Introduction

The last decade has seen an increase in the implementation of biofloc technology (BFT) in the cultivation of tilapia and shrimp since it increases production at a low cost and reduces the environmental impact of aquatic wastewater discharges (Martínez-Córdova et al. 2014). BFT promotes microbial proliferation in culture units, such that they exploit the nitrogenous waste and recycle nutrients, improving the quality of the water throughout the culture process (Avnimelech 1999). Although the microorganisms sustain the system, information that refers to their composition and abundance is scarce (Browdy 2014).

Previously, studies of the microbial community of the biofloc have primarily focused on the composition and abundance of bacterial communities due to their function as matter degraders, as a source of food and/or probiotics (Kim et al. 2015; Nevejan et al. 2016). However, there are multiple microorganisms, in addition to the bacterial communities, which contribute to the system; for instance, the microalgae contribute oxygen and synthesise biomass from ammonia and carbon dioxide (Ebeling et al. 2006); the flagellates and ciliates control the bacterial biomass and favour the formation of floccules (Pandey et al. 2014); amoebas, rotifers and nematodes increase the precipitation of sedimentable solids (SS) and reduce the volume of mud. In addition, this microeukaryote community (MEC) is an important source of macro and micronutrients (Fiałkowska and Pajdak-Stós 2008; Loureiro et al. 2012). In particular, the composition and abundance of the MEC present in BFT systems have yet to be evaluated despite their important contribution to the efficient operation of the system (Manan et al. 2016).

Although it is important to evaluate the MEC in terms of composition and abundance to show that microorganisms prevail in the biofloc during the development of the aquatic culture, it is also necessary to identify both environmental and management variables that can influence the structure of the community (De Schryver et al. 2008; Ray et al. 2010b). This would establish the biological basis for obtaining cultures in biofloc with a consortium of specific microorganisms that would improve the performance of the species and the efficiency of the system (Browdy 2014).

The biofloc culture develops through two phases: the maturation or “start-up” phase and the culture phase. During the maturation phase and the first weeks of cultivation, a biofloc with a microalgae base is used (photoautotrophic). During the culture phase, the system is stocked with the objective species, nutrients are introduced and the addition of organic carbon provokes the transition of a microalgae biofloc to a bacterial biofloc (heterotrophic) (Hargreaves 2013). In both phases, the generation of microbial biomass and organic material increases the concentration of biofloc; however, when the biofloc concentration exceeds the maximum tolerable value for species in cultivation, it could put the culture at risk and should, therefore, be eliminated (Ray et al. 2010b).

The biofloc that is removed both during and at the end of the culture is rich in nutrients which, in accordance with previous reports, has a protein content between 25 and 50%, and a lipid content between 0.5 and 15% (based on dry matter) (Hargreaves 2013). The variability in these nutrients depends on diverse factors such as the carbon source, concentration of SS, size of the flocs, salinity, retention time of the biofloc and microbiota present (Azim et al. 2008; Emerenciano et al. 2011; Liu et al. 2016). Despite these variations, the biofloc is considered as a protein alternative to fish meal in aquatic feeds (De Schryver et al. 2008); therefore, it is important to evaluate changes in the proximate composition during the biofloc production with the purpose of recognising the conditions that maximise the nutritional value.

On the other hand, at the end of an aquatic culture with biofloc, there is an uncertainty of what to do with the water of the system; however, after the removal of the biofloc, the water can be either partially or totally reused in new biofloc cultures, this allows part of the microbial community to be reused and reduces the period of maturation, decreasing the consumption of water and prolonging its useful life (Ray et al. 2010a; Krummenauer et al. 2014). Despite the benefits of reusing the water from BFT systems, to date, there are no known studies about the microeukaryote community present in aquatic cultures of biofloc with reused water.

In view of the importance of the microorganisms for the function of the BFT system and the scarcity of information about the MEC present in the biofloc, the objective of the study was to evaluate the composition and abundance of the microeukaryote community during the cultivation of Nile tilapia in water-reusing biofloc systems. Moreover, the study sought to indicate environmental and management variables that affect the community and to evaluate the nutritional composition of the biofloc during the transition to a bacterial biofloc. The purpose of this study is to contribute to the knowledge of the BFT systems, particularly the knowledge of the microeukaryote community present and of the biofloc as a source of nutrients.

Materials and methods

The study was carried out in the Laboratory of Genetic Improvement and Aquatic Production of the Technological Institute of Boca del Rio, Veracruz, Mexico. The production of the biofloc was constituted of two phases: maturation (10 weeks) and culture (15 weeks).

Experimental design and system

A completely randomised design was used with two treatments: tilapia culture in biofloc (TB) and tilapia culture in biofloc with reuse water (RWTB), each treatment had three replicates. The experiment was conducted using two macrocosm-microcosm systems (Emerenciano et al. 2011), which were installed in a greenhouse (30 m × 10 m × 4.5 m) and covered with shade mesh (10% brightness). The systems consisted of one main culture tank or macrocosm (3.8 m2 in area × 0.8 m deep water) and three experimental units or microcosm (0.26 m2 in area × 0.8 m deep water), in addition to one settling tank (0.26 m2 in area × 0.9 m deep water).

The system for treatment TB was filled with water from an artisanal well (3.66 m3), and the system for treatment RWTB was filled with water originating from a semi-intensive culture of Nile tilapia with BFT (3.66 m3). Prior to the use of the reused water, it was allowed to sediment for 96 h and the supernatant was transferred to the experimental system. In both cases, the water was filtered through a 50 μm mesh opening. The initial characteristics of water for TB and RWTB were temperature (T), 31.45 ± 0.06 and 31.50 ± 1.16 °C; dissolved oxygen (DO), 4.54 ± 1.21 and 3.80 ± 0.80 mg L−1; total dissolved salts (TDS), 156 ± 4.55 and 1543.75 ± 11.26 mg L−1; pH, 7.55 ± 0.10 and 6.88 ± 0.05; NH4-N, 8.79 ± 0.42 and 7.98 ± 0.30 mg L−1; NH3-N, 0.03 ± 0.01 and 0.02 ± 0.01 mg L−1; NO3-N, 13.22 ± 0.13 and 14.49 ± 0.44 mg L−1; SS, 0.00 ± 0.00 and 0.20 ± 0.10 mL L−1, respectively.

The systems maintained constant aeration by air stones (23 cm × 2.5 cm × 2.5 cm) connected to aerators 3 HP (Sweetwater, Apopka, FL). In both experimental systems, the water lost by evaporation or sludge removal was replaced with water from the artisanal well. During the study, floc-rich water was maintained in recirculation, running from the main culture tank into the experimental units by means of a submersible pump (40 L min−1) and was returned by gravity.

Production and maintenance of the biofloc in the culture systems

During the first day of maturation, 0.15 mL L−1 of liquid fertiliser (Bayfolan Forte, Bayer Crop Science Division, Bayer, Morelos, Mexico; composition % w/w = total N 11.47%, potassium as K2O 6.00%, B 0.036%, Cu 0.040%, Fe 0.050%, Mo 0.005%, Zn 0.080%, thiamine hydrochloride 0.004%, phosphorus as P2O5 8.00%, S 0.23%, calcium as CaO 0.025%, Co 0.002%, Mn 0.036%, magnesium as MgO 0.025%, indoleacetic acid 0.003%) was added to the culture systems to stimulate primary productivity and zooplankton (Pérez-Fuentes et al. 2016). A week before the conclusive phase of maturation 150 g of molasses day−1 was added to promote the growth of heterotrophic bacteria.

The production of biofloc during the culture of Nile tilapia was maintained with a C/N ratio of 12.5:1 using molasses as a carbon source (40% C determined by elemental analysis). The quantity of molasses was estimated from the feed supplied to the fish (carbon content of the feed was 41% based on dry matter) and the protein percentage; in addition, it was assumed that protein contains 16% nitrogen and that the fish will excrete 70% of protein nitrogen (Avnimelech 1999). The molasses were divided into two fractions and were added every third day to avoid an abrupt reduction in the concentration of dissolved oxygen in the water (Pérez-Fuentes et al. 2016).

The biofloc was monitored every week, and when the SS exceeded 50 mL L−1, maximum concentration that tolerates Nile tilapia (Hargreaves 2013), it was extracted from the culture systems through the sedimentation tank (flow 3 L min−1) until a concentration of between 25 and 35 mL L−1 was obtained. The biofloc removed from both systems was concentrated in an independent tank (3 m3) and was maintained with aeration until the conclusion of the experiment; this biofloc waste was used to evaluate the proximate composition of the biofloc concentrate and was not representative of the biofloc cultures.

Cultivation of Nile tilapia

At the end of the maturation period, the system was stocked with Nile tilapia juveniles, Oreochromis niloticus (mean ± SD, weight: 79.28 ± 14.44 g, and standard length: 12.44 ± 0.70 cm), maintained at a density of 100 tilapia m−3. During the study, the tilapia were fed a balanced diet (El Pedregal Silver Cup, Toluca, México; protein 32%, lipids 5%) using a feeding rate equivalent to 2% of the biomass. Feed was supplied to the tilapia daily in two rations (9:30 and 15:30 h) and it was adjusted after each biometry.

Water quality

During the study, the physicochemical parameters of the water, temperature (T, °C), dissolved oxygen (DO, mg L−1), total dissolved salts (TDS, mg L−1), ionised ammonia (NH4-N, mg L−1), non-ionised ammonia (NH3-N, mg L−1) and nitrate (NO3-N, mg L−1), were recorded daily at 14:00 h, using multiparametric equipment (Professional Plus model, YSI, Yellow Springs, OH), in addition to the pH (EcoTestr™ pH 2). The following parameters were evaluated weekly: nitrite (NO2-N, mg L−1), general hardness (GH, mg L−1 as CaCO3) and carbonate hardness (KH, mg L−1 as CaCO3) by colorimetry (Hagen Nutrafin, Mansfield MA) and settleable solids (SS, mL L−1) with 1000-mL Imhoff cones after 20 min of sedimentation (APHA 1995).

During the culture phase, calcium hydroxide was used to increase the carbonate hardness (≥ 100 mg L−1 as CaCO3) and the pH (≥ 7.0) when these parameters decreased to values of KH < 80 mg L−1 and pH < 6.5 (Furtado et al. 2011). To do this, 1 kg of hydrated lime was diluted in 450 L of water, left to sediment for 2 h and the supernatant (150 to 200 L) was added to the main tank of the culture system during the removal of the SS.

Evaluation of the microeukaryote community

To evaluate the microeukaryote community, the microorganisms were classified as follows: microalgae, ciliates, flagellates, amoebas, rotifers, nematodes, annelids, copepods and cladocerans (Ray et al. 2010b). The presence (1)-absence (0) of each group of microorganisms per experimental unit was evaluated weekly, this was determined from a sample of water (10 mL) collected from the top 10 cm of the water, the sample was fixed immediately in formalin (4% v/v). Subsequently, a millilitre of the sample was placed in a Sedgewick Rafter camera and 30 random fields were analysed (1 mm2) with an inverted phase contrast microscope (Motic, AE31Elite, 20× and 40×) coupled to a photographic camera (Moticam 5). The identification of the genus of each group was conducted using taxanomic keys (Patterson and Hedley 1992; Foissner and Berger 1996; Abolafia 2006; Holovachov and De Ley 2006; Govedich et al. 2009; Wallace and Snell 2009; Jeuck and Arndt 2013; Bellinger and Sigee 2015).

Response variables

The abundance of microorganisms present in the biofloc was evaluated by group from the determination of the most probable number of organisms or cells mL−1 of sample using the following equation: \( NMP\ \left( org\ {mL}^{-1}\right)=\frac{D\bullet {A}_1}{\left({A}_2\bullet \left({V}_1+{V}_2\right)\right)} \); , where D = theoretical density (obtained value to transform the frequency of occurrence \( =\left(\frac{NOG}{NCO}\right)\times 100 \), by use of Fracker and Brischle (1944)) (Bellinger and Sigee 2015), A1 = = area of the counting chamber (mm2), A2 = area of the observed fields (mm2), V1 = volume of the sample (mL), V2 = volume of the preservative (mL), NOG = frequency of occurrence (∑ presence), NCO = total number of quadrants/number of observed quadrants. The ecological indices were estimated using the following equations: Shannon’s diversity \( \left({\mathrm{H}}^{\prime}\right)=-{\sum}_{i=1}^S\left({p}_i\ln {p}_i\right) \), Pielou’s evenness (J′) = H′/ ln S, Simpson’s dominance \( \left({\mathrm{D}}^{\prime}\right)=\sum {p}_i^2 \); , where pi is the relative abundance per group = abundance per group/total abundance of microorganisms (TAM), S = richness (∑ groups) (Bellinger and Sigee 2015).

Determination of the proximate composition of the biofloc

During weeks 14, 18 and 25, 150 g (wet weight) was taken from the biofloc of each experimental unit, and on the 25th week, a sample was taken from the concentration tank. The samples were obtained after the filtration of approximately 100 L of water from the culture through a 50 μm mesh opening. Each sample was dried at 60 °C for 24 h in a convection oven (Memmert UNP 400); subsequently, they were packed into pre-tagged bags and maintained in refrigeration at − 5 °C until analysis. The proximate composition (dry weight basis) of the biofloc was evaluated as follows: protein (%) was determined using an elemental analyser (Flash 2000, Term Fisher Scientific, UK) considering N × 5.9 as a conversion factor (Liu et al. 2016); total lipids (%) were analysed using the soxhlet method with methanol–chloroform, in a ratio of 2:1, as a solvent; ash (%) was analysed through incinerating the sample in a furnace muffler (Felisa, Jalisco, Mexico) at 600 °C for 2 h; nitrogen-free extract (NFE) (%) was determined by the difference of 100 with respect to the summative values of protein, lipids and ash, each analysis was triplicated (AOAC 1990).

Statistical analysis

A principal component analysis was performed to determine the physicochemical parameters of the water that influenced the TAM; in addition, the intensity of the relationship was evaluated through a Pearson correlation analysis. In the analysis, the variables that had been registered throughout the trial and that presented variance were considered. The TDS were considered as a supplementary variable for a better interpretation of the components.

The number of observed genera per treatment was contrasted with rarefaction curves based on the Mau Tao samples, and the number of estimated genera with the non-parametric estimators of the Chao2, Jacknkife1 and Jacknife2 richness with the program EstimateS 9.1 (100 randomizations without replacement) (Colwell 2013).

To determine the differences in the physicochemical parameters of the water, TAM, the indices H′, J’ and D’, and for the proximate composition of the biofloc, a one-way ANOVA was performed. Prior to the analysis, the assumptions of normality and homoscedasticity were verified with the Kolmogorov-Smirnov and Levene tests, where the assumptions were unfulfilled, the data was log10 transformed and the proportions were arcsine transformed. The a posteriori comparisons were performed with the Tukey test. All analyses were performed using a nominal significance of 5% (Zar 2010) with the Statistics 10 program (Stat Soft, Inc).

Results

Water quality and its relationship with the total abundance of microorganisms

The temperatures of the water in the two culture systems were maintained within similar ranges; however, during the 17th week, the water temperature dropped to about 20 °C. A progressive decrease in the NH4-N concentration was observed, which recorded the maximum value at the beginning of the maturation phase. The mean values of TDS, NO2-N, NO3-N, pH and KH recorded from the RWTB system were significantly greater than the mean values recorded from the TB system (p < 0.05) (Table 1).

Table 1 Physicochemical parameters of the water during the culture of tilapia Oreochromis niloticus in biofloc
Full size table

The SS in both systems exceeded 50 mL L−1 after the first 2 weeks of tilapia cultivation; therefore, the biofloc extraction was performed on the sedimentation tanks to reduce the value of SS to between 25 and 35 mL L−1. The SS were extracted every 2 weeks until the 20th week of the study; subsequently, the extraction of SS was performed weekly since values up to 65 mL L−1 were observed.

The variability of the physicochemical parameters of the water and the TAM in both systems was explainable by the first two principal components (Table 2). Principal component 1 (PC1) clustered together six variables, while principal component 2 (PC2) clustered together two variables, the supplementary variable was associated with PC1 (Fig. 1a, b).

Table 2 Variation of the principal components: relationship between physicochemical parameters of the water and total abundance of microorganisms of the tilapia culture in biofloc
Full size table
Fig. 1
figure 1

Principal component analysis to the related physicochemical parameters of the water and total abundance of microorganisms of the tilapia culture in biofloc TB (a) and tilapia culture in biofloc with reuse water RWTB (b). T = temperature, DO = dissolved oxygen, TDS = total dissolved salts, NH4-N = ionised ammonia, NO3-N = nitrate, NH3-N = unionised ammonia, SS = sedimentable solids, TAM = total abundance of microorganisms

Full size image

The TAM in both treatments demonstrated a highly significant relationship with SS, T, NH4-N and NH3-N, whilst showing a moderately significant relationship with NO3-N. The TAM in TB demonstrated a highly significant relationship with the TDS (Table 3).

Table 3 Correlation of the total abundance of microorganisms with respect to the physicochemical parameters of the water of the tilapia culture in biofloc
Full size table

Structure and diversity of the microeukaryote community

During the maturation phase, four groups of microorganisms were observed: microalgae, ciliates, flagellates and amoebas. During the culture phase, in addition to these microorganisms, rotifers, nematodes and annelids were present. At the beginning of the study, the microalgae of the TB treatment required 3 weeks to reach a relative abundance > 95% (Fig. 2a), while the relative abundance of the microalgae of the RWTB treatment was > 95% during almost the entire maturation phase (Fig. 2b). During the cultivation phase, in both treatments, it was observed that the microalgae and protozoa accumulated a relative abundance > 80%. The remaining relative abundance (≤ 20%) constitutes rotifers as the most abundant, in addition to annelids and nematodes (Fig. 2a, b). The TAM in RWTB (161.52 ± 80.60 org mL−1) was significantly greater than the value observed in TB (139.54 ± 81.39 org mL−1) (p < 0.01).

Fig. 2
figure 2

Relative abundance of the microeukaryote community of the tilapia culture in biofloc TB (a) and tilapia culture in biofloc with reuse water RWTB (b)

Full size image

The diversity indices of Shannon and Pielou’s evenness for the microeukaryote communities were similar between treatments (p > 0.05) (Table 4). The maximum diversity in both treatments was observed during week 15, whilst the maximum evenness value was recorded during week 14 in TB and in week 16 in RWTB. At the end of the study, a decrease in values of diversity and evenness was observed in both treatment groups (TB, H′ = 1.38 nats ind−1, J′ = 0.71) (RWTB, H′ = 1.14 nats ind−1, J′ = 0.58). Simpson’s dominance for the microeukaryote community in RWTB was significantly greater than that observed in TB (p < 0.05), the minimum value of the dominance index was observed during week 15 for both treatments.

Table 4 Group of microorganisms and ecological indices of the microeukaryote community present in the tilapia culture in biofloc
Full size table

The number of accumulated genera of microorganisms was higher in RWTB (74 genera) compared to TB (63 genera) (Fig. 3). The number of accumulated genera in TB (63) represented 89, 86 and 80% richness with respect to the non-parametric estimators Chao2, Jacknife1 and Jacknife2, whilst the number of genera accumulated in RWTB (74) represented 95, 89 and 89% with respect to the same estimators.

Fig. 3
figure 3

Accumulation curves of observed genera (Mau Tao) of the microeukaryote community in the tilapia culture in biofloc (TB) and tilapia culture in biofloc with reuse water (RWTB). The dotted lines represent 95% confidence intervals

Full size image

A total number of 79 genera distributed across the seven microorganism groups in the evaluation were identified (Table 5). Several microorganisms present in the biofloc were also identified in the water for a period ≥ 10 weeks. Therefore, 23 genera prevailed in the RWTB treatment while 13 genera prevailed in the TB treatment. The genera of microalgae that were present for the longest duration in the culture included Chlorella (19 weeks), Coelastrum (19 weeks), Dictyosphaerium (21 weeks), Monoraphidium (25 weeks), Pediastrum (17 weeks), Scenedesmus (25 weeks) and Nitzschia (22 weeks); in the amoeba group: Arcella (22 weeks) and Euglypha (20 weeks); and for the ciliates group the genera Paramecium (20 weeks) (Table 5).

Table 5 Generic distribution of the microeukaryote community present during the tilapia culture in biofloc
Full size table

Proximate composition of the biofloc

The proximate composition of the biofloc during the study varied between treatments and over time (p < 0.05) (Table 6). The moisture of the biofloc was similar between treatments. At the end of the study (week 25), the protein content of the biofloc had increased in both treatments when compared with values observed during weeks 14 and 18, and the value obtained in the biofloc concentrate. In contrast, the content of lipids in the biofloc of TB and RWTB in the initial sample (week 14) was significantly greater with respect to the value recorded during other time periods and in the biofloc concentrate. The concentration of ash in the biofloc of both treatment groups increased with time, but the percentage of ash in RWTB was statistically greater by the end of the culture period. The NFE concentration of the biofloc sample taken from the TB treatment during week 18 was significantly greater in comparison with the samples derived from RWTB and the biofloc concentrate.

Table 6 Proximate composition of the biofloc (mean ± SD, n = 3) from the culture systems and the concentrated biofloc during the study
Full size table

Discussion

Water quality and its relationship with the total abundance of microorganisms

During the study, a decline in ambient temperature provoked a reduction in the water temperature outside the optimum range for the Nile tilapia culture (28–30 °C) (Emerenciano et al. 2017); however, it is possible that the reduction in the water temperature assisted the formation of stable flocs (De Schryver et al. 2008).

The variation in pH and the alkalinity of the water in both treatments was due, in part, to the consumption of CO2 by microalgae for the synthesis of biomass, which according to Ebeling et al. (2006) provokes an increase in the pH of the water. This effect was observed on the pH of the water in treatment RWTB which increased from 6.8 to 8.1 during the maturation phase. The variation in the alkalinity and pH of the water also occurred due to the ratio of C/N (12.5:1) used and because of prolonging the removal of solids, conditions that lead to occurrence of nitrification processes (Liu et al. 2016; Xu et al. 2016). These processes involve nitrifying bacteria that use the alkalinity as a source of primary carbon (7.5 g of alkalinity as CaCO3/g N) and produce elevated levels of CO2 which can be problematic, as it reduces the pH of the water if it is not offset (Ebeling et al. 2006).

In the last weeks of the culture phase, the pH of the water was observed to decline from 7.5 to 5.9 in treatment TB; therefore, calcium hydroxide was added to stabilise the alkalinity (≥ 100 mg L−1 as CaCO3) and pH (≥ 7) of the water (Furtado et al. 2011). However, despite the addition of calcium hydroxide in the water at intervals of between 7 and 15 days, the increase in the biomass of the heterotrophic and nitrifying bacteria continued and did not allow the values of alkalinity and pH to be maintained. Therefore, the addition of alkaline composites should be made at shorter intervals of time (3 to 5 days) to maintain stability in the culture systems.

The PC1 clustered together six variables (T, NH4-N, NH3-N, SS, NO3-N and TAM), which were associated with the “biofiltration” that occurred in the culture systems. This controlled the ammonium through three routes: conversion to microalgae biomass, oxidation by nitrification (NH4-N → NO2-N → NO3-N) and transformation to bacterial biomass (Ebeling et al. 2006; Hargreaves 2013). The PC2 clustered together two variables (DO and pH), which were associated with the “environmental control” that is carried out continuously or periodically in the culture systems to meet the requirements of the species in culture (Hargreaves 2013).

The positive relationship of TAM and TDS in TB derived from the increase in dissolved salts in the water, which favoured the proliferation of microalgae. In addition, the inverse relationship of TAM with respect to T, NH4-N and NH3-N suggests that the decrease in the water temperature in the systems provoked an increase in the bacterial biomass, as a product of the assimilation of nitrogenous compounds, which increased the abundance of the MEC. Similar results were reported by Asaduzzaman et al. (2010) who observed an inverse relationship between the nitrogenous compounds and the abundance of the plankton with an increase in the C/N (20:1) ratio. Moreover, Luna-Pabello et al. (1992) reported that at temperatures between 20 and 30 °C under aerobic conditions, high proliferation rates of ciliates, amoebas, flagellates, rotifers and nematodes are present as a result of improved degradation of organic matter.

A positive relationship was observed in TAM and SS in both treatments throughout the duration of the experiment since the microorganisms that constitute the MEC are associated with the biofloc. During the study, the concentration of SS increased quickly, which could be related to a slight increase in the C/N ratio (~ 14:1) of the water caused by the carbon present in the feed. The SS exceeded the maximum recommendation of 50 mL L−1 for tilapia culture (Hargreaves 2013) on repeated occasions at the end of the study and its extraction of the culture tanks affected the structure of the MEC. According to Ray et al. (2010a), the extraction of SS, although it affects the community of microorganisms, favours biofiltration and reduces the concentration of NO3-N in the system.

The relationship of TAM with respect to DO and pH was not significant, it is probable that this result is due to the continuous control of these parameters in the culture systems, since an endeavour of the study was to maintain a concentration of DO > 4 mg L−1 and a pH ≥ 7. Emerenciano et al. (2017) reported that the values of DO and pH are adequate for meeting the requirements of tilapia and the microbiota present in BFT systems.

Structure and diversity of the microeukaryote community

By monitoring the MEC, it was possible to gain an understanding of its structure during the culture of Nile tilapia in biofloc with different water sources. The follow-up of the MEC through general microorganism classification saves time and is less susceptible to biases, their application is simple and, in addition, it permits the observation of changes that assist the decision-making processes that contribute to the optimization of the system (Ray et al. 2010b).

In this study, microalgae, ciliates, flagellates, amoebas, rotifers, nematodes and annelids were observed in both treatments; however, it is important to consider that the cultivated species in biofloc influences the presence of some groups of microorganisms. For instance, during the cultivation of prawn or shrimp in BFT systems, in addition to microalgae, protozoa, rotifers and nematodes, the presence of gastrotrich copepods and cladocerans was reported (Asaduzzaman et al. 2010; Manan et al. 2016). During this study, there was a clear absence of copepods and cladocerans, this observation can be attributed to these groups being a natural food source for tilapia and they are easily predated because of their larger size (Ibrahim et al. 2015). In addition to the species, the presence of different groups of microorganisms also depends on the geographical location of the culture and the endemic biota (Ray et al. 2010b).

The greatest concentration of TAM was observed in the treatment RWTB, a result that corresponds to the use of reuse water, which comprised microalgae, ciliates and amoebas on entry into the system. In treatment TB, a period of 3 weeks was necessary before microalgae predominated in the system; this result highlights the importance of using reuse water from BFT like inoculum when establishing new culture with BFT, which reduces the period of maturation of the water.

During the culture phase, the introduction of nutrients and the addition of molasses to the system favoured the growth of heterotrophic bacteria and the proliferation of amoeba, flagellates and ciliates, which, together with the microalgae, dominate the system. The fluctuation in the relative abundance of these groups is associated more with the management of SS than the presence of tilapia, a result that is different to observations made by Lara et al. (2016), who attributed the variation in the abundance of ciliates, flagellates and amoebas to the depredation of the species in culture and to the influence of the aeration equipment on the formation of stable flocs.

During the last 8 weeks of the study, the presence of nematodes and annelids was observed, whilst rotifers appeared during week 15 of the experimental period. The late appearance of these groups of microorganisms is related to the time required to increase the relative abundance of protozoa in culture systems, since these microorganisms, together with the bacteria, are part of the diets of nematodes and rotifers (Wallace and Snell 2009).

The ecological indices, Shannon’s diversity and Pielou’s evenness were similar in both systems but they varied with time. It was apparent that as the biofloc bacteria developed, the MEC tends to equilibrate and the groups of microorganisms exhibit a semi-constant relative abundance, giving rise to high indices of diversity (1.54–1.58 nats ind−1). In contrast, when the development of the biofloc is affected by the extraction of SS, the relative abundance of the groups of microorganisms display increased variability and a decreased index of diversity (1.14–1.38 nats ind−1), which causes some groups of microorganisms to become dominant, such as those observed at the end of the study, when the weekly removal of SS favoured the relative abundance of microalgae, ciliates and amoebas, and thus increases the index of dominance in the systems.

In this study, a total of 79 genera of microorganisms were identified, where RWTB display the greatest richness of genera (74) in comparison with the TB group (63), the advantage is associated with the biota being already established in the reuse water. In accordance with the non-parametric estimators, the observed richness in the systems reached a value greater than 80% with respect to the estimated value, which denotes a good representative sample. In both systems, the same pattern of genera by group was observed: microalgae > ciliates > amoebas > rotifers > nematodes = flagellates > annelids. This observation was different to that reported by Asaduzzaman et al. (2010) during the cultivation of Malaysian prawn in BFT, who identified 61 genera of microorganisms including microalgae, rotifers and small crustaceans, which followed the pattern: microalgae > crustaceans > rotifers. The difference between the studies was related to the culture species, integration of nutrients and the management of the system.

During the study, the microalgae genera Chlorella, Coelastrum, Monoraphidium, Pediastrum, Scenedesmus and Nitzschia were present for ≥ 19 weeks. The microalgae genera, except Monoraphidium, have been reported to be present for the longest period in Malaysian prawn culture with BFT (Asaduzzaman et al. 2010). Monroy-Dosta et al. (2013) reported that the ciliate genera, including Paramecium, Epistylis and Colpidium, and the rotifer Philodina predominated during the culture of tilapia with BFT. In contrast, in this study, the ciliates that were present for the longest period were Paramecium, Acineta, Aspidisca, Coleps, Litonotus and Phialina whilst the genera of rotifers included Lecane and Lepadella. The microorganisms that were present for the longest period in this study could be considered as possible candidates to establish cultures in biofloc with a consortium of specific microorganisms.

Proximate composition of the biofloc

In the present study, the protein content of the biofloc was estimated with a conversion factor of N × 5.9 in order to prevent an overestimation of the protein value of the biofloc composition due to the presence of nitrogen fractions that are not associated with protein, such as DNA (Liu et al. 2016). The protein content of the biofloc was similar to the value reported in studies where a factor of N × 6.25 was used (Azim et al. 2008; Emerenciano et al. 2011). The percentage of protein in the biofloc was observed to increase (31 to 47%, based on dry matter) when extraction of SS took place at short intervals (7 days), this result suggests that the longer the retention of the biofloc in the system the greater the decrease in the protein value of the biofloc (Liu et al. 2016).

It was established that the nutritional value of the biofloc is affected by its microbial composition due to the influence of the biochemical constitution of the microorganisms (Fugimura et al. 2014). Therefore, it is assumed that the protein in the samples during week 25 corresponds to the greatest bacterial source, compelled by the decrease in abundance of the groups that constitute the MEC through the removal of SS. The protein content in the biofloc concentrate reached 36%, this result demonstrated that whilst efficient aeration is maintained, it is possible to accumulate the biofloc and reduce the quantity of water, which facilitates the harvest for its later use as a protein input in aquaculture feeds (De Schryver et al. 2008).

The lipid content of the biofloc varied with culture time, high values of lipids were related to the establishment of microalgae in the biofloc during the maturation phase; however, during the transition to bacterial biofloc, the concentration of lipids decreased to values as low as 1.1% (based on dry matter) despite the increase in relative abundance of the microalgae group at the end of the study. On the other hand, in both treatments, an increase in the ash content of the biofloc was observed, possibly derived from the entry of calcium hydroxide to control the alkalinity and pH of the water; however, the percentage of ash in the biofloc was less than that reported by Lara et al. (2016) (34%) who attributed the result to excess faecal matter in the water column. The increase in the NFE content of the biofloc of TB treatment in week 18 results mainly from a lower ash content in the sample.

Conclusion

The RWTB treatment displayed the greatest richness of genera and abundance of microorganisms due to the water-reusing biofloc systems. The physicochemical parameters with the most influence on the TAM were SS, T, NH4-N and NH3-N, where the extraction of SS exerted the greatest influence on the structure of the microeukaryote community. In the study, 79 genera of microorganisms were distributed between the groups: microalgae, ciliates, flagellates, amoebas, rotifers, nematodes and annelids; some of which remained in culture for a period exceeding 16 weeks. The nutritional composition of the microorganisms which constitute the MEC, in addition to the concentration and the time interval of SS removal affect the proximate composition of the biofloc. Further investigation is necessary to improve knowledge of the microeukaryote community present in BFT systems and regarding the factors that may have repercussions on the proximate composition of the biofloc.