Original Article Acta Limnologica Brasiliensia, 2016, vol. 28, e6 http://dx.doi.org/10.1590/S2179-975X3115 ISSN 0102-6712 This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Can a one-sampling campaign produce robust results for water quality monitoring? A case of study in Itupararanga reservoir, SP, Brazil Estudos baseados em uma única campanha de amostragem podem produzir resultados robustos para o monitoramento da qualidade da água? Um estudo de caso no reservatório de Itupararanga, SP, Brasil Frederico Guilherme de Souza Beghelli1*, Marcelo Luiz Martins Pompêo2, Mônica Páscoli3, Vivian Silva Lira1, Renata de Lima3 and Viviane Moschini-Carlos1 1 Universidade Estadual Paulista – UNESP, Avenida Três de Março, 511, Alto da Boa Vista, CEP 18087-180, Sorocaba, SP, Brazil 2 Departamento de Ecologia Instituto de Biologia – IB, Universidade de São Paulo – USP, Rua do Matão, Travessa 14, 321, Butantã, CEP 05508-090, São Paulo, SP, Brazil 3 Universidade de Sorocaba – UNISO, Rodovia Raposo Tavares, Km 92,5, Artura, CEP 18023-000, Sorocaba, SP, Brazil *e-mail: fred_sb@hotmail.com Cite as: Beghelli, F.G.S. et al. Can a one-sampling campaign produce robust results for water quality monitoring? A case of study in Itupararanga reservoir, SP, Brazil. Acta Limnologica Brasiliensia, 2016, vol. 28, e6. Abstract: Aim: The aim of this study was to conduct a rapid and comprehensive analysis of the reservoir using physical, chemical and biological variables comprising water and sediment as also to verify if a one-sampling campaign can produce information that is robust enough for monitoring purposes. Methods: One-sampling campaign was performed along the central body of Itupararanga reservoir comprising five sampling points. The following physico-chemical and biological variables were measured in each of them from water samples: temperature, pH, concentration of dissolved oxygen, electric conductivity, suspended matter, total nitrogen and phosphorus and fecal coliforms. A genotoxicity assay with Allium cepa was also performed with water from each one of the five sampling points. As a complement, the phytoplankton and benthic macroinvertebrates communities were also analyzed in the two most distant sampling points (P1 and P5). Results: Results indicate a scenario with eutrophication and toxicity. High concentrations of fecal coliforms were recorded and there is a worrying prevalence of cyanobacteria - potential toxin producers - in phytoplankton community. Benthic macroinvertebrates analysis suggests poor quality of the water and sediments. Genotoxicity tests suggest the presence of toxic substances along the waterbody. Conclusions: It is concluded that Itupararanga reservoir has water with poor quality that could be related to the land uses around the reservoir. The comprehensive approach performed here - where distinct sources of monitoring variables was considered - despite the rapid protocol (one campaign), was enough to respond to the distinct anthropic impacts received by the reservoir. Considering it, this approach is recommended as a rapid diagnostic tool. Keywords: water quality monitoring; bioindicator; biomonitoring; neurotoxins; hepatotoxins. Resumo: Objetivo: O objetivo deste estudo foi realizar uma análise rápida e abrangente do reservatório usando variáveis físicas, químicas e biológicas, compreendendo água e sedimento e verificar se esta abordagem baseada em uma única campanha de coleta pode fornecer informações suficientemente robustas para o monitoramento. Métodos:  Uma campanha de amostragem foi realizada ao longo do corpo central do reservatório de Itupararanga compreendendo cinco pontos de 2 Beghelli, F.G.S. et al. Acta Limnologica Brasiliensia, 2016, vol. 28, e6 types of stresses in aquatic environment need the appropriate monitoring tools. When considering water quality, the main impacts are the eutrophication, usually related to sewage discharges, agriculture and pasture as its consequences as oxygen depletion, bad smell, fish kill and the increase in the probabilities of cyanotoxins release (Vasconcelos, 2015; Beghelli et al., 2016); toxicity caused by metal pollution (Di Veroli et al., 2014; Cardoso-Silva  et  al., 2016) and organic pollutants (Zelníčková et al., 2015), usually related to industrial activities, urbanization and agriculture. When considering aquatic ecosystems, vast areas and distinct environments with their own peculiarities must to be monitored but the human and economic resources are limited and the time is a key factor when the objective is the maintenance of the ecosystems. Considering this scenario, monitoring tools and protocols are desirable and they must be the most accurate as possible but at the same time it is also desirable tools with low cost and time consuming. Among the predictors there are some ones with well-established protocols as the coliforms and eutrophication analyses. The first one is mainly related to sewage discharges or to the entrance of organic matter from livestock (Muniz et al., 2011). The latter is related to the entrance of nutrients in the ecosystems, mainly phosphorus and nitrogen, usually coming from sewage or agriculture activities (Trindade & Mendonça, 2014; Ni  et  al., 2015). Other variables as electric conductance, dissolved oxygen, pH, transparency and suspended matter can be related with the problems previously related, but they are also related with the generic conditions of the environment (Muniz et al., 2011; Esteves et al., 2015). 1. Introduction The water bodies provide many services for humanity but the misusage of them or of the surroundings can seriously compromise water quality and availability (Cunha  et  al., 2013; Suriano & Fonseca-Gessner, 2013; Silva & Costa, 2015). Some of the most widespread impacts are: riparian forestry removal (Torres & Ramirez, 2014); the contamination promoted by agrochemicals (Giraldo  et  al., 2014; Díaz-Barriga  et  al., 2015), industrial and domestic wastes (Maceda-Veiga et al., 2013) as also the exotic species introduction (Beghelli  et  al., 2014a; Fernandez  et  al., 2014; Rietzler et al., 2014). There is a growing number of pollutants detected in different kinds of environments. This situation highlights the requirements for precaution and treatment measures as also for warning for the possible consequences that these substances may cause in the environment and human health (Jongh et al., 2012; Wongsasuluk et al., 2014; Testai et al., 2016). This scenario may be worst when the water body is inserted in an urban area where it may receive many different impacts from multiple sources and where there is high demands for water. In Brazil, some reservoirs as Billings (Moschini-Carlos et al., 2010; Pompêo  et  al., 2013; Duarte & Borges, 2016) and Guarapiranga (Fontana  et  al., 2014) are emblematic situations of urban reservoirs that are seriously damaged by water contamination and surrounding environment misusing. Once there are many types of anthropic impact and considering the complexity of ecosystems, the environmental assessment must take into account distinct predictors and its interactions. Distinct amostragem. As seguintes variáveis biológicas e físico-químicas foram determinadas em cada ponto a partir de amostras de água: temperatura, pH, concentração de oxigênio dissolvido, condutividade elétrica, material em suspensão, nitrogênio e fósforo totais e coliformes fecais. Um  ensaio de genotoxicidade com Allium cepa também foi realizado com água proveniente de cada um dos cinco pontos de amostragem. Como complemento, as comunidades fitoplanctônica e de macroinvertebrados bentônicos também foram analisadas, mas apenas nos dois pontos de amostragem mais extremos (P1 e P5). Resultados:  Os resultados indicam um cenário de eutrofização e toxicidade. Altas concentrações de coliformes fecais foram registradas e há uma prevalência preocupante de cianobactérias – potenciais produtoras de toxinas - no fitoplâncton. A análise dos macroinvertebrados bentônicos indica baixa qualidade da água e dos sedimentos. Os testes de genotoxicidade sugerem a presença de substâncias tóxicas ao longo do corpo d’água. Conclusões: Conclui-se que as águas do reservatório de Itupararanga apresentam baixa qualidade. Tal situação pôde ser relacionada aos usos do solo no entorno do reservatório. A abordagem abrangente ora realizada onde fontes distintas de variáveis para monitoramento foram consideradas, apesar do protocolo expedito (uma campanha), foi suficiente para responder aos diferentes tipos de impactos antrópicos recebidos pelo reservatório sendo uma opção viável para um diagnóstico rápido. Palavras-chave: monitoramento da qualidade da água; bioindicadores; biomonitoramento; neurotoxinas; hepatotoxinas. 3 Can a one-sampling campaign produce robust results… Acta Limnologica Brasiliensia, 2016, vol. 28, e6 The biological communities are important tools in environmental assessment once they can integrate the environmental condition in opposition to the physico-chemical analyses, which are more specific. Furthermore, distinct communities have their specific roles in the ecosystems and may present differential sensitivities to pollution (Bonada et al., 2006). Genotoxicological and citotoxicological assays can bring information about toxicity in the environment. These assays can also give information to quantify the degree of toxicity, considering the cellular and chromosomal levels of organization. As consequence, they work as an alert about the risks to the biota or to the human health (Martins et al., 2016; Singleton et al., 2016). A rapid protocol for environmental diagnosis is proposed in this work, based on the hypothesis that a one-sampling campaign can generate robust information when different types of environmental indicators are considered and compared. To test this, the results from different predictors were compared witch other and with the land uses in an approach based on the weight of evidence (i.e. concordance in results). To do so, in the present work, each result was considered as an evidence that can give support or contradict the others (Lowell et al., 2000; Benedetti et al., 2012; Brandt et al., 2016). The aim of this work was to take a rapid and comprehensive analysis of the situation of the Itupararanga reservoir and to verify if a one-sampling campaign with distinct indicators are informative for monitoring purposes. For this, physical, chemical and biological analysis were performed. Study site: The Itupararanga is a tropical reservoir inserted in an urbanized region. It is formed by three main rivers (Figure 1) that form, posteriorly, the Sorocaba River – the main former river of the left margin of Tietê River. The waters from Itupararanga reservoir are used for many purposes as water supply for around a million people, power supply and to the irrigation of crops (Smith et al., 2005). The reservoir also suffers some negative impacts. The surrounding lands had the original vegetation suppressed in a great extension that passed to be occupied for gated communities, small farms and pastures. Sewage discharges, especially in the former rivers are also reported (Manfré et al., 2011; Taniwaki et al., 2013). 2. Material and Methods Samplings and in situ analyses were performed at September 3rd 2013 at five sampling points along the transitional and lacustrine zones of Figure 1. Itupararanga reservoir localization. 4 Beghelli, F.G.S. et al. Acta Limnologica Brasiliensia, 2016, vol. 28, e6 the central body of the reservoir (Beghelli  et  al., 2012). A schematic map of the reservoir with the localization and geographic coordinates is drawn in Figure  2. The predominant anthropic uses were qualitatively determined based on former researches (Freitas  et  al., 2008; Manfré  et  al., 2011; Taniwaki et al., 2013) as also it was based on the interpretation of satellite images and field observations. Sampling point P1 can be described as an area over intense multiple rural uses, mainly cattle raising, nude soil and eucalyptus plantation. In this area it can be also noticed the lower native vegetation proportion. The P2 is near of the greatest area with natural vegetation. However, the left margin in P2 area is entirely converted for agricultural purposes. Sampling point P3 area has also the right side well preserved but the left one is mainly occupied by an urbanized zone but in lower intensity then P4 and P5. The P4 and P5 are in the most urbanized areas along the reservoir wherein P5 is the one that must receive the greatest pressure from this category of land use (Figure 3). There were determined physical, chemical and biological variables for water quality characterization in five sampling points in the Itupararanga reservoir. The performed analysis as also the abbreviations considered for the present work, the method and references are described in Table 1. As a complement, phytoplankton and benthic macroinvertebrates communities from the first (P1) and last (P5) sampling points were analyzed. Samples of surface water were collected using nets with a mesh size of 20 µm and fixed with formaldehyde (4%) for qualitative analysis of phytoplankton. For quantitative analysis, raw water samples were fixed with acetic lugol (1%), and counting was performed by the method described by Utermöhl (1958), using sedimentation chambers and an inverted microscope (Zeiss Axiovert Scope A1). Taxonomic identification was performed until class level according to Bicudo & Menezes (2006) and Sant’Anna et al. (2006). Benthic macroinvertebrates were collected by using a Van Veen dredge (area=0.04 m2). The dredge was launched twice to compose one cumulative sampling per sampling point. Samples were sieved in the field (0.212 mm) and fixed with formaldehyde 4%. Benthic macroinvertebrates were counted and then identified under optic microscope (Zeiss Scope A1) until subfamily level for Oligochaeta, tribe for Chirnomidae and genus Figure 2. Schematic map of the Itupararanga reservoir showing sampling points. The arrow indicates the approximate dam position in the reservoir. 5 Can a one-sampling campaign produce robust results… Acta Limnologica Brasiliensia, 2016, vol. 28, e6 to Chaoboridae. The identification was based on the following identification keys: Froehlich (2007), Marchese (2009), Trivinho-Strixino (2011). A multimetric index was recently suggested as an alternative when aiming to a rapid assessment tool. The index is based on features that can be recognized by trained people but without the requirement of an expertise in taxonomy. The index is based on coarse taxonomic resolution and it was applied to benthic macroinvertebrates assessment. The final index is the simple mean of seven other index previously standardized to a 0-1 scale where 1 represent the best condition and 0 the worst. The indexes comprehended in the final multimetric index are mainly based on the ratios between sensitive and tolerant taxa or groups as bloodworms Figure 3. Satellite picture showing land uses around the sampling points: (A) P1; (B) P2; (C) P3; (D) P4 and P5. Source: Google Earth. Table 1. Variables analyzed, methods and references. Analysis1 Methods Reference pH Portable phmeter (in situ) ----------- EC (µS.cm–1) Portable condutivimeter HI 98304 (DiST4) (in situ) ----------- DO (mg.L–1) Titulometria Winkler (1888) Secchi (m) Secchi disk Esteves (2011) T (oC) Digital thermometer ----------- SM / DM (mg.L–1) Gravimetric Wetzel and Likens (1991), Paranhos (1996) TN (µg.L–1) Spectrophotometry HACH TP (µg.L–1) Spectrophotometry HACH Chl-a (µg.L–1) Spectrophotometry Wetzel & Likens (1991) Phyto (ind.L–1) Phytoplankton mesh (0.25 µm opening), four times dragging. Wetzel & Likens (1991) Utermohl (1958) IMMIB Two dredging in the littoral region, sieved over 0.2 mm mesh opening, multimetric approach. Beghelli et al. (2015) Col (MPN.100mL) Colorimetric (Colillert - IDEXX Quanti-Tray 2000) APHA (2005) TSI of Lamparelli Spectrophotometry Beghelli et al. (2012), Cunha et al. (2013) IM / IA Allium cepa Tedesco & Laughinghouse (2012) 1EC = electric conductivity; DO = dissolved oxygen; Secchi = Secchi disk lecture; T= temperature; SM / DM = suspended matter / dissolved matter; TN water total nitrogen; TP = water total phosphorus; Chl-a = chlorophyll-a; Phyto = phytoplankton analysis; IMMIB = multimetric index for benthic macroinvertebrates; Col = fecal coliforms; IM and IA = index of mitosis and index of chromosome alterations based on tests with Allium cepa. TSI = trophic state index. 6 Beghelli, F.G.S. et al. Acta Limnologica Brasiliensia, 2016, vol. 28, e6 chironomids. The taxonomic diversity was also considered as a generic measure of the quality of the environment. This approach was previously suggested for the environmental assessment in the Jundiaí-Mirin basin (Beghelli et al., 2015). The taxonomic resolution considered here was thought to be achieved without the need of any technique of fixation or lightening. As a result, the identification could be performed by using temporary slides with just water. The groups considered as “tolerants” or “sensitives” were defined by considering previous works performed in tropical environments (Moreno & Callisto, 2006; Moretti & Callisto, 2005; Beghelli et al., 2012, 2014b). Tolerant groups were: bloodworms chironomids (red chironomids); tubificids and Chaoborus sp. Tanytarsinae was considered as a sensitive group. In the current format, the multimetric index is very similar to that suggested for Jundiaí-Mirin basin but it includes the taxon Chaoboridae as a tolerant group and the metric related to exotic species is not considered once that there was no exotic taxons in Itupararanga reservoir. The Hilsenhoff’s Family Biotic Index, FBI (Mandaville, 2002) adapted to a 0-1 scale was also considered. The calculation of the indexes are based on relative abundance (with exception to diversity and FBI). Density of benthic macroinvertebrates was also calculated by dividing the number of organisms collected by the dredge area. For genotoxicity test assays with Allium cepa Linnaeus (Amaryllidaceae) were performed according to Dietz & Herth (2011) and Tedesco & Laughinghouse (2012). Seeds of A. cepa were germinated until the roots achieve 2 cm. After this procedure, seedlings were transferred to tubes containing the treatments: control, P1 P2, P3, P4 and P5 where remained exposed to the treatments for 24 hours. Then, the roots were fixed in a solution of methanol: acetic acid 3:1 (v/v) for 24h. Following fixation, the roots were hydrolyzed with an HCl 1M solution at 60oC for 9 minutes. The material was colored by Schiff’s reagent. Slides were mounted by slicing meristematic region of the roots that was fixed by the technique of squashing with aceto carmin. There were mounted three slides per treatment, each of them containing, approximately 500 cells, totaling 1500 per assay from each treatment. The slides were analyzed under optic microscope (Zeiss Axiovert 60) and the number of cells in division and total of cells per slide was counted for the index of mitosis calculation (IM = total number of cells in division /total number of counted cells). The  number of cells with chromosome alterations – cells with retardatary chromosomes, chromosomes bridges and micronucleus - was also counted for index of chromosome alterations calculation (IA = sum of the cells with alterations / total number of cells in division) (Taniwaki et al., 2013). Considering the data obtained, a matrix with values standardized by z score calculation was produced for 13 variables: Fecal coliforms, suspended solid matter, dissolved matter, Secchi disk lecture, pH, electric conductivity, dissolved oxygen, total nitrogen, total phosphorus, index of mitosis, index of alterations and water temperature. By using z scores, the variance of each variable is considered as the fluctuations around the respective mean. As  consequence, the variation is standardized by this procedure that makes all of them be comparable each other (Triola, 2008). From this standard matrix a Principal Component Analysis (PCA), considering the variances, was performed by using PAST 2.17c software (Hammer et al., 2001). 3. Results In a general view, the waters from Itupararanga reservoir are alkaline, well oxygenated; it has moderate values of electric conductivity and transparence as also low quantities of suspended matter (Table 2). The values recorded for TN were low, below limit of detection in four sampling points. Nevertheless, the concentrations of total phosphorus were high and varied from 270 to 500 µg.L–1. By TSI calculations, the Itupararanga reservoir could be classified as in supereutrophic or hypereutrophic conditions (Table 3). Biological data: The values recorded for fecal coliforms were high. Values were above Brazilian limits for recreational bathing or ingestion (Brasil, 2012) at P1 and P5 (Figure 4). Table 2. Physico-chemical results. Variables1 P1 P2 P3 P4 P5 T 18.51 18.50 19.50 19.50 19.50 pH 8.90 8.70 7.80 8.20 8.80 EC 77.00 68.00 73.00 72.00 75.00 DO 12.52 12.48 9.84 11.80 9.83 Secchi 2.91 2.16 2.02 2.25 2.51 DM 21.50 42.90 30.30 25.80 32.70 SM 4.00 4.60 4.20 4.20 5.30 1T = temperature (°C); EC = electric conductivity (µS.cm–1); DO = dissolved oxygen (mg.L–1); DM = dissolved matter (mg.L–1); SM = suspended matter (mg.L–1) and Secchi = Secchi disk lecture (m). 7 Can a one-sampling campaign produce robust results… Acta Limnologica Brasiliensia, 2016, vol. 28, e6 The results of the multimetric index for benthic macroinvertebrates (IMMIB) calculated for the samplings from P1 and P5 points that the environmental quality is below satisfactory condition (Table 4). Density values varied between 495 in P1 and 1247 ind.m–2 in P5. The phytop lankton taxonomic c l a s s Cyanobacteria was predominant and the dominant species are Cylindrospermopsis raciborskii (Woloszynska) Seenayya and Subba Raju and Monoraphidium contortum (Thuret in Brébisson) Komárkova-Legnerová. (Table  5). The taxa Mycrocistis sp, Dolichospermum sp, Geitlerinema sp, Planktolyngbya sp, Planktothrix sp and Pseudoanabaena sp were also present in the samples. The mean values of the index of chromosome alterations (IA) were about 0.05 and 0.12. Values were above the control and the most significative differences occurred in samplings with water from P1 and P4. Values of index of mitosis (IM) were distinguishable from control only in the samplings treated with P3 water where highly IM was recorded. P1 and P4 presented mean values below control but with high variation. This can be interpreted as indicating a beginning genotoxicity (Figure 5). Principal Components Analysis (PCA) shows that the variables considered in this study could explain 69.77% of the observed variation. From this, 41.10% refers to component 1 (C1) and 28.67% to the component 2 (C2) (Figure  6). Table 6 presents r values in relation to the first three components. 4. Discussion Our results corroborate the hypothesis that it is possible to get sufficient information for monitoring based in on sampling campaign but considering distinct predictors. When comparing the results with satellite imagery it is clear that the most impacted areas (P1, P4 and P5) had poorer conditions. Furthermore, the predictors presented agreement or complementarity between them, but not disagreements. The present results corroborate in the sense that P1 and P5 are impaired. These sampling points were in eutrophic or supereutrophic Table 3. Trophic state. TN1 TP2 TN/TP3 Chl-a4 TSIChl 5 TSITP 6 TSIm 7 Classification8 P1 100 500 0.20 23.05 62.12 72.12 67.12 Hypereutrofic P2 ND 410 ND 22.45 61.99 70.92 66.45 Supereutrofic P3 ND 390 ND 12.83 59.24 70.62 64.93 Supereutrofic P4 ND 480 ND 12.83 59.24 71.87 65.56 Supereutrofic P5 ND 270 ND 20.72 61.59 68.39 64.99 Supereutrofic 1Concentrations of total nitrogen (TN, µg.L–1); 2 total phosphorus (TP, µg.L–1) the 3total nitrogen: total phosphorus ratio (TN/TP); 4chlorophyll-a concentration (Chl-a, µg.L–1) and the values calculated for the TSI based on 5chlorophyll-a – TSIChl – 6phosphorus - TSITP – as also the final mean 7TSI (TSIm). 8For purposes of classification it is considered: supereutrophic – 63