As we have shown in this report, the socio-political and ecological dynamics of the societies and lakes studied are quite complex. This complexity is not fully captured either by the assumptions underlying model-based management approaches exemplified by Common Property Theory, or by the neo-institutional framework underlying much of the co-management approaches. It is also becoming increasingly evident that the management measures implemented in the southern African freshwater fisheries are not adequately based on empirical knowledge.
In this chapter we will first summarize our main findings. Then we will discuss the implications of these findings for the development of a management approach in small- and medium-sized freshwater fisheries that takes both the natural and social complexity of the systems into account, at the same time as attempting to ensure equitable and sustainable utilization of the resources.
Even though great differences can be observed between ecosystems and countries, a number of lessons can be learned from our studies that have important implications for the way management and/or co-management is conceived and implemented in many African freshwater fisheries. The following are the main empirical findings from our studies of small- and medium-sized lakes in Zambia, Zimbabwe and Malawi:
· Growth in fishing effort in this region has mainly been population-driven.
· Environmental drivers, especially changes in water level, are most significant for variation in the productivity of fish stocks, and population-driven growth in fishing effort has not been too harmful for the productivity of the fish stocks in these lakes.
· There is considerable mobility of people in and out of the fishery sector.
· Population-driven growth in effort tends to increase when macro-economic conditions deteriorate and/or the biological productivity in a lake improves. Reductions in population-driven growth in effort take place when local access-regulating mechanisms exclude newcomers, and when investment-driven growth in effort results in rising entry costs for those who wish to join the fishery.
· Investment-driven growth rarely takes place in these fisheries: only two cases have been observed. When such processes occur, the investors are entrepreneurs from outside the fishing communities, or the investments are made possible by input of capital generated in sectors other than fisheries.
· Where investment-driven growth takes place it has led to the collapse of specific fish stocks.
· The main constraints on investment-driven growth of fishing effort in the southern African countries are that the local institutional landscape impedes the development of infrastructure and access to capital and markets.
· Peoples flexible adaptation to the ecological and economic environment through frequent entries into and exits out of the fishery sector, facilitates the function of SADC freshwater fisheries as a safety valve or, in other words, as a buffer against poverty.
How to manage fisheries effectively has been intensively debated for several decades. For quite some time the classical approach to fisheries management has leaned heavily on the prediction and estimation of optimal yield levels such as maximum sustainable yield (MSY), to which effort levels should be adjusted. Although models pertaining to this approach were initially developed for the management of large-scale fisheries on single stocks in relatively stable environments in temperate climates, it has become the dominating approach worldwide. However, setting optimal catch and effort levels on a species-by-species basis is highly problematic anywhere in tropical multispecies and multi-gear situations, and the information needs for such an approach in the small-scale SADC freshwater fisheries discussed here are prohibitively costly. Overall catches and generalized levels of effort (numbers of fishermen and gears) can be known, but the highly diverse and dynamic fishing patterns make the establishment of causal relations between catches and a specific level of effort virtually impossible. In many ways the approaches must be considered harmful to the development of management systems. Management proposals based on single-species model predictions give a strong incentive towards adopting rather drastic measures on effort control and gear regulations in situations where fishing may represent an important means of survival for many people. Moreover, the MSY approach has severe limitations in systems with large environmental variability, and for such systems it is extremely difficult to causally relate trends in fish stocks levels to fishing alone. Such trends are often unclear and hidden because of environmental variation acting over different timescales causing what is called in statistical terms coloured noise (Densen, 2001).
In fact, we have shown how environmental factors are often more significant than fishing effort in the explanation of changes in fish production in the lakes studied. In Chapter five it is argued that water levels are a key (environmental) driver explaining the productivity of fish stocks. Productivity in many of the smaller freshwater lakes is extremely high, and for various reasons many stocks are highly resilient to increased pressure and have the capacity to bounce back if the pressure is released. Lake Chilwa is an extreme example as it has dried up completely six times within the last 150 years, but as soon as the lake has filled up again, fish stocks have recovered. Another important observation in numerous multispecies fisheries, including some of the systems discussed here, is that levels of total yields appear to be surprisingly stable over a large range of fishing effort. Underneath this relative stability of total yields, changes in species and size composition occur, both as a result of fishing and as a result of environmentally driven processes. Lastly, fluctuations in the number of fishers are concluded to be to some extent a reflection of variations in the productivity of the ecosystems. Lake Chilwa again is the clearest example of how the size of the fishing population fluctuates with such changes in productivity according to the lake levels. But also in other cases, one can infer that changes in productivity and catch rates to a certain extent regulate the level of effort.
The prevailing biological approach, based on setting effort levels through stock assessment models, is therefore not a solution if the objective of fisheries management is to secure a livelihood for the population. From a biological point of view, the management of fish stocks has to be pragmatic and adaptive. In such an approach, the first step would be to undertake an analysis of trends and variability in catches, catch rates, fishing effort and water levels. This will provide knowledge on the particular behaviour of a system under various conditions. The second step would be to establish a set of important biological indicators that can form the basis for the provision of knowledge that can be used in the decision-making process by the various agents involved in management of these fisheries. A generalized approach towards devising such indicators has to be based on information on:
System variability. Freshwater lakes and rivers can be classified over a range from pulsed to constant environments. In Chapter 5 we have shown that the Relative Lake Level Fluctuation, which is based on the seasonal and annual amplitude in water levels related to the depth of the lake, is a simple index that may help in such a classification. For any particular system, changes in water levels can provide a number of other indices that can be related to changes in stocks. For instance, persistence of conditions could relate to short-term, decadal trends in stocks and species composition, while peaks in lake levels could be related to strong year classes of longer-lived species. Lastly, the size of inundated areas at different lake levels could provide another index and be related to variations in catch rates and species composition.
Susceptibility of species to fishing. Applying this concept will indicate that management becomes less relevant from a biological perspective the more resilient a species is to increases in fishing pressure. Resilient species, such as tilapias - dominate in many African freshwater fisheries - and more especially highly-resilient species, such as small-sized species like freshwater herrings and small barbs. Indicators of change relate to type and size of species in the catch and lag phases between changes in catch rates and water levels.
Selectivity and scale of operation of fishing patterns. There is a limited danger in increased diversification of fishing patterns with small operational scales, i.e. when fishermen use methods that catch the ubiquitous bucket of fish per day. Such fisheries come very close to an overall unselective and ecologically sound fishing pattern, highly adaptive to changing conditions. The danger increases with increased gear efficiency, whether this arises from investments in better technology or from more intensive use of existing technology. Examples of such indicators are increase in numbers of gear per unit, increased sizes of individual gears and higher investment levels in gear and vessels.
These classifications and associated indicators are highly relevant in designing management systems for SADC freshwater fisheries. Since the impact of fishing activities on the productivity of the stock has proven to be limited, it is difficult from a biological perspective, to see the need for management in pulsed environments or in fisheries targeting resilient or highly-resilient species with a multitude of small-scale individually selective, but overall unselective gears. As long as the increase in fishing effort is only driven by population growth, fish resources in the four lakes (with few but important exceptions) are unlikely to be overexploited. The fisheries will in other words, remain largely biologically sustainable both with and without detailed management measures based on stock assessment models. A truly multispecies/multi-gear fishery cannot be managed on the basis of single-species models alone, though important species for which much information can be obtained could act as important indicators of change, if their changes in biomass or length structure reflect changes pertaining to the whole fishery. Besides this, a multispecies/multi-gear fishery can only be managed on the basis of the aggregations of species in the catch. This means that instead of focusing on individual species, management should be directed to maintaining the integrity of fish communities and ecosystems, including the source of their variation.
For more than a hundred years, fisheries management in SADC freshwater lakes has primarily been about establishing operational rules for the fisheries. This subject has been dealt with in detail in Chapter 3. Here we would only repeat that already in 1952, C.F. Hickling, a fisheries biologist in the Colonial Office in London, considered the effectiveness of most of the restrictions and prohibitions that had been in place in the colonies for almost 90 years as doubtful. He pointed out that licensing of gear or nets requires a large and expensive enforcement staff. Other measures, such as closed seasons, mesh-size regulation and size of fish regulations were also questioned as general management instruments. Nevertheless, despite his reservations towards most of the applied regulations, they have continued to form the basis for management of freshwater fisheries in sub-Saharan Africa, even though variations between countries exist. Our findings generally support Hicklings critique, but, although the relevance of these measures may be debated, they might be both necessary and useful in specific situations in a more adaptive management context. From a biological point of view, what to manage in systems that are adapted to high environmental variability can be summarized in a small number of general rules on an ecosystem level. Specific measures either derive from these rules or relate to specific objectives in order to:
Maintain the integrity of the ecosystems including their natural, environmentally driven, variation;
Maintain variability in habitats and avoid habitat destruction;
Protect important refuges in periods of adverse conditions, for instance through temporary closures.
There is no need for detailed gear regulations, as long as there is no major increase in efficiency of particular gears. Specific management measures can, nevertheless, be relevant for specific reasons.
Mesh size and other gear regulations. As shown in Figure 5.5, the regulation of gears, including mesh sizes can be interventions that have highly different consequences for different trophic levels within an ecosystem. For instance, employing smaller mesh sizes could increase catch rates over the short term, but at the same time the fishery will become more vulnerable to environmental changes, because the variation in CpUE is larger for smaller-sized species. Large mesh sizes could be needed to protect larger species like the Tigerfish (Hydrocynus vittatus) in Lake Kariba, but they are also measures to avoid the fishing of undersized specimens of larger species (i.e. the fish for the future). Choice of mesh size thus depends on the purpose of the fishery. Nevertheless, from a management perspective it is difficult (indeed almost impossible) to redirect the fishery back again towards larger fish as it requires very restrictive management measures (e.g. a major increase in the mesh size employed or closing the fishery for a period of time). Measuring its success will take time, at least in the order of a few years to a decade (Pet, 1995; Pet et al., 1996; Densen, 2001). Because of the difficulty in showing the biological impact of applying more restrictive management measures, such measures will most likely undermine the support of fishers where they are imposed.
Closed or protected areas are management measures that can be applied to protect nursery grounds, spawning aggregations or avoid habitat destruction in certain areas. In relation to Lake Malombe, some biologists believe that habitat destruction - in this case the disappearance of submerged weed beds, due either to reduced water levels or to seining - explain why the Chambo (Oreochromis spp.) stocks have collapsed. In the case of Lake Malombe, protected areas could therefore perhaps be effective in the recovery of this species by protecting spawning grounds. In Lake Chilwa, it is important to protect pools with fish remaining when the lake has dried up. On the Zimbabwean side of Lake Kariba, closed areas are used to protect spawning areas and nursery grounds, despite the fact that there is no strong biological evidence to support this management measure (Jul-Larsen et al., 1997). In some cases, these management measures are introduced to pursue political objectives by excluding fishers from certain areas in order to protect the interests of more powerful actors (e.g. tourist operators). For instance, in Lake Kariba closed areas have been used to maintain social control over the fishers and actually been used as an instrument to move fishers from certain areas (Malasha, forthcoming). On the other hand, the establishment of a closed area in Lake Mweru was aimed to protect the stocks, although little evidence exists for its effectiveness. But even if these examples may indicate something else, closed areas or no-take areas can, as already mentioned, be useful both in an adaptive management context during periods of adverse conditions, or as a general precautionary measure.
The implementation of closed seasons is a general measure to protect fish during breeding periods. This instrument has been used in Lake Malombe, but has been difficult to enforce. The major effect of a three-month closure in Lake Mweru has been that trade of fish more or less comes to a halt, which reduces general effort levels considerably. Although being heavily contested, the measure is generally considered acceptable by fishermen because they are still able to fish for subsistence and because they then direct their labour towards agriculture (Gordon, 2000; Aarnink, 1997).
Licences are promoted based on a biological rationale. However, in most cases governments enforce licensing mainly to obtain political and social objectives. Hickling also argued that the introduction of licences was mainly about supporting the Native Authority. In line with legislation supporting traditional leaders, e.g. the Administration Act and the Indirect Rule Act, fishing licences became a means to generate revenue for this Authority. More than to control fishing effort, licensing remains an important tool for many local chiefs or local authorities to obtain their share of the income generated from fishing activities (Malasha, forthcoming). The biological rationale for a licensing system in small- and medium-sized freshwater lakes in SADC is thus hardly convincing.
Malashas research demonstrates that management of fisheries is not so much about managing fish stocks as about managing or controlling fishers and other users of the resource. According to our findings, freshwater fisheries in the SADC region will remain biologically sustainable without any management intervention initiated by governments under the present macro-economic situation and local institutional conditions. The fisheries in small- and medium-sized freshwater lakes in southern Africa can in many ways be considered to be self-managed from a biological perspective. This does not mean that there is no need for management. From a precautionary viewpoint - which we firmly support - there is a need to monitor the development of catches, catch rates, fishing effort, water levels and the maintenance of the integrity of habitats. The challenge will be to use existing data more effectively and to set up effective data collection and monitoring systems based upon longer time series. Biological management may also be required if one wants to promote particular types of fisheries (e.g. angling for tourism). In any case, management will have to be something different from the traditional approach taken in the SADC region so far.
As shown in Chapter 3, the aggregate fishing effort in SADC freshwaters has increased considerably in the last 50 years and conventional models relate the increase to general social factors like demographic growth and increased demand for fish. However, the lakes under study demonstrate great internal differences in their effort development. For example in Lake Mweru, the number of fishers has steadily increased during the last four decades. In Lake Kariba, the number of inshore fishers has fluctuated a lot and effort is probably not much higher today than it was when the lake was filled up. The number of fishermen increased in Lake Malombe during the 1970s, but already in the 1980s, it started to stabilize and lately it has decreased (Hara, 2000; Donda, 2000). In Lake Chilwa, major fluctuations in the number of fishers have been observed, mainly due to the major water recessions during which the lake dries up completely. During major recessions, fishing operations have been suspended for two or three years.
Only when we break down the concept of fishing effort, is it possible to discover regularized patterns in the development of these fisheries. It proves that most changes in fishing effort are population-driven. Most of the fisheries seem able to absorb great numbers of newcomers who mainly continue to fish in the same manner and with the same gear as their predecessors. Malombe is the only lake where a major investment-driven growth in effort can be observed. The expansion of the Mpumbu fishery in Lake Mweru in the late 1940s is another example of investment-driven growth, but this fishery stopped in the 1950s due to the collapse of Mpumbu. In all other cases, it is population-driven changes we notice and the fisheries are characterized by extensive rates of both in- and out-migration. New people seem not only to enter, but also to go back to other occupations when they consider it worthwhile. Fisheries in southern African freshwaters therefore do not function as a last resort in the sense that once someone has entered there is no way to return to other occupations. It remains a fact, however, that far more people have entered the fisheries over the past 50 years than have left them.
Demographic growth and/or better markets for fish can explain neither the variations between the lakes, nor the internal fluctuations in fishing effort. Even if we accept that these factors play a role, we need to look for additional factors to explain effort development in SADC freshwaters. Fluctuations in the number of fishers are to some extent a reflection of the variations in the productivity of the ecosystems. Lake Chilwa is the clearest example of how the size of the fishing population fluctuates according to lake levels, but also on Kariba we saw how effort fell when the productivity decreased. The fluctuations also reflect changes in the conditions and job opportunities in other sectors of the economy. Improved conditions and better job opportunities elsewhere will immediately result in reduced fishing effort and vice versa. In order to understand the dynamics of freshwater fisheries in the SADC region, they must be placed in a much larger macro-economic context. Only when we see an improvement in the general macro-economic conditions will we be fully able to judge the relative importance of the role of external opportunities.
Furthermore, the local access-regulating mechanisms, which exist everywhere, prove to be important. As argued by Brox (1990), rather than being a management means as such, local access-regulating mechanisms more often reflect how local class and power relations mediate and restrict peoples access to resources. In many cases, such mechanisms are not made relevant or they are not effective, but in the cases of Malombe and Kariba we saw how periods of stabilization and reduction can only be understood with reference to such mechanisms.
Finally, we have observed that the use of capital-intensive fishing methods sometimes prevents population-driven increase in effort: the cost of equipment becomes so high that only a few fishers can afford it. This observation is supported by experiences from other parts of the world, such as in the pelagic fisheries in the North Sea employing Dutch, Danish, Scottish and Norwegian vessels, or the off-shore cod fishery out of Iceland or Norway using freezer trawlers, where a reduction in nominal effort (number of fishers or boats) often follows a major change in the technology applied or in the organization within the fishery. The Lords of Malombe (Hara and Jul-Larsen, 2003) could remain Lords because so few could cope with the investment costs and just as in the tragedy of the Norwegian Herring fishery (Brox, 1990), the tragedy in Malombe was not caused by an unlimited access of the many, but by the competition between the selected few who remained in the race.
Investment-driven growth is not as inevitable as projected by Gordon (1954) and Hardin (1968). With the prevailing macro-economic conditions in the region, with very little growth and limited funds for investment, it is unlikely that freshwater fisheries will attract investment from the outside, at least in the short- and medium-term perspective. Furthermore, reduced possibilities in labour migration and a complicated, unclear and ambiguous institutional landscape at the local level seem to effectively prevent any investment-driven growth from within the fishing communities. Probably the only case at present where investment-driven growth dominates development is on Lake Victoria.
The geographical and occupational mobility characteristic of fisheries in southern African freshwaters, and the flexibility of the ecosystems is a key to understanding their importance for the countries, as well as for the people involved. They serve as buffers in the national and regional economy and they serve as an important economic safety mechanism for thousands of people. Few cases we know about better illustrate Broxs (1990) argument that the problem with common property theory is not what it says, but what it omits to say. There is no doubt that the commons can lead to ecological tragedies; we have seen it in the case of the Mpumbu fishery in Mweru and the Chambo fishery in Malombe, but the theory omits to tell us that the commons also may play a very positive role, both for the individuals as well as for society. We will argue that it is important to maintain the southern African freshwater fisheries as commons - as valuable local and even national assets through their function as buffers and safety valves - in particular if poverty reduction is taken into consideration. Instead of enforcing severe restrictions on resource use, the freshwater fisheries should continue to be used as a safety valve for local people as well as migrants - facilitating their adaptation to changes in the macro-economic environment and in their occupational surroundings. In order to do so, it is important that management can control, or more correctly prevent, large investment in the fisheries. If investment-driven growth is allowed to take place, the consequences may become serious. In such circumstances, well considered and timely interventions are required from decision-makers in order to avoid building-up fishing power.
Here it is important to distinguish between investments by people from outside the fishing communities and investments made by people from inside. Our research indicates that a combination of government prohibitions and local access-regulating activities are quite effective in limiting massive investment-driven growth from the outside. In such cases, governments may intervene and thereby protect local interests. This was in fact the situation in Lake Mweru some years ago, where trawling was forbidden in order to protect local fisheries and food security.
Paradoxically, government intervention is often more problematic if investments come from inside the communities; neither government nor the local communities seem particularly able to cope with such a situation. It is extremely difficult for the local institutions to prevent insiders from investing in the fisheries, because insiders who are in a position to do so are usually powerful people and therefore in a strong position in the decision-making processes. Government intervention to avoid an investment-driven increase of effort is thus likely to be unsuccessful because of lack of local support. However, we see that investments from within the community often require financial resources from labour migration or other external sources, as well as an institutional landscape very different from what is generally found in rural southern Africa. The case of Malombe also reminds us that investors tend to diversify their activities. As a consequence, they consider themselves as much as farmers and/or businessmen than as fishermen. As long as the macro-economic conditions do not improve, we do not consider investment from inside the fishing communities to be a major management problem. In a longer term perspective, or where investment-driven growth already dominates the development, cooperation between government and the fishing communities appears to be the only way to improve the situation.
From a social and economic perspective, it is important to emphasize that population-driven growth in effort also reduces catch rates, even if by far less than investment-driven increases. Reduced catch per unit of effort (CpUE) certainly has negative economic consequences for the individual fishers. There is therefore a balance to strike between the need for individual incomes on the one hand, and the need for a buffer in the economy and a safety valve on the other. We are not certain, however, that government can strike such a balance in an effective way, but there is a good chance that the local access-regulating mechanisms in most cases can deal with it. The higher the population pressure becomes, the more we must assume that local access-regulating mechanisms are put into function. We saw in Box. 4.1 how they did in Kariba when the pressure on the land resources along the shore became too high.
Given that such mechanisms are deeply embedded in the local institutional landscape, there are few possibilities for government to play an effective role. But because population-driven growth in effort leads to an intensification in the level of conflicts, government has an important role to play in conflict mediation and resolution. As long as the conflicts are between groups that are economically or politically equal, disputes will probably best be resolved locally. However, in situations where conflicts are less embedded in local conditions, or where the lines of conflict follow clear differences in economic or political status and become more permanent (e.g. as in the Mweru fishery on the Congolese side, where the great trader Katebe Katoto had a lot of control), there is - ideally speaking - also a need for active support by governments of the weakest party in the conflicts. This is also the case if conflicts are between actors making alternative use of freshwater resources (e.g. fishers vs. tourist operators in Lake Kariba).
This type of social and/or economic management requires sets of information and data that at present are virtually non-existent. It is interesting that despite the change in rhetoric about the need for economic and social considerations in fisheries management, this has had very little impact on how the authorities organize their fisheries services, how the service is staffed or what sort of data they collect. To manage the fisheries in the manner just outlined will require the elaboration of some economic (e.g. cost-earning and investment) and social indicators (e.g. dependency of fishing and migration) and their integration in the data collection and analysis systems.
We started out in this report by referring to some uncomfortable feelings that we had related to the debate between management and co-management approaches. Before we end, we would like to explicitly address the question: To what extent can the management and the co-management approaches be useful for the governance of southern African freshwater fisheries? What kind of role should governments have vis-à-vis local communities, or Should the fisheries be left to manage themselves? Throughout the report, the findings demonstrate that the answers to these questions are somewhat more complicated than what the first reaction may indicate. Even if the fisheries under study in many ways can be seen as self-regulatory with little need for external intervention, we have also shown that there is a need to monitor their development, and that management measures soon may be needed if major investment-driven growth in effort were to take place. Changes in existing investment levels and patterns may change the situation dramatically in a limited period of time and lead to serious bio-ecological problems. This shows that the management approach and common property theory on which it relies must still be considered relevant. Even limited investment-driven changes may easily lead to tragedy scenarios as we have already shown in some limited cases. However, despite the lack of effective resource management regimes, investment-driven growth in fishing effort has not been characteristic of the fisheries in southern African freshwaters. As has been the situation over the past 50 years, nothing indicates that a no management solution on the part of the governments will lead to biological tragedies in these commons. On the contrary, too much government intervention may result in the fisheries losing some of their important social and economic functions as buffers and safety valves for a great number of people with limited possibilities in periods of stress. In that respect, too much reliance on CPT is of little use and may prevent us from capturing the present strengths of the freshwater commons.
The centralized scientific management approach often taken by the various SADC governments has been inappropriate for a number of reasons. But in order to understand why they have chosen such strategies, it is impossible not to emphasize the role of colonialism. Many of the fisheries regulations can in practice still be traced back to the colonial administrations from the first part of last century, (Hara, 2000; Malasha, forthcoming). For Zambia and Zimbabwe, Malasha concludes that: despite the numerous amendments attempted by the two countries, the principal objectives of fisheries management have remained the same since the advent of colonialism about a hundred years ago. After political independence, the centralized approach was maintained for at least two reasons. First, it was an approach that had been applied by the industrialized countries for a long time. The people in the new administrations had all received their training and professional experience from colonial institutions and staff who were convinced that their way of conceptualizing the relationship between man and nature was the only natural one. Secondly, the new administrations had also learned another lesson from those previously in control: under the pretext of concern for natural resources, the centralized management approach had proven to be a well-suited policy to control people in rural areas. The different political regimes that succeeded colonial rule, whether they called themselves socialist as in Zambia and Zimbabwe, or autocratic, which was the situation in Malawi for many years, therefore had good reasons not to give up this policy. The history of one hundred years of resource management in the region reveals that government, and its individual representatives, often act on the basis of their own interests rather than from a consideration of what type of regulation is appropriate.
Our report also emphasizes how increased population-driven growth in effort has already increased the level of conflicts between different groups of users and that this often requires monitoring as well as active involvement from government. This type of fisheries management can certainly not be solved by a centralized approach and can only be effective with an active participation and influence from the people directly involved. The neo-institutional framework of analysis that inspires the co-management approach has some great advantages compared to CPT. We have shown in Chapter 4 how a neo-institutional framework may be of help in understanding why some of the assumptions in CPT do not apply in the southern African fisheries. But its conception of institutions is somewhat too simplistic to fit the realities of the African rural communities. Only by taking empirical analyses of the existing institutional landscape into account, does the framework become more useful as a guide for concrete action. CPT is a conceptual model and can only be useful for designing management measures if it is integrated into an empirical analysis.
Most people involved will agree that a uniform concept of co-management does not exist and that co-management requires a reorganization of the institutional set-up. However, this is not a question of constructing local institutions that can cooperate with central authorities in managing the fisheries. Such an approach was already tried in what the colonial powers labelled indirect rule. The term community-based co-management has been suggested in order to get some meaning into the concept (Holm et al., 2000; Jentoft, 2000). In this approach, fishing communities are considered the platform to ensure local participation. There are some serious problems connected to this concept, since our study shows that pure fishing communities hardly exist along the lakes in southern Africa. Most communities have a very diverse occupational structure, and fishing is only one among several occupations that provide the livelihood for the local people. In addition, many fishers have two homes - one at the lakeshore and another somewhere else. It is therefore almost impossible to create boundaries between fishers and non-fishers with regards to the representation of the community in co-management arrangements.
Furthermore, according to the results and lines of thought in the research of Berry (1989, 1993) and others, the main reason for local actors to get involved in government-initiated projects is to secure and improve their access to resources. Granting access rights to specific groups is therefore problematic. Who should be granted such rights? Who is a genuine fisher? Our study shows that everybody is a (potential) fisher, and therefore that no one should be excluded from the decision-making process. This complicates a management system based entirely on the voice of the (existing) fishing communities because it will not reflect the concern of keeping the fisheries as safety valves in peoples strategies towards changing (and until now deteriorating) macro-economic conditions. If co-management primarily becomes a tool for allocation of access rights to local communities, it will prevent the freshwater lakes from remaining commons and safety valves. Since a policy including all potential users will be impossible, the only solution to this dilemma is that government takes upon itself to assure that the collective concerns are voiced and taken into account. We are not arguing that co-management cannot be an appropriate approach at all. Our concern is rather the importance of distinguishing between the allocation of access rights, the making of operational rules to control fishing activities and active participation in conflict resolution. In relation to the last two aspects, co-management can arguably be a useful tool to manage fisheries in the region (Normann et al., 1998).
In most cases where co-management initiatives have been launched, they have coincided with a change of political regime towards more democratic principles of rule, and with the initiation of Structural Adjustment Programmes on request of the World Bank and supported by most of the international donor agencies. This process has led to the initiation of decentralization policies in order to reduce public spending and the assumption that it will bring more local influence in the decision-making processes. Many studies of the Structural Adjustment Programmes (e.g. Mosly et al., 1991) have shown how quickly recipient governments find strategies to by-pass the conditions imposed upon them by the donors. Studies on decentralization processes (e.g. Bierschenk and Olivier de Sardan, 1998) demonstrate how the main result of political decentralization is often a shift of influence and decision-making from one elite to another.
Also, the experiences of setting up co-management in the region have not been very encouraging up till now. Most arrangements have tended to exclude user groups from the decision-making process and from influencing who should participate in the making of operational rules for the fisheries (Sverdrup-Jensen and Raakjaer Nielsen, 1998). Generally speaking, management of freshwater fisheries is still very much in control of governments, and the negotiating position of user groups versus that of governments is consequently comparatively low. As Chirwa (1997:69) points out: The FDs [Fisheries Departments] position of patronage means that the local user communities are the recipients rather than the initiators of decisions. They, themselves, are managed, together with their resources, by the Fisheries Department. This statement seems to be applicable to most of the examples of co-management within the region. Malashas comparative study (forthcoming) on the introduction of co-management in Zambia and Zimbabwe (Lake Kariba) demonstrates that, even if the co-management processes developed very differently in the two countries, control of people seems to have been a more important concern than considerations for the resources. The only difference was that it was the desire for control by the chiefs and the Kapenta operators which became important in Zambia, while it was the wishes of government which have dominated in Zimbabwe. It is probably correct to say that there is little difference between much of todays co-management arrangements and those of indirect rule that the British colonial power established a century ago.
In Chapter 5, we said that a major challenge in the management of small- and medium-sized freshwater lakes in SADC countries is to improve the knowledge base, and we argue that participation of the fishing communities is a precondition for implementing the monitoring systems that we advocate. It is simply not possible to move away from the classical management paradigm towards a more pragmatic and adaptive approach to management without involving the direct users in the process. But instead of opting for advanced social engineering - favoured by many donor agencies and non-governmental organizations - we propose a more modest strategy. Co-management must imply a process of mutual adaptation, where one tries to establish a convergence between government policies and the local institutional structures. Given the great institutional ambiguity and lack of clarity, which we have shown to exist at the local level, even this is by no means an easy task, but rather a long-term process with a lot of muddling through. It will have to be a learning process, and unfortunately, the specific design has to be tailor-made to suit the needs in each specific fishery or lake.
In this respect, it is a problem that government departments and their fisheries research extension services have not been reorganized, and the inputs to be accommodated by government have not been changed as a result of the changes in emphasis towards a more social and economic fisheries management (Donda, 2000). Co-management is mainly an arrangement to ensure communication between governments and communities. From a resource management perspective, it is important to establish a dialogue on the necessity of various management measures, as well as how to measure their impact. It involves an attempt at grass root level to develop new indicators for the development of the fisheries, which can be accepted by government as well as by the communities.
Irrespective of the political systems experienced in the past, fishers trust in government authorities has always been (at best) moderate. They have hardly ever found themselves at the winning end of the relationship and therefore, initiatives to establish co-management taken by government authorities are likely to be met with profound scepticism. As we just said, this is not without good reason; fishermen are suspicious about the motives and the sincerity of government authorities when they propose collaboration and the sharing of responsibilities. In their experience, the government always sets the rules and regulations, and co-management has entered a trajectory where fishers are only involved in the enforcement of the operational rules set by the government. The democratization discourse may have given fishers and other local actors an incentive to give collaborative management arrangements a try. But, if co-management continues on its present path, excluding stakeholders from defining management objectives or general support from government agencies, it may be pertinent to ask what benefit they can have from co-management arrangements.
To conclude, we support Broxs argument (1990) that common property theory still may have its strength as a conceptual model, but that it is definitely not a management tool. Governments active participation in management is certainly needed to avoid major investment-driven growth of fishing effort and to ensure that the small- and medium-sized freshwater lakes can remain a safety valve or a buffer for people in their struggle against macro-economic deterioration. In this way, the fisheries can continue to serve as important means to alleviate poverty and redistribute economic resources. We are aware of the fact that this tends to contradict mainstream management approaches that focus on limiting peoples access to the fisheries. As argued in this report, it is however, clear that this approach falls short for small- and medium-sized freshwater lakes in southern Africa.
But governments involvement has also to be tuned to voices and needs at the local level and co-management can certainly represent an option as long as todays approach is changed. Co-management is fine as long as it focuses on the real or appropriate issues (i.e. conflict resolution and avoiding investment-driven growth of effort) and it is seen as a process of empowering communities and a means to integrate the regulation of fisheries in the general development of the communities. However, if it only continues to be seen through the lens of biological conservation, co-management is not really required in these systems characterized by high system variability and with fisheries targeting species highly resilient to fishing. From a precautionary perspective, there is a need to monitor the fishery and to communicate/evaluate the results to the various users.
 This concept is
presented in more detail in Chapter 5 in the section on Susceptibility of
fish stocks and species to fishing under environmental
 The acts may have had different names in Zambia and Zimbabwe, and Malawi, but the approach of the British government was almost identical for the three countries, at the time of the Federation of Rhodesia and Nyasaland.