Burt et al 2010 Ecological Engineering.pdf

Viewer
Transcript

G Model ECOENG-1769;

No. of Pages 8

ARTICLE IN PRESS Ecological Engineering xxx (2010) xxx–xxx

Contents lists available at ScienceDirect

Ecological Engineering journal homepage: www.elsevier.com/locate/ecoleng

Benthic development on large-scale engineered reefs: A comparison of communities among breakwaters of different age and natural reefs John Burt a,∗,1 , Aaron Bartholomew b , Peter F. Sale a,c a

Department of Biological Sciences, University of Windsor, Windsor, Ontario N9B3P4, Canada Department of Biology and Chemistry, American University of Sharjah, P.O. Box 26666, Sharjah, United Arab Emirates c United Nations University-International Network on Water, Environment and Health (UNU-INWEH), 175 Longwood Road South, Suite 204, Hamilton, Ontario L8P0A1, Canada b

a r t i c l e

i n f o

Article history: Received 11 February 2010 Received in revised form 6 September 2010 Accepted 19 September 2010 Available online xxx Keywords: Breakwater Coastal development Coastal defense Coastal defense Benthos Succession Age Persian Gulf

a b s t r a c t Breakwaters represent large-scale engineered artiﬁcial reefs that can develop diverse and abundant communities and are likely to play an increasing role in marine ecosystems as human populations grow in coastal urban areas. Information on how these communities develop and if and when these communities begin to resemble those on natural hard-bottom habitat is essential for marine management, but is not well understood. In this study, benthic communities on six breakwaters ranging from 1 to 31 years of age were compared to provide an understanding of patterns of community development on engineered coastal defenses, and these were compared to communities on natural reefs to gain an understanding of how communities develop on artiﬁcial structures relative to those in natural habitats. Multivariate analyses indicated that benthic communities on breakwaters became more similar to natural reefs with increasing age, but that communities on even the most mature (31 years) breakwater were distinct from those on natural reefs (ANOSIM p < 0.001). Generally, breakwaters ≤5.5 years had higher abundance of turf algae, sponges, bivalves, and bare pavement, while more mature (≥25 years) breakwaters were dominated by corals. Coral cover on 25 and 31 years old breakwaters (46% and 56%, respectively) was signiﬁcantly higher than on natural reefs (37%; HSD test p < 0.05 and p < 0.001, respectively). These results indicate that breakwaters develop benthic communities that continue to change over periods exceeding 31 years, and that although they become more similar to communities on natural reefs with increasing age, these communities remain distinct. © 2010 Elsevier B.V. All rights reserved.

1. Introduction Man-made coastal defense structures such as breakwaters, jetties, and groynes now dominate the near-shore environment in many areas, in some cases more than doubling the length of the natural coastline (Bacchiocchi and Airoldi, 2003; Airoldi et al., 2005). Such structures represent large-scale unplanned artiﬁcial reefs on which marine communities develop (Svane and Peterson, 2001; Airoldi et al., 2005), with abundance and diversity of ﬁsh, corals, and other benthic organisms often exceeding that of nearby natural reefs (Pondella et al., 2002; Burt et al., 2009a,b). Coastal defense structures, and other artiﬁcial reefs, are usually added to soft sediment environments and provide hard substrates for the attachment of organisms that usually recruit through the settlement of plank-

∗ Corresponding author. Tel.: +971 502219269. E-mail address: [email protected] (J. Burt). 1 Present address: Faculty of Science, New York University, Abu Dhabi, P.O. Box 129188, Abu Dhabi, United Arab Emirates.

tonic larvae. In most cases, epibenthic organisms do not seem to be recruitment limited (Svane and Peterson, 2001), so the addition of hard substrates may lead to an increase in epifaunal recruitment and biomass, providing food resources that may support increased regional ﬁsh production (Bohnsack, 1989). Fish and crustacean communities may also be inﬂuenced by the increased regional habitat complexity and heterogeneity provided by breakwaters and their associated benthic assemblages (Svane and Peterson, 2001). Although not designed for ecological purposes, the relatively large size and ubiquity of engineered coastal defenses suggests that they are likely to play an increasingly important ecological role in coastal marine ecosystems as human populations continue to grow, particularly given the increasing pressure on natural reef habitats (Jaap, 2000; Sale et al., 2010). Most studies of community development on engineered coastal structures have focused on ﬁsh (Pondella et al., 2002; Guidetti et al., 2005; Clynick, 2006). However, the benthic community often contains species considered important to marine management for their aesthetic or nuisance value (Airoldi et al., 2005; Sheehy and Vik, 2009), and is of particular ecological importance in providing food,

0925-8574/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.ecoleng.2010.09.004

Please cite this article in press as: Burt, J., et al., Benthic development on large-scale engineered reefs: A comparison of communities among breakwaters of different age and natural reefs. Ecol. Eng. (2010), doi:10.1016/j.ecoleng.2010.09.004

G Model ECOENG-1769;

No. of Pages 8

2

ARTICLE IN PRESS J. Burt et al. / Ecological Engineering xxx (2010) xxx–xxx

settlement habitat, and shelter for many reef associated organisms, including commercial ﬁsheries species (Qian, 1999; Crossman et al., 2001; Burt et al., 2009a,b). Despite this importance, few studies have examined benthic community development on coastal defense structures, particularly in tropical regions, and these have typically focused either on the very early stages of recruitment and colonization (Osman and Whitlatch, 2004; Bulleri, 2005a,b; Burt et al., 2009a) or on mature communities (Moschella et al., 2005; Burt et al., 2009a,b). Although there is some evidence to suggest that benthic communities on coastal defense structures can differ from those of natural reefs or rocky shores (Connell and Glasby, 1999; Knott et al., 2004; Bulleri, 2005b; Moschella et al., 2005; Burt et al., 2009a,b, 2010), there is little knowledge of how these communities develop over time, nor whether they become increasingly similar to natural reefs as they mature (Bacchiocchi and Airoldi, 2003; Airoldi et al., 2005). Studies of sessile community development on artiﬁcial structures in temperate regions have observed contrasting patterns of community development. In some cases, there were rapid increases in species richness and abundance during the ﬁrst year (Woodhead and Jacobson, 1985; Relini et al., 1994), followed by declines in richness as competitively dominant fauna excluded early colonists (Carter et al., 1985; Woodhead and Jacobson, 1985), while others observed increased species richness as communities developed over time (Chapman and Clynick, 2006). In other cases there were no consistent changes in biomass, abundance, or species richness with increasing artiﬁcial reef submergence time (Wendt et al., 1989). An understanding of the types of communities to expect and the sequence of development is essential for addressing the ecological and management implications of these large-scale artiﬁcial reefs. Engineered coastal defense structures of differing age can be viewed as ‘natural experiments’ on which to observe the development of communities. Such structures are often designed with standardized materials, relief, and complexity, and are generally deployed in comparable near-shore environments within a region. Their staggered construction results in structures immersed for different lengths of time, providing an opportunity to infer temporal patterns of community development by comparing assemblages on structures of different known ages. Such natural experiments cannot account for confounding factors to the extent that is possible with true manipulative experiments, but they do allow examination of ecological processes at scales that would not be possible with a more controlled approach (Diamond, 1986). In this study, the benthic communities on six breakwater reefs ranging from 1 to 31 years of age were compared to provide an understanding of community development on large-scale artiﬁcial structures over time. These were compared to benthic communities on natural reefs to gain an understanding of community development on these artiﬁcial structures relative to mature natural habitats.

2. Materials and methods This study was conducted in Dubai, United Arab Emirates, in the south-eastern basin of the Persian Gulf. Six rocky-reef breakwaters ranging from 1 to 31 years of age were selected for study (Fig. 1). Two sites were sampled on each breakwater to account for variability in benthic assemblages known to occur within structures (Burt et al., 2009b, 2010). All breakwaters were constructed of approximately 4–5 ton quarried rocks, with the exception of the Jebel Ali Port breakwater which was made of concrete jacks of comparable mass. These breakwaters represent all large (>2 km long) breakwaters in Dubai. Six natural reef sites were also sampled for comparison to communities on breakwaters. These natural reef patches occur only to the south-west of the Palm Jebel Ali

in an area of exposed caprock (see Fig. 1) and the remainder of Dubai has a sea-bottom that is dominated by mobile sand and silt unsuitable for most reef fauna (Burt et al., 2008). Maximum depth to the natural substratum at all sites did not exceed 11 m. Water conditions were generally comparable at all sites with the exception of the Dubai Waterfront development, where slightly higher turbidity was associated with reclamation activities occurring several hundred meters from the sampling sites at the time of study. Predominant winds and waves in Dubai are from the north-west (Smit et al., 2008), and sampling was performed on wave exposed portions of breakwaters and unsheltered natural reefs. All areas experience limited tidal amplitude (0.8 m neap to 1.95 m spring) and are in close phase throughout the Dubai coast, while currents are relatively slow (range 0.25–0.40 m s−1 ) and consistent throughout Dubai (Cavalcante et al., 2010). Distribution of contaminants along the coastline in Dubai are spatially comparable and at low levels, except for limited areas located near heavy industrial sites distant from the breakwaters and natural reefs sampled here (AlDarwish et al., 2005). Composition of benthic communities was estimated using photo-quadrats. Six replicate 30 m line transects were photographed at 3 m intervals using a Nikon D80 10 mega-pixel digital camera mounted on a PVC quadrat frame enclosing a 0.25 m2 area, for a total of 66 photo-quadrats per site. Each photoquadrat was oriented parallel to the exposed substratum, such that data from each transect represented a range of horizontal to vertical surfaces. Sampling was standardized to approximately 5–6 m depth on each structure. CPCe image analysis software, version 5 (Kohler and Gill, 2006), was used to analyze benthic communities in each photoquadrat. A total of 50 points were randomly distributed across each photoquadrats image, and the benthos under each point was classiﬁed into its appropriate benthic/taxonomic category, following the methods of Burt et al. (2008) and Burt et al. (2009a,b). Data from photoquadrats were then pooled to provide a mean estimate of coverage of each benthic category for each transect at each site. Analyses were performed on substrate categories occurring in more than 5% of samples (McCune and Grace, 2002), including coverage of scleractinian corals, turf algae, coralline algae, porifera, solitary ascidians, bivalves, gastropods, the echinoids Diadema setosum and Echinometra mathaei, as well as the amount of bare pavement. Preliminary analysis indicated that one natural reef site was an extreme outlier (NR2: Mean Bray-Curtis distance = 0.77, SD = 2.62). Here, live benthic fauna made up <8% of the total cover, with the remainder of the benthos dominated by mobile sands and pavement. Although naturally interspersed throughout the area, such habitat is not characteristic of patch reefs in Dubai. As such, this site was excluded from further analyses on the basis that it was not ecologically or statistically representative of natural reefs (Tabachnick and Fidell, 2001). All data were pre-treated with arcsine square-root transformations prior to multivariate analyses. Non-metric multidimensional scaling (NMS) on Bray-Curtis distances was used to ordinate benthic data for each site based on six replicate transects per site. The NMS autopilot mode of PC-ORD (McCune and Mefford, 1999) performed a Monte Carlo signiﬁcance test on the best of 40 runs of real data with 50 runs of randomized data to optimize the number of axes. To ease interpretation, ordinations were rotated to load the age of breakwaters on the ﬁrst axis, with natural reef points incidentally rotated along with the breakwater points to maximizing the breakwater-age spread on this axis. Joint plots were superimposed on the ordination to illustrate the strength and direction of correlation of benthic members with ordination axes. Only benthos with a Pearson’s r > 0.5 were included in the joint plot. A one-way analysis of similarity (ANOSIM) was performed to test for differences in benthic communities among breakwaters

Please cite this article in press as: Burt, J., et al., Benthic development on large-scale engineered reefs: A comparison of communities among breakwaters of different age and natural reefs. Ecol. Eng. (2010), doi:10.1016/j.ecoleng.2010.09.004

ARTICLE IN PRESS

G Model ECOENG-1769;

No. of Pages 8

J. Burt et al. / Ecological Engineering xxx (2010) xxx–xxx

3

N29°30’ Iran 5 km

N26°30’ Dubai

Saudi Arabia

Study area

U.A.E. E48°30’

E54°30’

E51°30’

N23°30’

E57°30’ WD2

PA2

WF2 WF1

WD1

PA1 PJ2

NR1 - 6

PJ1

JP2

DD2 DD1

JP1

Fig. 1. Map of breakwater (䊉) and natural reef () sampling sites. Site names and age (at the time of study): NR, natural reefs; WF, Dubai Waterfront (1.5 years); PA, Palm Jebel Ali (3.5 years); JP, Jebel Ali port (31 years); PJ, Palm Jumeirah (5.5 years); WD, The World (1 year); DD, Dubai dry docks (25 years).

of different age as well as those of the natural reef. The natural reef and each breakwater were identiﬁed as separate groups a priori, and the test of differences between locations was performed on the transect data from each. ANOSIM is a multivariate ran-

domization test analogous to a standard one-way ANOVA being performed on a distance matrix, but with a minimum of assumptions (Clarke and Gorley, 2006), and is appropriate for assessing groups that have been assigned a priori. ANOSIM produces a test

Stress: 3.8

PJ1(5.5) WD1(1) Pavement PA1(3.5)

DD2(25) DD1(25) D. setosum

Bivalvia

Axis 2

PJ2(5.5)

Porifera Turf algae

Coral JP1(31)

PA2(3.5) JP2(31) WF1(1.5)

Coralline algae

WF2(1.5)

NR6

NR5 NR3

WD2(1) NR4

NR1

Axis 1 Fig. 2. Ordination of benthic communities at sampling sites on breakwaters (䊉) and natural reefs (), with a joint plot of benthic components that were strongly associated with either axis. Site name-number is given in capital letters, with age since construction (years) provided in parentheses for breakwaters.

Please cite this article in press as: Burt, J., et al., Benthic development on large-scale engineered reefs: A comparison of communities among breakwaters of different age and natural reefs. Ecol. Eng. (2010), doi:10.1016/j.ecoleng.2010.09.004

ARTICLE IN PRESS

G Model ECOENG-1769;

No. of Pages 8

4

J. Burt et al. / Ecological Engineering xxx (2010) xxx–xxx

Table 1 Results of comparisons between benthic communities on breakwater sites of consecutive age and natural reefs, pooled for each location/age. The R-statistic and its p-value for comparisons between breakwater communities are shown, along with average percent dissimilarity (Mean ı (%)). Where signiﬁcantly different, percent cover of benthos which contributed >10% to dissimilarity are listed. Age-groups compared

R

p

Mean ı (%)

Benthic variable

Group 1 cover (%)

Group 2 cover (%)

Contribution ı (%)

Cumulative ı (%)

1.0 and 1.5 years

0.57

<0.001

33.9

1.5 and 3.5 years

0.50

<0.001

25.5

36.0 50.5 4.3 0.03 66.8 21.9 1.7

21.9 66.8 0.3 2.6 44.0 39.9 7.5

28.5 24.3 12.0 11.3 28.1 25.8 15.8

28.5 52.6 64.8 76.2 28.1 53.9 69.7

3.5 and 5.5 years 5.5 and 25 years

0.11 0.92

>0.05 <0.001

23.4 42.2

25 and 31 years

0.75

<0.001

25.1

31 years and NR

0.66

<0.001

25.6

Pavement Turf algae Porifera Coral Turf algae Pavement Bivalves n.a. Coral Turf algae Bivalves Pavement Pavement Turf algae Coral Turf algae Coral Pavement

6.8 33.8 7.9 44.2 34.5 4.8 45.8 10.1 56.0 13.8

45.8 4.8 0.2 34.5 13.8 10.1 56.0 1.0 36.5 7.6

29.6 25.4 14.1 12.1 26.7 20.6 12.3 25.8 25.3 20.8

29.6 55.0 69.1 81.2 26.7 47.3 56.7 25.8 51.1 71.9

statistic, R, that assesses the null hypothesis that there are no differences among groups. R ≈ 0 when there are no signiﬁcant differences among groups, with greater differences among groups indicated as R approaches −1 or 1. The signiﬁcance of the R statistic is generated from randomization tests on the distance matrix. Where benthic communities were found to differ signiﬁcantly between breakwater or natural reef groups a similarity percentage analysis (SIMPER) was used to determine the percent dissimilarity between these benthic communities, and to identify the benthic members which were driving these differences. Any benthic members which were related to community differences in multivariate analyses were also examined with univariate analyses. One-way ANOVA with post hoc Tukey’s HSD tests were used to identify signiﬁcant differences in benthic members among breakwaters of different age and those of natural reefs.

3. Results Differences in benthic communities were indicated by separation of sites in NMS ordination (Fig. 2), where a two-dimensional representation reduced stress signiﬁcantly compared with randomized data (Real 2-D stress: 3.8, Stress in randomized data: 15.6; p < 0.05), with no signiﬁcant further reduction in stress in a third dimension. The ﬁrst axis represented 88% of variation and was aligned with differences in benthic communities related to the age of breakwaters. There was a relatively tight clustering of breakwaters less than 5.5 years old to the left on this ﬁrst axis, while breakwaters ≥25 years old and the natural reef sites were clustered to the right. In general, the distribution of sites on this axis changes sequentially with the age of breakwaters, becoming more similar to natural reefs with increased age. Pearson’s r indicated that younger breakwaters were strongly associated with turf algae (r = −0.90), bivalves (r = −0.61), and sponges (r = −0.55), while older breakwaters and the natural reefs were strongly correlated with higher coverage of corals (r = 0.76) and coralline algae (r = 0.51) in the ordination. The second axis represented 9% of the variation, with sites spread across this axis based on the relative abundance of bare pavement (Pearson’s r = 0.95), bivalves (r = 0.58), and the urchin D. setosum (r = 0.58). The three sites most strongly associated with this axis (Site 1 of PA, PJ, and WD) represent the section of breakwater most exposed to the predominant wind-driven wave action on each of these structures (Smit et al., 2008), and this wave

action may explain the higher bare pavement at these sites. In addition, differences in the relative abundance of D. setosum among sites was also associated with this axis, and its grazing action may have also contributed to differences in the relative abundance of bare pavement among sites on this axis. The spread of points in the NMS ordination (Fig. 2) indicated that community structure among breakwaters differed progressively with difference in age, such that natural reefs were most dissimilar to the youngest breakwaters. As such, for brevity, tests for differences in community structure between reefs focused on comparisons between reefs of consecutive age (i.e. those with the highest similarity as illustrated by NMS ordination), rather than on all 21 possible pair-wise comparisons. Benthic communities on breakwaters of each consecutive age generally differed from one another, and communities on the oldest artiﬁcial structure also differed from the natural reef (Table 1). Of all pair-wise comparisons, only reefs of 3.5 and 5.5 years (PA and PJ) had a non-signiﬁcant R-value; benthic communities on all other reefs compared differed signiﬁcantly. Average percent dissimilarity values indicated that the youngest breakwaters (1.0–1.5 years) had dissimilar communities, but that benthos became more similar through 5.5 years, where the 3.5 and 5.5 years breakwater communities did not differ from one another. The 5.5 and 25 years breakwaters, however, had highly dissimilar communities, likely as a result of the 20 years difference in immersion time. Benthic communities associated with the natural reef were signiﬁcantly different from those on the oldest breakwater (31 years). In terms of speciﬁc benthic members which discriminated community groups, the results of SIMPER analyses reﬂected the patterns indicated by joint plot in the NMS (Table 1; Fig. 2). Dissimilarity among breakwaters less than 3.5 years resulted mainly from differences in the relative abundance of bare pavement and turf, together contributing more than 50% of the cumulative dissimilarity between 1 and 1.5 as well as 1.5 and 3.5 years old breakwaters. Because no difference was identiﬁed between breakwaters aged 3.5 and 5.5 years, no attempt was made to identify discriminating taxa. An increase in coral cover and a decline in turf algae with age were primarily responsible for the high dissimilarity between the 5.5 and 25 years reefs. On the more mature 25 and 31 years reefs, dissimilarity was caused mainly by differences in the relative abundance of pavement, turf, and coral. Benthic differences between the oldest breakwater reef (31 years) and the natural reefs were mainly driven by higher cover of coral, turf,

Please cite this article in press as: Burt, J., et al., Benthic development on large-scale engineered reefs: A comparison of communities among breakwaters of different age and natural reefs. Ecol. Eng. (2010), doi:10.1016/j.ecoleng.2010.09.004

ARTICLE IN PRESS

G Model ECOENG-1769;

No. of Pages 8

J. Burt et al. / Ecological Engineering xxx (2010) xxx–xxx

a

60

b

50

Turf algae

5

Pavement

40 40

30 20

20 10 0

0 0

5

10

6

15

c

20

25

30

NR

Porifera

0

5

10

10

Cover (%)

5

20

d

Bivalvia

25

30

NR

25

30

NR

8

4

6

3 2

4

1

2

0

15

0 0

5

10

15

20

25

30

NR

0

5

10

15

20

Age (yr) 60

e

Coral

50 40 30 20 10 0

0

5

10

15

20

25

30

NR

Age (yr) Fig. 3. Mean percent cover (±SE) of benthic variables associated with breakwaters of different age (䊉) and natural reefs (, NR). Note difference in scale on vertical axis.

and bare pavement on the breakwater compared with the natural reef. Changes in the cover of the benthic variables that were identiﬁed as discriminating breakwaters of different age are illustrated in Fig. 3. One-way ANOVAs indicated signiﬁcant difference in the cover of each benthic variable among reefs (ANOVA F(6,94) : Turf = 59.5, Pavement = 15.1, Porifera = 8.4, Bivalves = 44.5, Coral = 295.7; p < 0.001 for each). Post hoc comparisons between reefs used Tukey’s unequal-N HSD tests. Turf algae generally declined with the age of breakwaters (Fig. 3a). There was no signiﬁcant difference in turf on 1 and 1.5 years old breakwaters, but turf coverage declined signiﬁcantly from 1.5 to 5.5 year old breakwaters (Tukey’s unequal N HSD test: p < 0.01) and again on the 25 year old breakwater (p < 0.001). The oldest breakwaters (25 and 31 years) and the natural reefs did not differ in turf cover. The highest turf coverage occurred on 1.5 years old breakwaters, with an average of almost 20% more cover than on the temporally adjacent 1 and 3.5 years breakwaters. This increased turf cover corresponds with a proportional decline in the total of bare pavement, porifera, and bivalve cover on the same breakwater (Fig. 3b–d), suggesting that their decline in percent cover is a function of turf dominance on this breakwater reef rather than age. It is possible that the high turf cover on this 1.5 year old breakwater resulted from increased nutrient loads associated with coastal

reclamation activities occurring several hundred meters from this breakwater, rather than from age-related changes in community structure. The cover of bare pavement did not differ signiﬁcantly between any of the ﬁve breakwaters aged 1–25 years, with the exception of the 1.5 year old breakwater discussed above having less pavement than the 5.5 years breakwater (Fig. 3b; p < 0.05). The oldest breakwater (31 years) had the lowest bare pavement of all artiﬁcial structures, and this was signiﬁcant compared with breakwaters aged 3.5–25 years (p < 0.05 each). Only this oldest breakwater reef (31 years) had comparable bare pavement to the natural reef; all younger artiﬁcial structures had signiﬁcantly more bare pavement than natural reefs (p < 0.05 each). Porifera cover was highest on the youngest breakwater (1 year), which had signiﬁcantly more sponge than any other location (p < 0.05 for all). However, porifera was not a dominant member of the benthos, with maximum coverage of only 4.3 ± 0.8%. There were no signiﬁcant differences among any other locations, with the exception that the 1.5 years reef that was dominated by turf had less porifera cover than the 5.5 years breakwater (p < 0.05). This explains the relatively weak, single appearance of porifera as a discriminating taxon in the SIMPER analysis (Table 1). The cover of bivalves was low on the 1 year breakwater and declined signiﬁcantly on the 1.5 years breakwater (p < 0.05), where

Please cite this article in press as: Burt, J., et al., Benthic development on large-scale engineered reefs: A comparison of communities among breakwaters of different age and natural reefs. Ecol. Eng. (2010), doi:10.1016/j.ecoleng.2010.09.004

G Model ECOENG-1769;

No. of Pages 8

6

ARTICLE IN PRESS J. Burt et al. / Ecological Engineering xxx (2010) xxx–xxx

the benthos was dominated by turf. Bivalve cover increased and peaked on the 3.5 and 5.5 years breakwaters, which had signiﬁcantly more bivalves than all other locations examined (p < 0.05 for all). The lowest bivalve cover was on the mature 25 years old breakwater, which did not differ from the 31 years artiﬁcial breakwater or the natural reef. In contrast to all other benthic variables which tended to decline with age, coral cover increased with the age of breakwaters. Coral coverage was virtually absent on 1 year old breakwaters with only one juvenile coral observed, but increased signiﬁcantly on breakwaters 1.5 years old where they occupied 2.4 ± 0.4% of substrate (p < 0.001). Coverage increased signiﬁcantly again on 5.5 year old breakwaters, as well as on breakwaters at each successive age compared (p < 0.001 for each). Coral cover was highest on the oldest breakwater (31 years of age), at 56.0 ± 1.4%, and the rate of coral cover increase showed no indications of becoming asymptotic on older breakwaters. Natural reefs had 36.5 ± 1.7% coral cover, which was higher than on young breakwaters aged 5.5 years or less (p < 0.001), but was signiﬁcantly lower than on the mature 25 and 31 years breakwaters (p < 0.05 and p < 0.001, respectively).

4. Discussion Breakwaters represent large and ubiquitous engineered ecosystems in coastal urban areas, and the development of associated benthic communities can have positive or negative implications for marine management. Given the growing urbanization of coastal systems worldwide, developing an understanding of the sequence of community development on these structures is essential to allow prediction of if and when management implications are likely to occur, and to understand if and when communities on these structures begin to resemble those in natural habitats. Although early benthic development on recently submerged structures has been relatively well studied, there is inadequate knowledge of how benthic assemblages develop over the long-term and how they compare with natural reefs over time. The results of this study suggest that benthic communities associated with breakwaters continue to change over periods exceeding 31 years, and that while there is a general convergence in community structure towards that on natural reefs with increasing breakwater age, the overall benthic community remains distinct. This reﬂects patterns observed in temperate regions, where benthic assemblages on natural rocky reefs generally differed from those of artiﬁcial structures associated with coastal developments (Connell and Glasby, 1999; Glasby and Connell, 2001; Knott et al., 2004). Such differences between artiﬁcial and natural habitats could lead to an overall increase in regional diversity, provided that natural communities are not negatively impacted by development (Connell and Glasby, 1999). Both in Dubai and elsewhere, there are examples of species which are observed only on artiﬁcial structures but not on natural reefs (Connell and Glasby, 1999), and visa versa (Knott et al., 2004; Burt et al., 2009a,b), and these habitat-associated differences may potentially contribute to a larger regional species pool. In addition, coastal defense structures are often deployed at different times and contain communities at different stages of successional development, and successional differences in communities among structures may also contribute to higher regional diversity (Connell and Glasby, 1999). In this study, for example, turf algae and bivalves were prevalent on young breakwaters, but largely absent from natural reefs (Fig. 3). Thus, the development of breakwater communities over time, and their difference from natural reefs, may have important implications beyond management alone.

Like all natural experiments the results of this study are qualiﬁed. This is not a longitudinal study but a one-time comparison of breakwaters that were constructed at different times in different physical and biological conditions. Although there appears to be general trends in community development among structures of different age in this study, long-term monitoring of community development within and among breakwaters would be necessary to determine the temporal and spatial consistency of these patterns. We are currently performing such monitoring. The youngest breakwaters examined here were dominated by turf algae, bivalves, and sponges. Turf algae are among the earliest colonists on artiﬁcial structures, and are often dominant members of the benthos during the ﬁrst few years (McClanahan, 1997; Aseltine-Neilson et al., 1999). Here, turf algae cover was greater on 1.5 year old reefs, where it occupied over two-thirds of the substratum, than on 1.0 year old reefs or on reefs of greater age. This might suggest a temporal pattern in which algal abundance increases through at least the ﬁrst 1.5 years of age, before declining. It is unlikely, however, that the spike in abundance on 1.5 years reefs is related to age as much as differences in environmental conditions. Cover of turf algae is usually highest shortly after immersion, and cover generally declines within a few months and remains relatively stable for several years thereafter (McClanahan, 1997; Bacchiocchi and Airoldi, 2003). Instead, the high coverage of turf algae on the 1.5 years reef was likely due to coastal reclamation occurring within several hundred meters of this location during sampling. The suspension of sediments during dredging is known to increase nutrient loads in the water column (Lohrer and Wetz, 2003), and such nutrient pulses are associated with increased turf algae cover (McClanahan et al., 2003). The spike in turf algae was associated with a decrease in the amount of bare pavement available for colonization, and it is likely that competitive interference for settlement space explains the concomitant decrease in the abundance of sponges and bivalves on this 1.5 year old reef. Competition with algae has been suggested as one of the primary factors structuring sessile invertebrate communities (Miller and Etter, 2008), and algal cover has been negatively associated with benthic invertebrate abundance in a variety of natural and artiﬁcial habitats (Glasby, 1999; Irving and Connell, 2002). Overall on the reefs younger than 5.5 years as a whole, changes in the abundance of bivalves generally reﬂect patterns of bivalve development on artiﬁcial structures elsewhere. Bivalves are often common members of the early benthic community (Perkol-Finkel and Benayahu, 2005), and generally increase in dominance over the ﬁrst several years on artiﬁcial structures (Woodhead and Jacobson, 1985; Aseltine-Neilson et al., 1999; Nicoletti et al., 2007). Bivalve cover peaked on the 3.5 and 5.5 year old reefs examined here, and then declined on older reefs. Similarly, sponge abundance was highest on the youngest artiﬁcial reef and generally declined thereafter, likely as a result of increased abundance of the dominant spongeeating angelﬁsh Pomacanthus maculosus on older breakwaters in this area (Burt et al., 2009a,b). The decline in turf, bivalves, and sponges on the oldest breakwaters coincided with an increase in coral cover. Corals were virtually absent on the 1 year old artiﬁcial reef with only a single juvenile observed, but increased to 2.4 ± 0.4% cover on a reef just six months older. This relatively rapid increase in coral cover on young breakwaters is not unusual. Although coral recruits are known to settle to artiﬁcial materials within a matter of months, they are rarely observed in situ on artiﬁcial structures during their ﬁrst year as a result of their small size and cryptic habitat preference (Abelson and Shlesinger, 2002). However, observed densities generally increase rapidly within the ﬁrst few months of the second year, as increased recruitment balances loss to mortality and the surviving juveniles grow to sufﬁcient size for observation (Abelson

Please cite this article in press as: Burt, J., et al., Benthic development on large-scale engineered reefs: A comparison of communities among breakwaters of different age and natural reefs. Ecol. Eng. (2010), doi:10.1016/j.ecoleng.2010.09.004

G Model ECOENG-1769;

No. of Pages 8

ARTICLE IN PRESS J. Burt et al. / Ecological Engineering xxx (2010) xxx–xxx

and Shlesinger, 2002). Our results are consistent with the suggestion that fouling assemblages tend to become dominated by long-lived, competitively superior colonial organisms over time (Jackson, 1977; Butler and Connolly, 1996), with coral cover showing a nearly linear pattern of increase on reefs of increasing age, and no indication of an asymptote through 31 years in this study. Abelson and Shlesinger (2002) observed a similar linear increase in coral colonies over an 8-year period on artiﬁcial reefs in the Red Sea. However, a series of studies of other artiﬁcial reefs in the Red Sea indicate that hard coral cover does tend to plateau over time. There, hard coral cover was 5.7% on 10 year old artiﬁcial reefs, increasing to 16.6% on 14 year old structures, but stabilizing on artiﬁcial reefs 20 to >100 years of age (mean cover: 33.4%; Perkol-Finkel and Benayahu, 2004; Perkol-Finkel et al., 2005). The lack of such an asymptote on reefs studied here may be due to reduced competition for space. In the Red Sea, soft corals preferentially recruit to artiﬁcial structures and their cover increases signiﬁcantly with age (Perkol-Finkel et al., 2005; Perkol-Finkel and Benayahu, 2007), frequently resulting in soft coral dominance on these man-made structures (Perkol-Finkel and Benayahu, 2004). The competition for space between soft and hard corals may explain the asymptotic cover of hard corals on older artiﬁcial structures in the Red Sea. However, soft corals do not exists in the study area examined here (Riegl, 1999), and we speculate that the absence of these competitive interactions may have allowed the continued increase in hard coral cover on older breakwaters examined here. There has been no plateau in coral cover up to 31 years on breakwaters, and coral cover on the 31 year old breakwater is higher than on the nearby natural reef suggesting that it may continue to increase on more mature breakwaters. Overall, benthic communities on breakwater reefs became more similar to those on natural reefs with age. It has been suggested that at least 10 years are required for artiﬁcial structures to develop communities comparable to natural reefs (Aseltine-Neilson et al., 1999; Abelson and Shlesinger, 2002; Perkol-Finkel and Benayahu, 2005). The oldest 31 year old reef had a community distinct from those on natural reefs in this study, however. Cover of turf, bivalves, sponges and bare pavement were comparable between the oldest artiﬁcial reef and the natural reefs but there was signiﬁcantly higher coral cover on the mature breakwater reef, and this drove multivariate differences between communities on these reef types. This reﬂects the ﬁndings of an earlier study showing that the composition of the coral communities associated with breakwaters and natural reefs in Dubai differed from one another, with breakwaters dominated by Cyphastrea microphthalma, Platygyra daedalea and Porites lutea while natural reefs were dominated by P. lutea, Porites harrisoni, C. microphthalma, Acropora downingi and Acropora clathrata, in order of abundance (Burt et al., 2009a,b). It is not surprising that mature breakwaters and natural reefs contain different benthic communities as they experience different environmental conditions. These natural reefs are in a low-relief area with substrates dominated by mobile sands and silts (Burt et al., 2008) while the high-relief rocky breakwaters provide substantial opportunity for coral settlement, presumably with lower impacts from sedimentation stress. Sedimentation has been associated with lower coral cover (Rogers, 1990), and has been suggested as a leading cause of coral mortality on Dubai reefs (Riegl, 1999). These habitat associated differences may result in continued divergence of natural reef and mature breakwater coral communities in this area as corals develop higher cover and different communities on breakwaters. In addition, impacts from the planned development of coastal real estate, ports, and desalination facilities near the natural reefs (Sale et al., 2010) are likely to contribute to further divergence of these communities.

7

5. Conclusions Relatively few studies have examined the development of benthic communities on man-made coastal defense structures, and these have typically focused on young structures in temperate environments (Osman and Whitlatch, 2004; Airoldi et al., 2005; Bulleri, 2005a,b). This study provides important data on benthic communities associated with tropical breakwaters across a wide range of ages (1–31 years), and contrasts these with communities associated with natural reefs. This information is essential for predicting the management implications of coastal defense structures and for determining whether breakwaters can act as artiﬁcial surrogates for natural reefs. Overall, the results of this study show that breakwaters transition from being dominated by ephemeral fouling fauna during their early stages of development to one dominated by coral on breakwaters >25 years, with benthic communities likely to continue to change on breakwaters more than three decades in age in tropical environments. Although breakwater benthos tends to become more similar to that of natural reefs with increasing age, they were found to remain distinct at 31 years and this is likely to persist as a result of differences in environmental conditions. These results indicate that engineered coastal ecosystems provide important habitat for the development of abundant and unique benthic communities, but that these structures should be constructed in as ecologically a sensitive manner as possible to conserve natural reef habitats in the tropics. Acknowledgements The authors would like to thank A. Bauman, M. Bernardo, D. Feary, and P. Usseglio for assistance in the ﬁeld, as well as K. M. Jenahi for assistance with image analysis. Nakheel PJSC provided funding and logistical support for this project while Major A. AlSuwaidi of Emirates Marine Environmental Group provided ﬁeld assistance; their support is appreciated. This study was part of the Nakheel – UNU-INWEH research program undertaken to understand the ecological dynamics of ecosystems in the vicinity of Nakheel marine projects. References Abelson, A., Shlesinger, Y., 2002. Comparison of the development of coral and ﬁsh communities on rock-aggregated artiﬁcial reefs in Eilat Red Sea. ICES J. Mar. Sci. 59, 122–126. Airoldi, L., Abbiati, M., Beck, M.W., Hawkins, S.J., Jonsson, P.R., Martin, D., Moschella, P.S., Sundelof, A., Thompson, R.C., Aberg, P., 2005. An ecological perspective on the deployment and design of low-crested and other hard coastal defence structures. Coastal Eng. 52, 1073–1087. Al-Darwish, H.A., El-Gawad, E.A., Mohammed, F.H., Lotfy, M.M., 2005. Assessment of organic pollutants in the offshore sediments of Dubai, United Arab Emirates. Environ. Geol. 48, 531–542. Aseltine-Neilson, D., Bernstein, B., Palmer-Zwahlen, M., Riege, L., Smith, R., 1999. Comparisons of turf communities from Pendleton Artiﬁcial Reef Torrey Pines Artiﬁcial Reef, and a natural reef using multivariate techniques. Bull. Mar. Sci. 65, 37–57. Bacchiocchi, F., Airoldi, L., 2003. Distribution and dynamics of epibiota on hard structures for coastal protection. Estuar Coast Shelf Sci. 56, 1157–1166. Bohnsack, J., 1989. Are high densities of ﬁshes at artiﬁcial reefs the result of habitat limitation or behavioral preference? Bull Mar. Sci. 44, 631–645. Bulleri, F., 2005a. Experimental evaluation of early patterns of colonisation of space on rocky shores and seawalls. Mar. Environ. Res. 60, 355–374. Bulleri, F., 2005b. Role of recruitment in causing differences between intertidal assemblages on seawalls and rocky shores. Mar. Ecol. Prog. Ser. 287, 53–65. Burt, J., Bartholomew, A., Usseglio, P., 2008. Recovery of corals a decade after bleaching in Dubai United Arab Emirates. Mar. Biol. 154, 27–36. Burt, J., Bartholomew, A., Bauman, A., Saif, A., Sale, P.F., 2009a. Coral recruitment and early benthic community development on several materials used in the construction of artiﬁcial reefs and breakwaters. J. Exp. Mar. Biol. Ecol. 373, 72–78. Burt, J., Bartholomew, A., Usseglio, P., Bauman, A., Sale, P.F., 2009b. Are artiﬁcial reefs surrogates of natural habitats for corals and ﬁsh in Dubai United Arab Emirates? Coral Reefs 28, 663–675.

Please cite this article in press as: Burt, J., et al., Benthic development on large-scale engineered reefs: A comparison of communities among breakwaters of different age and natural reefs. Ecol. Eng. (2010), doi:10.1016/j.ecoleng.2010.09.004

G Model ECOENG-1769; 8

No. of Pages 8

ARTICLE IN PRESS J. Burt et al. / Ecological Engineering xxx (2010) xxx–xxx

Burt, J., Feary, D., Bauman, A., Usseglio, P., Sale, P.F., 2010. The inﬂuence of wave exposure on coral community development on man-made breakwater reefs, with a comparison to a natural reef. Bull. Mar. Sci. 86, 839–859. Butler, A.J., Connolly, R.M., 1996. Development and long term dynamics of a fouling assemblage of sessile marine invertebrates. Biofouling 9, 187–209. Carter, J.W., Carpenter, A.L., Foster, M.S., Jessee, W.N., 1985. Benthic succession on an artiﬁcial reef designed to support a kelp reef community. Bull. Mar. Sci. 37, 86–113. Cavalcante, G., B. Kjerfve, D. Feary, A. Bauman, Usseglio, P., 2010. Water currents and water budget in a coastal mega-structure, Palm Jumeirah lagoon, Dubai, UAE. J. Coast. Res., in press. Chapman, M.G., Clynick, B.G., 2006. Experiments testing the use of waste material in estuaries as habitat for subtidal organisms. J. Exp. Mar. Biol. Ecol. 338, 164–178. Clarke, K., Gorley, R., 2006. PRIMER v6: User Manual. PRIMER-E Ltd, Plymouth, UK. Clynick, B.G., 2006. Assemblages of ﬁsh associated with coastal marinas in northwestern Italy. J. Mar. Biol. Assoc. U.K. 86, 847–852. Connell, S.D., Glasby, T.M., 1999. Do urban structures inﬂuence local abundance and diversity of subtidal epibiota? A case study from Sydney Harbour, Australia. Mar. Environ. Res. 47, 373–387. Crossman, D., Choat, J., Clements, K., Hardy, T., McConochie, J., 2001. Detritus as food for grazing ﬁshes on coral reefs. Limnol. Oceanogr. 46, 1596–1601. Diamond, J., Case, T. (Eds.), 1986. Overview: Laboratory Experiments, Field Experiments and Natural Experiments. Community Ecology. Harper & Row, New York, pp. 3–22. Glasby, T.M., 1999. Effects of shading on subtidal epibiotic assemblages. J. Exp. Mar. Biol. Ecol. 234, 275–290. Glasby, T.M., Connell, S.D., 2001. Orientation and position of substrata have large effects on epibiotic assemblages. Mar. Ecol. Prog. Ser. 214, 127–135. Guidetti, P., Verginella, L., Odorico, C., Boero, F., 2005. Protection effects on ﬁsh assemblages, and comparison of two visual-census techniques in shallow artiﬁcial rocky habitats in the northern Adriatic Sea. J. Mar. Biol. Assoc. U.K. 85, 247–255. Irving, A, Connell, S., 2002. Sedimentation and light penetration interact to maintain heterogeneity of subtidal habitats: algal versus invertebrate dominated assemblages. Mar. Ecol. Prog. Ser. 245, 83–91. Jaap, W.C., 2000. Coral reef restoration. Ecol. Eng. 15, 345–364. Jackson, J., 1977. Habitat area, colonization, and development of epibenthic community structure. In: Keegan, B., Ceidigh, P., Boaden, P. (Eds.), Biology of Benthic Organisms. Pergamon Press, Oxford, pp. 349–358. Knott, N.A., Underwood, A.J., Chapman, M.G., Glasby, T.M., 2004. Epibiota on vertical and on horizontal surfaces on natural reefs and on artiﬁcial structures. J. Mar. Biol. Assoc. U.K. 84, 1117–1130. Kohler, K., Gill, S., 2006. Coral Point Count with Excel extensions (CPCe): a Visual Basic program for the determination of coral and substrate coverage using random point count methodology. Comput. Geosci. 32, 1259–1269. Lohrer, A.M., Wetz, J.J., 2003. Dredging-induced nutrient release from sediments to the water column in a southeastern saltmarsh tidal creek. Mar. Pollut. Bull. 46, 1156–1163. McClanahan, T., 1997. Primary succession of coral-reef algae: differing patterns on ﬁshed versus unﬁshed reef. J. Exp. Mar. Biol. Ecol. 218, 77–102. McClanahan, T.R., Sala, E., Stickels, P.A., Cokos, B.A., Baker, A.C., Starger, C.J., Jones, S.H., 2003. Interaction between nutrients and herbivory in controlling algal communities and coral condition on Glover’s Reef, Belize. Mar. Ecol. Prog. Ser. 261, 135–147. McCune, B., Grace, J., 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, OR.

McCune, B., Mefford, M., 1999. PC-ORD: Multivariate Analysis of Ecological Data. MjM Software Design, Gleneden Beach, OR. Miller, R.J., Etter, R.J., 2008. Shading facilitates sessile invertebrate dominance in the rocky subtidal Gulf of Maine. Ecology 89, 452–462. Moschella, P.S., Abbiati, M., Aberg, P., Airoldi, L., Anderson, J.M., Bacchiocchi, F., Bulleri, F., Dinesen, G.E., Frost, M., Gacia, E., Granhag, L., Jonsson, P.R., Satta, M.P., Sundelof, A., Thompson, R.C., Hawkins, S.J., 2005. Low-crested coastal defence structures as artiﬁcial habitats for marine life: using ecological criteria in design. Coastal Eng. 52, 1053–1071. Nicoletti, L., Marzialetti, S., Paganelli, D., Ardizzone, G.D., 2007. Long-term changes in a benthic assemblage associated with artiﬁcial reefs. Hydrobiologia 580, 233–240. Osman, R., Whitlatch, R., 2004. The control of the development of a marine benthic community by predation on recruits. J. Exp. Mar. Biol. Ecol. 311, 117–145. Perkol-Finkel, S., Benayahu, Y., 2004. Community structure of stony and soft corals on vertical unplanned artiﬁcial reefs in Eilat (Red Sea) comparison to natural reefs. Coral Reefs 23, 195–205. Perkol-Finkel, S., Benayahu, Y., 2005. Recruitment of benthic organisms onto a planned artiﬁcial reef: shifts in community structure one decade postdeployment. Mar. Environ. Res. 59, 79–99. Perkol-Finkel, S., Benayahu, Y., 2007. Differential recruitment of benthic communities on neighboring artiﬁcial and natural reefs. J. Exp. Mar. Biol. Ecol. 340, 25–39. Perkol-Finkel, S., Shashar, N., Barneah, O., Ben-David-Zaslow, R., Oren, U., Reichart, T., Yacobovich, T., Yahel, G., Yahel, R., Benayahu, Y., 2005. Fouling reefal communities on artiﬁcial reefs: does age matter? Biofouling 21, 127–140. Pondella, D., Stephens, J., Craig, M., 2002. Fish production of a temperate artiﬁcial reef based on the density of embiotocids. ICES J. Mar. Sci. 59, S88–S93. Qian, P., 1999. Larval settlement of polychaetes. Hydrobiologia 402, 239–253. Relini, G., Zamboni, N., Tixi, F., Torchia, G., 1994. Patterns of sessile macrobenthos community-development on an artiﬁcial reef in the Gulf of Genoa (northwestern Mediterranean). Bull. Mar. Sci. 55, 745–771. Riegl, B., 1999. Corals in a non-reef setting in the southern Arabian Gulf (Dubai UAE): Fauna and community structure in response to recurring mass mortality. Coral Reefs 18, 63–73. Rogers, C.S., 1990. Responses of coral reefs and reef organisms to sedimentation. Mar. Ecol. Prog. Ser. 62, 185–202. Sale, P.F., Feary, D., Burt, J.A., Bauman, A., Cavalcante, G., Drouillard, K., Kjerfve, B., Marquis, E., Trick, C., Usseglio, P., van Lavieren, H., 2010. The growing need for sustainable ecological management of marine communities of the Persian Gulf. Ambio., in press. Sheehy, D.J., Vik, S.F., 2009. The role of constructed reefs in non-indigenous species introductions and range expansions. Ecol. Eng. 36, 1–11. Smit, F., Mocke, G., Giarusso, C., Baranasuriya, P., 2008. Coastal modelling of the Dubai coastline with emphasis on morphological model validation. Conference on Coastal and Port Engineering in Developing Countries VII, vol. Paper No. 71, PIANC-COPEDEC, Dubai, UAE, Paper No. 71. Svane, I., Peterson, J., 2001. On the problems of epibioses, fouling and artiﬁcial reefs, a review. Mar. Ecol. 22, 169–188. Tabachnick, B., Fidell, L., 2001. Using Multivariate Statistics. Allyn and Bacon, Needham Heights, MA. Wendt, P.H., Knott, D.M., Van Dolah, R.F., 1989. Community structure of the sessile biota on ﬁve artiﬁcial reefs of different ages. Bull. Mar. Sci. 44, 1106–1122. Woodhead, P.M.J., Jacobson, M.E., 1985. Epifaunal settlement, the processes of community-development and succession over 2 years on an artiﬁcial reef in the New York Bight. Bull. Mar. Sci. 37, 364–376.

Please cite this article in press as: Burt, J., et al., Benthic development on large-scale engineered reefs: A comparison of communities among breakwaters of different age and natural reefs. Ecol. Eng. (2010), doi:10.1016/j.ecoleng.2010.09.004