Environmental Management DOI 10.1007/s00267-015-0552-7

The Impact of Trampling on Reef Macrobenthos in Northeastern Brazil: How Effective are Current Conservation Strategies? Gleice S. Santos1 • Douglas C. Burgos2 • Simone M. A. Lira1 • Ralf Schwamborn1

Received: 16 October 2014 / Accepted: 8 June 2015 Ó Springer Science+Business Media New York 2015

Abstract Tropical reefs are used for intensive tourism in various parts of the world. However, few studies have investigated the effect of regular trampling on these fragile ecosystems. The aim of this study was to assess the effect of different conservation strategies (open access, partial protection, and total long-term closure) on intertidal reef tops in Porto de Galinhas and Tamandare´, Pernambuco State, Brazil. Analysis of the macrobenthic community was performed with photo transects and image analysis (CPCe). Twenty-seven transects were surveyed from January to August 2012, in intensively impacted (I) open-access sites, in partially protected (P) sites with occasional, illegal trampling, and in a permanently closed (C) site. In I sites, total live cover was half the cover found in adjacent P sites. The area of bare rock averaged 53.6 and 25.0 % in I and P sites, respectively. In the C site, the area of bare rock was only 19.8 %. In I and P sites, macroalgae (Palisada perforata) were dominating, while in the C site, the zoanthid Zoanthus sociatus was most abundant. Shell-bearing vermetids (Petaloconchus varians) and bivalves (Isognomon bicolor) were more abundant at the C site, being possible bioindicators for areas with zero or little trampling. Twelve years of total closure produced near-pristine communities in the C site, dominated by zoanthids and fragile mollusks. This study showed that trampling has severe and

& Ralf Schwamborn [email protected] 1

Department of Oceanography, Federal University of Pernambuco (UFPE), Cidade Universita´ria, s/n, Recife, PE 50670-901, Brazil

2

Department of Biology, Federal Rural University of Pernambuco (UFRPE), Dois Irma˜os, Recife, PE 52171-900, Brazil

long-lasting consequences for the structure of these ecosystems. Keywords Human impact  Reef environments  Marine protected areas  Tropical Atlantic

Introduction Coastal reef environments of shallow tropical waters are characterized by their high diversity and productivity, and the presence of habitat-forming organisms, such as macroalgae and sessile animals, that play key roles in these ecosystems (Kaplan 1982; Sheppard et al. 2009). These reefs are the target of various human activities, such as fishing, ports, urbanism, and tourism. Due to their ecological, social, cultural, and economic importance, researchers worldwide have been intensively investigating these environments. Several monitoring programs aim at assessing these ecosystems and at the search for optimal strategies for their sustainable management and use (Rogers et al. 1983, 1994; Westmacott et al. 2000; Hill and Wilkinson 2004; Bacalso and Wolff 2014). Only a few studies have investigated the responses of benthic fauna and flora to human impacts in Brazilian reefs. Recreational activities such as diving were shown to have severe consequences for the health of Brazilian reefs (Maida and Ferreira 1997; Lea˜o and Dominguez 2000; Oigman-pszczol and Creed 2011; Sarmento and Santos 2012). In their study, Oigman-Pszczol and Creed (2011) showed the drastic consequences for subtidal corals and macroalgae derived from recreational boat traffic and the inputs of solid and liquid waste at the subtropical beach of Bu´zios (Rio de Janeiro State). Furthermore, trampling has

123

Environmental Management

been shown to affect meiofaunal communities on the reef tops of Porto de Galinhas, Pernambuco State (Sarmento et al. 2011; Sarmento and Santos 2012). In spite of their intensive use and huge socio-economic and ecological importance, there are no data available on the effects of trampling on sessile macrobenthic organisms of Brazilian reefs. These communities include several species that are responsible for the building and erosion processes that determine the shape and dynamics of the coastline. There are no studies available that compare different conservation strategies for Brazilian tropical reef tops. These ecosystems usually do not present any fragile branching corals, but high abundances of macroalgae and zoanthids (Costa Jr et al. 2001), as compared to IndoPacific reefs, and could therefore possibly be less vulnerable to trampling than coral-dominated reef ecosystems. New methodologies have been developed to increase the accuracy and efficiency of research in reef environments, at low cost and being non-destructive. The photo transect method has been shown to be a useful tool that combines these advantages (Preskitt et al. 2004; Dumas et al. 2009), especially when used together with image analysis software for marine benthos, such as the CPCe package (Kohler and Gill 2006; Dumas et al. 2009). The objective of this study is to investigate whether trampling at present levels of intensity has an influence on sessile macrobenthos, using the photo transect method, and tests the hypothesis that this activity causes significant damage to nearshore reef ecosystems in northeastern Brazil. Furthermore, three different conservation strategies (open access, partial protection, and total closure) are being compared regarding their impacts on live cover, species composition, and diversity of sessile macrobenthos.

Materials and Methods Study Area Field activities were carried out on the reef tops of Porto de Galinhas (8° 59 0 00 00 –8° 330 3300 S and 35° 000 2700 –34° 590 0000 W) and Tamandare´ (8° 450 3600 and 8° 470 2000 S and 35° 030 4500 and 35°060 4500 W). Both sites are located in the State of Pernambuco, northeastern Brazil (Fig. 1). These two reef areas were chosen as they represent two extremes in trampling impact: Porto de Galinhas is one of the most intensively visited reef areas in Brazil, while the closed area in Tamandare´ is the only nearshore tropical reef top in Brazil that is totally protected from fishermen and tourists. The coral reefs of Porto de Galinhas are one of the main tourist attractions in Brazil, with a year-round intense visitation of reef tops and tidal pools. Porto de Galinhas beach receives about 900,000 tourists per year from Brazil

123

Fig. 1 Location of the study areas (Porto de Galinhas and Tamandare´) in northeastern Brazil

and abroad. In the peak season (December–February), up to 6200 tourists trample on the reef tops per month, while in the off-peak season, 1800 tourists visit the reef tops per month (E. Lacerda, Municipality of Ipojuca, personal communication). The number of tourists who walk on the reef tops of Porto de Galinhas has been estimated as high as 1020 visitors on a single low tide (Sarmento and Santos 2012), massed on a open-access reef area of about 300 m 9 50 m size (Fig. 2). The tourists reach the reef tops at low tides, by walking through the water up to their chest or on sailing rafts (at present there are 85 such tourist rafts registered in the local City Hall). To promote the conservation of these ecosystems, some parts of the reef tops have been isolated since 2009 for the creation of partially protected areas (Fig. 2). These partially protected areas (P) correspond to approximately 85 % of the total reef area of Porto de Galinhas. Before being allowed to access the reefs, all tourists have to visit a stand of environmental education, where they have to watch a video on the importance of the coral reefs and receive guidance about how to behave in the reef areas. After watching this video, each tourist receives an identification wristband that allows them access to the reef, that is monitored by environmental agents. The partially protected areas (P) are immediately adjacent to the intensively impacted, open-access areas (I), and thus suffer accidental trampling by tourists who stray into them from the impacted (I) areas because they do not realize or do not respect the signaling buoys and ropes, in spite of a frequent monitoring by environmental agents on the reefs. A recent study on trampling intensity in Porto de Galinhas showed values of up to 476 people trampling on a 6 m 9 10 m area per day inside the impacted area (i.e., up to 7.9 ind. m-2 day-2), while at the same time, only ten people per day were observed in the adjacent partially protected area (less than 0.002 ind. m-2 day-2, Santos et al., in prep.). The coral reef complex of Tamandare´ is located within the limits of the ‘‘Costa dos Corais’’ MPA, the first

Environmental Management

Fig. 2 Map of the study sites in Porto de Galinhas and Tamandare´, northeastern Brazil. I1 and I2 impacted sites under intense daily trampling by tourists. P1–P4 partially protected sites, where occasional trampling by tourists and fishermen occurs. C closed site with no trampling by tourists and fishermen. The position of the photo transects is indicated as white squares

Brazilian Federal Marine Protected Area with coastal reefs. There, a closed area (C) was defined for the reefs of Barra Island (Fig. 2), which has been fully protected from any recreational visitation and any fishing since 1999.

Fieldwork Seven field campaigns were conducted from January to August 2012. The reef sites were classified into three categories: I—intensively impacted (daily intensive trampling by tourists), P—partially protected (occasional trampling by tourists and fishermen), and C—closed (zero trampling by tourists or fishermen). In Porto de Galinhas, two heavily impacted sites (I1 and I2) suffer intense and

regular daily trampling by tourists. Four partially protected sites (P1, P2, P3, and P4) were defined in areas where visitation by tourists was prohibited since 3 years. Nevertheless, as they are immediately adjacent to openaccess areas, occasional (illegal) trampling occurs in these partially protected sites (Fig. 2). On the reef tops of Tamandare´, sampling was conducted in a closed site (C), where access was strictly controlled, and visitation has been limited to few scientists, with special permits, since 12 years. The percentage of macrobenthic live cover was assessed using the non-destructive photo transect method. Twentyseven fixed transects of 10 m length were placed in the impacted (6 transects), in the partially protected (12 transects) and in the closed (9 transects) sites, respectively (Fig. 2). Each transect consisted of ten regularly spaced PVC frames of 50 9 50 cm. Within each site, parallel transects were at least 2 m apart from each other, and at least 2 m away from large tidal pools. Transects placed in the partially protected areas were at least 2 m away from the border to the impacted areas. All transects were placed in flat, central areas of the reef tops, with similar topography and similar height above sea level. Beginning in the impacted area, two sites were chosen randomly (Fig. 2), and three transects were placed inside each of both impacted sites (six transects, i.e., sixty photo quadrats). Then four sites were chosen in the partially protected area, adjacent to the impacted sites, and three transects were placed inside each site (twelve transects, i.e., 120 photo quadrats). Conservation of the closed area in Tamandare´ prohibited a widespread placement of study sites throughout the reef top. Thus, to minimize trampling during the surveys, only one site, that best resembles the sites in Porto de Galinhas, was chosen randomly in a flat, central part of the reef top, where nine transects were placed (i.e., 90 photo quadrats). Only two well-trained researchers accessed the closed reef during the surveys. Once the transects and photo quadrats had been defined, the same plots were visited each time in all areas (fixed plot monitoring). A total of 270 quadrats along 27 transects were photographed during each field campaign. The inner area of a PVC frame (50 cm 9 50 cm) was photographed with a Canon 12 MP camera that was held at 1.2 m height above the reef surface (Rogers et al. 1994; Hill and Wilkinson 2004; Preskitt et al. 2004). Roughness of the reef tops along each transect was measured using a chain (size of the chain links: 32.0 mm 9 8.9 mm) and a metric tape, through the ratio of chain length and linear length. For each transect (10 m), five measurements of roughness were performed every 2 m (Rogers et al. 1983). To allow for a proper identification, small fragments of macroalgae were sampled close to the transects at each

123

Environmental Management

campaign and frozen. Also samples of vermetid mollusks (15 individuals) were taken and preserved in 80 % ethanol. Species Identification and Image Processing In the laboratory, macroalgae and vermetids were identified to species level under a stereo microscope. After an initial survey of all species present in the images, a database was created in the CPCe software, with all categories listed. Macroalgae that could not be identified to genus or species level in the analysis of images were classified as ‘‘turf.’’ For each image, 30 points were defined randomly, and each point was visually characterized within the 22 categories in the database (Kohler and Gill 2006). A total of 1890 photographs were taken (270 quadrats photographed on seven occasions), of which 1811 images were of sufficient quality for analysis with CPCe. A total of 54,330 points were randomly placed on these images (30 points per image) and visually analyzed. Macrobenthic species or types of substrate (bare rock, rubble, and sand) were visually identified on each point. Data Analysis The Shannon–Wiener index (H0 ) was calculated to assess the diversity of each photo quadrat, based on log2 abundance (i.e., live cover) of all taxa (Shannon 1948). Prior to analysis, all data were tested for normality and homoscedasticity using the Shapiro–Wilk and Bartlett tests (Zar 1996). Since several data sets did not show normal distribution and homogeneity of variances (even after arcsine transformation), non-parametric methods were used (Mann–Whitney U tests and Kruskal–Wallis ANOVA, Zar 1996). Kruskal–Wallis ANOVA was used to evaluate a possible effect of trampling on total live cover, the area of bare rock, live cover of the most abundant sessile macrobenthic organisms, and diversity. Variables were checked for collinearity (Person’s r \ 0.9) prior to all analyses (Zar 1996). These variables were compared between conservation strategies (n = 3) and sites (n = 7). Subsequently, Mann–Whitney tests were applied to compare groups (e.g., data from partially protected vs impacted sites) a posterior. ANOSIM with 999 permutations was used to evaluate whether the different groups (e.g., areas or months) found in the MDS are significantly different in their taxonomic composition. SIMPER was used to identify the species that most contributed to the similarities between images in the data set (Clarke and Warwick 2001). Analyses were performed using the ‘‘R’’ statistical computing environment (R Development Core Team 2005) and the PRIMER 6.0 software (Clarke and Gorley 2006).

123

Results Ten taxa of invertebrates were identified in the phototransects, including poriferans, cnidarians, mollusks, and echinoderms (Table 1). Among the cnidarians, three species of zoanthids were identified, including the dominant Zoanthus sociatus (Ellis, 1768). Two species of scleractinian corals (Favia gravida Verrill, 1868 and Siderastrea stellata Verrill, 1868) were identified, but the cover of these corals was negligible (average cover \0.2 %) in all areas (Table 1). These scleractinians were generally observed inside small, shallow tidal pools, whereas zoanthids were often exposed to the air on low tides. Seven categories of macroalgae could be consistently distinguished in the phototransects, including ‘‘turf’’ algae (Table 1). Microscopic examination of qualitative samples revealed a total of 27 taxa of macroalgae (Table 2), including six rhodophytes that contributed to multispecies algal turfs (taxa marked ‘‘T’’ in Table 2). The impacted sites (I1 and I2) were conspicuously dominated by bare, polished rock surfaces (Fig. 3). In these impacted sites, most (53.6 %) of the photographed area was filled with bare rock. The most abundant organisms in these impacted sites were two species of macroalgae, the rhodophyte Palisada perforata (Bory) KW Nam (17 %), and the calcareous alga Halimeda opuntia (L.) J. V. Lamour (6 %). The zoanthid Zoanthus sociatus (Ellis, 1768) represented only 5.3 % of the area in impacted sites (Fig. 3). A total of 16 species of macroalgae and 9 species of invertebrates were found in impacted sites, in addition to algae classified as ‘‘turf’’ and unidentified calcareous algae (Tables 1, 2). In the partially protected sites (P1–P4), bare rock was found on 25.0 % of the reef surface, about half the area of bare rock found in impacted sites (Fig. 3). The three most abundant organisms in these sites were the same as in impacted sites, but with considerably higher live cover: P. perforata (29.8 %), Z. sociatus (23.6 %), and H. opuntia (7.4 %). The vermetid mollusk Petaloconchus varians (2.4 %) was more than twice as abundant in partially protected sites than in impacted sites (0.6 %) (Fig. 3). Sixteen species of macroalgae and 10 species of invertebrates were also recorded in the partially protected sites. Furthermore, the bivalve Isognomon bicolor (C.B. Adams, 1845), that was absent in impacted sites, was detected in partially protected sites, but in low numbers (13 individuals) and with only 0.1 % average cover (Tables 1, 2). In the closed site off Tamandare´, the area of bare rock (19.3 %) was even smaller than in partially protected sites (Fig. 3). The most abundant organism was the zoanthid Z. sociatus (30.9 %). Other important organisms in the closed site were the algae P. perforata (16.2 %) and H. opuntia

Environmental Management Table 1 Live cover (mean, minimum–maximum, %) of the most abundant organisms and substrate types found on reef tops under different trampling intensities in Porto de Galinhas and Tamandare´, Pernambuco State, Brazil

Impacted

Protected

Closed

Caulerpa racemosa

0.01 (0–6.7)

0.14 (0–57)

0.33 (0–47)

Dictyosphaeria verluysii

0.63 (0–20)

2 (0–33)

2.86 (0–37)

Halimeda opuntia

6.29 (0–73)

7.54 (0–93)

3.9 (0–57)

Chlorophyta

Rhodophyta Calcareous algae

0.04 (0–6.7)

0.02 (0–3.4)

0.01 (0–3.3)

Gelidiella acerosa

1.67 (0–37)

1.83 (0–40)

0.33 (0–40)

Palisada perforata

17.7 (0–77)

27.5 (0–100)

15.9 (0–77)

Dictyota pulchella

0

0.008 (0–3.4)

0.01 (0–6.7)

‘‘Turf algae’’

4.16 (0–43)

4.72 (0–47)

6.56 (0–37)

Phaeophyta

Porifera Cinachyrella alloclada

0.01 (0–6.7)

0.12 (0–6.7)

0.05 (0–3.3)

Cliona sp.

0.007 (0–3.3)

0.03 (0–13)

0.67 (0–13)

Favia gravida

0.01 (0–3.3)

0

0.004 (0–3.3)

Siderastrea stellata

0.15 (0–6.7)

0.05 (0–6.7)

0.004 (0–3.3)

Cnidaria Scleractinia

Zoanthidea Zoanthus sociatus

5.18 (0–40)

23.3 (0–87)

30.5 (0–80)

Palythoa caribaeorum

0.007 (0–3.3)

0.004 (0–3.3)

0.52 (0–53)

Protopalythoa variabilis

1.78 (0–30)

1.26 (0–23)

3.02 (0–37)

Isognomon bicolor

0

0.14 (0–13)

0.41 (0–13)

Petaloconchus varians

0.61 (0–23)

2.32 (0–57)

13.4 (0–80)

0.9 (0–10)

0.42 (0–10)

0.9 (0–13)

Mollusca

Echinodermata Echinometra lucunter Substrate Bare rock

51.7 (0–93)

24.2 (0–87)

19.3 (0–70)

Rubble

8.63 (0–53)

3.83 (0–57)

0.96 (0–20)

Sand

0.16 (0–13)

0.19 (0–13)

0.009 (0–3.3)

Based on phototransects analyzed using the CPCe software

(3.8 %). The fragile, shell-bearing vermetid P. varians (13.8 %) was approximately seven times more abundant in the closed site than in partially protected sites and more than ten times more abundant than in impacted sites (Fig. 3). The fragile, shell-bearing bivalve I. bicolor was also more abundant in the closed site than in the other sites (86 individuals, 0.4 %). The closed site presented 13 species of macroalgae and 10 species of sessile animals (Tables 1, 2). Trampling intensity had a significant effect on the total live cover and on the area of bare rock (Kruskal–Wallis ANOVA, P \ 0.0001; Fig. 4; Table 3). The impacted area had a much lower live cover than the partially protected and closed areas (Fig. 4). This pattern was even clearer for the area of bare rock (Fig. 5). Pairwise comparisons showed that the impacted, partially protected, and closed

areas differed significantly from each other regarding these parameters (Mann–Whitney tests, P \ 0.0001; Figs. 4 and 5; Table 3). The area of bare rock and total live cover were significantly correlated (r = -0.7, P \ 0.0001, n = 1811) _ but they were not collinear ðr \ 0:9Þ. The rhodophyte P. perforata showed significant differences in cover between trampling intensities (Kruskal– Wallis ANOVA, P \ 0.0001, Fig. 6). Cover of this rhodophyte was significantly higher in partially protected sites than in impacted and closed sites (Mann–Whitney, P \ 0.0001), since it was replaced by bare rock in impacted sites and by zoanthids in the closed area. The calcareous algae H. opuntia also showed differences between trampling intensities (Kruskal–Wallis ANOVA, P \ 0.0001, Fig. 6). There were no differences in H. opuntia cover between impacted and partially protected

123

Environmental Management Table 2 List of macroalgae collected on reef tops under different trampling intensities in Porto de Galinhas and Tamandare´, Pernambuco State, Brazil I

P

C

Chlorophyta Acanthophora spicifera

9

Acetabularia sp. Anadyomene stellata

9 9

Bryopsis pennata

9

Bryopsis plumosa

9

9

Caulerpa racemosa

9

9

Cladophora vagabunda

9

9

Dictyosphaeria verluysii

9

9

Caulerpa microphysa

9

Dictyosphaeria sp. Halimeda opuntia

9 9 9

9

9

9

Rhodophyta Amphiroa sp.(T)

9

Antithamnion sp.

(T)

9

Bryothamnion seaforthii

9

Ceramium sp.(T)

9

9

Gelidiella acerosa Gelidiopsis variabilis

9 9

9

Gelidium pusillum

9

9

Gracilaria sp.

9

9

Hypnea musciformis

9

Jania sp.(T)

9

Jania adhaerens(T)

9

9

Palisada perforata

9

9

(T)

9

Polysiphonia sp.

9 9

Tricleocarpa cylindrica

protected and impacted sites. The differences between sites were highly significant (Kruskal–Wallis ANOVA, P \ 0.0001, Mann–Whitney tests, P \ 0.0001). The results of ANOSIM (R = 0.392) confirmed the existence of significant differences in species composition between all sites and months. The MDS plots generally presented a clear segregation between the impacted, partially protected, and closed sites, especially in January (Fig. 7). SIMPER analysis showed that the organisms that most contributed to the separation of sites with different degrees of trampling were P. perforata, Z. sociatus, H. opuntia, and P. varians. These are thus most likely indicators for trampling, where the macroalgae (P. perforata and H. opuntia) are likely indicators for trampled environments, while the fragile, sessile animals (e.g., Z. sociatus and P. varians) are likely indicators for less impacted reef tops in northeastern Brazil. The diversity of sessile macrobenthos showed to be significantly affected by trampling intensity (Kruskal– Wallis ANOVA, P \ 0.0001, Fig. 8). There were significant differences in diversity between all areas (Mann– Whitney tests, P \ 0.001). The roughness of the reef surface did not show any significant differences between sampling sites or areas (Kruskal–Wallis ANOVA, P [ 0.05), showing that the observed differences in live cover and species composition between areas were not due to any structural differences between areas, but most likely due to the different conservation status of the three areas.

9

Discussion

9

Phaeophyta Dictyota pulchella

9

9

Based on qualitative samples that were taken manually and analyzed under a stereo microscope. I: impacted, P: partially protected, C: closed sites. Taxa marked with ‘‘(T)’’ were part of algal turfs

sites. However, the closed site showed a lower cover of these robust algae than the remaining sites (Mann–Whitney tests, P \ 0.0001, Fig. 6). The cover of Z. sociatus showed a clear relation to the protection status of the reef areas. The impacted sites had the lowest Z. sociatus cover, while the closed site presented the highest cover of this zoanthid (Mann–Whitney tests, P \ 0.0001, Fig. 6). The difference between trampling intensities with different protection status (Fig. 6) was highly significant (Kruskal–Wallis ANOVA, P \ 0.0001). The protection status of the reef areas also had a clear and significant effect on the cover of the vermetid P. varians (Fig. 6). The closed site presented a cover approximately seven times larger than the partially

123

This study has brought new insights into the effectiveness of current conservation strategies for Brazilian reef top communities. Furthermore, the most vulnerable (mollusks and zoanthids) and the most robust (macroalgae) bioindicators could be identified. The different conservation strategies (open access, partial protection, and total closure) produced characteristic reef habitats and communities. A direct negative influence of trampling on sessile macrobenthos was evident in many parameters such as total live cover, area of bare rock, and total diversity. Similarly, the damage caused by trampling on Australian reefs also showed a strong relationship between the intensity of use of these environments and community structure (Kay and Liddle 1989). On reef tops with a predominance of scleractinian corals, trampling caused a long-lasting damage to these fragile reef-building organisms. The most conspicuous effect of trampling in the present study was the dominance of vast, polished rock surfaces without any macrobenthic organisms in the impacted sites

Environmental Management Fig. 3 Average relative abundance of the main organisms and area of bare rock on reef tops under different trampling intensities in Porto de Galinhas and Tamandare´ (Pernambuco State, Brazil). I impacted, P partially protected, C closed sites

Fig. 4 Area of total live cover on reef tops under different trampling intensities in Porto de Galinhas and Tamandare´ (Pernambuco State, Brazil). A Live cover by site. B Live cover by trampling intensities.

Table 3 Live cover (%), diversity (H0 ), and roughness (m) (mean, minimum– maximum) of the sampled areas on reef tops under different trampling intensities in Porto de Galinhas and Tamandare´, Pernambuco State, Brazil

I impacted. P partially protected. C closed sites. N number of analyzed photo quadrats. a–e Groups of samples defined by pairwise Mann–Whitney tests

Site

n

Live cover

Bare rock

Diversity (H0 )

n

Roughness

I1

210

39.4 (6.6–83.3)

59.1 (16.6–93.3)

1.3 (0.2–2.3)

15

1.1 (1.03–1.2)

I2

210

39.6 (3.3–96.6)

61.3 (3.3–96.6)

1.3 (0.2–2.5)

15

1.1 (1.03–1.1)

P1

210

88.6 (43.3–100)

11.4 (0–56.6)

1.2 (0.2–3.9)

15

1.1 (1.03–1.1)

P2

210

86.4 (36.6–100)

14.3 (0–63.3)

1.2 (0.2–2.3)

15

1.1 (1.04–1.2)

P3

180

47.1 (3.3–86.6)

54.6 (13.3–96.6)

1.5 (0.4–2.5)

15

1.1 (1.01–1.2)

P4

180

61.7 (16.6–96.6)

38.1 (3.3–83.3)

1.6 (0.2–2.4)

15

1.1 (1.04–1.2)

C

630

82.0 (30–100)

22.5 (0–93.3)

1.6 (0.2–2.6)

45

1.1 (1.03–1.2)

I1, I2 impacted sites, P1, P2, P3, P4, partially protected sites, C closed site

of Porto de Galinhas. Conversely, a dense and diverse macrobenthic community, with abundant zoanthids and fragile vermetid tubes was found in the closed site of Tamandare´.

A similar pattern was found on reefs in New Caledonia, where the amount of bare rock tended to increase, with little live cover on some reef tops (Dumas et al. 2009). Trampling has been shown to decrease the area of live cover by up to

123

Environmental Management

Fig. 5 Area of bare rock (%) on reef tops under different trampling intensities in Porto de Galinhas and Tamandare´ (Pernambuco State, Brazil). A Area of bare rock by site. B Area of bare rock by trampling

intensities. I impacted. P partially protected. C closed sites. N number of analyzed photo quadrats. a–e Groups of samples defined by pairwise Mann–Whitney tests

Fig. 6 Percentage of cover of the most abundant organisms (Palisada perforata, Halimeda opuntia, Zoanthus sociatus, and Petaloconchus varians) on reef tops under different trampling intensities in Porto de

Galinhas and Tamandare´ (Pernambuco State, Brazil). I impacted, P partially protected, C closed sites. a–c Groups of samples defined by pairwise Mann–Whitney tests

123

Environmental Management Fig. 7 Multidimensional scaling (MDS) plot of the macrobenthos on reef tops under different trampling intensities in Porto de Galinhas and Tamandare´ (Pernambuco State, Brazil), based on the images taken in January 2012. I impacted, P partially protected, C closed sites. n = 219 samples (i.e., photo quadrats) and 14 taxa

Fig. 8 Diversity of sessile macrobenthos on reef tops under different conservation strategies in Porto de Galinhas and Tamandare´ (Pernambuco State, Brazil). A Diversity by site. B Diversity by trampling

intensities. I impacted, P partially protected, C closed sites. N number of analyzed photo quadrats. a–g Groups of samples defined by pairwise Mann–Whitney tests

20 % on Australian coral reefs (Kay and Liddle 1989). The present study corroborates the results of Kay and Liddle (1989), who also noted that trampling inhibits the growth and abundance of macroalgae on Australian reef tops. In addition to the well-known negative effect on healthy scleractinian coral cover, as shown in previous studies, e.g., for the Great Barrier Reef, Australia (Woodland and Hooper 1977), this study showed that intensive trampling can cause a decrease in cover of macroalgae, and especially zoanthids and mollusks. Trampling is known to cause severe breakage, especially in fragile branched corals (Kay and Liddle 1989). In Hawaiian reefs under severe

trampling impact, a similar effect has been demonstrated for coral populations (Rodgers and Cox 2003). A similar result was obtained in a recent study on the macroalgae that grow on the offshore coral reefs of Maracajau´, northeastern Brazil, that are impacted by activities such as scuba diving and snorkeling (Silva et al. 2012). A protected area and an area that was impacted by tourism were compared for subtidal reef strata (mean depth: 2.5 m) in these reefs. Macroalgal biomass and diversity was lower in the area impacted by tourism as compared to the protected area, similarly to the results of the present study.

123

Environmental Management

The two coral species Favia gravida and Siderastrea stellata showed a negligible cover at all sites in our study, such as in other reef tops in northeastern Brazil (Costa Jr. et al. 2001). They were found inside shallow tidal pools on the reef tops, indicating that they are not well adapted to total exposure, in contrast to numerous Indo-Pacific intertidal coral species (Richards et al. 2015). Corals are not abundant on intertidal reef tops in Brazil, for reasons that probably have nothing to do with trampling, such as the biogeographic isolation from coral diversity hotspots and the proximity to estuaries. This study confirms that even when there was no trampling for 12 years, cover of scleractinian corals remained negligible on northeast Brazilian reef tops. High cover of zoanthids may indicate human impacts on reef tops in the Indo-Pacific. Conversely, in northeastern Brazil, where zoanthids compete mostly with macroalgae and not with corals, high cover of zoanthids seems to be an indicator of low trampling impact. When comparing the usefulness of various community parameters as indicators of trampling, the area of bare rock distinguished the partially protected areas very well from the closed area. Bare rock and total live cover are correlated, but not collinear, since there are also other types of bare substrate on the reef tops, such as rubble and sand. Similarly, a recent study also showed that the area of bare rock could be used as a straightforward indicator of trampling in intertidal areas off Southern California (Huff 2011). Similarly to the present study, species richness of macroalgae off Southern California was also higher in the impacted areas (Huff 2011). One possible explanation is that trampling regularly opens up new areas for pioneer species. The qualitative assessment of macroalgae that were manually sampled may indicate some possible differences in species composition and species richness between sites. However, a quantitative comparison of total algal species richness between sites was not possible, since the photo transects do not permit an accurate identification of all algal species. Macroalgae play an important role in these environments, with respect to primary production and food supply for herbivores (Kaplan 1982; Greenway 1995; Levinton 1995; Sheppard et al. 2009), and serve as shelter for numerous invertebrate taxa. The decrease in abundance of these autotrophs has a direct effect on the abundance and diversity of species associated to these algae, both for meio- and macrofauna (Ansari et al. 1991; Brown and Taylor 1999; Sarmento et al. 2011; Sarmento and Santos 2012). Meiofauna are the main food item for many macroinvertebrates (Gerlach 1971; Uthicke 1999; Schwamborn and Criales 2000; Jaafar and Dexiang 2014) and fish (Gerlach 1971; John et al. 1989; Ferreira et al. 2004). Therefore, through the reduction of the cover of

123

macroalgae, trampling directly affects associated invertebrates and fish communities through the loss of their natural habitat or the habitat of their prey. When reef tops are trampled on, the complex food webs are directly affected. The macroalga P. perforata is characterized by its robustness to environments that are regularly exposed to desiccation, being omnipresent on reef top environments in northeastern Brazil (Silva and Fujii 2012). The higher cover of these macroalgae in the partially protected sites of Porto de Galinhas than in the closed site of Tamandare´, may be related to the occasional trampling that competitively favors the macroalgae over the more fragile zoanthids that are dominant in the closed area. The organism with the highest cover in the closed site was the zoanthid Z. sociatus, while in partially protected and impacted sites, the rhodophyte P. perforata was most abundant. This may also be due to the faster growth of P. perforata than Z. sociatus. The partially protected sites of Porto de Galinhas are still trampled on occasionally and had been established for only 3 years, while the closed site of Tamandare´ suffers zero trampling and has been totally shut off for more than 12 years (since April 1999), thus having a longer succession period for the establishment of a more mature, zoanthid-dominated community. The slowgrowing zoanthids seem to have developed strategies for a long-term dominance in relation to macroalgae and other fast-growing organisms in non-disturbed environments. Cnidarians, specially zoanthids, have been shown to be successful long-term competitors with macroalgae in several parts of the world (Chadwick and Morrow 2011). However, some species of zoanthids, such as those of the genus Palythoa and Zoanthus, have physical and chemical mechanisms that help them to out compete scleractinian corals, and thus can monopolize extensive areas. These strategies are influenced mainly by factors such as light, temperature, and concentrations of nutrients (Chadwick and Morrow 2011). The present study showed not only a reduction of total live cover but also the disappearance of some fragile mollusk species in impacted sites as compared to partially protected and closed sites, such as bivalves and vermetids. The bivalve Isognomon bicolor is a native Caribbean species and is considered an invasive species in Brazil (Martinez 2012). Competition of this bivalve with zoanthids such as Palythoa caribaeorum has been recorded in southeastern Brazil (Mendonc¸a-Neto and Gama 2009). Mollusks of the family Vermetidae play an important role in Brazilian reef environments as important contributors for the build-up of reef structures (Kempf and Laborel 1968; Maida and Ferreira 1997). In Brazil, the most common vermetid genera are Petaloconchus and Dentropoma (Kempf and Laborel 1968). These organisms may exhibit dense aggregations of tubular shells, keeping their fragile

Environmental Management

tubes like chimneys upright above the surface (Hadfield et al. 1972), which keeps them exposed and vulnerable to trampling, that causes permanent and often lethal breakage of the shells. Since the cover of vermetids was less than 1 % in impacted sites and after 3 years of isolation of the partially protected sites in Porto de Galinhas, the cover of these individuals increased to 2 %, probably after the isolation of impacted sites, the biomass of the population of vermetids seems to recover at a doubling time of approximately 3 years, which is in good agreement with the cover of these organisms in the closed site (13.5 % after 12 years). Only a complete and long-term closure of reef areas did restore a level of diversity that is characteristic for these ecosystems under pristine conditions, as observed in the closed area. The creation of marine protected areas that include the management of fisheries and tourism has produced relevant progress toward the protection of habitats and stocks and yielded important studies on the resilience of reef ecosystems (Westmacott et al. 2000; Edgar and Barrett 2012; Maypa et al. 2012; Colle´ter et al. 2014). Our results show that in just 3 years of exclusion from intense trampling, the partially protected sites of Porto de Galinhas saw a significant increase in live macrobenthic cover and the presence of sessile macrobenthic organisms that were not recorded in adjacent impacted sites. This corroborates similar results for the meiofauna in these two areas (Sarmento and Santos 2012). However, the comparison with the closed site shows that in spite of being an important step toward the conservation of fragile species and ecosystem structure, the partial isolation of reef tops does not produce ecosystems that are even remotely similar to pristine or enduringly closed no-go areas. Species composition in the closed site was strikingly different from sites with occasional trampling, being dominated by completely different organisms (zoanthids, not macroalgae) and with a much higher cover of fragile vermetids. The partially protected sites of Porto de Galinhas and the closed sites in Tamandare´ presented a high level of recovery of the benthic cover, considering that before its closure, the closed area of Tamandare´ had been suffering, for many decades, the same impact of intensive tourism and fishery as in Porto de Galinhas. Recovery of benthic live cover is a much faster process than the build-up of a highly diverse community, that is characterized by animals, such as zoanthids and vermetids, that are more slowgrowing and more fragile than the dominant macroalgae. In conclusion, the complete closure of selected reef areas is the most effective way to recover damaged reefs and to preserve their biodiversity. Clearly, there is an urgent need for a long-term conservation policy, and a widespread awareness of the drastic effects of trampling on reef ecosystems in Brazil and many other tropical regions.

Acknowledgments We would like to thank the Federal University of Pernambuco (UFPE), the CEPENE research centre in Tamandare´, and the office for the environment of the City of Ipojuca for logistical support. Many thanks to Mauro Maida for methodological and logistical support and helpful hints. Thanks to Patrı´cia F. R. S. M. Nogueira and Daniele L. B. Mallmann for participating in fieldwork and helpful suggestions. Thanks to the Coordenac¸a˜o de Aperfeic¸oamento de Pessoal de Nı´vel Superior (CAPES) for the M.Sc. scholarship granted to the first author. We are greatly indebted to the anonymous reviewers whose invaluable suggestions greatly improved this manuscript.

References Ansari ZA, Rivonker CU, Ramani P, Parulekar AH (1991) Seagrass habitat complexity and macroinvertebrate abundance in Lakshadweep coral reef lagoons, Arabian Sea. Coral Reefs 10:127–131. doi:10.1007/BF00572170 Bacalso RTM, Wolff M (2014) Trophic flow structure of the Danajon ecosystem (Central Philippines) and impacts of illegal and destructive fishing practices. J Mar Syst 139:103–118. doi:10. 1016/j.jmarsys.2014.05.014 Brown PJ, Taylor RB (1999) Effects of trampling by humans on animals inhabiting coralline algal turf in the rocky intertidal. J Exp Mar Biol Ecol 235:45–53. doi:10.1016/S0022-0981(98)00186-5 Chadwick N, Morrow K (2011) Competition among sessile organisms on coral reefs. In: Dubinsky Z, Stambler N (eds) Coral reefs: an ecosystem in transition. Springer, Netherlands, pp 347–371 Clarke KR, Gorley RN (2006) Primer v6: user manual/tutorial. Primer-e, Plymouth Clarke KR, Warwick RM (2001) Change in marine communities: an approach to statistical analysis and interpretation. Primer-e, Plymouth Colle´ter M, Gascuel D, Albouy C, Francour P, Morais LT, Valls A, Loc’h LF (2014) Fishing inside or outside? A case studies analysis of potential spillover effect from marine protected areas, using food web models. J Mar Syst 139:383–395. doi:10.1016/j. jmarsys.2014.07.023 Costa Jr OS, Attrill MJ, Pedrini AG, De-Paula JC (2001) Benthic macroalgal distribution in coastal and offshore reefs at Porto Seguro Bay, Brazilian Discovery Coast. In: Proceedings of the 9th International Coral Reef Symposium, Indonesia, International Society for Reef Studies 23–27 Dumas P, Bertaud A, Peignon C, Le´opold M, Pelletier D (2009) A ‘‘quick and clean’’ photographic method for the description of coral reef habitats. J Exp Mar Biol Ecol 368:161–168. doi:10. 1016/j.jembe.2008.10.002 Edgar GJ, Barrett NS (2012) An assessment of population responses of common inshore fishes and invertebrates following declaration of five Australian marine partially protected areas. Environ Conserv 39:271–281. doi:10.1017/S0376892912000185 Ferreira CEL, Floeter SR, Gasparini JL, Ferreira BP, Joyeux JC (2004) Trophic structure patterns of Brazilian reef fishes: a latitudinal comparison. J Biogeogr 31:1093–1106. doi:10.1111/j. 1365-2699.2004.01044.x Gerlach SA (1971) On the importance of Marine Meiofauna for Benthos communities. Oecologia 6:176–190. doi:10.1007/ BF00345719 Greenway M (1995) Trophic relationships of macrofauna within a Jamaican seagrass meadow and the role of the echinoid Lytechinus variegatus (Lamarck). Bull Mar Sci 56:719–736 Hadfield MG, Kay EA, Gillette MU, Lloyd MC (1972) The vermetidae (Mollusca: Gastropoda) of the Hawaiian Islands. Mar Biol 12:81–98. doi:10.1007/BF00347431

123

Environmental Management Hill J, Wilkinson C (2004) Methods for ecological monitoring of coral reefs: a resource for managers. AIMS, Townsville Page Jaafar Z, Dexiang C (2014) Goby and shrimp associations: more than meets the eye. Coral Reefs 33:863. doi:10.1007/s00338-0141143-4 John JST, Jones GP, Sale PF (1989) Distribution and abundance of soft-sediment meiofauna and a predatory goby in a coral reef lagoon. Coral Reefs 8:51–57. doi:10.1007/BF00301803 Kaplan EH (1982) Ecology of the coral reefs. Field guide to coral reef of the Caribbean and Florida. Houghton-Mifflin Company, Boston Kay A, Liddle M (1989) Impact of human trampling in different zones of a coral reef flat. Environ Manage 13:509–520. doi:10. 1007/BF01867685 Kempf M, Laborel J (1968) Formations de vermets et d’algues calcaires sur les cotes du Brasil. Recl Trav Stn Mar Endoume 43:9–23 Kohler KE, Gill SM (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. doi:10.1016/j.cageo.2005.11.009 Lea˜o ZMAN, Dominguez JML (2000) Tropical Coast of Brazil. Mar Pollut Bull 41:112–122. doi:10.1016/S0025-326X(00)00105-3 Levinton JS (1995) Marine biology: function, biodiversity, ecology. Oxford University Press, New York Maida M, Ferreira BP (1997) Coral reefs of Brazil: an overview. In: Proceedings of the 8th International Coral Reef Symposium, Panama, Smithsonian Tropical Research Unit 263–274 Martinez AS (2012) Spatial distribution of the invasive bivalve Isognomon bicolor on rocky shores of Arvoredo Island (Santa Catarina, Brazil). J Mar Biol Assoc Uk 92:495–503. doi:10. 1017/S0025315411000865 Maypa AP, White AT, Can˜ares E, Raffy M, Eisma-Osorio RL, Alin˜o P, Apistar D (2012) Marine partially protected area management effectiveness: progress and lessons in the Philippines. Coast Manage 40:510–524 Mendonc¸a-Neto JP, Gama BAP (2009) The native Palythoa caribaeorum overgrows on invasive species in the intertidal zone. Coral Reefs 28:497. doi:10.1007/s00338-008-0449-5 Oigman-Pszczol SS, Creed JC (2011) Can patterns in benthic communities be explained by an environmental pressure index? Mar Pollut Bull 62:2181–2189. doi:10.1016/j.marpolbul.2011. 07.001 Preskitt LB, Vroom PS, Smith CM (2004) A rapid ecological assessment (REA) quantitative survey method for benthic algae using photoquadrats with Scuba. Pac Sci 58:201–209. doi:10. 1353/psc.2004.0021 Richards ZT, Garcia RA, Wallace CC, Rosser NL, Muir PR (2015) A diverse assemblage of reef corals thriving in a dynamic intertidal

123

reef setting (Bonaparte Archipelago, Kimberley, Australia). PLoS One 10(2):e0117791. doi:10.1371/journal.pone.0117791 Rodgers KS, Cox EF (2003) The effects of trampling on Hawaiian corals along a gradient of human use. Biol Conserv 112:383–389. doi:10.1016/S0006-3207(02)00336-1 Rogers CS, Gilnack M, Fitz HC III (1983) Monitoring of coral reefs with linear transects: a study of storm damage. J Exp Mar Biol Ecol 66:285–300. doi:10.2983/035.029.0201 Rogers CS, Garrison G, Grober R, Hillis Z, Franke MA (1994) Coral reef monitoring manual for the Caribbean and Western Atlantic. Virgin Islands National Park, Saint John Sarmento VC, Santos PJP (2012) Trampling on coral reefs: tourism effects on harpacticoid copepods. Coral Reefs 31:135–146. doi:10.1007/s00338-011-0827-2 Sarmento VC, Barreto AFS, Santos PJP (2011) The response of meiofauna to human trampling on coral reefs. Sci Mar 75:559–570. doi:10.3989/scimar.2011.75n3559 Schwamborn R, Criales MM (2000) Feeding strategy and daily ration of juvenile pink shrimp Farfantepenaeus duorarum in a South Florida seagrass bed. Mar Biol 137:139–147. doi:10.1007/ s002270000317 Shannon CE (1948) A mathematical theory of communication. Bell Syst Tech J 27:379–423. doi:10.1002/j.1538-7305.1948.tb01338. x Sheppard CR, Davy SK, Pilling GM (2009) The biology of coral reefs. Oxford Scholarship Online, New York Silva IB, Fujii MT (2012) The Genus Herposiphonia (Ceramiales, Rhodophyta) in the coral reefs environmental protection area, Northeastern Brazil, with new records for Brazil and the Atlantic Ocean. Braz J Bot 35:107–118. doi:10.1590/S010084042012000100011 Silva IB, Fujii MT, Marinho-Soriano E (2012) Influence of tourist activity on the diversity of seaweed from reefs in Maracajau´, Atlantic Ocean, Northeast Brazil. Rev Bras Farmacogn 22:889–893 Uthicke S (1999) Sediment bioturbation and impact of feeding activity of Holothuria (Halodeima) atra and Stichopus chloronotus, two sediment feeding holothurians, at Lizard Island, great barrier reef. B. Mar Sci 64:129–141 Westmacott S, Teleki K, Wells S, West J (2000) Management of bleached and severely damaged coral reefs. World Conservation Union (IUCN), Gland, Switzerland and Cambridge, Oxford Woodland DJ, Hooper JNA (1977) The effect of human trampling on coral reefs. Biol Conserv 11:1–4. doi:10.1016/00063207(77)90020-9 Zar JH (1996) Biostatistical analysis. Prentice Hall, New Jersey