Trace Elements in Mollusks, Crustaceans and Fish Commonly Consumed by the Catfish Chrysichthys nigrodigitatus Lacépède, 1803 from the Lake Togo-Lagoon of Aného Hydrosystem (Southern Togo)

The omnivorous fish species, C. nigrodigitatus feeds mainly on benthic organisms and may therefore consume contaminated food throughout its food web. This can lead to the bioaccumulation of contaminants such as trace elements in their tissues. However, fish consumption is a major pathway of human exposure to contaminants which may cause public health problems. The aim of the present study is to assess trace elements contamination in some species from the food web of C. nigrodigitatus. For this, 10 main food items of the silver catfish were collected at two sites from February to July 2017 and analyzed using an Atomic Absorption Spectrometer coupled with a hydride and cold vapour generator. The concentrations of trace elements varied greatly from one species to another and within each species. These values ranged from 0.007 mg/kg


Introduction
Most parts of terrestrial and aquatic ecosystems are now affected in one way or another by anthropogenic activities such as the rapid industrialization and demographic pressure during the last decades which lead to continental and aquatic ecosystems contamination by trace elements. Indeed, in aquatic environments, significant amounts of trace elements are introduced by industries, mining, fossil fuel combustion, run-off from agricultural lands, house hold sewages, atmospheric deposition and rocks weathering. These metals pose high environmental risks due to their longtime persistence in nature and possible bioaccumulation and biomagnification [1]- [6]. These inputs have greatly altered the biogeochemical cycles of trace metals and enhanced their bioavailability [7]. Consequently, it follows permanent disturbances in aquatic ecosystems leading to environmental and ecological degradation and which constitute a potential risk to a number of flora and fauna species, including human through food web [8] [9].
Aquatic organisms have been reported to contain higher concentrations of trace element in their tissues compared to the levels in the surrounding environment [10] [11]. Coastal aquatic ecosystems are of considerable ecological and socio-economic importance [12]. They are the habitats and nurseries for many larval and juvenile stages of fish species [13] [14]. Thus, trace element contamination of these ecosystems includes that of the fish food web species presenting a serious public health issue because these fish are finally consumed by humans.
The Silver Catfish C. nigrodigitatus from the Lake Togo-Lagoon of Aného hydrosystem is highly appreciated as protein source by local populations and contributes to their socio-economic well-being [15] [16]. It is well established that this hydrosystem and its basin includes the most contaminated coastal areas in Togo. Several workers reported the presence of harmful pollutants in waters, soils, sediments, biota and vegetables from the basin mainly due to phosphorites mining in the region [1] [4] [5] [17] [18] [19] [20] [21]. The Silver catfish which is known to be omnivorous feeding mainly on benthic organisms may therefore consume contaminated food throughout its food web. This can lead to the bioaccumulation of contaminants such as trace elements in its tissues. Knowing that fish consumption is the major pathway of human exposure to contaminants [22], human health may be highly threatened. The aim of the present study is to assess trace element contamination levels in some species from the food items of C. nigrodigitatus.
with fishermen from the lagoon complex [25]. The samples were wrapped in batches of species in sterile polyethylene bags and placed in coolers in the presence of a refrigerating equipment and then transported to the laboratory where they were stored at −20˚C [26]. Due to their small size, five batches of composite samples of 4 to 10 individuals for each species were carried out according to Pascal et al. [27] (the smallest making at least 75% of the largest). After identification, the samples were dried at 65˚C in an oven, finely ground in an agate mortar. The grinding equipment has always been cleaned before and after each sample. These samples were then digested using a mixture of reagents composed of 30% hydrogen peroxide (H 2 O 2 ) and 67% nitric acid (HNO 3 ) in the proportions of 1 H 2 O 2 : 3 HNO 3 at 90˚C on a hot plate [28] [29] [30] [31] [32]. For the determination of mercury, the samples were digested at room temperature for 72 -96 hours while stirring them regularly in order to allow good attack by the reagents. Simultaneously, the blanks were prepared and processed under the same conditions as the two series of samples. Then, each solution from the digestion was filtered, completed to 20 ml with distilled water and stored at room temperature. Trace elements were determined in these solutions, by atomic absorption spectrometer (AAS) with flame (Thermo Electron S. Series type), for Cd, Pb, Cr, Ni, Cu, Zn, Mn and by the same AAS coupled to a hydride and cold vapour generator (Thermo Scientific VP100 type), with flame for As and without flame for Hg. The reagents used for this purpose are analytical grade from Sigma-Aldrich for H 2 O 2 and HNO 3 and from SCPScience for trace element standards.

Accuracy and Quality Control
The quality of the analytical methods has been verified by internal control. A procedural blank was prepared with the same reagents and the same experimental conditions as the main samples. The blank allowed zeroing the device and was analyzed after each 10 samples batch during the analysis. This allowed to determine possible contaminations and eliminate the quantization errors. The standard solutions prepared for each trace element were also analyzed at regular intervals in order to verify the accuracy of the results. In addition, the repeatability of the results was checked by the analysis of duplicates which were randomly incorporated among the samples.

Bioconcentration Factors of Trace Elements
In order to assess the level of transfer of trace elements from the medium to the organism, bioconcentration factors (BCF) were calculated in relation to both water and sediment. These BCF are expressed according to the following equation [3] [10] [33] [34]. Trace element concentrations in waters and sediments (Table 1) used for the calculations were from simultaneous studies in the same hydrosystem [4] [5].

Statistical Analysis
The analysis of variances (ANOVA) followed by the Newman-Keuls test made it possible to evaluate the interspecies variations in trace element contamination of the prey species of C. nigrodigitatus [3] [10] [31]. Principal Component Analysis (PCA) was carried out to assess the typology of trace element contamination [35] [36] [37]. Pearson's correlation analysis demonstrated the links between trace elements [34]. These analyses were performed using the STATISTICA 6.1 software.    Figure 2 indicates that the trace element contents are unevenly distributed in the prey species. These interspecific variations in trace element levels are confirmed by analysis of variances (ANOVA) which were found to be significant at the 5% level for all elements (Table 3). However, this ANOVA is followed by the Newman-Keuls test which revealed several significant differences between the prey species considered in pairs and allowed them to be classified into homogeneous groups according to their levels of trace element accumulation. Thus, species bearing the same letter indices form a homogeneous group (Table 3).

Interspecific Variations in Trace Element Concentrations in Fish's Preys
In accordance with the trace element contents, the decreasing order of contamination of the prey species for each trace element is presented as indicated in Table 4. It is observed that the lowest levels are found most often in the fish spe-     Figure 3 shows that the lowest element accumulated in all species is Hg. Its levels vary from 0.007 mg/kg observed in C hippos to 0.146 mg/kg obtained in M. perna. The highest levels of trace elements recorded are those of Cu in F. notialis, of Zn in the species Eucinostomus melanopterus, E. fimbriata, C. hippos, M. perna and T. zillii and of Mn in C. amnicola, P. fusca, Pagurus sp. and G. paradoxa. It therefore emerges that there is a variation in the levels of accumulation of trace elements in all the species studied. The order of accumulation of trace elements in each species was therefore established and presented in Table 5.

Bioconcentration Factors Relative to Water (BCFw)
With the exception of Pb in the species F. notialis (BCFw = 0.90) and E. melanopterus (BCFw = 0.99), the other bioconcentration factors (BCFw) obtained are all greater than 1 and vary from 1, 26 for Cr in T. zillii to 12342.19 for Mn in P. fusca ( Table 6). The species which have accumulated more trace elements are those that lead a benthic life in direct contact with the sediments and are sedentary. These are the species P. fusca, M. perna, Pagurus sp., G. paradoxa. The least accumulators are exclusively composed of fish species (E. melanopterus, E. fimbriata, C. hippos, T. zillii).

Bioconcentration Factors Relative to Sediments (BCFsed)
The bioconcentration factors of trace elements relative to sediment (BCFsed) are presented in Table 7. They vary from 1.02 for Zn in C. hippos to 13.40 for Cu in Pagurus sp. Zn accumulation was observed in 80% of species with BCFsed between 1.02 obtained in C. hippos and 3.18 found in Pagurus sp. In addition, 60% of the species exhibited Cu BCFsed ranging from 2.48 in M. perna to 13.40 in Pagurus sp. Cd is accumulated by 50% of species with BCFsed varying from 1.63 in Pagurus sp. to 4.90 obtained in M. perna. Hg BCFsed greater than 1 were only found in M. perna (BCFsed = 3.33) and G. paradoxa (BCFsed = 1.84). Accumulation of Ni was observed only in Pagurus sp. with a BCFsed = 1.14.

Correlation Matrix between Trace Elements
Pearson's correlations between trace elements in C. nigrodigitatus prey are shown in Table 8. It indicates that all significant correlations between trace elements are positive. Thus, Cd is significantly correlated with Pb, Cr, Hg, As, Zn and Mn.
As for Pb, it is significantly correlated with Cr, Hg, As and Zn. In addition,   significant correlations were obtained between Cr and other trace elements except Cd and Pb. Significant correlations were recorded between Ni and each of Cu, As, Zn and Mn. Cu shows a good correlation with Zn and Mn. Also, significant correlations were obtained between Hg and the elements As, Zn and between As and Zn. Table 9 shows that the first 3 components explain 88.31% of the total variance with F1: 52.95%; F2: 24.97%; F3: 10.39%. However, the two components (F1 × F2) alone explain 77.92% of the total variance. Thus, this map can explain most of the information contained in the data regarding the distribution of trace elements in the different prey species of C. nigrodigitatus.

Principal Components Analysis of Trace Element in C. nigrodigitatus' Preys
The component F1 (52.95%) is determined in its negative part by the elements Cd, Pb, Cr, Ni, Hg, As and Zn with correlation coefficients presented in Table 8. Component F1 therefore indicates, from right to left, a contamination gradient in trace elements (Cd, Pb, Cr, Ni, Hg, As and Zn). The component F2 (24.97%) is determined in its negative part by the Cu (r = −0.77) and the Mn (r = −0.51) indicating, from top to bottom, a contamination gradient in Cu and Mn ( Figure  4(a)).
Four groups of species can be distinguished in Figure 4(b). The first group (G1) is essentially composed of fish (E. melanopterus, E. fimbriata, C. hippos and T. zillii) and shrimps (F. notialis). This group is characterized by the lowest levels of trace elements with Ni and Cu concentrations obtained in F. notialis which are much higher than those recorded in fish. The second group (G2) includes crabs (C. amnicola) and gastropods (P. fusca) which are characterized by fairly high levels of trace elements. However, the average Ni content of P. fusca is higher than that of C. amnicola. As for Cu contents, they are rather higher in C. amnicola than in P. fusca. The third (G3) and fourth (G4) groups have the high-Journal of Environmental Protection est levels of trace elements. The G3 group is formed from Pagurus sp. and presents the highest contents of Cr, Zn, Ni and Cu while group (G4) consists of M. perna and G. paradoxa characterized by the highest contents of Cd, Pb, Hg and As.

Discussion
The preys of C. nigrodigitatus in the Lake Togo-Lagoon of Aného hydrosystem are diverse and includes fish, gastropods, crustaceans, bivalves, plants etc. [24]. However, it is known that the accumulation of trace elements in the tissues of aquatic organisms, such as fish, also depends on their feeding behavior [10] [34] [38]. Thus, a few individuals from each taxon were assessed for their metallic contents. In fish species, the average concentrations recorded are higher than the quality standards of fishery products for Cd, Pb,   benthopelagic. This strong accumulation of trace elements in molluscs and crustaceans compared to fish has also been observed by Ali and Fishar [44] in Lake Qarun (Egypte), by Ouro-Sama et al. [45] in the Togolease lagoon system and by Zhang et al. [34] in marine environment in China. The difference in trace element accumulation between fish and benthic macroinvertebrates was demonstrated by principal component analysis (PCA  [52]. The high accumulation of all trace elements is due to the fact that mollusks have a poor ability to discriminate between elements which are similar in certain characteristics such as the valence of ions [53]. However, these mollusks have effective detoxification mechanisms that reduce the toxicity of the trace elements absorbed despite their high concentrations [44]. The intraspecific variations of the recorded element contents can be explained by the fact that the accumulation of trace elements in the tissues of aquatic organisms depends not only on the species but also on the chemical nature of the element concerned (molecular size, speciation chemical, bioavailability, etc.) and the physicochemical characteristics of the medium (temperature and pH, salinity) [33]. Indeed, whatever the route of entry of pollutants into the body, the intensity of absorption is extremely variable from one pollutant to another. This has been h attributed to the characteristics of the membrane barriers in contact  [55].
Results also indicate that bioconcentration factors relative to water (BCFw) are significantly higher than those relative to sediments (BCFsed). It can be inferred that, trace elements accumulated in the tissues of these aquatic organisms come mainly from water. These results corroborate those of other authors who believe that the levels of trace elements in tissues are largely influenced by their concentrations in water [3] [10] [56]. Overall, the BCF values were higher for Cu, Zn As and Mn. These high accumulations may be due to physiological needs since these elements are essential for the unfolding of biological processes [34] [57] [58] [59]. The significant correlations recorded between the trace elements indicate that these elements come from the same source and that their absorption, distributions and accumulations in each individual would respect the same physicochemical and biological processes [3] [60]. The correlations between the different trace elements were confirmed by the results of the PCA thus indicating their common accumulation processes.
Contamination of these species is a threat to their consumers in general and to C. nigrodigitatus in particular. Thus, the consumption of these species can contribute to the accumulation of trace elements in their tissues leading to problems of toxicity and survival of the species and a perturbation of the ecological balance [61]. Furthermore, humans at the end of the food chain cannot be spared the toxic effects of these trace elements.

Conclusion
It emerges from this study that most of the species overall present concentra- tion of preferred prey by catfish exposes the species and humans being to contamination by trace elements following its consumption. It is therefore imperative to pay particular attention to this ecosystem and to put in place a better management plan.