Exploring the vulnerability of a commonly exploited fish stock, Galeoides decadactylus in the artisanal mixed capture fisheries of Sierra Leone

Galeoides decadactylus is a commonly exploited fish stock in Sierra Leone, and the study of its vulnerability to capture in the artisanal mixed fisheries of Sierra Leone has not gained the needed attention of researchers, for which the present study has come in handy particularly when such information is critical for sustainable management of the stock. The study utilized length-frequency data obtained from 865 specimens as input datasets of analytical models inscribed for the FAO Fish Stock Assessment Tool (FiSAT II) software. Vulnerability indices for this study were the Length Structured Virtual Population Analysis (LVPA) and the Probability of Capture (PC). Results revealed minimum and maximum total length of 11cm and 36cm respectively for the study species, and LVPA showed highest catch (in numbers) for the length range, 20 cm (n = 94, 0000) - 21 cm (n = 100, 000). Also, the probability of capture evinced greater selection probability ( p = 0.9) of the larger sizes (17cm - 22 cm (mean, 19.5 cm TL) at a maximum mesh size of 8cm of bottom drift net utilised by the fishers. The results of LVPA implied that the smaller length groups of the assessed stock (TL ≤ 11 cm) are less vulnerable to fishing mortality, but most susceptible to natural mortality, and the probability of capture indicated that the gill nets used by the artisanal fishers for this study were highly selective to allow greater number (n = 4.8 x 10 6 ) of the smaller length group escape capture. In conclusion, G. decadactylus was efficiently exploited in the artisanal mixed fisheries. However, stringent measures on capture input and output controls are advised for the assessed stock.


INTRODUCTION
Galeoides decadactylus (Bloch, 1795) is commonly ref erred to as the Lesser Af rican threadf in f ish of the Polynemidae f amily with major records in marine waters ranging f rom Morocco to Angola, and sparsely exist in the North and Southern regions of the Af rican Continent (Wehye and Amponsah, 2017). G. decadactylus is typically salt tolerant, and conglomerate at depths range of 10 m -70 m (Carpenter et al., 2015). Occurs over sandy and muddy bottoms in brackish habitats and estuaries (Daget and Njock, 1986).
The artisanal f ishery of Sierra Leone operates in estuaries and inshore waters extending f rom the shoreline to a depth of 20-40 m (MFMR, 2020). This is sector is eulogized as a signif icant source of f ish protein to the vast majority of Sierra Leoneans alongside its critical support to employment and rural income (Neiland et al., 2016;. The artisanal f isheries of Sierra Leone are characterized by varying types of canoes such as (a) small dug-out canoes (commonly called Kru canoes) of length ranging f rom 4-6 m and moulded depth of 0.6m; (b) the standard canoes (usually planked and of length ranging f rom 5-10 m) with a crew of 1-3 men, 3-5men or 5-10men (c) the Ghana-type canoe (either dug-out or planked) with a length of about 12 m (Ndomahina and Chaytor, 1991;Thorpe et al., 2009) as given in Figure 5.
The Lesser Af rican threadf in f ish is a commercially viable stock that contributes vastly to the diet of Sierra Leoneans, as well as to national economy. Besides, G. decadactylus accounted f or 48.5% (0-30 m), 10.9% (31-50 m), 37.1% (51-100 m) and 3.5% (101-200 m) depth zones of ten commercially important demersal f ish stocks in Sierra Leone (Seisay, 2014). However, despite its critical contribution to national f ood security through capture f rom the industrial sector, Galeoides decadactylus is one of most commonly exploited f ish stocks in the artisanal mixed f isheries of Sierra Leone which vulnerability to capture by the artisanal f ishery sector has yet to gain scientif ic attention f or analysis, more so when the said f ishery is notorious of its unscrupulous f ishing methods such as the use of illegal mesh sizes of f ishing nets that may subsequently results into growth overf ishing, and hence, dampen sustaianabily of the stock.
The present study aimed at evaluating the vulnerability of G. decadactylus to capture in the artisanal mixed f isheries. This aspect of analysis has hardly attracted the attention of academics in Sierra Leone, and the inf ormation provided could f oster the enhancement of existing management strides f or such a commonly exploited economically viable f ish stock of Sierra Leone, in concordance with postulates of other scholars (Asadollah et al., 2017;Konoyima and Seisay, 2020).

Study Site
Sierra Leone lies between latitudes 7°10′ N and longitudes 10°14′ W on the west coast of Af rica, covering an area of 71 740 km 2 (Coutin and Payne, 1989), and a continental shelf area of about 30 000 km 2 (Neiland et al., 2016). There are two distinct climate systems: the dry season (November-April) and the monsoonal rainy season (May-October, Coutin and Payne, 1989).
Specimens were collected at Funkia Wharf , Goderich community located in the f ar West End of the capital Freetown as indicated in Figure 1.

Data Collection
Monthly f ish samples were collected between January and December 2020 using randomized data collection techniques f rom f ive semiindustrial f ishing boats that mainly employed surf ace drif t nets and bottom drif t nets of varying mesh sizes ranging f rom 4 cm -8 cm. Fish specimens were identif ied to the least taxonomic level using identif ication guide by FAO (2010) whereby G. decadactylus has nine dorsal spines, 13-14 dorsal sof t rays , body moderately elongate and compressed, snout very short, blunt and prominent; mouth inf erior; posterior edge of maxillary only slightly expanded, barely reaching past eye; two widely separated dorsal f ins, second dorsal f in and anal f in bases barely equal; pectoral f in inserted low on body; scales ctenoid; head and unpaired f ins partly covered with small scales, body unif ormly silvery, upper part greyish and belly white; a large rounded black spot, about equal to eye diameter, generally visible under lateral line, behind opercle (Daget, 2003).
Specimens were then preserved in ice boxes and taken to the laboratory of the Institute of Marine Biology and Oceanography (IMBO), Fourah Bay College, University of Sierra Leone f or analysis.

Figure 1. Location of Data Collection Site
Morphometric measurements of f ish samples included total length (TL) and somatic weight. The total length was measured to the nearest 1cm using a 1m f ish measuring board graduated in centimeters, whereas the weight was measured using an electronic top-pan weight scale (ADAM-ACBPlus600H) to the nearest 1 g. Overall, a total of 865 specimens of G. decadactylus were collected using randomized sampling procedures, avoiding disproportionate representation of a particular size of the assessed species in line with recommended best practices (Pauly, 1980).

Statistical analysis
The present study utilized time series length f requency data grouped by constant class size. Analysis of growth parameters f ollowed methods f itted in the FAO-ICLARM Fish Stock Assessment sof tware (FiSAT II) f or PCs (Gayanilo et al., 2005).
The Instantaneous growth rate (K) was estimated by method of K-Scan in ELEFAN 1 as implemented in the FiSAT II routine, using L∞ as input parameter.

Estimating mortality parameters
The instantaneous total mortality rate (Z) was computed using method of length converted

Data Collection Site
Major Road catch curve as implemented in the FiSAT II routine (Gayanilo et al., 2005). The Pauly's empirical relationship was employed in the estimation of the instantaneous natural mortality rate (M) at a mean surf ace temperature (T) of 28°C thus: LogM= -0.0066 -0.279 1og L∞ + 0.6543 log K + 0.4634 log T………… (2), using L∞, mean surf ace temperature (T°C) and K as input parameters (Pauly, 1980).

Length-structured virtual population analysis (LVPA)
The LVPA employed the modif ied method by Jones and van Zalinge (1981). Input parameters were Lꝏ and K.

Estimating probability of capture
The length-converted catch curve f itted in the FISAT II routine (Pauly, 1984;Gayanilo et al., 2005) provides an extended methodology for estimating the probability of capture of selected length groups (L25, L50 and L75) using the f ollowing relationships: ….. (13) (Where L25 = length at which 25% of the f ish were vulnerable to capture; L50 = length at which 50% of the f ish entering the trawl net were retained by the gear (Pauly, 1984) was taken to be equivalent to mean length of the f ish at f irst capture (Lc50); L75 = length at which 75% of the f ish were vulnerable to capture by the f leet; S1 and S2 = variables used f or estimating the probability of capture under the logistic model).

Length-structured virtual population analysis
Results revealed minimum and maximum total length of 11cm and 36cm respectively f or the study species. The Length Structured Virtual Population Analysis (LVPA) revealed highest catch (in numbers) f or the length range of 20 cm (n = 94 x 10 4 ) -21 cm (n = 10 x 10 4 ) with corresponding f ishing mortality of 0.68 yr -1 and 0.62 yr -1 (Table 1; Figure 2), and the number of specimens that survived gill net capture steadily decreased with increase in length ( Figure 2). Besides, natural death of the population of the assessed stock decreased with increase in total length ( Figure 2). Table 1 shows f urther, the relative steady state biomass of the population of G. decadactylus drawn f rom the artisanal mixed f ishery, and the greatest biomass (22.02 t) was obtained f or the length of 21cm-22cmTL.

Probability of capture
The normal distribution plot (Figure 3) evinced greater selection probability (p 0.9) of the length range of 17 cm -22 cm (mean of 19.5 cm TL) in the mixed mesh sizes employed by the f ishers.

DISCUSSION
The length-structured virtual population analysis f rom this study depicted that the smaller length groups of the assessed stock (TL ≤ 11 cm) , to a larger extent, survived f ishing mortality amidst a seemingly exerted f ishing pressure probably owing to some unique manoeuvring or protection strategy. It could also be that the gill nets used by the f ishermen were highly selective to allow the young specimens escape net capture, though highly susceptible to natural causes of death, and suspected limiting f actors to surplus production could be linked to prolif eration of their predators and/or pollutants, f or which pollution debarring measures are strongly advised by the authors. Retrospectively, the most capture-vulnerable length groups of G. decadactylus obtained f rom this study (17 cm-22 cm) coincided with the length at f irst capture (Lc50 = 17.2 cm) and maturity (Lm50 = 22.7 cm) of the same stock recorded by Amponsah et al. (2021) f rom the Ghanaian Coastal Waters. However, the inf erence of growth overf ishing in G. decadactulus prof f ered by Amponsah et al. (2021) was in contravention to the present analysis that portrayed colossal likelihood of selecting mostly the larger-sized individuals of the assessed stock by local f ishing nets. Such discrepancy depicted clearly the regional dif f erences in stock status and management, f ishing methods and study sample sizes. Slightly lower Lc50 values (Lc50 = 13.9 cm and 15.4 cm) have also been observed by Wehy e and Amponsah (2017) and Sossoukpe et al. (2016) f rom the Liberian and Benin jurisdictional marine zones respectively. Growth overf ishing occurs when smaller length classes of a f ish stock become the dominant population in catches (Amponsah et al., 2016 Further, the probability of capture f rom this study portrayed an increase in capture vulnerability with increase in size of G. decadactylus, also suggesting high selectivity in the gill nets of local f ishers that allowed greater number (n = 4.8 x 10 6 ) of the smaller length group (TL ≤ 11 cm ) escape capture, inter alia. Low capture of the larger sized individuals (> 27 cm) could, however, also imply that these could easily f ight their way out of the gill net utilized by the artisanal mixed f isheries or could easily escape f ishing grounds. Concurrently, several authors have suggested that appropriate f ishing nets could allow escapement of young f ishes and enhance optimal spawning stock biomass of a f ishery (Mehanna et al., 2012;Ghanbarzadeh et al., 2015;Amponsah et al., 2016Amponsah et al., , 2017Wahye et al., 2017).

CONCLUSION
The study depicted that the smaller length groups of the assessed stock (TL ≤ 11 cm) were less vulnerable capture by local f ishing nets, owing to high selectivity in gill nets utilized by the artisanal f ishers f or the stock to allow greater number (n = 4.8 x 10 6 ) of the smaller length group escape capture averse to the larger sized (> 21cm) specimens. Notwithstanding, natural death was eminent f or the young individuals.
The authors however recommend stringent input and output control measures on the exploitation of the assessed stock to ensuring sustainability in the spawning stock biomass and optimal production. Ensuring good habitat quality is also essential in minimizing natural cause of death in the study species, and is strongly recommended.