Contents lists available at ScienceDirect

Fisheries Research

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

Is catch-and-release shore angling compatible with the conservation goals ofmarine protected areas? A case study from the iSimangaliso Wetland Park inSouth Africa

Bruce Q. Manna,⁎, Paul D. Cowleyb, Stuart W. Dunlopa, Warren M. Pottsc

aOceanographic Research Institute, PO Box 10712, Marine Parade, Durban, 4056, South Africab South African Institute for Aquatic Biodiversity, PO Box 1015, Grahamstown, 6140, South Africac Department of Ichthyology and Fisheries Science, Rhodes University, PO Box 94, Grahamstown, 6140, South Africa

A R T I C L E I N F O

Handled by Chennai Guest Editor

Keywords:Recreational fisheries managementNo-take protected areasSurf-zone fishes

A B S T R A C T

Area management in the form of no-take marine protected areas (MPAs) has been criticised as a fisheriesmanagement tool because of its limited capacity to provide short-term benefits to local fisheries. This study useddata from a long-term fish monitoring and tagging project to assess whether catch-and-release (C&R) shoreangling could be compatible with the management objectives of a large, multiple-use, zoned MPA in SouthAfrica. Tag-recapture rates, trends in relative abundance and mean size of target species, sub-lethal effects andother potential environmental impacts suggested that C&R research angling, using best practise fish handlingtechniques, did not have an overall negative impact on protected fish populations. While positive from a sci-entific monitoring perspective, more sensitive species did show evidence of increased post-release mortality thatwould be exacerbated by higher intensity C&R angling conducted by members of the angling public. It is thusconcluded that C&R shore angling by members of the angling public is not compatible with MPAs zoned for no-take. However, because C&R angling does have substantially lower negative impacts compared to recreationalharvest fisheries, areas zoned for C&R offer good potential as buffer areas adjacent to no-take areas or as stand-alone areas where fish conservation can be improved. This concept is proposed for the improved conservation ofsurf-zone angling fish species within the iSimangaliso Wetland Park and further afield.

1. Introduction

The consequences of over a century of recreational shore anglingalong the South African coastline have been: 1) changes in catchcomposition with notable reductions in the abundance of species thatattain a large size, are slow growing, exhibit residency or have vul-nerable life history stages (e.g. estuarine dependency); 2) a reduction incatch rates or catch per unit effort (CPUE) of target species; and 3) areduction in mean size of target species (Bennett et al., 1994; Brouweret al., 1997; Dunlop and Mann, 2012). Management interventions suchas daily bag and minimum size limit restrictions have seldom abatedthese trends (Attwood and Bennett, 1995a; Maggs et al., 2012, 2016),but area closures in the form of well enforced no-take marine protectedareas (MPAs) have resulted in significant localised improvements in theabove mentioned stock status indicators (Bennett and Attwood, 1991;Attwood and Bennett, 1995b; Cowley et al., 2002; Mann et al., 2016a).

Despite these observed improvements, the use of no-take areas as a

management tool has sometimes been criticised as they provide limitedshort-term benefits to local fisheries (Hilborn et al., 2004; Alós andArlinghaus, 2013). Since no harvest is associated with recreational catch-and-release (C&R) angling, it has been suggested that it could be com-patible with management of no-take MPAs (Cooke et al., 2006). This,however, assumes that the effects of C&R angling do not cause declines infish populations or have other detrimental effects (e.g. litter, environ-mental damage, etc.) that are contrary to the goals of no-take MPAs.Given the uncertainty around this debate, and the fact that C&R anglinghas been proposed as a method to create revenue in some South AfricanMPAs (Attwood et al., 1997; Harris et al., 2011), we investigated theimpact of C&R angling on a surf-zone fish population in the St LuciaMarine Reserve, a zoned MPA within the iSimangaliso Wetland Park, aWorld Heritage Site in northern KwaZulu-Natal, South Africa.

In this paper, we used data from a long-term fish monitoring andtagging project (initiated in November 2001) aimed at comparing surf-zone fish catches within the exploited area of the St Lucia Marine

https://doi.org/10.1016/j.fishres.2018.07.021Received 22 August 2017; Received in revised form 3 July 2018; Accepted 29 July 2018

⁎ Corresponding author.E-mail address: bruce@ori.org.za (B.Q. Mann).

Fisheries Research 208 (2018) 179–188

0165-7836/ © 2018 Elsevier B.V. All rights reserved.

T

Reserve with those from an adjacent no-take sanctuary area and themovement patterns of dominant species (Mann et al., 2015, 2016a,2016b, 2016c). However, with the promulgation of a beach vehicle banin January 2002, which effectively reduced angling effort by virtue ofthe distance to the nearest access point (walking distance of ∼5 km),the primary objective of the project was changed to monitor the po-tential recovery of a previously exploited area using the no-take sanc-tuary as a benchmark (Mann and Tyldesley, 2013; Mann et al., 2016a).Although the primary aim of the project was to monitor the dynamics ofsurf-zone fish populations, it provided an opportunity to evaluate thelong-term impact of C&R angling and address the question: “Can C&Rangling be integrated into MPA management as an eco-tourism activity,without negatively impacting local fish stocks?”

To address this question, recapture rates of tagged fishes togetherwith trends in CPUE of commonly caught species were used to providean indication of post-release mortality in the iSimangaliso Wetland Parklong-term research monitoring project, which served as a proxy for anexperimental C&R fishery within a MPA. Similarly, changes in meanfish size (which should decline if there was substantial C&R mortality)were calculated over the 12-year study period. In addition, sub-lethaleffects of C&R angling such as deep-hooking, multiple captures andother environmental impacts such as tackle loss often associated withrecreational C&R shore angling (Lewin et al., 2006; Radomski et al.,2006; Brownscombe et al., 2017) were investigated. The findings of thisstudy are discussed in terms of opportunities for improved conservationand management of the iSimangaliso Wetland Park and other MPAs inSouth Africa and further afield. Some general considerations regardingthe future of C&R angling in South Africa are also discussed.

2. Methods

2.1. Study area

The St Lucia Marine Reserve forms part of the iSimangaliso WetlandPark and stretches from 1 km south of Cape Vidal (28° 08′S; 32° 33′E) toWhite Sands (27° 26′S; 32° 42′E) 11 km north of Sodwana Bay, a dis-tance of approximately 80 km, and extends three nautical miles(5.6 km) out to sea (Mann et al., 1998). A no-take sanctuary area issituated centrally between Leven Point (27° 55′S; 32° 35′E) and RedCliffs (27° 43′S; 32° 37′E), a distance of ∼25 km (Fig. 1). The shorelineis predominantly sandy and is backed by high vegetated dunes, but thesurf-zone itself consists of both sand and scattered low relief reefscomprised of sedimentary beach rock (Ramsay, 1996). No shore anglingby members of the public is permitted in the centrally situated no-takesanctuary area, but is allowed south of Leven Point and north of RedCliffs. However, a ban on the use of off-road vehicles in the coastal zonein January 2002 (Government Gazette No. 22960 promulgated in termsof the South African National Environmental Management Act No. 107of 1998) effectively precluded shore anglers from fishing more thanreasonable walking distance (∼5 km) north of Cape Vidal and south ofSodwana Bay. This ultimately created a buffer zone on either side of thesanctuary that was subjected to greatly reduced fishing pressure sub-sequent to the implementation of the beach vehicle ban.

2.2. Data collection

Standardised C&R angling by volunteer anglers using a stratifiedsampling approach was used in this study. Further details are providedby Mann et al. (2015, 2016a, 2016b, 2016c). A total of 32 anglers wereselected and trained on best C&R fish handling and tagging practices,based on the recommendations by Bartholomew and Bohnsack (2005),Cooke and Suski (2005), Cooke et al. (2006), Arlinghaus et al. (2010)and Brownscombe et al. (2017) and how to record the relevant data.The 32 anglers were used on a rotational basis and eight anglers par-ticipated on each field trip. In addition to the above pool of anglers, oneor two guest anglers were invited on each field trip.

Four quarterly field trips were conducted per year. During each fieldtrip, research C&R angling was conducted by eight anglers in four se-lected two kilometre areas. Two of the 2 km areas (SA and SB) wereinside the no-take sanctuary area (between Leven Point and Red Cliffs)and two were in the previously exploited area (EA and EB) south ofLeven Point (Fig. 1). On each day one team of four anglers fished in oneof the previously exploited areas (EA or EB) and the other team of fouranglers fished in one of the sanctuary areas (SA or SB). This was rotatedeach day so that at the end of four days each area had been fished byeach of the two teams, providing a statistically robust sampling protocolin order to compare catches between the sanctuary area and the ad-jacent previously exploited area (Mann and Tyldesley, 2013; Mannet al., 2016a).

While the sampling strategy was kept as consistent as possible, animportant change was made after the first 10 years of the project. InNovember 2011, each of the four 2 km sampling areas was enlarged tocover approximately 10 km of coastline. This change was im-plemented to reduce an inherent sampling bias whereby anglers ex-erted more effort at certain favoured localities within the 2 km sam-pling areas and to increase the scope of recapturing tagged fish over alarger area i.e. between the initial 2 km sample areas (Mann et al.,2015).

Standardised shore angling gear (i.e. 3–4.5m graphite surf rods,multiplier or fixed spool reels with braided or monofilament line ran-ging between 9–23 kg breaking strain) were used. A maximum of twobarbless single hooks per trace was allowed and hook sizes rangedbetween 1/0 and 10/0. No attempt was made to remove the hook if afish was “gut-hooked” (hook being lodged in the oesophagus) and theline was simply cut close to the eye of the hook (Schaeffer and Hoffman,2002; Butcher et al., 2010). Use of circle hooks was encouraged (Cookeand Suski, 2004) but not enforced due to the higher price of these hooksand the preferences of some anglers. In order to make the researchfishing comparable to that used by the general angling public and totarget as wide a range of species as possible, commercially availablebaits were used (Mann and Tyldesley, 2013; Mann et al., 2016a). Whileuse of certain artificial lures (i.e. surface plugs and spoons) was allowedto target various carangid species, soft plastic baits and the use of fly-fishing gear was not permitted.

All fish caught were landed using a plastic stretcher with asheathed stainless steel ruler down the centre and immediately cov-ered with a wet cloth, quickly measured, carried back to the waterand released. A bucket of fresh seawater was kept at the measuringstation and emphasis was placed on keeping the fish out of water foras short a time as possible and all surfaces were kept moist to reduceinjury and stress (Casselman, 2005; Brownscombe et al., 2017). Se-lected species greater than 300 mm fork length (FL), or total length(TL) in the case of serranids and sciaenids, were tagged using plasticdart tags. Two different types of tags were used namely PDAT tags(114 mm×1.6 mm) for fish over 600 mm and PDL tags (85 mm×1.6 mm) for fish between 300–600 mm. All tags were supplied byHallprint©, Australia and tags were obtained through the Oceano-graphic Research Institute’s Cooperative Fish Tagging Project(Dunlop et al., 2013).

All catch, effort and tagging data were recorded on a slate by eachangler. Effort data recorded included date, angler name, sampling area(i.e. EA, EB, SA or SB), time (hours) fished at each locality, targetingand tackle lost (i.e. number of hooks, swivels and sinkers lost whilefishing). Catch data recorded included tag number (if the fish wastagged or recaptured), species, length (mm fork or total length), date,time and locality. A note was made if the hook had been cut off and leftin the fish or if a tagging scar was present on the fish, indicating that ithad been previously tagged but the tag had been shed. Each fish’scondition was also assessed at the time of release (1 = swam awaystrongly; 2 = weak and struggling to regain equilibrium i.e. the abilityto maintain an upright orientation; 3 = bleeding as a result of hooking;4 = died as a result of capture).

B.Q. Mann et al. Fisheries Research 208 (2018) 179–188

180

2.3. Data analysis

All data were captured onto a MS Access database and analysis wasdone using MS Access queries and MS Excel spreadsheets. The numberof tag recaptures and tag scars were quantified for each species (in-cluding single and multiple recaptures). The CPUE (fish/angler/hour)was calculated for the top six most abundant species following themethods described by Mann et al. (2016a). Similarly, annual trends inmean length of the top six species were calculated for the duration ofthe study period. The overall objective of this study was to look at theimpact of C&R angling on the fish populations within the area as awhole. While differences in the demography and relative abundance of

the primary species between the previously exploited areas (EA & EB)and the no-take sanctuary areas (SA & SB) should not be ignored, thesewere described in detail by Mann et al. (2016a). For the purpose of thismanuscript, we first compared the relative abundance and mean sizebetween the previously exploited and sanctuary areas using a two-sample t-test assuming unequal variances. However, because all areassampled were effectively functioning as no-take areas (since im-plementation of the beach vehicle ban in 2002), per species CPUE andlength data from all sample areas were pooled to provide an indicationof overall trends in fish abundance and population size structure.Trends in CPUE and mean size were tested using linear regressions andone-way ANOVA. The growth rate of Lutjanus rivulatus determined from

Fig. 1. Map of the St Lucia Marine Reserve and Sanctuary within the iSimangaliso Wetland Park showing the sampling areas in this study (EA and EB are in thepreviously exploited area, while SA and SB are within the no-take sanctuary area). Note that the extended sampling areas (as of November 2011) simply consisted ofnorthward and southward extensions (shown by horizontal lines) of the original 2 km sampling areas.

B.Q. Mann et al. Fisheries Research 208 (2018) 179–188

181

tag-recapture data and implications of deep-hooking and multiple re-captures on growth, as described by Mann et al. (2016c), are presentedand discussed in context of C&R angling. In addition to the above, theoverall impact of C&R angling on the condition of fish at release and thebroader environmental impact of tackle loss were assessed.

3. Results

Between November 2001 and November 2013 a total of 59 fieldtrips were conducted with a total effort of 16,596 angling hours, ofwhich 8,212 h (49.5%) took place in the previously exploited area and8,385 h (50.5%) in the sanctuary area. During this period a total of15,324 fishes from 91 species and 38 families were caught and released.A total of 6,613 fish were tagged of which 701 individuals (10.6%)belonging to 16 species were recaptured at least once (Table 1).

3.1. Tag-recapture study

Of the dominant species with 100 or more fish tagged, five residentreef-associated species including Lutjanus rivulatus, Epinephelus margin-atus, Dinoperca petersi, Epinephelus andersoni and Pomadasys furcatus hadrecapture rates exceeding 10% (Fig. 2). In contrast, the recapture rate ofspecies associated with sandy substrata such as Trachinotus botla andRhabdosargus sarba, or more mobile species such as Caranx heberi, wassubstantially lower (Fig. 2).

3.2. Trends in abundance

CPUE of P. furcatus, L. rivulatus and R. sarba was significantly higherin the no-take sanctuary areas compared to the adjacent previouslyexploited areas (Table 2). Using pooled data for all areas combined,there was a significant (P < 0.05) increase in the CPUE of four (P.furcatus, L. rivulatus, Diplodus capensis and D. petersi) of the six mostabundant species. In contrast, there was a slight but insignificant(P > 0.5) decline in the CPUE of T. botla and a significant (P < 0.05)decline in the CPUE of R. sarba over the 12-year study period (Fig. 3).

3.3. Trends in mean size

The mean size of T. botla and D. capensis was significantly larger inthe sanctuary areas, while the mean size of D. petersi was significantlysmaller (Table 3). There was no significant difference in mean size ofthe other three species caught in the two zones (Table 3). Using pooleddata for all areas combined, trends in mean annual length of the sixmost abundant species caught all showed a slight increase in overallmean length during the study period, but none of these increases weresignificant (Fig. 4).

Table 1Tag and recapture data for 16 fish species tagged and recaptured in the St Lucia Marine Reserve between November 2001 and November 2013.

Family Species No. of fish taggedand recaptured

No. of single andmultiple recaptures

No. oftag scars

Total %Recaptureda

Rhynchobatidae Rhynchobatus djiddensis 55–5 4–1 1 9.09Serranidae Epinephelus andersoni 325–57 49–8 1 15.38Serranidae Epinephelus marginatus 295–73 49–24 2 17.29Serranidae Epinephelus tukula 236–11 12–0 1 5.51Haemulidae Plectorhinchus flavomaculatus 158–10 10–0 1 6.96Haemulidae Plectorhinchus playfairi 46–1 1–0 0 2.17Haemulidae Pomadasys furcatus 817–57 51–6 50 12.36Dinopercidae Dinoperca petersi 479–96 71–25 6 16.08Lutjanidae Lutjanus argentimaculatus 20–2 2–0 0 10.00Lutjanidae Lutjanus rivulatus 1308–652 412–240 33 34.02Sparidae Diplodus hottentotus 16–1 1–0 0 6.25Sparidae Rhabdosargus sarba 529–16 16–0 10 4.91Sciaenidae Argyrosomus japonicus 29–2 2–0 0 6.89Carangidae Caranx melampygus 30–3 3–0 0 10.00Carangidae Caranx heberi 141–5 5–0 0 3.55Carangidae Trachinotus botla 1328–13 13–0 2 1.13

a Including tag scars but excluding multiple recaptures.

Fig. 2. Recapture rates of 10 fish species with over 100 individuals tagged inthe St Lucia Marine Reserve between November 2001 and November 2013(dashed line shows the 10% recapture rate).

Table 2Statistical comparison of the catch per unit effort (fish/angler/hour)±standard deviation between the no-take sanctuary areas (SA & SB) and thepreviously exploited areas (EA & EB) for the six most abundant species caughtin the St Lucia Marine Reserve between November 2001 and November 2013.

Species Mean CPUESanctuaryareas

Mean CPUEPreviouslyexploited areas

t df p

Pomadasysfurcatus

0.220±0.101

0.163± 0.086 3.327 113 <0.001a

Lutjanus rivulatus 0.260±0.143

0.109± 0.084 6.984 94 <0.005a

Trachinotus botla 0.143±0.105

0.173± 0.109 −1.521 116 0.131

Diplodus capensis 0.053±0.042

0.049± 0.033 0.425 109 0.671

Dinoperca petersi 0.054±0.042

0.036± 0.032 2.714 108 0.008

Rhabdosargussarba

0.058±0.063

0.029± 0.033 3.043 88 0.003a

a denotes a significant difference.

B.Q. Mann et al. Fisheries Research 208 (2018) 179–188

182

3.4. Effects of deep-hooking and multiple captures on survival and growth

Of the 15,324 fish caught during this study (considering all species)only 1008 (6.6%) had ingested the hook. Of these fish 629 were taggedand released and 93 (14.8%) were recaptured. This recapture rate wasnot significantly different (P > 0.05) from the total of 5,984 fishtagged that had not ingested the hook, of which 911 (15.2%) wererecaptured (including multiple recaptures).

The growth rate of L. rivulatus that had ingested a hook when theywere tagged was slower but not significantly different to those that hadbeen hooked in the mouth and had the hook removed after capture(Table 4). Similarly, the growth rate of L. rivulatus that were only re-captured once was not significantly different to those that were re-captured on two or more occasions (i.e. multiple recaptures) (Table 4).

3.5. Impacts of C&R angling on fish condition at release

Of the 15,324 fish caught and released during this study, 31 (0.2%)were recorded as being “weak” on release (i.e. they had difficulty re-gaining equilibrium), 44 (0.29%) were recorded to be “bleeding”

Fig. 3. Trends in CPUE for the six most abundant species caught in the St Lucia Marine Reserve between November 2001 and November 2013.

Table 3Statistical comparison of the mean length (mm fork length)± standard devia-tion between the no-take sanctuary areas (SA & SB) and the previouslyexploited areas (EA & EB) for the six most abundant species caught in the StLucia Marine Reserve between November 2001 and November 2013.

Species Mean lengthSanctuaryareas

Mean lengthPreviouslyexploited areas

t df p

Pomadasysfurcatus

268±38(n= 1815)

270±40(n= 1373)

1.919 3236 0.055

Lutjanusrivulatus

322±83(n= 2195)

317±87(n= 923)

−1.744 1873 0.081

Trachinotusbotla

314±63(n= 1160)

298±59(n= 1395)

−7.069 2577 <0.005a

Diploduscapensis

224±31(n= 442)

215±33(n= 419)

−4.066 882 <0.005a

Dinoperca petersi 343±71(n= 467)

370±78(n= 295)

5.142 711 <0.005a

Rhabdosargussarba

387±87(439)

400±99(n= 225)

1.757 426 0.079

a denotes a significant difference.

B.Q. Mann et al. Fisheries Research 208 (2018) 179–188

183

(normally indicating that the fish had been deeply hooked or hooked inthe gills) and 58 (0.38%) died immediately as a result of capture. Noneof these fish were recaptured and based on the assumption that all ofthe above mentioned fish succumbed to the injuries sustained duringcapture, the total “immediate” mortality (133 fish) was extremely low(0.87%).

3.6. Associated environmental impacts of C&R angling

Strict discipline on littering was maintained throughout the projectand all litter was removed from the beach. However, a reality of shoreangling in rocky surf-zone areas is that some tackle will be lost. Overthe full 12-year duration of this project a total of 6,350 hooks, 5,840sinkers, 4,716 swivels and 10 lures were lost. This equates to approxi-mately 800 kg of lead sinkers assuming an average weight of 142 g (5ounces) per sinker.

Fig. 4. Annual trends in mean length (± standard deviation) of the six most abundant species caught in the St Lucia Marine Reserve between January 2002 andNovember 2013 (data from November 2001 was excluded).

Table 4Comparison between growth parameters estimated for single and multiple re-captures and for deep-hooked and unhooked Lutjanus rivulatus using a like-lihood ratio test. The terms g300 and g600 refer to fish of reference lengths300mm and 600mm FL, respectively (adapted from Mann et al., 2016c).

Treatment Sample size (n) Mean growth rate(mm y–1)

χ2 df p

Single recapture 291 g300=35.08g600=17.58

6.27 3 0.099

Multiple recapture 162 g300=35.42g600=16.51

Hook ingested 75 g300=34.58g600=17.32

2.34 3 0.504

Hook removed 357 g300=35.16g600=18.88

B.Q. Mann et al. Fisheries Research 208 (2018) 179–188

184

4. Discussion

4.1. Tag-recapture study

Recognizing that tag-recapture rates are linked to numerous factorsincluding natural mortality, C&R angling mortality, tag-induced mor-tality and tag shedding, it is generally not possible to estimate the re-lative contribution of all these processes unless a double-tagging pro-cedure or a captive study is used to quantify the respective components(Attwood, 2003; Pollock and Pine, 2007). Furthermore, it is well knownthat most of the mortality associated with C&R angling takes place post-release (i.e. delayed mortality) (Danylchuk et al., 2007; Brownscombeet al., 2017). Delayed mortality is divided into short-term mortalitywhich occurs within the first 72 h after release and is relatively easy toassess (e.g. Muoneke and Childress, 1994; Bartholomew and Bohnsack,2005; Cooke and Suski, 2005; Cooke and Wilde, 2007) and long-termmortality (> 72 h) which is more complicated to assess (Attwood,2003; Danylchuk et al., 2007; Pollock and Pine, 2007). Nevertheless,the high recapture rates achieved for the more resident reef-associatedspecies in this study (including up to seven multiple recaptures for someL. rivulatus individuals) is suggestive of limited immediate and delayedmortality following C&R angling. Such high recapture rates are seldomachieved in dart tag-recapture studies (Wilde and Sawynok, 2009;Dunlop et al., 2013) and this suggests that the fish capture and handlingmethods used in this study were successful in reducing mortality, par-ticularly for the more hardy species.

Despite the high recapture rates for some species, other species suchas R. sarba, T. botla and C. heberi had relatively low recapture rates.Although speculative, it is possible that these species have larger po-pulation sizes and/or exhibit more ranging movement behaviour (andthus moved out of the original 2 km sampling areas). However, re-capture rates did not improve after the sampling areas were extended inNovember 2011 (Mann et al., 2015). R. sarba appeared to be particu-larly susceptible to C&R mortality and accounted for 12 (21%) of the 58observed immediate mortalities recorded. This species fights particu-larly hard when hooked and quickly succumbs to air exposure unlessproperly handled (i.e. by immediately placing the fish into a bucket offresh seawater after capture). Mortalities also appeared to be higherwhen ambient conditions (air and water temperature) were warmer andlarger fish (> 450mm FL,∼1.5 kg) seemed to be more susceptible thansmaller fish (B. Mann, pers. obs.). Alternatively, most T. botla quicklyregained equilibrium and normally swam away strongly after release.However, it is known that after heavy physical exertion (i.e. the fight)and subsequent release, fish may require a substantial recovery periodduring which they exhibit reduced fitness or altered behaviour whichmakes them more susceptible to predation (Cooke et al., 2001;Danylchuk et al., 2007). T. botla are known to be important prey speciesfor large piscivores (e.g. Carcharhinus limbatus and Caranx ignobilis) andare frequently used as live-bait to target such predatory species (B.Mann, pers. obs.). Since the study area consisted of an open surf-zonealong an exposed coastline, these predators may have accounted for aproportion of post-release mortality as has been observed in otherstudies focusing on species such as Albula vulpes (Cooke and Philipp,2004; Danylchuk et al., 2007). Although relatively heavy tackle wasused to limit the fight time and reduce post release predation, theseresults indicate the C&R activities may have an impact on the speciescomposition in MPAs by reducing the numbers of species that are moresensitive to its impacts.

4.2. Trends in abundance

Use of nominal or raw CPUE data as an indicator of fish abundancehas been criticised as there are many other variables which can affectsuch trends (Maunder et al., 2006). To address this concern, Mann et al.(2016a) standardised the CPUE data and adjusted for selective targetingof the top four species caught during the first 10 years of this study.

These authors showed a significant increase in abundance of P. furcatus,T. botla, L. rivulatus and D. capensis in the previously exploited area andeither an increase (L. rivulatus and D. sargus), a stable trend (P. furcatus)or a slight decrease (T. botla) in the sanctuary area (the latter beingascribed to a subtle shift in targeting). Similarly in this study three ofthe six species analysed (P. furcatus, L. rivulatus and R. sarba) showed asignificantly higher CPUE in the no-take sanctuary areas (Table 2).Since the aim of this paper was simply to examine overall impacts of C&R angling, nominal CPUE data were pooled from all areas sampled toprovide a broad indication of trends in fish abundance. This showedsignificant increases in overall abundance of P. furcatus, L. rivulatus,D.capensis and D. petersi, a slight but insignificant decline in T. botla anda significant decline in R. sarba (Fig. 3). The slight decline in T. botla islikely to be as a result of a gradual change in targeting, with more effortbeing focused on surf-zone reef fish species as the angling team got toknow the area better (Mann et al., 2016a). Rather than being as a resultof angling mortality, the decline in R. sarba has been linked to the long-term closure of the nearby nursery area (St Lucia Estuary) as a result ofthe ongoing drought in northern Zululand and the negative impact thatthis has had on recruitment of this estuarine-dependent species (Mannand Pradervand, 2007).

4.3. Trends in mean size

A comparison between areas showed that two of the six speciesanalysed (T. botla and D. capensis) had significantly larger mean sizes inthe no-take sanctuary areas, while D. petersi were significantly larger inthe previously exploited areas (Table 3). Trends in mean size of the sixspecies pooled for all areas all showed a slight but insignificant increaseover the 12-year study period (Fig. 4). This is despite the observedlower recapture and higher mortality rates observed for R. sarba and T.botla. Mann et al. (2016a) showed a significant increase in mean lengthof P. furcatus, L. rivulatus and D. capensis in the previously exploitedareas over the first ten years of the study period indicative of stockrecovery, whereas no significant changes in mean size were detected inthe no-take sanctuary areas. These results indicate that the C&Rmethods used in this study did not have an overall negative impact oneither the abundance or size structure of the fish populations beingtargeted.

4.4. Hooking mortality

The observed capture mortality (i.e. immediate hooking mortality)was exceptionally low (0.87%) and tag-recapture results showed similarrecapture rates for both fish that had ingested the hook (14.8%) andthose that had not (15.2%). A more detailed study on the growth of L.rivulatus that had ingested hooks to those that had not, also revealed nosignificant difference (Mann et al., 2016c). While it is well known thattagging itself can have a negative effect on fish growth (Attwood andSwart, 2000), this result suggests that ingestion of barbless hooks has alimited effect on the mortality of fish subjected to C&R angling, aswitnessed by Wilde and Sawynok (2009).

4.5. Sub-lethal effects

It is known that capture by hook and line fishing may induce phy-siological stresses manifested by increased levels of lactate and cortisolin the blood, with recovery times being species dependent (Tomassoet al., 1996; Bartholomew and Bohnsack, 2005). Similarly, removing acaptured fish from the water and exposing it to air results in severebradycardia and elevated cardiac output (Cooke et al., 2001;Arlinghaus et al., 2007). Such sub-lethal effects may result in reducedgrowth rate and potentially reduced reproductive output (Cooke et al.,2006). While such impacts are difficult to quantify (Pollock and Pine,2007), it does suggest that C&R angling may have subtle long-termeffects on fish populations. This problem may be exacerbated in areas

B.Q. Mann et al. Fisheries Research 208 (2018) 179–188

185

exposed to intensive C&R angling where individuals might be capturedmultiple times. However, Mann et al. (2016c) revealed that growth of L.rivulatus recaptured multiple times compared to those that were onlyrecaptured once showed no significant difference. Although this findingis likely to be very species-specific, it does suggest that the impact ofeven intensive C&R angling can be mitigated to some extent by goodfish handling practises.

4.6. Associated environmental impacts of C&R angling

The environmental impacts of tackle loss during the current studywere considered to be relatively low (e.g. 69 kg of lead sinkers lost perannum over a ∼36 km stretch of coastline). Although this is minimal,when compared with the 24–190 sinkers/m2 which were found inSouth Wales (Cryer et al., 1987), the cumulative impacts of tackle losssuch as sinkers and fishing line (which have known negative environ-mental impacts) can be considerable (Jacks et al., 2001; O’Toole et al.,2009). Freshwater studies have suggested that lead can pose a threat towildlife such as birds which may ingest it (e.g. Scheuhammer et al.,2003). However, the primary concern in the coastal marine habitat,where waterfowl that could ingest lead are not present, would be dis-solution and the release of precipitates and chemicals into the water(Goddard et al., 2008). Fortunately, lead is stable under most en-vironmental conditions found in aquatic systems, suggesting that leadpollution in this fishery is unlikely to have negative impacts. Anotherseldom considered impact is that of broken-off traces with baited hookswhich then still hook fish which are unable to escape and either die orsuccumb to predation. Such impacts need to be considered in thecontext of the objectives of MPAs and again this is directly linked to theintensity of C&R angling that is permitted.

4.7. Is C&R angling compatible within South African MPAs?

Despite the positive results presented in this study, the answer tothis question is not simple or straightforward. Much discussion hasrecently taken place in South Africa around the issue of MPA zonation(NPAES, 2008; Harris et al., 2011; Sink et al., 2012). The most recentdecision (see discussion in Mills et al., 2015) has been to use four basiczonation categories, namely: Zone A) a no-take, no-go “sanctuary zone”where no human use is allowed except scientific research; Zone B) a“restricted use no-take zone” where only non-consumptive resource useis allowed (e.g. snorkelling, scuba diving, etc.); Zone C) a “controlled-pelagic use zone” where only harvesting of pelagic, migratory gamefishspecies is allowed; and Zone D) a general “controlled use zone” wherelimited consumptive use is permitted. Zones A and B above wouldcontribute towards achieving no-take protected area targets, whilezones C and D would only contribute towards achieving broader MPAtargets (NPAES, 2008).

Cooke et al. (2006) argue that C&R angling can be compatible withno-take MPAs under specific conditions and that with appropriateregulation and angler education, C&R can help enhance achievement ofconservation and management goals associated with MPAs, whilemaintaining public support and providing alternative tourism-basedrevenues for displaced fishers. Bartholomew and Bohnsack (2005) tooka more precautionary approach stating that while C&R angling is animportant activity, it does cause some disturbance, injury and mortalitywhich may conflict with some goals of no-take MPAs. In this respectthey mention the following points: 1) release mortality can jeopardizesurvival of large individuals which have a disproportionate contribu-tion to egg production; 2) cumulative mortality from C&R could bedamaging for long-lived species; 3) popularity of C&R may result in toomany anglers negating the population benefits of reduced mortalityfrom C&R compared to extractive fishing; and 4) allowing C&R in no-take MPAs makes surveillance and enforcement more difficult becausethe simple detection of the act of fishing is no longer sufficient evidenceof a violation.

Our findings support those of Bartholomew and Bohnsack (2005) inthat although the C&R angling we conducted was clearly a successfuland sustainable technique from a research monitoring perspective,there was some associated fishing mortality (particularly for sensitivespecies), sub-lethal effects and other environmental impacts. Theseimpacts would be exacerbated with increased angling effort, should thearea be opened to C&R angling by members of the public. These effectswould thus not be compatible with the conservation benefits associatedwith no-take MPAs (see review by Lester et al., 2009). However, ourfindings support the notion of designating zoned areas where C&R an-gling, as opposed to extractive harvesting, is allowed within MPAs(Bartholomew and Bohnsack, 2005; Cooke et al., 2006; Alós andArlinghaus, 2013). In this regard, C&R angling should be considered asan option in the controlled use zones C and D as described above.However, allowing C&R angling in Zone B (a no-take area where onlynon-consumptive use is permitted) is cautioned as there are some ne-gative impacts associated with C&R angling, which may be species-specific or go unnoticed due to delayed mortality.

Within the iSimangaliso Wetland Park and the St Lucia MarineReserve specifically, the area north of Cape Vidal to Leven Point iscurrently zoned to allow consumptive shore angling (IMP, 2016) butmost of this 22 km stretch of remote coastline is not accessible to shoreanglers as it extends beyond a reasonable walking distance (∼5 km).The same applies to the 23 km area south of Sodwana Bay to Red Cliffs(Fig. 1). While these areas may effectively be functioning as no-takeMPAs due to limited shore angler access (especially the southern area –Mann et al., 2016a), it is suggested that consideration be given to theestablishment of approximately 10 km “buffer areas” on either side ofthe no-take sanctuary area south of Leven Point and north of Red Cliffs.These buffer areas can be zoned as C&R areas and a concession can beprovided whereby a limited number of shore anglers are allowed accessto these areas under strict control (Mann et al., 2016b). While it isunderstood that there will be a degree of fishing mortality and sub-lethal effects as described above, with appropriate regulation and goodangler education (through the use of trained guides and implementa-tion of a strict C&R protocol), establishment of such zoned C&R areascould help to enhance the goals of this World Heritage Site. In thisrespect, a limited number of anglers would be allowed to access thisnear-pristine stretch of coastline and experience good quality angling.This would enhance tourism and revenue income to the Park but limitimpact on surf-zone fish populations. This could, therefore, be con-sidered to be an example of good contemporary conservation practice.

While the focus of this study addressed the compatibility of C&Rangling in MPAs, the results indicate that the designation of C&R onlyareas outside the boundaries of existing MPAs would also be beneficial.Management authorities are thus encouraged to consider this option inan attempt to reverse the over-exploited status of many coastal fisheryspecies in South Africa (Mann, 2013).

4.8. General comments regarding C&R angling

At this point it is important to highlight the fact that fishery man-agement agencies worldwide, including South Africa, have increasinglyput more reliance on measures that require mandatory C&R, such asminimum and maximum size limits, daily bag limits and closed seasons(Bartholomew and Bohnsack, 2005). Clearly the success of these mea-sures requires sufficient survival and subsequent reproduction by re-leased animals if they are to be successful. All marine recreational an-glers in South Africa have to abide by a suite of species-specificmanagement regulations (Mann, 2013). These regulations apply to allfish caught along the South African coast. Furthermore, the recreationalfishing industry in South Africa and elsewhere is increasingly en-couraging voluntary release of fish as means to improve sustainabilityand to expand recreational fishing opportunities (Bartholomew andBohnsack, 2005; Arlinghaus et al., 2007). For example, the SouthAfrican Shore Angling Association took a step forward in 1995 and

B.Q. Mann et al. Fisheries Research 208 (2018) 179–188

186

discouraged the killing of fish in shore angling competitions(Pradervand et al., 2007), recommending that members measure andrelease all fish caught with points determined using species-specificlength/weight regressions rather than by weighing individual fish. Thedevelopment of the new C&R only Rock-and-Surf Super Pro League isfurther evidence for the increasing trend of voluntary action by anglersthat is attempting to maintain sustainable recreational fisheries. Suchsteps should be supported and future attempts at improving the sus-tainability of recreational angling through C&R and other voluntarybehaviours should be strongly encouraged, including the involvementof recreational anglers in the design and management of MPAs(Danylchuk and Cooke, 2011). In this regard development of a C&Rprotocol, which promotes better survival of released fish, should bedeveloped and made widely available to improve angler support,awareness and behaviour (Cooke and Suski, 2005; Arlinghaus et al.,2010). For this, specific experiments further examining the impact of C&R on the physiology and survival of fish captured in marine recrea-tional shore fisheries are required.

Acknowledgements

The South African Association for Marine Biological Research sup-ported this project financially and logistically throughout its> 12-yearduration. We are also grateful to the following organisations for fundingreceived: Marine and Coastal Management (MCM)/National ResearchFoundation (2002-2005); MCM (2006-2009); iSimangaliso WetlandPark Authority (2011 onwards). Ezemvelo KwaZulu-Natal Wildlife isthanked for providing logistical support. Finally, this project would nothave been possible without the support and commitment of all 101anglers who participated during the> 12-year study period, we areextremely grateful to all of you.

References

Alós, J., Arlinghaus, R., 2013. Impacts of partial marine protected areas on coastal fishcommunities exploited by recreational angling. Fish. Res. 137, 88–96.

Arlinghaus, R., Cooke, S.J., Lyman, J., Policansky, D., Schwab, A., Suski, C., Sutton, S.G.,Thorstad, E.B., 2007. Understanding the complexity of catch-and-release in recrea-tional fishing: an integrative synthesis of global knowledge from historical, ethical,social and biological perspectives. Rev. Fish. Sci. 15, 75–167.

Arlinghaus, R., Cooke, S.J., Cowx, I.G., 2010. Providing context to the global code ofpractice for recreational fisheries. Fish. Manag. Ecol. 17, 146–156.

Attwood, C.G., 2003. Dynamics of the fishery for galjoen Dichistius capensis, with an as-sessment of monitoring methods. Afr. J. Mar. Sci. 25, 311–330.

Attwood, C.G., Bennett, B.A., 1995a. A procedure for setting daily bag limits on the re-creational shore-fishery of the South-Western Cape, South Africa. South Afr. J. Mar.Sci. 15, 241–251.

Attwood, C.G., Bennett, B.A., 1995b. Modelling the effect of marine reserves on the re-creational shore-fishery of the South-Western Cape, South Africa. South Afr. J. Mar.Sci. 16, 227–240.

Attwood, C.G., Swart, L., 2000. Discrepancy between otolith and tag-recovery estimatesof growth for two South African surf-zone teleost species. South Afr. J. Mar. Sci. 22,9–15.

Attwood, C.G., Mann, B.Q., Beaumont, J., Harris, J.M., 1997. Review of the state ofmarine protected areas in South Africa. South Afr. J. Mar. Sci. 18, 341–367.

Bartholomew, A., Bohnsack, J.A., 2005. A review of catch-and-release angling mortalitywith implications for no-take reserves. Rev. Fish Biol. Fish. 15, 129–154.

Bennett, B.A., Attwood, C.G., 1991. Evidence for recovery of a surf-zone fish assemblagefollowing the establishment of a marine reserve on the southern coast of South Africa.Mar. Ecol. Prog. Ser. 75, 173–181.

Bennett, B.A., Attwood, C.G., Mantel, J.D., 1994. Teleost catches by three shore-anglingclubs in the south-western Cape, with an assessment of the effect of the restrictionsapplied in 1985. South Afr. J. Mar. Sci. 14, 11–18.

Brouwer, S.L., Mann, B.Q., Lamberth, S.J., Sauer, W.H.H., Erasmus, C., 1997. A survey ofthe South African shore-angling fishery. South Afr. J. Mar. Sci. 18, 165–177.

Brownscombe, J.W., Danylchuk, A.J., Chapman, J.M., Gutowsky, L.F.G., Cooke, S.J.,2017. Best practices for catch-and-release recreational fisheries – angling tools andtactics. Fish. Res. 186, 693–705.

Butcher, P.A., Broadhurst, M.K., Orchard, B.A., Ellis, M.T., 2010. Using biotelemetry toassess the mortality and behaviour of yellowfin bream (Acanthopagrus australis) re-leased with ingested hooks. ICES J. Mar. Sci. 67, 1175–1184.

Casselman, S., 2005. Catch-and-Release Angling: A Review with Guidelines for ProperFish Handling Practices. Fish and Wildlife Branch, Ontario Ministry of NaturalResources, Peterborough, Ontario, pp. 26.

Cooke, S.J., Philipp, D.P., 2004. Behavior and mortality of caught-and-released bonefish

(Albula spp.) in Bahamian waters with implications for a sustainable recreationalfishery. Biol. Conserv. 118, 599–607.

Cooke, S.J., Suski, C.D., 2004. Are circle hooks an effective tool for conserving marine andfreshwater recreational catch-and-release fisheries? Aquat. Conserv. Mar. Freshw.Ecosyst. 14, 299–326.

Cooke, S.J., Suski, C.D., 2005. Do we need species-specific guidelines for catch-and-re-lease recreational angling to effectively conserve diverse fishery resources? Biodivers.Conserv. 14, 1195–1209.

Cooke, S.J., Wilde, G.R., 2007. The fate of fish released by recreational anglers. In:Kennelly, S.J. (Ed.), By-Catch Reduction in the World’s Fisheries. Springer, pp.181–234.

Cooke, S.J., Philipp, D.P., Dunmall, K.M., Schreer, J.F., 2001. The influence of terminaltackle on injury, handling time and cardiac disturbance of rock bass. N. Am. J. Fish.Manag. 21, 333–342.

Cooke, S.J., Danylchuk, A.J., Danylchuk, S.E., Suski, C.D., Goldberg, T.L., 2006. Is catch-and-release recreational angling compatible with no-take marine protected areas?Ocean Coast. Manag. 49, 342–354.

Cowley, P.D., Brouwer, S.L., Tilney, R.L., 2002. The role of the Tsitsikamma National Parkin the management of four shore-angling fish along the South-Eastern Cape coast ofSouth Africa. South Afr. J. Mar. Sci. 24, 27–35.

Cryer, M.J., Corbett, J.J., Winterbottom, M.D., 1987. The deposition of hazardous litterby anglers at coastal and inland fisheries in South Wales. J. Environ. Manag. 25,125–135.

Danylchuk, A.J., Cooke, S.J., 2011. Engaging the recreational angling community in theimplementation and management of aquatic protected areas. Conserv. Biol. 25,458–464.

Danylchuk, A.J., Danylchuk, S.E., Cooke, S.J., Goldberg, T.L., Koppelman, J., Philipp,D.P., 2007. Post release mortality of bonefish (Albula vulpes) exposed to differenthandling practises in South Eleuthera, Bahamas. Fish. Manag. Ecol. 14, 149–154.

Dunlop, S.W., Mann, B.Q., 2012. An assessment of participation, catch and effort in theKwaZulu-Natal shore-based marine line fishery, with comments on management ef-fectiveness. Afr. J. Mar. Sci. 34, 479–496.

Dunlop, S.W., Mann, B.Q., van der Elst, R.P., 2013. A review of the OceanographicResearch Institute’s Cooperative Fish Tagging Project: 27 years down the line. Afr. J.Mar. Sci. 35, 209–221.

Goddard, C.I., Leonard, N.J., Stang, D.L., Wingate, P.J., Rattner, B.A., Franson, J.C.,Sheffield, S.R., 2008. Management concerns about known and potential impacts oflead use in shooting and fishing activities. Fisheries 33, 228–236.

Harris, J.M., Livingstone, T., Lombard, A.T., Lagabrielle, E., Haupt, P., Sink, K., Schleyer,M., Mann, B.Q., 2011. Marine Systematic Conservation Assessment and Plan forKwaZulu-Natal – Spatial Priorities for Conservation of Marine and CoastalBiodiversity in KwaZulu-Natal. Ezemvelo KZN Wildlife Scientific Services TechnicalReport. 122 pp. .

Hilborn, R., Stokes, K., Maguire, J.-J., Smith, T., Botsford, L.W., et al., 2004. When canmarine reserves improve fisheries management? Ocean Coast. Manag. 47, 197–205.

IMP, 2016. iSimangaliso Wetland Park Integrated Management Plan (2011–2016).iSimangaliso Wetland Park Authority, St Lucia 207pp.

Jacks, G., Bystrom, M., Johansson, L., 2001. Lead emissions from lost fishing sinkers.Boreal Environ. Res. 6, 231–236.

Lester, S.E., Halpern, B.S., Grorud-Colvert, K., Lubchenco, J., Ruttenberg, B.I., Gaines,S.D., Airame, S., Warner, R.R., 2009. Biological effects within no-take marine re-serves: a global synthesis. Mar. Ecol. Prog. Ser. 384, 33–46.

Lewin, W.C., Arlinghaus, R., Mehner, T., 2006. Documented and potential biologicalimpacts of recreational fishing: insights for management and conservation. Rev. Fish.Sci. 14, 305–367.

Maggs, J.Q., Mann, B.Q., van der Elst, R.P., 2012. Long-term trends in the recreationalshore-fishery for elf (Pomatomus saltatrix) along the KwaZulu-Natal coast, SouthAfrica. Afr. J. Mar. Sci. 34, 401–410.

Maggs, J.Q., Mann, B.Q., Potts, W., Dunlop, S.W., 2016. Traditional management stra-tegies fail to arrest a decline in the catch-per-unit-effort of an iconic marine recrea-tional fishery species with evidence of hyperstability. Fish. Manag. Ecol. 23,187–199.

Mann, B.Q. (Ed.), 2013. Southern African Marine Linefish Species Profiles. SpecialPublication No. 9. Oceanographic Research Institute, Durban 343pp.

Mann, B.Q., Pradervand, P., 2007. Declining catch per unit effort of an estuarine-de-pendent fish, Rhabdosargus sarba (Teleostei: sparidae), in the marine environmentfollowing closure of the St Lucia estuarine system, South Africa. Afr. J. Aquat. Sci. 32,133–138.

Mann, B.Q., Tyldesley, M., 2013. Monitoring the recovery of a previously exploited surf-zone habitat in the St Lucia Marine reserve using a no-take sanctuary as a benchmark.In: Attwood, C., Booth, T., Kerwath, S., Mann, B., Marr, S., Duncan, J., Bonthuys, J.,Potts, W. (Eds.), A Decade after the Emergency: The Proceedings of the 4th LinefishSymposium. Langebaan, 16–20 April 2012. WWF South Africa Report Series – 2013/Marine/001. pp. 115–127.

Mann, B.Q., Taylor, R., Densham, D., 1998. A synthesis of the current status of marine andestuarine protected areas along the KwaZulu-Natal coast. Lammergeyer 45, 48–64.

Mann, B.Q., Cowley, P.D., Fennessy, S.T., 2015. Movement patterns of surf-zone fishspecies in a sub-tropical marine protected area on the east coast of South Africa. Afr.J. Mar. Sci. 37, 99–114.

Mann, B.Q., Winker, H., Maggs, J.Q., Porter, S.N., 2016a. Monitoring the recovery of apreviously exploited surf-zone fish community in the St Lucia Marine Reserve, SouthAfrica, using a no-take sanctuary area as a benchmark. Afr. J. Mar. Sci. 38, 423–441.

Mann, B.Q., Cowley, P.D., Kyle, R., 2016b. Estimating the optimum size for inshore no-take areas based on movement patterns of surf-zone fishes and recommendations forrezoning of a World Heritage Site in South Africa. Ocean Coast. Manag. 125, 8–19.

Mann, B.Q., Lee, B., Cowley, P.D., 2016c. Growth rate of speckled snapper (Lutjanus

B.Q. Mann et al. Fisheries Research 208 (2018) 179–188

187

rivulatus) based on tag-recapture data from the iSimangaliso Wetland Park, SouthAfrica. Afr. J. Mar. Sci. 38, 111–118.

Maunder, M.N., Sibert, J.R., Fonteneau, A., Hampton, J., Kleiber, P., Harley, S.J., 2006.Interpreting catch per unit effort data to assess the status of individual stocks andcommunities. ICES J. Mar. Sci. 63, 1373–1385.

Mills, M., Weeks, R., Pressey, R.L., Gleason, M.G., Eisma-Osorio, R.-L., Lombard, A.T.,Harris, J.M., Killmar, A.B., White, A., Morrison, T.H., 2015. Real-world progress inovercoming the challenges of adaptive spatial planning in marine protected areas.Biol. Conserv. 181, 54–63.

Muoneke, M.I., Childress, W.M., 1994. Hooking mortality: a review for recreationalfisheries. Rev. Fish. Sci. 2, 123–156.

NPAES, 2008. National Protected Area Expansion Strategy for South Africa. Priorities forExpanding the Protected Area Network for Ecological Sustainability and ClimateChange Resilience. Department of Environmental Affairs and Tourism and SouthAfrican National Biodiversity Institute Technical Report, 55pp.

O’Toole, A.C., Hanson, K.C., Cooke, S.J., 2009. The effect of shoreline recreational an-gling activities on aquatic riparian habitat within an urban environment: implicationsfor conservation and management. Environ. Conserv. 44, 324–334.

Pollock, K.H., Pine, W.E., 2007. The design and analysis of field studies to estimate catch-and-release mortality. Fish. Manag. Ecol. 14, 1–8.

Pradervand, P., Mann, B.Q., Bellis, M.F., 2007. Long-term trends in the competitive shore

fishery along the KwaZulu-Natal coast, South Africa. Afr. Zool. 42, 216–236.Radomski, P., Heinrich, T., Jones, T.S., Rivers, P., Talmage, P., 2006. Estimates of tackle

loss for five Minnesota walleye fisheries. N. Am. J. Fish. Manag. 26, 206–212.Ramsay, P.J., 1996. 9000 years of sea-level change along the South African coastline.

Quat. Int. 31, 71–75.Schaeffer, J.S., Hoffman, E.M., 2002. Performance of barbed and barbless hooks in a

marine recreational fishery. N. Am. J. Fish. Manag. 22, 229–235.Scheuhammer, A.M., Money, S.L., Kirk, D.A., Donalsson, G., 2003. Lead Fishing Sinkers

and Jigs in Canada: Review on Their Use Patterns and Toxic Impacts on Wildlife.Occasional Paper No. 108. Canadian Wildlife Service, Ottawa, Canada 48pp.

Sink, K., Holness, S., Harris, L., Majiedt, P., Atkinson, L., Robinson, T., Kirkman, S.,Hutchings, L., Leslie, R., Lamberth, S., Kerwath, S., von der Heyden, S., Lombard, A.,Attwood, C., Branch, G., Fairweather, T., Taljaard, S., Weerts, S., Cowley, P., Awad,A., Halpern, B., Grantham, H., Wolf, T., 2012. National Biodiversity Assessment2011: Technical Report. Volume: 4: Marine and Coastal Component. South AfricanNational Biodiversity Institute, Pretoria 325pp.

Tomasso, A.O., Isley, J.J., Tomasso, J.R., 1996. Physiological responses and mortality ofstriped bass angled in freshwater. Trans. Am. Fish. Soc. 125, 321–325.

Wilde, G.R., Sawynok, W., 2009. Effect of hook removal on recapture rates of 27 speciesof angler-caught fish in Australia. Trans. Am. Fish. Soc. 138, 692–697.

B.Q. Mann et al. Fisheries Research 208 (2018) 179–188

188