Algal blooms and trophic state in a tropical estuary blocked by a dam (northeastern Brazil) Ocean and Coastal Research

phytoplankton, algal blooms, events associated with increased eutrophication The Bacanga River Estuary is socioeconomically important due to artisanal fishing and aquaculture. It is blocked by a dam and is under human pressure along its drainage basin, intensifying the eutrophication process. This study reports on the occurrence of phytoplankton blooms and trophic state (TSI and TRIX) at six sampling sites during the annual cycle. The estuary was divided into downstream and upstream regions. Higher salinity, turbidity, depth, and lower dissolved oxygen levels were found downstream; whereas, high levels of chlorophyll a and nutrient concentrations were observed in both regions. There were blooms of Leptocylindrus danicus (1.45 × 10 6 cells L -1 ) and Skeletonema costatum (1.89 × 10 6 cells L -1 ) downstream; whereas phytoflagellate proliferation, such as those of Chlamydomonas sp. (13.17 × 10 6 cells L -1 ), Euglena gracilis (7.84 × 10 6 cells L -1 ), and Euglena proxima (1.03 × 10 6 cells L -1 ) were recorded upstream, with Chlamydomonas sp. as the discriminant species of this zone. Both trophic indices (TSI; TRIX) indicated elevated trophic conditions for the estuary, classifying it as hypereutrophic. Nevertheless, TSI only showed a significant relationship with some specific phytoplankton blooms. Thus, TSI seems to be the trophic index with a better response in the assessment of estuarine ecological functioning.


INTRODUCTION
Eutrophication in estuaries has been a major ecological problem that is caused by the inputs of natural and anthropogenic nutrients through the discharge of sediments, the atmosphere, the ocean, or even through the exchange of nutrients with the coastal region (Ferreira et al., 2011;Statham, 2012;Cloern et al., 2014;Hayn et al., 2014;Monteiro et al., 2016). The estuarine ecosystems are under heavy anthropogenic pressures that compromise their water quality and biodiversity, causing a loss of habitat and a reduction in the quality of life of riverside communities. Because of these impacts, it is necessary to characterize and assess the current conditions of these environments based on an ecosystemic and trophic approach (Cloern, 2001;Herrera-Silveira et al., 2009;Smith and Schindler, 2009;Ferreira et al., 2011;Cotovicz-Júnior et al., 2012a).
Therefore, it is important to measure the excessive growth of phytoplankton, also known as algal blooms, because these events are associated with increased eutrophication caused by anthropogenic activities (Napiorkowska-Krzebietke et al., 2017).

Ocean and Coastal Research
http://doi. org/10.1590/2675-2824069.20-006akddss ORIgINal aRTICle ISSN 2675ISSN -2824 The Bacanga River Estuary is socioeconomically important due to artisanal fishing and aquaculture. It is blocked by a dam and is under human pressure along its drainage basin, intensifying the eutrophication process. This study reports on the occurrence of phytoplankton blooms and trophic state (TSI and TRIX) at six sampling sites during the annual cycle. The estuary was divided into downstream and upstream regions. Higher salinity, turbidity, depth, and lower dissolved oxygen levels were found downstream; whereas, high levels of chlorophyll a and nutrient concentrations were observed in both regions. There were blooms of Leptocylindrus danicus (1.45 × 10 6 cells L -1 ) and Skeletonema costatum (1.89 × 10 6 cells L -1 ) downstream; whereas phytoflagellate proliferation, such as those of Chlamydomonas sp. (13.17 × 10 6 cells L -1 ), Euglena gracilis (7.84 × 10 6 cells L -1 ), and Euglena proxima (1.03 × 10 6 cells L -1 ) were recorded upstream, with Chlamydomonas sp. as the discriminant species of this zone. Both trophic indices (TSI; TRIX) indicated elevated trophic conditions for the estuary, classifying it as hypereutrophic. Nevertheless, TSI only showed a significant relationship with some specific phytoplankton blooms. Thus, TSI seems to be the trophic index with a better response in the assessment of estuarine ecological functioning.
2 Sá et al. The use of trophic indices to assess eutrophication status has been recommended by the Expert Group on the Scientific Aspects of Marine Environmental Protection (GESAMP) (Moncheva et al., 2001), as they can reflect the anthropogenic influence on water quality and the ecological functioning of rivers, lakes, and reservoirs (Cunha et al., 2013).
Trophic status indices provide information on how the availability of nutrients and light controls the development of phytoplankton, and how they can serve as useful tools for monitoring and protecting coastal ecosystems (Lemley et al., 2015). For example, the trophic state index (TSI) is used to evaluate the degree of eutrophication in tropical environments (Toledo et al., 1983), and the following variables are used to calculate the index: concentrations of chlorophyll a, total phosphorus, and Secchi depth. This index, according to the external inputs of nutrients, such as domestic sewage, industrial, and agricultural waste, contribute to the classification of water bodies according to the degree of eutrophication (Maia et al., 2015).
The trophic index (TRIX) is used to determine the trophic status of coastal waters based on biological (chlorophyll-a) and physical-chemical (oxygen, nutrients) metrics (Nasrollahzadeh et al., 2008). TRIX was originally developed for Italian coastal waters and then applied in many European seas. It is currently used in Brazilian coastal waters (Cotovicz-Júnior. Et al., 2013;Nascimento-Filho et al., 2013;Monteiro et al., 2016;Araújo et al., 2017). This trophic evaluation with phytoplankton indicator species and algal blooms contributes to the comprehensive assessment of phytoplankton, one of the biological quality elements in coastal waters (Pettine et al., 2007;Garmendia et al., 2013;Brugnoli et al., 2019).
These indices assume that the eutrophication processes depend mainly on algal biomass (Giordani et al., 2009), which considered a key element in determining water quality. Structural changes in their community and are often expressed by ecological indices (Spatharis and Tsirtsis 2010). Quantifying eutrophication in estuarine environments in Brazil has been a challenge. There are no conceptual models and standards to determine the trophic status according to the national environmental legislation (Cotovicz-Júnior et al., 2012b;Nascimento-Filho et al., 2013).
Based on this assumption, this study was carried out in a shallow, urbanized, and dammed tropical estuary. A large amount of sewage is discharged into the estuary daily, leading to contamination by toxic substances, thus, putting aquatic life at risk (Duartedos-Santos et al., 2017). This study aimed to relate the occurrence of phytoplankton blooms and trophic state using the TSI and TRIX indices to assess the ecological functioning of the Bacanga River Estuary (BRE) and the effects of nutrient enrichment on the phytoplankton community.
The estuary of the Bacanga River is inserted in the basin of the same name. It is considered the second largest in length and demographic density on São Luís Island, with a deficient sewage network, totaling 850 discharge points of untreated domestic sewage. It is worth mentioning that potentially polluting industrial plants are located in this basin, such as a port complex, beyond the surrounding population using the water resources available in the area. The effects on the eutrophication process in the phytoplankton communities were explored to identify the relationships of such indexes with nutrients concentrations in the tropical estuary blocked by a dam.

MeTHODs sTUDy aRea aND saMplINg DesIgN
The Bacanga River basin is in the northwestern part of the city of São Luís, in the Brazilian state of Maranhão, between 2°32'26" and 2°38'07" S and 44°16'00" and 44°19'16″ W. The basin occupies an area of 11,030 ha, which accounts for 12.33% of the island of São Luís, and it is divided into five sub-watersheds. There are approximately 58,844 households and a population estimate of 223,343 inhabitants distributed across 41 neighborhoods according to the São Luís Institute of City, Research and Urban and Rural Planning (INCID). It is worth mentioning that part of the Bacanga River basin is an environmental preservation area, according to State Decree nº 7545, on March 2, 1980.
The Bacanga River estuary (BRE) has a length of 22 km from its source to São Marcos Bay (Figure 1). This estuary is an inlet that penetrates the continental region near the city of São Luís and has two tributaries, Bicas and Gapara. Despite its perennial hydrological regime, Bacanga river flow is discrete (low flow) and is Sá et al. subject to variations according to the season, which favor the advance of marine waters that periodically flood the wide fluvial-marine plain, reaching basically the low and medium course of the river. As a result, the water body has estuarine characteristics and behaves according to the semidiurnal tidal regime, with tidal heights ranging between 4 and 6 m.
The connection with the sea is blocked by a dam that controls the entrance and the level of the seawater up to the 4-m level, in addition to reducing the inflow of the estuary waters. The hydraulic movement follows a circular, unidirectional, and irreversible pattern, which begins in the floodgates and moves upstream through the left side, returning through the right side of the dam. This behavior is repeated in both tidal periods (Pitombeira and Morais, 1977).
According to the rainfall historical average of the ten years (2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013), the seasonal cycle is well defined. The rainy season is from January to July, and the dry season is from August to December. During the sampling campaigns, the highest monthly rainfall rate was registered in February (296 mm) and the lowest in September (2.8 mm), according to the Brazilian Institute of Meteorology (INMET).
The samples were collected from six locations ( Figure  1) along the estuary during the annual cycle between 2012 and 2013. The sampling sites B1, Bicas stream (B2), Jambeiro stream (B3), and Coelho stream (B4) are located downstream, influenced mostly by seawater. On the other hand, sites B5 and B6 are located upstream, influenced mostly by freshwater. Six bimonthly samplings were carried out, three of them during the rainy season (April 2012, June 2012, and February 2013) and three during the dry season (August 2012, September 2012, and December 2012). The sampling was carried out from the downstream to the upstream region during the flood tide, according to the lunar cycle and the macrotidal dynamics in the BRE.

pHysICal aND CHeMICal paRaMeTeRs
The assessment of water quality was based on physical and chemical parameters, such as water temperature, pH, salinity and total dissolved solids (TDS), measured in the surface layer (50 cm deep) with a multiparametric probe (HI-9828). The water depth was measured with a digital probe (Speedtech), the transparency of the water with a Secchi disk, and the turbidity with a turbidimeter (Lamotte-2020). 4 Sá et al. Dissolved oxygen and biochemical oxygen demand (BOD 5 ) were determined according to the chemical method of Winkler, modified by Golterman et al. (1978) with 5 days of incubation at 20°C for BOD 5 .

NUTRIeNT MeasUReMeNTs
For the analysis of nutrients, 2 liters of water was collected in the surface layer (50 cm deep) with a Van Dorn bottle. The quantification of total phosphorus (TP) and ammonium ion (NH 4 + ) was made following the methodology described by Koroleff (1983). Nitrate (NO 3 -) and nitrite (NO 2 -) were determined according to Strickland and Parsons (1972); silicate (SIO 2 -) and orthophosphate (PO 4 -) according to Grasshoff et al. (1983). Dissolved inorganic nitrogen (DIN) was obtained from the sum of NH 4 + -N, NO 2 --N, and NO 3 --N; Dissolved inorganic silicate (DIS) from the SIO 2 -, and for dissolved inorganic phosphorus (DIP) only the PO 4 -3 was used. To evaluate the stoichiometric limitations, Redfield ratios were used, according to Pavlidou et al. (2004) and Xu et al. (2008).

pHyTOplaNkTON bIOMass aND abUNDaNCe
The phytoplankton biomass was estimated by chlorophyll a concentration. Water samples (2L) for chlorophyll a (µg L −1 ) determination were filtered through Whatman GF/F glass fiber filters, and pigment extraction was performed with 90% acetone. The volume of the filtrate ranged from 0.10 to 0.25 L. Pigment concentration was measured by spectrophotometry (Parsons et al., 1984;Greenberg et al., 1992) and calculations were done according to Strickland and Parsons (1972).
To analyze the structure of the phytoplankton community and the phytoplankton abundance, subsurface water samples (250 mL) were collected with a Van Dorn bottle and fixed with Lugol solution. The identification and counting of phytoplankton cells were carried out with an inverted microscope at 400 × magnification, following Utermöhl's technique (Utermöhl, 1958). At least 100 fields were counted using sedimentation chambers of 25 ml . The phytoplankton community at each sampling point was analyzed taxonomically, at the lowest taxonomic level, according to the specialized literature and updated according to AlgaeBase database. The phytoplankton were considered a bloom when a particular taxon reached or exceeded a density of 1 × 10 6 cells L −1 (Livingston, 2007).

TROpHIC sTaTe INDexes
The trophic index (TRIX) was adapted by Cotovicz-Júnior et al. (2012b), where a correction was made for a temporal data series. Vollenweider et al. (1998) recommended the mean ± 2.5 standard deviation as an appropriate statistical tool to define the upper and lower limits of the parameters used in the TRIX index. Thus, extreme or inconsistent values were excluded. The values were converted to log base 10. The TRIX for BRE was calculated according to the following formula (1): Where Chla is the chlorophyll a concentration (µg L -1 ); aD%O is the absolute deviation of dissolved oxygen saturation; DIN is the dissolved inorganic nitrogen, DIP is the dissolved inorganic phosphate, k is equal to 1.5, and m is equal 1.2. Trophic scales and descriptors for water quality were defined according to Giovanardi and Vollenweider (2004), Penna et al. (2004), Nasrollahzadeh et al. (2008) and Cotovicz-Júnior et al. (2012b). Numerically, the index is scaled from 0 to higher than 8, covering a wide range of trophic conditions from ultra-oligotrophic to hypereutrophic (Table 1).
The trophic state index (TSI) was proposed by Carlson (1977) and was modified for tropical environments in this paper. The formulas used for calculating the TSI for lotic environments are given below (2 and 3): In Chla x x 2 10 6 0 7 0 6 20 2 = ---where TP is the total phosphorous concentration measured at the water surface (µg L -1 ); Chla is the Chlorophyll a concentration measured at the water surface (µg L -1 ); ln is the natural logarithm. Thus, the results are equal to the arithmetic average of the total phosphorous and chlorophyll a indices that were used to classify the estuary into different trophic

2-4
Water poorly productive and low trophic status High (Oligotrophic)

5-6
Water moderate to very productive and high trophic status Moderate (Mesotrophic to Eutrophic)

6-8
Water very productive and high trophic status Poor (Eutrophic)

sTaTIsTICal aNalysIs
For the numerical analysis and spatial graphical representation, the mean of the spatial data from both downstream and upstream in the estuary in an annual cycle was used. Before these analyses, one-way PERMANOVA was performed to determine the significant differences (p<0.05) in the physicalchemical factors between the downstream and upstream regions. To test the differences (p <0.05) of the environmental variables in relation to the locations, parametric (ANOVA-F unilateral) or nonparametric (Kruskal-Wallis-H) tests were performed according to normality (Kolmogorov-Smirnov) and homogeneity (Levene's test) of variances. Based on the correlations established between the trophic indices (TSI and TRIX) and the environmental variables (R > 0.7), a Cluster analysis was performed using the cluster group average method to determine the dissimilarity, based on Euclidean distance (square root transformed).
Non-metric multidimensional scaling (nMDS) was applied to determine the similarity of the trophic states concerning phytoplankton blooms. A SIMPER analysis, based on phytoplankton abundance, was also applied to identify the "characterizing species" and the "discriminating species" with regards to similarity and dissimilarity, respectively (Clarke and Gorley 2006), and then an ANOSIM was applied to test the significance of the similarities. The multiple linear regression model (GLM), with a progressive forward selection of the variables (using p<0.05), was applied to assess the relationships between the trophic indices and the algal blooms. Trophic indices (TRIX and TSI) were used as predictors. The statistical analyses were performed using SPSS (version 24.0), STATISTIC (version 10.0), and Primer (version 6.0).

ResUlTs pHysICal aND CHeMICal paRaMeTeRs
The spatial distribution of the physical and chemical parameters of the water monitored in the estuary are shown in Figure 2 and Table 2. Significant differences were registered between the downstream and upstream regions of the study area based on the oneway permutational multivariate analysis of variance (PERMANOVA) analysis (F = 3.64, p = 0.02).
This difference between the upstream and down     Table 2). When considering the Redfield ratio in BRE, the DIN:DIP ratio was higher downstream (F = 4.04 and p = 0.05), specifically in B2. Likewise, the DIN:DIS ratio was higher in the lower reaches of the estuary (F = 5.36, p = 0.02), mainly in B4. Regarding the DIS:DIP ratio, there were no significant differences (F = 0.28 and p = 0.58) among the sampling sites, with lower values at B1 (near the dam) and B6 (upper region) ( Table 2).

pHyTOplaNkTON COMMUNITy
A total of 66 species of phytoplankton belonging to five major taxonomical groups were identified: Bacillariophyta (41 species), Euglenophyta (nine species), Cyanobacteria (eight species), Dinophyta (four species), and Chlorophyta (four species). Diatoms were the dominant group in both downstream and upstream zones, followed by cyanobacteria with seven species identified downstream, and Euglenophyta with nine species identified upstream.
The similarity percentages (SIMPER) analysis of the density of the phytoplankton species revealed a similarity contribution of 33.86%, with high similarity between the species observed downstream (30.57%), while the upstream percentage was only Sá et al.  (Table 4). In addition to the SIMPER analysis, the analysis of similarity (ANOSIM) global R test determined the level of significance of the differences in the structure of the phytoplankton community between the BRE zones. The results revealed a dissimilarity between the downstream and upstream regions with a global R of 0.71 and a significance level of 6.7%. The chlorophyll a concentration was not significantly different along the BRE (F = 0.006 and p = 0.33), with values ranging from 34.26 ± 32.70 µg L -1 downstream to 32.20 ± 46.33 µg L -1 upstream ( Table 2).

TROpHIC sTaTe INDICes
The TRIX ranged from 8.39 ± 0.23 (upstream) to 8.52 ± 0.32 (downstream), suggesting hypereutrophic conditions, characterizing the BRE as a very poor environment in terms of water quality. There was no significant difference between the downstream and upstream regions (F = 0.06 and p = 0.80). The TSI ranged from 67.23 ± 0.23 (upstream) to 68.28 ± 1.10 (downstream), suggesting high trophic conditions (hypereutrophic), with no significant difference between the two regions (F = 0.29 and p = 0.59) (Figure 3).

RelaTION Of THe TROpHIC sTaTe aND algal blOOMs
The association between the sampling sites through the hierarchical cluster analysis with the application of Euclidean distance showed the formation of two groups with a cut-off value of 0.73. The combinations indicated a similarity (0.42) among the sites B1, B2, B3 and B4 (downstream) forming group A, and between B5 and B6 (0.73) forming group B (upstream). The estuary is divided into two regions, with group A representing the downstream region and group B (Figure 4) in the upstream region, considering the variables of dissolved oxygen, salinity, turbidity, depth, and trophic state (TRIX and TSI).  The contribution of phytoplankton species (characterizing and discriminating) and the trophic status are represented in the non-metric multidimensional scaling (nMDS) with a stress level of 0.01 ( Figure 5).
The analyses showed a well-defined trophic state (TRIX and TSI) in both zones, with evident blooms of S. costatum and L. danicus downstream, specifically at B1 and B2 (Bacanga Dam). In the upstream Sá et al.  region, a high density of E. proxima and dominance of Trachelomonas sp. were observed at B5. The proliferation of Chlamydomonas sp. and E. gracilis is specific to site B6.

RespONses Of algal blOOMs TO THe TROpHIC sTaTe
The generalized linear models (GLM) showed that the bloom-forming species of phytoplankton in the BRE did not respond to the trophic state indicated by the TRIX index, with low Adjusted Squared Multiple (R = 0.21) and non-significant correlation (F = 2.29, p = 0.05). However, S. costatum, Protoperidinium sp., E. gracilis, Chlamydomonas sp., and L. danicus showed a direct and significant correlation (F = 9.17, p = 0.000007) with the total trophic status index (TSI Total) and an adjusted squared multiple of Sá et al.

INTeRaCTION beTweeN NUTRIeNT lOaD aND TROpHIC sTaTe
Tropical estuaries are subject to many impacts that affect biodiversity and ecosystem services (Herrera-Silveira et al., 2009;Cloern et al., 2014;Monteiro et al., 2016;Lepšová-Skácelová et al., 2018). These impacts are mainly owing to discharges of domestic sewage, industrial wastewater, and runoff from agricultural fertilizers that pollute many estuaries around the world (Pei et al., 2019). The damming of these environments is another factor that strongly modifies water biogeochemistry and flow patterns, which compromise their ecological and biological dynamics (Lima et al., 2020).
The impacts mentioned can be seen along the drainage basin of the Bacanga River; according to the trophic indices, the BRE has been classified as hypereutrophic, which is a relevant factor for the formation of algal blooms. Both TRIX and TSI had the same behavior in the BRE, with no significant differences between zones (downstream-upstream).
Thus, the TRIX index can be considered the one that best defines the eutrophication conditions of the BRE, owing to the high concentrations of NH 4 + that follow the low oxygen saturation levels throughout the estuary. This leads to a highly productive environment in terms of water quality (Cotovicz-Júnior et al., 2012b) in response to the interdependent relationship between nutrient availability and biomass and consequent higher transference energy to various trophic levels (Cloern et al., 2014).
Based on the Redfield molar ratio (N:P and N:Si), this pattern can be expressed by the co-limitation Sá et al. of phosphorus and silicate, to the excess of anthropogenic nitrogen (NH 4 + ) observed downstream (Pavlidou et al., 2004;Xu et al., 2008). This excess is a response to an increase in the DIN concentrations, which was probably the result of industrial wastewater and domestic sewage discharge in the study area, in addition to the indiscriminate use of fertilizers upstream that contributed to an increase in phosphorus in that region (Silva et al., 2014;Duarte-dos-Santos et al., 2017). Therefore, the limitation of phosphorus may have a low impact on the reduction of the trophic level of the water body, being compensated by the conservative behavior of nitrogen concentrations as a function of salinity (Anderson et al., 2002;Cloern et al., 2014;Paula-Filho et al., 2020).
The TSI was the index that best responded to phytoplankton blooming events. The TSI showed the spatial variation of total phosphorus and the high concentrations of chlorophyll a, promoting interaction between the conditioning factor and the response variable. In lotic environments usually rich in phosphorus, the shorter residence time and low light availability may restrict the development of phytoplanktonic biomass (Cunha et al., 2013;Lemley et al., 2015;Madhu et al., 2017). Therefore, the TSI results are satisfactory because the estuary has characteristics much more similar to that of a reservoir. The BRE was homogeneous with respect to the trophic indices owing to the shorter time of water renewal. Residence time in the BRE is longer because of the dam, providing conditions and facilitating decisions about water management (Cunha et al., 2013, Pitombeira andMorais, 1977). The presence of the dam causes the environment to have a common lentic, water-body behavior, in which the salinity depends on the control and administration of the gates that regulate the freshwater outflow and entry of seawater (Duarte-dos-Santos et al., 2017).
According to Kang et al. (2019), the construction of a dam in estuaries affects water quality owing to slower water flow rates, providing an increase in water residence time and primary productivity because of retention of anthropogenic constituents, including nutrients and organic compounds in estuarine reservoirs.

RespONses Of algal blOOMs TO THe TROpHIC sTaTe
Diatoms dominated downstream, being responsible for S. costatum blooms in this part of the estuary, followed by Melosira varians and L. danicus. In the upstream region, the proliferation of phytoflagellates, such as Chlamydomonas sp., E. proxima, and E. gracilis, was observed, under conditions of low salinity and phosphorus availability. According to Lepšová-Skácelová et al. (2018) and Cloern et al. (2014), the phytoplankton community responds to salinity tolerance and nutritional preferences, which could be used as key indicators of ecological status in eutrophic estuarine environments (Napiorkowska-Krzebietke et al., 2017).
Considering the occurrences of these blooms, it can be observed that the salinity and total phosphorus concentrations were the main factors that determined the pulses of S. costatum and L. danicus downstream, and Chlamydomonas sp., E. gracilis, and E. proxima upstream, organisms which could potentially be considered as key species and indicators of the trophic state in BRE.
According to the GLM, the TRIX does not respond well to the occurrences of these algal blooms in the BRE; values of this index are weakly correlated and insignificant. However, the total TSI showed a moderate correlation, including with the main bloomforming species (S. costatum, Protoperidinium sp., E. gracilis, Chlamydomonas sp., and L. danicus). The adaptive strategies presented by phytoplankton, in response to the limitations of nutrients and light that control their growth, divides the community into two groups, namely, the bloomers and the survivors. Algal bloomers are microphytoplankton such as diatoms and larger dinoflagellates, which grow in eutrophic systems. Survivalists, however, are predominantly smaller algae that grow in low-nutrient conditions (Franz et al., 2012;Pei et al., 2019).
The mixture of freshwater and seawater created two environmental gradients (downstream and upstream), with specific physical-chemical characteristics, considering salinity and depth as the most important characteristics. Under these conditions, microalgae, such as chain-forming diatoms, dominated the lower region of the estuary. This group of Sá et al. species characterizes the downstream region, which is subject to high salinity, greater depths, and low transparency as well as a high organic matter load (BOD 5 ), alkaline pH, and high concentrations of nutrients of anthropogenic origin (NH 4 + ).

CONClUsION
The Bacanga River estuary is a coastal system with variations in salinity, depth, and algal blooms in the two estuarine regions (downstream and upstream). The high load of nutrients of anthropogenic origin and restricted circulation of BRE waters may have contributed to the increase in chlorophyll a levels and bloom events.
The density peaks downstream were associated with blooms of L. danicus and S. costatum, which were limited by silicate. Meanwhile, in the upstream zone, phytoplankton proliferation was caused by a high density of phytoflagellates of the genus Chlamydomonas, as well as E. gracilis, and E. proxima, limited by phosphorus.
The TRIX was the index that best reflected the anthropogenic influence on the water quality of BRE, providing information on reduced nitrogen surplus. However, the TSI showed the trophic state as a function of chlorophyll-a concentrations, an evident response to phytoplankton growth and the formation of specific algal blooms. This index proved to be an excellent tool that better assesses the ecological functioning of a dammed estuary, with the confinement of nutrients caused by anthropogenic actions.

aCkNOwleDgMeNTs
We thank FINEP for the funding grant (01.10.0714-00 CTHIDRO) and CAPES (Coordination for the Improvement of Higher Education Personnel) for the Master's scholarship granted to the first author.