Marine protected areas export larvae of infauna, but not of bioengineering mussels to adjacent areas

Biological Conservation 144 (2011) 2088–2096

Contents lists available at ScienceDirect

Biological Conservation journal homepage: www.elsevier.com/locate/biocon

Marine protected areas export larvae of infauna, but not of bioengineering mussels to adjacent areas V.J. Cole a,⇑, C.D. McQuaid a, M.D.V. Nakin b a b

Coastal Research Group, Department of Zoology and Entomology, P.O. Box 94, Rhodes University, Grahamstown 6140, South Africa Department of Zoology, Walter Sisulu University, P/Bag x 1, Mthatha 5100, South Africa

a r t i c l e

i n f o

Article history: Received 4 October 2010 Received in revised form 1 February 2011 Accepted 27 April 2011 Available online 20 May 2011 Keywords: Colonisation Connectivity Ecosystem engineer Harvesting Mussels

a b s t r a c t Populations of organisms that create habitat can often be fragmented throughout landscapes by anthropogenic disturbances such as harvesting and loss or change to the identity of such bioengineers may lead to large changes in biodiversity. Using the fauna associated with a bioengineer, the intertidal mussel Perna perna, we tested hypotheses about the relative importance of larval export from protected populations in marine reserves. Harvesting led to the replacement of P. perna and the domination of shores outside reserves by turf-forming coralline algae, mostly Corallina spp. We determined whether the diverse fauna recruiting onto artificial units of habitat placed within mussel beds differed between reserves and non-reserve areas or whether shores outside reserves, and open to harvesting, received recruits through larval export from reserves. Furthermore, we determined whether this was affected by the distance away from reserves and whether colonisation was achieved by movement of adults from surrounding biogenic habitats or via the plankton. Overall, we found no effect of increasing distance away from a reserve on the cover of adult mussels or associated fauna. We found strong effects of the presence of marine reserves on abundances of molluscs and polychaetes but not crustaceans. There were greater densities of molluscs in sites with a reserve (i.e. inside reserves, and up to 5 km outside reserve boundaries), but more polychaetes in exploited sites. For molluscs, this pattern was driven by gastropods rather than bivalves. Furthermore, although reserves had greater cover of adult mussels than non-reserve areas, recruitment of mussels was not greater inside or near to reserves. Our study illustrates the effectiveness of these reserves in protecting stocks of adult mussels, and although there was no evidence that reserves provided export of the larvae of mussels (the target species), they did provide larval export of non-targeted associated species. By protecting a harvested bioengineer and through export of the larvae of its associated fauna, these reserves fulfil some, but not all the conservation aims of a marine protected area. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction There are concerns about how anthropogenic transformation of landscapes will lead to loss of species (Lindborg and Eriksson, 2004), so that we need to understand how fragmentation and connectivity affect ecological processes and the abundance and distribution of species in time and space. Usually habitat destruction refers to loss of biogenic habitat and consequent alteration of both habitat structure and abiotic conditions, rather than direct alteration of environmental conditions. Work has begun clarifying the effects of landscape fragmentation and the addition of corridors among habitat patches by focussing on individual species and the processes that threaten them (e.g. Beier and Noss, 1998; Bélisle, 2005), with previous studies illustrating delayed responses to habitat destruction (Hanski, 1998). Models of metapopulation dynamics propose that the probability of re-colonisation of patches that ⇑ Corresponding author. Tel.: +27 46 603 8525; fax: +27 46 622 8959. E-mail address: [email protected] (V.J. Cole). 0006-3207/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocon.2011.04.030

have been destroyed and of ‘‘rescue-effects’’ depend on the proximity of neighbouring patches (see Hanski (1998) for review). Extinction cascades are likely if populations of keystone species or entire functional groups become fragmented (Fischer and Lindenmayer, 2007). Organisms that create habitat are an important functional group in many landscapes and their populations can often be fragmented by anthropogenic or natural disturbances (Fahrig and Merriam, 1994). These habitat-forming species often facilitate increased diversity of smaller organisms (Bruno et al., 2003). Such biogenic habitats have also been referred to as resulting from ‘‘ecosystem engineering’’ (Jones et al., 1997). Ecosystem engineers are important because they may have widespread impacts on the structure of assemblages in time and space (Jones et al., 1994). On intertidal rocky shores, mussels and turf-forming coralline algae provide habitats for diverse assemblages (e.g. Seed, 1996; Kelaher et al., 2001; Borthagaray and Carranza, 2007). In addition to supporting greater diversity than substrata devoid of biogenic habitats, different types of bioengineers support distinctly different assemblages (Borthagaray and Carranza, 2007; Cole et al.,

2089

V.J. Cole et al. / Biological Conservation 144 (2011) 2088–2096

2007). If bioengineers are lost, the overall biodiversity of rocky shores will change due to the loss of the habitat they provide and their replacement by other taxa (Paine and Levin, 1981), leading to changes in habitat quality (Hanski, 1998). This is particularly problematic if populations of bioengineers become highly fragmented as small populations are more likely to go extinct than larger ones. Risk of extinction of small populations may, however, be decreased due to immigration from larger neighbouring populations (Hanski, 1998). An important management response to concerns about loss of species is the establishment of reserves (Halpern, 2003; Halpern and Warner, 2003). The use of marine protected areas (MPAs) is often couched in terms of the conservation of biodiversity, although a large proportion of studies on MPAs have focussed on harvested or target species rather than biodiversity per se (Russ and Angel, 2011). The consensus of these studies is that reserves have rapid and long lasting positive effects (Roberts et al., 2001; Halpern and Warner, 2002; Halpern, 2003), although recovery of benthic communities after the establishment of reserves may be seriously delayed by indirect effects (Parravicini et al., 2010). Reserves, particularly in marine systems where there are commercial fisheries, have often been considered as important sources of fish that ‘‘spill-over’’ beyond reserve boundaries (Rowley, 1994; Palumbi, 2004). The size of spill-over relative to residual populations outside the reserves is important, because managers may rely on such information to determine the size and extent of future reserves (McNeill and Fairweather, 1993). Previous studies that have inves-

tigated spill-over effects have mostly been on large mobile taxa, e.g. fish (see Roberts et al., 2001 for review). The few studies that have investigated similar effects due to the larval export of sessile taxa (e.g. Pelc et al., 2009) have been on habitat forming taxa (e.g. mussels and barnacles). Investigation of multispecies assemblages, particularly those that are associated with harvested taxa, is more complex. This is important because harvesting will have indirect effects on the biodiversity associated with bioengineers, as well as on the target species themselves. In the Transkei region on the east coast of South Africa (Fig. 1), intertidal organisms are harvested as a major source of protein (Lasiak, 1991). Within this region, there are a number of reserves where harvesting is prohibited (Hockey et al., 1988). Previous studies have shown that intertidal molluscs are the main target taxa for artisanal fishers (Hockey et al., 1988; Lasiak and Dye, 1989), and the brown mussel, Perna perna (Linnaeus 1758), often comprises greater than 80% of total shellfish offtake (Lasiak, 1992). Through the investigation of middens from coastal households in the region, Lasiak (1991) found that P. perna, the abalone, Haliotis spadicea, the turban snail, Turbo sarmaticus, and various patellid limpets were the preferred food for many people. Other molluscs were often (greater than 75% of the time) found in middens and included Oxystele sinensis, Fissurella natalis, Thais capensis, Burnupena cincta and Burnupena lagenaria, and Dinoplax validofossus (Lasiak, 1992). These taxa are also associated with beds of P. perna (Cole and McQuaid, 2010). Outside reserves, space freed by mussel collection is generally occupied by macroalgae (Lasiak and Field, 1995) and as

SOUTH AFRICA

200 km

Port St John’s

Hluleka

Hole-in-the-Wall

Dwesa

Shixini 15 km

Fig. 1. Map of South Africa showing the Transkei Region. Shores in each of the sites with marine reserves at Hluleka and Dwesa and exploited sites at Hole-in-the-Wall and Shixini were sampled. Shores were also sampled 2 km and 5 km south of each of the reserves and exploited shores.

2090

V.J. Cole et al. / Biological Conservation 144 (2011) 2088–2096

a result, algal associated species such as amphipods, ophiuroids and nudibranches, are more abundant in exploited areas than in areas inside reserves (Lasiak, 1998). Connectivity on scales of 10s to 100s km plays a critical role in determining the structure of assemblages (Borthagary et al., 2009), and the export of the planktonic larvae of sedentary species will depend on both local oceanography and larval dispersal abilities and so will depend on both geography and taxon (Cudney-Bueno et al., 2009), as well as the seasonality of reproduction (Carson et al., 2010). Marine reserves are considered a potential source of colonists for neighbouring populations subject to harvesting (Pelc et al., 2009). We therefore proposed that: (1) marine reserves have a positive effect on densities of mussels and their associated assemblages; (2) there is strong connectivity among patches of rocky shores along the coast so that marine reserves export fauna to areas open to harvesting; and (3) distance from a reserve is an important factor influencing the number of colonising fauna. We also proposed that the effects of reserves on populations of associated fauna would not be consistent, but would differ among and within taxa because, (a) some taxa are affected directly by harvesting while others are only indirectly affected, (b) taxa have different abilities to recolonise due to differences in mobility, mode of reproduction and capability for dispersal as adults or larvae. Consequently, we hypothesised that the assemblages associated with the same ecological engineer would differ in response to proximity and degree of habitat destruction or harvesting of the ecological engineer. To test these hypotheses, we used mussels and their associated fauna as a model system. Specifically, we predicted that the effects of marine reserves and distance from a reserve would have the greatest influence on molluscs because they would be affected directly (through targeted harvesting) and indirectly (through changes to the surrounding habitat). Crustaceans would be intermediately affected, with some taxa (e.g. crabs and rock lobsters) being targeted for harvesting and influenced directly whereas others (e.g. amphipods) would only be influenced indirectly. Polychaetes are not harvested and would, therefore, only be affected indirectly through changes to the surrounding habitat. The effect of distance away from potential sources may also be influenced by the dispersal distance of these taxa (Pelc et al., 2009, 2010), but this differs within the broad taxonomic groups (Kinlan and Gaines, 2003). We therefore also predicted the effects of marine reserves and distance from a reserve would differ between those species that are targeted directly by harvesting and those that are indirectly affected by harvesting.

2. Methods The research was done over a 100 km stretch of coast on the eastern coast of South Africa (Fig. 1), during the austral autumn of 2008. The experiment was set up in the Transkei region where there are multiple shores in marine parks and shores that are open to harvesting. The experiment was done in two sites with marine parks, Dwesa and Hluleka, and two randomly chosen exploited sites, Hole-in-the-Wall and Shixini (Fig. 1). Within each of these sites, there were shores that formed the ‘‘In’’ treatment (either ‘‘in’’ a reserve or ‘‘in’’ a randomly selected area open to exploitation). Sites with marine parks were interspersed along the coast with sites that were open to harvesting (Fig. 1), avoiding possible spatial confounding due to along-shore currents or other environmental gradients. ‘‘Near’’ shores were 1.5–2 km from ‘‘In’’ shores and ‘‘Far’’ shores were 5 km away; for both reserve and non-reserve sites, near and far shores were open to harvesting. The distance between the marine parks and exploited sites was 20– 30 km (Fig. 1). This design was necessary to successfully compare larval export at fixed distances from a marine reserve with changes

at similar spatial scales in areas where there is no marine reserve (i.e. open to harvesting) (see Chapman et al., 1995). We tested whether the cover of mussels and turf-forming coralline algae differed between reserve and non-reserve sites. Notably, within the reserves, there were extensive beds of mussels but outside the reserves there were only isolated individuals or small patches of mussels. Furthermore, space on shores outside of reserves appeared to be dominated by turf-forming coralline algae. Percentage cover of mussels, P. perna, and turf-forming coralline algae (a complex of Corallina spp., Jania spp. and Arthrocardia spp.) were estimated using the point intercept method, with six 50  50 cm quadrats (each with 100 evenly spaced points), for each taxon on each shore. For the fauna associated with mussels, an asymmetrical environmental impact assessment design was used, as described by Underwood (1991, 1992, 1993). The design was asymmetrical because, although the experiment was initially designed with two marine reserves, all experimental units were lost from one reserve site, leaving Dwesa as the only reserve site. It was necessary to keep two exploited sites to successfully investigate whether the presence of a marine reserve had an effect over and above that of natural spatial variability. A reserve by distance interaction was predicted assuming no effect of distance in exploited sites. In contrast, in the site with a reserve, it was expected that there would be more species, more individuals and different assemblages inside the reserve than at the two distances away from the reserve. If, however, the marine reserve supplied larvae to neighbouring shores (near and far) at equal rates to within the reserve, there would be a main effect of reserve but not distance. Hypotheses were tested using multivariate and univariate analyses. Multivariate data of assemblages were analysed using Permutational Analysis of Variance (PERMANOVA+ for PRIMER, Anderson et al., 2009). The design was based on that described by Anderson et al. (2009) which allows an asymmetrical design to be set up directly and yields the correct test, rather than having to build it in steps (as required for the univariate analyses). This was achieved by using an experimental design with a first factor ‘‘MPA versus Exploited’’ (fixed, 2 levels), a second factor ‘‘Site’’ that was random and nested in MPA versus Exploited (1 site nested in the level MPA and 2 sites nested in Exploited), and a third factor ‘‘Distance’’ that was orthogonal and fixed (3 levels: in, near or far) (see Anderson et al., 2009). Pair-wise a posteriori comparisons were done for significant sources of variation to determine patterns of difference relative to the hypotheses of interest. When there was a significant interaction between the factors ‘‘Site’’ and ‘‘Distance’’, pair-wise tests were selected to investigate the effect of distance for each site separately. Univariate analyses of numbers of species and numbers of individuals were carried out using asymmetrical analysis of variance (ANOVA) following the procedures outlined by Underwood (1992). These analyses dealt with the lack of replication of reserve sites by first analysing the data with a two-factor ANOVA with the random factor ‘‘Site’’ (3 levels) orthogonal to the fixed factor ‘‘Distance’’, then reanalysing the data omitting the reserve site. From the two separate analyses, the entire asymmetrical analysis was calculated by subtractions and additions of components (Underwood, 1992) to test the effects of the two exploited sites against the marine reserve and distance. When sources of variation for univariate analyses were shown to be significant, Student–Newman– Keuls (SNK) tests were used to compare means relative to the hypotheses of interest. For univariate and multivariate analyses, n = 4 replicates were used. Although ten replicates were attached to the shore at the beginning of the experiment, a minimum of 4 replicates at each shore were retrieved at the end of the experiment.

2091

V.J. Cole et al. / Biological Conservation 144 (2011) 2088–2096

Standardised units of habitat (plastic pot scourers) were used to collect colonists over a period of 3 months. Scourers have been successfully used to test hypotheses about colonisation of organisms in biogenic habitats (Gobin and Warwick, 2006; Underwood and Chapman, 2006) and have been shown to be colonised by assemblages matching those on nearby natural substrata (Smith and Rule, 2002; Cole et al., 2007). Use of scourers also removes any confounding influences due to duration of colonisation. Each scourer was secured to the rocky substratum, on the low shore, within a patch of mussels or adjacent to mussels (if there were few adult mussels). After 3 months, scourers were collected and fixed in ethanol. In the laboratory, the scourers were unrolled and washed through a 0.5 mm sieve. All organisms larger than 0.5 mm (i.e. macrofauna) were identified to species under a dissecting microscope and counted. To test for specific differences among different groups of taxa, species were then grouped into the broad taxonomic groups: Polychaeta, Mollusca and Crustacea. Mobile (e.g. gastropods, errant polychaetes, crabs) and sessile or sedentary taxa (e.g. bivalves and tube-dwelling amphipods) were also separated to determine the mode of colonisation (settlement of larvae or adult movement from surrounding substrata). Taxa were also separated according to whether they were directly or indirectly affected by harvesting (i.e. target versus non-target species), regardless of taxonomic grouping.

100

Mean percentage cover (+SE)

(a) Perna perna 80

60

40

20

0 Exploited 1

Exploited 2

Mean percentage cover (+SE)

100

MPA 1

MPA 2

(b) Coralline algae

80

60

40

20

3. Results The cover of both mussels and coralline turf showed a significant interaction between site and distance. There was a significantly greater cover of mussels inside the two reserve sites than near or far from the reserves (Fig. 2a, F4,60 = 31.72, P < 0.001, SNK Hluleka P < 0.05, Dwesa P < 0.01). In the exploited site at Shixini, there was no effect of distance (Fig. 2a, SNK P > 0.05). In the exploited site at Hole-in-the-Wall, there were significantly more

0 Exploited 1

Exploited 2

MPA 1

MPA 2

Fig. 2. Mean (+SE; n = 6) percentage cover of (a) Perna perna, and (b) coralline algal turf at Hole-in-the-Wall (Exploited 1), Shixini (Exploited 2), Hluleka (MPA 1) and Dwesa (MPA 2). Shores were sampled in a reserve site or in exploited sites (In, black bars), 2 km away (Near, light grey bars) and 5 km away (Far, dark grey bars).

Table 1 Permutational analysis of variance (PERMANOVA) of assemblages of (a) Mollusca, (b) Crustacea, and (c) Polychaeta comparing marine reserves and exploited sites (MvE), with nested sites (MPA nested in reserves and Exploited 1 & 2 nested in Exploited) and distance from a reserve or exploited site (In, Near, Far), n = 4 replicate artificial habitats per site. Source

df

(a) Mollusca MvE 1 Distance 2 Site (MvE) 1 MvE  distance 2 Site(MvE)  distance 2 Residual 27 Total 35 Pairwise: Exploited 1 & 2: In = Near = Far, MPA: In – Near = Far (b) Crustacea MvE 1 Distance 2 Site(MvE) 1 MvE  distance 2 Site (MvE)  distance 2 Residual 27 Total 35 Pairwise: Exploited 1 & MPA: Near – In = Far, Exploited 2: In = Near = Far (c) Polychaeta MvE 1 Distance 2 Site(MvE) 1 MvE  distance 2 Site (MvE)  distance 2 Residual 27 Total 35 Pairwise: Exploited 1: Near – In = Far, Exploited 2 & MPA: In = Near = Far ns, p > 0.05; ⁄p < 0.05;

⁄⁄

p < 0.01;

⁄⁄⁄

p < 0.001.

MS

Pseudo-F

P (perm)

8928 3306 19,436 2166 2383 924

0.46 1.39 21.03 0.91 2.58

ns ns

7733 1550 6805 2169 3288 1047

1.14 0.47 6.50 0.66 3.14

ns ns

5762 4008 6372 6081 4500 1454

0.90 0.89 4.38 1.35 3.09

ns ns

⁄⁄⁄

ns ⁄⁄

⁄⁄⁄

ns ⁄⁄⁄

⁄⁄⁄

ns ⁄⁄⁄

2092

V.J. Cole et al. / Biological Conservation 144 (2011) 2088–2096

mussels at the far shore than in or near (SNK P < 0.01). There was also a significant interaction between distance and site for the cover of coralline turf (F4,60 = 11.46, P < 0.001). In the reserve site, at Dwesa, there was significantly greater cover of coralline turf at near and far shores than inside the reserve (Fig. 2a, SNK P < 0.01). There was no effect of distance for the two exploited sites or the other reserve site at Hluleka (Fig. 2b, SNK P > 0.05). PERMANOVA analyses for assemblages of molluscs, polychaetes and crustaceans, showed a significant interaction between site and distance (Table 1). At the first exploited site, Hole-in-the-Wall, there was no significant difference among assemblages of molluscs according to distance, but near assemblages of crustaceans and polychaetes significantly differed from assemblages that were in or far (Table 1a–c). At exploited site 2, Shixini, distance had no sig-

Mean (+SE) number of individuals

400

(a) Mollusca 300

200

100

0 Exploited 1

Exploited 2

MPA

nificant effect on assemblages of any of the three taxa (Table 1a–c). In the reserve site, there were significantly different assemblages of molluscs in the reserve compared to near or far from the reserve (Table 1a). Near to the reserve had significantly different assemblages of crustaceans from inside the reserve or far from it (Table 1b). Assemblages of polychaetes did not differ according to distance from the reserve (Table 1c). There were significantly more molluscs in the reserve site than in the exploited sites (Fig. 3, Table 2a). There was also a significant difference in the numbers of individuals of polychaetes between the reserve site and the exploited sites (Table 2c). Specifically, there were more polychaetes in the exploited sites than in the site with a reserve (Table 2c, Fig. 3). The number of individuals of crustaceans was significantly greater at the shore near the reserve than within the reserve or far from the reserve (Table 2b). Both organisms that are directly harvested and those that are indirectly affected showed a significant interaction between distance and the presence of a reserve (Target: F2,27 = 28.27, Non-target: F2,27 = 10.14, both P < 0.001). SNK tests showed that for the two exploited sites and the site with a marine reserve, there were no clear patterns with respect to distance (SNK P > 0.05, Fig. 4). There was no significant effect of marine reserves on the number of species of molluscs (Fig. 5a, F2,27 = 2.30, P > 0.05), crustaceans (Fig. 5b, F2,27 = 0.63, P > 0.05) or polychaetes (Fig. 5c, F2,27 = 1.15, P > 0.05). There was also no effect of distance for molluscs (F1,27 = 0.91, P > 0.05), crustaceans (F1,27 = 2.12, P > 0.05) or polychaetes (F1,27 = 0.19, P > 0.05). Numbers of individuals of mobile and sessile taxa were separated to determine if colonisation of artificial habitats was due to larval settlement or of adult/juvenile movement from the surrounding substratum. The number of individuals of molluscs, crustaceans and polychaetes were separated into those that are mobile

Mean (+SE) number of individuals

1000

(b) Crustacea 800

600

Source

400

200

0 Exploited 1

Exploited 2

MPA

250

Mean (+SE) number of individuals

Table 2 Analysis of variance (ANOVA) of the number of individuals in artificial habitats at each of three distances (In, Near and Far) in two exploited sites (E1 and E2) and a site with a marine reserve (MPA) using a Beyond BACI sampling design (details in text), n = 4.

(c) Polychaeta 200

150

100

50

0 Exploited 1

Exploited 2

MPA

Fig. 3. Mean (+SE; n = 4) number of individuals of (a) Mollusca, (b) Crustacea, and (c) Polychaeta at Hole-in-the-Wall (Exploited 1), Shixini (Exploited 2), and Dwesa (MPA 2). Shores were sampled in a reserve site or in exploited sites (In, black bars), 2 km away (Near, light grey bars) and 5 km away (Far, dark grey bars).

df

MS

(a) Mollusca Location 2 59,769 Between E1 and E2 1 MvE 1 Distance 2 27,831 Location  distance 4 9224 Di  between E1 and E2 2 Di  MvE 2 Residual 27 3147 SNK: Exploited 1 < Exploited 2, Exploited < MPA

26,800 92,737

8237 10,211

(b) Crustacea Location 2 162,683 Between E1 and E2 1 98,902 MvE 1 226,465 Distance 2 43,686 Location  distance 4 214,517 Di  between E1 and E2 2 88,170 Di  MvE 2 340,865 Residual 27 SNK: Exploited: In = Near = Far, MPA: In = Far < Near (c) Polychaeta Location 2 59,769 Between E1 and E2 1 2427 MvE 1 117,110 Distance 2 27,831 Location  distance 4 9224 Di  between E1 and E2 2 10,250 Di  MvE 2 8197 Residual 27 SNK: Exploited > MPA ns, p > 0.05; ⁄p < 0.05;

⁄⁄

p < 0.01;

⁄⁄⁄

p < 0.001.

F

P

8.52 29.47 3.02

⁄⁄

2.62 3.24

⁄⁄⁄

ns ns ns

5.86 13.41 0.20

⁄

5.22 20.19

⁄

⁄⁄⁄

ns

⁄⁄⁄

0.77 37.21 3.02

⁄⁄⁄

ns

3.26 2.60

ns ns

ns

2093

V.J. Cole et al. / Biological Conservation 144 (2011) 2088–2096

12

(a) Target 30 25 20 15 10 5

(a) Mollusca Mean (+SE) number of species

Mean (+SE) number of individuals

35

0

8 6 4 2 0

Exploited 1

Exploited 2

MPA

Exploited 1

1400

Exploited 2

12

(b) Non-target 1200 1000 800 600 400 200

Mean (+SE) number of species

Mean (+SE) number of individuals

10

0

MPA

(b) Crustacea

10 8 6 4 2 0

Exploited 2

MPA

Fig. 4. Mean (+SE; n = 4) number of individuals of: (a) target species, or (b) nontarget species affected at Hole-in-the-Wall (Exploited 1), Shixini (Exploited 2), and Dwesa (MPA 2). Shores were sampled in a reserve site or in exploited sites (In, black bars), 2 km away (Near, light grey bars) and 5 km away (Far, dark grey bars).

and those that are either sedentary or sessile. Sessile or sedentary crustaceans and polychaetes showed no clear effects of marine reserves (Crustacea F1,27 = 1.43, Polychaeta F1,27 = 2.84, both P > 0.05), or distance (Crustacea F2,27 = 1.71, Polychaeta F1,27 = 0.49, both P > 0.05). In the case of sessile or sedentary molluscs (these were nearly all P. perna), there were more in the exploited sites than in the reserve site (F1,27 = 12.59, P < 0.001). There were significantly more mobile molluscs (primarily gastropods) in the reserve site than in the exploited sites (Table 3). For polychaetes and crustaceans, there was a significant interaction between distance and MPA versus exploited sites (Table 3). Neither of these two taxa showed clear patterns of difference with respect to the presence of reserves or distance (Table 3).

Exploited 1

Exploited 2

18

Mean (+SE) number of species

Exploited 1

MPA

(c) Polychaeta

16 14 12 10 8 6 4 2 0 Exploited 1

Exploited 2

MPA

Fig. 5. Mean (+SE; n = 4) number of species of: (a) Mollusca, (b) Crustacea, and (c) Polychaeta at Hole-in-the-Wall (Exploited 1), Shixini (Exploited 2), and Dwesa (MPA 2). Shores were sampled in a reserve site or in exploited sites (In, black bars), 2 km away (Near, light grey bars) and 5 km away (Far, dark grey bars).

4. Discussion By studying a targeted bioengineer and its associated assemblages, we investigated a large proportion of diversity on intertidal rocky shores in South Africa (Hammond and Griffiths, 2004). The MPAs we studied were demonstrably effective at protecting standing stocks of adult mussels. There was a clear difference in covers of mussels and coralline algae between areas open to harvesting and the reserves, and the domination of space by these two taxa may represent two distinct alternative stable states (sensu Petraitis and Dudgeon, 1999), established as a result of anthropogenic disturbance through harvesting. Macroalgae and filter-feeders often compete for space (McQuaid and Branch, 1984), and red algal turfs can be maintained by the removal of mussels by predators such as whelks, lobsters and fish (e.g. Robles and Robb, 1993). Coralline turfs could persist outside reserves due to the collection of mussel

predators such as crayfish (e.g. Côte and Jelnikar, 1999; Siddon and Witman, 2004; Nicastro et al., 2007). Natural predation of adult mussels is, however, believed to be generally low on this coastline (Griffiths and Hockey, 1987), as experimentally cleared plots within reserves (i.e. with no further mussel collection) remain dominated by algae for >9 years (Dye, 1992), these do seem to represent alternative stable states (see Connell and Sousa, 1983). Although previous studies have shown marked differences between the species assemblages associated with mussels and with coralline algae (Lasiak, 1998; Chapman et al., 2005), we found no direct relationship between the cover of the two biogenic habitats and the associated fauna. The reserve site (Dwesa reserve and up to 5 km outside the reserve) was significantly different from the exploited sites in terms of densities of associated molluscs and polychaetes. For molluscs,

2094

V.J. Cole et al. / Biological Conservation 144 (2011) 2088–2096

Table 3 Analysis of variance (ANOVA) of the number of individuals of mobile taxa in artificial habitats at each of three distances (In, Near and Far) in two exploited sites (E1 and E2) and a site with a marine reserve (MPA) using a Beyond BACI sampling design (details in text), n = 4. Source

df

(a) Mollusca Location Between E1 and E2 MvC Distance Location  distance Di  between E1 and E2 Di  MvE Residual SNK Exploited < MPA

MS

2

2 4

p < 0.01;

2 2 27

⁄⁄⁄

9.21 31.50 3.44

ns

6919 9608

2.21 3.06

ns ns

101,183 251,932

6.66 16.59 0.26

⁄

71,845 330,432

4.73 21.76

⁄

980 6549

1.45 9.70 0.21

ns

18.27 4.29

⁄⁄⁄

28,889 98,765 28,453 8263

(c) Polychaeta Location 2 Between E1 and E2 1 MvE 1 Distance 2 Location  distance 4 Di  between E1 and E2 2 Di  MvE 2 Residual 27 SNK: Exploited 1: In < Far < Near, Exploited 2, MPA: In = Near = Far ⁄⁄

P

63,828 1 1

(b) Crustacea Location 2 Between E1 and E2 1 MvE 1 Distance 2 Location  distance 4 Di  between E1 and E2 2 Di  MvE 2 Residual 27 SNK Exploited 1: Near < In = Far, Exploited 2: In = Near = Far, MPA: In = Far < Near

ns, p > 0.05; ⁄p < 0.05;

F

⁄⁄⁄

ns

3135

176,558

52,758 201,139

⁄⁄⁄ ⁄

⁄⁄⁄

15,184

3764

1574 7614 12,334 2894

⁄⁄

ns

⁄

675

p < 0.001.

this pattern was driven by mobile species (mostly gastropods) rather than sessile or sedentary molluscs (only bivalves). Although gastropods are able to crawl into the artificial habitats from the surrounding biogenic habitats (e.g. Olabarria, 2002), this is unlikely to explain the observed pattern because: (1) inside the reserve, habitat is provided mostly by mussels, and near and far from the reserve, the habitat is mostly coralline algae, (2) mussels and coralline turf support distinctly different assemblages of associated fauna (Chapman et al., 2005; Borthagaray and Carranza, 2007), and (3) colonisation of artificial habitats is strongly influenced by the surrounding habitat (Smith and Rule, 2002; Cole et al., 2007). We propose that reserves act as a source of gastropods to shores outside and within 5 km of reserves, and that this is due to the export of gastropod larvae because adults are unlikely to travel up to 5 km (Branch, 1971; Underwood, 1977). There was, however, no evidence of export of bivalve larvae (particularly the target species, P. perna). These results suggest that marine reserves can be successful in maintaining and enhancing the overall biodiversity of nearby exploited areas, without necessarily increasing larval supply or densities of the target species outside reserves. This is likely to be linked to the relatively shorter dispersal distances of gastropods compared to bivalve larvae (Kinlan and Gaines, 2003). In other words, reserves provide export of larvae dispersing over relatively short scales, but longer-lived larvae are likely to be transported beyond the region where intertidal harvesting occurs. This highlights the importance of the match, or mismatch, between local oceanography, larval duration and the timing of reproduction (Cudney-Bueno et al., 2009; Carson et al., 2010). Overall, recruitment of mussels to artificial collectors was relatively low. This was not unexpected as recruitment of mussels on biogeographic scales within southern Africa is particularly low in the Transkei region (Harris et al., 1998), probably because of the low abundances of adults at mesoscales of 10s to 100s of kilome-

tres. There was no evidence of export of mussel larvae, with recruits arriving in similar numbers at reserves and non-reserves. The arrival of mussel recruits at non-reserve areas may be due to the presence of the reserve (or other reserves within 200 km of the study sites, Shanks, 1995), or to the existence of inaccessible stocks of adults outside reserves (e.g. subtidal populations or populations at the base of cliffs, Pelc et al., 2009). The observed distribution of adult mussels cannot, therefore, be explained by recruitment and must be caused by post-recruitment processes. This will obviously include direct mortality from harvesting, but may also reflect the inability of mussel recruits outside reserves to move from the most available settlement substratum (coralline algae) to the primary substratum or existing adult mussels. Recruits of P. perna show ontogenetic changes in mobility, with large recruits moving from algae to join adults in the laboratory (Erlandsson et al., 2008). Field evidence indicates that, however, large recruits disappear from macroalgae, but do not migrate to the primary substratum, suggesting they are lost from the population (Erlandsson et al., 2011). It was not expected that reserves would affect polychaetes because they are not harvested in the Transkei region. There were, however, fewer polychaetes in the site with a reserve than in the exploited sites. The presence of polychaetes may be an indirect effect of harvesting with changes to the biogenic habitats increasing densities of polychaetes compared to undisturbed habitats inside reserves. Previous studies have not shown greater densities of polychaetes to be associated with macroalgae than with mussels (Chapman et al. 2005; Borthagaray and Carranza, 2007), and polychaetes are also indicators of disturbed environments (Olsgard et al., 2003; Giangrande et al., 2005), such as sites open to harvesting. It was strictly necessary to have more than one exploited site to confirm that any differences attributed to the presence of the

V.J. Cole et al. / Biological Conservation 144 (2011) 2088–2096

reserve, were not due to spatial variability alone (Underwood, 1991, 1992, 1993). There were, indeed, strong differences between the two exploited sites for the entire assemblage of taxa, numbers of species and numbers of individuals. The exploited sites focussed around Hole-in-the-Wall were different from those at Shixini. It is unlikely that these differences were caused by an effect of the surrounding substratum as the cover of mussels of and coralline turf was similar between the two exploited sites. The observed differences may be due to spatial factors such as geography, with sites further north naturally having different assemblages from those in the south. Port St John’s lies only 55 km from the most northern sites at Hole-in-the-Wall, and has been defined as the biogeographic boundary between the warm temperate south coast province and the subtropical east coast province (Brown and Jarman, 1978). Moreover, there are regional differences within Transkei in the intensity of harvesting, with greater numbers of shellfish gatherers in the central region (Hole-in-the-Wall) than farther south, including Shixini (Hockey et al., 1988; Lasiak, 1997). Variability between the exploited sites may therefore be due to different intensities of harvesting and/or the influence of biogeography. Nevertheless, our experimental design allowed us to determine that differences between the site with a reserve and the exploited sites were due to the presence or absence of a reserve (Underwood, 1993). It is important to recognise that the effectiveness of reserves depends on compliance and that this in turn partly depends on how they are perceived by local fishing communities (Sale et al., 2005). It is also important to recognise that different reserves can have different aims so that their effectiveness should be measured against the expectations for that particular reserve (Halpern, 2003). In the case of target species that are bioengineers, such as mussels, MPAs can serve two functions: protection of the target species and promotion of biodiversity through the export of associated species. The effects of reserves are likely to differ dramatically among taxa, yet the potential effects of harvesting of intertidal resources on associated, non-target biodiversity are rarely considered. We found that the indirect effects of harvesting mussels on the diverse assemblages of associated molluscs, crustaceans and polychaetes were considerable. Rather than viewing the success of reserves, including evidence of larval export beyond reserve boundaries, from the perspective of only the target species, we should consider the broader effects of reserves on overall biodiversity. Here, we found that reserves were highly effective in protecting the primary target species, thus providing completely different habitats from areas outside reserves and enhancing overall regional biodiversity. Importantly, while there was no evidence that reserves provided export of larvae of the target species, they did provide larval export of non-target organisms. Thus, in the context of the conservation of biodiversity, we must consider these MPAs to be effective by providing export of the larvae of invertebrate taxa, even though they do not export larvae of the target species to nearby exploited shores. Thus, they fulfil some, but not all of the conservation aims of MPAs.

Acknowledgements We thank M. Nkaitshana & B. Plaatjie for assistance in the field, and Eastern Cape Parks Board and reserve managers for use of the reserves at Dwesa-Cwebe and Hluleka. We are grateful to M.J. Anderson for assistance with multivariate analyses. This work is based upon research supported by the South African Research Chairs Initiative of the Department of Science and Technology and National Research Foundation. Funding was provided by a Rhodes University Postdoctoral Fellowship (V.J.C.). We are grateful

2095

to G.M. Martins and an anonymous reviewer for suggestions that greatly improved an earlier version of the manuscript. References Anderson, M.J., Gorley, R.N., Clarke, K.R., 2009. PERMANOVA+ for PRIMER: Guide to Software and Statistical Methods. PRIMER-E, Plymouth, UK. Beier, P., Noss, R.F., 1998. Do habitat corridors provide connectivity? Conservation Biology 12, 1241–1252. Bélisle, M., 2005. Measuring landscape connectivity: the challenge of behavioral landscape ecology. Ecology 86, 1988–1995. Borthagary, A.I., Brazeiro, A., Gimenez, L., 2009. Connectivity and patch area in a coastal marine landscape: disentangling their influence on local species richness and composition. Austral Ecology 34, 641–652. Borthagaray, A.I., Carranza, A., 2007. Mussels as ecosystem engineers: their contribution to species richness in a rocky littoral community. Acta Oecologica – International Journal of Ecology 31, 243–250. Branch, G.M., 1971. The ecology of Patella Linnaeus from the Cape Peninsula, South Africa I. Zonation, movements and feeding. Zoologica Africana 6, 1–38. Brown, A.C., Jarman, N. (Eds.), 1978. Coastal Marine Habitats. The Hague. Bruno, J.F., Stachowicz, J.J., Bertness, M.D., 2003. Inclusion of facilitation into ecological theory. Trends in Ecology and Evolution 18, 119–125. Carson, H.S., Lopez-Duarte, P.C., Rasmussen, L., Wang, D., Levin, L.A., 2010. Reproductive timing alters population connectivity in marine metapopulations. Current Biology 20, 1926–1931. Chapman, M.G., People, J., Blockley, D., 2005. Intertidal assemblages associated with natural Corallina turf and invasive mussel beds. Biodiversity and Conservation 14, 1761–1773. Chapman, M.G., Underwood, A.J., Skilleter, G.A., 1995. Variability at different spatial scales between a subtidal assemblage exposed to discharge of sewage and two control assemblages. Journal of Experimental Marine Biology and Ecology 189, 103–122. Cole, V.J., Chapman, M.G., Underwood, A.J., 2007. Landscape and life-histories influence colonisation of polychaetes to intertidal biogenic habitats. Journal of Experimental Marine Biology and Ecology 348, 191–199. Cole, V.J., McQuaid, C.D., 2010. Bioengineers and their associated fauna respond differently to the effects of biogeography and upwelling. Ecology 91, 3549– 3562. Côte, I.M., Jelnikar, E., 1999. Predator-induced clumping behaviour in mussels (Mytilus edulis Linnaeus). Journal of Experimental Marine Biology and Ecology 235, 201–211. Cudney-Bueno, R., Lavín, M.F., Marinone, S.G., Raimondi, P.T., Shaw, W.W., 2009. Rapid effects of marine reserves via larval dispersal. PLoS ONE 4, e4140. Dye, A.H., 1992. Experimental studies of succession and stability in rocky intertidal communities subject to artisanal shellfish gathering. Netherlands Journal of Sea Research 30, 209–217. Erlandsson, J.F., McQuaid, C.D., Stanczak, S., 2011. Recruit/algal interaction prevents recovery of overexploited mussel beds: indirect evidence that post-settlement mortality structures mussel populations. Estuarine Coastal and Shelf Science 92, 132–139. Erlandsson, J., Porri, F., McQuaid, C.D., 2008. Ontogenetic changes in small-scale movement by recruits of an exploited mussel: implications for the fate of larvae settling on algae. Marine Biology 153, 365–373. Fahrig, L., Merriam, G., 1994. Conservation of fragmented populations. Conservation Biology 8, 50–59. Fischer, J., Lindenmayer, D.B., 2007. Landscape modification and habitat fragmentation: a synthesis. Global Ecology and Biogeography 16, 265–280. Giangrande, A., Licciano, M., Musco, L., 2005. Polychaetes as environmental indicators revisited. Marine Pollution Bulletin 50, 1153–1162. Gobin, J.F., Warwick, R.M., 2006. Geographical variation in species diversity: a comparison of marine polychaetes and nematodes. Journal of Experimental Marine Biology and Ecology 330, 234–244. Griffiths, C.L., Hockey, P.A.R., 1987. A model describing the interactive roles of predation, competition and tidal elevation in structuring mussel populations. South African Journal of Marine Science 5, 547–556. Halpern, B.S., 2003. The impact of marine reserves: do reserves work and does reserve size matter? Ecological Applications 13, S117–S137. Halpern, B.S., Warner, R.R., 2002. Marine reserves have rapid and lasting effects. Ecology Letters 5, 361–366. Halpern, B.S., Warner, R.R., 2003. Matching marine reserve design to reserve objectives. Proceedings of the Royal Society of London Series B – Biological Sciences 270, 1871–1878. Hammond, W., Griffiths, C.L., 2004. Influence of wave exposure on South African mussel beds and their associated infaunal communities. Marine Biology 144, 547–552. Hanski, I., 1998. Metapopulation dynamics. Nature 396, 41–49. Harris, J.M., Branch, G.M., Elliot, B.L., Currie, B., Dye, A., McQuaid, C.D., Tomalin, B.J., Velasquez, C., 1998. Spatial and temporal variability in recruitment of intertidal mussels around the coast of Southern Africa. South African Journal of Zoology 33, 1–11. Hockey, P.A.R., Bosman, A.L., Siegfried, W.R., 1988. Patterns and correlates of shellfish exploitation by coastal people in Transkei: an enigma of protein production. Journal of Applied Ecology 25, 353–363.

2096

V.J. Cole et al. / Biological Conservation 144 (2011) 2088–2096

Jones, C.G., Lawton, H., Shachak, M., 1994. Organisms as ecosystem engineers. Oikos 69, 373–386. Jones, C.G., Lawton, J.H., Shachak, M., 1997. Positive and negative effects of organisms as physical ecoystem engineers. Ecology 78, 1946–1957. Kelaher, B.P., Chapman, M.G., Underwood, A.J., 2001. Spatial patterns of diverse macrofaunal assemblages in coralline turf and their associations with environmental variables. Journal of the Marine Biological Association of the United Kingdom 81, 917–930. Kinlan, B.P., Gaines, S.D., 2003. Propagule dispersal in marine and terrestrial environments: a community perspective. 84, 2007–2020. Lasiak, T., 1991. The susceptibility and/or resilience of rocky littoral molluscs to stock depletion by the indigenous coastal people of Transkei, Southern Africa. Biological Conservation 56, 245–264. Lasiak, T., 1992. Contemporary shellfish-gathering practices of indigenous coastal people in Transkei: some implications for interpretation of the archaeological record. Suid-Afrikaanse Tydskrif vir Wetenskap 88, 19–29. Lasiak, T., 1997. Temporal and spatial variations in the pattern of shoreline utilization in a region subject to subsistence exploitation. International Journal of Environmental Studies 52, 21–46. Lasiak, T., 1998. Multivariate comparisons of rocky infratidal macrofaunal assemblages from replicate exploited and non-exploited localities on the Transkei coast of South Africa. Marine Ecology Progress Series 167, 15–23. Lasiak, T., Dye, A., 1989. The ecology of the brown mussel Perna perna in Transkei, Southern Africa: Implications for the management of a traditional food resource. Biological Conservation 47, 245–257. Lasiak, T., Field, J.G., 1995. Community-level attributes of exploited and nonexploited rocky infratidal macrofaunal assemblages. Journal of Experimental Marine Biology and Ecology 185, 33–53. Lindborg, R., Eriksson, O., 2004. Historical landscape connectivity affects present plant species diversity. Ecology 85, 1840–1845. McNeill, S.E., Fairweather, P.G., 1993. Single large or several small marine reserves? An experimental approach with seagrass fauna. Journal of Biogeography 20, 429–440. McQuaid, C.D., Branch, G.M., 1984. The influence of sea temperature, substratum and wave exposure on rocky intertidal communities: an analysis of faunal and floral biomass. Marine Ecology Progress Series 19, 145–151. Nicastro, K.R., Zardi, G.I., McQuaid, C.D., 2007. Behavioural response of invasive Mytilus galloprovincialis and indigenous Perna perna mussels exposed to risk of predation. Marine Ecology – Progress Series 336, 169–175. Olabarria, C., 2002. Role of colonization in spatio-temporal patchiness of microgastropods in coralline turf habitat. Journal of Experimental Marine Biology and Ecology 274, 121–140. Olsgard, F.T.B., Holthe, T., 2003. Polychaetes as surrogates for marine biodiversity: lower taxonomic resolution and indicator groups. Biodiversity and Conservation 12, 1033–1049. Paine, R.T., Levin, S.A., 1981. Intertidal landscapes: disturbance and the dynamics of pattern. Ecological Monographs 51, 145–178. Palumbi, S.R., 2004. Marine reserves and ocean neighborhoods: The spatial scale of marine populations and their management. Annual Review of Environmental Resources 29, 31–68.

Parravicini, V., Thrush, S.F., Chaintore, M., Morri, C., Croci, C., Bianchi, C.N., 2010. The legacy of past disturbance. chronic angling impairs long-term recovery of marine epibenthic communities from acute date-mussel harvesting. Biological Conservation 143, 2435–2440. Pelc, R.A., Baskett, M.L., Tanci, T., Gaines, S.D., Warner, R.R., 2009. Quantifying larval export from South African marine reserves. Marine Ecology Progress Series 394, 65–78. Pelc, R.A., Warner, R.R., Gaines, M.S., Paris, C.B., 2010. Detecting larval export from marine reserves. Proceedings of the National Academy of Sciences of the USA 107(43), 18266–18271. Petraitis, P.S., Dudgeon, S.R., 1999. Experimental evidence for the origin of alternative communities on rocky intertidal shores. Oikos 84, 239–245. Roberts, C.M., Bohnsack, J.A., Gell, F., Hawkins, J.P., Goodridge, R., 2001. Effects of marine reserves on adjacent fisheries. Science 294, 1920–1923. Robles, C., Robb, J., 1993. Varied carnivore effects and the prevalence of intertidal algal turfs. Journal of Experimental Marine Biology and Ecology 166, 65–91. Rowley, R., 1994. Marine reserves in fisheries management. Aquatic Conservation: Marine and Freshwater Ecosystems 4, 233–254. Russ, G.R., Angel, C.A., 2011. Enhanced biodiversity beyond marine reserve boundaries: the cup spillith over. Ecological Applications 21, 241–250. Sale, P.F., Cowen, R.K., Danilowivcz, B.S., Jones, G.P., Kritzer, J.P., Lindeman, K.C., Planes, S., Polunin, N.V.C., Russ, G.R., Sadovy, Y.J., Steneck, R.S., 2005. Critical science gaps impede use of no-take fishery reserves. Trends in Ecology and Evolution 20, 74–80. Seed, R., 1996. Patterns of biodiversity in the macro-invertebrate fauna associated with mussel patches on rocky shores. Journal of the Marine Biological Association of the United Kingdom 76, 203–210. Shanks, A.L., 1995. Mechanisms of cross-shelf dispersal of larval invertebrates and fish. Biological Bulletin 170, 429–440. Siddon, C.E., Witman, J.D., 2004. Behavioral indirect interactions: Multiple predator effects and prey switching in the rocky subtidal. Ecology 85, 2938–2945. Smith, S.D.A., Rule, M.J., 2002. Artificial substrata in a shallow sublittoral habitat: do they adequately represent natural habitats or the local species pool? Journal of Experimental Marine Biology and Ecology 277, 25–41. Connell, J.H., Sousa, W.P., 1983. On the evidence needed to judge ecological stability or persistence. The American Naturalist 121, 789–824. Underwood, A.J., 1977. Movements of intertidal gastropods. Journal of Experimental Marine Biology and Ecology 26, 191–210. Underwood, A.J., 1991. Beyond BACI: experimental designs for detecting human environmental impacts on temporal variations in natural populations. Australian Journal of Marine and Freshwater Research 42, 569–587. Underwood, A.J., 1992. Beyond BACI: the detection of environmental impacts on populations in the real, but variable, world. Journal of Experimental Marine Biology and Ecology 161, 145–178. Underwood, A.J., 1993. The mechanics of spatially replicated sampling programmes to detect environmental impacts in a variable world. Australian Journal of Ecology 18, 99–116. Underwood, A.J., Chapman, M.G., 2006. Early development of subtidal macrofaunal assemblages: relationships to period and timing of colonization. Journal of Experimental Marine Biology and Ecology 330, 221–233.