The Impact of Lake Ecosystems on Mineral Concentrations in Tissues of Nile Tilapia (Oreochromis niloticus L.)

Simple Summary Fish are a source of minerals that is highly favored by consumers in most parts of the world. However, these minerals become toxic upon high-level intake and can accumulate toxic trace elements in different tissues. Nevertheless, mineral distribution in fish tissues is poorly evaluated. Analyzing tissue mineral distribution would help us to understand the physiological role of each tissue and the impact of the ecosystem on mineral and toxic trace element accumulation in the tissues. We evaluated the differences in mineral and toxic trace element concentrations of Nile tilapia (Oreochromis niloticus) tissues from three aquatic ecosystems. Distinct differences were observed between tissues in Nile tilapia; in addition, these concentrations were substantially affected by the lake the fish were caught from. The accumulation of elements toxic to humans, such as aluminum, should be monitored and, in particular, controlled when rearing these fish in aquaculture. Further investigation is warranted to identify the origin of the very high intestinal Fe concentration in all fish samples, which coincided with high concentrations of Al. Abstract This study evaluates the differences in mineral and toxic trace element concentrations of Nile tilapia (Oreochromis niloticus) tissues from three aquatic ecosystems in Ethiopia—Lake Ziway, Lake Langano, and Gilgel Gibe reservoir—with a focus on edible (fillet) and discarded (digestive tract, gills, skin, and liver) parts. A total of sixty (n = 60) Nile tilapia samples were collected, comprising twenty (n = 20) fish from each lake, and analyzed by inductively coupled plasma mass spectrometry. All elements varied markedly among tissues and between the lakes. Some differences in element concentrations were attributed to differences in nutrient load in the ecosystems and the function of the tissues. For instance, the calcium concentrations in skin and gill were distinctly higher in fish from calcium-rich Lake Langano. The d iscarded parts were richer in essential trace elements, showing an opportunity to promote their use in human nutrition to increase the intake of important minerals. However, the accumulation of elements toxic to humans, such as aluminum, should be monitored and, in particular, controlled when rearing these fish in aquaculture.


Introduction
Food insecurity and malnutrition remain a great problem worldwide, with a prevalent burden on developing countries [1,2]. Hidden hunger critically affects human health when diet fails to meet nutrient requirements [3]. Fish are a source of minerals that is highly favored by consumers in most parts of the world [4]. Moreover, eating fish provides polyunsaturated fatty acids (PUFAs) that help reduce the risk of cancer and cardiovascular and water pump irrigation by local farmers have rapidly expanded on the shoreline of Lake Ziway and along its tributary rivers [26]. The lake is also under heavy pollution pressure because of expanding urbanization: Batu and Maki are the fast-growing towns near Lake Ziway [26]. Compared to other lake basins, Lake Langano experiences only small seasonal water level variations. However, in the surrounding lake, there is an increasing number of resorts and tourists. As a result, untreated effluents are directly discharged into the lake; this could have negative effects [22]. The water in Gilgel Gibe Reservoir is collected from three agricultural streams (Merewa, Gibe, and Gulufa), three urban streams (Kito, Kochi, and Awetu), and one forest stream. These streams pass through fast-growing urban and intensive agricultural areas. Untreated wastewater, generated by Jimma town inhabitants, is directly discharged into the urban stream, which is the main tributary of the reservoir. Municipal waste discharge, overgrazing, brick preparation, vegetation removal and land conversion to cropland, drainage, and crop cultivation are the major threats from human activities around the tributaries of the reservoir [27].

Description of the Study Area
Fish samples (Nile Tilapia) from LZ, LL, and GR were used in this study, and a summary of limnological information is given in Table 1 [18,26,27]. Lake Ziway is located in the central Rift Valley zone of Ethiopia. It has open water and flat swampy margins on all sides except the south and southwest. It fills a depression at an elevation of about 1636 m above sea level. Meki and Ketar Rivers are the main tributaries of the lake. The lake is the shallowest lake in the country and drains into Lake Abiyata via Bulbula River. It is the third-largest freshwater lake in the country. It has a surface area of 434 km 2 and has five islands: Gelila, Debre Sina, Tulu Gudo, Tsedecha, and Fundro. The climate of the lake basin is dry to subhumid. The lowland area surrounding the lake is arid or semiarid, and the highlands are subdry humid to humid. Lake Langano is one of the northern Rift Valley lakes of Ethiopia. It is situated at an altitude of 1582 m and has a surface area of 241 km 2 and a mean depth of 17 m. The lake is highly turbid, and the water is usually reddish-brown; it is known by its nickname "Golden Lake". It is mainly fed by runoff and hot springs. The lake water discharges into Lake Abijata via Hora Kello River. The climate is rainy between June and September, followed by a dry season from November to February. Gilgel Gibe reservoir is located 283 km southwest of the capital and 70 km northeast of Jimma town. The dam is constructed at an altitude of 1640 m above sea level. Its maximum and minimum water levels during wet and dry seasons are 1671 and 1653 m above sea level, respectively. Its depth ranges between 2 and 35 m, with a mean depth of about 17.6 m, and it covers a total surface area of 62 km 2 . It is bordered by three districts (woredas): Omo-Nada, Kersa, and Tiro-Afeta. The area has a subhumid, warm to hot climate and receives between 1300 and 1800 mm of annual rainfall; it has a mean annual temperature of 23.7 • C. In addition to hydropower generation, the reservoir plays an important role for the local community as a fishery resource.  [18,26,27]. GR: Gilgel Gibe reservoir; LZ: Lake Ziway; LL: Lake Langano.

Water Quality Parameters of the Three Lakes
On-site physicochemical water quality measurements of the three lakes were performed. Day-time dissolved oxygen, water temperature, electrical conductivity, and pH were measured using a multiprobe meter (HQ40d Single-Input Multi-parameter Digital Meter; Hach Company, Loveland, CO, USA). In addition, the concentration of nitrate and ammonia were also measured on-site using a Palin test photometer (Photometer 7500; Tyne and Wear, UK, NE11 0NS ) ( Table 2).

Fish Sample Collection and Preparation
Nile tilapia (O. niloticus) samples were collected from LZ, LL, and GR during the dry season (October 2018). Nile tilapia was selected based on its relevance for commercial fishing in the country and fish consumption by the local population. The target tissues were chosen based on the edible and discarded parts. A total of sixty (n = 60) fish samples were collected, comprising 20 (n = 20) fish from each lake, caught using standard gillnets. Fish samples were thoroughly washed with tap water, separately wrapped in plastic foil (parafilm), placed on ice, and brought to the laboratory. Fish from Gilgel Gibe Reservoir were transported to Jimma University Zoological Sciences, whereas fish from Lake Ziway and Lake Langano were brought to the Batu Fisheries and Other Living Aquatic Resource Research Center. Upon arrival, total body weight and total body length of all fish were measured to the nearest 0.1 g and 0.01 cm, respectively. Because of the too-small sample amount of tissues such as the liver, the fish samples were randomly pooled into three groups (6, 6, and 7 fish). Each pooled sample was analyzed in triplicate. All materials used in the handling and dissecting of the fish samples were washed in distilled water in advance. Plastic tools (knife and forceps) were used to collect the target tissues, and the samples were stored at −20 • C until further analysis. The digestive tract was analyzed with its contents to have an idea about the mineral intake of the fish. The skin was analyzed with scales, and, in this study, both skin + scales are referred to as "skin"; this is because of an assumption that scales can potentially be a rich source of minerals. The concentrations of macrominerals, microminerals, toxic trace elements, and ash content were determined in all targeted tissues.

Percentage of Ash Determination
Three grams of the dried samples were placed in a preweighed crucible. Before placing in the oven, the samples in the crucible cups were charred on hot plates for 30 min to prevent excess smoke production. For ashing, the weighed samples were incinerated in a muffle furnace (550 • C) for 6 h, and the percentage of ash content was calculated [28].

Mineral Analysis
The fillet and discarded tissue parts of the fish samples were packed and shipped to Ghent University, Belgium, for the analysis of minerals and toxic trace elements. Following the methodology described by Association of Official Analytical Chemists, AOAC (1990), the samples were weighed on a microbalance (Mettler Toledo, AT21 comparator), homogenized, and allowed to react for 12 h in a mixture of 3 mL HNO 3 and 3 mL H 2 O 2 . Subsequently, the samples were heated in a microwave oven (CEM, MARS6) with the following program: 10 min at 55 • C and 400 W, followed by 10 min at 75 • C and 600 W, and, finally, 40 min at 120 • C and 1200 W. The extracts were analyzed by inductively coupled plasma mass spectrometry (ICP-MS; PerkinElmer, Elan DRCe) with a calibration range of 0-50 mg/L. The instrument of detection (LOD) for ICP-MS was calculated as the concentration associated with 3.3 times the standard deviation of the background noise recorded on 9 measurements of the procedural blank and given in online resources. Extracts were diluted 1:10 (v/v) with Ga as an internal standard solution. The signal of the sample analyte was always within the standard curve, and the ISO 11885 international standard method was used to calibrate the instrument. The quantification limits of the analyzed elements were set at 0.005 mg/L for Al, Cr, Cu, Fe, Zn, and Mn, 0.01 mg/L for Cd and P, 0.05 mg/L for Co, Mg, Na, S, Ca, and K, 0.1 mg/L for Ni, and 0.2 mg/L for Pb.

Nutritional Contribution Assessment
Possible mineral contributions in a 100-g serving of Nile tilapia tissues from the three Ethiopian lakes to a percentage of the daily value of foods were simulated based on a dry-matter basis of the samples. For the nutritional contribution assessment, the calculated intake of the analyzed essential elements per 100 g serving of Nile tilapia tissues was compared to the dietary reference intake (Recommended Daily Allowance and Adequate Index) for women aged between 19 and 30 years as an example that is set by the Food and Nutrition Board, Institute of Medicine, United States National Academy of Sciences [29][30][31].

Statistical Analysis
All data were displayed on a scatterplot to check for normality. The effects of tissues on mineral, heavy metal, and ash concentrations were evaluated with a general linear model-repeated measure analysis of variance (ANOVA). One-way analysis of variance (ANOVA) was also used to determine the effect of locality on mineral and toxic trace element concentrations. For both analyses, a subsequent posthoc comparison using Tukey's test was performed. All analyses were done using the statistical package of SPSS version 26.0. Significance was evaluated at p < 0.05 confidence level.

Results
Across the lakes, the concentrations of macrominerals Na, Mg, and K were higher in the digestive tract compared to other tissues. However, in the three lakes, the concentrations of Ca and P were markedly higher in the skin and gill tissues compared to other tissues. Gills from Lake Langano contained more Ca compared to the other two lakes. There was no considerable difference in muscle ash content between the lakes. The ash content in gill and skin was lowest in Gilgel Gibe fish compared to the other two lakes. However, higher ash content was observed in the digestive tract and liver from Lake Langano fish compared to the other two lakes ( Table 3).
The concentration of the evaluated microminerals (Fe, Zn, Cu, and Mn in mg/kg dry matter) varied greatly between the tissues and the lakes (Table 4). A remarkably higher concentration of Fe was observed in the gill tissue from Lake Ziway and Gilgel Gibe fish than Lake Langano fish. However, Fe concentration in the liver tissue from Gilgel Gibe fish was higher than from the other two lakes. There was no significant variation in Zn concentration in digestive tract tissue between the three lakes, but it varied in gill and skin. Mn concentration was higher in the skin and digestive tract from Gilgel Gibe fish. Muscle from Lake Langano had a distinctly higher Cu concentration compared to the other two lakes. However, the liver samples from the same lake showed lower Zn concentrations than the two other lakes.
The measured concentrations of toxic trace elements (Al, Cr, Co, Ni, and Cd in mg/kg dry matter) varied depending on the analyzed tissue and the origin of the sample ( Table 4). The highest concentrations were found in liver and digestive tract tissues in the three lakes. For instance, the liver and the digestive tract from Gilgel Gibe fish contained higher concentrations of cobalt than the same tissues from the other two lakes, but tissue cobalt concentrations did not vary between lakes. Muscle, gill, and skin contained lower concentrations of Al and Cr than in digestive tract and liver tissues across the three lakes. The concentration of cadmium was below the detection limit in all analyzed tissues and lakes, except in the liver tissue from Lake Ziway. Digestive tract samples from Gilgel Gibe and Ziway contained notably higher concentrations of Al than other tissues, but these were several-fold lower than samples from Lake Langano.
The percentage contribution of Nile tilapia tissues from the three Ethiopian lakes per 100 g serving to a mineral's dietary recommended intake (DRI) is summarized in Table 5 for women aged between 19 and 30, as an example. Gill and skin provided a higher percentage of DRI of Ca and liver provided a higher percentage of DRI for Fe, Na, and Cu than other tissues per 100 g of serving for women. Table 3. The concentration of ash and macrominerals in Nile tilapia tissues from three lakes in Ethiopia (g/kg dry matter; pooled samples (n = 3 from each lake)). Mineral concentrations of Nile tilapia tissues in the present study were also compared with United States Department of Agriculture (USDA) composition data for raw Tilapia fillets, as presented in Table 6. All Nile tilapia tissues, particularly gill, skin, and liver, had pronounced higher concentrations of all minerals than USDA composition data.   Mg  45  38  38  54  29  48  35  32  26  22  16  19  68  Na  3  12  3  30  7  17  3  32  6  22  15  27  23  K  69  19  23  100  53  23  13  65  57  26  18  53  64  Ca  22  490  460  30  23  660  580  20  24  760  950  50  8  P  143  386  371  357  114  529  442  214  114  629  714  186

Discussion
Ash is the indication of overall mineral content in the samples [32]. From the ash and individual element concentrations, it is clear that muscle (fillet) is not a rich source of minerals. Therefore, encouraging the consumption of other tissues, or the entire fish, could substantially increase mineral intake in the human diet.
Higher ash content in the digestive tract likely reflects high mineral concentrations in the diet, potentially including sediment. In several commercially important fish species [22,33], ash content responded to changes in localities. The sum of all analyzed elements only represents a fraction of the total ash in the digestive tract, suggesting that silicon-containing soil dominates total ash content in the digestive tract samples. The higher ash concentration in LZ digestive tract may, therefore, indicate the eutrophic status of the lake and a more particulate-suspension-feeding strategy of the fish in that lake [26], resulting in a high intake of insoluble, nonabsorbable minerals, as can be seen from the lower ash concentration in the liver of LZ fish, as well as the high intestinal Ca:P ratio, typical for calcium-based soil material.
The "bony" structures in gills and skin can explain the higher deposition of Ca and P in those tissues. Both external tissues deposit hydroxyapatite in the vitrodentin of the scales, composed of collagen covered with Ca [34]. In particular, in these tissues, the highest concentrations of these minerals were observed in fish from LL, showing an adaptation of these fish to the high electric conductivity of this lake. Likely the high Na concentrations in gills and skin in the same lake adhere to that principle as well. This means that the environment affects macromineral concentrations in fish tissues.
Lake Langano is indeed a lake with high Na and Ca concentrations in the water [35] compared to the two other lakes. In contrast to gills and skin, the low Ca concentration in the digestive tract of LL fish means that diet items are not necessarily higher in Ca in contrast to the water. However, reports have indicated that the elevated concentration of minerals in gills reflects a higher concentration in the diet [36], which may still hold for some of the microminerals, as will be discussed further.
Most S in the body is found in S-containing amino acids such as methionine and cysteine. The latter is especially needed to make strong protein structures through disulfide bonds. It is, therefore, logical to find the highest concentrations, again, in skin and gills and fairly low concentrations in muscle, despite its high protein content. Here, too, the high S concentration in LL skin and gills corresponds with the apparent need for stronger structures in the gills and skin to resist the high conductivity and mineral load in that lake. The high S concentrations in the digestive tract -and especially those from GG-are, again, likely to reflect the high S concentrations in the diet items because of the importance of disulfide bonds in the strength of chitin [37], a structure dominating the exoskeleton of zooplankton [38].
Further investigation is warranted to identify the origin of the very high intestinal Fe concentration in all fish. Since it coincides with high concentrations of Al, the intake of suspended soil or the accidental intake of particulate matter during suspension feeding may be a plausible explanation. The three lakes are surrounded by Al-and Fe-rich soils [26,27,39], increasingly challenged by soil erosion [40]. Although Fe is an essential mineral for animals, extremely high concentrations may become a burden to the fish; although Fe uptake is strongly regulated, hepatic accumulation seems to occur [41,42]. Fish are known to regulate elevated Fe concentrations in the whole body by controlling the level in blood, subsequently transferring Fe to the liver for storage [43]. As LL has more Ca-rich soils, the problem is less apparent than in LZ and GG. The overall picture regarding Fe confirms the report by [44] that essential minerals can be found in fish tissues, with the highest concentrations in the liver, gills, kidneys, and spleen and the lowest concentrations in skin and muscle. The liver has a crucial function in basic metabolism, (exchangeable) mineral storage, redistribution, and detoxification or transformation [45,46].
Iron deficiency causes anemia, and fish are a major source of Fe [47]. In the present study, the Fe concentration of the liver, digestive tract, and gills of Nile tilapia is severalfolds higher than in muscle and skin. Therefore, the liver and gills of Nile tilapia may be a good source of iron for human nutrition, depending on dietary needs. Similarly, Cu, Zn, and Mn were also higher in those tissues compared with muscle.
Fish liver and gills accumulate not only essential elements like Fe, Cu, Zn, and Mn, which may also become toxic at too-high uptake levels but also toxic trace elements like Cr, Al, Cd, Ni, and Co (Table A1) [10]. It is, therefore, important to be careful in the consumption of Nile tilapia liver due to the high content of these toxic trace elements. For this reason, only the liver and gills of Nile tilapia that are harvested or cultured under controlled environmental conditions can be considered safe for consumption.
According to the US Food and Drug Administration [48], food that provides more than 10% of nutrients per serving is considered a good source of that nutrient. In this study, mineral contribution to DRI per 100-g serving was calculated (Table 5). Nile tilapia tissues from the three Ethiopian lakes of this study can provide more than 10% of minerals (Ca, P, Fe, Zn, Cu, Mg, Mn, Na, K) for women aged between 19 and 30, hence making Nile tilapia tissues (other than muscle) good sources of these minerals. This comparison is made to show that small inclusions of such parts (tissues) in the diet may substantially improve the intake of certain minerals. Moreover, a comparison of mineral concentrations in tilapia tissues from the three Ethiopia lakes to USDA composition data for raw tilapia showed that Nile tilapia tissues from the three lakes had different concentrations of minerals compared to USDA composition data [48] (Table 6). This possibly indicates that wild tilapia may have higher mineral concentrations than fish in cultured conditions, and this should be considered in the contribution of fish to human mineral intakes. Several factors could contribute to these differences, such as diet, geographical variations, the environment that the fish were grown in, and the methods that were used to measure the minerals.
The concentration of heavy metals in water influences the content of toxic trace elements in fish tissues (Table A3) [49]. In our study, the highest concentrations of these toxic trace elements were recorded in the liver and digestive tract, followed by gills, and the lowest in muscle and skin. This agrees with other reports that the metal accumulation capacity of muscles in fish is lower than that of other tissues such as gills, gut, and liver (Table A2) [50]. Cadmium is a nonessential element found in natural waters, with potential toxicity to fish even at low levels [13]. The detection of Cd only in the liver tissue from LZ could be related to intensive human activities carried out in every corner of that lake. A recent publication reported that the lake is under increasing agricultural and urban pressure and is exhibiting deteriorating trends in several water-quality and ecological parameters; additionally, toxic trace element concentrations of the lake have shown increasing temporal trends [39]. Similarly, the higher Al accumulation in the digestive tract from Gilgel Gibe and Ziway fish is likely caused by the high inflow of eroded topsoil due to intensive unprotected farming activities in the surrounding ecosystems. Whereas the other toxic trace elements were still at acceptably low concentrations (Table A1), the high concentrations of Al in the digestive tract may be a concern for human consumption as it exceeds the maximum tolerable intake [51].

Conclusions
As a conclusion, distinct differences in minerals and toxic trace element concentrations were observed between tissues in Nile tilapia. In addition, these concentrations were substantially affected by the lake these fish were caught from. Some differences could be attributed to the water quality of the lakes, whereas others may refer to different food webs. Compared to fillet (muscle), the typically noneaten parts were much richer in minerals, leaving an opportunity to fortify human nutrition with essential minerals. However, caution is needed due to the accumulation of some toxic trace minerals such as Al, likely arising from anthropogenic soil erosion. Moreover, the route by which these toxic trace elements enter into fish tissue should be optimally identified, and responsible bodies should pay attention to public health safety. Quality control for toxic trace elements in water and diet is therefore warranted when culturing these fish. his field guidance at Gilgel Gibe Reservoir and the Batu Fish and Other Aquatic Life Research Center and its technical assistants for laboratory provision and sample preparation, respectively.

Conflicts of Interest:
The authors declare that they have no conflict of interest. GR, Gilgel Gibe reservoir; LZ, Lake Zeway; LL, Lake Langano.