Determination of water quality, trace elements contamination in Oreochomis niloticus and Clarias gariepinus reared in 2 types of ponds and health risk assessment in Cameroon

Background: Fish farming in Cameroon is growing very rapidly and sh available in the markets is mostly coming from sh farms, but domestic sh production is still low to meet demand. Intensication of production could lead to the occurrence of various types of contaminants that can affect the sanitary quality of farmed sh and consumer health. This study aimed to assess the quality of farmed sh collected in sh farms located in four regions of Cameroon (Center, South, West, Littoral). Results: Concentrations of arsenic, cadmium, lead, chromium, cobalt, and nickel were determined in pond water and muscles of Oreochromis niloticus and Clarias gariepinus raised in concrete and earthen ponds to evaluate health risk for consumers. Trace elements were determined using Inductively Coupled Plasma Optical Emission spectrometry (ICP-OES). Heavy metals in water for concrete and earthen ponds shows there are no signicant differences (p<0.05) between the rearing systems except for cobalt and nickel. Heavy metals concentrations in water were lower when compared to the WHO recommended limits except for cadmium in all the ponds. For concrete ponds, heavy metal concentrations decreased in water in the sequence of Cd > As > Co > Ni > Cr > Pb. For earthen ponds the concentrations decreased as Cd > Co > As > Ni > Cr > Pb. Conclusion: Concentrations recorded in the muscles of the 2 species were above the safety limits recommended by FAO/WHO. The target hazard quotient of As was highest compared to other metals and higher than the acceptable limits according to US Environmental Protection Agency guidelines in all sh species. the level of each investigated metal, highlighted a very low health risk for consumers. and them Furthermore, and were all present, in some cases at higher levels than The levels of heavy metals concentrations in concrete sh ponds and earthen were not signicantly different when compared the two species except for Ni in Nile tilapia and cobalt for African catsh. Concentrations of Arsenic in the esh of Nile tilapia and African catsh do not varies so. These values were above the maximum limits of FAO/WHO (0.2 µg/g). Cadmium in all sh samples were the maximum limits set by the commission of 0.05 mg/kg. they were no signicant differences at p<0.05. For Lead, Pb content in all samples from earthen ponds was below the maximum limits (0.3mg/kg), for concrete ponds the mean values slightly above the limit. The permissible limit for cobalt set by EU is 0.04 mg/kg, Co concentrations the were above the limits. The values do not differ for each rearing system, concrete (0.37 µg/g) earthen µg/g) in the esh of Nile tilapia but for African catsh signicantly


Introduction
During recent decades the global consumption of sh has increased due to growing population and rapid urbanization (Wim et al. 2007). Furthermore, sh is a major part of the human diet because it provides a healthy, low cholesterol source of protein and other nutrients including omega 3 fatty acid that reduce cholesterol levels and the incidence of heart disease (Ruxton et al. 2004 ). It has been predicted that sh consumption in developing countries will increase by 57 percent, from 62.7 million metric tons in 1997 to 98.6 million in 2020 (Delgado et al. 2003). Over the last 20 years, stocks of many sh species are in decline attributed to illegal and over-shing. Therefore sh farming appears to be a great solution to the massive demand for animal protein. Many production system exits for sh farming, in Cameroon it's mainly earthen ponds and concrete ponds that are use to cultivate shes. It's well known that aquatic organisms are characterized by accumulation of many harmful elements in the body amongst them heavy metals. The water used for sh farming contains very high concentrations of heavy metals which can affect adversely the quality of the nal product cultivated sh and consumer health (Staykov 2001). The accumulation of heavy metals in sh mainly depends on their concentrations in food and water, in this aspect the production system has an essential role. According to many authors high concentrations of heavy metals in water may have devastating effect on the ecological balance in the environment and in the diversity of aquatic organisms (Farombi et al. 2007). The consequence of this is a large loss of sh production on sh farms.
The presence of contaminating heavy metals in freshwater ponds at concentrations above natural loads has become a growing concern. Recent studies have investigated for instance sh contamination by lead, cadmium and mercury which are very toxic (Zaza et al. 2015;Zheng Zhang et al. 2007;Stancheva et al. 2013).Various studies in different countries reported the contamination of different farmed sh species with a different type of pollutants, trace metals are of particular concern, due to their potential toxic effect and ability to bioaccumulate in aquatic ecosystems (Vutukuru 2005;Dirilgen 2001;Voegborlo et al. 1999;Canli et al. 1998). In many developing countries, metals pollution problems in the sh farming sites were serious as re ected by high metal concentration recorded in Malawi (Puchase and Jamu 2009), Zimbabwe (Teta, Ncube and Naik 2017). Other studies conducted in india and Nigeria (Nath and Bhoumik 2013;Ibemenuga et al. 2013;Kumari et al. 2017) indicated high level of cadmium, arsenic and lead in water and sh tissues. Some regulatory agencies have categorized them as pollutants of high importance (US Enviromental Protection Agency 2006). Trace metals occur naturally in the environment and are present in low concentrations in freshwater (VanLoon and Duffy 2011). Rapid population growth, increased urbanization, the expansion of industrial activities, the exploration and exploitation of natural resources, the extension of irrigation and the spread of modern agricultural practices should be blamed for the occurrence of heavy metals. Increases use of fertilizers, which contain metals could also result in high concentration of trace metal in sh ponds. Fishes have been widely used as bio-indicators of pollution by metals in aquatic system because they occupy different trophic levels (Burger et al. 2002).
Cameroon like many other countries, is facing severe problem of freshwater pollution as almost 99% of the industrial waste water is discharged into streams and rivers without any treatment (Khan et al. 2012) which is undesirably distressing natural resources of the country (Rauf et al. 2009). Domestic sh production is still low to meet demand. Fish farming appears as a solution to ll this gap. These last years there is a tendency to increase the number of sh farm and intensify the production to ll the gap.
Intensi cation of production could lead to the occurrence of various types of contaminants that can affect the quality of farmed sh and consumer health. Few studies exist concerning sh ponds contamination by heavy metals in Cameroon. The studies have mainly concerned wild sh species commonly found in rivers and lake. For instance, Fonge et al. 2011 assessed the contamination of Dibamba river and Arius heudelottii by Co, Cd, Pb, Fe, Cu. Another study reported the contamination of heavy metals (Cr, Cd, Hg, Pb and Zn) in sh, mud and water from two urban lakes in Yaoundé (Demanou and Brumett 2010). Nkwelle et al. (2012) works on the determination of water quality, and trace metals in endemic Sarotherodon linellii, Pungu maclareni and Clarias maclareni, in crater lake Barombi Mbo in the south west region of Cameroon. There is no report on water quality in shponds and heavy metals contamination in farmed species; furthermore no data exist on the potential health risk related to the consumption of contaminated shes with respect to the rearing system. There is a need to investigate on the in uence of intensi cation of production system on farmed sh in order to limit the risks and to improve the quality of water and nal product. O. niloticus and C. gariepinus are amongst the most common shes reared and consumed in Cameroon. Therefore metal accumulation in these sh species presents serious public health. The fact that metals accumulate in the environment necessitates their continuous monitoring and assessment in respect of both ecological and human health impact. The purpose of the present study was to determine the level of heavy metal contamination (As, Cd, Cr, Pb, Co, Ni) in water, and muscle tissues of O. niloticus and C.gariepinus reared in earthen and concrete ponds to assess the ecological and public health risks and to identify possible sources of the metals for future remedial action. Additionally, to compare the two systems with regards to heavy metals contamination and health risks.

Study area
The study was conducted in four regions of Cameroon: the center, the south, the littoral and the west regions as shown in Fig. 1. The different regions were selected as they represent a high potential for sh farming due to favourable resources and climatic conditions. The center region covers 69.000 km 2 and is composed of rolling hills on a vast plain with a mean altitude of 700-800 m, with lowered mounds. The climate has two wet seasons. The population density is low, with about 36 inhabitants/km 2 (BUCREP 2010;NIS 2006). The south region covers an area of 47.110 km 2 , with a population of about 534.900 inhabitants and a density of 13.4 inhabitants per km 2 (NIS 2006). The Littoral region is covering an area of 20.239 km 2 and housing more than 2.202.340 inhabitants. The west region covers 13.872 km 2 and is mountainous, marked by highlands with a mean altitude of 1600 m and narrow valleys with catchments separating them. The climate has a unimodal wet season. The population density is relatively high, with about 143 inhabitants/km 2 (BUCREP 2010). The population density is 124 inhabitants per km 2 .

Sample collection
The samples were collected between April and August 2017, a total of 70 sh samples (35 samples of O. niloticus and 35 of C.gariepinus) and 35 water samples were randomly collected in 14 ponds respectively 8 earthen ponds (plate 1) and 6 concrete ponds (plate 2). The pond sizes varies between 200 and 500 m 2 , each pond was sampled twice directly from sh farmers based in four regions of Cameroon as mentioned in table 1. These regions were selected according to their importance in farmed sh production in the country. They have the largest proportion of sh farms in Cameroon.

Water
Triplicate 0.5 litre water samples from concrete and earthen were collected from sh farmers in each region, in different locations and depths of the ponds to obtain both representative and reproducible samples. Low-density plastic bottles were rinsed thrice with ambient water prior to sampling. One representative 1.5 litres composite for each sh pond was then obtained. For trace metals analysis samples were stored in acid-washed low density polyethylene bottles. Raw water samples were further transported in icebox to laboratory and were ltered through pre-cleaned cellulose acetate membrane lter paper (0.45 mm) and kept cold prior to trace metal analysis.

Fish
Fish were collected to obtain information on the bioaccumulation of trace metals in sh pond area. Two species Clarias gariepinus and Oreochromis niloticus were chosen because they are amongst the farmed species mostly reared and consumed in the four regions. High consumption of these species was attributed to the eshy nature and sweet taste. 70 sh samples, 5 each of C. gariepinus and O. niloticus were procured in seven sh farms located in each region. C. gariepinus samples have a mean weigth of 200 ± 47.68g and mean length of 20 ±1.5cm . O. niloticus have a mean weigth of 130 ± 18g and mean length of 16 ± 1.45cm. These sh species were put in sterile polythene bags and taken in icebox to the laboratory where they were washed with running tap water to remove dirt. All the sh samples were then separately stored inside deep freezer at about -4 °C and were allowed to thaw, scales were removed and washed with tap water before dissected with sterile scissors to remove gills, operculum, vertebrae, heart, muscles and kidney. Only the muscles tissue were transferred into sterile sample bottles, labeled and kept for digestion and analysis of heavy metals.

Sample preparation and analysis
Determination of physico-chemical parameters The physico-chemical parameters of the water, including the temperature, pH, electrical conductivity (EC), and total dissolved solids (TDS) were determined and analysed, Temperature was measured with glass mercury thermometer, conductivity was measured using Hanna Potable conductivity meter, pH was measured using a digital pH metre and total solids, dissolved solids and suspended solids was measured using gravimetric methods.

Digestion of samples
Fish samples were dried in an oven at 80-85 °C overnight. The samples were removed from the oven, allowed to cool, and ground in a clean mortar and pestle. Approximately, 0.5 g of each sample was placed in a te on microwave digestion bomb with 10 mL of concentrated HNO3. The samples were allowed to ramp to 180 °C for 5 min, digest at 180 °C for 9.5 min, and cool down for 5 min in a MARS 5 microwave digestion system (CEM, Matthews, NC). The samples were then transferred to clean volumetric asks, and diluted with H 2 O to 10 mL. The samples were stored at 5 °C until ready for analysis of metals.
An aliquot of 1 L water sample was digested with 5 mL concentrated nitric acid to a nal volume of about 25 mL. The digest was left to cool. It was then ltered into 50 mL volumetric ask and diluted to 50 mL mark with distilled water.

Trace elements determination
The digested sample was then ltered through Millipore membrane lter of 0.45 μm (Type HV) and diluted up to 25 ml by addition of distilled water. This ltrate of each sh sample was then processed for analysis of sh muscle content of arsenic (As), cadmium (Cd), chromium (Cr), cobalt (Co), lead (Pb), and nickel (Ni), were using Inductively Coupled Plasma (Perkin Elmer, Model: Optima7000 DV ICP-OES) following the Standard Methods (AOAC, 2012). Detection limits for Cu, Pb, Cr and Cd were 0.01 ppm. High purity metal standards were used for instrument calibration and also for accuracy checks after analysing every ve samples. Metal levels in sh tissues were expressed as mg kg -1 dry weight, while metal levels in water were expressed as mg l -1 (ppm). The wavelengths used for the detection and measurement of the trace elements are 193,696 nm; 228,802nm; 267,716nm; 228,616nm; 220,353 nm; 231,604 nm respectively. The levels of metal contamination of water in relation to sh were expressed as bioconcentration factor (BCF) according to the methods of Lau et al. (1998) calculated as: where, C sh is the mean concentration of the heavy metal in muscle of sh, C water is the mean concentration of the metals in water and BCF sh is the bioconcentration factor of the metals in the sh.
Health risk assessment ( where E F is the exposure frequency (365 days/year); ED is the exposure duration, equivalent to average lifetime, 59 years for Cameroonian population WHO (2017); FIR is the fresh food ingestion rate (g/person/day), which was considered to be 42g/person/day for sh (Ali & Hau, 2001); Cf is the conversion factor (0.208) for fresh weight (FW) to dry weight (DW), Cm is the heavy metal concentration in foodstuffs (μg/ kg FW); BW is the average body weight (average adult body weight was considered to be 75 kg) and TA is the average exposure time for non-carcinogens (equal to E F . ED) (Saha & Zaman, 2012). Health risk of consumers due to intake of metal in sh was assessed by using THQ in accordance with US-EPA Region III risk-based concentration table (US EPA, 2009) and Bortey-Sam et al. (2015). If THQ is less than 1 then the exposed population is unlikely to experience adverse effects; whereas THQ greater than 1 indicates a likelihood of non-carcinogenic effects, with an increasing probability as the value increase (Wang et al., 2005). THQ was calculated using the Equation

Statistical analyses
The results obtained were subjected to statistical evaluation. The concentration of trace metals in water and tissue samples were reported as the average means ± standard deviation. From the normality test, the variables (metals in water and sh tissues) were not normally distributed. Consequently, a non parametric test Kruskal Wallis was used to examine differences in mean values among the two species and rearing systems with signi cance at p < 0.05, using statistical package R.

Results And Discussion
Physicochemical parameters of pond waters and concentrations of toxic metal in water Physicochemical parameters (temperature, pH, conductivity, total dissolved solids) and concentration of toxic metals of the pond waters are shown in Table 2. There were no signi cant differences in the physico-chemical characteristics of the water from earthen and concrete, only for TDS and conductivity (p < 0.05). The physicochemical parameters and level of trace elements in pond water were compared with international standards. According to Madu et al. (2017), water quality parameters in uence the heavy metals levels in freshwater bodies and their accumulation in aquatic biota. The ndings of the study revealed that the mean temperatures for both earthen (29.5°C) and concrete (26°C) ponds were within the WHO recommended limits. Fish is a cold blooded animal, its temperature is dependent on the temperature of its environment. It changes with the temperature of the surroundings. The optimum water temperature for sh survival has been reported to be between 20-30 o C (Bhatnagar and Devi 2013; Ntenegwe and Edema 2008). The temperature changes affect both the metabolism and physiology of shes, and so its productivity (Agbaire et al. 2015). It was also observed that there is no signi cant difference in temperature in the culturing facilities used for the experiment. The reasons can be ascribe to the fact that the earthen and concrete facilities are subjected to the same light intensity and maybe the type of holding structure does not in uence temperature change. The present results are in line with recent studies conducted by Moshood (2017) and Agbaire et al. (2015) which reported that the temperature of ponds ranges between (30.9±0.1-29.4±0.1°C) and (26.73±1.730°C) respectively.
The pH measurement helps to determine if the water is a proper environment for sh, plants and algae (Olukunle et al. 2017). Fish are known to have an average pH of 7.4 therefore pond water average within this range is optimum. In the current study, the results obtained for pH were respectively in concrete ponds and earthen pond where within the limits. The desirable range for pond pH is 6.5 -9.5 and acceptable range is 5.5 -10.0 (Stone and Thomforde 2003). It has been reported that pH value between 6 and 9 was appropriate for sh production (Bhatnagar and Devi 2013). Thus, good pond productivity and sh health can be maintained. The ndings are slightly lower than those obtained by Olukunle and Oyewumi (2017) in Akure (Nigeria) which reported a pH (7.10 ±0.06 -8.39±0.01) for concrete ponds and earthen ponds. A comparative study conducted by Moshood (2017) to assess the water quality of four types of aquaculture ponds under different culture systems reported pH values of 7.88 ±0.2 for earthen and 6.75±0.2 for concrete.
The conductivity values of the ponds gave a good estimate of the condition of ponds under the different culture systems. Both TDS and electrical conductivity are usually positively correlated. The earthen pond has a lower conductivity value of 123 ms/cm, and lower value of TDS than concrete ponds. The FAO acceptable limit for conductivity in aquaculture is between 20 and 1500 μs/cm (DWAF. 1996). The conductivity and total dissolved solids in earthen and concrete ponds probably came from the source of water for the ponds, absence of phytoplankton and aquatic vegetation to assimilate the mineral salts and the effects of residual feed in water which on decomposition added some mineral salts into the water.
The presence of algae and macrophytes which utilized the salts and rainwater as source of water might be the reason for the slightly lower concentrations of conductivity and TDS in both earthen and natural ponds. The conductivity and TDS range however in the ponds were ideal for sh culture. When compared with recommended limits, all the parameters were within the prescribed water quality standards for sh farming. Furthermore, there was no signi cant difference (P<0.05) obtained among the parameters, except for TDS and conductivity showing the effects of the different culture systems. All these results in the present study compare reasonably well with other results obtained ( Olukunle and Oyewui 2017) . Water quality describes physical, chemical, biological and aesthetic properties of water which determine its tness for use and its ability to maintain the health of farmed aquatic organisms. It is imperative to have all the physical, chemical and biological factors in the ponds in a balanced proportion and tolerable limit for optimum sh production. Table 2 shows the concentration of trace element (As, Cd, Cr, Pb, Ni, Co) in water from the earthen and concrete ponds. Heavy metal concentrations decreased in concrete ponds water in the sequence Cd > As >Co > Ni > Cr > Pb. For earthen ponds, the concentrations decreased as Cd > Co > As > Ni > Cr > Pb. The analysis of heavy metals in water showed that the concentrations did not exceed WHO guidelines. The concentrations of each trace element in water did not vary signi cantly either in concrete or earthen ponds. Comparatively the values obtained were lower than other studies in same type of sh ponds carried out by Echor and Okaliwe (2017) that reported As (1.36-1.30 µg/L), Cr (0.72-0.90 µg/L) and Pb (2.00-1.60 µg/L). Aladasanmi et al. (2014) also reported high values above optimal levels for Pb (3.30 µg/L), Cr (9.11 µg/L), Co (10.53 µg/L) and Ni (90.22 µg/L). Fish performs all their body functions in water therefore the quality of water is very important, the type of ponds in uenced the water quality and the absorption of heavy metals in the tissue of shes, in the present study no signi cant differences occurs concerning the contamination by heavy metals. Thilza and Muhammad (2010) reported that concrete ponds pose lesser threat to the livehood of the sh as compared to the earthen pond types. This is due to the management practices on sh farms; the constant change of water in concrete pond type increases the chances of heavy metals to be absorbed in the tissues of shes causing health threat to both sh and sh consumers. Water is the rst source of contamination for sh, since all production materials move into the water, the food they consume, their waste products such as faeces and therefore the origin of water should be control, in the present study water comes mainly from borehole for concrete ponds and for earthen ponds from rivers (Bouelet Ntsama et al. 2018). Fish can be affected by contaminants present in these different sources. However, there should be periodic or constant water quality control of sh ponds in order to ensure sh consumer's safety. The knowledge of heavy metal concentrations in water is very important with respect to management, human consumption of these aquaculture products and to determine the useful monitor and remediation of the most polluted area (Pravin et al. 2011). The concentrations of heavy metals in concrete and earthen sh ponds were slightly higher than permissible limits for these metals. This underscores the need to periodically study our sh rearing environment.
Trace metal concentrations of tested heavy metals (dry weight) found in Clarias gariepinus species raised in concrete and earthen ponds have been presented in Figure 3.
The present study revealed the presence of heavy metals in both rearing systems. This highlights the fact that pollution could affect concrete sh ponds as well. The results obtained suggest that the presence of heavy metals such as As, Cd, Pb, Cr, Ni, Co and some of which are essential for life at trace levels are well above permissible concentrations making them a signi cant threat to ecosystems. Furthermore, toxic trace elements, arsenic, lead and cadmium were all present, in some cases at higher levels than acceptable concentrations. The levels of heavy metals concentrations in concrete sh ponds and earthen sh ponds were not signi cantly different when compared the two species except for Ni in Nile tilapia and cobalt for African cat sh. Concentrations of Arsenic in the esh of Nile tilapia and African cat sh do not varies so. These values were above the maximum limits of FAO/WHO (0.2 µg/g). Cadmium concentrations in all sh samples were above the maximum limits set by the European commission (EC, 2006) of 0.05 mg/kg. they were no signi cant differences at p<0.05. For Lead, Pb content in all samples from earthen ponds was below the maximum limits (0.3mg/kg), for concrete ponds the mean values slightly above the limit. The permissible limit for cobalt set by EU is 0.04 mg/kg, Co concentrations obtained in the present study were above the limits. The values do not differ for each rearing system, concrete (0.37 µg/g) and earthen (0.402 µg/g) in the esh of Nile tilapia but for African cat sh these values varies signi cantly respectively concrete (0.24 µg/g) and earthen (2.08 µg/g). Ni Concentrations in all sh samples were signi cantly different for Nile tilapia, concrete (12.75 µg/g) and earthen (0.570 µg/g). For African cat sh, there was no difference regarding the contamination in concrete and earthen (0.47; 0.22 µg/g). The concentration obtained were all within the limits of 0.4 mg/kg set by the European commission and 1 mg/kg set by the USEPA (2000) except for Nile tilapia raised in concrete ponds. Cr concentrations were above the international limits (1 µg/g) for all the sh samples, for Nile tilapia when compare the two system concrete (4.133 µg/g) and earthen (2.17 µg/g) no differences was but for African cat sh the difference was signi cant (concrete: 4.55 µg/g, earthen: 9.23 µg/g). The result obtained from this study was similar to the ndings of Daniel and Mathew (2016) and Taweel et al (2013). Madu et al (2017) reported higher values compares to this study, for instance Pb (23.20 µg/g) and Ni (17.55 µg/g), these values were above the permissible values. These variations may be attributed to the differences between the localities, and the amount and source of pollution from an area to another, ecological needs, metabolism and feeding patterns of sh, and also the season in which studies were carried out in this study. Trace metals found commonly in sh are attributed to water, sediment or sh feed contaminated by the raw ingredients and by a mineral pack added by the manufacturer. In the current study, the presence of trace elements in sh were related to pond water, fertilizers and animal droppings use frequently in the study area (Bouelet Ntsama et al. 2018) and probably to sh feed, some authors reported the contamination of raw materials use to produce the feed ( Das et al. 2017). The source of chromium in this research may have originated primarily from the feeds Cadmium is released to the environment in wastewater, and diffuse pollution is caused by contamination from fertilizers and local air pollution. The potential hazards of metals transferred to humans are probably dependent on the amount of sh consumed by an individual.
The bioconcentration factor of each studied metal was evaluated for muscle tissues of C. gariepinus, O. niloticus (Table 5). The trend of the bioconcentration factors was similar to the trend of the metals in the tissues of the sh . In addition, the bioconcentration factor of Mn in the gill and liver of the three sh was signi cantly different (p < 0.05). Among the sh, BCF values recorded for tissues were higher than those recorded for the tissues of H. niloticus and C. gariepinus. The overall highest BCF was recorded for Ni (67.99 ± 9.58) and the overall lowest BCF was recorded for Mn (8.79 ± 1.59) in the muscle of H. niloticus. In H. niloticus, the BCF of Pb was highest in all tissues studied, whereas in C. gariepinus, Pb had the highest BCF in the gills and muscle and Ni had the highest in the liver. Bioconcentration factors of trace elements give a measure of the metal concentration in an organism relative to its concentration in the medium (Ajima et al. 2015). The BCF of heavy metals in muscle of the sh species in the present study showed that there was appreciable bioaccumulation of the various heavy metals in the sh tissues. Because all BCF values determined for the two sh species studied were above 100, according to USEPA (1991) there is no risk to human health associated with consumption of the esh of these sh. The results of this study are consistent with the report by Abdallah and Nweeze et al., (2008) in which pelagic sh (herbivores) recorded higher metal concentrations than benthic (carnivorous) sh. Khalid argued that since species such as tilapia are herbivores, they bioaccumulate a higher metal concentration in the esh than carnivorous cat sh. This suggestion is in agreement with the current study, the cat sh being generally more contaminated than tilapia. This may be due to dietary habits of carnivorous sh observed during dissection, a low tendency for sequestration (higher BCF) of trace metals and a greater susceptibility of this species to the presence of trace metals pollutants relative to in Tilapia.

Relationships between trace element levels in shes
Inter-metal correlations of sh species were assessed and presented in Table 3 and 4 for both species.
The correlations between the different metals may result from the similar accumulation behavior of the metals in the shes and their interactions. Noted signi cant correlations among metals may re ect a common source of occurrence and indicative of similar biogeochemical pathways for subsequent accumulation in the muscle tissue of shes. In the present study, signi cant correlation was observed between studied some heavy metals, in Nile tilapia correlations were between Lead and Nickel, Lead and Arsenic, Nickel and cobalt. In African Cat sh, correlations were less strong between Cadmium and Nickel, Chromium and Nickel, Lead and Nickel, Cobalt and lead. Correlation matrixes were used to understand the signi cant differences between selected heavy metals in sh. Higher correlations between metals may reveal similar contamination levels and primarily originate from a similar source and be transported together (Yi et al. 2011;Maanan et al. 2014). Metals with highly positive correlations are believed to have similar sources, while those with highly negative correlations are believed to have different origins (Al-Alimi and Alhudify 2016).

Health risk assessment
The mean concentrations of heavy metals in the muscle of the two sh species were used to evaluate the human health risk from sh consumption. The average concentration of each metal in the muscle tissues of the sh was converted from dry weight to wet weight, in order to evaluate the health risk (Table 6). The EDI of Ni was highest in all sh species (8.76 for concrete pond in C. gariepinus and 6.34 for earthen ponds) followed by that of Cr (3.13 in O. niloticus for earthen ponds). The target hazard quotient of As was highest in all sh species (for C.gariepinus 6.36 in concrete ponds and 3.96 for earthen ponds; For O. niloticus, 3.33 in concrete ponds and 3.32 for earthen ponds), all these values were above 1. The target hazard quotient of Pb was lowest in both species and rearing systems.
The THQ for most metals was less than 1, which indicates that consumers may experience minor health effects by these metals consumption, except for As. Arsenic had THQ values between 3.32 to 6.76 for species and sh ponds. Hazard indices calculated for ve metals (Pb, Cd, Cr, and Ni) in samples from both specie were all below 1, indicative of a non-carcinogenic effect on human health for all the four metals. This means that the exposure of these metals to the human population from the consumption of farm -raised will not result in any considerable health risks that may be associated with the metals. The accumulation of metals in sh depends on equilibrium between absorption and depuration rates, and thus may re ect localized bioavailability of these substances. It also has to do with the concentration of the trace metal in the surrounding water as well as the feeding habits of the sh species. The adverse effects of heavy metals on aquatic and human lives are well known. Fish lose sense of smell in polluted water and consequently this affects their feeding ability (Brian 2013). In humans, acute or chronic exposure to heavy metals can lead to various disorders such as cancer and can also result in excessive damage due to oxidative stress induced by free radical formation (Monisha et al. 2014).

Conclusion
The present investigation reveals that all the physicochemical parameters and heavy metals except for cadmium in surveyed sh ponds were within the recommended limits for good sh production. The concentration in the sh esh was slightly above the maximum limits and all the hazard quotient were below 1, except for Arsenic. There were no signi cant differences between concrete and earthen ponds. Thus, sh farming in the different regions of Cameroon investigated demands regular monitoring of physico-chemical parameters and metals status for effective management. Fish therefore, sh farmers should possess basic water quality kits to regularly monitor basic water quality parameters of their ponds. Principles of Hazard Analysis of Critical Control Point (HACCP) should be strictly adhered to in sh farm site location and production processes, so that our aquaculture product can go to the international market. The use of vascular plants that possess abilities to absorb these metals via their roots in ponds contaminated with pollutants can help to remedy such environment. Government should put in place a regulatory body which will control aquaculture production and ensure that licensed sh farmers are in line with the international norms. Comparison of Trace metal concentrations (μg.g−1 d.w.) in esh of Nile tilapia (Oreochromis niloticus) reared in concrete and earthen. For each species, the signi cances of the level differences between both rearing systems are indicated below the boxplots. KW: Kruskal Wallis test; NS: Non signi cative.