The study areas for this research were the villages of Samochima, Mohembo West, and Ngarange, situated in the Okavango Delta’s panhandle (see Figure 2). The Okavango Delta is the largest fresh water body in Botswana [9] and supports the largest fishery in the country [10] . Furthermore, the panhandle has the highest density of fishers in the Delta [11] . The Delta is possibly an extension of the African Rift Valley system [12] , covering an area of 16,835 km² [13] . It is a major tourist attraction [14] [15] and the main water source for north central Botswana [13] . The Okavango River crosses the Botswana border at Mohembo as a single broad river from Namibia with

its origin in Angola, where it is created by the confluence of the Cubango and Cuito Rivers [13] . The annual floodwaters from Angola arrive in the northern panhandle around December/January [16] . The Okavango Delta is characterized by a semi-arid climate, with a mean rainfall of approximately 490 mm∙yr⁻¹ occurring between November and March [17] .

According to the 2001 census [18] , Ngarange village has a population of 948 people; Samochima village has a total population of 847 people while Mohembo West has a population of 1299 people. Essentially, the Delta’s panhandle supports the highest density of fishers [10] and several tourism facilities [14] all within a relatively small area. The fishing season is open from March 1 until December 31 [19] , which coincides with peak fishing effort in the Delta when most fishers return to fishing from arable farming activities [6] . This period also coincides with the tourism season when the lodges in the panhandle receive regional and international tourists.

A structured questionnaire was used to collect primary data on fisheries management and conflict issues in the three study villages (Samochima, Ngarange and Mohembo West) from commercial, basket fishers and tour operators. The research team consisted of one researcher, one associate researcher and two research assistants. Before administering the questionnaire, the research assistant and the associate research were trained in the survey design by pre-testing it in one village outside the study area. This helped the assistants to gain confidence and proficiency in administering the questionnaire, and also allowed for the team to edit and update the survey instrument. Questions were administered among the respondents using their local language, while English was used among the tour operators because they were English speaking. The total population method was used where all the commercial fishers in the study villages were interviewed (Samochima = 16; Mohembo West = 9; Ngarange = 9). However, a random sample of fifteen basket fishers in each study village was determined, resulting in a total of 45 basket fishers therefore 48 respondents were interviewed. According to Mosepele [11] , there are 47 basket fishers in Ngarange, 118 in Mohembo West and 66 in Samochima. Socio-economic and demographic data on commercial and basket fishers was derived from two comprehensive surveys (i.e. using the total count method) conducted by Mosepele [11] in 1998 and Bokhutlo et al. [20] in 2006. Three tour operators, who happened to be the only ones in the study area (during the 2007 field survey), were also interviewed.

All data from structured questionnaires was entered into a Microsoft Excel database for analysis. Frequencies were then calculated to identify the possible causes and remedial measures for conflicts in the panhandle. Relative proportions were also calculated to establish various variables such as age distribution, income distribution, level of sharing of fishing grounds, preferred fishing grounds, and incidence of conflicts as a function of season. A Spearman Rank order correlation implemented in STATISTICA [21] was also used to explore the relationship between conflicts’ incidences and other variables such as flooding and fishing season. Significance levels were set at 90% confidence level for all statistical analysis.

Figure 3 illustrates age distribution of basket and commercial fishers in the panhandle where the youngest fishers were found in Samochima and Ngarange while the older fishers were found in Mohembo West. Samochima commercial fishers have the widest range of age among the study villages. Notwithstanding, the majority of commercial fishers in both Samochima and Ngarange were younger than Mohembo West where most commercial fishers were older.

There were spatial variations in the marital status of basket fishers among the study villages (Table 1). The highest proportion of unmarried basket fishers was observed in Mohembo West while Samochima had the highest proportion of married basket fishers. Furthermore, Table 1 also shows that while the dominant family size in Samochima villages was 5 - 9 people per household, the largest family size of over 15 people per household was also found in this village. Incidentally, the dominant family size in all the three study villages was 5 - 9 people per household.

The majority of commercial fishers in the study villages were married. Commercial fishers in Samochima village had the highest proportion of married fishers (87%), while Mohembo West had both the lowest proportion (40%) of married commercial fishers and the highest proportion (20%) of unmarried commercial fishers among the study villages. Conversely, Ngarange had the lowest proportion of unmarried commercial fishers among the study villages (Table 1). As summarized in Table 1, the common household size for commercial fishers in all the study villages was a family size of 5 - 9 people. Furthermore, a relatively high proportion of the fishers also had a family size of over 15 people, even though Ngarange commercial fishers had the highest proportion of family size of over 15 people.

In terms of income, Table 2 shows that all the basket fishers had an income of less than US$ 85.00 month⁻¹

Where: 1 = Basket fishers; 2 = Commercial fishers.

Where: 1 = Basket fishers; 2 = Commercial fishers.

(using the 2007 BWP vs. US$ exchange rate). Notwithstanding, the basket fishers in the three study villages had diverse livelihood activities. The majority of basket fishers in all study villages indicated that fishing is their major livelihood activity. The second most important livelihood activity in all the three villages was arable agriculture, while pastoral farming and the government’s old age pension scheme were marginally important.

All commercial fishers in the three villages earned more than US$ 85.00 month⁻¹. More than 60% of commercial fishers in the three study villages perceived fishing as their major source of income. More than 53% of commercial fishers in the three study villages also practiced arable agriculture, while a substantial proportion of fishers in Samochima and Ngarange practiced pastoral farming. A small proportion (<13%) depended on the old age pension scheme while another small proportion of fishers were engaged in other livelihood activities. Furthermore, more than 60% of fishers in the three villages regarded fishing as a very important activity (Table 2).

As summarized in Table 3, basket fishers in the three villages had a higher preference for lagoon and floodplain habitats, while only a small proportion preferred fishing from the main river channel. Commercial fishers had a higher preference for all the three habitats (i.e. lagoon, floodplain and channel) while tour operators showed a higher preference for lagoon and channel habitats only. Perhaps due to preference for similar fishing habitats, Table 3 also shows that there were high levels of interaction among these different fisher groups in their preferred habits.

The majority of all the fisher groups in the study villages (except for Samochima basket fishers) indicated that they had full access to their preferred fishing areas. However, a small proportion of Samochima commercial fishers (13%) indicated that they do not have unfettered access to their fishing grounds. The majority of Samochima basket fishers (60%) indicated that they did not have full access to their preferred fishing grounds (Table 4), while a small proportion (13%) of Samochima commercial fishers indicated the same (12% of commercial fishers did not respond to the questions). The majority of Samochima commercial fishers and basket fishers indicated that access was the major source of fishing conflict in their area, while a small proportion of Samochima commercial fishers indicated gear destruction by tour operators as a major source of conflict (Table 5).

Note: T/O = Tour operators.

Note: Tour operators indicated that they did not experience any type of conflicts during this survey.

Some commercial fishers in Samochima and Mohembo West felt that basket fishing was detrimental to fish stocks (Table 6). Moreover, Table 6 also shows that a large proportion of tour operators, and smaller proportions of commercial and basket fishers felt that drive fishing is detrimental to fish stocks. Tour operators and basket fishers in all the study villages, with a higher proportion of basket fishers at Samochima felt that drive fishing is detrimental to fish stocks. About a third of Mohembo West commercial fishers (33%) and a tenth of Ngarange commercial fishers (11%) indicated that mosquito nets were detrimental to fish stocks. Two thirds of the tour operators also indicated that mosquito nets were detrimental to fish stocks, while some basket fishers from Mohembo West and Samochima expressed similar concerns. Some commercial fishers and tour operators also indicated that hook and line fishing was detrimental to fish stocks (Table 6).

More than 50% of Samochima commercial fishers and basket fishers indicated that they experience conflicts during the low flood period in the Delta, while only a very small proportion (<50%) of commercial fishers indicated that they experienced conflict at high flood periods (Figure 4). These two fisher groups from Samochima

Note: T/O = Tour operators.

felt that the conflicts were very serious. Furthermore, a small proportion of Mohembo West basket fishers also indicated that they experience conflicts at low floods as illustrated in Figure 4, though they felt that these were not serious. Table 7 shows that while conflict incidences and flooding are significant at a 90% confidence level (p = 0.051), there is nonetheless a strong negative relationship between conflict and flooding in the panhandle.

The majority of fishers in all the villages indicated that summer is the best fishing season while most of them felt that winter is the worst fishing season (Figure 5). However, a proportion of commercial fishers from Samochima and Ngarange felt that winter was a good fishing season.

The majority of all fisher groups in all the study villages indicated that a spatial zoning of fishing grounds is not a feasible management tool because of limited fishing grounds (Figure 6). Nonetheless, the majority of fishers (i.e. commercial and basket fishers) felt that the Delta’s fisheries need to be regulated by both government and traditional authority (Table 8).

According to this study, commercial and basket fishers from Samochima experience more conflict in the Delta’s panhandle, compared to the two other villages. This conflict is caused by the high interaction between the three major fisher groups (small-scale commercial, subsistence and recreational) in their main fishing grounds (i.e. lagoons, main channel, etc.). This observation makes sense, taking into account that the major contributing factor to this conflict is lack of access, while gear destruction is experienced by a relatively small proportion of commercial fishers. The fact that most of the conflict occurs during the low flood season suggests that there is increased competition for limited space which subsequently accentuates access (due to competition) as the major cause of conflict at Samochima. Tweddle et al. [8] also found that commercial and recreational fishers in the

Delta compete for the same space and resource. Furthermore, Mosepele et al. [22] observed that recreational and commercial fishers target the same fish species, the former targeting trophy sized individuals, and the latter target market sized fish. Exploiting the same species also creates conflict among fishers.

Furthermore, several conflict issues highlighted by tour operators in the Delta [8] suggest the existence of a fishing conflict caused by contradictory multi-use of water which has been observed in other fisheries elsewhere [23] [24] . Understandably, Welcomme [25] observed that recreational fishers normally come from “a wealthy professional urbanite” group that does not depend on fish as a source of livelihoods but regard it as a recreational activity. Therefore, it is on the basis of divergent philosophical orientations [26] that the most intense conflict in fisheries occurs between recreational and commercial fishers [27] . Nybacka [28] also found that there was a seasonal variation in conflict between recreational fishers and the rest of the fishery (in Finland) where conflict normally occurs during the warm summer vacation months. However, conflict between commercial and recreational fishers can also develop due to lack of communication [5] , which agrees with findings from this study for the Okavango Delta fishers.

The perception by commercial fishers that the use of fishing baskets is detrimental to fish stocks, especially by the Mohembo commercial fishers, coupled with some low level conflict experienced by Mohembo basket fishers suggests the existence of some conflict between these two groups. It is possible that commercial fishers (gill net fishers) could be concerned about potential recruitment overfishing of their target species because, by their nature, fishing baskets harvest smaller sized species (but not juveniles of commercial species) as observed by Mosepele et al. [22] . This type of conflict might be based on what Haraldsdottir [29] describes as gender- based resource tenure issues where women’s “de facto rights” are trampled by men’s “de jure rights”. Past research in the Delta [22] [30] showed that basket fishers target smaller-sized species in shallow water on the fringes of lagoons and channel habitats. Therefore, while basket fishing is a major source of household nutrition especially during times of food scarcity [30] its value is being compromised through conflict.

Generally, conflict is an intrinsic attribute of multiple-gear fisheries that share the same space and time, or compete for the same species, or both. According to Hilborn [26] , conflicts in fisheries can also be caused by the four objectives of fisheries management (social, economic, political, and biological) outlined in Hilborn and Walters [31] . Hilborn [26] argues against one of the basic tenets of classical fisheries management which advocates for fishery exploitation levels to achieve maximum sustainable yields (MSY). He argues that fishing to achieve MSY generally contributes to over-capitalization in some fisheries, which is created by government subsidies as a socio-economic objective (of job creation). Similar observations were in made in the Delta with the advent of funding for fishing activities through the Financial Assistance Policy in Botswana. This effectively commercialized the Delta’s fishery through the provision of grants to purchase modern fishing equipment [32] . Increased capitalization increased competition and access issues which contributed to conflict, especially at low flood levels when there are limited fishing areas. This argument is based on the observation that Samochima commercial fishers cover large distances in their fishing expeditions [8] ostensibly in search of “better” fishing grounds due to localized over-fishing. Inevitably, they encroach on other fishing grounds downstream, which might be frowned upon by resident fishers (despite the prevailing open access regime).

The current legal framework in the Delta’s fishery is based on a classical paradigm [33] that does not however, address conflict resolution in its entirety. Furthermore, fisheries management in Botswana is an institutionalized top-down approach [34] that has not worked effectively in most fisheries elsewhere. While it is possible that the lack of a national fisheries policy [33] might contribute towards the existence of conflict in the fishery, it is equally possible that the lack of well-defined user rights and lack of a holistic management approach that includes most or all of the stakeholders is also a major contributor.

Three main conflict issues have been identified in this study. Firstly, the level/intensity of conflict is flood pulse driven and characterized by seasonality due to the flood regime. Secondly, the observed conflict between commercial (i.e. gill net) fishers and basket fishers is possibly caused by misguided concerns based on fishing gear used. Lastly; the recreational fishers’ aversion to drive fishing is possibly based on concerns of potential over-exploitation due to the high catches normally associated with drive fishing (see [34] ). Different approaches can be used to manage these three major conflicts in the fishery. In other conflict-ridden fisheries, conflict is resolved through spatial zoning of fishing grounds as a self-regulatory measure [35] . A similar observation was made by van Ginkel and Steins [24] that fishers sometimes use self-regulation (based on informal regulations) for spatial zoning as a conflict resolution mechanism.

For the Delta, a spatio-temporal zoning pattern is proposed as a conflict resolution mechanism, taking into consideration the additional factor of the flood pulse. Recreational fishers can use the main channel and the large lagoons immediately adjacent to the main channel for their fishing activities at low floods, while commercial fishers can fish in the large floodplain lagoons and side channels during this period.

This proposed approach can be based upon a co-management paradigm as the main conflict resolution tool for the Delta’s fishery. This approach is based upon the FAO Code of Conduct for Responsible Fisheries [36] that advocates the inclusion of local communities in management of their fisheries resources. Subsequently, co- management is used extensively to achieve a more holistic approach to fisheries issues [37] even though Caddy [38] acknowledges that setting up a “decision-making process” is the most difficult (and yet important) aspect of the fisheries management process. Sipponen and Greboval [5] highlight that a forum based on a co-management paradigm can be created for commercial and recreational fishers to communicate as a conflict resolution mechanism. One such involving and highly democratic approach was adopted in Samoa by King and Faasili [39] which culminated in a village fisheries management plan which was developed by most community groups, including women. Therefore, co-management has indeed “become the buzzword of contemporary fisheries management” [40] and has subsequently been adopted/proposed as a major strategy towards a holistic management paradigm in fisheries around the world (e.g. [5] ).

Jul-Larsen et al. [41] argue that co-management is one of the best tools for conflict resolution in African freshwater fisheries. Moreover, van Ginkel and Steins [24] highlight that compliance for regulations to solve multiple-use conflicts can be sought and negotiated at the local level before these can be eventually codified into law, which agrees with Overby [35] about the eventual integration of zoning (based on self-regulation) into federal law in the Gulf of Mexico. A more fundamental step towards co-management is to recognise the interests of groups that have been historically disadvantaged, such as women fisher groups [29] . Indeed women basket fishers in the Delta are not organized in any formal grouping and are therefore not part of any management process [42] . While subsistence fishing is a major source of livelihoods in the Delta’s communities [43] , especially basket fishers to whom fishing is also a form of cultural expression [30] , their concerns need to be integrated into the proposed co-management paradigm.

According to Siar et al. [44] , educational attainment is positively related to the attainment of fishing communities towards territorial user rights in fisheries management. Therefore, this suggests that intense education campaigns are critical towards a successful implementation of a co-management fisheries regime and should therefore be done in the Delta to increase the probability of success of this initiative. Mosepele [11] observed that conflict in the fishery was also caused by lack of a common understanding and appreciation of the value of the fishery to each stakeholder in the fishery. This conclusion was based on Mosepele’s [6] observation that access and property rights were the major precursors of conflict in the fishery and not fish stock over-exploitation. Therefore, Mosepele et al. [45] through the BIOKAVANGO project [46] created a committee whose mandate was to implement an adaptive co-management regime in the fishery. The basic philosophy behind this adaptive management regime was to create a feed-back system whereby the committee would get feedback on its management interventions through a dynamic process.

The Okavango Fisheries Management Committee (OFMC), formed in the Delta’s panhandle during the Biokavango project [45] , is similar to Boisneau and Mennesson-Boisneau’s [4] intervention, which advocates for a participatory fisheries management regime that includes all the major stakeholders. This Committee essentially creates a forum that enhances communication between tour operators (recreational/sport fishers), local fishers (commercial and subsistence) and the regulatory authorities (key government departments) in the Delta’s panhandle. This management intervention is also aligned to Sipponen and Grevobal’s [5] analysis that lack of communication is usually the root cause of conflict in fisheries. The OFMC then developed a Code of Conduct as a conflict resolution tool through a democratic process (the process leading to the development of the OFMC is described fully in Mosepele et al. [45] ). It is hoped that this participatory approach to fisheries management will contribute towards a sustainable utilization of the resource that will benefit all the stakeholders. According to Soma [47] , sustainable development of natural resources can be achieved through the inclusion of all stakeholders through a participatory approach. It is envisaged that the OFMC is the vehicle for that participatory approach in the Delta’s fishery.