See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/250017352

Artificial Reefs: The Importance of Comparisons with Natural Reefs

Article  in  Fisheries · April 1997

DOI: 10.1577/1548-8446(1997)022<0028:ARTIOC>2.0.CO;2

CITATIONS READS
232 1,612

2 authors:

Mark Carr Mark A. Hixon

University of California, Santa Cruz University of Hawaiʻi at Mānoa
80 PUBLICATIONS   8,083 CITATIONS    159 PUBLICATIONS   12,592 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Lionfish Invasion View project

Improved Management of the California Coastal Zone View project

All content following this page was uploaded by Mark A. Hixon on 13 February 2019.

The user has requested enhancement of the downloaded file.

 ARTIFICIAL REEF MANAGEMENT

Artificial Reefs: The Importance of

Comparisons with Natural Reefs
By Mark H. Carr and Mark A. Hixon

ABSTRACT
Methods used to evaluate the performance of an artificial reef will vary according to the purpose
for which the reef was built. To determine how well artificial reefs mitigate losses due to human
activities on natural reefs, the performance of artificial reefs should be evaluated using contempora-
neous comparisons with relatively undisturbed natural reefs. Unfortunately, comparisons between
artificial and natural reefs are typically confounded by differences in reef size, age, and isolation.
We compared colonization and subsequent assemblage structure of reef fishes on coral and artificial
(concrete block) reefs in which reef size, age, and isolation were standardized. Species richness and
fish abundance (all species combined) were greater on reefs of natural rather than artificial struc-
ture, but substantial differences in species composition were not detected. Our results suggest that
artificial reefs with structural complexity and other abiotic and biotic features similar to those of
natural reefs will best mitigate in-kind losses of reef fish populations and assemblages from natural
reefs. Because of the open nature of most reef fish populations, estimating the contribution of artifi-
cial reefs in attracting v producing reef fishes will require a regional assessment of rates of demo-
graphic processes on both artificial and nearby natural reefs.

he two primary goals of artificial reefs in and the concern of fisheries managers about main-
coastal habitats have been, first, to enhance taining stocks of exploited species at harvestable lev-
the production of reef-associated species els, and the desire of conservation biologists to miti-
(i.e., macroalgae, invertebrates, and fishes) gate losses to reef-associated species caused by
and, second, to increase the convenience or efficiency human activities. Greater scrutiny stems from the
of harvesting reef-associated species (Seaman et al. paucity of unequivocal evidence that artificial reefs
1989; Seaman and Sprague 1991; Pratt 1994). Most fulfill their intended objectives, which has led to
often, the purpose of increasing production is to miti- debate regarding their roles in producing versus sim-
gate losses from overfishing or other human activi- ply attracting (i.e., redistributing) organisms, espe-
ties (e.g., pollution, habitat destruction, entrainment cially reef fishes.
and impingement by cooling water systems of Methods used to evaluate the performance of an
coastal power plants). The second goal has been to artificial reef will vary according to the purpose for
create reef habitat both attractive to reef species and which the reef was built. If the primary objective of
easily accessible to harvesters, thereby increasing an artificial reef is to compensate for anthropogenic
catch-per-unit effort (and per-unit cost), at least tem- impacts on particular features of a natural fish popu-
porarily. However, with the ever-increasing concern lation, community, or its habitat, the performance cri-
for conservation and enhancement of reef-associated teria must include information on those specific fea-
species, this second goal has received less priority tures (e.g., population abundance, size structure,
and, in fact, has recently been perceived as a poten- species composition of fish and other reef biota;
tial problem rather than a desired objective (Alevizon Ambrose 1994). If the objective is to mitigate a loss in
and Gorham 1989; Bohnsack 1989; Polovina and fish production, then the performance criteria also
Sakai 1989). should include information from which production
During the past decade, artificial reef programs can be estimated (e.g., larval recruitment, immigra-
have received greater interest and scrutiny (Seaman tion, growth, reproduction, mortality and emigra-
Of l1 1989Q Seaman nnl Snrciip 1991: Pratt 1994). In- tion). Alternatively, performance criteria required to
creased interest stems from growing fishing pressure mitigate in-kind losses in communitywide production

Mark H. Carr is an assistant research biologist at the Marine Science Institute, University of California, Santa
Barbara, CA 93106; [email protected] A. Hixon is a professor at the Department of Zoology, Oregon State
University, Corvallis, OR 97331-2914. Authorship decided by a coin toss.

28 * Fisheries Special Issue on Artificial Reef Management Vol. 22, No. 4

 ARTIFICIAL REEF MANAGEMENT

include the structural (e.g., species composition and more conceptual discussion of the effects of artificial
abundance) and functional (e.g., productivity) attrib- reefs on regional fish production, the necessity of ex-
utes of communities. In any case, the performance of plicitly defining the region in question, and the
artificial reefs should be evaluated using contempo- importance of comparing local production on artifi-
raneous comparisons with relatively undisturbed cial and natural habitats at that scale.
natural reefs nearby. It also is important to know how
quickly artificial reefs are colonized and what factors An Experimental Comparison of Artificial
influence rates of colonization. For example, if long and Natural Reefs
lag times exist between construction of the reef and To determine the effectiveness of small artificial
establishment of the targeted assemblage, a different reefs in mimicking fish assemblages associated with
reef design or other methods that expedite the colo- natural coral patch reefs, we compared the rate of
nization of reef biota may be necessary to more colonization and resultant fish assemblages on repli-
quickly compensate for lost resources. It may be nec- cate natural and artificial reefs of roughly similar
essary to provide or enhance settlement habitat to in- size, age, and isolation. Near the Caribbean Marine
crease the rate of recruitment of planktonically dis- Research Center at Lee Stocking Island, in Exuma,
persed propagules (i.e., larvae and spores) or to Bahamas, we translocated 16 natural coral patch reefs
transplant adults to provide a local spawning source (ca 6 m2 ) to an expansive, shallow (< 4 m depth)
of species with limited dispersal capabilities (e.g., sand bank, where we also constructed 16 artificial
some invertebrates or macroalgae). Obviously, patch reefs of nearly the same size (ca 4 m2) but of
knowledge of the dispersive potential of targeted different structure (taller profile [0.8 m vs. 0.4 ml and
species and the mechanisms of recruitment in natural less variable size of shelter holes) and initially with-
habitats is critical to predicting rates of colonization out associated food organisms (Figure 1). Descrip-
of artificial structures, as is understanding the effects tions of the transplanted coral reefs and the artificial
of reef isolation and the surrounding habitat on the reefs are provided in Carr and Hixon (1995) and
rates and species composition of
colonization. Reef size, age, and K
isolation have not been controlled .
in the few studies that have com- I
pared fish assemblages between
natural and artificial reefs (e.g.,
Molles 1978; Ambrose and
Swarbrick 1989; DeMartini et al.
1989). This is because artificial
reefs are typically much smaller,
younger, and far more isolated
than their natural counterparts.
In this article, we suggest that
evaluating the role of artificial
reefs will benefit markedly from,
first, more detailed comparisons of
the populations and assemblages
of reef species that use artificial
reefs with those on natural reefs
and, second, a determination of
the spatial scales over which artifi-
cial reefs act to attract or produce This natural coral pal tch reef was translocated near similar artificial patch reefs in order to com-
reef species. Conceptually, many pare assemblage of reef fishes during an experiment inthe Bahamas.
of the issues we raise are applica-
ble to most reef-associated species, with some differ- Hixon and Beets (1993, Figure 2; 24 large holes
ences based on relative mobility. However, we focus design), respectively. Our purpose for choosing con-
specifically on reef fishes because of our greater famil- crete blocks was to use standard materials commonly
iarity with these species. We first address the impor- employed in other studies of reef fish assemblages
tance of comparing artificial and natural reefs, using (e.g., Talbot et al. 1978; Bohnsack and Talbot 1980;
information gleaned from our recent experimental Hixon and Beets 1989, 1993). Both reef types were
studies conducted in the Bahamas. We finish with a isolated from their nearest neighbor by 200 m of

April 1997 Special Issue on Artificial Reef Management Fisheries 29

ARTIFICIAL REEF MANAGEMENT

following two years. (The

reefs are still being mon-
itored at this writing.)
mm mm MIN Comparison of the
mm MIN* - mm
- , fish assemblages associ-
ated with our artificial
mm _m _u z z ,z , and natural reefs
mm IN - UK
- I- v vv showed that overall net
rates of fish recruitment
(all species combined)
1 meter were nearly equal at
approximately five new
Figure 1 is a scale illustration of the artificial reefs and translocatedI natural coral reefs used in the larval recruits per reef
larval recruits per reef
experiment at Lee Stocking Island, Bahamas. per census (Hixon and
Carr, unpublished data).
Assuming our estimates
open sand and constructed from August 1991 to of instantaneous rates of recruitment sampled on the
December 1992. However, eight of the artificial reefss six census dates are representative of relative re-
were constructed at exactly the same time as eight c)f cruitment rates throughout the two-year sampling
the natural reefs were translocated (14-19 Decembe:r period, net colonization rates of older fish were likely
1992), so we compare only these reefs in this article due to differential post-recruitment mortality, emigra-
We conducted complete visual censuses of the reef tion, and/or immigration. During the first two years
fishes colonizing each reef six times during the of the experiment, natural reefs accumulated fishes
(both number of individuals and number of species)
A. Fish Abundance more rapidly than artificial reefs (Figure 2). Although
species richness was higher on natural reefs through-
80 out the sampling period, we found no strong differ-
ence in the species composition of fishes (Figure 3).
Of 38 reef fish species observed on either reef type,
.n 60 only three occurred primarily on natural reefs and
LL
0 four occurred primarily on artificial reefs (Figure 3).
2 40 All seven of these species were among the least abun-
E dant on the experimental reefs.
z For the particular structural features of the natural
20 and artificial reefs that we studied, these results sug-
gest that, even when reef age was controlled, and the
0 surrounding habitat and degree of isolation was stan-
dardized, the resulting number of individuals and
0 5 10 15 20 25
species of fishes on natural reefs after two years was
greater than on artificial structures. Thus, the greater
20 vertical relief and shelter availability (number of
holes) of artificial reefs did not compensate for the
greater structural complexity (variety of hole sizes)
0) 15 and natural forage base provided by the corals and
(A
associated benthos of the natural reefs. These results
0 suggest that artificial reefs intended to mitigate the
10 degradation of natural reefs should be structurally as
co
a)

E
Figure 2 illustrates the mean (SEM) number of fish individuals and
z
Z 5 species colonizing 8 artificial and 8 natural patch reefs translocat-
ed to the same sand-bottom habitat off Lee Stocking Island, Baha-
mas, at the same time. In both plots, the final values are signifi-
0 cantly different (P < 0.05, t-tests). Note that the final apparent
0 5 10 15 20 25 decline in species richness was correlated with a winter census
following a late summer census. After the summer settlement
Time (months since start of experiment) period, rare colonists disappear during the winter.

30 Fisheries Special Issue on Artificial Reef Management Vol. 22, No. 4

 ARTIFICIAL REEF MANAGEMENT
similar as possible to natural reefs, especially by
promoting the development of naturally dominant
16 Reefs
benthos (such as corals or macroalgae). 0 Artificial reefs 0 Natural reefs

Attraction v Production
Comparison of fish assemblages associated with nat-
ural and artificial reefs also is fundamentally pertinent
to the "attraction-production" question. Although
logistically difficult to quantify, attraction is a relatively
straightforward concept, which we define as the net
movement of individuals from natural to artificial
habitats. Production is a more problematic concept.
Best quantified as a change in biomass through time
(integrating both the number and mass of individuals),
it reflects births (typically via recruitment of plank-
tonically dispersed larvae), immigration, growth,
death, and emigration. Additionally, gamete produc-
tion is critical to understanding the reproductive con-
tribution of a local population to regional production.
Without measurements of these demographic rates,
< 38 Species
O
estimates of production are difficult to interpret. Li

Recent studies have begun to examine local produc-

tion of fishes on artificial reefs (Polovina and Sakai
1989; DeMartini et al. 1994; Johnson et al. 1994). How- H.
ever, because of the dispersive potential of pelagic lar-
vae of most reef fishes and the additional mobility of S
benthic juveniles and adults of many species, the rela- coruscus
tive fate and performance of fishes in natural v artifi-
cial habitats is critical to understanding the contribu-
tion of artificial reefs. The general question of interest monnga
is: Does an artificial reef provide a habitat for in-
creased production that would otherwise not be possi-
ble? More specifically, could fishes that recruit to artifi-
cial reefs (by either larval settlement or immigration of
older benthic stages) have recruited instead to natural
reefs, and had they done so, what would be their rela-
tive rates of growth, mortality, and emigration, and how DCA Axis 1
would their recruitment influence these demographic Figure 3 summarizes results of detrended correspondence analysis
rates for resident conspecifics and other species? Im- (DCA) of fish assemblages on 8 artificial and 8 natural patch reefs in
portantly, these questions require assessing production the Bahamas inDecember 1994. Inthe upper plot (reefs inspecies
on an artificial reef in the context of regional produc- "space'), reefs are plotted by their species composition "scores"
tion on natural reefs. along DCA axes 1 and 2, which represent linear combinations of
species based on correlated patterns of abundance. The distance
By emphasizing production on an artificial reef in between reefs isproportional to the dissimilarity of fish species
the context of regional production, it is essential to composition and relative abundances. Note that most of the natural
define explicitly the "region" in question. Two possi- reefs cluster above the diagonal, whereas most of the artificial reefs
ble regions are of particular interest. First, there is the occur below the diagonal, with interspersion of reef types in the
area enclosing a collection of local populations that center of the plot. There was no significant difference inthe distrib-
ution of DCA scores between the two reef types (t-test, P= 0.38). In
influence one another's production, i.e., the meta- the lower plot (species inreef "space"), the position of each of 38
population within which an artificial reef has been fish species corresponds with their abundance on reefs depicted in
inserted. Second, and more tangible, is the area of the upper plot, such that the 3 species named above the diagonal
interest to fisheries managers, regardless of how its were mostly found on natural reefs and the 4 species named below
boundaries are defined. Although the two kinds of the diagonal were mostly found on artificial reefs. The fact that most
species cluster inthe center of the plot indicates that most species
regions are inextricably linked, we focus our discus- were similarly abundant on both natural and artificial reefs, espe-
sion on the latter, in which the contribution of an artifi- cially the two dominant grouper, Epinephelus striatus and E.gutta-
cial reef is to be assessed in a defined management area. tus (indicated by crossed symbols).

April 1997 Special Issue on Artificial Reef Management Fisheries * 31

ARTIFICIAL REEF MANAGEMENT

from each other such that there was no larval or mi-

gratory connection (Figure 4B). If the management
area included only these two reefs, then the artificial
reef would necessarily enhance regional production
even if the productivity of the artificial reef was less
than that of the natural reef. In this scenario, larvae
that settled on the artificial reef would be lost from
the management area if that reef did not exist. On
the other hand, if recruitment to an artificial reef re-
duces recruitment to natural habitats by intercepting
larvae (Figure 4C), and if survival and growth are
greater on natural reefs, the effect of an artificial reef
will be to reduce the regional production of fishes,
counter to its purpose. In both examples, knowledge
of larval transport, settlement, and subsequent
growth and mortality on both natural and artificial
reefs would be required to answer the attraction-
Figure 4 illustrates example scenarios depicting the importance of com-
paring natural and artificial reefs. In case A,where the management production question.
area includes only an artificial reef, the reef obviously increases regional In summary, only if production of obligate reef or-
production. In case B,a strong longshore current precludes any larval or ganisms is greater on artificial reefs than on natural
migratory transport between an artificial reef and a natural reef within reefs within an explicitly defined management area
the management area. Here, larvae that settle on the artificial reef can one conclude unequivocally that artificial reefs
would be lost from the management area if the artificial reef did not
exist. Thus, the artificial reef enhances local production even if its pro- enhance production. If otherwise, then artificial reefs
ductivity is less than that of the natural reef. in case C,an artificial reef may enhance production, but testing this possibility
intercepts larvae destined for a natural reef down current. If the inter-
cepted larvae grow less productively on the artificial reef than they oth-
erwise would have on the natural reef, then only attraction has ... if no natural reefs occur in a
occurred, and the artificial reef has decreased regional production. management area containing
an artificial reef, then any obligate
The size of the management area, and the spatial
distribution of reefs within that area, can profoundly
reef organism on the artificial reef
influence interpretation of an artificial reef's effects. has necessarily enhanced production
For example, if no natural reefs occur in a manage- within that management area.
ment area containing an artificial reef, then any
obligate reef organism on the artificial reef has neces- involves distinguishing between cases B and C of
sarily enhanced production within that management Figure 4, which requires knowledge of the fates of
area (Figure 4A). Clearly, the smaller the manage- settlement-stage larvae passing near artificial reefs.
ment area, the greater the contribution of the artifi- Currently, this task is extremely difficult, if not impos-
cial reef to that area. sible, and emphasizes the importance of our under-
If natural reefs do occur in a management area standing of mortality and transport of larvae. All studies
containing an artificial reef, and if larvae that settle of artificial reef communities, whether for mitigation
and survive on the artificial reef would have either purposes or examining the attraction-production
(1) not settled on a natural reef in the management question, would benefit by careful comparisons with
area if the artificial reef was not present or (2) settled natural reef systems. Such studies would necessarily
but experienced either lower growth or lower sur- include both local and regional spatial scales. )
vival on a natural reef than on the artificial reef, then
the artificial reef again will have enhanced produc- Acknowledgments
tion. Failure to recruit to natural reefs in the absence
We thank J. Beets, R. Gomez, B. Head, T. Kaltenberg,
of an artificial reef can be caused by sufficiently high S. Swearer, S. Thompson, and C. Tinus for help in translo-
mortality in the plankton or advection of larvae away cating and constructing reefs in the Bahamas. Logistical
from natural reefs. Relatively poor survival may re- support was provided by the administration and staff of
flect resource limitation on natural reefs or suscepti- the Caribbean Marine Research Center. J. Harding and
bility to higher predation rates. For example, suppose D. Reed provided valuable comments on the manuscript.
that an artificial reef was built offshore of a natural Support for this study was provided by the National
reef and a strong longshore current isolated the reefs Science Foundation grant OCE-92-17163 and the National

32 * Fisheries Special Issue on Artificial Reef Management Vol. 22, No. 4

 ARTIFICIAL REEF MANAGEMENT

Underseas Research Program grants CMRC-92-46, 93-12, dance at an artificial rock reef and a cobble-bottom
94-15, and 95-3042. M. Carr also received support from kelp forest. Bull. Mar. Sci. 44:881-892.
Minerals Management Service grant 14-35-0001-30758. Hixon, M. A., and J. P. Beets. 1989. Shelter characteris-
tics and Caribbean fish assemblages: experiments
References with artificial reefs. Bull. Mar. Sci. 44:666-680.
Alevizon, W. S., and J. C. Gorham. 1989. Effects of arti- . 1993. Predation, prey refuges, and the structure
ficial reef deployment on nearby resident fishes. Bull. of coral-reef fish assemblages. Ecol. Monogr.
Mar. Sci. 44:646-661. 63:77-101.
Ambrose, R. F. 1994. Mitigating the effects of a coastal Molles, M. C., Jr. 1978. Fish species diversity on model
power plant on a kelp forest community: rationale and natural reef patches: experimental insular bio-
and requirements for an artificial reef. Bull. Mar. Sci. geography. Ecol. Monogr. 48:289-305.
55:694-708. Polovina, J. J., and I. Sakai. 1989. Impacts of artificial
Ambrose, R. E, and S. L. Swarbrick. 1989. Comparison reefs on fishery production in Shimamaki, Japan.
of fish assemblages on artificial and natural reefs off Bull. Mar. Sci. 44:997-1,003.
the coast of southern California. Bull. Mar. Sci. Pratt, J. R. 1994. Artificial habitats and ecosystem
44:718-733. restoration: managing for the future. Bull. Mar. Sci.
Bohnsack, J. A. 1989. Are high densities of fishes at arti- 55:268-275.
ficial reefs the result of habitat limitation or behav- Seaman, W., R. M. Buckley, and J. J. Polovina. 1989.
ioral preference? Bull. Mar. Sci. 44:631-645. Advances in knowledge and priorities for research,
Bohnsack, J. A., and F. H. Talbot. 1980. Species packing technology and management related to artificial
by reef fishes on Australian and Caribbean reefs: an aquatic habitats. Bull. Mar. Sci. 44:527-532.
experimental approach. Bull. Mar. Sci. 30:710-723. Seaman, W., and L. M. Sprague. 1991. Artificial habitats
Carr, M. H., and M. A. Hixon. 1995. Predation effects on for marine and freshwater fisheries. Academic Press,
early post-settlement survivorship of coral-reef fishes. San Diego, CA.
Mar. Ecol. Progr. Ser. 124:31-42. Talbot, F. H., B. C. Russell, and G. R. V. Anderson.
DeMartini, E. E., D. A. Roberts, and T. W. Anderson. 1978. Coral reef fish communities: unstable, high-
1989. Contrasting patterns of fish density and abun- diversity systems? Ecol. Monogr. 48:425-440.

April 1997 Special Issue on Artificial Reef Management Fisheries * 33

View publication stats