The purposes of the survey were threefold. First, to collect catch and effort data over a 12-month period from small-scale fisheries at Sebaboleng and Maqalika Dams in Maseru Town, as a complement to fish sampling data. Second, to collect selected socio-economic data on individual fishermen and thereby obtain their profile. Last, but not the least, the method and the survey design were themselves the focus for testing and evaluation, since documented experiences from similar surveys in southern Africa are scarce.
Catch and effort data were obtained for the whole survey period, although data were scarce for some months, as described below. A large number of fishermen were identified and socio-economic profiles obtained. Methodological experiences were evaluated as outlined below, and will contribute to further ALCOM work in this field.
The method chosen for the survey had not been tested earlier by ALCOM as a monitoring tool, which is why the survey was somewhat experimental.
The survey design was done by the author in collaboration with ALCOM's Senior Aquaculturist and Consultant Socio-economist.
An attempt was made to obtain full coverage of fishing activities during a 12-month period, so that it revealed seasonal variations. However, it was considered too costly and impractical to have enumerators at the dam every single day. So a set-up was selected with interviews being done every two weeks, during all seven days of each week (see Time Frame, Appendix 3). During this week, questions about the last two weeks' fishing activities were asked. This procedure would yield data on fishing activities throughout the year and reveal annual fluctuations.
A questionnaire was designed and field-tested with 12 respondents. It was reviewed after testing and used throughout the survey period (see Appendix 1). All questionnaires were completed by employed enumerators. Questions were formulated in English but the enumerators carried out interviews in Sesotho to ensure better communication. The enumerators were equipped with spring scales for weighing catches and a measuring stick for recording the length of fish. They were trained by staff from the Fisheries Section, in collaboration with ALCOM and the Department of Youth and Women's Affairs. On-site supervision was done mainly by the Department of Youth and Women's Affairs, with continuous support from ALCOM staff.
As the number and identity of persons fishing at the dam was unknown at the start of the survey, it was decided that two weeks be spent initially identifying fishermen (through use of a key sheet or identification sheet, see Appendix 2a). The key sheet was slightly altered after four months of survey work, by adding a few questions (see Appendix 2b). It was changed to make it simpler to fill in and to allow space for a fisherman's signature. Each fisherman was assigned a unique respondent number, which was used throughout the survey.
The survey then started focusing on the fishermen identified but also continuously including new fishermen. Every time an unknown fisherman was met, a key sheet was completed for him and he was thereafter included in the survey. At this stage, an attempt was made to monitor more closely a sample of fishermen. The purpose of this was to gain more detailed information on people that fished often, apart from monitoring all other fishermen as well. A sample of 20 fishermen was then selected from the group of fishermen initially identified, using the criteria of “regular fisherman” (see 2.3). A “regular fisherman” was defined simply as a person who fishes more than once a month.
The enumerators were told to make sure that they interviewed the fishermen every two weeks even if they did not meet them at the dam site. As physical addresses had been obtained, interviewing the fishermen at their homesteads, if they did not come to the dam, was considered quite simple. However, this proved difficult. Some people did not stay at the addresses indicated, and some were not known by the names they had given the enumerators. Some of the fishermen were never at home when the enumerators called, despite repeated attempts by the latter and with forewarning too. In a nutshell, these problems made it impossible to closely follow a selected group of persons fishing more intensively than others -- a procedure which would have provided welcome detail on fishing practices.
In the light of this background, a census survey approach was chosen. One of the shortcomings of the survey design is that it limits the use of statistical inference techniques, e.g. for extrapolation of catches. It has to be assumed that the number of fishermen interviewed corresponds well to the actual number of persons fishing at the dam. Theoretically, a number of fishermen could have been missed if they happened to go fishing only on the non-interview weeks. However, this is unlikely. Enumerators sometimes went to the dam during non-interview weeks either by mistake, or to compensate for not going on working weeks. On such occasions they mainly met the same individuals. All in all, 240 fishermen were identified, of whom 209 fished exclusively at Sebaboleng dam.
The enumerators indicated that they knew who the regular fishermen were. Therefore it seems unlikely that substantial catches were taken out by unknown persons. Of course the possibility of illegal fishing at night with nets did exist, but it was never reported by fishermen or people living around the dam, and not heard of by the enumerators during the survey period. Nor did the Fisheries Section report such activities at the survey site.
The main weaknesses in the data collected lay in a probable under-reporting of catches due to (a) the respondents being suspicious of the enumerators (b) enumerators often stopping their work before the actual fishing stopped, or (c) simply missing people. The large size of the Sebaboleng dam made it difficult for enumerators to stop all fishermen for interviews when they were through with fishing. Sometimes fishermen had already left the dam site before the enumerator arrived. The survey design also made analysis of the fisheries data somewhat complicated.
Another weakness of the data is that only one dam was surveyed -- instead of the two initially planned. This was because of circumstances beyond the control of supervisors and enumerators, as described under Section 2.2.1
Some of the problems met during the survey affected the quality of data collected. These are outlined below.
- Interviews done outside agreed dates: During some survey months, the enumerators did interviews outside the agreed interview weeks, instead of sticking to the time frame. This was unfortunate, as these questionnaires could not be used. When interview weeks were not adhered to, catches recorded by “right” and “wrong” questionnaires (questionnaires used outside the agreed weeks) got mixed up.
- Weighing and recording weight of fish: Questions relating to “fishing today”, i.e. the interview day, were formulated with a view to collecting length and weight data of fish actually seen by enumerators. Length of fish was measured in size classes, with enumerators using a stick with size classes marked on them. Size data are considered fairly accurate; it is easy to measure fish seen, and also quite easy for individuals to remember the sizes of fish disposed of by pointing them out on the measuring stick. However, collecting weight data proved more complicated, as fishermen had often already disposed of part of the catch when the enumerators arrived. It was in those cases impossible for enumerators to record weight, as fishermen could not assess it. Only the weight of fish actually seen by the enumerators was recorded.
- Suspicion about use of results and purpose of survey: Even though supervisors and enumerators tried hard to gain the fishermen's trust, by carefully and repeatedly explaining the purpose of the survey, it is possible that they were perceived as representatives of authority collecting information on the fishermen in order to control them. This explains the refusal by some of the fishermen to be interviewed. It would also have biased some of the correspondents' answers -- they may have understated actual catches for fear of taxation and the like.
The survey was initially planned for two adjacent dams, Sebaboleng and Maqalika: it was known that these two dams were being fished. However, individuals fishing at Maqalika tended to walk away (or even run off!) when enumerators approached them. Some Maqalika fishermen told the enumerators that they did not want to be interviewed as they were constantly being chased off the shoreline by the authorities. The latter wanted to restrict access to the dam, which constitutes an important source of water supply for Maseru town. The enumerators were identified with this action, although they did not have anything to do with it. This made it impossible for them to work at Maqalika.
“Research fatigue”: The enumerators expressed concern during the second half of the survey period about some fishermen not wanting to be interviewed any longer. This can be an expression of the so-called “research fatigue” in the field, which is a result of communities being saturated with inquiry. During the survey design, this problem was foreseen and ideas on how to limit its impact were discussed. It was considered appropriate that the implementors of the survey “reward” the respondents by organizing a training seminar about half-way during the survey period. The training seminar would include information on reservoir fisheries subjects; improved handlines and other gear would be tested; Assistance to self-help projects would also be included. However, despite repeated calls through various communication channels, less than 10 fishermen turned up for a preparatory meeting of this seminar. At later meetings too, attendance was equally poor. The training seminar was therefore not feasible, and had to be dropped.
Research fatigue affects the accuracy of answers to survey questions. As the motivation of fishermen decreases, the likelihood of answers being incorrect increases. Enumerators suspected this to be the case with some fishermen, a minority. As the second half of the survey period was the peak fishing period, October–February, this problem may have affected the results more than if the survey had started later in the year.
All fishermen at Sebaboleng Dam were classified as part-time fishermen during the survey period. Part-time fishing is here understood as taking up 75% or less of a normal full-time working week (40 hours) of the fisherman.
The following definitions are used in the report to distinguish between different categories of fishermen:
“Regular part-time fisherman” is a person who goes fishing more than once a month on any reservoir.
“Occasional part-time fisherman” is a person who goes fishing less than once a month on any reservoir.
ALCOM's test fishing programme in Lesotho was initiated in January 1991 when five dams were selected for regular test fishing (Thamae, Sebaboleng, Maqalika, Mejametalane and Thaba-Phatsoa). The first four are all within the Maseru town area, and the last one in the lowlands of Leribe District. The criteria for selection of Sebaboleng and Maqalika dams for the monitoring survey were:
permanent water bodies;
individuals fishing regularly;
water bodies were part of the test fishing programme; and
easy access for Fisheries Section and Department of Youth and Women's Affairs' staff.
Furthermore, Maqalika dam had been identified as one of the more important dams for the test fishing programme, being the largest of the selected water bodies, and being most likely to have a good potential for sustainable gill-net fisheries.
The two dams are linked to each other by Sebaboleng spilling into Maqalika when it overflows (usually after every rainy season). Maqalika has approximately 64 ha surface area when full, and Sebaboleng 18 ha. The two dams are located in the northern part of Maseru Town, approximately one kilometre from the town centre.
Sebaboleng dam is surrounded by settlements and has mainly a flat or gently sloping shore. One stretch of the northern shore is steep and quite inaccessible, making it difficult to fish from. The dam also goes by the name of “Lakeside”.
Maqalika is a fenced dam under the control of the Water and Sewerage Branch of the Ministry of Water, Energy and Mining, with fewer settlements around than in Sebaboleng. It is an important source of municipal water supply for Maseru town. Sebaboleng on the other hand is a dam with open access under the control of the Ministry of Agriculture, Co-operatives and Marketing. Sebaboleng is widely used for cattle watering and domestic purposes, apart from the fishing.
Since the authorities impeded access of fishermen to the shore of Maqalika dam during the survey period, as explained earlier, the dam had to be excluded from the survey at an early stage. A few Maqalika fishermen are included in the summary of key sheet data, but none in the questionnaire data.
