1. SUSTAINABLE EXPLOITATION AND SENSIBLE MANAGEMENT OF SOUTH AFRICA DEVELOPMENT COMMUNITY (SADC) FRESHWATER FISHERIES

1.1 Introduction

From the hilltop of Kariba town in Zimbabwe, the spectacular view of Lake Kariba is instructive for someone concerned with fisheries management. An overview of the lake’s eastern basin, from the Bumi Hills in the west, towards the Sanyati and Gatchegatche estuaries further east and all the way around to the Kariba dam in the north, reveals that the space allocated to fisheries along this lakeshore is very limited. Most of the shore is allocated for other purposes, such as natural reserves and parks, tourism, urban development and generation of hydroelectric power. A large portion of “state land”, however, is not reserved for any particular purpose. The areas permanently allocated for fishing are divided between the premises of the offshore Kapenta fishery (mainly operated by white entrepreneurs) and limited areas of communal lands where other people are allowed to fish. Approximately 50 inshore fishermen organized in cooperatives (of whom many are foreigners from Malawi, Mozambique and Zambia easily expelled by a simple political decision) fish there. For the Kapenta fishing units, the spatial limitations caused by the land allocation system is a minor problem, since their modern fishing rigs can be moved to all areas of interest in the basin. The inshore fishermen, on the other hand, who are much poorer and operate from non-motorized canoes and small boats, are effectively prevented from harvesting many of the most productive fishing grounds in the basin. Their fishing effort is therefore spatially concentrated in small areas where fish become scarce and profits are consequently low.

Looking at Lake Kariba over time, one notices that the lake acts like a “living organism”. Water levels and lakeshore formations constantly change - seasonally and annually. During the latter half of the 1990s, for example, the maximum lake level increased by 11 m and brought the lake back to the size and shape it had in 1963 when it was filled up for the first time.

This complex and variable situation in both time and space - sensed from the hilltop, but concretely experienced by the fishermen every day - reminds us of issues that tend to be forgotten in the literature and debates on fisheries and fisheries management. Firstly, fish is often only one of many resources in a freshwater ecosystem. The management of fish resources is therefore also affected by how the other resources are exploited. Secondly, the situation described above shows that it is the small-scale fishermen that are most negatively affected by existing resource management measures: compared with the Kapenta operators, the tourist agents, hotel managers and electricity suppliers, the poor farmers and fishermen in the communal areas around Lake Kariba have few chances (if any) of making their voices heard. In other words, fisheries management is (and must be) politics: it regulates and differentiates between people’s and groups’ access to vital resources. Hence the management system in Lake Kariba thus inevitably reflects local as well as national relations of power.

1.2 Our point of departure - some uncomfortable feelings

Over the years, the notions of “management” and “co-management” have come to carry considerable symbolic content. Sometimes these notions are perceived as “metaphors” representing different schools of thought in the governing of the world’s renewable natural resources. Since many of these resources are not privately owned, “management” usually refers to ideas emerging from the classical work generally known as common property theory (CPT), as this evolved on the basis of models developed by M.B. Schaefer (1954), H.S. Gordon (1954) and later G. Hardin (1968). Through the development of concepts like carrying capacity, maximum sustainable yield (MSY), resource rent and maximum economic yield (MEY), we are all familiar with how adherents of CPT argue that, unless an external agent intervenes and regulates access to and exploitation of common resources, the contradiction between commonly owned resources and privately owned means of production will lead to intensified exploitation beyond nature’s carrying capacity. The ultimate result is a tragedy characterized by serious reductions in the resource base (including both biomass and biodiversity) and the depletion of potential economic profits that could have been generated. The emphasis on external resource regulation has also led to an image of the “management” protagonists as particularly strong supporters of the role of the state in resource management, even if CPT as such does not automatically imply this viewpoint.

Unlike the “management” position, which became extremely influential in both academic and policy-oriented debates, the “co-management” approach is not based on one particular theory or model. In many ways it emerged as the result of different arguments contradicting the “management” view. Already in the 1970s - long before anyone had heard about the notion of “co-management” - social anthropologists and human ecologists (e.g. J.M. Acheson, B. McCay and F. Berkes) had started to challenge some of the assumptions underlying CPT. These scholars based their criticism on empirical evidence from various types of production systems and from many parts of the world. They aimed mainly to demonstrate that many production systems contain mechanisms that effectively regulate people’s access to common resources. They argued that, by taking for granted that a common property regime implies free access to the resources, CPT fails to analyse and understand the social institutions that are often crucial to the management of resources in local communities. Knowledge of how these institutions work is, according to these scholars, a prerequisite for the improvement of local resource management systems. The emphasis on local management regimes as “community-based” represented an alternative to the state-centred solutions advocated by the protagonists of CPT.

In the 1980s, management studies were influenced by a new interest in institutional analysis, particularly among political scientists like E. Ostrom. Though these scholars were inspired by anthropological studies of local communities’ rules for the management of common resources, the “new institutionalism” viewed natural resource management from a slightly different angle. Instead of the strong emphasis on local communities in the management of natural resources that many anthropologists had advocated, management became more of a question to be resolved in the interface between the state, the civil society and the market. The term “co-management” thus slowly came to substitute that of “community-based” management, even if the critique of and the disagreements with CPT thinking in certain ways remained the same.

Although we have chosen “Management, co-management, or no management?” as the title of our work, our aim is not to enter into the debate between the “managers” on the one side and the “co-managers/community-based managers” on the other. Instead, our approach emerges as a consequence of some uncomfortable feelings connected with this debate. Through our experience in research and other work related to African fisheries, we have increasingly realized the extent to which the issues and problems that emerge (to us) as the most important ones, are not addressed in the management debate. Sometimes contradictions are “produced” in the debate that do not result in particularly useful discussions and, in other instances, the debate is often incapable of establishing alternative points of view, which in our view are highly needed. Let us outline three reasons that to some degree explain why we do not find the current “management” versus “co-management” controversy particularly fruitful.

1. The “truth” problem. In our view, the debate on the management of common property resources has to a large extent centred on whether CPT as a model is “true” or not. For example, many of the intensive and detailed empirical analyses of how local institutions regulate people’s access to the resources have explicitly aimed to show one of the main arguments of CPT about the need for state intervention to be false. The contributions of the community-based and co-management literature have undoubtedly been important. In a field which 20 years ago still remained completely dominated by natural sciences, they correctly make the point that CPT as a model lacks any reference to social and political content in the relations between the social actors and therefore that CPT fails to capture the fact that communities in ‘real life’ often develop precautions and limitations through the establishment and maintenance of a range of social institutions or “unwritten laws”.

Nevertheless, we agree with those who argue that the debate has been complicated by the general drive to show CPT to be false (Brox 1990). Briefly put, Brox’s argument is that much of the critique is confused because CPT is perceived (by its defenders as well as by those who criticize CPT) as an empirical model, in other words a model that pretends to say something about reality. However, since the model is void of social content it can only be regarded as an analytical model - i.e. a model that claims to have empirical relevance only if the assumptions on which it is based are correct. The question in this work is therefore not whether common property theory is “true” or “false”, but rather whether and to what extent CPT can be seen as a useful tool in explaining the way in which SADC freshwater fisheries are developing. And this leads us to the second problem.

2. The “effort” problem. Since both ‘managers’ and ‘co-managers’ tend to consider CPT as an empirical model, few of them have seemed particularly interested in empirical investigations about the dynamics of fishing effort, or about its causes and its consequences. Like the management supporters, the protagonists of co-management accept CPT’s assumption that growth of effort in a commons is inevitable, unless effective management measures are put in place. Our own experience from research and commissioned studies in many areas in Africa (marine as well as freshwater) has been far more varied. While we felt that certain fisheries in Africa arguably have been growing over a long time, it seemed that others have experienced a much more static and less dramatic development. In the latter case we also had problems identifying any effective management measures in place - initiated by the state or by the local community - and we wanted to investigate in some detail what had been the case in SADC freshwater bodies.

The causes for increased fishing effort in specific areas are seldom investigated, but merely reflect the simplistic economic and demographic arguments underlying CPT. Fisheries are seen as closed systems and so it is not envisaged that people may adapt in unexpected ways to changing circumstances, and that the extent to which they will overfish often depends on whether they have other and better economic options. Approaches such as the ‘last resort’ perspective of D. Pauly do consider fisheries as part of a wider social and economic context. However, whereas growth in effort is seen as a result of artisanal fisheries’ ability to absorb superfluous labour from other sectors, ‘last resort’ thinking tends to assume that once marginalization has forced people into fisheries they are forced to remain there until they have undermined their own livelihoods. Empirical evidence, however, shows that poor people in variable and fragile environments tend to diversify their sources of income and that fisheries is often only one of several income-generating activities, or a temporary option.

Furthermore, it is interesting to notice how the co-management approach does not always incorporate important findings of the new institutional researchers. For example, the economist J.-P. Platteau (1989a and b) has developed an argument on how high ‘transaction costs’ in many developing countries may explain why small-scale fisheries, based on a simple technology, often seem to prevail despite the emergence of new and more capital intensive fisheries. Even though the argument has clear implications for effort development compared to the underlying assumptions of CPT, it has not led to revisions among the supporters of co-management about how they conceive effort growth.

3. The “management belief” problem. In the management/co-management controversy, the focus has often been more on who should manage than on what should be managed. Whether the enforcement of management rules is put in the hands of the state, or of the local community, or one promotes a cooperation between these two levels in the management process, both approaches are based on the assumption that fisheries can always be fruitfully managed. However, from both a natural science and a social science perspective, the truth of this assumption is not evident. New literature in both fields indicates a much more complex situation and reference needs to be made to at least three types of interrelated scientific discourses where such problems are discussed.

The first discourse relates to what has been termed the ‘new ecology’. During the 1990s, a range of detailed empirical studies representing an alternative approach in ecological research emerged. Interestingly, while both conventional management thinking and the common property debate have been richly inspired by research in fisheries, this has not been the case in the new ecology, where biological and social assumptions in conventional management thinking are being re-questioned as an entity. The new ecology discourse has mainly derived from studies of pastoralism (Ellis and Swift, 1988; Scoones, 1995; Sanford, 1983; and Swift, 1988) forestry (Fairhead and Leach, 1994, Cline-Cole, 1994) and other types of land use (Adams, 1992; Scoones and Cousins, 1994; Tiffen et al., 1994). This does not mean that similar positions in biology and in social science are not found within the domain of fisheries research. During the course of this work it was certainly astonishing to come across the Memorandum of Dr C. F. Hickling^[1] (see chapter 3), a fisheries biologist in the Colonial Office in London, who already in 1952 emphasized many of the same viewpoints as one finds with the new ecologists. But in more recent debates on fisheries management, very few works have tried to relate to the arguments. One notable exception is J. Quensière and others’ study of the Niger’s Central delta in Mali (Quensière, 1994).

Representatives of the new ecology question the long-reigning image of considering ecosystems (without human intervention) as closed entities in equilibrium (or in a process towards re-establishing equilibrium). Instead they insist that local ecosystems in most cases must be considered to be in a constant and ever-changing state of non-equilibrium due to considerable climatic variation (rainfall, temperature, wind and evaporation) which are variables defined as external to the system.

The protagonists of the new ecology are concerned with the implications of these views for how one conceives the relationship between human behaviour and the regenerating ability of ecosystems. Management models based upon CPT represent an equilibrium approach and consider human intervention as the only significant external variable. This makes it possible to establish a very close correlation between human intervention (effort) and the generating capacity of the system. For example, Schaefer’s (1954) dome-shaped model of the relation between effort and production and the concept of ‘maximum sustainable yield’ are in fact based upon the assumption that such a correlation exists. However, if the other variables are not considered as only minor ‘disturbances’, but as factors that may alter the dynamics of the ecosystem in a significant and perpetual manner and keep it in a state of non-equilibrium, the picture becomes far more complicated. Then one cannot a priori say how changes in effort will affect the ecosystem, since the effect of effort must be expected to vary according the state of the abiotic variables. An example of this line of reasoning is Ellis and Swift’s (1988) demonstration of how increased effort among Kenyan pastoralists during times of abundant pastures (due to favourable rainfall) is a sound adaptation since the grass that is not utilized must be expected to die when the good rainfalls cease. There is therefore not much rationale in trying to regulate effort under these conditions.

Secondly, comes the important debate about the politics of resource conservation. Even if it may appear as a truism, it is important to remember that management and conservation measures always reflect political choices. As mentioned above, the rules of land tenure along the Zimbabwean shores of Lake Kariba affect various groups of fishermen and users of the lake differently. Studies of colonial and post-colonial conservation policies (Anderson and Grove, 1987; Beinart, 1989; Carruthers, 1993; Hoben, 1995) have demonstrated how these policies were basically rooted in the prevailing images about man and nature, of the colonial administrators: “...the function of Africa as a wilderness in which European man sought to rediscover a lost harmony with nature and the natural environment”; “This perceptual polarisation of a despoiled Europe and a ‘natural’ Africa has held sway since the nineteenth century”; “The problem is rooted in the nature of the colonial relationship itself, which allowed the Europeans to impose their image of Africa upon the reality of the African landscape” (Anderson and Grove, 1987: 4-5).

At the same time the studies show how regulation and conservation policies, also directly reflect the economic and political interests of the colonial state and/or the dominant groups of people influencing its decisions (e.g. white settlers). However, this does not mean that conservation policies were the same as conservation practices. Various local groups soon established patterns of behaviour that became crucial to practice even if they were ignored as policies (McCracken, 1987). Interestingly, it seems as if the post-colonial states to a large extent have prolonged many of the regulations introduced during colonial times, irrespective of changes in policy (Malasha, 2003).

Thirdly, a discourse in anthropology and sociology, with reference to sub-Saharan Africa, questions the concept of ‘institution’ and the distinction between so called ‘modern’ and ‘traditional’ institutions among the new institutionalists (e.g. Berry, 1993, 1997; Peters, 2000). These studies have shown that the ways access to vital resources are obtained and secured, are far more complex than the new institutionalist co-managers tend to think. By conceiving institutions as simple ‘rules of the game’, the new institutionalists tend to emphasize the existence in the local community of existing rules (traditional) which often oppose the formalized rules of the state (modern). The works of, for instance, Berry (1989, 1993), and Peters (2000), show how the new institutionalists fail to grasp the historical dimension in how rules have been constructed and they convincingly argue that local institutions, crucial in the regulation of people’s access to vital resources, often reflect a number of different meanings, are unclear and sometimes even lack coherence. Through detailed historical analyses, their research shows that so-called ‘local management systems’ have often emerged as a result of negotiation and accommodation in long-term power struggles. Various actors including a variety of external powers operating at different periods of time have based their claims and legitimacy on different ‘logics’ and values.

The lack of hegemonic power at the local level has facilitated a continued co-existence of different repertoires of values. When people seek to legitimate their claims to resources, this situation gives them a possibility of referring to a variety of logics which easily leads to the introduction of new ones without replacing those already existing (Berry, 1989). The neo-institutional image of a clear-cut distinction between modern and traditional institutions is challenged by the demonstration that history has entailed a much more complex institutional landscape where both the local as well as government institutions are continuously being constructed in the interface between different actors and where ambiguity and sometimes even contradiction are common. This challenge of the neo-institutional approach is therefore often referred to as a ‘constructivist’ approach.

These discourses have strong implications for how to legitimize, conceive and implement resource management in rural Africa. The new ecology’s emphasis on non-equilibrium in ecological processes and on the role of the abiotic variables for regeneration of stocks, tends to reduce the biological rationale for a great number of detailed regulations which have been introduced into the African commons over more than a century. As shown by Ellis and Swift (1988), it may sometimes be futile to reduce pressure on pastures because the vegetation -if not eaten - will die anyhow. However, this does not mean that the new ecology considers all management measures as biologically futile. The question is merely to distinguish measures based on empirical observations and those based on dubious assumptions. Policy analyses of natural resource management sometimes question both its biological and its social legitimacy. When interests of power become one of the basic principles for regulating resource access, it is not evident that these principles reflect either ecological or social equity concerns.

Finally, the anthropological view on institutional change in Africa renders the conception of ‘how to manage’ or ‘who should manage’ much more complicated. According to this approach, it is not possible - as the neo-institutionalists would do - to distinguish between a set of modern and a set of traditional institutions that are more or less internally coherent. The concept of the ‘traditional management system’ - utilized extensively in the community-based management and co-management literature - becomes problematic because it is more a question of many access-regulating mechanisms that often are based on very different systems of norms and values and therefore may be both ambiguous and contradictory (Fay, 1994; Jul-Larsen and Kassibo, 2001). To implement management in the classical sense of a top-down set of government regulations or co-management in the sense of a co-existence between two sets of institutions (modern and traditional) therefore becomes difficult.

1.3 Research questions and the organization of the report

The preceding section has revealed a number of empirical and theoretical problems that are seldom dealt with in a systematic manner in applied research of this kind. It is therefore the intention of this work to analyse the following questions in some detail:

• How have yields and effort developed in SADC freshwater bodies?
• What have the causes behind this development been?
• How does increased fishing effort affect the regeneration of stocks?
• What are the management implications?

The objective is not so much to provide clear answers to all these questions, as to establish new bases of knowledge for the debates and indicate directions for further investigations. As the subtitle indicates, resource management is always embedded in dilemmas to which there are no simple answers. The style may therefore sometimes appear as more of a discussion than a presentation of clear-cut arguments and conclusions.

After some definitions and conceptual clarifications in the next chapter, Chapter 3 addresses ‘the effort problem’. Instead of assuming that fishing effort is continuously growing, the chapter investigates the effort dynamics and to a minor extent the yield dynamics over the last 50 years, on both an aggregate level as well as on selected water bodies. In order to analyse the causes behind the development in fishing effort, the chapter distinguishes between effort growth caused by increased numbers of people (population-driven) and growth caused by increased financial investments (investment-driven). Furthermore, it investigates the occupational mobility of fishers, both in terms of getting access to the sector, and leaving it. Finally we look at the role government regulation seems to have played in explaining the effort development.

Chapter 4 is a more detailed analysis of the causes behind the changes in effort we have observed and gives us an opportunity to discuss in some detail the relevance of the new institutionalist and the constructivist approach with reference to the ‘management belief problem’. The first section deals with causes behind population-driven changes, while the second investigates the requirements for investment-driven changes. Chapter 5 first discusses the results and value of classical stock-assessments, where effort is the only explanatory variable, in the freshwater systems studied. Then, taking into account the often overlooked aspects of environmentally driven natural variability in fish stocks, it examines the evidence for the long-term effects of increased effort. The impact of fishing on ecology and stock regeneration is addressed step by step from three perspectives: system variability, susceptibility of fish species to fishing and selectivity and scale of operation of fishing patterns. Finally in Chapter 6 we summarize and combine the findings of the previous chapters and try to identify the major implications for management. First we consider it from a biological perspective, before we turn to aspects more based on social considerations.

1.4 Methods and data

The report is based on ten case studies and a literature review. The case studies are to be considered as a part of the report, but they are published in a separate volume. The water bodies included in the case studies are: Lake Mweru-Luapula and Lake Bangweulu in Zambia; Lake Kariba covering both the Zambian and Zimbabwean sides; Lake Chilwa and Lake Malombe in Malawi. Ideally we would have liked to do combined fieldwork covering both biological and social science research on all the water bodies, but for practical and methodological reasons this was not possible. Each of the five case studies focusing on biological issues covers one of the five water bodies, while those focusing on social science issues only cover three lakes; three cover effort development issues in Kariba (Zambian side), Mweru-Luapula and Malombe, one covers the role of trade in Kariba (Zambian side), while the last case study is a comparative study of the history of fisheries regulations in Zambia and Zimbabwe.

The authors have considerable experience in management-related research related to small-scale fisheries development in the region. Kolding and van Zwieten have extended research experience from Kariba and Mweru, while Raakjær Nielsen has worked on South African fisheries. Overå and Jul-Larsen have most of their research experience from West African small-scale fisheries - marine as well as freshwater. In addition, all have short-term experience through missions, research and postgraduate teaching and supervision from a number of other lakes including Victoria, Tanganyika, Malawi, Cahora Bassa, Chilwa and Itezhi-tezhi, as well as the Bangweulu and Okavango swamps, just to mention water bodies in the SADC region.

In social science, historical data have been problematic to obtain. Except for Kariba, few studies have been done in the areas we cover and survey data are scarce and difficult to get hold of. Since the studies on Mweru and on fisheries regulations in Zambia and Zimbabwe are connected to projects for doctoral theses, archive studies have been undertaken in these, but not for the others. Effort data from the last two decades have mainly been collected from frame surveys where these existed and later adjusted according to other effort data compiled by the biologists (see below). The four case studies dealing with local conditions are based upon one or more pieces of fieldwork using various types of interview methods in combination with observation and discussions with assistants in the local areas. The study on trade followed traders between Lake Kariba and Lusaka. A small survey covering 460 fishing households on the Zambian side of Kariba was organized and made a valuable contribution to understanding the occupational mobility among the fishers. A follow-up of an inventory of fishing gear, first undertaken in 1993 and covering all the fishing units in Malombe was also organized.

Data used for the biological case studies are all from existing data collections stored in various formats at the field offices and statistical departments of the respective fisheries departments or research institutes. Biological and fishery data entail both time series of catch and effort data from Catch and Effort Data Recording Systems (CEDRS) and experimental survey data (Table 1.1). The CEDRS of the various lakes are either sampling based, i.e. two stage systems where catch rates are sampled in monthly (Kariba, Malombe, Chilwa), quarterly or tri-annual (Mweru, Bangweulu) surveys and total effort is counted in annual (Malombe, Chilwa) or irregular (Kariba, Mweru, Bangweulu) frame surveys. Other CEDRS’s are based on logbooks (Kapenta fishery in Lake Kariba). Experimental data contain more biological detail but are mainly shorter or longer time series based on gillnet-surveys using experimental fleets of gillnets with a range of different mesh sizes (Kariba - Zimbabwe and Zambia - and Mweru). In both Bangweulu and Mweru, local fishermen were involved in experimental fishery data collections (see Ticheler et al., 1998 for methodology). Fishermen recorded data daily for the pelagic light fishery on Chisense (Microthrissa moeruensis) in Lake Mweru. We refer to the individual case studies for a detailed discussion of the data and the various data collection procedures used in each lake. Lake-level data were obtained from the respective hydrological departments (Kariba, Mweru, Bangweulu, Malombe and Chilwa) supplemented by own measurements (Mweru) and literature (historical data).

Table 1.1 Time series used in the biological case studies and data sources. CEDRS = Catch and Effort Data Recording System. In Kariba logbooks were used in the Kapenta fishery on both sides of the lakes. The logbooks of Lake Mweru refer to the sample of 21 Chisense fishermen who filled in daily logbooks over the period indicated.

A point of concern for many African small-scale fisheries, and equally for the cases in this study, is data reliability. A general tendency exists among researchers and development workers to dismiss catch and effort data as completely unreliable, possibly forged, and in most cases useless for anything but the roughest indications of catch and effort levels. This rejection is too easy. In all cases in this study it is possible to point out where and how data collections become unreliable, and in particular for what purposes they can and cannot be used (see also Chapter 5). Little evidence exists for data forging or false reporting in the cases studied; most errors, at least with the sampling systems, are “administratively induced”, and as the Malombe case shows (Zwieten et al., 2003a) they can be quantified based on the original sample data. We refer to each case study for a more detailed discussion of data reliability.