Seasonal dynamics and diversity of cyanobacteria in a eutrophied Urban River in Brazil

The Jundiai River, located in the Brazilian State of São Paulo, is environmentally and historically significant to the region. The intense development of anthropic activities (i.e., urbanization, industry, and agriculture), municipal and industrial effluent discharge, urban and rural runoff, and lack of proper vegetative protection of the banks have caused deterioration in the water quality. As a result of water pollution, the presence of cyanobacteria has been observed next to the Jundiai River basin outlet. In a regional context of water scarcity, the Jundiai River has become a strategic source for several municipal water supply systems; therefore, the knowledge of the seasonal dynamics of cyanobacteria is essential for proper water treatment and water safety.

Massive cyanobacterial development in water is associated with organoleptic concerns, toxicological issues (Newcombe et al. 2010), and poor water quality (Kuo et al. 2018). Nutrient availability, although directly related to cyanobacterial development, does not act in isolation for biomass growth, as other factors such as light, temperature, and water velocity are also essential in the growth process (Cunha et al. 2017). Tropical water bodies experience certain peculiarities, such as intense solar radiation and high water temperatures, which encourage phytoplankton nutrition absorption and enable high nutrient assimilation (Ludolf Gomes et al. 2012).

Cyanobacteria proliferation in rivers is a warning sign because it indicates deterioration in water quality. Evidence indicates an exponential increase in the abundance of cyanobacteria in subtropical aquatic ecosystems of South America, primarily driven by nutrient enrichment associated with the change in land-use patterns (Kruk et al. 2023). Increased nutrient concentrations associated with watershed fertilizer runoff, sewage discharge, reduced flow due to the installation of artificial structures, and increased water temperatures are often the causes of this disturbance (Park et al. 2021). Particularly in the subtropics, for example, the State of São Paulo, Brazil, riverine ecosystems are becoming increasingly susceptible to cyanobacterial development as temperatures rise owing to climate change (Haakonsson et al. 2017).

Periodic monitoring of cyanobacteria levels in water systems is critical to prevent and manage potential health risks associated with toxin exposure. This is particularly important in Brazil, where a large section of the population depends on surface water sources for water supply. In addition, currently, insufficient emphasis is being placed on monitoring and counting cyanobacterial diversity in river water systems that possess the potential to serve as a public water supply source. Therefore, this study aims to identify the seasonal dynamics and species diversity of cyanobacteria on Jundiai River (São Paulo, Brazil) and explore the relation of climate, water quality, and hydrological parameters with cyanobacteria abundance. This research provides valuable data for current and future users of the water of the Jundiai River, as well as contributes to a broader understanding of cyanobacterial dispersion globally.

The data on freshwater variables were collected from December 2018 to December 2022, according to the schedule in Table 1. The water sampling procedure followed rigorous guidelines, including laboratory internal Standard Operating Procedures (SOP 238, 230, and 239), Brazilian technical standards (ABNT NBR 9898/1987), and the Standard Methods for Examination of Water and Wastewater (23rd Edition 2017, Methods 1060 and 9060). These procedures were selected based on specific analysis objectives. The sampling was conducted at a similar schedule, approximately around noon, to ensure the reliability and consistency of the collection process.

Table 1

Sampling schedule by freshwater variable (2018–2022)

^aAmmoniacal nitrogen, biochemical oxygen demand (BOD_5,20), chemical oxygen demand (COD), color, dissolved oxygen, dissolved aluminum, nitrogen-nitrate, nitrogen-nitrite, pH, total dissolved solids, total phosphorus, turbidity, thermotolerant coliforms, total zinc, water temperature.

View Large

The water quality parameters included one monthly sampling of cyanobacterial species diversity and total cell counting (CYAN cell mL⁻¹), chlorophyll-a (Chl-a), and other relevant parameters (such as ammoniacal nitrogen, biochemical oxygen demand (BOD_5,20), chemical oxygen demand (COD), color, dissolved oxygen (DO), nitrogen-nitrate, nitrogen-nitrite, pH, total dissolved solids, total phosphorus (TP), turbidity, thermotolerant coliforms, and water temperature). All analyses were performed in an ISO/IEC 17025 accredited laboratory that conforms to the general standards for testing and calibration laboratory competence. The laboratory methods for all the analyses can be found in the Supplementary Material.

The total number of samples obtained for each parameter were as follows: 35 (DS: 23; WS: 12) for cyanobacterial diversity, 41 (DS: 26; WS: 15) for cell counting, 43 (DS: 26; WS: 17) for chlorophyll-a, and 15 (DS: 10; WS: 5) for other water quality parameters. Furthermore, climatological variables including daily maximum, minimum, and mean temperatures, monthly precipitation, precipitation within the last 24 and 72 h, and hydrological parameters (i.e., flow rate) were collected. The climatic data were obtained from CIIAGRO (2023), while the hydrological data were acquired from DAEE (2023).

Descriptive statistics tools were used to organize, summarize, and characterize the data set. The seasonal variations were represented by box-and-whisker plots for the cyanobacteria cell counting and physicochemical parameters. The analysis of variance (ANOVA) was used to study the statistical significance of water quality metrics and cyanobacteria indicators across seasonal assessment (i.e., dry and wet seasons). Based on the data features and aiming to explore the correlation between the environmental factors (such as water quality parameters, climate variables, and flow rate) and cyanobacteria abundance, Spearman's rank correlation coefficient (ρ) was calculated. Univariate linear regression analysis with a confidence level limit of 95% (p-value < 0.05) was used to determine statistically significant relationships. For discussion purposes, p-values up to a confidence level of 10% (p-value < 0.1) are supplied. To minimize leverage in correlation and significance analyses, the outliers identified during preliminary data analysis were excluded. Outliers were identified through visual analysis of the data distribution and creation of scatter plots while considering the use of z-scores and modified z-scores, which were ultimately deemed unnecessary given the prominent deviations discernible through visual examination. All data visualization and statistical analyses were executed using the R programming language on RStudio (© 2009–2022 RStudio, PBC).

The presence of cyanobacteria cells was detected in all samples; however, the months of January and February had the lowest cell counts, with 339 cells mL⁻¹ in January 2021, 215 cells mL⁻¹ in February 2021, and 370 cells mL⁻¹ in February 2020. The cyanobacteria cell counts showed a clear tendency for cell growth throughout the DS of the year, particularly between May and August (Figure 3). A significant difference (p-value < 0.05) in the concentration of cells is observed between the DS and WS, with averages of 5,722 and 2,486 cells mL⁻¹, respectively.

Throughout the year, Pseudanabaena sp. was consistently detected in the Jundiai River, while in nine other months, Raphidiopsis sp. (renaming from Cylindropermopsis) (March–August and October–December), Aphanocapsa sp. (February–May and July–November), and Geitlerinema sp. were also identified.

In general, at least 10 of all 17 reported species are potential producers of a wide range of toxins. Among those, Aphanocapsa, Dolichospermum, Oscillatoria, Phormidium, Pseudanabaena, Merismopedia, and Microcystis are microcystin-genera-producers (Chorus & Welker 2021). Furthermore, Dolichospermum and Raphidiopsis produce cylindrospermopsin (Chorus & Welker 2021). Dolichospermum, Phormidium, Planktothrix, and Cuspidothrix are producers of anatoxin-a (Chorus & Welker 2021). In addition, Raphidiopsis and Planktothrix also produce saxitoxin (Chorus & Welker 2021). Nevertheless, all cyanotoxin samples collected during the entire monitoring period exhibited concentrations below the detection limit (<0.1 μg L⁻¹ for microcystin and cylindrospermopsin; < 0.08 μg L⁻¹ for saxitoxin). Consequently, despite the presence of cyanotoxin-producing strains, the water of the Jundiai River did not exhibit detectable levels of microcystin, saxitoxin, or cylindrospermopsin.

During December and August, Raphidiopsis sp. was the dominant species, comprising >50% of the total number of cells in the monthly samples analyzed (77.05 and 58.35%, respectively). Notably, in the samplings conducted in October and December of 2022, this species accounted for 100% of the cells detected. Raphidiopsis sp., a cylindrospermopsin-producing cyanobacteria, was the most prevalent species during both sampling periods, accounting for 23.49 and 33.68% of total cell abundance during the DS and WS, respectively (Table 2).

On the DS, the Pseudanabaena sp. cells constituted 20.39% of the total cell count, while in the WS, this percentage decreased to 14.31%. Pseudanabaena organisms, which were the second most abundant cyanobacteria during both DS and WS, are able to synthesize toxins (Chorus & Welker 2021) and are geosmin and 2-methylisoborneol (MIB)-producing species (Chorus & Welker 2021).

Geosmin and MIB are odorous compounds, which can be perceived even at concentrations as low as nanograms per liter (ng L⁻¹) (Chorus & Welker 2021) and detected in small numbers (∼500 cells mL⁻¹) of cyanobacteria cells (Newcombe et al. 2010). In contrast to cyanotoxins, taste and odor compounds are not directly associated with human health risks (Chorus & Welker 2021). However, the sensory perception of water by the population can lead to complaints (Newcombe et al. 2010) and a loss of confidence in the delivered water quality.

Raphidiopsis sp., Pseudanabaena sp., and Geitlerinema sp. collectively constitute >57% of the total cyanobacterial cells during both seasons. Notably, during the DS, these species contribute even more, accounting for 62.08% of the total cyanobacterial population. Notably, certain species were exclusively present during the DS, such as Cuspidothrix sp., Merismopedia tenuissima, and Coelomoron sp., which were found in October 2019, September 2021, and September 2020, respectively.

During the WS, not only fewer cells but also lesser species richness was observed. Raphidiopsis showed higher abundance during the WS, suggesting a greater tolerance to less favorable flow and ammonia nitrogen conditions. Other species that increased in abundance during this season include Merismopedia sp., Phormidium sp., Planktothrix sp., and Chroococcus sp. Particularly, Raphidiopsis racibiorskii is an invasive species in Brazil, known for its ability to fix atmospheric nitrogen and high adaptability to various environmental conditions (Padisák 1997). This species occurs in numerous riverine environments and poses a risk to public health by producing toxic compounds (Padisák 1997).

TP concentrations were assorted, within the range of 0.37–2.62 mg L⁻¹ over the study period, and with no statistically significant difference between the dry and wet seasons. According to the phosphorous concentration-based trophic state assessment for river systems used by the Environmental Company of the São Paulo State (CETESB 2013), the Jundiai River can be categorized as supereutrophic and hypereutrophic based on the aforementioned threshold values. In this context, phosphorus concentration in the Jundiai River does not appear to be a limiting factor for the growth of phytoplankton species. Nitrogen forms also exhibit similar behavior.

Studies have suggested optimal nutrient values for increased biomass and bloom development, including TP > 0.175 mg L⁻¹ and ammoniacal nitrogen > 0.5 mg L⁻¹ (Zhao et al. 2019), TP > 0.10 mg L⁻¹ (Kuo et al. 2018), and TP ≥ 0.14 mg L⁻¹ and total nitrogen < 1.34 mg L⁻¹, which are associated with potentially toxic cyanobacteria (Giblin & Gerrish 2020).

Phosphorus is widely considered to be the most limiting nutrient for cyanobacterial growth (Loewen et al. 2020). However, in several environments, including the one investigated in this study (ρ = 0.068; p-value > 0.05), TP and its constituents do not exhibit a statistically significant correlation with the cyanobacterial biomass, likely because of the already high nutrient loading and eutrophication (Li et al. 2018; Kim et al. 2019; Zhao et al. 2019).

The high nutrient content of the Jundiai River is attributed to its land-use and land-cover features. Water resources in anthropized areas are exposed to significant pollution loads from urban, industrial (Yang et al. 2019), and agricultural sources (Loewen et al. 2020). For example, the direct discharge of untreated wastewater and microbial degradation of organic compounds increase ammonia nitrogen concentrations in the water (Xing et al. 2007). Surface runoff from agricultural areas and discharge from domestic and industrial sewage are the primary sources of phosphorus (Xing et al. 2007; Kuo et al. 2018; Kim et al. 2019).

Previous investigations conducted in Brazilian rivers have revealed a positive correlation between cyanobacterial density and the occurrence of point sources of pollution, while the presence of riverbank protective vegetation has been associated with a reduction in cyanobacterial abundance (Cunha et al. 2017). Therefore, to mitigate the exacerbation of cyanobacteria proliferation in water resources, particularly in the Jundiai River, it is essential to minimize nutrient inputs.

The Jundiai River exhibited water temperatures in the range of 18.2–26.2 °C over the entire data collection period. The average water temperature in the DS (20.9 °C) was lower than that in the WS (24.7 °C), consistent with the DS occurring in the winter months and WS in the summer months. The number of cyanobacteria cells was more during the DS with temperatures in the range of 18.4–25.1 °C, despite the higher water temperatures observed during the WS (within the range of 23.4–26.2 °C). Water temperature is directly affected by air temperature (p-value < 0.05). This trend of high cyanobacteria cell counts at lower temperatures was statistically significant (p-value < 0.05) and exhibited a negative correlation with the daily maximum (ρ = −0.39), minimum (ρ = −0.46), and average air temperatures (ρ = −0.45) (Figure 6).

Moreover, the region's climatic conditions result in only a few days per year with minimum air temperatures <10 °C. As previously mentioned, the climate characteristics of the region where the Jundiai River is located consist of mild to hot winters, with minimum daily air temperatures recorded during DS sampling days ranging from 10 to 20 °C and maximum daily air temperatures ranging from 18 to 37 °C. Additionally, the region experiences predominantly sunny days. Consequently, the water temperature, which remained between 18 and 25 °C, still fits into the range supported by cyanobacteria species. Certain cyanobacterial species exhibit varying degrees of thermal tolerance. For instance, select species within the Aphanizomenonaceae family have been shown to prefer mild temperatures (Park et al. 2021) and thrive in temperatures as low as 14 °C (Xing et al. 2007). In addition, Pseudanabaena sp., a highly adaptable species, can withstand a broader range of temperatures within 15–35 °C (Gao et al. 2018).

Research conducted in a prominent South American river reveals that the impact of climate and hydrological factors on cyanobacterial abundance is minimal when compared to the nutrient influence caused by land-use change (Kruk et al. 2023). Water resources located in tropical regions are particularly prone to cyanobacteria abundance owing to the region's climatic characteristics, intense solar radiation, and high temperatures that accelerate the consumption and assimilation of nutrients by primary producers (Ludolf Gomes et al. 2012). This phenomenon results in the consequent dominance of cyanobacteria over other phytoplankton species (Li et al. 2018). Furthermore, these environments typically lack significant temperature oscillations throughout the year (Dao et al. 2016), thus enabling blooms to persist.

During the WS, which encompasses the summer months, the recorded minimum daily air temperatures ranged from 14 to 23 °C, with maximum daily air temperatures between 21 and 36 °C. The water temperature remained between 23 and 26 °C, conditions that closely resemble those observed during the DS. Therefore, similarly to TP, we can assume that the pattern of values recorded during the monitoring period remained within a range of non-critical or limiting values for cyanobacteria development.

A negative correlation was observed between cyanobacteria cell counts and flow (ρ = −0.22), which, despite its weak magnitude, was statistically significant at a confidence level of 10% (Figure 6). The increased flow rates associated with greater precipitation (p-value < 0.05 for monthly, 24-h, and 72-h rain) during the hot and humid months likely led to a more turbulent and unstable environment for cyanobacteria cells, in addition to nutrient depletion. Consequently, less favorable conditions for cell reproduction are enforced during the WS. The findings of this study align with the previous research in the field, which has shown that flow is the hydrological factor that most frequently compromises the development of cyanobacteria proliferation and is among the most significant factors triggering and affecting the cyanobacteria abundance (Nelson et al. 2018). Indeed, rivers exhibit a higher degree of sensitivity to precipitation than that of lentic ecosystems (Haakonsson et al. 2017).

During periods of low flow in this river ecosystem, the growth of cyanobacterial cells can be attributed to two main synergistic relationships. Firstly, the diminished flow results in the accumulation of crucial nutrients, such as nitrogen and phosphorus, due to the restricted dilution of pollutants discharged into the Jundiai River. This observation is supported by data, which demonstrate a negative association between water flow and TP (ρ = −0.40) and, although weak, a significant correlation at the 10% confidence level with ammonia nitrogen (ρ = −0.23). Thus, the reduction in the volume of water available for nutrient dilution allows cyanobacteria to have greater access to essential nutrients, promoting their growth. Additionally, during the WS, higher water flow associated with increased water turbidity (ρ = 0.50 and p-value < 0.05) reduces access to sunlight and indicates increased hydraulic turbulence. Under these conditions, cyanobacterial cells, which typically prefer still waters, are carried away by the river current. Therefore, the combination of these factors – the positive relationship between flow and turbidity and the negative relationship between flow and nutrients – creates favorable conditions for cyanobacterial growth during periods of low river flow.

High flow generally leads to intense agitation, which prevents blooms by increasing water turbidity and reducing light availability at deeper layers, ultimately disrupting populations (Cha et al. 2017). This disruption causes the cells to dissipate before the proliferation process is initiated (Cha et al. 2017). Alterations to the flow profile can impact the distribution of cyanobacterial genera and lead to changes in their composition (Giblin & Gerrish 2020). Some species of the genera Anabaena, Microcystis, Aphanizomenon, and Dolichospermum are dominant in low-discharge waters (Nelson et al. 2018; Giblin & Gerrish 2020; Park et al. 2021). By contrast, other phytoplankton groups are more resilient under high flow conditions, dominating the environment (Giblin & Gerrish 2020). Nonetheless, the implementation of efficient phosphorus control measures can mitigate the risk of cyanobacterial proliferation, even in situations characterized by lower flow rates and higher temperatures (Haakonsson et al. 2017).

Each aquatic ecosystem has unique characteristics, and the growth of cyanobacteria is influenced by a complex interrelationship of diverse environmental factors. The development of cyanobacteria in a river experiencing its lower flows may be specific to the context and should not be extrapolated as a universal trend applicable to all low flow environments. Thorough investigations and continuous monitoring are essential to comprehensively understand the dynamics of cyanobacterial growth.

Overall, air temperature is strongly and positively correlated with cyanobacteria cell growth, and global warming will result in more incidences of such episodes (Yan et al. 2020). However, as revealed by the current results, the temperature increase is not the primary issue for the Jundiai River, as water temperature remained above 18 °C. The growth of cyanobacteria in the studied location of the Jundiai River is associated with high concentrations of nutrients, primarily phosphorus, and reduced flow that is observed during drought periods.

Future climate projections emphasize, in addition to warming, the expansion of drought-affected areas in South America and other regions (Feng & Fu 2013). Notably, when comparing the 2010–2020 data with the 1980–2010 data, a 3.68% reduction in flow was observed in the Jundiai River (ANA 2021). The scenario was even more severe in other nearby regions, with an average flow reduction of >30% (ANA 2021). In recent years, the population increase in the region, coupled with the reduction of surface water flows, has contributed to water insecurity that is exacerbated by the degradation of surface source water. The impact of urbanization on water resource quality depends, among other factors, on the effectiveness of public sanitation services, which are still deficient in many regions of Brazil, particularly in terms of tertiary treatment.

In general, unregulated changes in human activities have the greatest impact on the onset of cyanobacterial proliferation, significantly affecting the protection of water quality, aquatic ecosystem balance, and human health (Loewen et al. 2020). The scarcity of riverbank protective vegetation due to unregulated urban development along riverbanks significantly affects the water resource conditions in terms of temperature, solar radiation, as well as the retention of nutrients and sediments from runoff (Cunha et al. 2017). Therefore, water resources in anthropized watersheds that are inherently vulnerable to cyanobacteria because of the physical and hydrological characteristics of the system (Burford et al. 2007) are likely to experience even more severe blooms in accordance with climate change projections (Paerl & Paul 2012).

The occurrence of algae and cyanobacteria in water treatment for public supply can have significant implications for various aspects of the treatment process, including coagulant dosage, performance of the clarification stage, filter clogging and performance, generation of toxins, release of taste and odor compounds in the water, and disinfection demand. To meet regulatory standards, upgrades to the water treatment systems are often necessary, such as the use of activated carbon, oxidant readjustment, or the implementation of additional processes like nanofiltration and reverse osmosis to establish multiple barriers for effective contaminant removal (Merel et al. 2013).

In the context of water safety, a strategy to minimize the release of intracellular compounds into the water treatment train leads to the removal of preserved cyanobacteria cells (Newcombe et al. 2010); without membrane damage during clarification. The effective removal of cyanobacterial cells during the DS poses a critical challenge in the treatment of water from the Jundiai River. This season not only witnesses the highest concentration of cells but also exhibits the lowest levels of turbidity. The limited presence of inert particles in the water avoids the generation of sufficiently heavy flocs for sedimentation during the clarifying step. As an alternative solution, flotation could be employed.

Source water quality fluctuations threaten the treatability of water for human consumption. The design and operation of a water treatment plant must be efficient to deal with different matrices over time, such as treating water having low turbidity and high concentration of organic matter and nutrients, favoring the proliferation of cyanobacteria, which is the object of this study. Thus, seasonal periodic variations in quality, as well as the variations resulting from increasing anthropogenic pressure and climate changes over the years of operation of a municipal water supply system (usually two or three decades), must be considered when selecting treatment technologies. Efficient and effective treatment plant operation will thus depend on refined monitoring. Protection and recovery of sources, multiple green and gray barriers, and the combination of all these management practices are essential to guarantee the safety of water for human consumption.

Lakes and reservoirs are characterized by hydrological conditions that are conducive to phytoplankton growth and are vital sources of water for human populations. Consequently, concerns over the safety of these water resources have intensified, and most studies on cyanobacterial proliferation have been conducted in lakes and reservoirs. Over time, the quality of riverine water resources that serve as vital sources of water for populations worldwide as well as in Brazil has declined, leading to a growing cyanobacterial proliferation in these environments.

In the context of this exploratory study, it is important to highlight the limitations associated with the statistical correlations. These limitations include a small sample size, diminishing the statistical power of the analysis and restricting the generalizability of the results to a broader population. Furthermore, the absence of multivariate analysis may result in overlooking potential interactions and complex relationships among multiple independent variables. In addition, the limited precision stemming from a smaller number of observations contributes to wider confidence intervals, making it more challenging to draw definitive conclusions regarding the strength and direction of the associations.

Over the course of 3 years of the Jundiai River (Salto, São Paulo State, Brazil) water monitoring, a consistent pattern of significantly elevated cyanobacterial cell counts during the DS was identified, which also coincided with the presence of the highest number of species. Furthermore, most of the species identified in the river can produce toxins and generate unpleasant taste and odor in the water, posing a threat to water safety. Raphidiopsis sp., Pseudanabaena sp., and Geitlerinema sp. were the dominant species throughout the dry and wet seasons.

The DS is characterized by higher average nutrient concentrations and lower water flow. The finding of the current study provides evidence that the reduction in river flow creates the primary precursor condition to the cyanobacteria cell growth. Because the Jundiai River is highly eutrophic, phosphorus is not a limiting factor.

Effective management of the Jundiai River basin requires control of points and diffuse pollution sources to improve water quality. The main challenges include the control of pollution of Brazilian rivers through the implementation of tertiary treatment of effluents and increased riparian protection of streams. Therefore, monitoring the presence of cyanobacteria in the river sources of drinking water, such as the Jundiai River, is essential for ensuring water safety and implementing effective management measures.

Notably, the current study adopted an exploratory approach. Therefore, the obtained findings should be interpreted with caution, as they do not aim to establish definitive conclusions or generalize the insights. Further research, encompassing larger sample sizes and more comprehensive analyses, is imperative to validate and reinforce the results, leading to broader inferences.

This project was supported by the Graduate Program in Civil Engineering of the University of Campinas. The authors wish to thank the local water company SAAE Salto for its cooperation and assistance.

This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001, and the Autonomous Service of Water and Sewage of Salto (translated from Portuguese: Serviço Autônomo de Água e Esgoto de Salto – SAAE Salto).

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

The authors declare there is no conflict.