Fishery managers face difficulties in making choices due to a scarcity of information on the state of fisheries stocks for virtually all the species caught worldwide [1,2]. Despite the importance of fisheries resources in terms of regional revenue and employment for coastal communities, adequate assessments of their state are hampered by a lack of data of sufficient quality and quantity [3,4]. For fisheries evaluation, information on growth and age is crucial, particularly for characterizing population structure, recruitment, and mortality [5].

The amount of energy expended to develop body weight and length is referred to as growth [6,7]. Growth studies have long been used to investigate the structure and dynamics of fish populations. Growth is also an indication of fish habitat quality in fisheries ecology [8]. Habitat factors have an impact on fish growth patterns [9-11]. Age and physiological condition [6]. Sex and seasons are among elements that might influence fish development [10,11].

Growth is a fundamental yet crucial biological activity that connects many processes and affects the life history of fishes. Other life history factors like natural mortality and fecundity can be directly connected to growth [5,13]. Growth parameters, for example, are required to construct age-structured catch curves and calculate natural and total mortality, which correspond to the negative components of stock dynamics. In population dynamics, the capacity to precisely forecast fish development offers a wide range of uses [13]. Individual growth, rather than recruitment, causes increases in the biomass of a (closed) stock.

Mathematical models that relate a fish species' size (usually length) to its age are required inputs to other models, such as those used in stock assessments, which are used to monitor a population and advise management choices such as harvest length limitations and other regulatory actions. Furthermore, growth data may be used to assess a fishery's state and predict how fisheries will respond to exploitation in the future [14]. Growth parameter estimations, for example, can be used to compare various populations (or stocks) across time to assess density-dependence or prey availability [15,16]. Fish growth varies by species and within species across gradients (e.g., latitude, temperature), and using mathematical models to depict growth enables for comparison [17,18].

There are several ways of modeling the growth of fish [19]. However, the Von Bertalanffy Growth Function (VB) has been by far the most studied and most used of all length-age models in fish biology because (1) it is based on bioenergetics principles; (2) it is useful in other fishery assessment models [20] and (3) of its empirical success in describing growth [21,22]. However, Lester et al [23] showed that the VB equation is rather a simple function of age at maturity and reproductive effort and provides a good fit of somatic growth only after maturity. Despite this, and since most of the studies use this model to describe growth in fish, we adopt this model in the present work for intra- and interspecific comparisons.

Nevertheless, it is becoming increasingly clear that new approaches, models, and model selection procedures may be necessary to account for growth variability caused by a variety of species-specific and environmental variables, as well as to prevent issues like sampling bias and incomplete data [18].

Small pelagic fish constitute approximately a quarter of all fish caught across the world, making them commercially and environmentally significant [24-26]. Among such fish species, the common horse mackerel, Trachurus trachurus [27] is a major commercial fish in the northeast Atlantic and Mediterranean Seas, with commercial landings of about 500,000 tonnes in the mid-1990s down to around 250,000 tonnes in the 2000–2002 period [28]. In Morocco, After the sardine Sardina pilchardus Walbaum, 1792, and the chub mackerel Scomber colias Houttuyn, 1782, the horse mackerel has an important place in landings of small pelagic fish; catches have been estimated at 52000 tonnes each year over the previous 10 years, with 83 % of those caught between Cape Boujdour and Cape Blanc [29]. The horse mackerel contributes mostly to the national production of small pelagic fisheries, although little is known about its growth patterns, which were investigated only in northern [30] and southern Morocco [31]. There are no simultaneous investigations on variation in growth parameters among the recently discriminated stock-units off the Moroccan Atlantic coastal waters [32].

Given the spatial changes in habitat factors due to variability of hydrodynamic and trophic conditions off the Moroccan Atlantic coast waters [33-35]. We hypothesized that growth parameters in T. trachurus would vary among the recently identified stock-units [36]. In addition, some estimated life history traits were taken into account as well, such as longevity and mortality. The obtained data were compared among the previously studied populations of T. trachurus, using an auximetric analysis for correcting the instantaneous growth rate (k) for differences in the asymptotic body length.

A total of 235 specimens of T. trachurus were obtained in summer 2018 (July –September) from commercial fishing vessels (Table 1) in three distant landing ports on the Moroccan Atlantic coast: Larache, Safi and Dakhla, from the north to the south (Figure 1). Intact fresh specimens were frozen soon after collection and defrosted for further laboratory analysis about one month later to ensure that all fish were analyzed after a similar period of being frozen.

Table 1: Summary of information for the samples of Moroccan Atlantic horse mackerels Trachurus trachurus used for growth parameters analysis; N indicate sample sizes.

Figure 1: The geographic locations of the sampling fishing areas (asteriks) for Trachurus trachurus, off the Morccan Atlantic coastal waters.  

All specimens were weighed (BW) and measured for total length (TL) to the nearest 0.1 mm and 0.1g, respectively. The gender was determined by macroscopic examination of gonads.

The power function W=aLb is used to illustrate the relationship between a fish's length and weight relation (LWR), where W is body weight (g), L is total length (cm), ‘a’ is the body shape coefficient, and ‘b’ is an exponent that implies isometric or allometric growth [37-39]. The least squares approach was used to derive the LWR parameters ‘a’ and ‘b’ from the log-transformation of the data, and the coefficient of determination (r2) was used to determine the degree of correlation between the weight-length variables. The 95 percent confidence limits of the parameters ‘a’ and ‘b’, as well as the statistical significance of r2, were calculated [40]. Values of the exponent ‘b’ provide data on the growth of fish. Weight gain is isometric when b=3. Weight gain is allometric when the value of ‘b’ differs from 3 (positive allometric if b>3, negative allometric if b<3). The null hypothesis of isometric growth (H0: b = 3) was tested using a t-test with the statistic ts = (b-3)/Sb, where Sb is the standard error of the slope ‘b’, for the threshold α =0.05 for testing significant differences among slopes (b) between two regressions for the same species [41].

Otolith pairs (sagittae) were extracted, and their ages were estimated by interpreting the growth rings. The broken burnt/method [42,43] was used to prepare otoliths for reading, and the ageing criteria were those described in [44]. The edges of opaque and translucent bets along the otolith's edge were inspected for all samples. The same person performed all the otolith readings.

Growth was expressed in terms of the von Bertalanffy equation [45] the parameters of which determined by the Beverton-Holt’s method [46] is one of the most frequent models used by fisheries biologists to investigate fish development and its interpretations, such as fish population dynamics and the impact of fishing laws on the catch.

L_t= L_∞ [1-exp ^{(-k (t-t0))}]

Where Lt is the total length of the fish at age t (years), L∞ (mm) is the asymptotic length that fish could achieve, k (year-1) is the growth constant rate, which determines how fast the fish approach L∞ and t0 is the hypothetical time at which the length of the fish is zero.

Longevity and Natural Mortality Rate:

The growth constant rate k has units of reciprocal time and is difficult to interpret [47]. It is then easier to interpret k in terms of half-lives (ln 2/k) with units of time. The time it takes to attain the fraction x of L∞ is as follows:

Fabens (1965) defined longevity based on x = 0.9933 assuming L0 = 0, and obtained:

t_0.99 = 5/k = 7.21*ln 2/k

Natural mortality of fish could occur due to predation, diseases, age and environmental factors. suggests a relationship between the natural mortality and water temperatures. An increase in water temperature will lead to the increase in natural mortality of fish. The natural mortality (M, yr-1) can be estimated using Pauly empirical equation [48].

Log M = -0.0066 – 0.279 log L_∞ + 0.6543 log k + 0.4634 log T

Where T is the mean surface temperature of the water (°C)

Hoeing (1983) developed a method to estimate population parameters such as total mortality (Z)

Ln Z = 1.46 - 1.01 ln t_max

Where t_max is the maximum observed age (years), in the considered sample as determined by otolithometry.

Pauly [48] stated that the total mortality rate is summation of natural mortality and fishing mortality (F) is then as follows:  

F = Z – M

Rate of exploitation (E) is the ratio of fishing mortality (F) and total mortality (Z) [48] as follows: 

E = F / Z

Gulland [49] states that the optimal exploitation for a fish stock occurs when fishing mortality (F) is proportional to the natural mortality:

F_optimum = M

Thus, the optimal rate of exploitation (E_optimum) is 0.5. Resource is considered suffering from overfishing if the rate of exploitation is greater than 0.5.

Intraspecific comparison

For the four growth parameters L∞, k, t0, and t0.99, we collected published data on other populations of the same species from the literature for an allometric analysis. In addition, for an objective comparison, we corrected the effect of body size (L∞) on the observed constant growth rate (k, yr-1) and Longevity (t0.99, yrs) as follows:

k/ L∞^b,   t_0.99/ L∞^b

Where ‘b' is the slope of the LWR for the whole T. trachurus of all studied populations including those in the present work; data are provided in table 4.

We also estimated the age at first sexual maturity of females (Tm, years) using the mean value of the body length range (20 – 25 cm) indicated by Letaconnoux [50] and Abaunza et al. [51] and the value of 22.75 cm recently reported for T. trachurus off the north Atlantic Moroccan coast [27]. The corresponding age at first maturity at a mean body length of 22.5 cm was derived from the Von Bertalanffy equations for the three sampling areas.

Length-Weight Relationships

The LWR was established for each of the three localities, was found to be isometric for Dakhla (b non-significantly different of 3, t-test, P > 0.05), and hypo-allometric for both Safi and Larache (b significantly lower than 3, t-test, P < 0.001).

Age-Length Keys

The otolith reading and the annual mark counting made it possible for each class of size to obtain a data couple; the number of marks of size corresponding to an age-length couple (years, mm), then to establish age-length keys for the three fishing areas.

The youngest horse mackerel sampled from Larache and Safi belongs to the age group 0+. The oldest are above the age of 5+ for the two areas. Most of individuals are aged 2+ and 3+ in the two areas: respectively 36% and 26% for Larache and 41% and 26% for Safi.

As for the Dakhla area, the age ranges from 0+ to 4+, with the most observed age classes being 2+ and 3+ with respective percentages of 28% and 43%.

Growth Parameters

The estimated age of specimens was used to determine the average size of fish for all the horse mackerel sampled in the study areas. We then adjusted the Von Bertalanffy equation to the couples of mean lengths at age values calculated for all individuals.

 The growth in the length curves of horse mackerel for each area shows that the growth is higher for the individuals from Safi compared to those from Larache and Dakhla. The asymptotic body length (L∞: infinite size, cm) is higher in the north (Larache) compared to center (Safi) and south (Dakhla) (52.37 vs. 36.23 and 37.58, respectively)

Intraspecific Comparison: Auximetric Analysis

Given that the growth rate constant k does not account for size differences among populations, an allometric analysis of this growth rate is required and will allow for finer inte-population comparisons by removing the effect of body size. The allometric relationship between growth rate k (yr-1), according to the VBERT model, and asymptotic body size L∞ (mm), derived from growth data of 41 populations, is statistically significant (r2 = 0.105; F1, 39=4.54; P = 0.034; n = 41) (Figure 2). The relationship obtained (k = 0.680*L∞-0.351±0.1645) corresponds to a hypo-allometry (b < 1) indicating a decrease in the growth rate in length with the asymptotic body size in the population considered. According to the coefficient of determination (r2) of this relationship, only 10.5% of the variation of k would be due to differences in body size and the rest (89.5%) related to other biological and ecological attributes of the populations compared.

Figure2: The allometric relationship between growth rate k (yr^-1), and asymptotic body size L_∞ (mm), according to the VBERT model, derived from growth data of 41 populations of Trachurus trachurus.

The couples of both [log10 k and log10 L∞] and [log10 t0.99 and log10 L∞], are largely dispersed along their corresponding regression lines with slopes of ±0.351 (S.E. = 0.1645) intercepts of -0.167 (S.E. = 0.2553) and 0.866 (S.E. = 0.1552), respectively, which explains a low (but significant) percentage of the total variance of the data (r2 = 0.105, N= 38, p < 0.05) (Figure 2). The size-corrected growth constant rate (k/L∞b) obtained in this study, are higher in Safi area (0.077) than in the Dakhla (0.045) and Larache (0.036). The corresponding size-corrected longevities (t0.99/ L∞b) were similar in Larache and Dakhla (8.3 yrs), and lower in lower in Safi (5.1 yrs) (Table 2).

Table 2: Horse mackerel’s growth parameters from Von Bertalanffy Growth Equation. With along with parameters of longevity, and the estimated age at first sexual maturity, corrected for asymptotic body size.

Age of Females at Sexual Maturity

The estimated age at first sexual maturity in females (Tm, yrs), for a body size range of 20–25 cm as explained above, is of 7.8 for Dakhla, which is higher than the similar values close to 5 for both Larache and Safi.

Longevity (t_0.99) and Mortality

Calculated longevity (t0.99) was 33.1, 17.9 and 30.1 years, respectively for Larache, Safi and Dakhla. The values corrected for body size (L∞), are 8.4 years for both Larache and Dakhla, and 5.1 years for Safi.

The estimated instantaneous natural mortality rate (M, yr-1) was lower (0.22) for Safi, intermediate for Dakhla (0.35) and higher for Larache (0.42). The instantaneous total mortality (Z, yr-1) rate was 0.85 for Larache and Safi and 1.06 for Dakhla. The corresponding fishing mortality rates (F, yr-1) were higher in 0.71 and 0.63, respectively for Dakhla and Safi and lower in Larache (0.43). Stocks revealed to be overexploited in Safi and Dakhla (E=0.74 and 0.66, respectively) whereas Larache stock is rather its optimal rate of exploitation (E=0.50) (Table 3).

Table 3: Atlantic horse mackerel’s parameter values for the instantaneous total mortality (Z, yr-1), naturel mortality M (yr-1) fishing mortality rates (F, yr-1), and the rate of exploitation (E), off the three fishing areas off the Moroccan Atlantic coastal waters.

According to Lester  et al’s analysis [23] von Bertalanffy (VB) model, commonly used in fish growth studies, revealed to provide a good description of somatic growth after maturation, but not before. These authors showed that the parameters of the VB equation are simple functions of age at maturity and reproductive effort. They assumed these two life-history traits adjust to changes in adult mortality to maximize lifetime offspring production, so that the model predicts that: (i) the optimal age at maturity is reversely correlated with the adult mortality rate; and (ii) the optimal reproductive investment is almost equivalent to the adult mortality rate. The age at maturity in T. trachurus is not known for the Moroccan Atlantic populations, but only the size at sexual maturity (22.75 cm) was recently reported for north Moroccan Atlantic and data on maximal gonado-somatic index in females (4.2%: in 2015-2016) as an indicator of reproductive effort was obtained for Larache [27]. This value is lower than those reported for the Portuguese Atlantic populations (6-9%: 1978-1990) and North Sea and English Channel (13%: 1968-1969) [42]. The individuals, which are more than two-year-old reach the time at their first sexual maturity from a body length of 19 cm. Between 20 and 25 cm, the growth rate undergoes important changes. This period is reflected in significant variations in both weight and linear growth [50]. For our three studied populations and using an average size at maturity of 22.5 cm, the calculated age at maturity was about 8 years for Dakhla, and 5 years for Safi and Larache. For this latter, and using the reverse VB equation, data on VB growth parameters for females [52] and those on the size at maturity [27]. The age at maturity was estimated to 6.9 years. This value is slightly higher than that obtained in the present study for both Larache and Safi (5 years). On the other hand, a delayed age at maturity around 8 years was found for Dakhla. However, all these values are higher than those reported for European Atlantic populations (1-2 years). Such differences may be attributed to those in size-corrected growth rate and asymptotic body length, which in turn, can reflect variations in ecological conditions (hydrodynamism and climate, food availability, population dynamics…) among habitats.  The largest asymptotic body size in Larache [52] would explain the slower somatic growth, but higher longevity, in this population. However, further investigations are needed to confirm this, particularly for spatial variation in reproductive traits and food availability.

As stated by Kerkich et al [52] and given the difficulty of sampling of both smaller and larger horse mackerel individuals, the computed values of VB parameters are questionable. This would affect the estimates of t0 and L∞, and consequently of k [53]. These parameters of growth only apply to the length range sampled [42] and not the entire population sample. Variations in growth patterns in the same species can also be explained by biotic (abundance and quantity and / or quality of the food, age, maturity) and abiotic (sampling method, climate factors, technique of age determination).

Apart from the problems of sampling, age determination and the methods used for estimating the VB parameters, we can however make relative inter-population comparisons based on a similar growth model.

The only data available on VB growth parameters for the Moroccan Atlantic horse mackerel populations off the coastal Atlantic waters, are those of Kerkich et al [52] in Larache fishing area. These authors reported a k value (0.14 yr^-1) close to that in the present study for the same sampling area (0.15 yr^-1), but with a lower L∞ (44.67 vs. 53.37 cm). In a close NW Mediterranean population in the bay of M’diq, Morocco, Kerkich et al [30] obtained a lower k (0.10 yr^-1) with a similar L∞ (43.9 cm) as in Larache. In south Portugal, Arruda recorded values of 0.132 and 41.05, which are lower than those obtained in this study, but quite similar to those of Kerkich et al. [52] for Larache. In Mauritanian populations BA et al., (1988) reported k and L∞ values respectively of 0.181-0.241 and 41.39-41.76, which are higher than those we obtained for Dakhla population (0.166 and 37.58, respectively). The difference in L∞ values could be attributed to sampling bias, age determination and/or body variation in body size and environmental conditions among seasons/years.

Table 4: Data on growth parameters and longevity (absolute and body size corrected values for 41 studied populations of Trachurus trachurus. Sampling period, fishing area, growth model and aging method are indicated.

Populations were in the range of the corresponding values in other populations (0.025 – 0.128). Safi population showed the highest value compared to Larache and Dakhla (0.077 vs. 0.036 and 0.045, respectively). The body size-corrected longevities (t0.99, years), were 5 for Safi and about 8 for both Dakhla and Larache. For this latter, the obtained value is close to that calculated from data of kerkich et al [52] which was about 9, lower than that from M’Diq (13), Moroccan western Mediterranean [30]. The corresponding values for European Atlantic and Mediterranean populations ranges from 6 to 11 and from 4 to 13, respectively (see table 4). The value for Dakhla is slightly higher than that calculated for Mauritania (about 7).

The body size corrected growth rate values (k’) for the three studied populations, are within the range of all the studied populations in both Mediterranean and Atlantic coastal waters (0.027 to 0.128 yr-1; see table 3) with the extreme values increasing from western to eastern Mediterranean populations (0.027 to 0.128) and from northern to southern NE Atlantic area (0.031 to 0.089). However, Safi exhibits a usual k’ of 0.077 while Dakhla has a lower value of 0.045, which is much lower than the value from Maxim (1995) (0.075) and Mauritanian stock unit (Ba et al., 1988) (0.063 and 0.047, respectively in 1987 and 1988).

Using scales and length frequency analysis to estimate growth parameters, the values obtained for k are in the range of 0.13 to 0.33 [54,55]. It is also strange to assume estimates of L∞ to be greater than 50 cm [54,56-58]. When fish larger than 44 cm are rarely reported from commercial and research vessel catches (ICES, 2002a). On the other hand, negative values of t0 lower than –2 may indicate problems in the sampling process, usually in inadequate sampling of the very young specimens[54,57,59]. This is the case of the Dakhla sample with t0 = -2.26, and in which all the specimens were approximately within a same narrow size range.

The disparity in growth rate, hence in longevity, could be attributed to the changes in environmental factors, length, age, sex, and gonad development, as well as distinct phases of ontogenetic development. Food size, quantity, and quality, as well as water temperature, are all directly related to a population's growth factors [60].

Variations in growth could be related to differences in habitat, temperature, and even feeding habits [61]. Geographic location and several environmental factors, including water temperature, organic matter, food quality, catch date and time, stomach fullness, disease, and parasite infection, can all influence weight-at-age estimations [62].

The range values in our samples are different from the relevant literature probably due to dif¬ferences in ontogenetic development, body condition, sex, maturity as well as variations in geographic locations, seasonality, and small sample sizes.

Indeed, the Larache population showed the lowest fishing mortality rate (0.43 vs. 0.63 and 0.71, respectively for Dakhla and Safi). On the other hand, the high natural mortality (M) is the highest among the three studied populations. This might be due to the impacts of several factors (genetic and/or environmental conditions).

Fishing mortality rates for T. trachurus off the Moroccan Atlantic coastal waters were significantly larger than the rates of natural mortality in Dakhla (0.71 vs. 0.35) and Safi (0.63 vs. 0.22). This indicated that the mortality of horse mackerels in these two stock-units was largely caused by fishing activities. The low natural mortality and high fishing mortality rates may indicate the occurrence of growth overfishing, in which more young fish were captured than old fish [63]. The high fishing mortality of the species might be related to the increase in fishing activities in these areas.

Therefore, controlling and surveillance on fishing activities including restriction on the minimum fish size of the horse mackerels. On the hand, the rate of exploitation in Larache was of 0.50 indicating that the exploitation was at its optimal level for 2018 summer fishing season. In contrast, during the longer period of June 2005 to November 2006, this stock unit was globally rather overfished (E=0.67 with M=0.14 and F=0.26) [30]. Suggests the optimum rate of exploitation of a resource is 0.5. Therefore, precautionary management approaches are necessary by controlling and restricting the number of fishing fleets targeting horse mackerels to maintain the sustainability of the fish stock in this fishing area [49].

The difference in the exploitation rate of horse mackerel in the Larache fishing area between the years 2005-2006 [30] and 2018 (this study) would be related to a substantial decrease in fishing effort (50-60 vs. 20-30 boats/year) and catches (about 1700 vs. 333 tonnes), in relation with a drastic decline between 2006 and 2019 in the fishing effort targeted by purse seiners as well as a strong decrease in the number of active trawlers targeting horse mackerel in the port of Larache [64]. It appears that the body size-corrected growth rate increases directly with the exploitation rate (0.036 vs. 0.50; 0.077 vs. 0.74, and 0.045 vs. 0.66, respectively for Larache, Safi and Dakhla), which is reversely related to the body-size corrected longevity (t0.99); the corresponding respective pairs of values are 8.3 vs. 0.50; 5.1 vs. 0.74 and 8.4 vs. 0.66).

The obtained results indicate the existence of difference in the growth patterns among the three studied T. trachurus stock-units off the Moroccan Atlantic waters. In fish, variations in life history traits, including somatic growth, may constitute an adaptive strategy to change in reaction to variations of the environmental/genetic or physiological conditions [65,22]. As well as to a south-to-north gradient [66,67]. The fish growth is affected by numerous environmental factors, namely food as the driving force, and temperature and body size rate-controlling and scaling factors, respectively [68]. In the NE Atlantic Ocean, studies dealing with the relationship between T. trachurus growth rate and habitat conditions are rather rare, but several comparative investigations of populations of other pelagic species revealed that a high proportion of the variance of growth parameters among fishing areas was imputed to changes in the temperature and chlorophyll [69,42].

References