, 742 (2007);
et al.Glenn R. Almany,
in a Marine Reserve
Local Replenishment of Coral Reef Fish Populations
www.sciencemag.org (this information is current as of May 3, 2007 ):
The following resources related to this article are available online at
version of this article at:
including high-resolution figures, can be found in the online
Updated information and services,
can be found at:
Supporting Online Material
, 4 of which can be accessed for free:
cites 24 articles
This article appears in the following
in whole or in part can be found at:
permission to reproduce
of this article or about obtaining
Information about obtaining
registered trademark of AAAS.
c 2007 by the American Association for the Advancement of Science; all rights reserved. The title SCIENCE is a
CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005.
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
on May 3, 2007
8. S. W. Squyres et al., J. Geophys. Res. 111, E02S11,
9. S. W. Ruff et al., J. Geophys. Res. 111, E12S18,
10. R. V. Morris et al., J. Geophys. Res. 111, E02S13,
11. R. G. Burns, in Remote Geochemical Analysis: Elemental
and Mineralogical Composition, C. M. Pieters,
P. A. J. Englert, Eds., (Cambridge Univ. Press, Cambridge,
1993), pp. 539–556.
12. R. V. Morris et al., Abstract #2014, presented at the 31st
Lunar and Planetary Science Conference, Houston, TX, 13
to 17 March 2000 (available on CD-ROM).
13. B. Scaillet, B. Clemente, B. W. Evans, M. Pichavant,
J. Geophys. Res. 103, (B10), 23937 10.1029/98JB02301
15. R. V. Fisher, A. C. Waters, Am. J. Sci. 268, 157 (1970).
16. K. H. Wohletz, M. F. Sheridan, Geol. Soc. Am. Spec. Pap.
180, 177 (1979).
17. D. M. Rubin, Cross-Bedding, Bedforms, and Paleocurrents
(Society of Economic Paleontologists and Mineralogists,
Tulsa, OK, 1987).
18. J. A. Grant, S. A. Wilson, S. W. Ruff, M. P. Golombek,
D. L. Koestler, Geophys. Res. Lett. 33, L16202,
19. N. A. Cabrol et al., J. Geophys. Res. 108, (E12), 8076
20. This research was carried out for the Jet Propulsion
Laboratory, California Institute of Technology, under a
contract with the NASA.
Supporting Online Material
Figs. S1 to S7
19 December 2006; accepted 26 March 2007
Local Replenishment of Coral Reef
Fish Populations in a Marine Reserve
Glenn R. Almany,1* Michael L. Berumen,1,2Simon R. Thorrold,3
Serge Planes,4Geoffrey P. Jones1
The scale of larval dispersal of marine organisms is important for the design of networks of marine
protected areas. We examined the fate of coral reef fish larvae produced at a small island reserve,
using a mass-marking method based on maternal transmission of stable isotopes to offspring.
Approximately 60% of settled juveniles were spawned at the island, for species with both short
(<2 weeks) and long (>1 month) pelagic larval durations. If natal homing of larvae is a common
life-history strategy, the appropriate spatial scales for the management and conservation of coral
reefs are likely to be much smaller than previously assumed.
vation rely on untested assumptions about the de-
gree to which fish populations are connected by
larval dispersal (1–3). Connectivity is a critical
parameter in models for optimizing the size and
any of the desired outcomes of ma-
rine protected areas (MPAs) in fisheries
management and biodiversity conser-
spacing of MPAs (4–6), but the scarcity of direct
information on larval dispersal limits the models’
utility (7). Because larvae typically spend times
ranging from days to months in the pelagic envi-
ronment before seeking suitable habitat to begin
adult life, direct measurements of connectivity
are challenging (7–9). Thus, although larvae have
the potential to travel far from their birthplace,
realized dispersal distances are seldom known.
We studied populations of two species of
coral reef fishes with different reproductive strat-
egies, occupying a 0.3-km2coral reef surrounding
a small island that was recently designated an
MPA (Kimbe Island in Kimbe Bay, Papua New
Guinea). Orange clownfish (Amphiprion percula;
Pomacentridae) spawn demersal eggs that hatch
after several days of parental care, and larvae
then spend ~11 days in the pelagic environment.
In contrast, vagabond butterflyfish (Chaetodon
vagabundus; Chaetodontidae) release gametes
directly into the water column (that is, there is
no parental care), and larvae spend an average
of 38 days in the pelagic environment (Fig. 1).
The reproductive characteristics of the vagabond
butterflyfish are found in most marine fish spe-
cies and in nearly all species targeted by fish-
eries throughout the world’s oceans.
In December 2004, we tagged larvae using a
method whereby mothers transmit stable barium
(Ba) isotopes to their offspring before hatching
and dispersal (10). A total of 176 clownfish fe-
males and 123 butterflyfish from the reef sur-
rounding Kimbe Island (Fig. 2) were captured
and injected with a BaCl2solution that was highly
enriched in137Ba and depleted in135Ba as com-
pared to natural Ba isotope values. In February
2005, we returned to Kimbe Island and col-
lected 15 clownfish and 77 butterflyfish that had
recently settled into benthic reef habitats after
completing their pelagic larval phase. The analysis
of daily growth increments of sagittal otoliths
(ear bones) confirmed that each of the recent
settlers was born after the injection of the adults
with BaCl2. We then quantified Ba isotope
ratios in the otolith cores of settlers, using laser
ablation inductively coupled plasma mass spec-
trometry (ICP-MS). Ba isotope ratios in the
otoliths of all individuals fell on the theoret-
ical mixing curve between the enriched isotope
spike and natural Ba, with values that were sim-
ilar to those from otoliths of larvae from three
reef fish species injected with enriched
1Australian Research Council Centre of Excellence for Coral
Reef Studies and School of Marine and Tropical Biology,
James Cook University, Townsville QLD 4811, Australia.
2Department of Biological Sciences, University of Arkansas,
Fayetteville, AR 72701, USA.3Department of Biology, Woods
Hole Oceanographic Institution, Woods Hole, MA 02543,
USA.4Joint Research Unit 5244, Ecole Pratique des Hautes
Etudes–CNRS, Centre de Biologie et d’Ecologie Tropicale et
Mediterraneenne, Universite de Perpignan, F-66860 Perpignan
*To whom correspondence should be addressed. E-mail:
Fig. 1. Study species. An adult (A) A. percula (photo by S. R. Thorrold) and (B) C. vagabundus (photo by R. Patzner).
4 MAY 2007VOL 316
on May 3, 2007
under controlled laboratory conditions (Fig. 3).
Definitive identification of tagged fish was based
on otolith138Ba/137Ba ratios because they con-
verged on the natural ratios much more slowly
than did138Ba/135Ba values (Fig. 3). A fish
was considered tagged if the138Ba/137Ba ratio
of its otolith core was more than 3s lower than
the mean value from measurements of control
otoliths (n = 86) assayed throughout the ICP-
Otoliths of nine clownfish and eight butterfly-
fish were identified as tagged based on138Ba/137Ba
ratios, providing incontrovertible evidence that
these fish had returned to their natal reef (Fig. 3).
Assuming that we tagged all clownfish larvae
produced from Kimbe Island, 60% of juveniles
made the return journey—a conservative estimate
if we failed to locate any breeding pairs. For
butterflyfish, we estimated the proportion of the
total adult population captured and injected with
Ba as 17.3% [95% confidence interval (CI): 14.4
to 20.0%) via a population survey conducted after
123 adults had been injected with Ba and exter-
nally tagged. Scaling the proportion of tagged
juveniles (8 of 77) to the proportion of adults in-
jected with Ba indicated that a remarkable 60.1%
(95% CI: 52.0 to 72.2%) of juvenile butterflyfish
returned to their natal reef. Tagged juveniles were
found in a variety of locations scattered around
Kimbe Island, although the larvae of both species
that returned settled in the greatest numbers at the
southeastern corner of the island (Fig. 2).
Our direct estimate of ~60% self-recruitment
for these two species demonstrates that larvae are
capable of returning to a very small target reef
(only 0.3 km2), even after an extended larval
duration. Although there is much recent indirect
evidence for the limited dispersal of marine lar-
vae (11), our results, in combination with two
previous mark/recapture studies of larval disper-
sal (12, 13), suggest that self-recruitment in ma-
rine fish populations may be common and take
place on a smaller scale than previously realized.
For example, a recent Caribbean-wide biophys-
ical model of population connectivity in reef
fishes highlighted larval capabilities as a key
factor determining levels of self-recruitment (14).
When active larval behavior was introduced in
the model within a few days after hatching, self-
recruitment of virtual larvae averaged ~21% to
reef areas delineated by 450 km2. In our study,
the proportion of self-recruitment was three times
greater to a reef more than three orders of mag-
The observation that parental habitat is de-
monstrably of sufficient quality for survival and
reproduction provides a compelling argument for
the presence of some degree of self-recruitment
Fig. 2. (A) Satellite image of the Kimbe Island MPA (taken by the IKONOS-2
satellite at a resolution of 1 m). (B and C) Schematic diagrams of Kimbe Island
showing the locations of tagged (red circles) and untagged (white circles) juveniles
collected in February 2005. The locations of juvenile (B) A. percula (n = 15) and (C)
C. vagabundus (n = 77) are shown. In (C), the number in each circle corresponds to the number of juveniles collected from that location.
Fig. 3. (A) Ba isotope ratios in otolith cores from
juveniles of three reef fish species: A. melanopus
(diamonds), A. percula (squares), and Centropristis
striata (triangles), spawned from females injected
with137BaCl2up to 3 months after the injections
and reared in the lab through settlement (10); and
mean values (±3s) from controls for each species
(open symbols), along with the theoretical mixing
curve (dashed line) between the enriched Ba
isotope spike and natural Ba ratios. (B) Ba isotope
ratios in otolith cores of tagged A. percula (squares)
and C. vagabundus (circles) juveniles and mean
(±3s) values from the otoliths of all unmarked
mixing curve (dashed line) between the enriched
Ba isotope spike and natural Ba ratios.
VOL 316 4 MAY 2007
on May 3, 2007
in fish populations. Selection may therefore favor
the retention of many larvae, especially if the
probability of encountering better adult habitat
by dispersing is low (15) or advantages accrue
through local adaptation (16). A number of mech-
anisms may be used by larvae to avoid being
swept away from natal reefs. Field evidence sug-
gests that reef fish larvae migrate vertically in the
water column to exploit currents at different depths
and thereby avoid dispersal away from spawning
locations (17). Larvae are also capable of sustained
directional swimming soon after hatching (18),
and possess a range of well-developed sensory
systems to locate and orient to reefs, including
sight, smell, and sound (18–21).
Despite the high levels of self-recruitment we
detected, ~40% of juveniles of both species came
from outside the MPA. The reef nearest to Kimbe
Island is 10 km away, and reefs in this region are
typically separated by 5 to 20 km. Ecologically
important larval exchange must occur between
populations at these scales. Thus, the Kimbe
Island MPA is likely to be self-sustaining as well
as providing recruitment subsidies to populations
beyond its boundaries. Although levels of reten-
tion and connectivity may differ where reefs are
closer and populations are less isolated, the Kimbe
Island example sets a new boundary condition for
the scale at which self-recruitment can occur.
Ideally, the size and spacing of marine re-
serves should be predicated on an understanding
of larval dispersal distances (3–6, 22). The op-
timal design should be one in which individual
MPAs are large enough so that populations with-
in reserves can sustain themselves, yet small
enough and spaced so that a proportion of larvae
produced inside the MPA is exported to un-
protected areas (3, 5, 12). Our study suggests that
the spatial scale at which coral reef MPAs can
achieve these dual goals may be relatively small.
However, if natal homing and larval retention
are common, some MPAs may fail to deliver
substantial recruitment subsidies to locations
beyond their boundaries. We therefore support
recent suggestions (23, 24) that MPA networks
should be combined with conventional man-
agement strategies to both protect threatened
species and ensure the sustainability of fisheries
on coral reefs.
References and Notes
1. E. Sala et al., Science 298, 1991 (2002).
2. F. R. Gell, C. M. Roberts, Trends Ecol. Evol. 18, 448 (2003).
3. B. S. Halpern, R. R. Warner, Proc. R. Soc. London Ser. B
270, 1871 (2003).
4. L. W. Botsford, A. Hastings, S. D. Gaines, Ecol. Lett. 4,
5. A. L. Shanks, B. A. Grantham, M. H. Carr, Ecol. Appl. 13,
6. S. R. Palumbi, Ecol. Appl. 13, S146 (2003).
7. P. F. Sale et al., Trends Ecol. Evol. 20, 74 (2005).
8. R. R. Warner, R. K. Cowen, Bull. Mar. Sci. 70, 245 (2002).
9. S. R. Thorrold et al., Bull. Mar. Sci. 70, 291 (2002).
10. S. R. Thorrold, G. P. Jones, S. Planes, J. A. Hare, Can.
J. Fish. Aquat. Sci. 63, 1193 (2006).
11. S. E. Swearer et al., Bull. Mar. Sci. 70, 251 (2002).
12. G. P. Jones, S. Planes, S. R. Thorrold, Curr. Biol. 15, 1314
13. G. P. Jones, M. J. Milicich, M. J. Emslie, C. Lunow, Nature
402, 802 (1999).
14. R. K. Cowen, C. B. Paris, A. Srinivasan, Science 311, 522
15. R. R. Strathmann et al., Bull. Mar. Sci. 70, 377 (2002).
16. D. O. Conover, L. M. Clarke, S. B. Munch, G. N. Wagner,
J. Fish Biol. 69, 21 (2006).
17. C. B. Paris, R. K. Cowen, Limnol. Oceanogr. 49, 1964
18. J. M. Leis, Adv. Mar. Biol. 51, 57 (2006).
19. G. Gerlach, J. Atema, M. J. Kingsford, K. P. Black,
V. Miller-Sims, Proc. Natl. Acad. Sci. U.S.A. 104, 858
20. D. Lecchini, J. Shima, B. Banaigs, R. Galzin, Oecologia
143, 326 (2005).
21. S. D. Simpson, M. Meekan, J. Montgomery, R. McCauley,
A. Jeffs, Science 308, 221 (2005).
22. L. W. Botsford, F. Micheli, A. Hastings, Ecol. Appl. 13,
23. R. Hilborn, F. Micheli, G. A. De Leo, Can. J. Fish. Aquat.
Sci. 63, 642 (2006).
24. L. W. Botsford, Bull. Mar. Sci. 76, 245 (2005).
25. We thank S. Sheppard of The Nature Conservancy for
satellite imagery analyses; E. Laman-Trip for aging juvenile
butterflyfish; J. Almany, D. DeVere, N. Gardiner, V.
Messmer, M. Srinivasan, C. Syms, and H. Walsh for field and
lab assistance; and the Global Environmental Fund,
Connectivity Working Group, for providing a forum for
discussing different approaches to evaluating marine
connectivity. This project was financially supported by an
Australian Research Council (ARC) Discovery Grant
(DP0208120), the ARC Centre of Excellence for Coral Reef
Studies, a Coral Reef Initiative for the South Pacific grant,
and NSF through an International Research Fellowship to
G.R.A. and grants OCE-0215905 and OCE-0424688.
Supporting Online Material
Materials and Methods
30 January 2007; accepted 2 April 2007
Developmentally Regulated piRNA
Clusters Implicate MILI in
Alexei A. Aravin, Ravi Sachidanandam, Angelique Girard,
Katalin Fejes-Toth, Gregory J. Hannon*
Nearly half of the mammalian genome is composed of repeated sequences. In Drosophila, Piwi
proteins exert control over transposons. However, mammalian Piwi proteins, MIWI and MILI,
partner with Piwi-interacting RNAs (piRNAs) that are depleted of repeat sequences, which raises
questions about a role for mammalian Piwi’s in transposon control. A search for murine small RNAs
that might program Piwi proteins for transposon suppression revealed developmentally regulated
piRNA loci, some of which resemble transposon master control loci of Drosophila. We also find evidence
of an adaptive amplification loop in which MILI catalyzes the formation of piRNA 5′ ends. Mili mutants
derepress LINE-1 (L1) and intracisternal A particle and lose DNA methylation of L1 elements,
demonstrating an evolutionarily conserved role for PIWI proteins in transposon suppression.
birth (P14) (1–4). However, Mili expression
begins in primordial germ cells at embryonic
day 12.5 (5, 6), and transposons, such as L1,
can be expressed in both premeiotic and meiotic
germ cells (7, 8). We therefore probed a con-
nection between Mili and transposon control by
nown piRNAs are not expressed until
spermatocytes first enter mid-prophase
(pachytene stage) at ~14 days after
examining MILI-bound small RNAs in early-
stage spermatocytes. Notably, MILI-associated
RNAs could be detected at all developmental
time points tested (Fig. 1 and fig. S1). Northern
blotting revealed that pre-pachytene piRNAs
join MILI before pachytene piRNAs become
expressed at P14 (Fig. 1B). The appearance of
pre-pachytene piRNAs was MILI-dependent,
suggestinga requirementfor thisproteinineither
their biogenesis or stability (Fig. 1C). These
results raised the possibility that MILI might be
programmed by distinct piRNA populations at
different stages of germ cell development.
lated MILI complexes from P10 testes and deeply
sequenced their constituent small RNAs. Like
pachytene populations, pre-pachytene piRNAs
were quite diverse, with 84% being cloned only
once. The majority of both pre-pachytene (66.8%)
and pachytene (82.9%) piRNAs map to single
genomic locations. However, a substantial fraction
(20.1%) of pre-pachytene piRNAs had more than
10 genomic matches, as compared to 1.6% for
Annotation of pre-pachytene piRNAs re-
vealed three major classes (Fig. 2A). The largest
(35%) corresponded to repeats, with most match-
ing short interspersed elements (SINEs) (49%),
long interspersed elements (LINEs) (15.8%), and
long terminal repeat (LTR) retrotransposons
(33.8%). Although pachytene piRNAs also
match repeats (17%), the majority (>80%) map
Watson School of Biological Sciences, Cold Spring Harbor
Laboratory, Howard Hughes Medical Institute (HHMI), 1
Bungtown Road, Cold Spring Harbor, NY 11724, USA.
*To whom correspondence should be addressed. E-mail:
4 MAY 2007 VOL 316
on May 3, 2007