PRELIMINARY ANALYSIS OF THE FLYINGFISH FISHERY OF SAINT LUCIA AND INTERIM RECOMMENDATIONS

by
Williana Joseph
Department of Fisheries
Water Front, Castries, Saint Lucia
[email protected]

BACKGROUND

The majority of fishing vessels in Saint Lucia catch flyingfish during the flyingfish season however, many fishing vessels from the West coast communities of Gros Islet, Castries. Anse la Raye, Canaries, Soufriere, Laborie and Choiseul target flyingfish compared to communities on the East coast.

Under the current data collection system, data are not collected from at least three primary sites where flyingfish landings are important components of the landings. The annual landings for the island vary considerably from year to year but on average contribute between 3 - 10 percent of the total landings (George, 1999).^[7]

2. OBJECTIVES

The main objectives of the workshop include:

1. Analysis of flyingfish landings and fishing effort of participating countries.
2. Assess the quality of data, data collection and storage procedures.
3. Estimate regional total landings and abundance trends.
4. Develop an outline of a regional management plan for flyingfish.
5. Discuss and compare social and economic importance of flyingfish to each island.

3. DATA USED

The data used in the analyses for Saint Lucia included information from observed fishing trips (landings and effort) from various landings sites over a five - year period (1995 - 1999). Data also included information on vessel type but all were classified as dayboats, number of crew, season, vessel length and engine horse power. In addition to, annual estimated flyingfish landings from 1983 to 1999. Hard data sets for 1990 to present are available on flyingfish landings for Saint Lucia whereas data sets from 1995 to 2000 are stored in TIP vs. 3.4.

4. STATISTICAL AND ASSESSMENT ANALYSES

4.1. Exploratory Analyses

4.1.1 Objectives

The main objectives of this procedure were to examine the data set more closely for errors and omissions, including seasonal, spatial and annual trends.

4.1.2 Method/Models

Summaries of data were undertaken in EXCEL and SPSS which included summarizing observed and estimated catch and effort by month, site location gear season and year. Graphical representation of data in various formats including box plots, leaf and stem plots, bar charts and line graphs were done. In addition, conversion of landings (weight) to a log scale, then performing the above summaries and construction of the various plots.

4.1.3 Results and Discussion

Considerable variation in estimated annual landings was observed over the 17 - year period (Fig. 1). A general trend of decrease in estimated landings was noted between 1985 and 1997, however an increase in landings was observed in 1998. The reasons for this trend are not known but may be due to differences in annual abundance of fish, in addition to other social, economic and environmental factors such as: availability of higher valued species, market and ex-vessel price of flyingfish and weather conditions (hurricanes).

The predominated gear used in the fishery is gillnet and it contributes at least 67 % of the observed sampled effort however, hand lines and dipnets are also popular gears contributing 11 % and 6 % respectively (Table1). Trolling (TROL), bottom long line (BLIN) and surface long line (SLIN) were observed collectively to contribute at least 11 percent of the observed effort. Verification checks will be conducted for the relevant trips and gear information corrected for future analyses since gears such as BLIN, SLIN and TROL were the main gears used during the fishing trip.

A plot of the estimated annual effort (1995 - 1999) and landings over the period under consideration is presented in Figure 2. The total estimated effort was calculated using the formula:

Total estimated landings/total estimated effort = CPUE
If CPUE = observed landings/observed effort then
Solving for Total annual effort having estimated annual landings and annual CPUE
Estimated annual effort = Estimated annual landings/annual CPUE.

Values obtained ranged from 3 000 - 8 000 trips (Fig. 2). These values are quite large and do not reflect the true effort for this fishery. A box plot was constructed for observed landings to determine trends in landings by month throughout the period 1995 - 1999 (Fig. 3). Throughout the period under consideration, landings in April, July and August were low (below 50 kg); landings were not recorded for September and October. The highest recorded landings occurred in January - March and May - June. For all months, except April, the distribution for each month was negatively skewed - very few small values. Hence, the Fishery shows a seasonal trend primarily occurring between November and July.

Sampling coverage over the five-year period was greater in 1995 than the other years possible due to the complete census data collection system that was in place at this time. Four sites, Anse la Raye, Micoud, Gros Islet and Soufriere contributed the greatest effort for flyingfish (Table 2). Frequencies of the observed effort by year and by gear are shown in Table 3 and Table 4 respectively.

4.2 Standardization of Effort

4.2.1 Objective

To assess the interaction (effects) of various factors which may affect effort but do not represent changes in abundance of the stock in an effort remove these factors or to account for them in the analyses.

4.2.2 Method/Model

Attempts to standardize catch rates (landings per trip), using factors such as gear, season, month and site were undertaken using the formula: log landings = log (Landings +1) to convert the observed weight of flyingfish. Log landings were then used in a GML model (SPSS v.8.0.).

4.2.3 Results and Discussion

The contribution of each factor (column Type III Sum of Squares) was quite low resulting in insignificant effect by each of these factors. Both R squared and adjusted R squared values are extremely low probably as a result of the fact that the assumptions of the model not being satisfied.

Comparison and aggregation of CPUE data for the Eastern Caribbean region were not done due to several issues that needed to be resolved such as:

i) Difficulties in standardising effort regionally; for example equating a Tobago or Barbados day-boat with a St Lucian or Dominican pirogue or canoe;

ii) Differences in data collection strategies: and

iii) Differences in flyingfish fishery throughout the region.

5. RECOMMENDATIONS

5.1 Statistics

1. For future analyses and to obtain a longer time series of catch and effort for the flyingfish fishery of Saint Lucia, the Department of Fisheries should consider converting hard copy data sets (archived) into digital data sets.

2. Conduct thorough gear verification checks for the current digital data set for this fishery due to the results of the exploratory data analyses.

3. Revise the present data collection plan to include at least two primary sites where flyingfish are targeted. At one of these primary sites collection of fish landing data commenced in October 2000 and the Department considering commencing the collection of fish landing data at another site in April 2001.

4. Train data collectors in flyingfish species identification due to the multispecies nature of the flyingfish fishery. At least two flyingfish species are captured by this fishery, but both species are recorded as one species.

5. Complimentary data social, economic and environment data should be collected to assist in explaining some of the trends observed and to ensure that social and economic aspects are in the development of management strategies.

5.2 Management

In order to have a better idea of effort, in terms of the number of fishers/vessels involved (target) in this fishery, the Department should consider conducting a survey to gather the data during their annual vessel licensing programme.

Figure 1. Estimated annual landings of flyingfish in Saint Lucia

Table 1. Frequency of gear types sampled at each site (1994 - 1999)

Figure 2: Estimated annual landings and estimated annual effort

Table 2. Frequency of sampling at each site for each year.

Figure 3: Box plot of observed landings for each month (1995 - 1999)

Table 3 Monthly sampling frequency for each year.

Table 4 Annual sampling frequency of gear-type.

Figure 4. Annual monthly observed effort (trips) for Flyingfish in Saint Lucia.