Is gear-based management of herbivorous fish a viable tool to prevent or reverse
phase shifts in coral reefs?
Linking resilience theory to practice
Quentin Dilasser
Is gear-based management of herbivorous fish a viable tool to prevent or reverse phase shifts in coral reefs?
Linking resilience theory to practice
Quentin Dilasser
Is gear-based management of herbivorous fish a viable tool to prevent or reverse phase shifts in coral reefs?
Linking resilience theory to practice
Master thesis (60 hp)
Stockholm Resilience Center, Stockholm University
Supervisors:
Dr. Magnus Nyström, Stockholm Resilience Center, Stockholm University Dr. Narriman Jiddawi, Inst. Marine Science, University of Dar es Saalam Phd student Matilda Thyresson, Dept of Systems Ecology, Stockholm University
Stockholm Resilience Center Stockholm University
Kräftriket 2B 10691 Stockholm
Sweden
Acknowledgment
This thesis is dedicated to my parents that always supported me.
This master thesis gathers the scientific results of this year-based project but does not capture the full picture of this journey and especially its wonderful human experience. Therefore, I would like to express my grateful to the people with whom I had the chance to work and live with.
First of all, I would like to sincerely thank my two supervisors Dr. Magnus Nyström and Matilda Thyresson for giving me the opportunity to work on a project I always dreamed about, that involve oceans and humans. Thank you also for giving me the possibility to work in a creative atmosphere, for sharing your knowledge and time and to believe in me enough to carry out this study. Also thanks to Mickael Tedengren, associate professor at the department of Systems Ecology for his help for the administrative work.
I would also like to thank my local supervisor, Dr. Narriman Jiddawi, for all the advises, contacts and sharing of information and also the research team from the Institute of Marine Sciences in Zanzibar for letting me using their materials and buildings as my base for the study.
Thank you to my translator Yussuf for being a friend and for its pertinent advices throughout the field study.
Special and sincere thanks to the Dugeish family for letting me the privilege to stay with them during my fieldwork and for considering me as a member of their family. Thank you for your moral support, for your contacts, for the fun you gave me and for taking care about me during difficult times. Without you this research would not have been possible to accomplish.
Thanks to the fishers communities of Malindi, Mazizini, Buyu and Nyamanzi that let me interview them but also to let me live good moments with them, thank you for your time and your willingness to share the valuable information you have.
Thanks to the Swedish International Development Agency (SIDA) that supported this thesis with a scholarship (MFS).
Abstract.
Herbivorous reef fish are a key functional group for the ecological resilience of coral reefs. As they feed on algae, a major resource competitor of coral polyps, they can prevent and reverse coral-macroalgal phase shifts. The resilience of the reefs against such phase shifts is given by the ability of herbivores to keep the system in a cropped state from filamentous algae or by their capacity to feed on macroalgae. Most of the management plans that aim to protect coral reefs have been focusing on the establishment of marine protected areas or no-take areas where fishing activities are strictly restricted or prohibited. In low-income countries, such managed areas can be difficult to accept from a fisher´s perspective and lack of money also tends to lead to limited surveillance capabilities and lowered compliance. These challenges are important to address when managing small-scale fisheries and where fish are considered as both, a marketable commodity and a subsistence good.
A perhaps less contentious strategy for fishers is gear-based management, where the use of fishing gears that are detrimental to coral reef resilience are restricted and at the same time gears that do not compromise resilience are promoted. This study aims to investigate how nine different fishing gears (i.e. different lines, traps, nets and spears) used in the coral reef fisheries of Zanzibar (Tanzania) capture herbivorous reef fish that can prevent (preventers) or reverse (reversers) coral-macroalgal phase shifts. Two interesting findings emerged from the study.
First, different fishing gears had different impacts on these two functional groups where lines, large traps and seine nets fisheries had most impacts. Second, there were monsoonal differences in the catch of preventers and reversers. These findings are discussed in relation to i) similar studies conducted in different reef environments and ii) the feasibility of gear-based management in Zanzibar.
Key words: Coral reefs, resilience, gear-based management, herbivores, preventers and reversers, phase shifts, fisheries, Zanzibar
Table of Contents
Acknowledgement………...3
Abstract………...4
1. Introduction………..…………...…...7
1.1 Problem statement………...…………7
1.2 Objectives………....8
1.3 Hypothesis………...9
1.4 Research question………....9
2. Conceptual framework……….…...10
2.1 Social-ecological systems and coral reefs resilience ………...……...10
2.2 Coral-macroalgal phase shifts………...……...…...11
2.3 Herbivores: a key functional group in coral reef dynamics………...…...12
2.4 Preventers and Reversers: a functional dichotomy………...…...………....14
2.5 Gear-based management………...………...…...15
3. Method………..…………..……...16
3.1 Study area……….………...……..…..16
3.1 a) Study site……….………...16
3.2 Coral reefs in Zanzibar………..……...17
3.2 a) Description of the reefs………...17
3.2 b) Multiple uses, multiple ecosystem services………..……...17
3.2 c) Threats to Zanzibar coral reefs ………..………18
3.3 Coral reef fishery in Zanzibar…..……….………..……...18
3.3 a) Generalities………...18
3.3 b) Monsoonal patterns………..20
3.3 c) Spatio-temporal variability……….…..20
3.3 d) The coral reef fish catch composition……….. 21
3.3 e) Signs of overfishing………... 21
3.3 f) Existing protected areas……….. .22
3.3 h) Fishing gears………...22
3.4 Epistemological frame……….…..……23
3.5 Field study……….………..…..24
3.5 a) Fishers interviewed (questionnaires).………...…….…….24
3.5 b) Procedure of the questionnaires..……….…….25
3.5 c) Categorization into preventers and reversers……….. 26
3.5 d) Data preparation and statistical analysis……….………….27
3.6 Limitations of the methodology……….………...28
4. Results………..30
4.1 Average daily catches (based on annual results)…………...………...30
4.2 Monsoonal variability……….…..33
5. Discussion……….…36
5.1 Fishing gears versus functional groups……….36
5.1 Monsoonal variability………37
5.2 Is gear-based management a real option?...37
6. Conclusion………..…..42
Literature cited………..……..44
Appendix 1: Fishing vessels in Zanzibar………...58
Appendix 2: a) Total catch of fish in Zanzibar. b) Fish catch in the three regions of Zanzibar………..60
Appendix 3: Coral reef fish composition of the catch and their name in Swahili……..61
Appendix 4: Fishing gears in Zanzibar………...62
Appendix 5: The different spatio-temporal drivers affecting the coral reef fisheries in Zanzibar………..…....67
Appendix 6: Interview outline………66
Appendix 7:Species list of herbivorous reef fish used in the study and their functional group affiliations………...71
Appendix 8: Average Bray-Cutis dissimilarities (<50%) between gears for the catch of preventers and reversers obtained by the SIMPER test…….…..74
Appendix 9: Average abundances of preventers and reversers caught by each gear and per day……….75
1. Introduction
1.1 Problem statement
Coral reefs are tropical, biogenic and shallow-clear water biophysical structures. They are primarily composed of scleractinian (stony) coral species that build a calcium carbonate skeleton and occur as colonies of polyps or in solitary form (Achituv & Dubinsky 1990;
Richmond 1997). Even though they compose less than 0.1% of the ocean floor (Spalding et al.
2001), these ecosystems are among the most complex and biologically diverse ecosystems (Odum & Odum 1955; Spalding et al. 2001; Green & Bellwood 2009) and are sometimes referred to as underwater tropical rain forests (Reaka-Kugla 1997).
Coral reefs provide essential ecosystem services (e.g. food supplies, aesthetic, cultural, recreational, tourism, physical barrier against storm) for millions of people (Costanza et al.
1997; Moberg & Folke 1999; MA 2005). Many coral reefs are located in developing countries (Donner & Portere 2007) and are thus capital ecosystems that generate socio-economic and ecological assets (Hughes et al. 2010). However, coral reefs are under multiple natural and anthropogenic stressor, primarily through overfishing, destructive fishing practices, pollution, climate change (increase sea surface temperature, ocean acidification, water dilatation) storms and diseases (Johnstone et al. 1998; Jackson et al. 2001; Hughes et al. 2003; Bellwood et al.
2004; Hoegh-Gulberg et al. 2007; Green & Bellwood 2009; Burke et al. 2011) that erode their ecological resilience and threaten their future survival. At present, 75% of the world´s coral reefs are threatened and one-third of the reef´s building species are under “elevated risk of extinction” (Carpenter et al. 2008; Burke et al. 2011).
Herbivory has been recognized as a key ecosystem process that underpin coral reef resilience as it maintains the balance between corals and algae for resources (e.g. Ogden et al. 1978; Hughes et al. 2003; Bellwood et al. 2004; Mumby et al. 2010; Norström et al. 2009; Cheal et al. 2010;
Burkepile et al. 2010; Lokrantz et al. 2010). Indeed, the loss of herbivory due to overfishing has been suggested as one of the most important drivers leading to undesirable ecological phase shift from coral dominated reefs to macroalgal dominated reefs (Hughes 1994; Nyström et al.
2000; Jackson et al. 2001; Hughes et al. 2003; Mumby 2006; Norström et al. 2009, Obura et al.
2009; Cheal et al. 2010) this is the most common type of shift observed on reefs (e.g. Bellwood et al. 2004; Hughes et al. 2007; Nyström et al. 2008). In a recent study, Mumby et al. (2010)
argued that reducing the exploitation of reef herbivores is an appropriate management goal to enhance coral reefs resilience. Additionally, coral reef herbivorous have the dual capacity of preventing (i.e. preventers) and reversing (i.e. reversers) coral-macroalgal phase shifts (e.g.
Hughes. 2007; Green & Belwood 2009; Mumby et al. 2010). Consequently, management strategies are needed to sustain abundant stocks of herbivores (i.e. preventers and reversers) that sustain the resilience and integrity of coral reefs, and hence the flows of goods and services.
Marine Protected Areas (MPA´s) or No-Take Areas (NTA´s) where fishing activities are restricted or prohibited have been suggested as a tool to manage herbivorous reef fish and protect/enhance coral reef resilience (Mumby et al. 2008 and 2010; Hughes et al. 2010).
Despite good results from the manager´s perspective (e. g.; Hughes et al. 2003 and 2007) these management strategies are not always appreciated by local resource users (McClanahan 1999;
Christie 2004; McClanahan et al. 2005 and 2006; Davies et al. 2009). Especially in low-income countries, marine reserves can exhibit unrealistic goals if it does not incorporate alternative sources of livelihoods for coastal communities that are already under economic pressures.
Additionally, in this context there are little capacities for their surveillance mainly because of a lack of financial support and compliance (Cinner et al. 2009a). Gear-based management (McClanahan et al. 2004, 2005 and 2008; Cinner et al. 2009a; Davies et al. 2009), that promote fishing gears that does not compromise coral reef resilience, has been put forth as a less contentious management strategy.
1.2 Objectives
This study has two distinct objectives. The first aims to explore one way to prevent and reverse coral-macroalgal phase shifts through the use of gear-based management that target herbivorous reef fish. The second objective was to investigate the relationship between the catch of functionally important herbivorous fish and gears used in artisanal coral reef fisheries in Zanzibar (Tanzania).
Coral reefs on the coast of Zanzibar are one example of reefs being under increasing anthropogenic stress and signs of overfishing and destructive fishing practices that are likely to affect key herbivorous fish have been observed (Johnstone et al. 1998; Tanzania State of the Coast 2001; Muthiga et al. 2008; Lokrantz 2009; Masalu 2009; Lokrantz et al. 2010).
Additionally, a recent study by Lokrantz et al. (2010) supports that herbivorous reef fish are
“susceptible to gears” and “are probably being impacted by artisanal fisheries” in Zanzibar.
The focus here is made on artisanal multi-species and multi-gear coral reef fisheries since they represent the main fishing activity in Zanzibar (Jiddawi et al. 2007). In the context of gear- based management the multi-gears features of the coral reef fisheries in Zanzibar is of importance, if restricting one specific gear fishers could shift to another gear.
Species compositions and abundances of the catch for eight different fishing gears (i.e. large and small traps, long- and hand lines, seine nets, gill nets with small and large mesh sizes and spear guns and octopus spearing) were assessed and discussed within the context of ecological resilience (Holling 1973), and coral-macroalgal phase shifts.
1.3 Hypothesis
The hypothesis used here is that catches of herbivorous reef fish that either prevent or reverse phase-shifts as well as catch yields vary with the selection of fishing gears used in the coral reef fisheries of Zanzibar. Hence, gear-based management could be used to either prevent or reverse phase shifts on coral reefs.
1.4 Research questions
Three questions will be addressed:
Are there any differences between fishing gears used in the coral reef fisheries in Zanzibar and their catches of different species of functionally important herbivorous fish that can prevent or reverse coral-macroalgal phase shifts?
Are there monsoonal differences in the catch of herbivorous fish that can prevent or reverse coral-macroalgal phase shifts?
Is gear-based management a plausible strategy to implement for coral reef fisheries management in Zanzibar?
2. Conceptual framework
2.1 Social-ecological systems and coral reefs resilience
This thesis considers both social and ecological systems as a unity, so called social-ecological systems (SES) (Berkes & Folke 1998). SES recognizes the complex interconnectedness and interdependence between coral reef ecosystem dynamics and processes and human societies and activities; linkages that are nested across spatio-temporal scales. This reflects that humanity is part of ecosystems and shapes them but also that societal development and well-being rely on the provision of ecosystem services (Costanza et al. 1997; Moberg & Folke 1999; Folke 2006;
MA 2005). Here, the SES of interest is the Zanzibar coral reef fisheries. There is growing scientific evidence suggesting that both social and ecological components are crucial to understand in order to attain sustainable coral reef fisheries (e.g. Cinner et al. 2009b).
Evidences also show the need to look at human activities as a major driving force that influences coral reef development trajectories of change (e.g. Nyström et al. 2000; Nyström 2001b; Hughes et al. 2003; Bellwood et al. 2004).
This thesis is built on the intertwined concepts of ecological resilience (Holling 1973) and complex adaptive systems (CAS) (Levin 1998). Ecological resilience (hereafter resilience) describes “the capacity of a system to absorb disturbances and re-organize so as still as to retain the same ecosystem processes, feedbacks and identity” (Walker & Salt 2006). Resilience is acknowledged by a rising number of scholars and they recognize coral reefs as more dynamic and complex ecosystems than previously thought (Holling 1986; Hughes 1994; Nyström et al.
2000; Scheffer et al. 2001; Folke et al. 2004; Hughes et al. 2005; Kinzig et al. 2006; Mumby et al. 2008; Nyström et al. 2008) and that they are becoming increasingly vulnerable to changes (e.g. storms) by losing their resilience (e.g. Nyström et al. 2000; Hughes et al. 2003; Bellwood et al. 2004; Nyström et al. 2008). Such changes are translated into reality as through the existence of critical thresholds, tipping points where coral reefs can shift to undesirable alternative ecosystem regimes (e.g. from coral to macroalgal dominated states) and are locked in this new ecological configuration by positive feedbacks (e.g. Hughes 1994; Nyström et al 2008; Norström et al. 2009; Mumby & Steneck 2008). This view of the dynamics of ecosystems challenges the standard Clementsian successional paradigm in ecology (Norström et al. 2009) that advocates that ecosystems evolve towards one single regime (equilibrium dynamics) where climax communities are established (Clements 1916).
Both coral and algal dominated reefs can be resilient. Nevertheless, from an anthropocentric perspective algal dominated reefs exhibit a negative resilience whereas the resilience of coral dominated reefs is considered as positive (see also Box 1 in section 2.2). Therefore, in this thesis the use of resilience refers to the ability of coral reefs in the face of change to resist phase shifts towards a macroalgae dominated state and reorganize after disturbances but also in its capacity to reverse such phase shift when it has already occurred.
Phase shifts are difficult to predict (Scheffer et al. 2003; Nyström et al. 2008). To avoid unwanted ecological surprises, such as coral-macroalgal phase shifts, safeguarding coral reef resilience has been promoted as a key management goal (Hughes et al. 2003; Bellwood et al.
2004). Nonetheless, little is known about how to prevent and reverse such phase shifts (Bellwood et al. 2006; Cheal et al. 2010; Hughes et al. 2010) and about how to manage coral reef fishing activities that allow fishing practices but at the same time retain the provision of key ecosystem processes such as herbivory (Mumby et al. 2008; Cinner et al. 2009a and b).
Finally, there are also no studies assessing particular fishing gears and their roles in coral- macroalgal phase shifts. This study brings reflections on how to fill these gaps using gear-based management (McClanahan et al. 2005; McCalanahan & Cinner 2008; Cinner et al. 2009a), i.e.
the restriction on specific types of fishing gears that erode the resilience of the reefs but also the promotion of others that do not comprise resilience.
2.2 Coral-macroalgal phase shifts
A shift from coral to macroalgal dominated state provides the archetypical example of a coral reef phase shift (Box 1), although other shifts occur (Norström et al. 2009). Field-based, experimental and modeling work have built a strong evidence base for coral-macroalgal phase shifts. Work on Jamaican reefs in the beginning of the eighties documented a dramatic shift towards fleshy brown macroalgal (Sargassum sp.), unpalatable to most herbivores, (Carpenter 1990; Hughes 1994; Steneck & Dethier 1994; Nyström et al. 2000; Scheffer et al. 2001 and 2003; Mumby et al. 2008; Norstöm et al. 2009) after the die-off of the Caribbean sea-urchin Diadema antillarum. Experimental evidence in the Great Barrier reef using cages (exclusion of herbivorous fish) (Hughes et al. 2007), video bioassay approach, transplanted algal assay (Bellwood et al. 2006; Mantyka et al. 2007; Cvitanovic & Bellwood 2009; Cheal et al. 2010) and algal removal experiments in Belize (McClanahan et al. 2000) have shown that there is a negative relationship between the increase in macroalgae cover and fecundity, recruitment rates
and survival of coral polyps. Modeling also suggests (Mumby 2009) a positive relationship between space-opening and macroalgal growth at the expense of polyps growth and the predominant role of grazing intensity to remove macroalgae for colonization and growth success of polyps.
Shifts to macroalgae dominated states, considered as undesirable (Box 1), exhibits positive feedbacks that lock the system in this new ecological configuration and makes it even more difficult to reverse. Positive feedbacks are loops of
reciprocal actions between two components that tend to amplify each other´s roles (Chapin et al. 2009;
Cinner et al. 2011; Hoey and Bellwood 2011;
Nyström et al. in review). As an example of such positive feedbacks, William et al. (2001) in their study about reefs in Belize suggested that once there is enough space opened up for algal colonization this can exceed the “grazing capacity of herbivorous to crop down fleshy macroalgae” (McClanahan et al.
2001; Williams et al. 2001) that lead inevitably to temporally stable coral-macroalgal phase shifts. Some authors theorized that, in some cases, the level of the driving variable (i.e.
herbivory), needed to reverse a non desirable phase shift towards a coral reef dominated state, might differ between forward and backward shift; this is known as hysteresis (e.g. Sheffer et al.
2003; Nyström et al. 2008).
2.3 Herbivores: a key functional group in coral reef dynamics
Functional groups are “species that perform a similar function, irrespective of their taxonomic affinities” (Bellwood et al. 2004). This definition can be seen as a synonym of ecological guild but the use of the functional groups approach in this study is not only driven by the common diet of certain species but on performing particular functions that underpin ecosystem processes upon which coral reefs depend (e.g. bioerosion).
Box 1: Desirability versus undesirability
Un- and desirability are anthropocentric concepts that quantify the benefits for societies from the provision of ecosystem services between alternative states. In this context, coral dominated reefs are widely recognized as a desirable ecological state for the fulfillment of required ecosystem services for societal development (Moberg
& Folke 1999; Folke et al. 2004, Nyström et al. 2008; Nyström et al. in review).
Field-based, experimental and modeling evidences (Ogden et al. 1978; Bellwood et al. 2004; Arthur et al. 2006; Hughes et al.
2007; Nyström et al. 2008; Mumby 2009) show the central role that herbivorous reef fish (e.g. parrotfish, rabbitfish) play with regards to resilience in coral reefs. They perform one of the most important ecosystem functions on reefs through the controlling and removal of filamentous algae that constantly grow on coral structures and macroalgae (e.g. Sargassum sp.) that colonize reefs after a disturbance. Both types of algae compete with coral polyps mainly for space and light (e.g. Green & Bellwood 2009; Lokrantz 2009).
Herbivorous reef fish play different but complimentary roles and are usually categorized into four functional groups: scrapers/small excavators (e.g. small parrotfish), bioeroders/large excavators (e.g. large parrotfish) that feed on epilithic filamentous algae that grow on coral, grazers/detritivores (e. g surgeonfish, rabbitfish) that feed on macroalgae and on epilithic material and browsers that feed only on macroalgae (e.g. rudderfish, rabbitfish) (Green &
Bellwood 2009; Lokrantz 2009).
Studies show that overfishing of key functional herbivorous groups is one of the major factor behind erosion of reef resilience, through the loss of top-down algal controllers that might lead to macroalgal phase shifts (e.g. Bellwood et al. 2004; Pauly 2005; Hughes et al. 2007; Mumby et al. 2007; Norstöm et al. 2009; Green & Bellwood 2009; Lokrantz 2009; Cheal et al. 2010;
Lokrantz et al. 2010). For instance, Mumby et al. (2007) argued that an unexploited parrotfish community in Caribbean reefs can maintain “approximately 40% of the reef in a permanently grazed state but overfishing reduces this capacity to about 5%”.
Box 2: Functional redundancy and response diversity
Functional redundancy (Walker et al 1992) and response diversity (Elmqvist et al. 2003) are two key properties within functional groups (Nyström. 2006).
Functional redundancy describes the capacity of species within a group to functionally compensate for the loss of species and still perform the same ecological contribution.
Response diversity refers to the variety of response carried by different species to oppose impacts of disturbances to the system.
Functional redundancy and response diversity are important for coral reefs resilience as they both act as biological insurance to successfully cope with disturbances.
2.4 Preventers and Reversers: a functional dichotomy
Experimental evidence has shown that reef herbivorous fish have both the capacity to prevent and reverse coral-macroalgal phase shifts (Cheal et al. 2010; Bellwood et al. 2006a; Mantyka et al. 2007;
Cvitanovic & Bellwood 2009). In a recent study Cheal et al. (2010) stressed the need to make “the distinction between reversal and prevention of phase shifts” since different species are likely to be involved in prevention and reversion processes.
These experiments also highlight that the number of species involved in reversion processes are fewer than the number of species involved in the prevention processes. Finally, herbivorous species richness and abundance are also important factors for insuring herbivore function on reefs. Indeed, complimentary feeding behaviors are profitable for preventing and reversing coral-macroalgal phase shifts (Burkepile et al. 2008). Consequently, herbivores can be divided in two distinct functional groups, i.e.
preventers and reversers (Fig. 1). Preventers (e.g. Scarus ghobban, Scarus niger, see Fig. 2) feed on filamentous algae that constantly compete for resources with coral polyps, in that sense preventers are important against coral-macroalgal phase shifts as they help insure the presence of coral polyps on reefs. Reversers can feed on fleshy algae and coral epilithic material (e.g.
parrotfish of the genus Calotomus and Leptoscarus, see Fig. 3; Bellwood et al. 2009). For example, in their study Bellwood et al (2006a) discovered a reverser species: the batfish (Platax pinnatus) that was able to remove major large macroalgae (Sargassum) in the Great Barrier Reef even when if it has reached a canopy of several meters in height.
Fig. 2: Two different species of preventers: Scarus Ghobban (left) and Scarus niger (right). Source: J. E. Randall.
In other words, preventers are species that prevent coral-macroalgal phase shift and reversers are species reversing coral-macroalgal phase shifts when it has already occurred (Fig. 1, 2 and
Platax pinnatus
3). Subsequently, fishing that target preventers and reversers can compromise coral reefs resilience to cope with such phase shifts. Hence, managing fishing gears could be one route to balance the distribution and abundances of preventers and reversers in the reefs.
Fig. 3: Two different species of reversers Calotomus carolinus (left) and Leptoscarus vaigiensis (right). Source: J.
E. Randall.
2.5 Gear-based management
Gear-based management (McCalanahan et al. 2004, 2005 and 2008; Cinner et al. 2009a) refers to the restrictions of specific types of fishing gears that target ecologically important fish and the selectivity of non-detrimental gears for the resilience of coral reefs since the species composition of the catch is correlated with the type of fishing gear used (e.g. Jiddawi et al.
2002; McClanahan et al. 2004). Indeed, fishing gears have their intrinsic strength and weaknesses in terms of use and conservation of resources (McClanahan et al. 2008), thus this element could provide a framework for resources management.
A gear-based management provides an alternative to marine protected areas (MPA´s) and no- take areas (NTA´s) that could be of great interest for managers because it is suggested to be less contentious to the fishers (McClanahan 2005). This is due to the fact that coral reef fisheries are multi-gears activities; hence fishers could shift to another type of fishing gears.
This is especially relevant when looking at small-scale/artisanal fisheries where fishers are highly dependent on the fishery for their subsistence and when there are few livelihoods options (Davies et al. 2009).
3. Methods
3.1 Study area
The study was conducted between November 2010 and February 2011 on the southern island of Zanzibar, known as Unguja (hereafter Zanzibar) (Fig 4a). Zanzibar is located in the Western Indian Ocean area (6°S and 39°E) about 35 km off mainland Tanzania. Zanzibar is a semi- autonomous political unit within the United Republic of Tanzania and is administrated by its own constitution and government (Constitution of the United Republic of Tanzania 2010).
3.1 a) Study sites
The study was conducted in four landing sites at the western side of Zanzibar; Malindi, Mazizini, Buyu and Nyamanzi following a southward direction (Fig. 4b). Buyu and Nyamanzi are relatively small landing sites (between 110 to 125 vessels/site) compared to Malindi and Mazizini (between 230 to 395 vessels/site). These sites were chosen with regards to a previous study conducted by Lokrantz et al. (2010).
Most of the fishing gears used in the four landing sites are: large and small traps, lines (hand lines and long lines), nets (gill and seine nets) and spear guns (Fig. 5).
Fig. 4: a) Location of Zanzibar island (Unguja) on the coast of Tanzania.
b) Zanzibar island and the four landing sites (Malindi, Mazizini, Buyu and Nyamanzi). Adopted from Lokrantz et al. (2010).
a) b)
3.2 Coral reefs in Zanzibar
3.2 a) Description of the reefs
Most reefs in the western side of Zanzibar are shallow (<10m) and they fringe small islands and sandbanks that are scattered along the coast (Lokrantz et al. 2010). Live coral cover is dominated by scleractinian species (i.e. stony corals) and reefs have a low macroalgal cover.
Like in many other reefs some algal turfs have been recorded (Lokrantz et al. 2010).
3.2 b) Multiple uses, multiple ecosystem services
Coral reefs in Zanzibar are primarily used as, fishing grounds and as a major tourist attraction (e.g. Tanzania Coastal management partnership 2000, Tanzania State of the Coast 2001).
Living and dead corals are also exploited for lime collection through coral mining activities to produce cement, building material and white color paint for houses (Muthiga et al. 2008). Reef areas are also indirectly used for seaweed farming particularly cultivated in lagoons, protected by coral reefs, and/or within reef grounds and for intertidal shell harvesting for food and ornamental purposes (Lirasan & Twide 1993; Tanzania Coastal management partnership 2000;
Tanzania State of the Coast 2001; Jiddawi et al. 2002).
0%
50%
100%
Malindi Mazizini Nyamanzi Buyu
Fishing gear distribution (%)
Handline and longlines
Large and small traps
Gill nets and seine nets
Spear
Others (fish weirs, scoop net, beach seine, purse seine, ring net, cast net)
Fig. 5: Fishing gears in the four sites studied (Buyu, Mazizini, Nyamanzi and Malindi) Source: Jiddawi et al.
(2007).
3.2 c) Threats to Zanzibar coral reefs
Zanzibar’s reefs face multiple anthropogenic stressors. Reefs show clear signs of overexploitation and destructive fishing practices can occur (Johnstone et al. 1998; Tanzania State of the Coast 2001; Masalu 2009; Lokrantz et al. 2010). Climate change has caused coral bleaching due to increasing seawater temperature, eustatism, acidity and reduced salinity and oxygen, which have resulted in significant coral mortality. Human-induced disturbances also include disease outbreaks, land-based inputs of sediments and nutrients that reduce water quality (Hughes et al. 2003; Bellwood et al. 2004; IPCC 2007; Cinner et al. 2009b; Green &
Bellwood 2009; Masalu 2009), tourism degradation of the reefs (e.g. diving, snorkeling, reef walking), ship groundings (Muthiga et al. 2008; Masalu 2009) and overuses of coral resources (Johnstone et al. 1998; Gösseling et al. 2004). In addition, biological perturbations such as outbreaks of corallivorous crown-of-thorn sea stars (Acanthaster planci) also act as important sources of disturbances on Zanzibar reefs (Muhando & Lanshammar 2008; Muthiga et al. 2008;
Lokrantz et al. 2010).
3.3 Coral reef fishery in Zanzibar
3.3 a) A general overview
The coral reef fishery in Zanzibar is both a traditional and artisanal/small-scale activity that requires a relatively low capital investment. It embraces multiple gears that catch a myriad of species for commercial and subsistence purposes and this fishery represented 96% of the total fish catch in Zanzibar in 2007 (e.g.
Haekstra et al. 1990; Jiddawi et al. 2002 and 2007; Jiddawi &
Pandu 1987). The fishing mainly occurs in shallow (<20m depth) and easy accessible waters close to fishing villages and landing sites (Jiddawi & Pandu 1987; Lokrantz et al. 2010). Lokrantz et al (2010) argued that most of the reef fishing activity occur within a narrow strip of 1.5 to 2 km to the landing site and is based on daily movements because of fuel costs or time constrains for sailing boats to go further offshore. Indeed, the artisanal fishery is predominantly based on traditional and relatively small unmotorized crafts and “low-tech”
fishing gears (e.g. hand lines, traps) (e.g. Review of marine fisheries for Tanzania 2003;
Jiddawi et al. 2007). These fishing crafts are: dugout (mtumbwi) and outrigger canoes (ngalawa), dhows (dau), mashuas and boat (boti) (Hoekstra et al. 1990; Jiddawi et al. 2002 and 2007; Jiddawi & Pandu 1987) (Appendix 1). The most frequently used type of boat for fishing is ngalawa (Haekstra et al. 1990). These fishing vessels are wooden planked boats (except for mtumbwi and ngalawa which are made from a single log) and usually propelled by poles, paddles and sails. Nevertheless, Haekstra et al. (1990) argued that about one quarter of the fishery operates without fishing vessels. In addition, freezing facilities in boats and at landing sites rarely exist thus preventing fish storage processes (Jiddawi et al. 2002). Therefore, the fish is mostly sold fresh, smoked, salted or cooked (Jiddawi et al. 1997).
Fish represent the major source of animal protein in Zanzibar; the annual consumption per capita is high and ranges between 17 and 40 kg (Johnstone et al. 1998; Jiddawi et al. 2002 and 2007; Jiddawi & Pandu 1987). To put this in contrast, the annual consumption in Sweden is around 6 kg (FAO 2003). The majority of fish are caught, sold and consumed locally (Thyresson et al. in press). However, fish can be regionally transported within Zanzibar boundaries particularly to Zanzibar Town (Urban region) or to Tanzania mainland and primarily in the economic capital Dar Es Salaam (MBCA 2005) and for some species like the sea cucumber internationally exported (Eriksson et al. 2010).
Fishers in Zanzibar are among the poorest in Tanzania (Suleiman 1999) and are more dependent on fish resources than in the mainland, primarily because of the limited availability of arable land (Jacquet & Zeller unpublished data). Fisheries also represent one of the most important coastal occupation and sources of income for coastal populations in Zanzibar (Jiddawi et al. 2007). It is a predominant economic driver and represents a large part of the Zanzibar GDP (2,2-10,4%) (Jiddawi et al. 2002). The number of full-time fishers has been estimated to between 18 000 and 23 000 in Zanzibar archipelago (Tanzania State of the Coast 2001; Jiddawi et. 2002 and 2007). However, a large number of people are indirectly committed to the fishery through for example: boat construction-repairing, trading (middlemen and retailers), making fishing gears and mending and fish processing (dryers, salters, etc.) (Masalu 2009).
Within the fishing communities more or less everyone is involved in fisheries. Whereas coral reef and pelagic fisheries are practiced by men only, intertidal fishery (mainly bivalve collection) and seaweed gleaning are performed by women, children and elders (Tanzania State
of the Coast 2001; Jiddawi et al. 2002 and 2007).
Women also play a central role in the processing and trade with fish (Tanzania State of the Coast 2001;
Jiddawi et al. 2002 and 2007; personal observations).
Fig. 7: Children and women collecting bivalve in Mazizini
3.3 b) Monsoonal patterns
The island is characterized by two distinct monsoonal seasons that involve a complete reversal in wind direction. The southeast monsoon (Kusi) occurs from June to September and is distinguished by a southeast wind whereas the northeast monsoon (Kaskazi) occurs from November to March and is defined by a northeast wind. This latter is characterized by a short rainy period, higher air temperature as well as a constant and light wind and weaker currents;
thus it is the period where most of the fishing activities take place (Richmond et al. 1997;
Jiddawi et al. 1997; Tobisson et al. 1998; Jiddawi et al. 2002).
3.3 c) Spatio-temporal variability in coral reef fisheries
Although fishing is practiced all year around, the extraction activity exhibits a temporal variability in fishing pressure. The fishing peaks between November and February (Kaskazi) when local and migrating fishermen (dago) (Richmond et al. 1997; Jiddawi et al. 1997 and 2002) benefit from mild environmental conditions.
Coral reef fishery is an open-access activity where anyone can participate (Jiddawi et al. 2002).
This has resulted in that the number of fishers has increased overtime, including also dago fishers (Jiddawi et al. 1997). There are three types of dago fishers: the fishers coming from mainland Tanzania and fishers coming from the island of Pemba (north of Unguja) and the fishers coming from Zanzibar (Unguja) itself but that travel locally within the island. Dago fishers increase the fishing effort on coral reefs in Kaskazi (Jiddawi et al. 2002) and add significant quantities of fish in the local market. There is also spatial variability in fishing patterns in Zanzibar. The western part of the island displays a three times higher fishing activities than in the South Zanzibar region and almost two times higher than in the north of Zanzibar (Appendix 2). These differences are likely to be due to the distribution of coral reefs since the highest coral covers are found around Zanzibar Town and the north of the island, but
also due to the distribution of human communities on the island where the western part is by far the most populated area (FAO CountrySTAT b).
3.3 d) The coral reef fish catch composition
Coral reef fishing is a multi-species activity and the most common fish caught are found within families like Labridae1 (parrotfish/Pono, wrasses/Chewa), Siganidae (rabbitfish), Lutjaenidae (snappers), Acanthuridae (unicornfish/Kangaja, Puyiu), Lethrinidae (emperors/Changu) and Mullidae (goatfish/Mkundaji) (Appendix 3) (personal observations; Johnstone et al. 1998;
Jiddawi et al. 2002; Jiddawi & Khatib 2007)
3.3 e) Signs of overfishing
The coral reefs around Zanzibar exhibit signs of overexploitation (Johnstone et al. 1998;
Tanzania State of the Coast 2001; Jiddawi et al. 2002; Review of marine fisheries for Tanzania 2003, Muthiga et al. 2008; Lokrantz 2009; Masalu 2009) in terms of annual decline in coral reef fish catch (Haekstra et al. 1990; Jiddawi et al. 2002), high abundance of sea urchins (i.e.
indicator of loss of fish predators) and lower abundance of large fish (>50cm body size) (Lokantz et al. 2010). Tyler et al. (2009) also demonstrated a positive relationship between fishing pressure on commercial species and their depth distribution (i.e. depth refuge effect) on reefs in Zanzibar.
3.3 f) Existing protected areas
With the exception of the Chumbe Island Marine Park and the Menai Bay Conservation Area most of the reefs around Zanzibar are open-access fishing grounds (Jiddawi et al. 2002), i.e.
anyone, having a fishing license, is allowed to fish.
1 Family Scarines that is better known as Labridae, formerly called Scaridae (Westneat et al. 2005)
Chumbe Reef Sanctuary (CHICOP) located at Chumbe Island was established in 1994 through a private initiative (Riedmiller 1998). It is a privately managed “no-take” area and fishing activities have been banned since 1994 (Riedmiller 1998; Reef sanctuary-Chumbe Island Coral Park 2010). CHICOP is open for eco-tourism (e.g. snorkeling and scuba diving) since 1998 and is also used for educational purposes.
In response to the increasing number of fishers and the use of destructive fishing practices, the Menai Bay Conservation Area (MBCA) was established in 1997 in the southwest part of Zanzibar (Assessment of MBCA 2005; Torell et al. 2006). The area is managed simultaneously by the seventeen local communities living around the MBCA, the Zanzibar Department of Fisheries2 and it receives financial and technical supports from the World Wildlife Fund (WWF). This area is open for tourism (e.g. whale and dolphins watching) and fishing activities are under tougher regulations than in other parts of the island.
The Zanzibar Stone Town Conservation Area forms a square management domain and gathers four reefs, i.e. Changuu, Bawe, Pange and Bat Grave. It is a new project but management strategies and the boundaries of the conservation area are still under debate (Muthiga et al.
2008).
3.3 h) Fishing gears
Fishing gears used in Zanzibar can be divided in lines (hand lines/Mshipi, long lines/Dhulumati), nets (gill/Jarife, cast/Kimia, seine/Nyavu, ring/Kuzunguke, beach/Juya, mosquito/Mtundo), traps (portable/Dema and Towe and fixed/Uzio) and spear fisheries (spear guns/Bunduki and octopus spearing/Umangu) (Appendix 4). Spear guns and beach seine nets are illegal in Zanzibar but are still practiced (Jiddawi
et al. 2002). Moreover, dynamite and poison fishing have been sporadically reported (Johnstone et al. 1999). Both are illegal. Lines are the most commonly practiced fishing types followed by moveable
2 In 2005 the Marine and Coastal Environments Management Program (MACEMP) was created by the Global Environmental Fund and the World Bank. MACEMP is an important organization controlling fishing regulations in the MBCA (MACEMP PAD Draft 2005; Torell et al. 2006)
Reminder:
Fishing gears are important aspects to study since the catches of different species are correlated with the selection of fishing gears (section 1. 2.)
Fig. 9: Towe traps.
trap fisheries (Hoekstra et al. 1990).
Portable traps are generally home-made and built from flexible pieces of woods (e.g. Muenzi, Mehi kichi) finely cut and attached together with plastic raffia interwoven in hexagonal patterns (personal observations, Fig. 9). In some cases, they can also be made from steel or aluminium (in this case they are called dema wire).
Hand lines are made of single nylon lines where one or two hooks are baited. In contrast, long lines consist of two types of nylon lines; a thick main line along which hundreds of secondary thinner lines are attached to it with baited hooks. A single long line is generally operated by two fishermen (personal observation).
Gill- and seine nets both represent a wall of netting; usually gill nets are made from a thicker nylon than seine nets. Both gears are destructive when used on coral reefs. There are different types of gill nets, with small, medium, large and mixed mesh sizes. Gill netters either set their net on the surface or at the bottom (personal observation).
Most spear guns and octopus spearing gears are home-made and both propel a spear to a prey.
They differ because spear gunners propel their spear by pulling the trigger of the spear gun and octopus spearers manually throw their spear. Spear gun is a more expensive gear to make than octopus spear (personal observation). Octopus spearing mainly target octopus but opportunistically catch fish rather than spear guns that mainly focus on fish.
3.4 Epistemological frame
This research strives to be holistic and integrative as to incorporate the multifaceted (socio- economic and ecological) problem of preventing and reversing coral-macroalgal phase shifts.
Nevertheless, only herbivores and their catches demarcate the scope of this study. Indeed, it is methodologically difficult to integrate and collect reliable data about all the drivers (e.g.
climate change, terrestrial run-off, tourism, population growth) affecting the reefs in a short period of research (Appendix 5). Hence, one has to retain the complexity of the coral reef fisheries system in Zanzibar when interpreting and discussing results from this study. These limitations will be addressed in the discussion (section 5.).
In this thesis a hypothetico-deductive methodology was applied. A hypothesis was formulated and empirically verified through the collection of data framed with an experimental protocol (based on interviews).
This thesis is both post-normal-oriented (Popper 1970; Funtowicz and Ravetz 1992) and normal-oriented (Kuhn 1962). In essence, this reflects that the thesis deals with uncertainties in the use of its theoretical framework, as all scientific work, but decisions are also urgent to take in the face of ecosystem change. Paradigms of resilience, phase shifts and functional redundancy are used here as unquestioned concepts but, despite a growing recognition of their relevance to scientifically depict reality, these concepts are still under debate (e.g. Fong et al.
2006; Nyström et al. 2008). Hence, “critical rationalism” (Popper 1945) is needed when reading this monograph so to remember that these questionable concepts constitute its backbones.
To finish, this research assumes being positivist in its Comtian sense; a scientific methodology to collect data has been applied to answer the question this study ask. In addition, the study was conducted in an effort of epistemological realism and objectivism to depict, as close as possible, the “reality”.
3.5 Field study
3.5 a) Fishers interviewed (questionnaires)
Interviews were conducted with active coral reef fishermen practicing a multi-species coral reef fishery and operating with fishing gear(s) from a boat. Questionnaires specifically aims to collect data with regards to presence/absence and number of herbivores (i.e. preventers and reversers) caught by each gear per day. In total 175 questionnaires were conducted in Malindi (n = 47), Mazizini (n = 46), Buyu (n = 43) and Nyamanzi (n = 39). The fishers being interviewed ranged from young to elder and only men were questioned since it is mainly men practicing this type of fishery. Only one fisher per boat was interviewed. Questionnaires were conducted following a stratified sampling strategy based on type of gear to ensure a representation of all reef-associated gears used at each site.
Each questionnaire took between 15 min to 1, 5 hours depending on the gear used and the number of herbivorous species captured by the fisher.
3.5 b) Procedure of the questionnaires
The questionnaire (Appendix 6) was written in English but held in Swahili (official language spoken in Zanzibar) with the assistance of a translator. Pilot questionnaires were conducted to assess the relevance of the questions.
The interviews were designed to elicit information with regards to i) fishing gears, ii) their links to catch of different species, and iii) monsoonal variations. The core of the questionnaire was based on displaying pictures to the fishers of 29 species of herbivores (see also section 3.5 c).
The task for the fishers was partitioned in two sets of actions:
A) To determine which species were caught with which gears (presence/absence). The fishers were asked to split up the herbivores pictures in two piles, those that were being caught and those that were not. This was done for each gear used.
B) When the phase A was finished, questions to assess the number of (abundance) herbivores being captured by each gear and per day were asked. Two different methods were used due to progress of the methodology during the course of the study (see Box 3).
i) In the first method (for n = 34 interviews), the interviewer asked for the total number of fish that were usually captured with the gear per day (not only herbivores). The fisher was then asked to estimate the number of fish caught per each herbivorous species being captured (derived from A) per gear and per day. Finally, fishers were asked to assess the actual percentage these herbivores represent from their total catch per gear and per day. Because estimated numbers of preventers and reversers individuals caught (i.e. numbers given by the fishers for each herbivores) in some cases exceeded the total catch per day of the gear, the value of preventers and reversers was mathematically adjusted to fit the total catch (see Box 3).
The interview was made for both the Kusi and the Kaskazi season.
ii) In the second method (for n = 113 questionnaires), the interviewer also asked for the number of individuals in total catch per gear and day. But instead of asking the fisher to estimate the number of fish per herbivorous species that they get, beans were used to estimate the proportion of different herbivorous species caught in different gears (Fig. 10). This was done as follows: each picture of herbivorous species captured per fisher (derived from A) was displayed on the ground, and then a known number of beans was put beside the pictures. This amount of beans was representing the total catch (all the fish captured and not only the herbivores included in this study) for a normal day with a particular gear. The fisher was then asked to split up the beans in one pile representing the proportion of these herbivores from the
total catch and a second pile representing the proportion of the fish that the fisher catches but that are not represented here (Fig. 10). The fisher was then asked to distribute the beans, from the first pile, onto the different herbivorous cards to represent the proportion of the different species being caught. It was carefully explained that one bean did not represent one individual but actually meant a proportion, i.e. relative abundance.
Fig. 10: Three different steps for the method using beans to collect abundance data.
If several gears were used by the fisher, data regarding the presence and absence of the herbivores for each gear were collected. However, information regarding the abundances were collected only for the major gear, from the fishers perspective (see also Box 3 section 3.5 d)) or when data with regards to some specific gears (e.g. gill nets) were scarce then they were chosen by the interviewer. The same procedure was repeated for both Kusi and Kaskazi.
Data derived from the two different methods did not diverge with regards to relationships between gears and the proportion of preventers and reversers caught except for spear fisheries, which I assume is due to few spear fisher interviews analyzed for the first method (n = 3) compared to the second (n = 20). Hence, the data were aggregated and spear fisher data were also included.
3.5 c) Categorization of preventers and reversers
The categorization of herbivores (i.e. preventers and reversers) was based on data derived from Lokrantz (2009). Triangulation of information on feeding behavior was done, using the literature (Green & Belwood 2009; Froese & Pauly, Fishbase 2010; Berkström unpublished data) and scholars’ opinions (Appendix 7). A preventer was categorized as: an herbivore that mainly feeds on filamentous algae. Whereas a reverser was defined as an herbivore feeding on fleshy algae. Once herbivores were classified into these two functional groups, photos for illustrations were taken from www.fishbase.org (Froese & Pauly 2010). This step was critical since the abundance estimations by fishers were based on the recognition of these photos. Even though strong sexual dimorphisms and sex reversal is common among reef fish only photos of males were used due to time constraints.
3.5 d) Data preparation and statistical analysis
The data analysis was divided into three parts. First preparation of data (calculations), second the visualization of data and third the multivariate statistical analysis.
Box 3: The two different methods used to calculate fish catch with different gears
The preparation of data was performed by creating a matrix (using Excel) to enable the coding of the collected data. From this matrix, graphs were drawn up, after recalculating abundances estimated by fishers (see Box 3 in section 3.5 b), to visualize the results.
A multivariate statistical analysis of abundances of preventers and reversers in different gears was performed since this study deals with the analysis of multiple variables simultaneously (gears and functional groups). A one-way analysis of similarities (ANOSIM, software P6 e- primer 6.0) was done to statistically assess the significance of dissimilarities between fishing gears and abundances of both preventers and reversers captured for each gear and per day. The factors were the gears (i.e. large and small traps, gill nets with small and large mesh sizes, spear guns and octopus spearing, seine nets, hand- and long lines) and the functional groups (i.e.
preventer and reverser). One gear for each respondent were included, which means that when the interviewer collected data regarding abundances for multiple gears used by the respondent only one of them was used for the analysis. The choice of an ANOSIM was motivated by the fact that it is a powerful and commonly used statistical tool to compare the variation in species composition and abundance among sampling units (here gears) but also because it is a tool for non-parametric tests (that does not require normal distribution and homogeneity of variance) (Anderson et al. 2001). Following the ANOSIM, a Bray-Curtis similarity test in percentage (SIMPER test, software P6 e-primer 6.0) between groups of gears was performed to assess specific pairwise dissimilarities in percentage (Clarke & Gorley 2006).
3.6 Limitations of the methodology
The hypothetico-deductive method implies auxiliary assumptions (AU) to set in order to verify (or falsify) the hypothesis. AU, such as the selection of fishing gears to study, were based on the literature and personal reflections (also based on experience accumulated during the field work). Hence, if such AU were wrong or incomplete then results from this study might be biased.
Two methods have been used to assess abundances. Even if the largest part of the study was done with the second method (n = 113 of ntotal = 147), this is a limitation since questionnaires were analyzed following a different method. Nevertheless, since relationships between gears did not show any significant differences (except for spear fisheries) comparing the two methods, the data were aggregated.
The herbivore species composition of this study was based on a study by Lokrantz et al. (2010) which does not present an exhaustive species list on herbivores. Corollary, some herbivores that are important for preventing and reversing phase shifts might be missing from this list (e.g.
Bellwood et al. 2006a). Hence, it is difficult to evaluate the full effect of preventers and reversers species in the coral reef ecosystems in Zanzibar. Furthermore, strong sexual dimorphisms as well as sequential hermaphrodism (sex reversal), are common for reef fish (Lieske et al. 2002). To keep the study as simple as possible only photos of males were used which could generate confounded results (e.g. if mostly females in some species are usually captured). The categorization of herbivorous reef fish into preventers and reversers was done using triangulation of information for fish diet collected from the literature (www.fishbase.org;
Bellwood et al. 2006; Lokrantz et al. 2009; Green & Bellwood 2009; Berkström unpublished data). This is a limitation since this phase did not follow a systematic methodology.
Furthermore, only herbivorous reef fish were chosen, hence excluding omnivorous species that could contribute to grazing of algae (e.g. Belwood et al. 2006a).
All gears used in the coral reef fishery in Zanzibar have not been investigated. Excluded gears included for example: cast nets, beach seine, machete and dynamite fishery. There are several reasons for this, but it was mainly due to time constraints and because they are illegal (i.e. few fishers would admit using them). Gears like ring nets, mosquito nets, gill nets with mixed mesh sizes and fixed traps (fixed nets, weirs, fences) have been investigated but the number of questionnaire conducted associated with these gears were too low to include them in the analysis.
Fishers in Zanzibar tend to use the same name for seine-, gill- and fixed nets (known as nyavu) which are officially labeled for seine nets only. When it was possible for the interviewer to see the gear while interviewing the fisher this aspect could be counteracted, when it was not possible, a short description of the gear was being asked to the respondent. This is a limitation since some of the data collected might have been wrongly put under the gear category seine nets.
4. Results
The results will be presented in two parts: the first part compiles annual results, i.e. findings on presence/absence of herbivores and abundances of species in the different gears and the second part presents the monsoonal differences of the catch.
4.1 Average daily catches (based on annual results)
The annual results showed that there was difference in terms of catch of preventers and reversers (Fig. 11 and 12) and only 37% of the herbivorous species included in this study were caught by all the gears investigated. , different gears targeted a different proportion of preventers (hereafter P) and reversers (hereafter R).
According to the fishermen, large traps targeted the largest number of different P and R species per day (based on annual results, Fig. 11). (based on annual results, Fig. 11). Large traps were followed by gill nets with small mesh sizes, small traps and spear guns that capture approximately the same number of P and R species. Seine nets, hand lines, long lines and gill nets with large mesh size target a smaller number of P and R species (Fig. 11). Although, there were overlaps with regards to species being caught by different gears, results also showed that large traps catch more R species than long lines for instance.
Fig. 11: Preventers (P) and reversers (R) captured per gear and per day (n = number of questionnaire). The section
« Do not get » refers to the percentage of herbivorous species the fishers reported that they did not capture.
0%
50%
100%
Large traps (n=48)
Gill nets small mesh (n=6)
Small traps (n=34)
Spearguns (spearguns
and Octopus spearing)
(n=26)
Seine net (n=30)
Handline (n=55)
Longlines (n=20)
Gill nets
large mesh (n=10)
Presence of P and R species (%) caught /gear/day (annual results)
R P Do not Get
Reminder: in all cases there is more P than R being captured because there are nearly two times more P species.
When looking at the total number of individual P and R caught (abundance based on average annual results) per gear and day a different pattern emerged (Fig. 12). There was still a difference between gears in terms of catch composition, also confirmed by the ANOSIM (R = 0,095; p<0.2%). However, the highest number of P individuals was being captured by seine nets, hand- and long lines (equally effective), followed by large traps and to a lesser extent by gill nets with small mesh sizes (Fig. 12a). In contrast, the largest amount of R individuals was captured by hand- and long lines. Two gears caught a moderate number of both P and R individuals; i.e. large traps and gill nets with small mesh sizes. Spear fisheries, gill nets with large mesh sizes and small traps had the lowest catch of both P and R. The low value of the Global R in the ANOSIM (R = 0.095) depicted that there was strong overlapping in the catch of preventers and reversers between gears (Appendix 8 and 9). These overlaps are normal since the catch of only two functional groups is compared between gears. The SIMPER test gave information about gears pairwise dissimilarity between gears (Table. 1).
Table. 1: Average Bray-Curtis dissimilarities (>50%) obtained by the SIMPER test between different groups of gears.
When data on abundances per gear were multiplied by the number of gears present at the four different sites studied (Fig. 12c), two gears clearly stood out from the other gears for their predominant roles in the capture of both P and R individuals. When comparing these gears hand lines caught by far more P and R individuals than large traps. The other gears, that played important roles for the catch of P and R when not multiplying by the number of gears for each
Groups of gears Average dissimilarities (in %)
Hand lines & Gill nets large mesh 72,81 Spear guns & Gill nets large mesh 70,79 Long lines & Gill nets large mesh 70,69 Seine nets & Gill nets large mesh 70,31 Large traps & Gill nets large mesh 68,53 Gill nets large mesh & Gill nets small mesh 66,55 Small traps & Gill nets large mesh 66,54 Gill nets large mesh & Octopus spearing 65,18
Hand lines & Spear guns 54,91
Hand lines & Small traps 53,16
Hand lines & Octopus spearing 52,64
Seine nets & Spear guns 52,28
Small traps & Long lines 51,20
