Evaluating the effect of acute ammonia toxicity in a freshwater ecological environment on the growth and hematological functioning of Clarias gariepinus

Ammonia is one of the main environmental pollutants in aquatic ecosystems (Xia et al. 2018). It enters the aquatic environment from several sources such as sewage effluent, industrial waste, agricultural run-off and decomposition of biological waste, resulting in pernicious toxic impacts on aquatic organisms and soils (Randall & Tsui 2002; Sinha et al. 2012; Attah et al. 2021). In natural surface waters, ammonia occurs in two forms: ionized ammonia, NH₄⁺, and unionized ammonia, NH₃⁰ (Francis-Floyd 2009; Sinha et al. 2012). Excessive ammonia can cause a reduction in growth performance, tissue erosion and degeneration, immune suppression and high mortality in aquatic animals, and increasing ammonia levels in blood and tissues leads to toxicity (Li et al. 2014). Stress has been considered to be the main contributing factor of fish disease and mortality in aquaculture (USDA 2015).

Hematological parameters are one of the commonest and most important biomarkers for diagnosing the structural and functional status of fish exposed to effluents and pollutants (Faggio et al. 2014; Burgos Aceves et al. 2019). They are considered to be a reliable approach in the assessment of toxicities of different single chemicals and their mixtures to fish (Vosylienė et al. 2003). Changes in hematological parameters depend on the magnitude of the impact of contaminant (concentration), the duration of exposure, fish species, age and health status (Alimba et al. 2019). Disrupted hematological patterns appear very quickly and precede changes in fish behavior and visible lesions (Brucka-Jastrzebska & Protasowicki 2005). Alterations in white blood cell (WBC) numbers might be regarded as a prognostic tool or an early-warning signal of disturbance in homeostatic defense abilities of fish (Oladokun et al. 2020).

Fish, which is an aquatic dweller, is an important source of animal protein and lipids for humans and domestic animals (Banaee 2013). In Nigeria, about 10% of the population depends wholly or partly on fish and on the fisheries sector for their livelihood. It also provides a rich source of protein for human consumption. Clarias gariepinus is one of the most cultured fish in Nigeria (FAO 1997; Offem & Ayotunde 2010; Adekunle 2011). In the study areas, soil erosivity and high slope events are frequently leading to erosion and washing agricultural ammonia toxic substances downstream, and these have become a regular feature in Ebonyi state, especially in Afikpo Axis, thereby increasing toxicity in freshwater ecological environments (Ota et al. 2018b). Ota & Eyasu (2020) have carried out many studies on the impact of slope gradient, land-use and land-cover (LULC) changes on soil erosion in the study region. In their findings, they discovered that slope gradient and LULC changes have drastically led to significant sediment yield and variation in soil physico-chemical properties, soil toxicity and moisture content changes in the study areas. Ammonia (NH₃) concentrations are present in the aquatic environment in Nigeria, especially in the study area, due to agricultural run-offs and decomposition of biological wastes.

Ebonyi State was among the six new states created in Nigeria by the then federal military government of General Sani Abacha, with the largest being the city of Abakaliki, comprising 13 administrative local government areas. As a new state, it has changed over the years, and population increase in the state has led to an increase in built-up expansion from 1996 to 2018. As a new state, it has witnessed massive, rapid and robust infrastructural transformation, institutional expansion, large-scale land-use conversion and population density increase (Ota et al. 2021). Against this background, this study was designed to investigate the effect of NH₃ toxicity in freshwater on the growth and hematological function of C. gariepinus subjected to sub-acute exposure. The aim was to (a) determine the tolerance of C. gariepinus to NH₃-concentrated environments, (b) determine the growth and survival rates of these fingerings when exposed to sub-lethal concentrations of NH₃ in water, and (c) assess the hematological responses of fish to ammonia toxicity. Fish are vulnerable to toxicity, especially when exposed to several nitrogenous compounds such as ammonia in their environments. A lot of work has been done on acute and chronic ammonia toxicities in many marine fish, especially scaly species, such as Megalobrama amblycephala (Zhang et al. 2017), Oreochromis niloticus (Benli et al. 2008), Salmo salar (Kolarevic et al. 2012) and Premnas biaculeatus (Rodrigues et al. 2014), but little research has been attempted on scaleless fish. In addition, C. gariepinus is a scaleless freshwater fish that may be more susceptible to the effects of toxic chemicals than fish with scales. Increase in fish toxicity and aquatic pollution in the study area has become, and is becoming, a recurrent phenomenon, and no research attempts have been made in these areas to address the challenge of toxicity. Hence, this work is designed to investigate the effects of NH₃ toxicity on freshwater in the study area. Evidence-based recommendations from this study will support efforts to understand and to evaluate the possible mechanisms underlying the toxic response of C. gariepinus to ammonia exposure, thus achieving the United Nations Sustainable Development Goals (UN SDGs 11) to sustain life below water, particularly in developing countries like Nigeria.

The research was carried out at the wet laboratory of the Department of Fisheries and Aquaculture, Ebonyi State University, Ebonyi State, Abakaliki, Nigeria. The area is located at latitude 06°361 N and longitude 08°161 E (Figure 1) and has an elevation of about 100 m above sea level (Figure 2) in the derived Savannah of the Southeast agro-ecological zone (Ota et al. 2018a, 2020). Rainfall is fairly distributed throughout the year, with a minimum annual rainfall of 1,800 mm and a maximum of 2,000 mm.

Three hundred active C. gariepinus fingerlings (1.52 ± 0.03 g) were collected from the ‘Regina pacis fish farm’ and transported to the wet laboratory in well-oxygenated containers filled to capacity with water from the farm pond. The fish obtained were of the same genetic background and appeared healthy. They were immediately acclimated for 14 days under laboratory conditions and fed a 2-mm multi-feed diet as maintenance ration at 5% body weight per day (bw·d⁻¹). Two distinct experimental periods were adopted for this study: the acute toxicity period lasted for 4 days (96 h) and the sub-acute toxicity period was spread out for 21 days.

Records of fish weight were taken with an electronic sensitive weighing balance to ascertain the weight of fish before the commencement of the study. Fish were randomly stocked in 15 plastic containers (25 liters capacity) at 10 fish per container and inundated with 20 liters of dechlorinated tap water. The experiment was designed to have five treatments of graded concentrations of various levels of NH₃ and control (without NH₃) for ease of comparison of response by C. gariepinus. Each of these treatments was replicated three times to provide a total of 15 experimental treatments for both the acute and sub-acute toxicity studies. NH₃ was obtained from ammonia solution S.G. 0.88, manufactured and packaged by Griffin and George, 28 Ealing Road, Wembley Middlesex, England Fisons Plc, Scientific Equipment Division.

A preliminary test was carried out using the trial and error method to establish the ranges of lethal concentration values. The toxicity period lasted for 4 days (96 h). After the acclimation for 14 days, feeding was discontinued 24 h before the commencement of the bioassay. For the acute toxicity study, five graded concentrations of NH₃ solution were used: T₀ (0.00 g·l⁻¹), T₁ (0.40 g·l⁻¹), T₂ (0.60 g·l⁻¹), T₃ (0.80 g·l⁻¹) and T₄ (1.00 g·l⁻¹). Ten fingerlings were stocked per replicate and estimated for their mortality rate after 24, 48, 72 and 96 h, respectively. Dead fish were removed immediately to prevent water contamination. At the end of the toxicity period, the surviving fish were washed with clean tap water and then subjected to the sub-acute toxicity period.

A 96-h lethal median concentration (LC₅₀) was determined by a probit analysis method as described by Finney (1971) using an arithmetic method to derive probit mortality and plotting the former with the logarithm concentrations. LC₅₀ was subsequently extrapolated from probit 5 to the log concentrations. The antilog value gave LC₅₀ in M·L⁻¹. It is a parametric method in which after calculating percent mortalities, the values of empirical probit from Table 1 are noted depending upon the straight line obtained in the graph. The values of empirical probit were followed by calculating the expected/provisional probit to determine the values of mean and standard deviation of mortalities.

The threshold toxicity was estimated from a graph of log medium survival line versus log concentration by extrapolation.

The safe concentration of NH₃ solution was estimated by multiplying LC₅₀ by a factor of 0.01 (Koasomedinate 1980).

Ten groups of C. gariepinus fingerlings (1.41 ± 0.10 g) were batch-weighed and randomly stocked in 15 plastic containers (15 liters capacity each). Fish weights were taken at the beginning and end of the experiments. Exposure to sub-lethal concentrations of NH₃ lasted 21 days after which five fingerlings per container were subjected to hematological analysis.

• (a)
  The total feed intake (TFI) was estimated from the following relationship (Windham 1996):

  

  (3)
• (b)
  The food conversion ratio (FCR) was calculated from the following equation:

  

  (4)

• where F is the weight of food supplied to fish during the test period, W₀ is the weight of fish at the beginning of the test period and W₁ is the weight of fish at the end of the test period.

• (c)
  Weight gain (WG) was calculated from

  

  (5)
• (d)
  The percentage survival rate (PSR) was estimated thus:

  

  (6)
• (e)
  The specific growth rate (SGR) was estimated from the following relationship:

  

  (7)

Blood samples were collected from the caudal peduncle of the fish by using 2.5 ml capacity syringes and hypodermal needles treated with an anticoagulant (ethylenediaminetetraacetic acid). The following hematological parameters were analyzed in the laboratory: packed cell volume (PCV), WBCs, red blood cells (RBCs) and hemoglobin concentration (HBC).

The results obtained were subjected to one-way analysis of variance (ANOVA) to determine significant differences between the treatment means and the control (T₀) using Statistical Package for Social Science (SPSS) software. The Duncan's multiple range test was used to separate the differences among the treatment means. The differences were considered significant at P ≤ 0.05 (Steel & Torrie 1990).

The total number of C. gariepinus per day at various NH₃ concentration levels is shown in Table 1. The mortality of the exposed fingerlings for acute toxicity studies increased from T₁ to T₄. The result shows that the mortality rates were 30% in T₁, 56.67% in T₂, 60% in T₃ and 73.33% in T₄, with no record of mortality seen in control (T₀), and the overall mortality rate ranged from 30 to 73%. It was observed that the mortality of fish in T₄ increased at 24 h and gradually reduced at 96 h, while mortality in T₁ increased at 96 h, although the values obtained were lower than those recorded with T₄. The 96 h median lethal concentration (96 h, LC₅₀), determined graphically with profit versus log concentration, gave a mean value of 3.2 g·l⁻¹ with upper and lower confidence limits of 6.28 and 1.45 g·l⁻¹, respectively. The threshold concentration, also determined graphically from a plot of log of survival time versus log concentration (Figure 3), gave a value of 1.00 g·l⁻¹. The safe concentration was determined by calculating from the product; this was done by multiplying LC₅₀ by 0.01 (Koasomedinate 1980), and a value of 0.03 g·l⁻¹ was obtained. Graphically from a plot of log of survival time versus log concentration (Figure 2), a value of 1.00 g·l⁻¹ was obtained. The safe concentration was determined by calculating from the product, by multiplying LC₅₀ by 0.01 (Koasomedinate 1980), and a value of 0.03 g·l⁻¹ was obtained.

Table 2 illustrates the growth performance of C. gariepinus exposed to sub-acute concentrations of NH₃, namely, T₀ (0.00 g·l⁻¹), T₁ (0.40 g·l⁻¹), T₂ (0.60 g·l⁻¹), T₃ (0.80 g·l⁻¹) and T₄ (1.00 g·l⁻¹), respectively, within a 14 -day study period. Statistical analyses were applied to determine the effect of sub-acute NH₃ levels on fish growth and survival. After analysis, the data indicated that the mean number of fingerlings differed significantly (P < 0.05) between all NH₃ treatments, but there was no significant difference (P > 0.05) between T₁ (0.40 g·l⁻¹) and T₄ (0.60 g·l⁻¹). Analyses of the final number of fish indicated that C. gariepinus fingerlings had final numbers that differed significantly (P< 0.05) in all treatments, while no significant difference (P > 0.05) was recorded between T₁ and T₂. Similarly, significant differences (P < 0.05) were recorded in the final weight of fish, but no significant difference (P > 0.05) was recorded between T₀ and T₁, T₂, T₃ and T₄ each.

The TFI showed no significant difference (P > 0.05) in all treatments, including control (T₀). The SGR showed a significant difference (P < 0.05) in all treatments, but the control (T₀) treatment did not show any significant difference (P > 0.05) with T₁ and T₂. Similarly, no significant (P > 0.05) difference was recorded between T₃ and T₄ (Table 1). From the graph analysis plot in Figure 1, it can be observed that the TFI increases at controls (T₀) and T₁ and starts to decrease at T₂ and T₄, showing a negative trend. Plotting the SGR against different levels of NH₃ concentration indicates a positive trend; however, at T₀ (control), the SGR shows a sharp decrease at T₁ and starts to pick up (increase) in T₂, but a flat curve is observed at T₃ and T₄ (Figure 3). The PSR results show irregular variation among treatments; the graph shows a negative relationship between PSR and deferments levels of NH₃ concentration (treatments) and a decrease in T₀ and T₄, but a slight rise is seen in the graph line in T₁ and T₂ (Figure 3). The FCR results show an upside-down v-shaped curve where the FCR starts to increase in T₀ and suddenly starts to decrease sharply in T₂ and T₄, leading to a tentative variation relationship between the FCR and the different levels of NH₃ concentration application. The results show that the relationship between the MWG and the different levels of NH₃ concentration is positive (Figure 3). At the initial stage (T₀), the MWG starts to show a slight decline (from T₀ to T₁), and expectantly, it begins to witness a sharp and robust increase in T₂, T₃ and T₄. Furthermore, significant differences (P < 0.05) exist among the MWG in all treatments, while no significant difference (P > 0.05) is derived between the controls (T₀) and T₂. The same trend is found for the MWG per fish in the various treatments. The MWG decreases as the level of NH₃ in water increases. The FCR of fish is significantly (P < 0.05) higher (i.e. –0.12 ± 0.02 in NH₃ treatments with low concentration) as shown in Table 2. The same trend is seen with the survival rate (%) where values differ significantly (P < 0.05) in all treatments, but no significant difference is observed (P > 0.05) between T₀ and T₂ (Table 2). The control (T₀) has the highest survival rate, followed by T₂, T₁, T₃ and T₄.

The hematological parameters of the test fish when exposed to sub-acute concentrations of NH₃ for 14 days are presented in Table 3. The following is the result when compared with the effect of NH₃ on the blood of the exposed fish and the ammonia-free treatments: statistical data on the above showed that the PCV showed a significant difference (P < 0.05) in all treatments, but no difference (P > 0.05) was observed in the NH₃-free treatments and T₁ and T₃. The same trend as in the PCV above was obtained from the values of HBC, where significant differences (P < 0.05) were recorded in all treatments and none (P > 0.05) between T₁ and T₃. In addition, significant differences (P < 0.05) were obtained by analyzing data on WBCs. The same pattern was observed with the RBCs. The relationship analysis between WBCs and different levels of NH₃ concentration on C. gariepinus fish in this study showed an ‘up and down’ relationship with a ‘W’ curve shape (Figure 4). WBCs were observed to have increased in T₀, T₂ and T₄, with the highest value recorded in T₂, while a decrease in WBCs was seen in T₁ and T₃, with the highest decline value recorded in T₃. The correlation between the PCV and the different levels of NH₃ concentration on C. gariepinus also showed a ‘W’ curve shape graph. The results revealed that there was a rise in PCV values in T₀, T₂ and T₄, with the highest value recorded in T₂, while a low value of the PCV was seen in T₁ and T₃, with the lowest PCV recorded in T₃ (Figure 4). Graph analysis of Hb against different levels of NH₃ concentration showed both positive and negative trends. A positive trend was observed when Hb began to rise in T0, T₁ and T₃ and a negative trend was seen when there was a sudden fall in T₃ and a pick up again in T₄. The RBC outputted graph was observed to be relatively steady (flat curve) at T₀ and T₁, and suddenly, the peak at T₂ and began to fall sharply at T₃ and T₄, with a relatively similar ‘V’ shape curve (Figure 4).

The exposure to toxic substances in aquatic environments can induce negative effects on production and growth performance in fish (Kim & Kang 2015). Barbieri & Bondioli (2015) reported a lower LC₅₀ for the ammonia exposure of Pacu fish, Piaractus mesopotamicus, by increasing the water temperature; in other words, an increase in temperature leads to ammonia toxicity. In this study, an assessment of the total number of dead fish (Table 1) when exposed to various NH₃ concentrations under acute toxicity situations indicated that the exposed fish showed signs of toxicity. A similar study conducted by Francis-Floyd (2009) also reported a high rate of fish malfunction when exposed to NH₃ concentrations and an increase in death rate. It has been reported that high levels of copper exposure and NH₃ cause a toxic syndrome including growth depression, increased mortality (Mohseni et al. 2014) and oxidative stress (Berntssen et al. 2000). Studies have shown that exposure to copper sulfate NH₃ increases oxidative stress in the livers of fish (Vutukuru et al. 2006; Trivedi et al. 2012).

In this study, fish subjected to an acute NH₃ stress (T₁–T₄) caused more than 29%–73% of fish mortality. The cumulative mortality linearly decreased as NH₃ increased, suggesting that NH₃ induced mortality. The activity stress of NH₃ on fish as reflected by its mortality is likely attributed to the effects of NH₃ on improving antioxidant status by increased antioxidant enzyme activities (Ural et al. 2015). This physiological malfunction involves uncoordinated movements, air gulping, attempted escape, convulsion and quiescence before death. The results from this study agree with those of other researchers on the issue of fish toxicity (Kim et al. 2019; Shabrangharehdasht et al. 2020).

The astringent nature of NH₃ is considered to reduce palatability and feed intake (Mueller-Harvey 2006), and the variation in the TFI among treatments in the present study showed that the NH₃ concentrations used in this study altered and affected feed intake of fish. There was high TFI in T₀, and this may be directly attributed to the low effect of NH₃ concentrations on the TFI. It has been reported that NH₃ reduces feed intake and the nutrient digestibility of fish (Azaza et al. 2009; Omnes et al. 2017) mainly due to its toxicity that gives feed an unpleasant taste and decreased palatability (Becker & Makkar 1999).

In the sub-acute exposure of fish, the differences recorded in the final number of fish may be attributed to environmental factors and/or NH₃ concentration levels, leading to mortality (T₁ and T­₄). Furthermore, the available results on the final weight of fish indicated that the lowest observable effect gave a value of 0.40 g·l⁻¹ for the final weight. This result is inconsistent with the report of Saber et al. (2004), which indicated that the observable effect of NH₃ concentration on the growth performance of Nile tilapia was 0.14 mg·l⁻¹ unionized ammonia nitrogen (UIA-N). This variation may be due to the different concentration levels used for the fish species. On the other hand, the final weight of the marine fish decreased when UIA-N·L⁻¹ concentration increased (Foss et al. 2003; Lemarie et al. 2004; Sten et al. 2004). These authors reported that weight decreased with increasing concentration of UIA-N·L⁻¹. This trend was attributed to the decrease in daily feed intake and decrease in food conversion efficiency. In Table 2, it is revealed that food intake reduces with increases in NH₃ concentration, although with very little variation. This result agrees with the reports of Foss et al. (2003, 2004) and Saber et al. (2004).

Significantly lower (P < 0.05) SGR values were observed in T₃ and T₄ than that in the controls (T₀), T₁ and T_2. This is in agreement with the report of Saber et al. (2004). Similarly, other researchers such as Harris et al. (1998) and Foss et al. (2003, 2004) reported that the SGR of fish decreased with increasing the concentration of NH₃, and this was attributed to decreases in food intake. Significant decreases in food intake and a significant decrease in the MWG exerted on the fish by the various NH₃ treatments as NH₃ levels increased are also consistent with the reports of Foss et al. (2003, 2004), Lemarie et al. (2004) and Saber et al. (2004). Furthermore, the significant decrease (P < 0.05) of the WMG between T₃ and T₄ compared with the control (T₀) is attributed to a decrease in daily food consumption. Wang & Walsh (2000) reported a reduction in the MWG of fish under investigation and attributed this to some physiological disturbances in the fish.

Higher FCR values were observed in this study and lower conversion of food to muscles (Table 2). This result is in agreement with the report of Foss et al. (2004), which stated that the mean FCR of fish decreased as the UIA-N concentration increased. The first stage of ammonia freshwater aquatic environment is the nitrification cycle that involves nitrogenous contaminants of nitrites and nitrates. Its accumulation is a serious problem, especially in aquatic ecological environments. Hematological parameters are considered a good indicator for assessing toxicity effects in fish (Yang et al. 2010; Burgos-Aceves et al. 2019; Kim et al. 2019; Shabrangharehdasht et al. 2020). Hematological analyses showed significant (P < 0.05) decreases in the PCV, RBC and Hb levels and were in line with the results obtained by Fayioye (2002). The study noted significant (P < 0.05) decreases in hematological parameters, especially fish without amomonia toxicity and the fish of both C. gariepinus and O. niloticus exposed to aqueous and ethanoic extracts of Raphia vinifera and Parkia biglobosa. The decreased PCV value may be due to anemia, hemodilution or hemolysis of RBCs (Fayioye 2002). Saeed (1997) also recorded a marked reduction in RBC count and Hb concentration due to exposure of O. niloticus to acute NH₃ concentration. Secondly, the significant (P < 0.05) decreases in the WBC levels upon exposure of fish to sub-acute levels of NH₃ may be due to the WBCs in animals acting as soldiers of the body.

Kim et al. (2020) reported that ammonia exposure significantly reduced the hematocrit and hemoglobin levels in African giant fish (C. gariepinus). Yang et al. (2010) also suggested that ammonia exposure led to significant decreases in blood RBCs and hemoglobin. In the present study, there was an irregular increase and decrease in the Hb content, PVC, WBCs and RBCs among the treatments after exposure to high and low ammonia levels. This suggests that the fish had anemia due to ammonia exposure. Hoseini et al. (2019) found that an increase in free radicals from exposure to ammonia might cause an attack on RBCs and WBCs, resulting in RBC and WBC destruction. Studies have shown that ammonia toxicity induces hematopoietic inhibition by destroying the RBC and WBC production sites (Zeitoun et al. 2016). Thus, the occurrence of anemia symptoms might be attributed to RBC and WBC destruction or hematopoietic tissue injury after ammonia exposure. In contrast, Zhang et al. (2017) found no significant differences in RBC production, and hemoglobin synthesis was observed in all treatment groups. Ammonia exposure led to variations in the PCV (Hoseini et al. 2019). These results may indicate that different chemicals can act as hematological disruptors via a variety of different mechanisms. In addition, the decrease in the WBC and RBC count in some of the treatments in this study might be associated with non-specific immune suppression (Gholami-Seyedkolaei et al. 2013).

It is widely observed that ammonia can lead to kidney impairments of fish, thereby further causing a massive loss of WBCs and RCBs through renal excretion (Abidi 1990). Previous reports have also found that higher ammonia concentrations cause protein deamination, thereby increasing protein degradation (Priyadarshini et al. 2011). Ammonia toxicity not only affects hematological characteristics, it is a well-known fact that external environmental pressure can also cause large increases in the blood glucose levels of fish (Wells & Pankhurst 2010). Sriwastava & Srivastava (1985) reported that stress exerted on fish activates the secretion of catecholamine, which enhances the conversion of hepatic glycogen to blood glucose to supply the greater energy demands of cell metabolism.

The response of C. gariepinus fingerlings to survival, growth and mortality in this study provided some basis for predicting the overall impact of NH₃ to fingerlings. Exposure of test fish to NH₃ concentrations above 1.00 g·L⁻¹ could result in hyperventilation, symptoms of toxicosis and increase in mortality rate. The exposure of O. niloticus to NH₃ concentration may affect growth and cause a reduction of Hb, RBC and WBC levels. It is, therefore, recommended from this study that properly managed NH₃ levels in intensive aquaculture systems can go a long way in increasing food intake by fish and can help ameliorate the hazards posed by NH₃ buildup from unconsumed foods. Increasing pond aeration through aeration could be ineffective in reducing the overall pond NH₃ concentration due to the relatively small area of pond to be aerated. However, aeration increases the DO levels, causing fish to be less stressed. It is, therefore, prudent to avoid vigorous pond aeration to prevent the stirring of bottom sediments, which consequently increases NH­₃ concentration.

The authors are thankful to the Department of Fisheries and Aquaculture, Faculty of Agriculture and Natural Resource Management, Ebonyi State University, Abakaliki, Nigeria, and all laboratory staff members for providing supporting facilities to carry out this research work successfully.

All relevant data are included in the paper or its Supplementary Information.