tmp8506.tmp

African Journal of Marine Science

ISSN: 1814-232X (Print) 1814-2338 (Online) Journal homepage: http://www.tandfonline.com/loi/tams20

Reef fishes recruited at midwater coral nurseries
consume biofouling and reduce cleaning time in
Seychelles, Indian Ocean
S Frias-Torres, H Goehlich, C Reveret & PH Montoya-Maya
To cite this article: S Frias-Torres, H Goehlich, C Reveret & PH Montoya-Maya (2015):
Reef fishes recruited at midwater coral nurseries consume biofouling and reduce
cleaning time in Seychelles, Indian Ocean, African Journal of Marine Science, DOI:
10.2989/1814232X.2015.1078259
To link to this article: http://dx.doi.org/10.2989/1814232X.2015.1078259

View supplementary material

Published online: 24 Sep 2015.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tams20
Download by: [197.157.123.12]

Date: 28 September 2015, At: 21:39

African Journal of Marine Science 2015: 1–6
Printed in South Africa — All rights reserved

Copyright © NISC (Pty) Ltd

AFRICAN JOURNAL OF
MARINE SCIENCE

ISSN 1814-232X EISSN 1814-2338
http://dx.doi.org/10.2989/1814232X.2015.1078259

This is the final version of the article that is published
ahead of the print and online issue

Short Communication

Reef fishes recruited at midwater coral nurseries consume biofouling and
reduce cleaning time in Seychelles, Indian Ocean
S Frias-Torres1,2*, H Goehlich1,3, C Reveret1,4 and PH Montoya-Maya1
Nature Seychelles, Island Conservation Centre, Amitie, Praslin, Republic of Seychelles
Smithsonian Marine Station, Fort Pierce, Florida, USA
3
Institute of Biological Sciences, University of Rostock, Rostock, Germany
4
CREOCEAN, La Rochelle, France
* Corresponding author, e-mail: [email protected]
1

Downloaded by [197.157.123.12] at 21:39 28 September 2015

2

In coral reef restoration, coral gardening involves rearing coral fragments in underwater nurseries prior to
transplantation. These nurseries become fish-aggregating devices and attract biofouling. We hypothesised
that: (1) the presence of corals at a nursery is critical to recruit fish assemblages and (2) the recruited fish
assemblages control biofouling, reducing person-hours invested in nursery cleaning. Three midwater coral
nurseries were deployed at 8 m depth for 27 months within the marine protected area of Cousin Island Special
Reserve, Seychelles, Indian Ocean. Each nursery consisted of a 6 m × 6 m PVC pipe frame, layered with a recycled
5.5-cm-mesh tuna net. Human cleaning effort was calculated based on daily dive logs. Nursery-associated fish
assemblages and behaviour were video-recorded prior to harvesting corals after a 20-month growth period and
seven months post-coral harvesting. The density (ind. m–2) of blue-yellow damselfish Pomacentrus caeruleus was
12–16 times higher when corals were present than when corals were absent at the nurseries. Fish assemblages
recruited into the nurseries included three trophic levels, from herbivores to omnivores, in six families: Ephippidae,
Pomacentridae, Labridae (Scarinae), Gobiidae, Siganidae and Monacanthidae. Higher abundance of large fish (total
number of individuals) resulted in 2.75 times less person-hours spent in nursery cleaning. These results have
important implications for cost-effective coral reef restoration.
Keywords: animal-assisted cleaning, coral gardening, coral reef restoration, endangered species, floating ecosystem
Online supplementary material: Supplementary video footage of (a) humphead parrotfish Bolbometopon muricatum and (b) multiple coral
reef fishes at midwater coral nurseries can be found online at http://dx.doi.org/10.2989/1814232X.2015.1078259.

Introduction
As coral reefs continue to decline worldwide (Hughes et al.
2003; Pratchett et al. 2014), combining traditional conservation with active restoration is increasingly perceived as a
strategy to support resilience in these threatened ecosystems (Hughes et al. 2005; Rinkevich 2008). The emerging
field of ecosystem restoration is expected to become a
dominant discipline in environmental science in the 21st
century (Hobbs and Harris 2001).
Coral reef restoration by ‘coral gardening’ incorporates a
two-step protocol. First, coral ‘seedlings’ (from fragments,
nubbins or settled larvae) are raised in underwater
nurseries. Second, the nursery-reared corals are harvested
and transplanted onto damaged reef areas (Rinkevich
2006). In contrast to terrestrial nurseries – which are
isolated, physically separated from natural ecosystems, and sterilised as a pest-control measure (South
and Enebak 2006; Vercauteren et al. 2006) – midwater
coral nurseries are open to recruitment of reef organisms.

Thus, coral nurseries can also function as artificial reefs or
fish-aggregating devices, due to an increase in available
substrate for reef organisms, structural complexity and
attraction of organisms from natural reefs (Abelson 2006;
Shafir and Rinkevich 2010). Cleaning the nurseries from
biofouling algae and sessile invertebrates (including
sponges, hydroids, bivalves, barnacles and tunicates) is
essential to avoid space competition with corals and coral
death. Nursery cleaning requires a considerable allocation
of the time invested in the total restoration project (Shafir et
al. 2010; Johnson et al. 2011; Hyde et al. 2013).
Fish assemblages (grazers and invertivores) at the
nursery site offer the possibility of animal-assisted cleaning
of biofouling algae and invertebrates. Such biofouling
removal by fishes could reduce human hours invested in
nursery cleaning (Shafir et al. 2010). However, quantification of animal-assisted cleaning by coral reef fishes at
midwater coral nurseries has not been investigated.

African Journal of Marine Science is co-published by NISC (Pty) Ltd and Taylor & Francis

2

We investigated the ability of midwater coral nurseries
to recruit fish assemblages and the potential for animalassisted cleaning at those nurseries in Seychelles,
Indian Ocean. Specifically, we hypothesised that: (i) the
presence of corals at the nursery is critical to recruit fish
assemblages; and (ii) if a coral nursery evolves into a
functional ecosystem, the recruited fish assemblages
control biofouling, thereby reducing person-hours invested
in nursery cleaning.

Downloaded by [197.157.123.12] at 21:39 28 September 2015

Material and methods
As part of a large-scale coral reef restoration project by
‘Nature Seychelles’, three midwater net nurseries were
deployed within the marine protected area of Cousin Island
Special Reserve, Seychelles (04°19′35″ S, 055°39′24″ E).
Each nursery consisted of a 6 m × 6 m frame constructed
from PVC pipe, layered with a recycled 5.5-cm-mesh
tuna net. The three net nurseries (N9, N10, N11), located
about 1 km from the nearest coral reef, were attached
to the sandy bottom at 17 m by anchor lines and kept at
a depth of 8 m using recycled plastic jerrycans as buoys.
In July and August 2012, nurseries were filled with coral
fragments (nubbins) obtained from corals of opportunity
(corals dislodged due to anchor damage or storms) and
donor corals. Each nursery contained ~480 coral fragments.
The corals were a mix of different growth-forms, roughly
50% branching/tabular (Acropora hyacinthus, A. cytherea,
A. vermiculata, A. abrotanoides, A. appressa, Pocillopora
indiania, P. damicornis, P. grandis, Stylophora pistillata,
S. subseriata), 30% massive/submassive (Acanthastrea
brevis, Astreopora myriophthalma, Coscinaraea monile,
Cyphastrea sp., Dipsastraea lizardensis, Favites vasta,
Galaxea fascicularis, Goniastrea edwardsi, Goniopora
tenuidens, Hydnophora microconos, Lobophyllia hemprichii,
Astrae curta, Paramontastraea serageldini, Pavona
decussata, P. explanulata, Platygyra acuta, Porites lobata)
and 20% encrusting (Echinophyllia aspera, Echinopora
hirsutissima, Favites pentagona, Leptastrea purpurea,
Leptoseris incrustans, Psammocora haimiana, Turbinaria
irregularis). Scientific names of corals follow the World
Register of Marine Species (WoRMS) (Hoeksema and
Cairns 2015).
We regularly attached GoPro Hero™ underwater
cameras to the nurseries to observe nursery-associated fish
assemblages and behaviour for the last eight months of a
20-month nursery period. Prior to harvesting the corals for
transplantation (April 2014), we swam 10 video transects
6 m long and 60 cm wide at each nursery to record fish
species composition, abundance and behaviour. We chose
the blue-yellow damselfish Pomacentrus caeruleus as an
indicator species to quantify the effect of coral presence
vs absence at the nurseries. This species is an obligate
coral reef fish; it swims near the substratum, does not
usually move away from the reef, and feeds mostly on
zooplankton (83% of the diet) but also on benthic algae and
vagile benthic invertebrates (17% of the diet) (Frédérich
et al. 2009). Two experienced observers counted all
video-recorded P. caeruleus at each nursery. A t-test of
matched pairs was used to evaluate the precision of the two
independent observers. Differences in P. caeruleus density

Frias-Torres, Goehlich, Reveret and Montoya-Maya

(ind. m−2) between the coral nurseries were assessed
by performing a one-way ANOVA using Anscombetransformed data (√(x + ⅜)) with Tukey HSD post hoc
comparisons (Anscombe 1948). We also used the video
transects to quantify large fishes feeding on the nurseries.
Differences in large-fish abundance (total number of individuals) between nurseries were tested using a Kruskal–Wallis
ANOVA with non-parametric post hoc multiple comparisons
(Mann–Whitney tests with the Bonferroni correction) (Sokal
and Rohlf 1995).
Feeding behaviour of nursery-associated fish species
was compared with trophic-level and food-item information
reported by FishBase (http://www.fishbase.org; Table 1). In
FishBase, a mean trophic position is calculated based on
all food items consumed by a species and weighted by their
relative abundance. The trophic level is obtained by adding
1 to the mean trophic position (Froese and Pauly 2014).
The range of trophic levels includes: 1 for primary producers,
2–2.19 for primary consumers (herbivores consuming mainly
plants or detritus), 2.2–2.79 for omnivores (consuming plants
or detritus and animals), 2.8–3.99 for secondary consumers
(consuming mostly animals) and ≥4 for tertiary consumers
(obligate carnivores).
After coral harvesting, the empty nurseries (N9, N10)
were left in place. One nursery (N11) was towed to the
transplantation site and destroyed during the coral extraction process, so it was unavailable for baseline comparisons. Seven months after coral removal (November 2014)
video transects were repeated at each remaining nursery,
using the procedures explained above, to record fish species
composition, P. caeruleus density (ind. m–2) and large-fish
abundance (total number of individuals, transformed to ranks
for analysis). We used these post-harvesting surveys as our
control, providing the baseline of how much fish recruitment
could be attributed to the structure of the nursery alone.
Nursery cleaning started nine months after filling with
corals and continued as needed until transplantation.
Cleaning involved removal of biofouling around each coral
for a width of 15 cm (above and below the net) with a wire
brush and toothbrush, and removing barnacles from the
frame, ropes and jerrycans using hammers.
Based on our daily dive logs that recorded the number of
divers and days spent cleaning each nursery, we calculated
cleaning effort by adding the number of hours spent by
each diver in the team, resulting in total diver-hours. We
also calculated the time elapsed between cleanings, based
on the dive logs. Differences in cleaning effort between the
nurseries were tested with a one-way ANOVA with Tukey
HSD post hoc comparisons. Statistical analyses were
performed using Statistica 6.0 software.
Results
Two types of reef fishes recruited to midwater coral nurseries
– transient and resident. Transient species included
Ephippidae – longfin batfish Platax teira; Labridae (Scarinae)
– the threatened humphead parrotfish Bolbometopon
muricatum; and Siganidae – forktail rabbitfish Siganus
argenteus. Resident species included Pomacentridae –
two-bar damselfish Dascyllus carneus, regal damselfish
Neopomacentrus cyanomos and blue-yellow damselfish

3

African Journal of Marine Science 2015: 1–6

Table 1: Transient and resident fish species recruited to the midwater coral nurseries, nursery feeding behaviour observed, and published
trophic levels and diet (FishBase information [http://www.fishbase.org] derived from Bellwood and Hughes 2006; Frederich et al. 2009;
Froese and Pauly 2014; Kuo et al. 2015)

Family/species (common name)

Ephippidae
Platax teira* (longfin batfish)
Labridae (Scarinae)
Bolbometopon muricatum* (humphead parrotfish)
Siganidae
Siganus argenteus (forktail rabbitfish)

Downloaded by [197.157.123.12] at 21:39 28 September 2015

Pomacentridae
Dascyllus carneus (two-bar damselfish)
Neopomacentrus cyanomos* (regal damselfish)
Pomacentrus caeruleus* (blue-yellow damselfish)
Gobiidae
Gnatholepis cauerensis (eyebar goby)
Pleurosicya micheli (Michel’s ghost goby)
Monacanthidae
Oxymonacanthus longirostris (longnose filefish)

Observed nursery
feeding behaviour
Transients

FishBase information
Trophic
level (SE)

Food items

Macroalgae

4.0 (0.64)

Zooplankton, jellyfish, macroalgae

Macroalgae, barnacles

2.7 (0.41)

Benthic algae, live corals, shellfish

Epilithic algal turfs
Residents

2.2 (0.0)

Macroalgae, turf algae, soft corals

Plankton
No feeding observed

2.7 (0.29)
3.4 (0.5)

Plankton

2.7 (0.30)

Zooplankton
Zooplankton
Zooplankton, benthic algae, vagile
benthic invertebrates

No feeding observed
No feeding observed

2.4 (0.1)
3.0 (0.5)

Inferred from similar species
Commensal on hard corals

Coral polyps

3.3 (0.6)

Acropora sp. polyps

* Species was found at the nurseries both with corals present and corals absent

P. caeruleus; Gobiidae – eyebar goby Gnatholepis
cauerensis, Michel’s ghost goby Pleurosicya micheli; and
Monacanthidae – longnose filefish Oxymonacanthus
longirostris. Large fishes that were observed feeding at
the net nurseries included B. muricatum, S. argenteus
and P. teira. Fish were observed feeding on macroalgae
(B. muricatum and P. teira), epilithic algal turfs
(S. argenteus), plankton (D. carneus and P. caeruleus) and
coral polyps (O. longirostris) (Table 1, Figure 1; supplementary video material, available online). Although we did
not complete a systematic survey of nursery invertebrates,
we found the following to be common: commensal crabs
(Trapezia spp.) in branching/tabular corals; brittle stars
(Ophiothrix spp.) in branching/tabular and massive/submassive corals; and bristleworms (Polychaeta) in all corals. Soft
corals (Dendronephthya spp.) recruited to the underside of
the nurseries. After coral harvesting, the nurseries had an
impoverished fish fauna, with only two transient species
(P. teira and B. muricatum) and two resident species
(N. cyanomos and P. caeruleus).
Two observers counted all video-recorded P. caeruleus
at each nursery, with corals (April 2014) and without corals
(November 2014). There were no significant differences
between observers at nursery N9 (t9 = 0.65, p = 0.53), N10
(t9 = 0, p = 1) and N11 (t9 = 0.5, p = 0.63). When corals
were present, P. caeruleus density at each nursery was
6.4 ind. m–2 (SD 0.35) (N9), 3.7 ind. m–2 (SD 0.02) (N10),
and 3.8 ind. m–2 (SD 0.13) (N11) (Figure 2). The differences
in P. caeruleus density were statistically significant (F2,3 =
116.7, p = 0.001). The post hoc comparisons showed that

(a)

(b)

(c)

Figure 1: Three reef fish species associated with midwater coral
nurseries: (a) Bolbometopon muricatum; (b) Pomacentrus caeruleus
on Acropora sp. coral colony; and (c) Siganus argenteus feeding on
epilithic algal turfs (photos: CR, SF-T and HG, respectively)

Adjusted in photoshop
Used *cmyk alt

4

MEAN DENSITY (ind. m−2)

Frias-Torres, Goehlich, Reveret and Montoya-Maya

Coral

7
6
5
4
3
2
1
N9

N10

CORAL NURSERY

N11

MEAN RANKED ABUNDANCE

Figure 2: Mean density of Pomacentrus caeruleus at each
midwater coral nursery (N9, N10 and N11) with corals present and
absent, respectively. Error bars denote SD

4
3.5
3
2.5
2
1.5
1
0.5
N9

N10

CORAL NURSERY

N11

Figure 3: Mean ranked abundance of Bolbometopon muricatum,
Siganus argenteus and Platax teira, combined, at each midwater
coral nursery (N9, N10 and N11). Vertical bars show minimum and
maximum values

MEAN CLEANING EFFORT
(diver-hours)

Downloaded by [197.157.123.12] at 21:39 28 September 2015

No Coral

25
20
15
10
5
N9

N10

CORAL NURSERY

N11

Figure 4: Mean cleaning effort at each midwater coral nursery (N9,
N10 and N11). Error bars denote SE

the differences between N9 and the other two nurseries (by
a factor of approximately 1.7 in each case) were significant
(p = 0.002).
When corals were absent, P. caeruleus density at each
nursery was 0.39 fish m–2 (SD 0.04) (N9), 0.32 fish m–2 (SD
0.02) (N10), and these values did not differ significantly
(t-test, p > 0.05) (Figure 2). When comparing nurseries N9
and N10, based on coral presence or absence, the total
number of P. caeruleus was ~12 times (N10) to ~16 times
(N9) higher when corals were present than when corals
were absent (two-way ANOVA, Fnursery1,4 = 142.72, p  =
0.0003; Fcoral1,4 = 3 528.54, p < 0.0001). The interaction term
of nursery and coral presence was significant (Finteraction1,4  =
97.46, p = 0.0006).
When corals were present, herbivorous fish abundance
was the highest at nursery N9 (H = 16.93, p = 0.0002),
compared to nurseries N10 and N11 (Figure 3). The
combined ranked abundance of B. muricatum, S. argenteus
and P. teira was highest at nursery N9 compared to
nurseries N10 and N11. The results were significant
(H corrected for ties = 16.93, df = 2, p = 0.0002). Post
hoc comparisons revealed significant differences in the
abundance of the three species only between nurseries N9
and N11 (U0.05(3,10) > 80.99).
Mean cleaning effort at each nursery was 8 diver-hours
(SE 1.15) (N9), 14.6 diver-hours (SE 1.3) (N10) and 22
diver-hours (SE 6) (N11), with a range of 6–28 diver-hours
for the 20-month nursery period. The differences in cleaning
effort were significant (F2,5 = 6.54, p = 0.04) (Figure 4). Post
hoc comparisons showed that the difference in the mean
cleaning effort between nurseries N9 and N11 (a factor of
2.57) was significant (p = 0.048). The differences in mean
cleaning effort between N9 and N10, and between N10 and
N11, were not significant (p = 0.22 and p = 0.28, respect­
ively). Time between cleanings was 2.2 months (SE 1.54)
(N9), 3.1 months (SE 1.80) (N10) and 3.6 months (SE 2.40)
(N11). These differences were not significant (F2,6 = 0.44,
p = 0.65).
Discussion
Coral presence at the midwater nurseries was a critical
factor in the abundance of Pomacentrus caeruleus. Corals
provide P. caeruleus, an obligate coral reef fish, with all
its shelter and nutritional needs (zooplankton, benthic
algae and vagile benthic invertebrates). In the empty
nurseries, P. caeruleus was found only at the highest threedimensional structural points – mainly buoys, near connections of net and nursery frames, and at points where the
frame had been enlarged due to encrustations. The peak
breeding season of P. caeruleus is May and June (Jung
et al. 2010). Although there are no data on the planktonic
duration of P. caeruleus, we assume it to resemble the
planktonic phase of other Pomacentrus species with similar
ecological requirements. The planktonic larval duration
of 17 species of western Pacific Pomacentrus has been
estimated to be between 17 and 19 days (Wellington and
Victor 1989). Therefore, the higher P. caeruleus abundance
in April 2014, when corals were present, and the lower
abundance in November 2014, when corals were absent,

5

Downloaded by [197.157.123.12] at 21:39 28 September 2015

African Journal of Marine Science 2015: 1–6

was not linked to a recruitment pulse typical of coral reef
fishes (Williams 1983), but to the presence or absence of
corals in the nurseries.
The presence of fish assemblages, and their predatory
effect on biofouling, reduced the amount of effort to clean
the nurseries. Diver cleaning effort for nursery N9 was
36.3% of the time required for nursery N11. Nursery N9
had the highest density of P. caeruleus (ind. m–2) and
the highest abundance (total number of individuals) of
B. muricatum, S. argenteus and P. teira. Nursery N11 had a
lower density of P. caeruleus and the lowest abundance of
B. muricatum, S. argenteus and P. teira.
Midwater coral nurseries often require weekly to monthly
cleaning during the first six months after filling with corals,
and monthly cleaning thereafter (Shafir et al. 2010; Johnson
et al. 2011; Hyde et al. 2013). If the nurseries under study
had required monthly cleaning over the 20-month coral
growth period, we would have invested 20 cleaningmonths (months during which some cleaning effort in these
nurseries would occur). However, we started cleaning
nine months after filling, and we completed eight cleaningmonths for the three nurseries combined, reducing our
projected time invested in cleaning by 60%.
Overall, the fish assemblages included three trophic
levels, from herbivores to omnivores (Table 1). Fishes
capable of limiting algal growth included B. muricatum
and P. teira feeding on macroalgae and S. argenteus
feeding on epilithic algal turfs. Based on video observations, B. muricatum did not predate on the nursery corals,
and there was a lack of the characteristic scarring left by
this parrotfish. The presence of P. teira underscores the
functional nature of the midwater nursery as an evolving,
floating, coral reef, because batfishes are critical in phaseshift reversal from a macroalgal-dominated to a coral- and
epilithic-dominated reef state (Bellwood and Hughes 2006).
Similar midwater coral nurseries evolved into floating-reef
ecosystems over five years in the Gulf of Eilat, Red Sea
(Shafir and Rinkevich 2010).
Our results have important implications for sustainable
and cost-effective coral reef restoration. The net nurseries
with seeded corals provide a three-dimensional patch
of habitat where a floating ecosystem can evolve over
time (Shafir and Rinkevich 2010; this study), including the
recruitment of fish assemblages of different trophic levels.
The presence of corals is critical for such recruitment to
occur. The fish assemblages that recruit to the nurseries
consume biofouling algae and invertebrates. As a result,
diver-time invested in cleaning the nurseries is reduced,
and can be allocated instead to additional coral transplantation, capacity building or scientific research, depending
on project needs. As corals are harvested, the net nursery
frame can be kept in place to receive new coral fragments
for gardening and start a new floating ecosystem.
Acknowledgements — We thank K Rowe, M Mullins, D Quinlan
and D Soto for help during fieldwork and N Shah for commenting
on an earlier draft of the manuscript. Funding to Nature
Seychelles was received through the United States Agency
for International Development (USAID) Reef Rescuers Project
674-A-00-10-00123-00.

References
Abelson A. 2006. Artificial reefs vs coral transplantation as
restoration tools for mitigating coral reef deterioration: benefits,
concerns, and proposed guidelines. Bulletin of Marine Science
78: 151–159.
Anscombe FJ. 1948. The transformation of Poisson, binomial and
negative binomial data. Biometrika 35: 246–254.
Bellwood DR, Hughes TP, Hoey AS. 2006. Sleeping functional
group drives coral-reef recovery. Current Biology 16: 2434–2439.
Frédérich B, Fabri G, Lepoint G, Vandewalle P, Parmentier E.
2009. Trophic niches of thirteen damselfishes (Pomacentridae)
at the Grand Récif of Toliara, Madagascar. Ichthyological
Research 56: 10–17.
Froese R, Pauly D (eds). 2014. FishBase. Available at www.
fishbase.org, version (04/2014) [accessed 8 December 2014].
Hobbs RJ, Harris JA. 2001. Restoration ecology: repairing the
Earth’s ecosystems in the new millennium. Restoration Ecology
9: 239–246.
Hoeksema B, Cairns S. 2015. Scleractinia. World Register of
Marine Species. Available at www.marinespecies.org/aphia.
php?p=taxdetails&id=1363 [accessed 29 May 2015].
Hughes TP, Baird AH, Bellwood DR, Card M, Connolly SR, Folke C
et al. 2003. Climate change, human impacts and the resilience of
coral reefs. Science 301: 929–933.
Hughes TP, Bellwood DR, Folke C, Steneck RS, Wilson J.
2005. New paradigms for supporting the resilience of marine
ecosystems. Trends in Ecology and Evolution 20: 380–386.
Hyde J, Chelliah AJ, Chen SY. 2013. Reef rehabilitation
experiments: a review of results and lessons learned. Reef
Check Malaysia, Jalan Ampang, Malaysia.
Johnson ME, Lustic C, Bartels E, Baums IB, Gilliam DS, Larson L
et al. 2011. Caribbean Acropora restoration guide: best practices
for propagation and population enhancement. Arlington, Virginia:
The Nature Conservancy.
Jung MM, Oh BS, Kim SY, Lee CH, Yang MH, Han SJ et al.
2010. Spawning volumes and times of blue devil Pomacentrus
caeruleus. Korean Journal of Ichthyology 22: 90–95.
Kuo FW, Fan TY, Liu MC, Kuo CY, Chen CA. 2015. Hidden
ecosystem function of rabbitfishes? Siganus fuscescens feeds
on the soft coral, Sarcophyton sp. Coral Reefs 34: 57.
Pratchett MS, Hoey AS, Wilson SK. 2014. Reef degradation and
the loss of critical ecosystem goods and services provided by
coral reef fishes. Current Opinion in Environmental Sustainability
7: 37–43.
Rinkevich B. 2006. The coral gardening concept and the use of
underwater nurseries; lesson learned from silvics and silviculture.
In: Precht WF (ed.), Coral reef restoration handbook. Boca
Raton, Florida: CRC Press. pp 291–302.
Rinkevich B. 2008. Management of coral reefs: we have gone
wrong when neglecting active reef restoration. Marine Pollution
Bulletin 56: 1821–1824.
Shafir S, Edwards A, Rinkevich B, Bongiorni L, Levy G, Shaish L.
2010. Constructing and managing nurseries for asexual rearing
of corals. In: Edwards AJ (ed.), Reef rehabilitation manual. St
Lucia, Australia: Coral Reef Targeted Research and Capacity
Building for Management Program. pp 49–72.
Shafir S, Rinkevich B. 2010. Integrated long-term mid-water coral
nurseries: a management instrument evolving into a floating
ecosystem. University of Mauritius Research Journal 16:
365–385.
Sokal RR, Rohlf FJ. 1995. Biometry: the principles and practices of
statistics in biological research. New York: WH Freeman & Co.
South DB, Enebak SA. 2006. Integrated pest management
practices in southern pine nurseries. New Forests 31: 253–271.
Vercauteren UC, Lavelle MJ, Hygnstrom S. 2006. Fences and

6

Frias-Torres, Goehlich, Reveret and Montoya-Maya

(Pomacentridae). Marine Biology 101: 557–567.
Williams DMB. 1983. Daily, monthly and yearly variability in
recruitment of a guild of coral reef fishes. Marine Ecology
Progress Series 10: 231–237.

Downloaded by [197.157.123.12] at 21:39 28 September 2015

deer-damage management: a review of designs and efficacy.
Wildlife Society Bulletin 34: 191–200.
Wellington GM, Victor BC. 1989. Planktonic larval duration of
one hundred species of Pacific and Atlantic damselfishes

Manuscript received January 2015, revised April 2015, accepted June 2015