lable at ScienceDirect

Ocean & Coastal Management 84 (2013) 130e139

Contents lists avai

Ocean & Coastal Management

journal homepage: www.elsevier .com/locate/ocecoaman

Fish as indicators of diving and fishing pressure on high-latitude coralreefs

C. Floros*, M.H. Schleyer, J.Q. MaggsOceanographic Research Institute, P.O. Box 10712, Marine Parade, 4056 Durban, South Africa

a r t i c l e i n f o

Article history:Available online

* Corresponding author. Tel.: þ27 31 3288229.E-mail address: cfloros@ori.org.za (C. Floros).

0964-5691/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.ocecoaman.2013.08.005

a b s t r a c t

Despite the proclamation of South Africa’s coral reef marine protected areas (MPAs) more than 20 yearsago, the effects of human activities on the fish communities have not been investigated. This study used amulti-species Fish-index to compare ecological indicators such as biomass, abundance, trophic structureand reproductive potential between multiple-use and no-take sanctuary zones. Seven study reefs weresurveyed; six in South Africa and a non-MPA reef in southern Mozambique. Randomly stratified un-derwater visual censuses (UVC) using the point count technique were used to survey fish communities.Environmental variables and habitat characteristics were also recorded. Nonmetric multidimensionalscaling ordinations were similar for abundance and biomass trends and revealed a high degree of overlapbetween all zones, except for the no-take Sanctuaries. The latter formed discrete clusters and weresignificantly different (Analysis of Similarity) to the other zones. Total abundance and biomass werehighest in the Sanctuary zones and lowest in the Open zone. Differences in trophic composition betweenzones were largely due to predatory species. This was supported by similarity percentages analysis(SIMPER) which identified six discriminating species. Length-frequency analysis of these species revealedconsistent trends with higher numbers of large individuals in the Sanctuary zones and reduced numbersof small individuals in zones open to human activity. These results along with those of the GeneralisedLinear Models (GLM) demonstrate that human activities are affecting the southern African coral reef fishcommunities. Marginal differences between the multiple-use MPA zones on the South African reefs andthe non-MPA reef in southern Mozambique suggest that MPA management objectives require re-evaluation.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Coral reefs provide worldwide benefits and services to the valueof US$ 29.8 billion every year (Cesar et al., 2003). The biggest pro-portion is generated by tourism (US$ 9.6 billion) and fisheries (US$5.7 billion). Yet, these activities have increasingly caused degrada-tion on many reefs throughout the world (Halpern et al., 2008;Wilkinson, 2008). Various measures to manage and mitigate coralreef damage have been proposed and, among those, marine pro-tected areas (MPAs) have been proposed as an ideal managementsolution because they were perceived to simultaneously addressissues of overfishing, habitat degradation, and tourismdevelopment(Kelleher and Kenchington, 1992; IUCN, 2004).

MPAs are specifically intended to limit human activities indesignated locations (Sale et al., 2005; Mora et al., 2006) and the

All rights reserved.

degree to which human activities are limited determines the typeof MPA. In most instances, MPAs can be classified into two broadtypes; areas that are open to resource use and areas closed toresource use. No-take MPAs are areas closed to exploitation (heretermed sanctuaries). The second type of MPA allows harvesting ofresources, but under protective regulations that pertain to eachspecies being harvested. In addition, the types of fishing or har-vesting gear may be restricted. Such MPAs are multiple-use zonesand most often permit recreational activities such as SCUBA diving,snorkeling, whale watching and fishing.

Nevertheless, MPAs are not ‘cure alls’ (Alder, 1996; Agardy et al.,2003;Mora et al., 2006) andmany face difficulty in implementationand enforcement due to poor governance, and lack of managementguidance and evaluation (White et al., 2006). Among the majorchallenges restricting effective MPA management is a lack of sci-entific information about the status and nature of activities oper-ating therein (Kelleher et al., 1995; Pomeroy et al., 2005;Wells et al.,2007). Assessment of MPA effectiveness is a matter of great urgencyand importance given the multitude of stressors threatening the

C. Floros et al. / Ocean & Coastal Management 84 (2013) 130e139 131

future of coral reefs (Bellwood et al., 2004; Knowlton and Jackson,2008). Furthermore, many MPAs are promulgated as multiple-usezones and complex decision-making processes are often requiredto balance conservation and socio-economic objectives. Thus, it isimperative to assess the impacts of human activities on the bio-logical components of coral reefs and whether these impacts areconsistent with the management objectives of the MPA.

Obtaining such information requires regular monitoring andassessment of ecological integrity within MPA boundaries.Ecological integrity refers to system wholeness, including thepresence of appropriate species, populations, and communities andthe occurrence of ecological processes at appropriate rates andscales (Karr, 1981; Angermeier and Karr, 1994). Evaluating changesin ecological integrity is often a major obstacle facing MPA man-agers due to a scarcity of technical skills, experience and funding(Pomeroy et al., 2005). Indicator-based monitoring programmesmay provide an effective solution to these challenges because

Fig. 1. Studied reefs on the north east coast of South Africa and southern Mozambique. Sitethree multiple-use zones (Two-mile, Seven-mile and Nine-mile reef) and a non-MPA zone

ecological indicators may represent inexpensive means of gath-ering scientific data that does not require skilled personnel. For thisreason, they have become important tools in coral reefs research,reporting and management (see Jameson and Kelty, 2004 forreview).

The challenge is determining which biological criteria areeffective measures of ecological integrity but are also simpleenough to monitor (Dale and Beyeler, 2001). Fish are conspicuousbiological components of coral reefs that have good indicator po-tential owing to their importance as a valuable protein source(Pauly et al., 2002; Bell et al., 2009) and functional ecological roleson reefs (Bellwood et al., 2004). Accordingly, this study selected amulti-species index (Fish-index) to assess the effects of humanactivities on high-latitude coral reefs in South Africa. The aim wasto use comparisons of ecological parameters such as biomass,abundance, trophic structure and reproductive potential as mea-sures of ecological integrity. Changes observed in functional

s included three Sanctuary zones (Leadsman Shoal, Red Sands Reef, and Rabbit Rock),(Shallow Malongane).

Table 1Zonation of the reefs based on SCUBA diving and fishing intensities. SCUBA diving statistics are averages for the period 2007e2008 (Pieters, 2009).

Zone Activity Reef name

SCUBA diving intensity Fishing intensity

Open Low (w4500 dives/year) High (unregulated angling and spearfishing) Shallow MalonganeDiving-Fishing Low (w2100 dives/year) Restricted (angling and spearfishing; gamefish onlya). Seven-mile Reef, Nine-mile ReefHigh-Diving High (w54,000 dives/year) Nil Two-mile ReefSanctuary Nil Nil Leadsman Shoal, Red Sands, Rabbit Rock

a Pelagic bony fish of the families Scombridae, Carangidae, Pomatomidae, Coryphaenidae, Rachycentridae, Xiphiidae, Ostiophoridae and Sphyraenidae, the species Aprionvirescens, as well as pelagic cartilaginous fish of the families Carcharinidae, Isuridae, Sphyrnidae, Alopiidae and Odontaspididae (Marine Living Resources Act, Section 3.1 (G)Regulation R1429).

C. Floros et al. / Ocean & Coastal Management 84 (2013) 130e139132

processes such as growth, reproduction and trophic functioningmay provide MPA managers with insight into current conditionsand aid in predicting future trends.

South Africa’s coral reefs are located within the boundaries oftwo longstanding, contiguous MPAs. Two types of conservationstrategies are recognized in the MPAs: no-take sanctuary zones andmultiple-use zones. Recreational fishing and SCUBA diving are themost common activities in the MPAs (Schleyer, 2000). Fishing isknown to have a direct effect on fish communities via the harvest oftarget and bait species, and the removal of functional groups, andhas the potential to cause significant changes in the structure ofreef fish communities (Cooke and Cowx, 2004; Dulvy et al., 2004).SCUBA diving, has been shown to affect coral communities(Tratalos and Austin, 2001; Zakai and Chadwick-Furman, 2002;Barker and Roberts, 2004; Hawkins et al., 2005), however, thepaucity of empirical studies investigating the effects of high divingintensity on fish communities in the literature represents a‘knowledge-gap’. It is thus both timely and pertinent that the ef-fects of human activities on the South African coral reef fish as-semblages are assessed for efficient MPA management.

2. Materials and methods

2.1. Study area

The South African coral reefs are located along the north-eastcoast of South Africa. Six study reefs were selected to representdifferent types of MPA zones (Fig. 1). The study area was extendedto include non-MPA reefs in southernMozambique for comparativepurposes. However, only one reef, located at Ponta Malongane,could be included as a study site due to political and logisticalconstraints. The southern Mozambican and South African coralreefs are the southernmost reefs in theWestern Indian Ocean (Rieglet al., 1995). Due to their close proximity they share numeroussimilarities. They are exposed to similar oceanographic conditions,which are dominated by the southwards flowing Agulhas Current(Lutjeharms, 2006). They are confined to the narrow continentalshelf (Ramsay, 1996) and can be classified as patch reefs (Ramsayand Mason, 1990). They are atypical coral reefs in that the coralcommunities are non-accretive and form a veneer on late-Pleistocene beach rock (Ramsay, 1996). Their benthic commu-nities are similar with corals contributing 50e70% to the total livingcover (Jordan and Samways, 2001; Pereira, 2003; Celliers andSchleyer, 2008). The fish assemblages in both regions are verysimilar, with approximately 300 reef-associated species; 80% ofwhich are Indo-Pacific (Pereira, 2003; Celliers and Schleyer, 2008;Floros et al., 2012).

2.2. Reef protection status and human activity

The South African coral reefs are situated within the Maputa-land MPA and St Lucia MPA, proclaimed in 1986 and 1979,

respectively. No changes have beenmade to theMPA zonation sincetheir proclamation. For this study, the reefs were zoned accordingto the intensity and type of human activity (Table 1). These wereSanctuary, High-Diving, Diving-Fishing and Open zones. Sanctuaryzones theoretically prohibited all forms of human activity. TheHigh-Diving zone permitted SCUBA diving only. The Diving-Fishingzone permitted SCUBA diving and restricted fishing (gamefishonly). Lastly, the Open zone had no active law enforcement at thetime of data collection which meant that activities on the reef wereunregulated. No data for fishing intensity were available for any ofthe zones.

2.3. Fish-index selection

Species were included as indicators based on literature reviewsand field observations. The criteria included vulnerability toexploitation by fishing or hobbyists, trophic status and ease ofidentification. The Fish-index species are listed in Appendix C alongwith the justification for their selection.

2.4. Surveys of reef fish communities

Surveys of fish communities were conducted from August2007 to February 2009 to include four summer and four wintersampling periods. The point count underwater visual censustechnique, adapted from Samoilys and Carlos (2000), was usedto estimate the abundance and biomass of the Fish-index spe-cies. A total of 60 point counts were conducted per reef. Eachcensus (point count) consisted of a 5 min count within a circle10 m in diameter. All Fish-index species observed on the reefand within the water column during each point count wererecorded and their sizes were estimated to the nearest 10 cm.Estimates of fish length were used to generate biomass usingknown length-weight regression coefficients from Fishbase(Froese and Pauly, 2011). Only two divers entered the water atany one time, a surveyor and a buddy diver. The same diverconducted all point counts to minimize variation and errorincurred by diver bias. Point counts within each reef wereseparated by at least 50 m. The time of each point count wasconducted between 0800 and 1400. All counts were conductedat a depth range of 12e15 m.

2.5. Trophic levels

Species were allocated to one of eight trophic guilds; top-levelpredators, medium-level predators, planktivores, omnivores, her-bivores, benthivores, corallivores and invertivores (Appendix C).Trophic allocationwas based on diet information retrieved from theliterature and supplemented by field observations (see Floros et al.,2012 for details).

2D Stress: 0.26 2D Stress: 0.25

Human activityHigh-DivingDiving-FishingSanctuaryOpen

Human activityHigh-DivingDiving-FishingSanctuaryOpen

a b

Fig. 2. Non-metric multidimensional scaling ordination of samples based on a) fourth-root transformed abundance data and b) log(xþ1) transformed biomass data.

Table 2Results of ANOSIM run on fourth-root transformed species abundance data fordifferences between zones. Global R ¼ 0.423. Significance of Global R < 0.001. Sig-nificant differences are in bold.

Pairwise tests R statistic Significance level %

High-Diving, Fishing-Diving 0.302 0.006High-Diving, Sanctuary 0.429 0.005High-Diving, Open 0.087 9.2Diving-Fishing, Sanctuary 0.461 <0.0001Fishing-Diving, Open 0.235 0.6Sanctuary, Open 0.6 0.0005

C. Floros et al. / Ocean & Coastal Management 84 (2013) 130e139 133

2.6. Habitat characteristics and environmental variables

On completion of each point count, aspects of the reef habitatwere described within the point count area; topography and coralcover. Coral cover was estimated using a rapid visual techniqueadapted from English et al. (1994), where estimates where dividedinto three categories (low, medium and high) according to thepercentage of coral (hard and soft). The topographic complexity ofthe substratum was visually estimated using a six point scale sys-tem adapted from Polunin and Roberts (1993). The adaptationswere specific to southern African coral reefs and included thefollowing categories: 1 e flat reef; 2 e low undulating spur andgroove; 3emedium slopes or ridges with no grooves or gullies; 4e

medium spur and groove or pinnacles; 5 e high slopes or ridges;and 6 e high spur and groove, overhangs or pinnacles. Depth wasmeasured using a dive computer and sea temperatures were ob-tained from an underwater temperature probe stationed at 18 m.

2.7. Statistical analysis

Univariate analyses were carried out on abundance and biomassdata using One-Way analysis of variance (ANOVA). Data were notnormally distributed and thus non-parametric KruskaleWallasOne-Way ANOVA on ranks was used. Dunn’s pairwise multiplecomparisons procedurewas used to detect differences between fishabundance and biomass parameters. Univariate analyses wereconducted using the statistical package Sigma Plot 11.0. The crite-rion for significance for all tests was p � 0.05.

Multivariate analyses were undertaken using PRIMER v.6 tofurther explore difference between zones. Abundance data werefourth-root transformed, while the log(xþ1) transformation wasused for biomass data due to large differences in the values for top-level predators. Non-metric multidimensional scaling (MDS) wasused to examine differences in spatial distribution of Fish-indexassemblages across the different zones. Analysis of Similarity(ANOSIM) was also used to confirm or refute trends observed in theMDS ordinations. R-statistics >0.45 were considered to signifysignificant differences with a limited degree of overlap betweenfish community structure and values <0.45 indicated large degreesof overlap between community structures (Clarke and Gorley,2006). SIMPER analysis was used to identify those Fish-index spe-cies responsible for the BrayeCurtis dissimilarity between MPAzones. Length-frequency graphs were generated for the sixdiscriminating species identified by SIMPER. Size at sexual maturityfor each of the six species was determined from the literature and

used as an approximation for reproductive potential. GeneralisedLinear Models (GLM) using R (R Development Core Team, 2011)were used to examine the influence of environmental variables(temperature, depth), habitat characteristics (coral cover andtopography) and activity on fish community parameters (abun-dance and biomass). Activity was classified according to the in-tensity of diving and fishing in the different MPA zones (Table 1).The GLM for both abundance and biomass data followed the form:

ðcountsÞ ¼ b0 þ b1ðactivityÞ þ b2ðtemperatureÞþ b3ðcoral coverÞ þ b4ðtopographyÞ þ b5ðdepthÞþ ε

Overdispersion in the abundance data invalidated the use of thePoisson distribution and the negative binomial distribution withlog-link functionwas used (Maunder and Punt, 2004). Biomass datawere continuous and were thus modelled using gamma distribu-tion with inverse-link function.

3. Results

3.1. Spatial distribution of species

The spatial orientation of the abundance and biomass datasuggested a high degree of overlap between the High-Diving, Div-ing-Fishing and Open zones (Fig. 2). In contrast, the Sanctuaryzones formed distinct clusters in both data sets. ANOSIM testsconfirmed the trends observed in both MDS plots (Tables 2 and 3).In terms of abundance, the Fish-index community differed signifi-cantly between Sanctuary zones and the Open zone, and betweenSanctuary and Diving-Fishing zones (Global R ¼ 0.423; p < 0.001).

Table 3Results of ANOSIM run on log(xþ1) transformed species biomass data for differencesbetween zones. Global R ¼ 0.423. Significance of Global R < 0.001. Significant dif-ferences are in bold.

Pairwise tests R Statistic Significance level %

High-Diving, Diving-Fishing 0.202 0.2Sanctuary, High-Diving 0.516 0.0005High-Diving, Open 0.162 1.7Sanctuary, Diving-Fishing 0.468 <0.0001Diving-Fishing, Open 0.236 0.4Sanctuary, Open 0.713 0.0005

Fig. 3. Total mean abundance (a), biomass (b) and trophic structure of the Fish-indexspecies in the four zones on the southern African study reefs.

C. Floros et al. / Ocean & Coastal Management 84 (2013) 130e139134

The biomass data revealed that the Fish-index structure in Sanc-tuary zones was significantly different to the other three zones(Global R ¼ 0.423; p < 0.0001).

3.2. Abundance, biomass and trophic guild composition

The total mean abundance and total mean biomass manifestedsimilar trends across zones with the highest values recorded inSanctuary zones and the lowest in the Open zone (Fig. 3). Differencesin mean abundances were significant between Sanctuary(18.01 fish 78 m�2) and High-Diving (13.86 fish 78 m�2), andSanctuary and Open zones (9.81 fish 78 m�2) (p < 0.001). In addi-tion, mean abundances between Diving-Fishing (15.93 kg 78 m�2)and Open reefs were also significantly different. Total biomass onSanctuary reefs (23.71 kg 78 m�2) was significantly different and atleast three times greater than in all other zones (p < 0.001).

Invertivores were the most abundant trophic guild in all MPAzones and did not differ significantly between zones. Planktivoresand herbivores were the next most abundant guilds. Abundancesfor herbivores were significantly different between High-Divingand the Open zone (P < 0.01). Top- and medium-level predatorswere significantly more abundant in Sanctuary zones (P < 0.001).Differences in biomass between MPA zones were most significantfor predator guilds. In the Sanctuary zones, the total biomass of top-level predators (8.32 kg 78 m�2) was seven times great than thatrecorded in the High-Diving zones and four times greater than inthe Diving-Fishing zone. No top-level predators were recorded inthe Open zone. Medium-level predators in the Sanctuary zones(10.6 kg 78 m�2) were at least six times greater than in the otherthree zones.

3.3. Discriminating species

Comparisons between Sanctuary zones and the other zonesrevealed that six species were the top contributors to the dissimi-larity in each pairwise comparison (Table 4). These species wereEpinephelus tukula, Lutjanus bohar, Aprion virescens, Caranx mel-ampygus, Variola louti and Oplegnathus robinsoni and their cumu-lative contribution to the dissimilarity between each comparativegroup was almost 50%. Excluding O. robinsoni, all the above speciesare important reef predators. All these species are potentially tar-geted in the Open zone; however, only C. melampygus andA. virescens may be targeted on the South African coral reefs.

3.4. Length-frequency analysis

Length-frequency graphs were generated for the six discrimi-nating species identified by SIMPER analysis. All six species wererecorded to have greater abundance and mean body size in theSanctuary zones (Fig. 4; Appendix A & B). There was a clear trendfor reduced abundances and mean size in zones open to fishing(Fishing-Diving and Open) such that only low numbers of sexually

mature individuals were recorded in these two zones. The Openzone had the lowest number and smallest mean body size of the sixspecies, and two species, A. virescens (Fig. 4a) and E. tukula (Fig. 4e),were absent from this zone. In the High-Diving zone, no sexuallymature individuals of A. virescens, Oplegnathus robinsoni andE. tukula were recorded, while sexually mature individuals of thethree other species were present in low numbers.

3.5. Environmental variables, habitat characteristics and humanactivity

The recorded environmental variables and habitat characteris-tics appeared to explain little of the deviance observed in the fishabundance and biomass between MPA zones. All variables were

Table 4Results of SIMPER analysis. Only species providing the highest percent contributiontowards the average dissimilarity in abundance and biomass data between zoneshave been included. Species highlighted in bold are those contributing >40% to theoverall dissimilarity.

High-Divingvs sanctuary

Diving-Fishingvs sanctuary

Openvs sanctuary

Epinephelus tukula 10.32 9.18 9.85Lutjanus bohar 9.84 8.65 9.58Aprion virescens 8.37 9.42 9.51Caranx melampygus 6.97 6.9 7.73Variola louti 7.19 6.96 7.18Oplegnathus robinsoni 6.31 6.63 6.68Odonus niger 5.88 7.1 6.55Pomacanthus imperator 5.44 5.45 4.94Acanthurus leucosternon 4.34 4.31 4.06Balistoides conspicillum 3.67 4.71 3.86Chaetodon meyeri 3.65 4.61 3.77Scarus rubroviolaceus 3.26 3.12 3.63Bodianus diana 3.74 3.99 3.35Siganus sutor 2.57 2.77 3.34Forcipiger flavissimus 3.44 3.25 3.25Chaetodon trifascialis 2.12 0 2.47Amphiprion allardi 0 3.28 2.36Average dissimilarity 45.01 44.11 52.49

C. Floros et al. / Ocean & Coastal Management 84 (2013) 130e139 135

initially modelled in the GLMs; however, only activity, topographyand coral cover were found to be significant andwere thus includedin the final models (Tables 5 and 6). Activity was most accountablefor the explained deviance in both the abundance (8.7%) andbiomass (18.8%) between zones. Topography explained 0.9% of thedeviance in abundance, while coral cover explained 2% of thedeviance in biomass between zones.

4. Discussion

This study used a pre-determined list of species to investigatedifferences in fish community parameters between zones of vary-ing protection levels. Most of the significant differences in abun-dance, biomass, trophic levels and reproductive potential wererecorded for predatory and target reef fish species. For all param-eters, values were highest in the Sanctuary zones and lowest in theOpen zone. These results are consistent with other studiescomparing reef fish communities across gradients of protection inthe Western Indian Ocean (McClanahan & Graham et al., 2005;McClanahan et al., 2007; Floros et al., 2012) and Indo-Pacific(Graham et al., 2003; Westera et al., 2003; Evans and Russ, 2004;Lester and Halpern. 2008) and strongly suggests that the differ-ences are a result of human activities. These findings wereconfirmed by the GLMs which demonstrated that the deviancebetweenMPA zones was explainedmore by human activity than byhabitat characteristics or environmental variables.

4.1. Fishing

Efforts have been made to regulate fishing on the South Africancoral reefs by restricting the extraction of species to gamefish. Inaddition to the species restrictions, there are daily fishing quotas foreach target species (C. melampygus ¼ 5 day�1, A. virescens¼ 10 day�1), but no minimum size limits. Despite the daily quotas,the results suggest that fishing is having an effect on populations ofthe aforementioned target fish species. Low densities and smallmean body sizes of target species are detectable effects of over-fishing on coral reefs (Russ and Alcala, 1989; Westera et al.,2003). In the Diving-Fishing zones, the reduced mean size oftarget species corresponded closely to theoretical predictions (see

Dulvy et al., 2004; Graham et al., 2005), with all target speciesshowing significantly smaller mean sizes. A similar, but more se-vere trend was evident in the Open zone where target species suchas L. bohar and C. melampygus were represented by very lowabundances of small individuals. In addition A. virescens andE. tukula were not recorded. At the time of data collection, therewere no fishing restrictions for C. melampygus, A. virescens, andL. bohar, while E. tukula was fully protected from recreationalfishing only (Decree 51/99 of 31 August). The absence of the latterspecies from the study records may due to the inclusion of only onereef in southern Mozambique, due to intensive fishing pressure(recreational, subsistence, or semi-commercial) or a combination ofboth factors. However, similar sampling effort on other reefsrecorded high abundances of E. tukula and the results are mostlikely indicative of high levels of exploitation.

A reduction in themean size of targeted species in fished areas istermed size selective fishing or ‘age truncation’ and is due to fisherstargeting larger individuals (Berkeley et al., 2004). Age truncationcan have an important effect on fish assemblage structure andfunction, potentially affecting the productivity and resilience of fishpopulations (Baskett et al., 2005). The low abundance and absenceof sexually mature individuals of A. virescens, C. melampygus,O. robinsoni, E. tukula and L. Bohar recorded in the non-Sanctuaryzones suggests reduced reproductive potential and could havesignificant ramifications for future generations. These populationsmay be reliant on juveniles from surrounding no-take zones toreplenish stocks. Sanctuary or no-take zones have been advocatedas areas of high reproductive output because there are greaterdensities of larger, sexually mature fish present. It has been furthersuggested that the increased reproductive output, whether in theform of eggs, larvae or juvenile fish, may repopulate areas open tofishing (Berkeley et al., 2004; Francini-Filho and Moura, 2008).Whether this occurs, depends on the location of Sanctuary zones,the oceanographic conditions in the region, larval dispersal andlarval life history characteristics (Watson et al., 2009).

The South African Sanctuary zones are situated north and southof the multiple-use zones. Given the greater biomass and abun-dance of fish within the Sanctuary zones, the potential for con-nectivity through adult spillover or larval dispersal between thezones is thus high (e.g. Russ, 2002; Tupper, 2007). It is most likelythat dispersal or movement would take place along the north-south gradient due to the predominantly south-flowing Agulhascurrent, making the reefs in the Northern sanctuary zones the mostimportant reefs due to their strategic position. On a largergeographic scale, it is also likely that some degree of connectivityexists between coral reef fish communities along the east Africancoast and the fish communities on the southern African coral reefs.The prevailing southward flowing currents in the MozambiqueChannel (Lutjeharms, 2006) have been suggested as a mechanismlinking populations of certain coral species (Ridgway et al., 2008;Macdonald et al., 2011). If the northern reefs are acting as sourcesof propagules for fish communities in the south, any threats ordisturbances to these reefs will have significant consequences forthe southern African coral reef communities.

4.2. Scuba diving

Due to increased mortality of target species associated withfishing, it was anticipated that the densities and biomass of thesespecies would be lowest in the Diving-Fishing zones and similarin the Sanctuary and High-Diving zones. Differences betweenSanctuary and Diving-Fishing zones were significant; however, sotoo were the differences between Sanctuary and High-Divingzones. Lack of compliance by the fishers in the High-Divingzone may provide an explanation for the low abundances of the

Fig. 4. Length-frequency distribution plots of discriminating species in the four different zones. a) Aprion virescens, b) Caranx melampygus, c) Oplegnathus robinstoni, d) Variola louti,e) Epinephelus tukula and f) Lutjanus bohar. Circles indicate size at sexual maturity.

C. Floros et al. / Ocean & Coastal Management 84 (2013) 130e139136

six discriminating species. However, the high volume of divers onthis reef would make it likely that suspicious fishing activitywould be reported to the authorities. While poaching cannot becompletely excluded, the high diving intensity in this zone may

Table 5Results of abundance GLMwith negative binomial distribution and log-link functionwhere d.f. ¼ degrees of freedom, AIC ¼ Akaike’s information criterion,P ¼ significance level.

Abundance model structure d.f. AIC Residualdeviance

Explaineddeviance %

p

Null 2951.6 473.32þActivity 3 2948.2 432.05 8.7 <0.001þTopog 1 2918.3 428 0.9 <0.05Full model 4 2916.3 428 9.6 <0.001

be having a negative effect on the abundance of these largespecies.

The low densities of the top-level predator, E. tukula on theHigh-Diving reef, were of particular concern as this species is oneof the most significant large-bodied predators on South African

Table 6Results of biomass GLM with gamma distribution and inverse function whered.f¼ degrees of freedom, AIC¼ Akaike’s information criterion, P¼ significance level.

Biomass modelstructure

d.f. AIC Residualdeviance.

Explained deviance % p

Null 9048.2 768.52þCoral cover 2 9041.6 753.53 2.0 <0.01þActivity 3 8930.8 608.01 18.9 <0.001Full model 5 8921.8 592.98 20.9 <0.001

C. Floros et al. / Ocean & Coastal Management 84 (2013) 130e139 137

coral reefs. Historic records and anecdotal accounts of higherabundances of E. tukula on the High-Diving reef during the 1980sto early 1990s (Koornhof, 1991; Chater et al., 1995) indicate that areduction in population numbers has occurred during the lastthree last decades. The results and observations of this studysuggest that this decline may be linked to high diving intensity.E. tukula is a large resident predator known to show aggressiveterritorial behaviour towards divers on reefs isolated from humanactivities (Debelius, 2001; Peschak, 2010). All E. tukula encoun-tered on Sanctuary reefs displayed aggressive or curious behav-iour towards divers which included open mouth displays,bumping of divers, biting of the buoy-reel and stalking of diversthroughout the dive. In contrast, E. tukula behaviour on theDiving-Fishing reef was cautious and the divers were seldomapproached. Furthermore, E. tukula were most commonlyobserved at the edge of diver visibility where they moved fromone overhang to another. Thus the persistent presence of SCUBAdivers on the High-diving reef may be causing a competitivedisturbance to E. tukula.

4.3. Implications for management

No-take MPAs carry a high socio-economic cost because of lostfishing grounds (Jones, 2008), but show great benefits and yieldsignificantly great biomass and densities (Gell and Roberts, 2003;Lester and Halpern, 2008). Multiple-use MPAs are typically seen asmore feasible and are often implemented as a compromisebecause they confer some benefits over open access areas (Lesterand Halpern, 2008). This study also provided evidence that no-take MPA have significantly greater benefits to coral reef fishcommunities in South Africa; however, the differences betweenthe multiple-use and open zones were considerably smaller thananticipated. The implications for this are significant in terms of theobjectives and goals of the different MPA zones. The open zone insouthern Mozambique had little legislation and no enforcement

Appendix A

Mean abundance (fish 78 m�2) of the Fish-index species and difference

Sanctuary High-Divi

Acanthurus leucosternon 0.904 1.200Amphiprion allardi 0.220 0.077Aprion virescens 0.617 0.062Balistoides conspicillum 0.072 0.046Bodianus diana 0.502 0.554Caranx melampygus 0.919 0.615Chaetodon madagaskariensis 0.804 0.754Chaetodon meyeri 0.885 0.400Chaetodon trifascialis 0.038 0.169Chaetodon trifasciatus 0.024 0Diplodus cervinus 0.019 0.385Epinephelus tukula 0.292 0.077Forcipiger flavissimus 0.512 0.292Labroides dimidiatus 1.507 2.338Lutjanus bohar 1.512 0.138Odonus niger 2.129 1.754Oplegnathus robinsoni 0.502 0.108Plectroglyphidodon johnstonianus 0.464 0.369Pomacanthus imperator 0.187 0.246Pygoplites diacanthus 0.005 0.015Scarus rubroviolceus 1.292 1.569Siganus sutor 0.072 0.215Thalassoma hebraicum 3.024 2.462Variola louti 0.531 0.185Zebrasoma desjardini 0.014 0.092

a Indicates species that were inconclusive in the ANOVA or Multiple comparisons pro

protecting the reefs at the time of data collection. In contrast, theSouth African MPAs have clear objectives outlined in the MPAmanagement plan (Ezemvelo KwaZulu-Natal Management Plan2007). However, the objectives are identical for all MPA zones,despite differences in the type and intensity of human activity ineach zone. Thirteen objectives are relevant to the coral reefs andfall within three functional categories; protection, fisheries man-agement and utilisation (see Floros, 2010 for details). Whenassessed using the indicator parameters (biomass, densities, tro-phic composition and reproductive potential), eight objectiveswere questionable with regard to achieving their managementgoals. This indicates that the multiple-use zones are not providingprotection to the reef fish communities under the current levels ofhuman activity. It also suggests that these zones require a morecomplex management approach which includes a reassessment oftheir objectives to balance the trade-off between sustainableresource use and conservation. In contrast, the no-take Sanctuaryzones appear to have achieved all management objectives. Theprohibition of all human activity in these areas has promotedgreater biomass, densities and reproductive potential of the reeffish communities. Considering their long history of closure tohuman activities, the Sanctuary zones potentially represent un-disturbed ecosystems on which future management plans shouldbe based.

Acknowledgements

We are grateful for the financial support provided by the Na-tional Research Foundation and the South African Association forMarine Biological Research. We also thank the staff of the Ocean-ographic Research Institute for their technical support in the field.The 4x4 vehicle used in this project was sponsored by the MazdaWildlife Fund. Ezemvelo KwaZulu-Natal Wildlife and the iSi-mangaliso Wetland Authority are acknowledged for their logisticalsupport.

s between zones shown by one-way ANOVA.

ng Diving-Fishing Open p

1.296 0.353 NS0.576 0.333 <0.0010.024 0 <0.0010.464 0.059 0.007a

0.464 0.333 NS0.288 0.098 0.0030.984 0.706 NS0.144 0.549 <0.0010 0.137 0.007a

0 0.078 0.007a

0 0 NS0.080 0 <0.0010.352 0.294 NS2.704 2.235 0.0050.184 0.118 <0.0014.024 1.137 NS0.104 0.039 0.0050.032 0.235 <0.0010.144 0.039 NS0.016 0 NSa

1.200 0.725 NS0.080 0.137 NSa

2.576 2.294 NS0.184 0.098 <0.0010.024 0 NSa

cedure due to their uncommon status.

Appendix C

Trophic guild allocation of the Fish-index species and the ratio-nale for their inclusion in the study. B ¼ benthivore, C ¼ corallivore,H ¼ herbivore, I ¼ invertivore, M-L P ¼ medium-level predator,O ¼ omnivore, P ¼ planktivore, T-L P ¼ top-level predator.

Trophic guild Rationale

Acanthurus leucosternon H Common herbivoreAmphiprion allardi O Iconic speciesAprion virescens M-L P Target species and important

predatorBalistoides conspicillum I Sensitive to diver presenceBodianus diana I Common prey speciesCaranx melampygus T-L P Target species and important

predatorChaetodon madagaskariensis O Generalist feederChaetodon meyeri C Specialist feederChaetodon trifascialis C Specialist feederChaetodon trifasciatus C Specialist feederDiplodus cervinus I Regional endemicEpinephelus tukula T-L P Largest reef predatorForcipiger flavissimus I Prey speciesLabroides dimidiatus I Specialist feederLutjanus bohar M-L P Important predatorOdonus niger P Trophic representativeOplegnathus robinsoni B Regional endemicPlectroglyphidodon johnstonianus C Coral specific speciesPomacanthus imperator B Specialist feederPygoplites diacanthus B Specialist feeder and

uncommon statusScarus rubroviolceus H Largest reef herbivoreSiganus sutor H Important regional herbivoreThalassoma hebraicum I Common prey speciesVariola louti M-L P Important reef predatorZebrasoma desjardinii H Conspicuous grazer

Appendix B

Mean biomass (kg 78 m�2) of the Fish-index species and differences between zones shown by one-way ANOVA.

Sanctuary High-Diving Diving-Fishing Open p

Acanthurus leucosternon 0.125 0.145 0.193 0.036 0.047Amphiprion allardi 0.004 0.001 0.016 0.012 <0.001Aprion virescens 1.914 0.062 0.017 0 <0.001Balistoides conspicillum 0.102 0.029 0.069 0.119 NSBodianus diana 0.045 0.056 0.038 0.025 NSCaranx melampygus 2.393 0.828 0.402 0.092 <0.001Chaetodon madagaskariensis 0.020 0.046 0.031 0.031 NSChaetodon meyeri 0.058 0.031 0.017 0.046 <0.001Chaetodon trifascialis 0.001 0.019 0 0.019 0.011a

Chaetodon trifasciatus 0.001 0 0 0.010 NSa

Diplodus cervinus 0.005 0.029 0 0 NSa

Epinephelus tukula 8.126 1.075 2.077 0 <0.001Forcipiger flavissimus 0.025 0.015 0.018 0.018 NSLabroides dimidiatus 0.010 0.016 0.025 0.015 0.004Lutjanus bohar 4.361 0.407 0.215 0.048 <0.001Odonus niger 1.079 0.731 1.244 1.000 NSOplegnathus robinsoni 1.197 0.134 0.190 0.114 0.04Plectroglyphidodon johnstonianus 0.002 0.002 0.001 0.003 <0.001Pomacanthus imperator 0.194 0.329 0.209 0.058 NSPygoplites diacanthus 0.000 0.001 0.001 0 NSa

Scarus rubroviolaceus 1.340 2.571 2.508 1.243 <0.001Siganus sutor 0.050 0.133 0.066 0.125 NSa

Thalassoma hebraicum 0.242 0.258 0.314 0.239 NSVariola louti 1.170 0.490 0.151 0.215 <0.001Zebrasoma desjardinii 0.005 0.021 0.009 0 NSa

a Indicates species that were inconclusive in the ANOVA or Post Hoc comparison due to their uncommon status.

C. Floros et al. / Ocean & Coastal Management 84 (2013) 130e139138

References

Agardy, T., Bridgewater, P., Crosby, M.P., Day, J., Dayton, P.K., Kenchington, R.,Laffoley, D., McConney, P., Murray, P.A., Parks, J.E., Peau, L., 2003. Dangeroustargets? Unresolved issues and ideological clashes around marine protectedareas. Aquat. Conserv. 13, 353e367.

Alder, J., 1996. Have tropical marine protected areas worked? An initial analysis oftheir success. Coast Manage. 24, 97e114.

Angermeier, P., Karr, J., 1994. Biological integrity versus biological diversity as policydirectives. BioScience 44, 690e697.

Barker, N.H.L., Roberts, C.M., 2004. Scuba diver behaviour and the management ofdiving impacts on coral reefs. Biol. Conserv. 120, 481e489.

Baskett, M.L., Levin, S.A., Gaines, S.D., Dushoff, J., 2005. Marine reserve design andthe evolution of size at maturation in harvested fish. Ecol. Appl. 15, 882e907.

Bell, J.D., Kronen, M., Vunisea, A., Nash, W., Keeble, G., Demmkea, A., Pontifex, S.,Andréfouët, S., 2009. Planning the use of fish for food security in the Pacific.Mar. Policy 33, 64e76.

Bellwood, D.R., Hughes, T.P., Folke, C., Nyström, M., 2004. Confronting the coral reefcrisis. Nature 429, 827e833.

Berkeley, S.A., Hixon, M.A., Larson, R.J., Love, M.S., 2004. Fisheries sustainability viaprotection of age structure and spatial distribution of fish populations. Fisheries29, 23e31.

Celliers, L., Schleyer, M.H., 2008. Coral community structure and risk assessment ofhigh-latitude reefs at Sodwana Bay, South Africa. Biodivers. Conserv. 17, 3097e3117.

Cesar, H.S.J., Burke, L., et al., 2003. The Economics of Worldwide Coral ReefDegradation. Cesar Environmental Economics Consulting (CEEC), 6828GHArnhem, The Netherlands.

Chater, S.A., Beckley, L.E., vander Elst, R.P., Garrat, P.A.,1995.Underwater visual censusof fishes in the St Lucia Marine Reserve, South Africa. Lammergeyer 43, 15e23.

Clarke, K.R., Gorley, R.N., 2006. PRIMER v6: User Manual/Tutorial. PRIMER-E,Plymouth.

Cooke, S.J., Cowx, I.G., 2004. The role of recreational fishing in the global fish crisis.BioScience 54, 857e859.

Dale, V.H., Beyeler, S.C., 2001. Challenges in the development and use of ecologicalindicators. Ecol. Indic. 1, 3e10.

Debelius, H., 2001. Indian Ocean Reef Guide. IKAN-Unterwasserarchiv, Frankfurt.Dulvy, N.K., Polunin, N.V.C., Mill, A.C., Graham, N.A.J., 2004. Size structural changes

in lightly exploited coral reef fish communities: evidence for weak indirecteffects. Can J. Fish Aquat. Sci. 61, 466e475.

English, S., Wilkinson, C., Baker, V. (Eds.), 1994. Survey Manual for Tropical MarineResources. Australian Institute of Marine Science, Townsville.

Evans, R.D., Russ, G.R., 2004. Larger biomass of targeted reef fish in no-take marinereserves on the Great Barrier Reef, Australia. Aquat. Conserv. 14, 505e519.

Floros, C., 2010. An Evaluation of Coral Reef Fish Communities in South AfricanMarine Protected Areas (PhD dissertation). Oceanographic Research Institute,University of KwaZulu-Natal.

C. Floros et al. / Ocean & Coastal Management 84 (2013) 130e139 139

Floros, C., Schleyer, M.H., Maggs, J.Q., Celliers, L., 2012. Baseline assessment of high-latitude coral reef fish communities in southern Africa. Afri J. Mar Sci. 34, 55e69.

Froese, R., Pauly, D., 2011. Fishbase. www.fishbase.org (accessed July 2011).Francini-Filho, R.B., Moura, R.L., 2008. Evidence for spillover of reef fishes from a no-

take marine reserve: an evaluation using the before-after control-impact (BACI)approach. Fish Res. 93, 346e356.

Gell, F.R., Roberts, C.M., 2003. Benefits beyond boundaries: the fishery effects ofmarine reserves. Trends Ecol. Evol. 18, 448e455.

Graham, N.A.J., Evans, R.D., Russ, G.R., 2003. The effects of marine reserve protectionon the trophic relationships of reef fishes on the Great Barrier Reef. Environ.Conserv. 30, 200e208.

Graham,N.A.J., Dulvy, N.K., Jennings, S., Polunin, N.V.C., 2005. Size-spectra as indicatorsof the effects of fishing on coral reef fish assemblages. Coral Reefs 24, 118e124.

Halpern, B.S., Walbridge, S., Selkoe, K.A., Kappe, C.V., Micheli, F., D’Agrosa, C.,Bruno, J.F., Casey, K.S., Ebert, C., Fox, H.E., Fujita, R., Heinemann, D., Lenihan, H.S.,Madin, E.M.P., Perry, M.T., Selig, E.R., Spalding, M., Steneck, R.S., Watson, R., 2008.A global map of human impact on marine ecosystems. Science 319, 948e952.

Hawkins, J.P., Roberts, C.M., Kooistra, D., Buchan, K., White, S., 2005. Sustainability ofSCUBA diving tourism on coral reefs of Saba. Coastal Manage. 33, 373e387.

IUCN, 2004. Managing Marine Protected Areas: a Toolkit for the Western IndianOcean. In: IUCN Eastern African Regional Programme. Nairobi, Kenya.

Jameson, S.C., Kelty, R.A., 2004. A review of indicators of land-based pollution stresson coral reefs. In: Joint EPA/NOAA/USGA/DOI Workshop: Assessing PollutionStress on Coral Reefs Aug 31eSept 2 Honolulu, Hawaii.

Jones, P.J.S., 2008. Fishing industry and related perspectives on the issues raised byno-take marine protected area proposals. Mar. Policy 32, 749e758.

Jordan, I.E., Samways, M., 2001. Recent changes in the coral composition of a SouthAfrican coral reef, with recommendations for long-term monitoring. Biodivers.Conserv. 10, 1027e1037.

Karr, J.R., 1981. Assessment of biotic integrity using fish communities. Fisheries 6,21e27.

Kelleher, G., Kenchington, R., 1992. Guidelines for Marine Protected Areas. IUCN,Gland, Switzerland.

Kelleher, G., Bleakley, C., Wells, S., 1995. A Global Representative System of MarineProtected Areas. The World Bank, Washington.

Knowlton, N., Jackson, J.B.C., 2008. Shifting baselines, local impacts, and globalchange on coral reefs. PLoS ONE 6, 215e220.

Koornhof, A., 1991. The Dive Sites of South Africa. Comprehensive Coverage ofDiving and Snorkelling. New Holland, Cape Town.

Lester, S.E., Halpern, B.S., 2008. Biological responses in marine no-take reservesversus partially protected areas. Mar. Ecol. Prog. Ser. 367, 49e56.

Lutjeharms, J.R.E., 2006. The Agulhas Current. Springer, Berlin.Macdonald, A.H.H., Lamb, J., Schleyer, M.H., 2011. Acropora austera connectivity in

the south western Indian Ocean assessed using nuclear intron sequence data.Mar Biol. 158, 613e621.

Maunder, M.N., Punt, A.E., 2004. Standardizing catch and effort data: a review ofrecent approaches. Fish Res. 70, 141e159.

McClanahan, T.R., Graham, N.A.J., Calnan, J.M., MacNeil, M.A., 2007. Toward pristinebiomass: reef fish recovery in coral reef marine protected areas in Kenya. Ecol.Appl. 17, 1055e1067.

Mora, C., Andrefouet, S., Costello, M.J., Kranenburg, C., Rollo, A., Vernon, J.,Gaston, K.J., Muyers, R., 2006. Coral reefs and the global network of marineprotected areas. Science 312, 1750.

Pauly, D., Christensen, V., Guénette, S., et al., 2002. Towards sustainability in worldfisheries. Nature 418, 689e695.

Pereira, H.M., 2003. Recreational Scuba Diving and Reef Conservation in SouthernMozambique (MSc thesis). Oceanographic Research Institute, University ofKwaZulu-Natal.

Peschak, T.P., 2010. Lost World: the Marine Realm of Aldabra and the Seychelles. JBPublishing.

Pieters, M., 2009. Dive Statistics for Sodwana Bay. Ezemvelo KZN WildlifeAuthorities.

Polunin, N.V.C., Roberts, C.M., 1993. Great biomass and value of target coral-reef fishes in two small Caribbean marine reserves. Coral Reefs 100, 167e176.

Pomeroy, R.S., Watson, L.M., Parks, J.E., Cid, G.A., 2005. How is your MPA doing? Amethodology for evaluating the management effectiveness of marine protectedareas. Ocean. Coast. Manage. 48, 485e502.

R Development Core Team, 2011. R: a Language and Environment for StatisticalComputing. R Foundation for Statistical Computing, Vienna, Austria, ISBN 3-900051-07-0. http://www.R-project.org/.

Ramsay, P.J., 1996. Quaternary marine geology of the Sodwana Bay continental shelf,Northern KwaZulu-Natal. Bull. Geo. Surv. S. Afr. 117, 1e85.

Ridgway, T., Riginos, C., Davis, J., Hoegh-Guldberg, O., 2008. Genetic connectivitypatterns of Pocillopora verrucosa in southern African marine protected areas.Mar. Ecol. Prog. Ser. 354, 161e168.

Riegl, B., Schleyer, M.H., Cook, P.J., Branch, G.M., 1995. Structure of Africa’s southern-most coral communities. Bull. Mar. Sci. 5, 676e691.

Russ, G.R., Alcala, A.C., 1989. Effects of intense fishing pressure on an assemblage ofcoral reef fishes. Mar. Ecol. Prog. Ser. 56, 13e27.

Russ, G.R., 2002. Yet another review of marine reserves as reef fishery managementtools. In: Sale, P. (Ed.), Coral Reef Fishes. Academic Press, San Diego, CA,pp. 421e443.

Sale, P.F., Cowen, R.K., Danilowicz, 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. Criticalscience gaps impede use of no-take fishery reserves. Trends Ecol. Evol. 20,74e80.

Samoilys, M.A., Carlos, G., 2000. Determining methods of underwater visualcensus for estimating the abundance of coral reef fishes. Environ. Biol. Fish 57,289e304.

Schleyer, M.H., 2000. South African coral communities. In: McClanahan, T.R.,Sheppard, C., Obura, D. (Eds.), Coral Reefs of the Indian Ocean. Oxford UniversityPress, New York, pp. 83e105.

Tupper, M.H., 2007. Spillover of commercially valuable reef fishes from marineprotected areas in Guam, Micronesia. Fish. Bull. 105, 527e537.

Tratalos, J.A., Austin, T.J., 2001. Impacts of recreational SCUBA diving on coralcommunities of the Caribbean Island of Grand Cayman. Biol. Conserv. 102,67e75.

Watson, D., Anderson, M.J., Kendrick, G.A., Nardi, K., Harvey, E.S., 2009. Effects ofprotection from fishing on the lengths of targeted and non-targeted fish speciesat the Houtman Abrolhos Islands, Western Australia. Mar. Ecol. Prog. Ser. 384,241e249.

Wells, A., Burgess, N., Ngusaru, A., 2007. Towards the 2012 marine protected areatarget in Eastern Africa. Ocean. Coast. Manage 50, 67e83.

Westera, M., Lavery, P., Hyndes, G., 2003. Differences in recreationally targetedfishes between protected and fished areas of a coral reef marine park. J. Exp.Mar. Biol. Ecol. 294, 145e168.

White, A.T., Meneses, A.B.T., Ovenden, M., Tesch, S., 2006. Sustaining marine pro-tected areas through continued monitoring and evaluation: the MPA reportguide and management rating system. In: Proc 10th Int Coral Reef Symp, Oki-nawa, Japan, pp. 1466e1470.

Wilkinson, C.R., 2008. Status of Coral Reefs of the World: 2008. Global Coral ReefMonitoring Network and Reef and Rainforest Research Centre, Townville,Australia.

Zakai, D., Chadwick-Furman, N.E., 2002. Impacts of intensive recreational diving onreef corals at Eilat, northern Red Sea. Biol. Conserv. 105, 179e187.