recovery
Levin, L.A. 2006. Recent progress in understanding
Environment and Resources 29:31­68.
and extinction risk. Bioscience 54:297­309.
larval dispersal: New directions and digressions.
Palumbi, S.R., S.D. Gaines, H. Leslie, and R.R.
Jackson, J.B.C., M.X. Kirby, W.H. Berger, K.A.
Integrative and Comparative Biology 46:282­297.
Warner. 2003. New wave: High-tech tools to help
Bjorndal, L.W. Botsford, B.J. Bourque, R.H.
Little, L.R., A.D.M. Smith, A.D. McDonald, A.E.
marine reserve research. Frontiers in Ecology and
Bradbury, R. Cooke, J. Erlandson, J.A. Estes,
Punt, B.D. Mapstone, F. Pantus and C.R. Davies.
Environment 1:73­79.
110
Oceanography Vol. 20, No. 3

---

Perez-Ruzafa A., M. Gonzalez-Wanguemert, P.
Sala, E., E. Ballesteros, and R.M. Starr. 2001. Rapid
Lenfant, C. Marcos, and J.A. Garcia-Charton. 2006.
decline of Nassau grouper spawning aggregations
Effects of fishing protection on the genetic struc-
in Belize: Fishery management and conservation
ture of fish populations. Biological Conservation
needs. Fisheries 26:23­30.
129:244­255.
Sale, P.F., R.K. Cowen, B.S. Danilowicz, G.P. Jones, J.P.
Powles, H., M.J. Bradford, R.G. Bradford, W.G.
Kritzer, K.C. Lindeman, S. Planes, N.V.C. Polunin,
Doubleday, S. Innes, and C.D. Levings. 2000.
G.R. Russ, Y.J. Sadovy, and R.S. Steneck. 2005.
Assessing and protecting endangered marine spe-
Critical science gaps impede use of no-take fishery
cies. ICES Journal of Marine Sciences 57:669­676.
reserves. Trends in Ecology and Evolution 20:74­80.
Pressey, R.L., C.J. Humphries, C.R. Margules, R.I.
Salm, R.V., T. Done, and E. McLeod. 2006. Marine
Vane-Wright, and P.H. Williams. 1993. Beyond
protected area planning in a changing climate. Pp.
opportunism: Key principles for systematic
207­221 in Coral Reefs and Climate Change: Science
reserve selection. Trends in Ecology and Evolution
and Management. Coastal and Estuarine Studies
8:124­128.
61, American Geophysical Union.
Purcell, J.F.H., R.K. Cowen, C.R. Hughes, and D.A.
Shanks, A.L., B.A. Grantham, and M.H. Carr. 2003.
Williams. 2006. Weak genetic structure indi-
Propagule dispersal distance and the size and spac-
cates strong dispersal limits: A tale of two coral
ing of marine reserves. Ecological Applications
reef fish. Proceedings of the Royal Society, Series B
13:S159­S169.
273:1,483­1,490.
Simberloff, D. 1988. The contribution of popula-
Reynolds, J.D., N.K. Dulvy, N.B. Goodwin, and J.A.
tion and community biology to conservation sci-
Hutchings. 2005. Biology of extinction risk in
ence. Annual Review of Ecology and Systematics
marine fishes. Proceeding of the Royal Society, Series
19:473­511.
B 272:2,337­2,344.
Steneck, R.S., M.H. Graham, B.J. Bourque, D. Corbett,
Roberts, C.M. 1997. Connectivity and management of
J.M. Erlandson, J.A. Estes, and M.J. Tegner. 2002.
Caribbean coral reefs. Science 278:1,454­1,457.
Kelp forest ecosystems: Biodiversity, stability,
Roberts, C.M., S. Andelman, G. Branch, R.H.
resilience and future. Environmental Conservation
Bustamante, J.C. Castilla, J. Dugan, B.S. Halpern,
29:436­459.
K.D. Lafferty, H. Leslie, J. Lubchenco, and others.
Stockhausen, W.T., and R.N. Lipcius. 2001. Single large
2003. Ecological criteria for evaluating candidate
or several small marine reserves for the Caribbean
sites for marine reserves. Ecological Applications
spiny lobster? Marine and Freshwater Research
13:S199­214.
52:1,605­1,614.
Roberts, C.M., and J.P. Hawkins. 1997. How small can
Swearer, S.E., J.E. Caselle, D.W. Lea, and R.R. Warner.
a marine reserve be and still be effective? Coral
1999. Larval retention and recruitment in an
Reefs 16:150.
island population of a coral-reef fish. Nature
Roberts, C.M., and J.P. Hawkins. 1999. Extinction
402:799­802.
risk in the sea. Trends in Ecology and Evolution
Thorrold, S.R., R.S. Burton, G.P. Jones, M.E. Hellberg,
14:241­246.
S.E. Swearer, J.E. Niegel, S.G. Morgan, and R.R.
Roberts, C.M., C.J. McClean, J.E.N. Veron, J.P.
Warner. 2002. Quantifying larval retention and
Hawkins, G.R. Allen, D.E. McAllister, C.G.
connectivity in marine populations with artificial
Mittermeier, F.W. Schueler, M. Spalding, F. Wells,
and natural markers: Can we do it right? Bulletin of
and others. 2002. Marine biodiversity hotspots and
Marine Science 70:291­308.
conservation priorities for tropical reefs. Science
Turpie, J.K., L.E. Beckley, and S.M. Katua. 2000.
295:1,280­1,284.
Biogeography and the selection of priority areas
Roberts, C.M., and N.V.C. Polunin. 1991. Are marine
for conservation of South African coastal fishes.
reserves effective in management of reef fisheries?
Biological Conservation 92:59­72.
Reviews in Fish Biology and Fisheries 1:65­91.
Vincent, A.C.J., and Y. Sadovy. 1998. Reproductive
Roberts, C.M., J.D. Reynolds, I.M. Côté, and J.P.
ecology in the conservation management of
Hawkins. 2006. Redesigning coral reef conserva-
fishes. Pp. 209­245 in Behavioural Ecology
tion. Pp. 515­537 in Coral Reef Conservation, I.M.
and Conservation Biology, T. Caro, ed., Oxford
Cote and J.D. Reynolds, eds, Cambridge University
University Press, New York.
Press.
Worthen, W.B. 1996. Community composition and
Rogers, C.S., and J. Beets. 2001. Degradation of marine
nested-subset analyses: Basic descriptors for com-
ecosystems and decline of fishery resources in
munity ecology. Oikos 76:417­426.
marine protected areas in the US Virgin Islands.
Environmental Conservation 28:312­322.
Sadovy, Y. 1993. The Nassau grouper: Endangered or
just unlucky? Reef Encounter 13:10­12.
Sala, E., O. Aburto-Oropeza, G. Paredes, I. Parra, J.C.
Barrera, and P.K. Dayton. 2002. A general model
for designing networks of marine reserves. Science
298:1,991­1,993.
Oceanography September 2007
111

---