In two pilot-scale experiments, fingerlings and juvenile of tilapia were reared in high rate algal pond (HRAP) effluent. The combination of three different total ammonia nitrogen (TAN) surface loading rates (SLR1 = 0.6, SLR2 = 1.2; SLR3 = 2.4 kg TAN·ha−1·d−1) and two fish stocking densities (D1 = 4 and D2 = 8 fish per tank) was evaluated during two 12-week experiments. Fingerlings total weight gain varied from 4.9 to 18.9 g, with the highest value (equivalent to 0.225 g·d−1) being recorded in SLR2-D1 treatment; however, high mortality (up to 67%) was recorded, probably due to sensitivity to ammonia and wide daily temperature variations. At lower water temperatures, juvenile tilapia showed no mortality, but very low weight gain. The fish rearing tanks worked as wastewater polishing units, adding the following approximate average removal figures on top of those achieved at the HRAP: 63% of total Kjeldahl nitrogen; 54% of ammonia nitrogen; 42% of total phosphorus; 37% of chemical oxygen demand; 1.1 log units of Escherichia coli.

It is estimated that in 2015 only 39 per cent of the global population (2.9 billion people) used a safely managed sanitation service (i.e., excreta safely disposed of in situ or treated off-site), and 2.3 billion people still lacked even a basic sanitation service (WHO & UNICEF 2017). In turn, the vast majority of the hungry lives in the developing regions, where an estimated 780 million people were undernourished in 2014–16, with an undernourishment prevalence of 12.9% (FAO et al. 2015). As a whole, this points out the need for low-cost sanitation and sewage treatment systems in low-income countries, preferably combined with wastewater use in agriculture or aquaculture, which is seen as method of water and nutrient recycling as well as improving household food and nutrition security (WHO 2006).

Under favorable climate conditions and where land is available, waste stabilization ponds (WSP) are one of the most appropriate methods of wastewater treatment in developing countries, as they are low-cost, low-maintenance, highly efficient, entirely natural and highly sustainable systems (Peña & Mara 2004). WSP are also one of the treatment options most compatible and used with aquaculture, since they provide a highly enriched environment that supports the growth of plankton, which serves as fish food (WHO 2006). High rate algal ponds (HRAP) are an advanced type of pond – paddlewheel-mixed, shallow, open raceway – developed for the treatment of wastewater and resource recovery. According to Craggs et al. (2012), HRAPs retain the advantages of conventional ponds (simplicity and economy), overcome some of their drawbacks, and have the additional benefit of recovering wastewater nutrients as harvestable algal/bacterial biomass for beneficial use as biofuel fertilizer or feed.

The literature reports several studies on wastewater use in fish culture in a variety of situations, looking both at toxic effects/fish survival and at fish productivity. For instance, there are quite a few reports on carp polyculture in sewage-fed ponds in India, either in large or experimental scale (e.g., Raychaudhuri et al. 2008; Mandal et al. 2018). As stated by Watanabe et al. (2002), a fundamental advantage of tilapia for aquaculture is that it feeds on a low trophic level; members of the genus Oreochromis are all omnivores, feeding on algae, aquatic plants, small invertebrates, detritus and associated bacterial films, as well as on a variety of feeds of animal origin. Thus, there also are a quite a few works on wastewater-fed tilapia culture, either in WSP (e.g., Silva et al. 2000 in Brazil; Abdul-Rahaman et al. 2012 in Ghana), or in fish rearing tanks/ponds fed with WSP effluents. In the latter case, some works are based on batch experiments (i.e., tanks intermittently fed with effluents), e.g., Gaigher & Krause (1983) in South Africa and Mota et al. (2009) in Brazil. Others, more closely resembling the experimental arrangement of the present work, used tanks/ponds continuously fed with effluents of domestic sewage treated in WSP or HRAP. For example, Edwards et al. (1981), in Thailand, reared Tilapia nilotica in 4 m3 concrete tanks and 200 m2 earth ponds using HRAP effluent; Moscoso (1996), in Peru, used the effluent of a tertiary WSP for rearing Nile tilapia in 370 m2 earth ponds; in Brazil, Mota et al. (2009) used 50 m3 concrete tanks fed with an effluent of anaerobic-facultative-maturation pond series for rearing Nile tilapia. In the same location of the present work, Bastos et al. (2003) reared Nile tilapia in two parallel 1 m3 plastic tank systems: (i) one receiving the effluent of a wastewater treatment system composed of an upflow anaerobic sludge blanket reactor followed by a series of three pilot-scale polishing ponds, where the plankton produced in the ponds was the only available food source; (ii) and another, of conventional aquaculture in freshwater and provision of artificial feed. Briefly, at low stocking densities fingerlings showed similar weight gains both in the wastewater-fed and in the conventional systems; however, during the more advanced fish growth stages higher weight gains were obtained in the conventional system. The authors concluded that the WSP phytoplankton proved to be of excellent nutritional value for the fingerlings but may have been insufficient to sustain high production of tilapia juvenile. Overall, these above-mentioned studies recommended using up to 10 fish per m3 or 2 fish per m2, and reported high fish survival as well as fish productivity estimates up to 6 t·ha−1·y−1.

One of the major concerns regarding wastewater use in fish culture is ammonia toxicity. Un-ionized ammonia (NH3-N) is the most toxic form of ammonia because of its ability to move across cell membranes; thus the toxicity of total ammonia nitrogen (TAN = NH4+ + NH3) is dependent on the percentage present in the un-ionized form, which, in turn, is pH-dependent. In spite of having no shortage of studies on ammonia toxicity to fish, and also more specifically to tilapia (e.g., Wrigley et al. 1988; Benli & Köksal 2005; Hanna et al. 2013), engineers still lack information on design criteria for wastewater-fed fish culture systems based on the very important variable of ammonia surface loading rates (kg TAN·ha−1·d−1). Actually, there are only a few reports on general design criteria, and not so recent. Mara et al. (1993), quoted by WHO (2006), based on the concept of minimal wastewater treatment in WSP for maximal production of microbially safe fish, suggested that a wastewater-fed fish pond receiving an effluent of a facultative pond should be designed on the basis of a surface loading of total nitrogen of 4 kg N·ha−1·d−1 (according to the authors, ‘too little nitrogen results in a low algal biomass in the fish pond and consequently small fish yields; too much nitrogen gives rise to high concentration of algae, with the resultant high risk of severe dissolved oxygen depletion at night and consequently fish kills’). Regarding ammonia toxicity the authors only stated that free ammonia in the fish pond should be <0.5 mg L−1 in order to avoid acute toxicity to fish. Edwards (1992), thinking of maintaining good dissolved oxygen conditions, indicated that fish ponds should receive maximum loading rates of 25 kg BOD5·ha−1·d−1 (BOD5: 5-day biochemical oxygen demand).

Aiming at filling some of these information gaps, in this study we evaluated the effect of three ammonia surface loading rates (SLR) and two fish stocking densities on the growth and mortality of fingerlings and juvenile of tilapia cultured in tanks fed with an effluent of HRAP, using plankton as the only food source

The experiment was carried out in Viçosa, State of Minas Gerais, Brazil (20°45′14″S, 42°52′54″W; average altitude = 648 m), at an experimental site in the University of Viçosa. This Brazilian town presents a humid subtropical climate (Köppen classification), with average, maximum and minimum temperatures of 19.8, 32.4 and 7.2 °C, respectively; annual average relative humidity and precipitation of, respectively, 81%, and 1,221.4 mm – spread over rainy (spring–summer) and dry (autumn–winter) seasons.

Domestic sewage was collected fortnightly, using a submersible pump, from a septic tank of a local small wastewater treatment plant. This anaerobically primarily treated wastewater was transported in tanks of 500 L and transferred to 3,000 and 5,000 L reservoirs at the experimental site, which then fed fiberglass HRAP with the following characteristics: length = 2.86 m, width = 1.28 m, free board = 0.2 m, pond depth = 0.3 m, surface area = 3.3 m2, volume = 1 m3 (Figure 1 left). The ponds' stainless steel paddlewheels were driven by a 1 hp electric motor, whose rotation was controlled by a frequency inverter (WEG, series CFW-10) to provide a mean horizontal water velocity between 0.10 and 0.15 m·s−1. The HRAP were operated at an average hydraulic retention time (HRT) of 7 days with an inlet continuous flow regulated two or three times a day.

Figure 1

High rate algal ponds and sewage reservoirs (left); and fish rearing tanks (right): (1) system inlet; (2) water level control and system outlet.

Figure 1

High rate algal ponds and sewage reservoirs (left); and fish rearing tanks (right): (1) system inlet; (2) water level control and system outlet.

Close modal

Under natural climatic conditions, fingerlings and juvenile of genetically improved farmed tilapia (GIFT) were reared in 18 plastic tanks with 210 L as useful volume (Figure 1 right). Three groups of six tanks received, each, a continuous flow rate of treated effluent, so that three different ammonia SLR were tested: SLR1 = 0.6, SLR2 = 1.2, and SLR3 = 2.4 kg of TAN·ha−1·d−1. The loading rates were chosen based on values published in the literature of sewage-fed fish ponds; the lowest value was similar to those reported by Bastos et al. (2003) and Moscoso (1996), and in order to evaluate the effect of higher levels of ammonia two and four times the lowest value were applied. Based on the values of wastewater flow applied to the tanks, the average HRT in the rearing tanks was: SLR1 = 55 d, SLR2 = 27.5 d, and SLR3 = 13.8 d. For each group of tanks, two fish stocking densities (D1 = 4, D2 = 8 fish per tank, corresponding respectively to 6 and 12 fish per m2) were randomly located. Taking into account the high phytoplankton concentration usually found in HRAP effluents, D1 was selected to be twice the values recommended by Edwards et al. (1981) and Bastos et al. (2003), which were based on conventional ponds.

The rearing units were filled with treated wastewater and left therein for at least 10 days to reduce the levels of ammonia and its toxicity. In order to avoid breeding grounds for Aedes aegypti mosquitos in the stored wastewater during that period of time, at the beginning of the experiments the HRAP effluent was diluted as follows: 60% of the tank's useful volume was filled with tap water free of chlorine and 40% with the effluent.

A first experiment was carried out for 12 weeks during the end of the summer and the first 3 weeks of the autumn. Tanks were stocked with GIFT fingerlings with an average initial weight of 2.60 ± 0.15 g in a randomized distribution of the three treatment replicates: T1: SLR1-D1; T2: SLR1-D2; T3: SLR2-D1; T4: SLR2-D2; T5: SLR3-D1 and T6: SLR3-D2. A second experiment was conducted for 12 weeks during autumn and the first week of winter. Tanks were stocked with juvenile GIFT with an average initial weight of 58.24 ± 7.61 g, according to the replicate's randomized distribution for the six treatments, as previously mentioned. Before starting the experiment, the juveniles were reared in a recirculating aquaculture system and were fed with commercial fish feed with 34% dietary protein.

Fish were weighed at the beginning and at the end of the experiments in order to calculate the average total weight gain for each treatment, as well as the daily weight gain (total weight gain/duration of the experiment). Fish tanks were checked twice a day in order to monitor fish mortality, as well as to remove and weigh dead fish. Analysis of variance was applied to look at differences between treatments (fish weight gain and fish mortality) using the software R 3.1.0. Correlation between fish weight gain and water quality parameters was evaluated by the Spearman's rank test using the software Excel.

During the experiments the following physical and chemical variables were monitored according to the recommendations of APHA et al. (2012) (the methods used for the respective analyses are within brackets): chemical oxygen demand (COD: 5220D), total phosphorus (TP: 4500-P C), ammonia nitrogen (AN: 4500 – NH3D), total Kjeldahl nitrogen (TKN: 4500-N D), dissolved oxygen – membrane electrode method (DO: 4500-O G), pH – electrometric method (4500-H+ B), temperature (2550 B). The pH, DO and temperature were monitored using a Hach portable meter, model HQ40d; the chromogenic-fluorogenic method (Colilert®) was used to measure Escherichia coli (enzyme substrate coliform test: 9323). Chlorophyll-a was extracted with 80% ethanol and measured by spectrophotometry (APHA et al. 2012). COD, TP, AN, TKN, DO, pH, temperature and E. coli were monitored in the septic tank and HRAP effluents once a month during the first experiment, and fortnightly during the second experiment; in the second experiment composite samples were also analyzed from the fish tanks of each SLR line. Chlorophyll-a was measured in the same frequency for the HRAP effluent during the two experiments, and in the fish tanks at the end of the two experiments. The pH, DO and temperature were measured every 5 days in the septic tank effluent, in the HRAP and in the 18 fish culture tanks.

Experiment with tilapia fingerlings

The average concentrations of TKN, AN, COD and TP recorded in the effluent samples were, respectively: (i) 77; 91.8; 1,454.7; and 16.2 mg L−1 in the septic tank (ST) effluent; and (ii) 31.7; 49.7; 295.2; and 7.6 mg L−1 in the HRAP effluent. Thus, average removal efficiencies of 50.8% for TKN, 49.1% for AN, 80.9% for COD and 53% for TP were recorded in the HRAP during the experiments The high values of COD in the septic tank effluent were mainly due to sludge accumulation in the treatment unit and the transport of sludge particles during the pumping of sewage samples. The average chlorophyll-a concentrations recorded in the treatment–fish culture system were: 1,736 μg L−1 in the HRAP effluent; 161.53 μg L−1 in SLR3; 660 μg L−1 in SLR1 and 680.4 μg L−1 in SLR2.

The average water temperature remained rather stable, usually around 23–24 °C. According to a local database, during the same period, the minimum, average and maximum values of air temperature were, respectively, 11.4, 22.3 and 33.5 °C; often, the air temperature varied as widely as 11.4 °C in 24 hours. The pH varied from 5 to 10.2 in the SLR treatments, with average values higher than 8.3; in the HRAP pH varied from 6.9 to 9.1. The DO varied widely, from 2.3 to 21.1 mg L−1 in the SLR treatments, and from 2.2 to 15.9 mg L−1 L in the HRAP, with lower values found at dawn; after midday, supersaturated concentrations of DO were recorded due to intense photosynthetic activity. Table 1 shows the wastewater DO, pH and temperature mean and standard deviation values found in the HRAP and in the SLR treatments during the first experiment.

Table 1

Average and standard deviation (within brackets) values of DO, temperature and pH in the HRAP and in the rearing tanks during the tilapia fingerlings experiment

Treatment or rearing unitDO (mg·L−1)Temperature (°C)pH
HRAP 6.5 (2.86) 22.4 (2.84) 7.9 (0.66) 
T1 (SLR1-D1) 12.4 (5.03) 23.7 (2.94) 8.7 (1.29) 
T2 (SLR2-D2) 9.4 (3.64) 23.6 (2.95) 8.8 (1.03) 
T3 (SLR2-D1) 9.2 (3.10) 23.7 (3.03) 9.1 (0.39) 
T4 (SLR2-D2) 9.2 (3.72) 23.7 (3.01) 8.6 (1.17) 
T5 (SLR3-D1) 10.6 (4.29) 23.8 (2.96) 8.5 (0.94) 
T6 (SLR3-D2) 11 (4.64) 23.8 (3.06) 8.4 (1.43) 
Treatment or rearing unitDO (mg·L−1)Temperature (°C)pH
HRAP 6.5 (2.86) 22.4 (2.84) 7.9 (0.66) 
T1 (SLR1-D1) 12.4 (5.03) 23.7 (2.94) 8.7 (1.29) 
T2 (SLR2-D2) 9.4 (3.64) 23.6 (2.95) 8.8 (1.03) 
T3 (SLR2-D1) 9.2 (3.10) 23.7 (3.03) 9.1 (0.39) 
T4 (SLR2-D2) 9.2 (3.72) 23.7 (3.01) 8.6 (1.17) 
T5 (SLR3-D1) 10.6 (4.29) 23.8 (2.96) 8.5 (0.94) 
T6 (SLR3-D2) 11 (4.64) 23.8 (3.06) 8.4 (1.43) 

Table 2 presents the results of the total fish weight gain and total mortality during the experiment – average values of the three replicates of each stocking density in each SLR treatment. In general, the daily weight gain was lower than those reported by other authors, e.g., Edwards et al. (1981) and Mota et al. (2009). Marked differences in fish weight gains were noticed between treatments, although not statistically significant, probably due to the large data variability (high standard deviation values). In fact, differences in growth within the same fish group are not unexpected. Johnsson et al. (2006) suggested that appetite suppression associated with stress-induced anorexia and decreased nutrient digestibility in subordinate fish can lead to decreased growth in the short- and long-term metabolism. Fernandes & Volpato (1993), studying Nile tilapia, registered a higher metabolic cost for the subordinates compared to the dominants.

Table 2

Average and standard deviation (within brackets) values of total fish weight gain and total mortality during the tilapia fingerlings experiment

TreatmentAverage total weight gain (g)Total mortality (%)
T1 (SLR1-D1) 4.86 (0.33) 66.7 (23.57) 
T2 (SLR1-D2) 7.76 (1.77) 41.7 (35.84) 
T3 (SLR2-D1) 18.90 (14.50) 41.7 (42.49) 
T4 (SLR2-D2) 10.07 (0.27) 62.5 (44.49) 
T5 (SLR3-D1) 9.96 (4.82) 16.7 (11.79) 
T6 (SLR3-D2) 5.91 (1.66) 54.2 (41.25) 
TreatmentAverage total weight gain (g)Total mortality (%)
T1 (SLR1-D1) 4.86 (0.33) 66.7 (23.57) 
T2 (SLR1-D2) 7.76 (1.77) 41.7 (35.84) 
T3 (SLR2-D1) 18.90 (14.50) 41.7 (42.49) 
T4 (SLR2-D2) 10.07 (0.27) 62.5 (44.49) 
T5 (SLR3-D1) 9.96 (4.82) 16.7 (11.79) 
T6 (SLR3-D2) 5.91 (1.66) 54.2 (41.25) 

With the exception of SLR1 treatments (T1 and T2), the lower the fish stocking densities the higher was the weight gain (T3 and T5, compared to T4 and T6, respectively). The highest weight gain (18.9 g, equivalent to 0.225 g per day) was obtained in T3 treatment, maybe due to the intermediate ammonia SLR level (thus less toxic ammonia effects) coupled with the also intermediate chlorophyll-a concentration (hence reasonable availability of plankton as food source). In SLR3 treatments, fish growth may have been inhibited due to the high ammonia concentrations, despite the highest chlorophyll-a concentration. El-Sherif & El-Feky (2008) also observed that tilapia weight gain decreased with increasing levels of ammonia. In turn, the low weight gain in SLR1 may have been associated with limited availability of plankton (in spite of the weight gain in T2 being higher than in T6). Overall, these results suggest the potential for rearing fingerlings at low stocking densities, even at moderate or high ammonia loading rates.

A strong negative correlation was found between fish weight gain and DO (Spearman's rank correlation coefficient rs = −0.90). Surprisingly at first sight, such finding may be explained by the larger variability of DO data in those treatments with higher DO concentrations. As suggested by Wang et al. (2009), it is likely that dynamic changes in oxygen levels would increase the routine metabolic rate and reduce the amount of energy available for growth. There was no correlation between fish weight gain and water temperature (rs = −0.014), certainly because the water temperature did not vary much between treatments. However, correlation was found between fish weight gain and both water pH (rs = 0.371) and chlorophyll-a (rs = 0.486). Notwithstanding these associations being only weak to moderate they suggest that higher plankton productivity may have promoted fish growth.

Fish mortality varied widely, and this certainly contributed to no statistically significant differences being found between the treatments' mortality values. Again, this may be in part explained by the behavioral and physiological consequences of the social order imposed between dominant and subordinate fish (Johnsson et al. 2006). Full mortality was registered in 7 out of the 18 rearing tanks. Mortality events occurred mostly, or more intensively, following wide daily temperature variations, which may have exerted a synergic effect with the ammonia concentrations. Fish mortality may also help to explain the above-mentioned exceptions in the weight gain results (T1 vs. T2 and T2 vs. T6), in as much as, in general, the lower the fish mortality the higher was the weight gain (T2, T3 and T5, compared to T1, T4 and T6, respectively).

Experiment with tilapia juveniles

Table 3 shows the average and the standard deviation values of the wastewater characteristics found in the septic tank and the HRAP effluents and in the fish rearing tanks during the second experiment.

Table 3

Average and standard deviation (within brackets) values of TKN, AN, TP, COD, E. coli and chlorophyll-a in the septic tank and HRAP effluents and in the rearing tanks during the tilapia juvenile experiment

Treatment or rearing unitTKN (mg·L−1)AN (mg·L−1)TP (mg·L−1)
Septic tank 113.2 (20.65) 147.3 (7.38) 16.4 (1.40) 
HRAP 40.1 (14.63) 33.9 (21.44) 12.7 (0.68) 
SLR3 20.3 (7.50) 24.1 (4.37) 9.8 (2.00) 
SLR1 10.5 (3.81) 10.7 (3.53) 5.5 (1.87) 
SLR2 14.1 (5.95) 11.9 (8.55) 6.9 (2.13) 
Treatment or rearing unitCOD (mg·L−1)E. coli (MPN.100 mL−1)Chlorophyll-a (μg·L−1)
Septic tank 1,064 (396.96) 6.6 × 105 (1.01 × 106– 
HRAP 231.5 (57.79) 3.2 × 104 (3.2 × 104475.3 (147.11) 
SLR3 167.8 (45.40) 3.7 × 103 (6.3 × 103234.1 (83.35) 
SLR1 124.7 (34.54) 1.8 × 103 (1.45 × 103178.4 (70.53) 
SLR2 145.5 (51.06) 2.8 × 103 (2.29 × 103174.7 (71.61) 
Treatment or rearing unitTKN (mg·L−1)AN (mg·L−1)TP (mg·L−1)
Septic tank 113.2 (20.65) 147.3 (7.38) 16.4 (1.40) 
HRAP 40.1 (14.63) 33.9 (21.44) 12.7 (0.68) 
SLR3 20.3 (7.50) 24.1 (4.37) 9.8 (2.00) 
SLR1 10.5 (3.81) 10.7 (3.53) 5.5 (1.87) 
SLR2 14.1 (5.95) 11.9 (8.55) 6.9 (2.13) 
Treatment or rearing unitCOD (mg·L−1)E. coli (MPN.100 mL−1)Chlorophyll-a (μg·L−1)
Septic tank 1,064 (396.96) 6.6 × 105 (1.01 × 106– 
HRAP 231.5 (57.79) 3.2 × 104 (3.2 × 104475.3 (147.11) 
SLR3 167.8 (45.40) 3.7 × 103 (6.3 × 103234.1 (83.35) 
SLR1 124.7 (34.54) 1.8 × 103 (1.45 × 103178.4 (70.53) 
SLR2 145.5 (51.06) 2.8 × 103 (2.29 × 103174.7 (71.61) 

The maximum sewage constituent values recorded in the wastewater treatment–fish rearing system were: (i) ST – 131.9 mg·L−1 TKN; 156.1 mg·L−1 AN; 18.7 mg·L−1 TP; 1,485 mg·L−1 COD; (ii) HRAP – 70.8 mg·L−1 TKN; 54.7 mg·L−1 AN; 13.3 mg·L−1 TP; 343 mg·L−1 COD; (iii) SLR3 – 33.1 mg·L−1 TKN; 30.6 mg·L−1 AN; 11.9 mg·L−1 TP; 248 mg·L−1 COD; (iv) SLR1 – 17.4 mg·L−1 TKN; 14.6 mg·L−1 AN; 8.2 mg·L−1 TP; 191 mg·L−1 COD; (v) SLR2 – 24.8 mg·L−1 TKN; 20.5 mg·L−1 AN; 9.4 mg·L−1 TP; 202 mg·L−1 COD. Again, the concentrations of chlorophyll-a varied widely in the HRAP effluent (from 280.3 to 723.6 μg·L−1) as well as in the SLR lines: 125.5 to 392.5 μg·L−1 in SLR3; 109.5 to 320.4 μg·L−1 in SLR1, and 101.5 to 325.7 μg·L−1 in SLR2.

Based on the values shown in Table 3, the average removal efficiencies in the HRAP and in the fish rearing system were: (i) HRAP – 64.6% TKN; 77.0% AN; 22.9% TP; 78.2% COD; 1.31 log units E. coli; (ii) SLR3 – 49% TKN; 29% AN; 23% TP; 28% COD; 0.94 log units E. coli; (iii) SLR1 – 74% TKN; 68% AN; 57% TP; 46% COD; 1.25 log units E. coli; (iv) SLR2 – 65% TKN; 65% AN; 46% TP; 37% COD; 1.1 log units E. coli. The removal results in the HRAP (from both the experiments with the fingerlings and the juvenile) were in accordance with previous reports, like in Young et al. (2017). The reduction of chlorophyll-a concentration from the HRAP effluent to the fish rearing tanks (again, in both experiments) may be a result of phytoplankton consumption by the fish. It is noticeable that the chlorophyll-a values found in the experiment with juveniles were lower than those from the experiments with fingerlings, which may be explained by the higher plankton consumption by bigger fish in the second experiment, as well as by the lesser average radiation recorded during the second study (less 17%). Clearly, the final wastewater quality was improved by rearing juvenile tilapias, which fed on particulate organic matter, plankton and detritus present in the HRAP effluent; fish also controlled the breeding grounds for mosquitos by eating the mosquito larvae.

Table 4 shows the mean and standard deviation values for DO, pH and temperature of the wastewater in the wastewater treatment–fish rearing system during the second experiment. The water temperature was lower than in the first experiment, usually under 20 °C. According to the local database the minimum, average and maximum air temperature values during the experiment were 7.3, 17.5 and 31 °C respectively. DO and pH values were also lower than those from the experiment with fingerlings, probably as a result of the lower temperatures and lesser photosynthetic activity. The pH varied from 5.5 to 8 in the rearing tanks, with average values below 7; in the HRAP pH varied from 6 to 7.9. DO varied from 0.7 to 11.8 mg·L−1 in the rearing tanks, and from 4.4 to 10.7 mg·L−1 in the HRAP, with lower values occurring early morning whereas supersaturated concentrations were found in the afternoon.

Table 4

Average and standard deviation (within brackets) values of DO, temperature and pH in the HRAP and in the rearing tanks during the experiment with tilapia juvenile

Treatment or rearing unitDO (mg·L−1)Temperature (°C)pH
HRAP 6.8 (2.07) 18.2 (3.39) 7.1 (0.53) 
T1 (SLR1-D1) 6.0 (2.38) 19.8 (2.87) 6.6 (0.63) 
T2 (SLR2-D2) 4.7 (2.32) 19.7 (2.82) 6.7 (0.53) 
T3 (SLR2-D1) 6.1 (2.49) 19.8 (2.85) 6.7 (0.64) 
T4 (SLR2-D2) 5.0 (2.50) 19.8 (2.82) 6.8 (0.64) 
T5 (SLR3-D1) 4.9 (3.37) 19.9 (2.78) 6.9 (0.53) 
T6 (SLR3-D2) 4.9 (3.00) 19.9 (2.89) 6.9 (0.64) 
Treatment or rearing unitDO (mg·L−1)Temperature (°C)pH
HRAP 6.8 (2.07) 18.2 (3.39) 7.1 (0.53) 
T1 (SLR1-D1) 6.0 (2.38) 19.8 (2.87) 6.6 (0.63) 
T2 (SLR2-D2) 4.7 (2.32) 19.7 (2.82) 6.7 (0.53) 
T3 (SLR2-D1) 6.1 (2.49) 19.8 (2.85) 6.7 (0.64) 
T4 (SLR2-D2) 5.0 (2.50) 19.8 (2.82) 6.8 (0.64) 
T5 (SLR3-D1) 4.9 (3.37) 19.9 (2.78) 6.9 (0.53) 
T6 (SLR3-D2) 4.9 (3.00) 19.9 (2.89) 6.9 (0.64) 

There were no significant differences of weight gain between treatments. Very low weight gain was recorded in treatments SLR2-D1 and SLR3-D1 – up to 3 g over 12 weeks; no weight gain was recorded in SLR2-D2 and SLR3-D2, whereas there were even weight losses in SLR1 treatments. These findings may be explained by the low water temperatures which prevailed in these second experiments since, as stated by El-Sayed (2006), tilapia feeding is sharply reduced below 20 °C, and stops at about 16 °C, while severe mortality occurs at 12 °C. In an experiment using a four pilot-scale WSP series in South Africa, Wrigley et al. (1988) also reported that low winter temperatures and low DO concentrations during spring restrict production. Small African sharptooth catfish (Clarias gariepinus) stocked in the first three ponds grew rapidly in summer (relative daily growth rate of 5.9%), but during winter growth rate was reduced to 0.2% d−1. The fourth pond was stocked with a polyculture of Oreochromis mossambicus; Hypopthamichthys molitrix, Cyprinus carpio and Labeo umbratus. Again, during summer, O. mossambicus and C. carpio had average relative daily growth rates of 1.4 and 2.1%, respectively, while in winter both H. molitrix and L. umbratus lost weight (respectively, 0.2 and 0.3% of body weight·d−1) and C. carpio grew slower.

Even though during this experiment water temperature and pH values varied only narrowly, and fish weight gain was very low, plankton productivity still seems to have somehow favored fish growth: strong positive correlations were found between fish weight gain, water temperature (rs = 0.786), chlorophyll-a (rs = 0.843) and pH (rs = 0.629). No fish mortality occurred during the second experiment, probably because the sensitivity to ammonia is reduced at the more advanced growing stages of the fish; as observed by Benli & Köksal (2005) in ammonia toxicity tests with tilapia, larvae were less tolerant than fingerlings to the same levels of ammonia.

Tilapia fingerlings were highly sensitive to ammonia and short term temperature variation, which led to high mortality. On the other hand, tilapia juveniles withstood the ammonia effects and, although no mortality was recorded, the low temperature affected drastically the fish weight gain. The rearing tanks worked as polishing units and contributed to improving the effluent water quality in terms of nutrients, organic matter and E. coli. The results obtained herein suggest that, under the specific circumstances of these experiments (for instance, the prevailing climatic conditions), tilapia fingerlings culture using high rate algal pond effluent is feasible, during summer–autumn, with fish stocking density of 6 fish per m2 and ammonia surface loading rate of 1.2 kg·ha−1·d−1. The feasibility of tilapia juvenile rearing is still to be confirmed over warmer temperatures than those tested here.

The authors would like to thank the Brazilian agency CAPES for providing a postgraduate scholarship and research financial support.

Abdul-Rahaman
I.
,
Owusu-Frimpong
M.
&
Ofori-Danson
P. K.
2012
Sewage fish culture as an alternative to address the conflict between hunters and hunting communities in Northern Region
.
Journal of Agriculture and Sustainability
1
(
1
),
1
22
.
APHA (American Public Health Association), AWWA (American Water Works Association) & WEF (Water Environment Federation)
2012
Standard Methods for the Examination of Water and Wastewater
, 22nd edn.
APHA, AWWA and WEF
,
Washington, DC
.
Bastos
R. K. X.
,
Pereira
C. M.
,
Pivelli
R. P.
,
Lapolli
F. R.
&
Lanna
E. A. T.
2003
Utilização de esgotos sanitários em piscicultura (Use of domestic sewage in fish culture)
. In:
Utilização de Esgotos Tratados em Fertirrigação, Hidroponia e Piscicultura – Projeto PROSAB (Use of Treated Sewage in Fertirrigation, Hydroponics and Fish Culture – PROSAB Project)
(
Bastos
R. K. X.
, coord.).
ABES, RiMa
,
Rio de Janeiro
, pp.
193
221
.
Benli
A. Ç. K.
&
Köksal
G.
2005
The acute toxicity of ammonia on tilapia (Oreochromis niloticus L.) larvae and fingerlings
.
Turkish Journal of Veterinary and Animal Sciences
29
,
339
344
.
Edwards
P.
1992
Reuse of Human Wastes in Aquaculture, a Technical Review
.
The World Bank, UNDP-World Bank Water and Sanitation Program
,
Washington, DC
,
USA
.
El-Sayed
A.-F. M.
2006
Tilapia Culture
.
CABI Publishing
,
Wallingford, Oxfordshire
,
UK
.
El-Sherif
M. S.
&
El-Feky
A.
2008
Effect of ammonia on Nile tilapia (O. niloticus) performance and some hematological and histological measures
. In:
8th International Symposium on Tilapia in Aquaculture
, 12–14 October 2008,
Cairo, Egypt
, pp.
513
530
.
FAO, IFAD & WFP
2015
The state of food insecurity in the world 2015
. In:
Meeting the 2015 International Hunger Targets: Taking Stock of Uneven Progress
.
FAO
,
Rome
.
Fernandes
M. O.
&
Volpato
G. L.
1993
Heterogeneous growth in the Nile tilapia: social stress and carbohydrate metabolism
.
Physiology & Behavior
54
(
2
),
319
323
.
Hanna
M. I.
,
El-Maedawy
S. A.
,
Kenawy
A. M.
&
Girgis
S. M.
2013
Sublethal effects of acute ammonia exposure on Oreochromis niloticus
.
Global Veterinaria
11
(
5
),
592
603
.
Johnsson
J. I.
,
Winberg
S.
,
Sloman
K. A.
2006
Social interactions
. In:
Behaviour and Physiology of Fish: Volume 24 (Fish Physiology)
(
Sloman
K. A.
,
Wilson
R. W.
&
Balshine
S.
, eds).
Academic Press Elsevier B.V.
,
Amsterdam
, pp.
151
196
.
Mandal
R. N.
,
Chakrabarti
P. P.
,
Paul
B. N.
,
Chattopadhyay
D. D.
,
Das
A.
,
Hussan
A.
,
Jayasankar
P.
2018
Recycling of sewage in aquaculture: decadal technical advancement
. In:
Wastewater Management through Aquaculture
(
Jana
B.
,
Mandal
R.
&
Jayasankar
P.
, eds).
Springer
,
Singapore
, pp.
95
118
.
Mara
D. D.
,
Edwards
P.
,
Clask
D.
&
Mills
S. W.
1993
A rational approach to the design of wastewater-fed fishponds
.
Water Research
27
(
12
),
1797
1799
.
Moscoso
J.
1996
Aquaculture Using Treated Effluents from the San Juan Stabilization Ponds, Lima, Peru – Executive Summary
.
OPS/CEPIS
,
Lima
,
Peru
.
Mota
S.
,
Santos
E. S.
&
Santos
A. B.
2009
Water and fish quality during the reuse in aquaculture of sewage treated in stabilization ponds: experience of Ceará state, Brazil
. In:
8th IWA Specialist Group Conference on Waste Stabilization Ponds
, 26–30 April 2009,
Belo Horizonte, Brazil
.
Peña
M.
&
Mara
D.
2004
Waste Stabilisation Ponds
.
IRC International Water and Sanitation Centre
,
Delft, The Netherlands
.
Raychaudhuri
S.
,
Mishra
M.
,
Nandy
P.
&
Thakur
A. R.
2008
Waste management: a case study of ongoing traditional practices at East Calcutta Wetland
.
American Journal of Agricultural and Biological Sciences.
3
(
1
),
315
320
.
Silva
F. J. A.
,
Mara
D. D.
,
Pearson
H. W.
&
Mota
S. E.
2000
Informal fish culture in the Maracanaú waste stabilization ponds in Fortaleza, Brazil
.
Water Science and Technology
42
(
10–11
),
393
398
.
Wang
T.
,
Lefevre
S.
,
Huong
D. T. T.
,
Cong
N. V.
,
Bayley
M.
2009
The effects of hypoxia on growth and digestion
. In:
Hypoxia: Volume 27 (Fish Physiology)
(
Richards
J. G.
,
Farrell
A. P.
&
Braunes
C. J.
, eds).
Academic Press Elsevier B.V.
,
London
, pp.
361
396
.
Watanabe
W. O.
,
Losordo
T. M.
,
Fitzsimmons
K.
&
Hanley
F.
2002
Tilapia production systems in the Americas: technological advances, trends, and challenges
.
Reviews in Fisheries Science
10
(
3–4
),
465
498
.
WHO (World Health Organization)
2006
WHO Guidelines for the Safe use of Wastewater, Excreta and Greywater. Volume 3. Wastewater and Excreta Use in Aquaculture
.
WHO
,
Geneva
.
WHO (World Health Organization) & UNICEF (United Nations Children's Fund)
2017
Progress on Drinking Water, Sanitation and Hygiene: 2017 Update and SDG Baselines
.
WHO and UNICEF
,
Geneva
.
Wrigley
T. J.
,
Toerien
D. F.
&
Gaigher
I. G.
1988
Fish production in small oxidation ponds
.
Water Research
22
(
10
),
1279
1285
.
Young
P.
,
Taylor
M.
&
Fallowfield
H. J.
2017
Mini-review: high rate algal ponds, flexible systems for sustainable wastewater treatment
.
World Journal of Microbiology and Biotechnology
33
(
6
),
117
.