Access to low priced small fish for livelihood support : a case study on omena (silver cyprinid) trade for local and domestic consumption in Kenya

(1)

Access to low priced small fish for livelihood support

A case study on omena (silver cyprinid) trade for local and domestic consumption

in Kenya

Author: Lola van der Voort, 10539557 Supervisor: Prof. J.M. Bavinck

Second reader: Dr. J. Scholtens University of Amsterdam

(2)

59

Annex 1: List of fieldwork locations 68

Annex 2: Annex 2: List of in-depth interview respondents 69

Annex 3: List of survey respondents 70

(4)

Acronyms

BMU Beach Management Unit DHC Direct Human Consumption

FAO Food and Agricultural Organization FMFO Fishmeal and fish oil

GIS Geographical Information System

LVEMP Lake Victoria Environmental Management Program LVFO Lake Victoria Fisheries Organization

KMFRI Kenyan Marine Fisheries Research Institute SES Social-ecological system

The image on the front-page is a painting made by Aron Vellekoop Leon. It is called “Stuck” and in the context of this research it illustrates the competition between omena traders and the lack of flexibility of the lower scale traders in comparison to larger scale newcomers. “Stuck” was used with permission of the artist and can be retrieved from: https://aronvl.com.

(5)

1. Introduction: Small forage fish as food and means of livelihood

In its recent fisheries and aquaculture report, the Food and Agricultural Organization (FAO) pronounced that small pelagic fresh water fish have the potential to boost Africa’s food security and people’s livelihoods. Small pelagic fish or ‘forage fish’ are small schooling species that feed on plankton and are eaten themselves by larger predators. By converting energy from lower trophic levels into food for larger fish, marine mammals and seabirds, they occupy a vital place in marine food webs (Isaacs, 2016; Cashion et al, 2017). Schools of forage fish form concentrated energy resources and migrate along shores and across oceans followed by other marine species. Besides oceans they also inhabit freshwater habitats. However, like plankton, the quantities of forage fish are variable and highly depend on climate conditions (Tacon & Metian, 2009). Therefore, scientists are concerned about the effects of climate change and marine degradation on these species. Furthermore, overfishing threats the survival of larger marine species as the availability of forage fish declines.

Fish and other aquatic foods also play an important role in human nutrition. According to BeEneE et al. (2015) and Kawarazuka & BeEneE (2011) small forage fish species contain many essential micronutrients, minerals, and are high in protein. They are high in vitamin A, iron and zinc. At present more than two billion people worldwide, in particular in developing countries, are estimated to be deficient in these essential vitamins and minerals. Various authors therefore argue that forage fish could play an important role in battling malnutrition (Tacon & Metian, 2009; Kawarazuka & BeEneE, 2011; Cashion et al., 2017). Traditionally, small pelagic fish have been a key source of food for low-income consumers across the world, especially in coastal regions (Alder, et al. 2008). Besides their nutritional value, small fish offer other advantages for these consumers. Because of their size and schooling behaviour they are generally easily accessible and affordable. They can be purchased in small quantities and they can be more evenly divided among household members. Furthermore they can be processed and stored for a long period of time to meet seasonal shortages (Maljuf et al., 2017, Kawarazuka & BeEneE, 2011). In addition, the trade of fish plays an indirect but essential role in achieving food security through livelihood support (BeEneE et al., 2015; FAO, 2018: 52). Fish and small scale fisheries have important social and economic benefits in developing countries where the sector generates employment and income for many people (Maljuf et al., 2017). Kawarazuka & BeEneE (2011) estimate that 95% of people in developing countries directly or indirectly depend on small scale fishing, aqua-culture or trade for their livelihood. This is especially the case in lower

(6)

people worldwide depend on fish-related activities in sustaining their livelihoods (BeEneE et al. 2015; Allison et al, 2013). This embodies more than 10 % of the world population. Although the revenues for most people are not that high, it is often their main source of income, which allows them to access food (Heck et al. 2007; BeEneE et al. 2011; FAO, 2018: 52). Furthermore, trade connects producers with (distant) markets and consumers and is therefore an essential link in the global food security debate (FAO, 2018: 52).

While there is a widely held perception that captured fish is moving directly to the tables of consumers, 27 % of global marine fisheries landings between 1950 and 2010 (around 20 million tons annually) were in reality destined for other uses (Cashion et al., 2017). An estimated 18 million tons of fish are annually used for reduction into fishmeal (Cashion et al., 2017). This is a commercial product generally used as fodder for farmed fish or poultry/livestock. While it is partly made from the bones and offal leftovers from processed fish used for human consumption, the biggest percentage is manufactured from wild-caught, small forage fish (Jacobs, 2017; Tacon & Metian, 2009). Because the moisture and oil of fish needs to be extracted in order to create fishmeal, the production of fish oil goes hand in hand with the fishmeal production.

Moreover, the increasing global population and expanding middle classes, particularly in Asia, are consuming more and more fish. In our global economy fish is increasingly sourced in the waters of developing countries where labour is cheaper (Crona et al, 2016; Maljuf et al., 2017; Tacon & Metian, 2009; Cashion et al, 2017). As a result, small nutritious fish – either whole or processed as FMFO – have been increasingly transferred from developing countries to western markets. Developed countries imported 80% of total traded fish in 2006 and were also the highest consumers of seafood at an average of 23 kg per person a year (Tacon & Metian, 2009). In contrast, the lowest consumers on a per capita basis were the countries in Sub-Saharan Africa at 6.94 kg a year. Here the probability of death for children under the age of 5 (mostly linked to malnutrition) is at the global highest too, varying on average between 100 to 300 per 1000 children (Tacon & Metian, 2009). From a food-security perspective the growing production of FMFO and small fish export from developing countries therefore raises questions. This research focuses specifically on the small pelagic silver cyprinid. It is a zoo-planktivorous, small sardine-like schooling fish (Latin name: Rastrineobola argentea) that is found all over Lake Victoria. In Swahili it is called Dagaa and in Luo language omena. Because the latter is the term widely used around the Kenyan side of Lake Victoria – the area where the fieldwork of this research took place – it will mostly be referred in this thesis as omena (image 2). The fish is cheap, nutritious and usually sold sun-dried, either for human consumption or for reduction into fishmeal. Its nutritional value is high as the whole fish is being eaten and therefore provides more micro elements and vitamins than fish flesh only. (Legros & Luomba,

(7)

2011). Omena is therefore considered to be an important source of protein for many local low-and medium-income fish consumers around Lake Victoria (Medard, 2015). Moreover, it plays an important role in people’s livelihoods through trade. In Kenya alone, about 850,000 people, especially from lower classes, directly depend on the fishery of omena.And because of its lower market value, it is currently the most dominant fisheries resource sold to end consumers in local fish markets (Kronen et al., 2015).

As omena offers livelihood support to many people in Kenya, this thesis focuses on the position of traders in the domestic supply chain. It examines small-scale fish trade, both in terms of how it links to resource use and consumption, and as a livelihood, by means of a case study of markets in Kenya. It starts by presenting a theoretical framework on the basis of former research and relevant literature in chapter 2. In chapter 3 the research methods, ethics and implication of the study will be outlined. Chapter 4 will provide the background story and a contextual perspective on the research space and chapter 5 offers an overview of the fieldwork results. Chapter 6 summarizes and concludes on the findings and reviews the theories used in this thesis. All maps, figures, tables and pictures in this thesis are the author’s own.

(8)

2. Literature Review

The theoretical basis of this thesis consists of two interwoven approaches: 1) a systemic approach and 2) an actor approach. It is assumed that there are interrelations and dynamics between societal systems and individual actors (Giddens, 1983, Burns, 2006; Folke, 2006; Rounsevell et al., 2012). According to Giddens (1983) they cannot be analysed separately as systems are created, maintained and changed by individual actors, while they are at the same time constrained by it. This study departs from that assumption and examines this nexus in the context of fish trade in Kisumu, Kenya. This chapter starts by offering a systemic approach (2.1) and an actor approach (2.2) on the subject, and explains various relating concepts within these approaches which are relevant to this study. Subsequently, a hypothesis on the interaction between system and agents in the context of this thesis is set out (2.3). In part 2.4 all concepts are combined and explained in a theoretical graphic figure whereupon this is translated into the research questions of this thesis in part 2.5.

2.1 Systemic approach to food supply: Social-ecological systems and small fish

supply chains

This paragraph provides an introduction to the systemic view on the supply of food. I regard the system behind small fish supply chains as a social-ecological system, which we will be explained in paragraph 2.2.1. In paragraph 2.2.2 the supply chains themselves are further examined from a social-ecological perspective. It defines the general concept of supply chains, sets out the specifics of wild seafood chains and the post-harvest phase, and explains how this is related to the main actor group in this study, the fish traders (the topic of paragraph, 2.3).

2.1.1 Social-ecological systems

A system is generally defined as a group of interacting or interrelated entities that form a unified whole (von Bertalanffy, 1972). A social system is a system constituted of a network of relationships that exists between individuals, social groups and institutions (Parsons et al., 2017). An ecosystem is a biological community of living organisms in concurrence with the abiotic components of their physical environment, interacting as a whole (Costanza, 2012) Worldwide there is a growing awareness on the embeddedness of the natural environment in the social, cultural and economic fabric of society, with which according to numerous scholars it forms a set of ‘social-ecological systems’ (Ostrom, 2009; Burns et al, 2002; Berkes et al., 2003; Young et al., 2006; Cote & Nightingale, 2012; Costanza, 2014; Glaser et al., 2008). Humans are an

(9)

integral component of essentially all ecosystems (Redman et al., 2004) and human society plays a fundamental role in designing them, as people worldwide have replaced many natural regulatory processes with those that they have engineered (Costanza, 2014). While unravelling complex societal problems, interactions between social and ecological structures therefore have to be considered and dealt with in an integrative and interdisciplinary way (Ostrom, 2009). Social-ecological system theory provides an approach from which to analyse social-ecological interaction.

Social-ecological systems (SESs) are complex and adaptive systems composed of two primary subsystems: (1) human society and economy on the one hand, and (2) a biological, ecological system on the other (Young et al., 2006; Ostrom, 2009; Costanza, 2014). Social-ecological system theory focuses on the interaction between these continuously changing subsystems. Although there are multiple ways to look at or define those systems, they are generally considered as being: (1) complex, because they consist of many parts interacting non-linearly; (2) adaptive, because the system’s components are responsive and co-evolve as they develop over time; (3) multidimensional, meaning they exists on multiple different levels and scales; and (4) delimited by spatial boundaries surrounding a particular ecosystem and their context problems (Glaser et al., 2008; Ostrom, 2009; Costanza, 2014).The ecological domain of the system consists of the various biological, geological and physical processes of an ecosystem, and the social domain covers the social, cultural, economic and technological institutions within (Ostrom, 2009; Cote & Nightingale, 2012; Costanza, 2014). According to Redman et al. (2004) a particular example of interaction between social and natural systems is natural resource extraction, trade and use. The resource studied in this research is small fish for human consumption, and the domain within social-ecological systems in which this study is therefore specifically interested are ‘food systems’. These systems include all processes, actors and infrastructure involved in feeding a population (Kneafsey et al., 2013; HLPE, 2017). A food system is influenced by environmental, economic, political and social contexts. Furthermore, economies have developed beyond national borders into an increasingly integrated global economy (Young et al., 2006). As food systems expand across international boundaries, it disconnects those for whom food is destined from the stages of its creation, making the social-ecological system more complex.

Many studies have showed the damaging environmental consequences of the global economy as resource extraction is increasingly intensified and industrialised and resources are distributed further distances. Moreover, as food chains expand, they increasingly separate those for whom the food is destined from the stages of its creation (Hendrickson & Heffernan, 2002;

(10)

small-scale ones, have seen the value added to food captured by large businesses”, creating inequality. They are supporting the notion that local food systems are delivering social, environmental and economic benefits. As opposed to global systems, local and domestic ones are believed to improve carbon footprints, strengthen local economies and contribute to the food security at household level, especially in developing countries (ENRD, 2012 in Kneafsey et al., 2013:, p.19). Here local refers to a situation where the area where an economic activity is controlled and takes place is the same locality as where it originates from. In a local food systems the production, processing, trade and consumption of food occurs in a defined reduced geographical area (Kneafsey et al, 2013, p. 28).

According the High Level Panel of Experts (HLPE) report from 2017 on nutrition and food systems, one way to understand a food systems is to simultaneously study the food supply chain, food environment and the consumer behaviour of a resource. This study will follow that understanding and focusses on the local supply chains of small fish for Direct Human Consumption (DHC) within the social-ecological system of Lake Victoria, and the relationship and interaction of the relevant actors with this system.

2.1.2 Domestic seafood supply chains

A supply chain represents a system of people, organizations, activity and resources involved in moving a product from producer to consumer (Burrows et al., 2015; Trienekens, 2011). It encompasses all links between a resource or material and an end product. Besides harvesting, production and consumption, this also entails the processing, fabrication, transportation, storage, trade and consumption of a product. Supply chains are highly variable, depending on the product characteristics and scale of production (Thorpe et al, 2005). Furthermore, they are in most cases part of a broader supply chain network (Harland, 1996). This means multiple chains from a specific product or resource are bundled, but go through different production processes ending up at different types of end users.

According to Burrows et al. (2015) key characteristics of wild seafood supply chains – which are the type of chains this study explores – are their unpredictability, low margins, minimal product tracking, high global demand, and the fact that they centre on is a highly perishable product. Furthermore chain structures can be variable. They can be vertically integrated, meaning they functions under single company ownership, or dispersed. In the latter case, every function is performed by an independent entity (Burrows et al, 2015). In general, as the chain increases in length, the margins get lower, forcing players to do whatever necessary to cut costs. Furthermore, its market access can be open or bottle necked. The second refers to a system where middlemen create a bottleneck for fishers, restricting direct access to the market and creating power asymmetry and relations of dependencies (Burrows et al, 2015).

(11)

In academics, supply chains are analysed in many different ways by different disciplines. This study is specifically looking at the network structure, power dynamics and geography of the resource and people involved in the post-harvest chain of fish. Thorpe et al (2005: p.111) find that: “much supply chain literature takes for granted that, by conceptualising the process as a chain, the interdependence of the constituent links is implicitly recognized”. They emphasise that no participant in the chain is an ‘island’ and that their livelihoods are determined as much or even more by the actions of others in the chain than by themselves. In supply chain literature, some scholars have therefore been advocating for an integrated approach, looking at both horizontal and vertical aspects of chain network structures. The horizontal dimension reflects the relationships between actors in the same chain link; and the vertical dimension reflects the relationships between actors of different chain links (Trienekens, 2011). Another error many supply chain studies contain according to Thorpe et al (2005: p.114), is the underlying assumption that such markets are perfectly competitive. Research has showed that there is big disparity between actors, and especially large retail traders can potentially exercise market power through their fish buying behaviour. When studying supply chains, it is therefore of relevance to investigate who has the power to determine the chain direction and shape and who experiences the burden of chain changes. Nevertheless, little research into market structures and arrangements within the post-harvest fish supply chain has been undertaken to date (Thorpe et al., 2005). This study is interested in the post-harvest chain and especially the trader’s position in it. While a supply chain analysis is used to outline the broader network of trade that actors/traders operate in, the next chapter will evaluate supply chains from an actor point of view.

2.2 Actor approach: Accessibility and adaptability

In order to expose possible consequences of changing systems for vulnerable groups in society, this research additionally approaches literature and data from an actor point of view. In the research the concepts of market accessibility and small fish supply chains will therefore be simultaneously studied from the perspective of local fish traders for DHC. While value chains (which has been defined in 2.2) will be used to outline the broader network, the concepts of access (2.2.1) and adaptive capacity (2.2.2) look at the position of a trader in this network from an individual trader/company point of view.

(12)

fisheries have become significantly capitalised and trade has been liberalised (Crona et al, 2016). A division has thereby arisen in which developing countries became mainly exporting countries of fish while developed countries are directly and indirectly consuming them. This is mainly due to rising demand for seafood in ‘developed’ and ‘transitional’ economies, evolving consumption patterns and the globalization of fish processing through reduced international transport costs. Also the ability of developed countries to meet consumption declined because of depletion and/or restrictive management of fish stocks in those countries (Medard et al, 2016; Crona et al, 2016).

Therefore, Tacon and Metian (2009) claim that: “free market economics and access are currently the main drivers that select whether small pelagic forage fish are fished for feed, domestic consumption or export” (p. 297). This means that the major obstacles for growth of fish for DHC are primarily economic, and relate to the accessibility of global and regional markets for industrial fishing companies and fisher folk. In return, this affects the accessibility of traders selling fish for DHC to the fish supply chain. For this research the concept of ‘supply chain access’ to traders of fish for Direct Human Consumption, will be studied. Additionally, accessibility will be studied in terms of network structures. This means that not only the relationships between traders and fishermen, but also those between ‘trader – processor’, ‘trader – trader’ and ‘trader – policy maker/executor’ will be investigated.

Ribot (1998, p. 310) defines access as “the freedom or ability to make use of”. According to him the term is closely related to ‘property’, which is a “right in the sense of an enforceable claim to some use or benefit”(p. 310). A ‘right’ implies an acknowledged claim society supports. But ‘ability’ is broader than that, resting on the fact of demonstration without any socially articulated approval (Ribot, 1998). Access combines both the de jure articulated systems and extra-legal mechanisms governing resource use. Ribot (1998) finds those extra-legal mechanisms to be: (1) ‘social identity’, which is a person’s status based on gender, age or nobility); 2) ‘social relations’, implying family, lineage friendship and historical ties among individuals or groups; 3) ‘coercion and trickery’, which includes corruption, misinformation, threats of violence, theft and other ways of deceive, 4) ‘material wealth’, which entails material and financial capital; and 5) ‘physical circumstances’, involving someone physical stature and location. Ribot (1998, p. 312) explains that if a person for example has: “capital or a particular status he or she can enter and use a given resource, maybe even against the rules produced by society”. Or when someone has family members and friends in a particular trade, he or she might be able to access the locations extraction, production or trade more easy.

Two related concepts that support the mechanisms of access are: ‘control’ and ‘maintenance’. Control means “the power and the ability to mediate others’ access”, and maintenance means “expanding those powers to keep access open for oneself or others” (Ribot,

(13)

1998, p.311). Therefore, the concept of accessibility enables us to disaggregate bundles of power and the mechanisms supporting them (Abate-Kassa & Peterson, 2011). In order to understand accessibility and place it in the wider context of trade, the concept will be studied along the ‘supply chain’ of small fish for local consumption.

2.2.2 Adaptive capacity

Following supply chain accessibility, a third concept which will be analysed in this study is the ‘adaptive capacity’ of traders. According to Fraser et al. (2005) adaptive capacities are “the social and technical strategies and skills of individuals and groups that are directed towards responding to environmental and socio-economic changes” (p. 467). It is the ability of actors to cope with system- change and dynamics (Milestad et al., 2010). In the context of food systems, adaptive capacity is usually exhibited or deployed to maintain livelihoods, food production, or food access (Fraser et al., 2005).

From the perspective of this study, the concept of adaptive capacity will be used to investigate business arrangements that are made to maintain or enhance traders’ access to the local small fish supply chain. Business arrangements are understood as the complete set of social as well as technical arrangements, made to maintain a sustainable business (Milestad et al., 2010). As this research focuses on traders’ perspectives, it will study the business arrangements along the dimensions of ‘supply’ and ‘demand’. Studying the concepts of ‘accessibility’, ‘small fish supply chains’ and market ‘adaptation’ together, allows for an understanding of who benefits from natural resources, how they benefit and how those patterns of benefit distribution might be changed (Ribot, 1998).

2.3 Hypothesis: Involution

There are numerous ways in which the particularities of social-ecological systems can influence the economy within. However, multiple case studies on population growth in agricultural economies in Asia have shown this can lead to a process of ‘involution’. While this concept has not yet been applied to the context of fisheries, there may be similarities between societies dependent on small scale agriculture and those dependent on fisheries. This research therefore uses the concept of involution, and examines if it is occurring in Lake Victoria’s fishing community, and if so, in what way. Before elaborating on the specifics of Lake Victoria’s fisheries in chapter 4, the concept will first be defined.

(14)

Indonesia. According to Geertz (1963) involution is a self-defeating process in which market changes as well as technological changes take place without actual alteration in living standards and social structure. He used the term as a description of the process in which Indonesian peripheral societies adapted to Dutch colonization and population growth by refining their existing social structures to further support small-scale intense labour. An increasing demand of rice and sugar cane led to an intensification of cultivation. But where in most Western European societies such increases would be accompanied with significant technological or political change, in Indonesia it created larger social complexity. According to Geertz (1963), communities did not change in fundamental ways but redistributed land management and labour into smaller scales and tasks. As such, more people were employed and labour intensity increased. This increased the output per area, while a growth in output per capita did not occur. According to Clifford Geertz, Javanese peasants faced with increased poverty, distributed what they had in an egalitarian fashion as a collective defence mechanism against risks that are associated with trade in a capitalist economy (Tamukamoyo, 2009).

Following Clifford Geertz, the term has been applied to a number of different cases. The historian Philip Huang slightly expanded the theory in his studies on Chinese development. In his book The Peasant Family and Rural Development in the Yangzi Delta, 1350-1988 (1990) he defines three types of agrarian economic change. The first type he calls ‘development’. This is a situation where output increases faster than labour input, which means that the output per worker grows. In the second type, output expands at the same rate as labour so that the output per worker stays the same. This situation he calls ‘intensification’. Finally ‘involution’ is a situation where output expands but at a slower rate than labour. In that case the output per worker declines. Here the concept of involution is about how per-worker or per-capita outcomes could stagnate or decrease despite expansions in absolute output.

In a similar case study on Indonesia, Nathan & Kelkar (1999) found that an increase in population depending on agriculture had two effects. First, in areas where increasing land for cultivation was impossible, the effect was an increase in the number of people working on the available land, without a marginal productivity of the increased labour. This resulted in the sharing of work and income which meant a sharing of poverty. On the other hand, in areas where some extension of agricultural land was possible, people that were thrown out of work went looking for other options in agriculture. This led to increased exploitation of natural resources to sell for fast cash eventually degrading natural ecosystems (Nathan & Kelkar, 1999). Finally, a study on small-scale businesses in Rural Java by Stein Kristiansen (2003), professor in Economics at the Agder University College of Norway, found that involution can also occur in non-agricultural settings. He describes how in rural Java increasing numbers of unemployed people gather into businesses with low entry barriers and in which success has

(15)

been observed. Millions of small-scale cottage businesses in which labour force consist of individuals or family units, operate along traditional and informally organised ways of production and marketing. Their products often suit local markets well and they are therefore concerned with maintaining harmonious relations with neighbours, local customers and traditions. Moreover, traders in these type of businesses are mostly lower educated and have limited access to information on technology and market developments. Therefore, those businesses are characterised by having low margins and a limited capacity to innovate. In the liberalised and increasingly open economy of Indonesia, a large part of those businesses therefore have according to Kristiansen (2003) difficulties to survive. Among those small-scale businesses in rural Java, a combined reaction to increasing population pressures and unemployment started the process of involution. Kristensen (2003): “In the Javanese context, social acceptance and neighborhood status are values that are considered extremely important and traditions often get higher priority than change” (p. 23). Traditional ideas of social responsibility and shared poverty that are part of the Javanese culture, therefore caused an influx of labour with a decrease in the output per person.

In summary, according to Kristiansen’s analysis on involution based on involution literature and a personal study in rural Java, it can be characterised by: 1) “A high labour power absorption ability”; 2) “Close social relations and mutual responsibility in local society”; 3) “Low entry barriers in the form of capital and skills”; 4) “Limited access to capital and credit for investments”; 5) “Limited access to non-local markets”; and 6) “Lack of information on markets and technological development” (Kristiansen, 2003, p. 24). Moreover, as Nathan & Kelkar (1999) described, involution can occur where economic growth is impossible due to ecological boundaries of production within an ecosystem. This study applies the concept of involution in relation to fish trade in local supply chains in Kenya, taking an actor’s point of view

2.4 Theoretical Figure

In figure 1 the connections between all the discussed concepts are visualized in one image. All phenomena take place in a certain social-ecological system. In this system societal factors as well as environmental factors determine the form, appearance and occurrence of a supply chain. A supply chain is typically part of a bundle of different supply chains (a supply chain network), but the type this research is interested in is the chain for domestic consumption. Vice versa, the supply chain, and the economic activities that take place within, affects the social-ecological system.

(16)

from an actor point of view. It is expected that both trader’s accessibility on the one hand and adaptability on the other define this interaction. Based on previous studies, this study expects that the dynamics between the specific social-ecological system and the actors participating in the domestic small fish supply chain cause a process of involution. This hypothesis will be tested in this study. Finally this study assumes that a social-ecological system with accessible resources and adaptive actors play an important role in the provision of domestic food security and livelihood sustenance.

Figure 1: Theoretical figure: Relations between concepts and where to find them in this thesis

2.5 Gap in knowledge and research questions

There is a lack of recognition of the importance of omena and other small pelagic fish species for nutrition, health and livelihoods which has led to inadequate investments in the resource (Kronen et al., 2015). Recently, various studies have pointed out the opportunity for this fish to feed local and regional populations in East and Central Africa. But there is still little known about the supply chain of omena, especially when it comes to trade for local and domestic consumption. As mentioned earlier, little research has been undertaken into market structures

(17)

and arrangements within the post-harvest supply chain of fish in general (Thorpe et al, 2005). In social research, the trader’s perspective on supply chains is therefore understudied, especially in trade for domestic use in the Global South. This research therefore engages in fish trade for local and domestic consumption from a trader’s perspective. The following questions form the basis of this study:

How do local traders selling fish in market places ensure the continuous supply of low-price small fish for direct human consumption?

1. How do natural fluctuations (climate/ecosystem) and harvesting patterns affect the availability of low-price fish in Lake Victoria, Kenya?

2. How is the supply chain of omena for direct human consumption within the social-ecological system of the Kenyan side of Lake Victoria organized from a trader’s perspective?

3. How do local fish traders adapt and enhance their accessibility to the market of low-price omena?

(18)

3. Methodology

This study relies on material gathered during a 2,5 week field study in April/May 2019 in Kisumu and at different locations around the Kenyan shore of Lake Victoria. In total, 10 open interviews and 75 semi-structured surveys were performed by Loes van der Plaat and Lola van der Voort. The two researchers have independently made use of this collective database for the writing of their individual Master thesis. In addition, observations and GPS tracking in combination with Loes van der Plaat, and literature study by Lola van der Voort herself are used to create an integral portrait of the case. In this chapter the research methods of this study will be explained and justified.

3.1 Research design, methods, instruments and supervision

This research is a case study of omena trade in Kisumu county in Kenya. The unit of analysis are Omena traders on markets in Kisumu county and at other trade locations on the Kenyan side of Lake Victoria. During the research, quantitative as well as qualitative methods were combined in a mixed method research design with a qualitative driven approach. This means the research is essentially qualitative, emphasizing the findings gained through in-depth interviews, open survey-questions and observation. In addition, structured survey-questions served to check and clarify the data obtained from the in-depth interviews and to explore unexpected outcomes. This means that more data of greater depth can be used than can typically be gathered by solely quantitative or qualitative research designs (Bryman: 619, 2012).

During the fieldwork the ‘Kenyan Marine and Fisheries Research Institute’ (KMFRI) offered supervision and local support. KMFRI is a national research institute in marine and freshwater fisheries, aquaculture, environmental and ecological studies, and marine research including chemical and physical oceanography. Within the institute, deputy director of freshwater systems, Dr. Christopher Aura Mulanda and scientist at the department of socio-economics, Horace Owiti Onyango, were our main contacts.

3.2 Data sources and sampling

In this research 10 key-actors in Lake Victoria’s fisheries policymaking, implementation and research were interviewed using semi-structured, in-depth interview questions. In addition, the 75 semi-structured surveys were conducted with different types of traders working on 5 different markets in Kisumu county and on 5 different landing sites around the lake.

(19)

Furthermore, staff members of two Beach Management Units (BMU’s) were interviewed in an open and informal way and observations were made in all markets and landing sites visited. Finally, different spatial databases from ArcGIS online, GPS tracking, policy rapports and scientific literature were used in addition to the interviews.

For this research a combination of snowball sampling and purposive sampling was used to select respondents and locations. Due to the limited time in which the fieldwork could take place, KMFRI proposed locations and a set of officials that would be interesting to interview. The characteristics of these officials, including their function, affiliation and the method and date of their interview can be found in Annex 2. In addition we surveyed 75 traders from different markets and landing sites in Kisumu and around Lake Victoria, with an emphasis on people trading on markets in Kisumu county. Respondents were all purposively sampled, meaning that respondents were selectively approached based on the judgement of the researcher and research assistant. We aimed to diversify the sample on the basis of stall size and the position of the stall at the trade location. Although it was attempted to minimize personal influences from the researcher, the respondents might also have been selected on the probability that they would be willing to cooperate based on their body language. A list of survey respondents can be found in Annex 3.

3.3 Data gathering procedure

All open-interviews were conducted by Loes van der Plaat and Lola van der Voort together. In advance they created interview plans for each respondent consisting of 5 to 10 open questions. During the interview, the planned questions were asked in a specific order but throughout the questions the researchers expanded on certain topics the respondent appeared to be familiar with. The interviews took between 45 minutes and 1,5 hours, and all were recorded and transcribed afterwards.

For the survey questions, both researchers were accompanied by a research employee from KMFRI. They offered support as translator, interpreter, guide and facilitator during the fieldwork. In order to reach the fieldwork locations, drivers employed by KMFRI drove them to and from the sites, which were not always easily accessible. Those drivers also offered support in facilitating other practicalities (like purchasing and handing out cold drinks to respondents and showing permits to police officers) and shared with us their local knowledge. All received wages, which were partly payed by KMFRI and partly by the researchers themselves. Finally, Horace Owiti offered support in designing and formulating the survey questions.

(20)

Map 1: Location of Kisumu county and city in the Lake Victoria region and the positioning of map 2, 3 and 5 in this region

(21)

3.4 Study area and research sites

The fieldwork took place on the Kenyan side of Lake Victoria. This coastal region partly consists of open waters and partly of the Nyanza Gulf stretching to the east side towards Kisumu city and county (Map 1). It lies within the equatorial region, and solar radiation and water temperatures are relatively constant throughout the year. There are two main rainy seasons in the area, the long rains occurring from March to May and short rains from November to December (Mucholwa Simiyu et al., 2018). Victoria lake has been topic of many studies on numerous topics from different disciplines since many years, which will be further elaborated in chapter 4. Annex 1 shows a list of the fieldwork locations which were visited in this research area. This were five inland markets selling omena in Kisumu county and six landing sites from which Omena is traded from. KMFRI selected these particular markets because they are the most known for selling omena in Kisumu and they selected landing sites on different locations around the lake in order to identify possible spatial variations. At all locations, except on Wichlum beach, surveys were conducted with both wholesale and retail traders. At all beaches observations were done and notes were taken. Map 3 shows the location of the visited markets in Kisumu county and map 2 shows the location of the visited landing sites around the Kenyan side of the lake. Both maps and their locations in the Lake Victoria region are displayed in map 1. The following subparagraphs give short descriptions of each research location.

(22)

Jubilee market

Jubilee market is one of the largest markets in Kisumu county. It is a market selling all sorts of goods, from food to furniture, textiles, cosmetics and hardware. It is partly roofed and partly in open air and densely occupied with stalls. According to multiple traders there is a problem with space at the market as the space for customers and suppliers to walk and for people to trade is limited. Specific goods are generally clustered together and there are two dried fish sections on the market with dozens of smaller and bigger stalls selling omena. Sizes of stalls generally vary between ½ a meter and 3 meters in length. On Jubilee market 15 traders were interviewed Kiboswa market

Kiboswa market (image 2) is a stretched out market situated outside of Kisumu city with a big zone for fish, especially omena. Stalls are the same sizes as at Jubilee market. There is a small roofed area but a large part of the market is in open air. At this market 9 traders were interviewed.

Chiga market

Chiga market (image 5) is a smaller and very compact market selling a lot of omena. Stalls are generally smaller then those at the other markets. There is also a small roofed area but a large part of the market is in open air. This market appeared to be the most polluted and least organized of all markets as there was much damage to the paths. It was also very muddy and there were many beggars and much litter on the ground. At this market 9 traders were interviewed.

Ahero market

Ahero market was the most clean and organized market we visited. It is situated outside Kisumu city and all stalls are well maintained, exactly the same size and have a roof. There is a lot of shade and there are no vehicles prohibited on the market. It sells all types of food and again dozens of omena stalls are present. Customers at the market appear more wealthy than customers from Kiboswa and Chiga market. Here 10 people were interviewed.

Kibuye market

This is one of the largest open-air markets in all of East Africa. It offers goods from groceries to clothing and services from carpeting to medical treatment. The market is situated in the Kisumu city centre and attracts many customers from all over Kisumu county. The organisation of the market and the stalls appear similar to Jubilee market. There is a lot of omena sold here. We interviewed 9 people at Kibuye market.

(23)

Asat landing site

This is an illegal (unregistered) and small site located in the Nyanza Gulf. Because the bay is shallow, fishermen fish through boat seine fishing which is an illegal fishing method which will be further explained in chapter 5. On the site mainly omena, caradina and juvenile Nile perch is caught and sold. Drying mainly takes place on nets, however we saw omena being dried on the sand. The site appears relatively unclean and unorganized as garbage is spread on the drying area. On Asat beach 8 people were interviewed.

Litare landing site

Litare landing site is a bigger site located outside of the Nyanza bay on a long stretched rock formation. According to the BMU chairman there are around 208 fishing boats targetting omena daily and approximately four fishermen per boat fishing. He thinks there are more traders than fishermen on the site on a daily basis. Landed omena is sold to traders all over the country but mainly goes to markets in Homa bay county. According to the Chairman there is a lack of space for drying on the site. The sites appear the most clean and organised of the sites that were visited for the research and the BMU members were open to be interviewed. Here 3 traders were interviewed.

Sindo landing site

This is a relatively big landing site near the Tanzanian border with a larger village directly bordering it. The site appears more polluted and impoverished then Litare, Wichlum and Marenga Omena beach. There was nobody from the BMU present to be interviewed about the site. At Sindo beach 4 people were interviewed.

Marenga Omena beach

This site is located very close to the border with Uganda and is the biggest site visited. It is divided in multiple beaches landing one type of fish. The landing site is simultaneously a large market selling all kind of goods. According to the BMU secretary it is an intersection of trade between Uganda and Kenya and many boats are owned by Ugandan fishermen. From Uganda they mostly trade fish and charcoal. He estimates that about 700 fishermen are fishing omena and 580 are trading it on a daily basis. Fish sold here is only consumed within Kenya and travels all over the country. Some traders imported fish from Uganda because it is cheaper there. On Marenga beach 4 traders were interviewed.

(24)

Uhanya landing site

Uhanya beach (image 4) is the smallest site visited with a clean landing surface and BMU staff that was very willing to assist us in finding respondents. A small village and market was bordering the site. All the omena was already sold for the day so there were not many people around. There were people landing Nile Perch. On this beach we interviewed 4 traders.

Wichlum landing site

Wichlum beach (image 3) appeared well organized and clean. It does not seem to be a big trading hub like Marenga beach but has a community centre on the terrain where many people were gathered drinking tea. The BMU staff was very helpful and guided us through the area, providing us information on the landing, drying and trading that happens on the beach. We were here during landing at 7 in the morning so we did observe all activities. We did not conduct any surveys here.

Map 2: Locations of visited landing sites around Map 3: Locations of visited markets in Kisumu county,

Lake Victoria, Kenya Kenya

(25)

3.5 Data analysis

In the data analysis phase SPSS was used to organize and analyse the quantitative survey responses. This enabled comparing the survey answers and define the general characteristics of our sample. The open interview questions form the surveys and the qualitative interviews were analysed partly using Atlas.ti and partly manually during transcribing. In both methods quotes were marked using different labels per topic. Spatial data was analysed using ArcMap. GPS data from different fieldwork locations was translated into maps using databases from ArcGIS online and ArcMap.

3.6 Ethics

There is a growing controversy around illegal fishing, fish trade and corruption in and around Lake Victoria. Furthermore, there is much competition in de fishing industry and fisheries communities are known to face poverty and associated social issues. Especially women are marginalized and vulnerable to be exploited. Hence there could be inequalities and power asymmetry among different actors in the fishing network. The people who were interviewed might be affected by these phenomena, since they rely heavily on the resources from the lake. Uncovering their activities, opinions, vulnerabilities and business strategies might make it more easy for people to take advantage of them or their business. And as we were also cooperating with state institutions, their possible engagement with illegal activities might make them vulnerable for prosecution as well. This means that research can have a negative impact on them and their issues, position, stories and answers should be treated with awareness and sensitivity. This was done in multiple ways.

Firstly, we were completely transparent about the research, its topic and purpose. Before every interview we gave an introduction about the researchers, with whom they cooperate, what the research is about, what kind of questions they can expect and why this subject is being studied. An important part of this introduction is the fact that the researchers are students and that the data will mainly be used for them to accomplish their Master-studies. Respondents were also told that the findings will be shared with KMFRI, the Fisheries Service and the county government of Kisumu, and that they might act upon it, but that this was not within our control. Accordingly, people were informed on what they could and could not expect of the outcome of the research and how this might or might not serve them.

Subsequently, people were explicitly asked for their permission to be interviewed to ensure informed consent. When cooperation was refused this decision was treated with respect. When people agreed to cooperate we automatically ensured their anonymity. Their identity was

(26)

decide to stop the interview or refuse to answer certain questions. On the advice of KMFRI, all survey respondents were offered a cold soda during the interview to show our appreciation for their collaboration, and afterwards the researchers local telephone number was handed out when requested. If the respondent would like to have or give more information afterwards, they would be able to reach us. To make sure nobody else could access our data documentation and the personal information of respondents, surveys and notes from interviews and observations were always stored in backpacks that were carried by ourselves. It was kept close to us together with other important personal belongings so we could always keep an eye on it.

Besides containing structured questions, the survey was designed in a way to make space for the own interpretation of the respondent without unnecessarily leading them. This was mainly in the form of open questions, open answer possibilities and space for notes when the respondent wanted to expand on a certain topic. At the same time topics were clearly introduced to guide the respondent to the type of information that was needed. Nevertheless, there was one topic that was chosen to not address in the survey. Jaboya, or the exchange of sex for trade, was left out. In the view of the researchers, this topic needed more trust, sensitivity and time to address, which was not possible due to limitations in time. The topic would have interfered with the integrity of the respondents in this research. Questions relating to Jaboya, were asked to people involved in policymaking, implementation and research instead.

3.7 Limitations

The major limitations of this research were time related, due to the short fieldwork period in Kenya. This is mostly due to misfortunes that occurred in arranging the practicalities for this research. The fieldwork would originally take place in Mwanza, Tanzania. Yet as a result of political and bureaucratic procedures and outdated information that was provided about those procedures, the research could not take place within the time period that was available for writing the thesis. After a month of unsuccessfully arranging the required documents, permits and visa, it was decided to continue the research in Kenya where entry requirements for researchers are less strict and the University of Amsterdam has more connections. However, less time was available for the fieldwork due to strict schedule in the masters and the financial consequences a longer delay would have for the students doing research.

As a result 2,5 weeks of actual fieldwork could take place instead of the planned 6 weeks. In this 2,5 weeks 8 in depth interviews and 75 surveys were conducted. This is not sufficient to give a completely representative picture of the case and the research population. With more time available, we would have liked to do at least the double. Moreover, we would have liked to include ‘illegal traders’ in the research. These are traders, trading outside official

(27)

marketplaces and possibly without traders permits. Also we would have liked to interview agents (people requiring large quantities for bigger traders and/or industries). But both groups probably needed more time to approach and gain trust.

The guidance KMFRI offered, had both advantages and disadvantages for the outcome of our research. It had advantages because KMFRI possess a lot of knowledge on the subject and could point us in the right direction so we could work as efficiently as possible in the short time period. But, it has disadvantages because KMFRI is not an independent organisation and we might have spoken to only one type of respondent that was part of their network, mainly other state institutions. This might have made the sample slightly homogeneous and results slightly one-dimensional because these respondents might have similar views. If there was more time, we would have liked to talk more to BMU’s, local NGO’s and market masters.

Another factor that might have limited our research findings has to do with the Western appearance, experience and outsider role. On the one hand we might have been biased because of our western background and world view. Because we are not familiar with Kenyan life and reality, we might not give the right interpretations to answers given by the respondents. Therefore we decided to collaborate closely with our local assistants and asked for their interpretation. These assistants might again have had their own interpretations of our questions and the answers of respondents. To make sure we were on the same page, we had a lot of contact during the interviews and briefed them extensively beforehand. Moreover, we mainly made and used personal notes of the interviews.

Furthermore, our western appearance and outsider position might have had an effect on the way we were perceived by respondents. Some traders appeared to be sceptical towards us. They might have questioned our intentions as there is a lot of Western interest in the resources from Lake Victoria. Also many people are involved in illegal fishing practices or are part of the informal economy. They might have held back information, answered incorrectly or were excluded from our research as they did not want to join. We also mostly interviewed traders who were trading on official markets and are possibly missing an important and vulnerable group: the group trading on informal locations. Their access to the supply chain is not included in this research.

(28)

4. Context: The social-ecological system of Lake Victoria, Kenya

This chapter provides background information about the research space and creates an ecological, social, economic and political context on the basis of scientific literature from different disciplines in relation to omena trade. It starts with an overview of the ecological changes that Lake Victoria endured and describes its current ecological state (4.1). Subsequently it describes demographic and employment trends in Kenya (4.2.) and omena fishing patterns (4.3). Finally, omena fisheries and fish trade governance will be shortly explained. This chapter therefore is a short description of the social-ecological system of the Kenyan side of Lake Victoria.

4.1 Species invasion, eutrophication and climate change

Lake Victoria, spanning over 68,800 square kilometres is the largest tropical lake in the world. It is the spring of the Nile river and its waters are shared by Kenya (6%), Uganda (43%) and Tanzania (51%), with a drainage basin that extends to Rwanda and Burundi (Mucholwa Simiyu et al., 2018). It is a source of food, employment and water for many people and it supports a valuable artisinal and commercial fishery. In recent decades the lake has undergone a series of profound ecological changes. According to a wide variety of studies, these interlinked changes are largely anthropogenic and the direct and indirect result of human interference in natural habitats (Ogutu-Ohwayo, 1990; Verschuren et al, 2002; Kobingi et al., 2009; Hecky et al, 2010; Kolding et al, 2008). The following paragraph presents an overview of literature on the current environmental state of Lake Victoria in relation to fisheries and how it came into being. In the 1950s, several fish species were introduced in Lake Victoria as an ecological experiment in order to boost the production of larger fish, for food consumption and export (Ogutu-Ohwayo, 1990). Among these fish were different Tilapia species (including Oreochromis leucostictus, Tilapia zillii and Tilapia rendalli) and the Nile Perch (Lates niloticus), a large, voracious predator feeding on fish, crustaceans and insects. Around twenty years later the population of Nile Perch increased significantly, also referred to as the ‘Nile Perch Boom’. Although the outcome of this ‘experiment’ is still disputed - the fish production indeed took an immense shift providing communities with revenue, employment and food – these fish have irreversibly transformed the lakes environment. At least 500 native fish species were either extinct or virtually eliminated from the lake after the 1980 Nile perch population explosion (Verschuren et al, 2002; Ogutu-Ohwayo, 1990; Kolding et al, 2008). These comprised mainly endemic cichlid fishes, which are commonly found in fresh water at relatively shallow depth. Up to now, a lot of attention has been given to the Nile Perch invasion in relation to the

(29)

severe extinction of the lakes native species. The case received much public interest after the award winning documentary Darwin’s Nightmare came out in 2004. The documentary recorded the Nile perch boom and the ecological deterioration on the Tanzanian side of lake Victoria. More information about the documentary can be found at: www.darwinsnightmare.com. The introduction of the fish became one of the most sited examples of the negative effects alien species can have on ecosystems. But studies show other factors might be playing an even larger role when it comes to the wave of extinction in the lake (Verschuren et al, 2002, Seehausen et al., 1997; Kobingi et al., 2009; Hecky et al, 2010).

One major problem in the lake is Eutrophication. This is the increase in the concentration of nutrients – most commonly nitrogen and phosphorus - to an aquatic ecosystem. This is a natural occurring phenomenon but is often dramatically increased by human activities (Jonge et al., 2002; Kolding et al, 2008). According to Jonge et al. (2002) there are three major sources of anthropogenic nutrient input. The first is erosion and fertilizer leaching from agricultural areas and the second factor is industrial waste water and sewage from cities. Thirdly atmospheric deposition of nitrogen, from for example the combustion of fossil fuels or animal breeding is known to contribute to the problem. Nutrients can thus enter ecosystems by surface water as well as ground water and air. Palaeolimnological studies on the Kenyan side of the lake show that nutrient loading begun increasing as early as the 1930s well before the Nile Perch was introduced in the lake (Hecky et al, 2010). This is expected to correlate with ongoing population growth in the lake area and the lake has currently become the most eutrophic of the world’s large lakes.

Eutrophication can have various ecological and social impacts. When the concentration of inorganic nutrients is too high, phytoplankton species grow and multiply fast. Phytoplankton are microscopic photosynthesising organisms are that require sunlight and those nutrients to multiply (Hecky, 2010). Although those planktons are the food source of numerous of other organisms and thus play a vital role in aquatic food webs, an overload of them can cause hypoxia, a common result of eutrophication (Kolding et al, 2008). Increases of phytoplankton are accompanied by the growth of bacteria that decompose them when they die. In this process the bacteria use a lot of oxygen which can lead to a deficiency. A direct effect of hypoxia are fish kills, and changes in fish population can disrupt whole ecosystems (Hecky, 2010; Kobingi et al., 2009; Verschuren et al., 2002; Kolding et al, 2008).

Another effect of eutrophication relates to the breeding of cichlid fish, that used to be abundant in Lake Victoria. Seehausen et al. (1997) describe how the cichlid fish species in Lake Victoria have a very specific sexual selection system and choose mates on the basis of

(30)

many endemic cichlid fishes in the lake to become reproductively isolated (Seehausen et al, 1997). This might have contributed to the catastrophic decline of these native fish species and thus to the degradation of the whole ecosystem.

According to Hecky et al, 2010 climate change is generally less included in the debate around Lake Victoria’s ecosystem degradation and change. There are no large-scale continuous observational records done in the area which makes it difficult to demonstrate a causal relation. Nevertheless, according to Awange et al (2006) Lake Victoria’s water levels have been receding rapidly, and since the lake receives 80 % of its refill through rainfall, climatic contributions cannot be ignored. A meta study from Odada et al, (2009) shows that recent climatic trends for the lake basin have shown 10-40% decreases in rainfall since 1960. A more recent study by Tong et al (2016) showed that the lake levels seemed to be gradually increasing again after 2010 and talk about ‘water level fluctuations’. They expect that besides population growth, fluctuations in rainfall are an important factor. Furthermore climate variability in itself is expected to contribute to eutrophication, decreasing oxygen levels and water transparency (Rijssel et al, 2016). Compared with the 1920s, the surface temperature of Lake Victoria is about 1°C warmer (Njiru et al, 2018). More research on the speed of this warming and the effects of it is required.

4.2 Demographics and employment

Within the Lake Victoria basin, population growth has steadily outpaced continental averages by between 2.5% and 11.3% per decade (Odada et al, 2009). As a result of the wealth of natural resources and the economic benefits the basin offers, the registered population growth within 100 km around the lake is significantly higher than that in the rest of Africa (Odada et al, 2009). Map 4 shows that the population density in the Victoria lake region in 2004 was significantly higher than areas around it. Within Kenya, the human population has doubled over the past 20 years. The country currently counts more than 52,5 million people and is expected to exceed 65 million in 2050 (Kibaru-Mbae, 2018). This is mainly attributed to high birth-rates and improvements in life expectancy. Between 2006 and 2018 the average level increased from 48.9 to 64 years of age. Still the Kenyan population is very young. The median age is currently 19,7 years old (Kibaru-Mbae, 2018) and 80% of its population is below 35 years old (Awiti & Scott, 2016).

Altough the Kenyan economy enjoys a leading position in the East and Central African region, an estimated 16,9% of its population lived below the world poverty line in 2015 (Shifa & Leibbrandt, 2017). According to Shifa and Leibbrandt (2017) and Mohajan (2013) this is mainly caused by social inequality, government corruption, an undiversified economy and health

(31)

problems. A nationwide study from Shifa and Scott (2016) showed that the overall unemployment of people under 35 is around 55%. Especially among rural women the unemployment is high (68%) and the employment rates were lowest among those without a post-secondary education (15%). Although a part of this group might be working in the informal sector as it is estimated that almost 80% of Kenya's total workforce is part of that (Okungu & McIntyre, 2019), these groups enjoy no social protection from the state or other entitlements to certain employment benefits (Okungu & McIntyre, 2019; Vuluku et al., 2013). The International Labour Organization defines the informal sector as “all types of non-formal employment involving workers who are not protected by or bound to national labour laws” (Okungu & McIntyre, 2019, p. 146).

(32)

4.3 Lake Victoria fisheries

As mentioned before, growing industrialization, increasing agricultural pollution and sewage from cities and settlements is now considered to be the main driver of Lake Victoria’s degradation. But besides these damaging depositions, this growing population has also been increasingly extracting the lakes resources (Balirwa et al, 2003; Kolding et al, 2008). Where at the beginning of the 20th century the lake was sparsely fished with a variety of simple traditional methods, at the end of the century the lake supported an intensive fishery. This intensification is considered to relate to the introduction of new fishing technologies (e.g. increasingly destructive fishing gear and outboard motors), diversification of markets (especially the export of Nile Perch), and an increased number of fishers (Balirwa et al., 2003). Simultaneously, the explosion of Nile Perch changed the formerly diverse and relatively small fisheries into one in which three species – two of them introduced – make up almost the whole catch (Balirwa et al., 2003; Kolding et al, 2008). This ‘faunal shift’ happened in the mid-1980s and Nile Perch, Nile tilapia and omena have been exploited since.

By the early 1990s, the total yield had increased by a factor six compared to ten years earlier and the fisheries production rose from around 100.000 tons per year in the 1980s to about 1 million in 2008 (Kolding et al, 2008; Mkumbo & Marshall, 2015). The Nile Perch fishery formed the basis of a large export industry and around 75% of the reported total landings of Nile Perch in Kenya were exported (Kolding et al, 2008). However, the Nile Perch stock has recently been showing signs of overfishing (Mkumbo & Marshall, 2015; Yongo et al., 2018). Where in the 1980s the catch made up 66% of total fish catches, in 2008 this was only 28%. Besides it having serious consequences for those whose livelihood started to depend on Nile Perch trade (Mkumbo & Marshall, 2015; Yongo et al., 2018), it had allowed prey species like omena, to increase. The small fish gained popularity around 1990 when other indigenous species became extinct due to the Nile Perch invasion (Isaacs, 2016). Since then local demand increased, and after the establishment of the East African Community (EAC) and the Common External Tariffs (CET) early 2005, so did regional demand (Medard 2015). Now this fish, which is known as Dagaa in Swahili and Omena in Luo, is the dominant species fished on and traded from Lake Victoria (Isaacs, 2016; Odongkara et al., 2018). It makes up 55% of the total catches (Odongkara et al., 2018).

4.4 Fisheries governance in Kenya

Since 2010 the national system of governance in Kenya changed to a devolved system of governance and 47 counties now hold certain individual political, administrative and fiscal

Access to low priced small fish for livelihood support : a case study on omena (silver cyprinid) trade for local and domestic consumption in Kenya