INTRODUCTION

Urbanization, industrial development and agricultural production carried out without environmental knowledge can have a great impact on water resources. The over-exploitation of water bodies and the deterioration of water quality are global trends ^{(Gutterer et al. 2009)}.

According to ^{Peña and Mara (2004)}, waste stabilization ponds (WSP) are usually the most appropriate method for domestic and municipal wastewater treatment in developing countries, where the weather conditions are favorable for the operation of the ponds. These ponds present high efficiency and totally natural and highly sustainable operations. They require low investment, low operating and maintenance costs, and minimal energy consumption ^{(Sah et al. 2012)}.

WSPs are typically shallow ponds and include aerobic, facultative, and anaerobic zones, allowing for the settling of suspended solids, degradation of organic matter, and pathogen removal by natural processes ^{(Inglis et al. 2021)}. Although polishing ponds also carry out complementary organic matter removal, they receive very low organic loads. However, based on the pretreatment performance, the main role of the ponds is the removal of pathogens (referred to here as maturation ponds) and nitrogen; design criteria for polishing ponds are very similar to those used for maturation ponds ^{(von Sperling 2007)}.

Microalgae play a fundamental role in wastewater treatment, due to the production of oxygen through photosynthesis, which becomes available to bacteria for organic matter decomposition; microalgae also contribute, directly and indirectly, to nutrient removal processes (volatilization, precipitation, and assimilation) and pathogen reduction ^{(Bento et al. 2005)}.

According to ^{Barthel (2007)} and ^{Uehara and Vidal (1989)}, zooplankton is also an important component in the dynamics of the aquatic environment. Rotifers, cladocerans and copepods mainly occur in secondary and tertiary ponds; their presence contribute to the treatment process, once they can clarify the effluent when feeding on bacteria, microalgae, protozoa and organic waste. At high densities, these organisms impact the oxygen balance, due their herbivory, once the algae content decreases as the respiratory demand increases.

Studies on the structure and dynamics of phytoplankton are important, as organisms that respond promptly to environmental change, working as ecological indicators and increasing knowledge about existing interactions between physical processes, on the one hand, and biological responses, on the other. Knowledge about planktonic organisms in stabilization ponds can improve system management. The present research mainly sought to conduct biomass analysis and investigate the time variations and composition of the planktonic communities (phytoplankton and zooplankton) in four polishing ponds used for wastewater treatment, through the qualitative and quantitative analysis of these organisms.

MATERIALS AND METHODS

STUDY AREA

The study was performed in the Integrated Unit for Sewage Treatment and Effluent Usage, located in the city of Viçosa, Minas Gerais, geographical coordinates 20°45´14´´ S and 42°52´54´´ W.

Pretreatment was performed by an upflow anaerobic sludge blanket (UASB) reactor and a submerged aerated biofilter, in full-scale (115 m³/d), made up of prefabricated steel units, followed by a series of four polishing ponds (P1, P2, P3 and P4), all in pilot-scale prefabricated fiberglass, with 16.3 m² individual surface area, 0.90 m average useful depth and 2.0 lenght/width ratio (Fig. 1).

The study was conducted from February 2009 to June 2010, interrupted in September, November and December, due to operational problems. Throughout the study, the pond system was operated at two different hydraulic retention times (HRT), in two operational periods: from February to October 2009 – Period I – the pond flow rate was 3.5 m³/d, with HRT of four days for each pond; from January to June 2010 – Period II – the flow rate was 2 m³/d, with a HRT of seven days for each pond.

Figure 1. Schematic illustration of the experimental series of polishing ponds 

RESULTS

BIOLOGICAL VARIABLES

Chlorophyll-a

The concentrations measured for chlorophyll-a during the two research periods showed higher values for Period II (Fig. 2).

The maximum measured concentrations of chlorophyll-a were 554.5, 453.9, 465.8 and 385.0 μg/L, respectively, in ponds P1, P2, P3 and P4; the mean values and the standard deviations were 200.3 ± 150.5, 222.2 ± 123.5, 149.8 ± 115.4 and 99.8 ± 120.6 μg/L; the minimum concentration values were 30.0, 69.7, 39.3 and 4.6 μg/L (Fig. 2). A pattern of variation similar to that found in TSS was observed, which corroborated the effect of microalgae on the variation of TSS values.

Phytoplankton

Throughout the study period, 103 taxa, divided into seven classes were found, namely, Chlorophyceae (53), Euglenophyceae (21), Cyanobacteria (18), Cryptophyceae (6), Zygnemaphyceae (2), Chrysophyceae (2) and Dinophyceae (1). Table 1 presents the taxa count in the ponds during the study period.

Total cell density in the series of ponds reported wide variation (2647 cell/mL to 151678 cell/mL), represented by four numerically representative classes: Chlorophyceae, Euglenophyceae, Cyanophyceae, and Cryptophyceae.

Zooplankton

During the research, the zooplankton community was composed of ten genera, and group Rotifer was the richest, with six identified genera. Cladoceran and Copepod were also inventoried. In addition to the organisms, young stage Copepod individuals were observed (nauplii and copepodites), besides Insecta larvae. Table 2 presents the zooplankton community found in the series of ponds during the research.

Figure 2. Variation of chlorophyll-a values in the series of ponds for the two operational periods. 

From February of the first year to June of the second year, Nauplii occurred more frequently, followed by the genus Thermocyclops and insect larvae. Colurella, Mesocyclops and Lecane were the least frequent genera, and the latter was only found in P4 (Table 3).

Table 1. Total taxa count in the ponds during the study period, divided into phytoplankton classes. 

Table 2. Zooplankton community found in the series of ponds. 

Table 3. Abundance (%) of zooplankton, identified in each polishing pond, during the research. 

PRINCIPAL COMPONENTS ANALYSIS

In the PCA for the ponds, the number of factors with eigenvalues > 1 that were identified during the first operational period was two, three, four, and three respectively for maturation ponds P1, P2, P3, and P4; for the second operational period, the number of factors was three, three, two and three for ponds P1, P2, P3 and P4.

In the eight principal component analyses performed, it was observed that the first two eigenvalues explained a high percentage of the total variability of the data. For the first operational period, the first two eigenvalues explained 65.7%, 67.9%, 57.9%, and 61.1% of the variability for the P1, P2, P3, and P4 ponds, respectively. By the second operational period, these values accounted for 68.8%, 60.6%, 74.4%, and 76.8% of the variability for maturation ponds P1, P2, P3, and P4, respectively.

For the first operational period, the PCA indicated that there was a high positive correlation between the variables: total phytoplankton, BOD, TSS, and chlorophyll-α in every one of the polishing ponds in the series. On the other hand, there was a high correlation between the variables: total zooplankton, transparency, DO, pH, and temperature; however, for the P3 and P4 ponds, there was also a slight correlation of total zooplankton concerning the first group of variables mentioned above.

The PCA for the second operational period again revealed that there was a high correlation between the variables in total phytoplankton, BOD, TSS, and chlorophyll-a in all polishing ponds. In addition to this, there was a high correlation between the variables of total zooplankton, transparency, DO, pH, and temperature; however, for the P3 pond, there was also a slight correlation of total phytoplankton with transparency.

The PCA with emphasis on phytoplankton organisms yielded the following results: in the P1 pond for the class Euglophyceae, a positive correlation was recorded concerning BOD and alkalinity, as well as a negative correlation with relation to temperature and DO; for the classes Cyanophyceae and Cryptophyceae there was no correlation with the parameters analyzed; for the class Chlorophyceae a positive correlation was recorded about TSS, COD and chlorophyll-a, as well as a negative correlation with transparency. In the P2 for the class Chlorophyceae, a positive correlation was observed with TSS and COD, as well as a negative correlation with temperature; for the classes Euglenophyceae and Cyanophyceae, no correlation was recorded with the parameters analyzed; for the Cryptophyceae class, there was a positive correlation with alkalinity, BOD, chlorophyll-a, pH, as well as a negative correlation concerning DO. In the P3 for the classes Cyanophyceae, Cryptophyceae and Chlorophyceae, a positive correlation was recorded concerning BOD, chlorophyll-a, TSS, pH and alkalinity, as well as a negative correlation with temperature and transparency; for the class Euglenophyceae there was no correlation concerning the parameters analyzed. In the P4 for the classes Chlorophyceae and Cryptophyceae, a positive correlation was observed about chlorophyll-a, BOD, alkalinity, TSS, and a negative correlation concerning transparency; for the class Euglophyceae there was a positive correlation concerning COD; finally, for the class Cyanophyceae, a positive correlation was recorded concerning pH and DO, as well as a negative correlation with COD.

DISCUSSION

According to ^{Nogueira and Matsumura-Tundisi (1996)}, climate factors, such as precipitation and wind, greatly affect plankton composition and biomass. In this study, a significant inverse correlation between precipitation and the total density of phytoplankton was verified only in P2.

In this study, the lowest Secchi transparency levels were associated with a great amount of suspended material in the water column, including phytoplankton. It was evidenced by the negative correlation between water transparency and TSS in the four studied ponds, namely, -0.646 in P1, -0.651 in P2, -0.648 in P3, and -0.629 in P4. The higher the transparency, the greater the light penetration in the entire water column, which increases photosynthetic activity, oxygen production and pH.

The water quality results revealed an alkaline condition in the water column, with a slight tendency to acidification at the bottom. Broadly speaking, the pH variation pattern was similar to the DO variation pattern, with an increasing tendency from the first to the last pond, in sequence. High concentrations of chlorophyll-a and TSS, as well as lower transparency in the water, are often the result of high densities of phytoplanktonic organisms, which have a greater capacity to survive in environments with high concentrations of BOD, while many zooplanktonic organisms do not. Some extreme values of DO and temperature are due to the timing of measurement of the parameters in situ, early in the morning, due to lower night temperatures, the absence of photosynthetic processes, and nocturnal respiration by planktonic organisms was observed. Likewise, as reported by ^{Pham et al. (2020)}, organic matter exerts a strong influence on DO concentrations, since the higher the loads of this pollutant, the greater the consumption of the dissolved gas by the heterotrophic bacteria of the stabilization ponds.

The photosynthesis performed by microalgae in stabilization ponds contributes to pH elevation, which can reach values over 9.0 in the event of high photosynthetic activity ^{(von Sperling, 2007)}.

Decreased alkalinity was generally observed during the treatment process. The mean values and standard deviations of alkalinity, calculated for ponds P1, P2, P3, and P4 were, respectively, 135.0 ± 53.3, 104.6 ± 53.9, 83.6 ± 40.2 and 70.9 ± 33.5 mg CaCO₃/L.

Increased chlorophyll-a concentrations indicate higher biomass density of green algae, confirming the photosynthetic activity of microalgae. The high primary productivity also contributes to the increase in suspended solids, as well as to the presence of organic debris, to the resuspension of sediments from the ponds.

During the study period, the highest number of taxa was found in P3 (91), followed by P2 (87), P4 (81) and P1 (73). Most organisms belonged to the class Chlorophyceae, which reveals the highest number of taxa in P2 and P3 (48 taxa each). The Euglenophyceae class presented the highest number of taxa in P1: 19 records, gradually decreasing towards P4, where 13 taxa were registered (Table 1). Classes Zygnemaphyceae, Dinophyceae, and Chrysophyceae provided the lowest number of taxa, from which organisms of class Chrysophyceae were only registered in P3 and P4.

In a phytoplankton survey conducted in a stabilization pond located in Sao Jose dos Campos, Sao Paulo-Brazil, ^{Cardoso (1979)} found that classes Chlorophyceae and Euglenophyceae were the most common, while the genera with the highest densities were: Chlorella, Golenkinia, Euglena, Oscillatoria, Merismopedia, Chlamydomonas, Scenedesmus and Monoraphidium. All of them were observed at the polishing pond system evaluated in this research. Merismopedia and Oscillatoria were identified by ^{Cruz et al. (2003)} in facultative ponds, which are commonly characterized by an environment rich in organic matter.

A positive correlation between cyanobacterial density with BOD, chlorophyll-a, and temperature.

The class Cyanophyceae is represented by organisms that are well-adapted to extreme conditions. Therefore, they are highly proliferative in eutrophic ponds. Eight abundant species were found, and only Aphanocapsa delicatissima was dominant in the count of May of Period I (7784 cell/mL). ^{Reynolds (1996)} described this genus as well adapted to calm waters, with moderate concentrations of nutrients. That might explain the fact that Aphanocapsa delicatissima has contributed the highest density in low precipitation periods. Rain might be a disturbing agent of the structure of the phytoplankton community, thus contributing to the reduced density of this class in the rainy season due to dilution, as reported by ^{D'allasandro et al. (2020)}. For the class Cyanophyceae, unlike studies such as those by ^{Chorus and Bartram (1999)} and ^{Komarek and Anagnostidis (2008)}, no correlation was observed in the P1 concerning BOD, chlorophyll-a, temperature, DO, pH or COD, according to ^{Paerl and Otten (2013)}, this could have been due to specific conditions of the pond such as pH variation, temperature, nutrients, as well as predation, competition with other algae or specific characteristics of the cyanobacterial community.

^{Goncalves (2008)} characterized the phytoplankton community in maturation ponds of sanitary sewage treatment system in Petrolandia, Pernambuco state, Brazil. The phytoplankton was composed of 20 taxa, mainly Cyanophyceae Oscillatoria sp. and Chlorophyceae Closteriopsis acicularis, both present in most samples; in this research, these taxa were recorded in ponds P2, P3 and P4.

The class Cryptophyceae presented important variations in the densities during the sampling period, from a minimum of 23 cell/mL in P3 (February of Period I) to a maximum of 14814 cell/mL in P1. There was a clear time separation, with low densities in the months of the first period of the research, and higher numbers over the second period. Cryptophyceae use available organic matter as an energy source and are abundant in environments with high organic load; ^{Im et al. (2023)} stated that Cryptophyceae showed significant correlations with BOD and total organic carbon in the artificial Lake Uiam, South Korea.

While investigating the effect of HRT and surface application rate on the composition of phytoplankton community in polishing ponds on a pilot scale, ^{von Sperling and Oliveira (2010)} verified the dominance of classes Chlorophyceae and Euglenophyceae. A decrease was observed in order Chlorococcales and genus Cryptomonas, due to reduced HRT. In the present study, statistical results pointed to significant differences in the density of class Cryptophyceae during the research. Low density was coincident with the period when the system was operated with a shorter hydraulic retention time (four days).

The lowest densities were observed in the classes Dinophyceae (38 to 506 cell/mL) and Zignemaphyceae (37 to 1502 cell/mL). Organisms from class Dinophyceae were only found in three collections (October of Period I and January and March of Period II). Class Chrysophyceae was only found in qualitative sampling, represented by two taxa (Dinobryon sertularia and Synura sp.).

To achieve greater accuracy in the comparison of the ponds, the density per class data were submitted to ANOVA, followed by the Tukey mean comparison test. The results were interpreted taking a 5% significance level reference. The density values of classes Cyanophyceae and Cryptophyceae differed during the two periods of the study; no significant differences were observed between classes Chlorophyceae and Euglenophyceae. The difference was observed between the four ponds only in class Euglenophyceae. No other class expressed significant differences from one pond to the other.

The zooplankton community plays an essential role in the aquatic system dynamics due to its characteristic of being a link in the food chain, once they feed on algae and serve as food for more complex organisms, such as fish.

In general terms, the zooplankton found in the series of ponds presented low density, with total numbers of 310 org/mL in P1, 2436 org/mL in P2, 1077 org/mL in P3, and 513 org/mL in P4.

The zooplankton community presented time variation over the study period, with the disappearance of some organisms. In P1, zooplankton was detected in 55% of the samples. In August of Period I, there was a peak in P2, mainly due to high Rotifer density (1049 org/mL).

The absence of cladocerans throughout the sampling period was indicative of the eutrophic environment.

Eutrophication is an event capable of producing deep transformation in the zooplankton community, which is associated with the replacement of palatable forms of phytoplankton by less-savoring ones, such as filamentous cyanobacteria. Thus, changes in the food chain cause the "bottom-up" effect on zooplankton and modify the community composition.

The study conducted by ^{Silva and Matsumura-Tundisi (2002)} demonstrated the sensitivity of the Copepod community to changes in water quality, which helps to explain the lower densities of these organisms in P1 (operated with higher organic loads).

The life cycle of Copepods is known to be long if compared to other zooplankton components (Rotifer and Cladoceran), which makes them susceptible to a high mortality rate before reaching the adult stage. In this research, Copepod exhibited low density, which ranged from 0.1 to 74.1 org/mL.

^{Moscoso et al. (1992)} observed intense proliferation of ciliate, rotifers, copepods, and cladocerans in stabilization ponds in Peru; ^{Guerrin (1988)} and ^{Nandini (1999)} registered populations of Rotifer and Cladoceran with respective values of 3000 and 300 organisms/L density in stabilization ponds.

The predominance of Rotifer could be related to its ability to reach maturity earlier, and to have a faster replacement rate if compared to microcrustaceans ^{(Nogueira and Matsumura-Tundisi 1996)}.

In the studied polishing ponds, the Rotifer density ranged from 0.1 to 1028.6 org/mL; the maximum density was registered in August of the first year in P2.

Rotifer and Cladoceran are particularly able to grow at high densities, feeding on organic waste and bacteria ^{(Roche, 1998)}, ^{Arevalo-Stevenson et al. 1998}.

While studying the same polishing pond system, ^{Bastos et al. (2010)} found five genera composing the zooplankton community, four of which were identified in this research: Brachionus, Moina, Mesocyclops, and Thermocyclops. Only Daphnia was not observed.

In the present study, no pattern was found to explain the behavior of zooplankton in the series of ponds. As these organisms depend mainly on DO in water for its maintenance, this may have probably been one of the limiting factors in the community, especially in P1. However, as the distribution of zooplankton is affected by several other factors, additional research is needed to elucidate doubts and help to understand the dynamics of these organisms in WSP; furthermore, knowledge of planktonic community fluctuations can assist in operation management and maintenance of the ponds through the anticipation of certain harmful phases, such as the cyanobacteria domain ^{(Amengual-Morro et al. 2012)}.

CONCLUSIONS

In conclusion, the evaluated polishing pond system was efficient in reducing both particulate and organic matter, which resulted in effluents with average concentrations of 62.4 mg/L (TSS), and BOD concentrations ranging from 4.1 to 37.5 mg/L. Most of the taxa identified in phytoplankton belong to classes Chlorophyceae (53), Euglenophyceae (21) and Cyanophyceae (18), besides Cryptophyceae (6), Zygnemophyceae (2), Chrysophyceae (2) and Dinophyceae (1). The total cell density of phytoplankton in the series of ponds presented a wide range (2647 cell/mL to 151678 cell/mL), with four numerically representative classes: Chlorophyceae, Euglenophyceae, Cyanophyceae, and Cryptophyceae. Class Euglenophyceae was associated with high levels of organic matter, mostly present in P1 and P2, and only one species of class Cyanophyceae was dominant during the study period (Aphanocapsa delicatissima), in the sampling from May of Period I. The zooplankton community presented low densities, and the group Rotifer was the most representative. No correlations were found between phytoplankton and zooplankton.