Part of the reason for such differences in targeted and landed species diversity is the fishing gear and/or boats being used. The analysis of survey data related to fishing boats revealed the following:

• The vast majority of boat owners owned just one boat: in Koukoudé, the majority of boat owners had one boat but a larger proportion of owners had multiple boats.
• Salan (traditional Guinean canoes) was the most popular type of boats in all three communities, while the second most popular type was flimbote (brought in from Senegal). The other two observed boat types were gbankenyi (traditional Guinean) and yoli (Senegalese).
• Over 95 percent of these boats were motorized by outboard engines.

Overall, the FISH2018 survey data show that fish production is more diversified in Bongolon. This may be an indication of greater resilience of Bongolon’s local fisheries to shocks in the ecosystem, as the conventional belief suggests. However, at this point the hypothesis is not confirmed. Nonetheless, the findings likely have implications for the design of fisheries co-management pursued in WARFP-GN.

Food Security in Target Fishing Communities in Guinea

The LSMS2018-FISH data were used to analyze the living standards and food security of the residents in the three communities. The survey asked about household expenditures on food and other items, as well as subjective questions about household food security. This analysis is based on the following two sets of variables.

• Household expenditures: (1) food expenditure, (2) total expenditure, and (3) proportion of food expenditure in total expenditure.
• Responses to the subjective evaluation questions of household food security asking whether the following conditions applied during the past year: (1) concerned about food shortage, (2) unable to eat healthy food, (3) ate but not much variety, (4) skipped a meal due to lack of food, (5) ate the amount less than ideal, (6) ran out of food, (7) hungry but did not eat, and (8) spent whole day without eating.

Table 5.1 summarizes the variables used in the analysis. (Note that “HH” denotes household.) The analysis presented here suffers from the small sample size of the LSMS2018-FISH dataset, which was partly driven by the small budget for the survey: altogether 84 observations are included, with 36 from Koukoudé and 24 each from Bongolon and Matakang. Looking across the demographic variables used in the analysis, the average household size was largest in Bongolon (5.7 people), followed by Matakang (4.9) and Koukoudé (3.2). On average, the households were older in Matakang (with an average age of 43 years) and Bongolon (41) than in Koukoudé (36). Overall, 40 percent of the households in the sample had at least one household member engaged in fishing (referred to as “fishing household” hereafter). The proportion of fishing households is highest in Matakang (67 percent), followed by Bongolon (46 percent) and Koukoudé (17 percent). Other variables are discussed in more detail below.

Table 5.1. Descriptive statistics of selected variables from the LSMS2018-FISH survey

	Total	Koukoudé	Bongolon	Matakang
Variables	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Food insecurity variables (0=no, 1=yes)
Concerned about food shortage	0.679 (0.470)	0.556 (0.504)	0.875 (0.338)	0.667 (0.482)
Unable to eat healthy food	0.524 (0.502)	0.444 (0.504)	0.875 (0.338)	0.292 (0.464)
Ate but not much variety	0.690 (0.465)	0.556 (0.504)	0.792 (0.415)	0.792 (0.415)
Skipped a meal due to lack of food	0.506 (0.503)	0.333 (0.478)	0.833 (0.381)	0.435 (0.507)
Ate the amount less than ideal	0.631 (0.485)	0.500 (0.507)	0.833 (0.381)	0.625 (0.495)
Ran out of food	0.369 (0.485)	0.306 (0.467)	0.750 (0.442)	0.083 (0.282)
Hungry but did not eat	0.333 (0.474)	0.250 (0.439)	0.708 (0.464)	0.083 (0.282)
Spent whole day without eating	0.286 (0.454)	0.167 (0.378)	0.625 (0.495)	0.125 (0.338)
Demographic variables
HH size	4.405 (2.602)	3.194 (1.737)	5.708 (3.581)	4.917 (1.640)
Adult-equivalent HH size (FAO)	3.254 (1.878)	2.402 (1.209)	4.127 (2.603)	3.658 (1.275)
Adult-equivalent HH size (Alternative)	2.614 (1.146)	2.069 (0.799)	3.180 (1.530)	2.865 (0.753)
HH head’s age	39.4 (11.3)	35.7 (10.1)	41.3 (12.8)	43.0 (10.3)
HH head’s highest education attained (0=none ~ 4=post-secondary)	0.714 (1.025)	0.611 (0.934)	0.875 (1.076)	0.708 (1.122)
HH engaged in fishing (0=no, 1=yes)	0.393 (0.491)	0.167 (0.378)	0.458 (0.509)	0.667 (0.482)
Expenditure variables (million Guinean franc)
1. HH annual food expenditure	24.6 (22.0)	23.6 (19.1)	33.5 (30.8)	17.0 (9.7)
2. HH annual total expenditure	45.2 (31.4)	44.7 (30.4)	57.5 (39.5)	33.8 (17.3)
3. HH annual Engel’s coefficient [(2)/(1)]	0.545 (0.190)	0.529 (0.179)	0.605 (0.204)	0.507 (0.186)
Sample size (N)	84	36	24	24

A series of statistical analyses to explain the variations in average household expenditures and subjective food security across communities revealed the following tendencies:

• Subjective “food insecurity” was strongest in Bongolon; statistically Matakang and Koukoudé were not different.
• Matakang had lower food and total expenditures than the other two communities on average; the food expenditure proportion was also lowest in Matakang. This may be because the general price level in Matakang (including the price of food) is likely lower as the village is remote relative to the other two communities.

Next, the data were used to analyze the variations in these variables at the household level. The results of the regression analysis are summarized as follow:

• As was expected, both food expenditure and total expenditure were positively related to household size.
• Food expenditure, total expenditure, and food expenditure proportion were larger for more educated households. This may mean that more educated households (1) earned higher incomes and (2) placed more importance on food and nutrition. To confirm this hypothesis, however, further analysis with a larger dataset is necessary.
• There was no evidence that a fishing household had higher or lower food or total expenditure. In fact, the proportion of food expenditure in total may be lower for fishing households. There was no statistical difference in the responses to the food security questions between fishing and non-fishing households.
• For both household expenditures and subjective food security questions, the most important variable explaining the variations was location (that is, Matakang, Bongolon, or Koukoudé).

Conditions for Fishery Co-Management

A single definition of co-management does not exist. Nevertheless, one common feature is the involvement of some level of coordination among the resource users, who otherwise can be, and have been, operating independently. Based mostly on FISH2018 survey data on fishing boat owners, conditions for successful fisheries co-management systems in the three communities were investigated.

1. Non-fishing occupations among fishing households

The survey asked what secondary occupation the boat owners were engaged in. In all three communities, a fishing household that only did fishing was relatively a minority; about 40 percent of fishing households in Matakang and Bongolon were full-time while less than 20 percent were so in Koukoudé. There were many fishing households that engaged in farming and other occupations; only a few in Matakang were engaged in livestock production.

This prevalence of secondary occupations can be either an advantage or disadvantage for fishery co-management. It is advantageous because the members of fishing co-management need not be privileged solely based on fishing, but from a combined benefit from all occupations. For example, cutting back the fishing days to curtail overharvesting allows fishing households to spend more time in farming, supplementing the loss from fishing in the short run while the fish stock recovers. On the other hand, the presence of an alternative occupation could reduce the relative importance of fishing as an income-earning activity, thereby diminishing the incentive among the fishing households to co-manage the fishing. For this reason, it will be important to understand the relative importance of fishing and other occupational activities among the fishing households.

2. Financial ties and co-management membership

The FISH2018 survey included a set of questions that are indicative of the existence of financial ties between fishers and fish buyers. The responses to these questions revealed the sources of finance that fishing households received and the buyers to whom they sold their catch. A majority or close to majority of fishing households financed their boats and equipment on their own. However, Bongolon is unique in that close to 50 percent of its fishing households were financed by fishmongers—the highest among the three communities. Matakang follows closely with approximately 30 percent of its fishing households financed by fishmongers. With this in mind, an interesting pattern emerges on the sales side. In Bongolon, a dominant share of sales went to fishmongers as opposed to the first-come, first-served situation in Koukoudé. The sales destinations in Matakang were more evenly spread, but the destination with the highest frequency is fishmonger as well. These results suggest that some sort of debt-and-sales relationship existed.

In the context of fishery co-management, this observation has two implications. First, fishing households that are in a debt-for-sale relationship may be hindered from pursuing other business opportunities (such as selling to buyers who offer higher prices). Second, the fishmongers and other actors in a similar financial relationship with fishing households need to be part of the co-management arrangement. Financers will have an incentive to support improved fishery management because the performance of the fishery directly relates to their loans being repaid, in addition to improving the supply of fish to financiers’ fishmonger businesses.

3. Cohesiveness of fishing households in community

Non-economic factors such as disputes among fishing households in a community matter for the success of co-management. In all three communities cooperative-like associations were present with varying degrees of activeness and success. The FISH2018 survey asked what incidents these associations helped resolve. There are considerable differences in the nature of incidents across the three communities. In Koukoudé, disputes among members dominate the incidents, while Matakang experienced both member and outsider disputes equally. Bongolon, on the other hand, had very few incidents of any kind that required their associations to help resolve. This is not surprising as Bongolon’s associations are considered to be well functioning. Co-management groups in Matakang and Koukoudé will need to be aware of the nature of member disputes so as not to let them undermine the co-management effort.

Conclusion

Overall, the activities related to survey data collection and analysis in Guinea led to the following conclusions.

• Survey data are not commonly used for informing projects in the fisheries sector. However, this study showed that survey data can provide important insights into how small-scale fishing is conducted on the production side, as well as insights into the well-being of residents of fishing communities more generally.
• The sample size of the dataset from the LSMS oversampling in Guinea was too small, and definitive inferences that inform WARFP-GN could not be obtained. This experience points to the importance of investing in data collection long before the phase of project design and implementation.
• The most important variable explaining the differences in expenditure-related variables was the community indicator. Therefore, understanding the condition and structure of the village economy is important when designing project activities in each location. In Chapter 6 of this e-book, the village economy of Koukoudé is analyzed in detail.