Design of a monitoring network and assessment of the pollution on the Lerma river and its tributaries by wastewaters disposal

Science of the Total Environment 373 (2007) 208 – 219 www.elsevier.com/locate/scitotenv

Design of a monitoring network and assessment of the pollution on the Lerma river and its tributaries by wastewaters disposal C. Fall ⁎, A. Hinojosa-Peña, M.C. Carreño-de-León Universidad Autónoma del Estado de México, Centro Interamericano de Recursos del Agua, Facultad de Ingeniería, Mexico Received 7 June 2006; received in revised form 20 October 2006; accepted 30 October 2006 Available online 19 December 2006

Abstract While the 2005 progress report of the United Nations Millennium Development Goals stresses out the need of a dramatic increase in investment to meet the sanitation target in the third world, it is important to anticipate about some parallel negative impacts that may have this optimistic programme (extension of sewer networks without sufficient treatment works). Research was initiated on Lerma River (Mexico), subjected to many rejects disposal, to design a monitoring network and evaluate the impact of wastewaters on its water quality. The discharges was inventorized, geo-positioned with a GPS and mapped, while the physico-chemical characteristics of the river water, its tributaries and main rejects were evaluated. Microtox system was used as an additional screening tool. Along the 60 km of the High Course of Lerma River (HCLR), 51 discharges, with a diameter or width larger than 0.3 m (including 7 small tributaries) were identified. Based on the inventory, a monitoring network of 21 sampling stations in the river and 13 in the important discharges (N 2 m) was proposed. A great similitude was found between the average characteristics of the discharges and the river itself, in both the wet and dry seasons. Oxygen was found exhausted (b 0.5 mg/L) almost all along the high course of the river, with COD and TDS average levels of 390 and 1980 mg/L in the dry season, against 150 and 400 mg/L in the wet season. In the dry season, almost all the sites along the river revealed some toxicity to the bacteria test species (2.9 to 150 TU, with an average of 27 TU). Same septic conditions and toxicity levels were observed in many of the discharges. Four of the six evaluated tributaries, as well as the lagoon (origin of the river), were relatively in better conditions (2 to 8 mg/L D.O., TUb 1) than for the Lerma, acting as diluents and renewal of the HCLR flow rate. The river was shown to be quite a main sewer collector. The high surface water contamination by untreated wastewaters that is depicted in this research should be taken into account in the Millennium Goals strategies, by promoting treatment plan works simultaneously, when sewer networks in the third world would extend. © 2006 Elsevier B.V. All rights reserved. Keywords: Lerma River; Microtox; Wastewater; Monitoring stations; Physical–chemical characteristics; Millennium goals

1. Introduction One of the targets of the United Nation's Millennium Development Goals (MDG,) is to halve, by 2015, the ⁎ Corresponding author. CIRA-UAEM, Apdo. postal 367, Toluca, C.P. 50091, Estado de México, México. Tel.: +52 722 2965550; fax: +52 722 2965551. E-mail address: [email protected] (C. Fall). 0048-9697/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2006.10.053

proportion of people without sustainable access to safe drinking water and basic sanitation (United Nations, 2005). While the 2005 progress report of the MDG stresses out the need of a dramatic increase in investment to meet the sanitation target, it is important to anticipate about some parallel negative impacts that may have this optimistic programme. Because the sanitation priority in developing countries often focuses in providing toilets in rural regions and

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

building sewer networks in large and medium cities, the delay that can suffer wastewater treatment plants construction, may have a drastic impact on surface water course subjected to receiving many of the rejects. To document the case of courses that have suffered from the effects of massive wastewater disposal is one way that has the international scientific community to advocate for simultaneous prioritizing of the construction of treatment plants. Untreated domestic wastewater, industrial discharges and agricultural activities, also with some treated effluents may contribute to surface waters pollution. Some recent examples of streams subjected to untreated wastewater discharges are the rivers Vrishabbavathy (India), Balatuin (Philippines), Lujin (Argentina), Nilufer (Turkey) and Ikpopa (Nigeria). In these cases, the negative impact resulted primarily from the high contents of organic matter (COD and BOD), nutrients (phosphorus and nitrogen), metals, and oxygen depletion (Ahipathy and Puttaiah, 2006; Dyer et al., 2003; Rodriguez et al., 2006; Karaer and Kuçukballi, 2006; Ekhaisei and Anyasi, 2005). Surface waters are also used for disposal of treated effluents from wastewater treatment plants (WWTP). These effluents usually contain only small amounts of various contaminants, but the harmful components accumulate over time in the rivers (Cotman et al., 2001). Kolpin et al. (2002) recognize that many industrial and household chemicals, pharmaceuticals and other consumables, as well as biogenic hormones are released directly to the environment after passing through wastewater treatment processes, which often are not designed to remove them. Roberts and Thomas (2006) showed that various pharmaceutical compounds are effectively reduced during their passage through a WWTP, whilst others are sufficiently persistent to occur in estuarine systems. The occurrence of organic wastewater contaminants (OWC) may probably be worst in rivers that receive untreated wastewaters. In the third world, many surface waters are known locally to be very contaminated but the cases rarely are documented, due to the lack of basic tools (such as established monitoring networks) as well as logistical and analytical capacities (specially for the persistent organic compounds). River quality data typically come from a monitoring network (Clement et al., 2006), which may be used, among many others objectives, to measure the beneficial effect of the installation of a WWTP, realize spatial–temporal studies on river water quality or evaluate the changes in trends of discharges practices. Unfortunately, to date, there is not a concise general strategy or methodology for designing monitor-

209

ing networks, especially when deciding the placement of sampling points (Strobl et al., 2006). Historically, there was a tendency to use experience, intuition and subjective judgement in locating the monitoring stations (Ning and Chang, 2004). Depending on the objectives, selection of the sampling sites may be biased (e.g, towards streams susceptible to contamination: Kolpin et al., 2002). Where it was necessary to design a more technically sound monitoring network, fuzzy multi-objectives evaluation framework and geographical information systems were alternatively used for identifying the critical sampling locations (Ning and Chang, 2004; Strobl et al., 2006). In the third world, there exists an important backward with respect to the capacity of analyzing the persistent organic pollutants (POPs), so it is necessary to reckon on some alternative methods that allow to easily detect the cases where serious toxic contamination occurred. One of the possibilities is to use the Microtox system (Azur Environmental Inc., California, USA) as a screening tool to evaluate the quality of surface waters and wastewater discharges. Compared to other bio-tests, it is admitted that the Microtox test is excellent about its sensibility, rapidity, simplicity, replicability, cost and discriminatory power. However, even if several studies have demonstrated its correlation with other bio-tests, its ecological relevance is frequently questioned (Bennett and Cuttage, 1992). Moreover, in the wastewaters ambit, Microtox was shown to be a proved sensitive test to toxicants but not at all representative of the effects of wastewaters on biological treatment plants (Fall et al., 2002; Gutierrez et al., 2002). The hydrologic system Lerma–Chapala constitutes one of the most important watersheds in Mexico. The region is drained by the rivers Lerma and Santiago as main collectors (1180 km long). The High Watershed of Lerma River is located in the State of Mexico (Fig. 1), subdivided in three portions called High Course, Medium Course and Low Course, with a total length of 175 km. The High Course (60 km) corresponds to the region near Toluca city that is located between the source lagoon and a point about 9 km down the river, at the José Antonio Alzate reservoir. In the past, the Lerma River ecosystem included different kinds of fish, crustaceans and marine bird species, of which what is left today, is only a memory. Demographic growth and concomitant industrialization have generated serious pollution problems in the Lerma River (De la Fuente et al., 2000). In the last years, some environmental studies were started on selected sites of the Lerma River (Avila-Pérez et al., 1999; Barceló-Quintal et al., 2000; De la Fuente et al., 2000), mostly in the Alzate reservoir, in order to better know the environmental conditions and suggest potential solutions.

210

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

currents zones, dead zones, sites of higher sediments accumulation, vegetation, habitats, river's width, in order to take them into account in the design of the monitoring network; c) Identify physically in the terrain the location of the sampling points used in previous studies to take them into account, when pertinent, for the design of the strategic network; d) Identify some particular zones that offer an ideal location and accessibility to receive the sampling stations for river and tributaries water, sediments and discharges; e) Take pictures and measure the geographical coordinates of several other points to make a map of the river. The visits were performed in both the wet and dry seasons to detect the changing configuration (temporal streams, submerged pipes in rainy season and vegetation). As a result, a detail map of the discharges was obtained, which was fundamental in the designing of the network. 2.2. Design of the monitoring network

Fig. 1. Location and configuration of the Lerma High Course River (HCLR).

Even if there is more information about the river, the combined integration and interpretation of these data will not be possible without a common data base, or that the sampling points used for each actor are not referenced or traceable. For this, the objective of this research was: 1st) to design a permanent strategic network of monitoring stations for the High Course of Lerma River (HCLR); 2nd) to use the stations of this network to evaluate the spatial and temporal variability of the course water quality and the impact of the discharges it receives.

It is not possible to always match the sampling points of different works, different objectives and different sampling densities. However, it is possible to make that all researchers select their sampling points within a same network with known coordinates. The monitoring network was integrated by stations in the heart of the river (type R) and stations located in the different discharges (type D). The type D stations were selected under the hypothesis that the larger discharges would be determinant to the rivers' water quality. On the other hand, the main criteria to set the stations location (type R) in the river were, that the sampling sites would be easily accessible and distributed along the HCLR, that there was a river station before and another after each important discharge or group of discharges, and whenever possible, to preserve most of the old sampling points used before by universities and water agencies. 2.3. Characterization of the river waters and of the discharges it receives

2. Methods 2.1. Sites survey, cartography and discharges census The zone of interest formed by the High Course of Lerma River (HCLR) goes from Chiconahuapan lagoon in Almoloya del Río, to 9 km down the exit of the Alzate dam (Fig. 1). A complete on-foot travel of the zone was taken in order to: a) Make an inventory and position in detail with a GPS the different wastewater rejects and the main tributaries; b) Identify the location of obstacles, river's infrastructure, as well as turbulence and strong

The characterization performed in this work was based on a limited number of discharge stations (type D) and river sites (type R), selected within the general monitoring network. The discharges to be monitored were selected among those that contribute with more flow to the river, including polluted or non-polluted tributaries (small rivers) that join in. All the stations in the river were 200 m or more from the large discharges, in places where the contaminant plumes have visibly finished mixing. The discharge sampling points were within the 30 m preceding their junction to the Lerma

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

river. Two sampling campaigns were performed in the selected locations for the monitoring, the first one in the wet season (October) and the other in dry season (March). Grab samples were taken in the center of the current, at a third of the depth. At the same time, the currents' velocity and parameters such as conductivity (Cond.), temperature (Temp.) and dissolved oxygen (D.O.) were measured in situ (portable equipment WTW P4 and YSI 57). The different samples were refrigerated at 4 °C with the adequate preservatives and were later analyzed to determine pH, turbidity (Turb), chemical oxygen demand (COD hatch kit), solids in all their forms (TTS: totals; TDS and VDS: total dissolved and volatile; TSS and VSS: total suspended and volatile), according to the standard methods (APHA, 1989). An aliquot of 50 mL, without a preservative, separated in glass bottles (free of air) since the sampling, was analyzed by the Microtox™ system (Azur Environmental Inc., California, USA) to trace the presence of toxic compounds (acute toxicity test). In general, the manufacturer's Standard Basic Test protocol was used with which each sample was analyzed in 4 dilution levels (between 5.6 and 45%). In cases where toxicity was low, aliquots were analysed again without dilution (81.9% Basic Test and 100% Whole Toxicity Test). The reactives and material (lyophilised Vibrio fisheri, osmotic adjusting solution, reconstitution solution, diluents, cuvettes and pipettor tips), along with the measurement instrument (M500 analyser) were acquired from the supplier's representative (SDI inc.). A standard of phenol and a blank were used as control, always complying with the quality control criteria. Toxicity data analysis and reports were performed automatically with MicrotoxOmni™ 1.18 Software. The obtained measurement was the effective concentration at 50% (EC50), later converted to toxicity units (TU = 100/ EC50). The samples showing hormesis or an EC50 higher than 100% (TU b 1) were considered non-toxic. 3. Results and discussions 3.1. Sites survey, cartography and discharges census 85 importance points were geopositioned, with a GPS-X, along the 60 km of the High Course of the Lerma River (HCLR) which allowed reconstructing the river into a map (Fig. 2). 51 discharges were inventorized, which included 21 canals (41% of the discharges), 19 tubes (37%), 4 brooks (8%) and 7 small rivers (14%) that flow into the Lerma. In Fig. 2, the discharges are identified as T1 to T52, also with a name that indicate the community or municipality of origin. The rejects were distributed along all the HCLR, from the Almoloya del

211

Rio lagoon (traditional origin of the river) to the Alzate dam (almost the HCLR limit). The set of information regarding the inventory was presented in an atlas with a more detailed individual description of each discharge or tributary of interest (Hinojosa-Peña, 2006). Most of the discharges come from different municipalities to evacuate the domestic and industrial wastewaters. Moreover, three important treatment plants in the Toluca city (2 municipal and 1 industrial WWTP, T22, T27, T14) discharge their effluents (approx. 2.5 m3/s) in the Lerma High Course. Also 7 small rivers (flag symbols), located all of them in the second half of the Course, do the same. The diameters or width of the more visible tubes, canals and tributaries vary from 0.3 m to 15 m, with an average of 2.8 m, a median of 0.75 m and a mode of 0.5 m. As shown by the Parreto diagram in Fig. 3, 40% of the discharges have a dimension smaller or equal to 0.5 m whereas those of 0.5 to 2 m are 25%, those from 2 to 5 m, 20% and those from 5 to 15 m, 15%. The small discharges (b0.5 m) usually come from very close zones to the river, whereas the tubes, canals and tributaries are collectors of several kilometers. It was inferred that the total contribution, in caudal and contaminant load, of the relatively large sized discharges (N2 m) could be determinant on the quality of the river. 3.2. Proposal of the monitoring network and its application to this work Based on the discharges data and the network design criteria, it was proposed a sampling network of 21 stations in the river (R1 to R20) and 13 in the important discharges (D1 to D13), all geo-positioned with a GPS so that every researcher could be able to select among them the ones that would be more relevant to their study in particular, without losing the traceability. The proposed stations appear in Fig. 2, first for the River's course (R1 to R21), and secondly for the discharges to be monitored (D1 to D13). From the application of the design criteria of the network that were mentioned in the methods, it resulted that some of the R stations are directly located under bridges or near the highways. Also the 8 R-stations traditionally used by the National Water Commission (C.N.A.) were kept, as well as two other points over the Alzate dam, where groups of researchers based their works in the past. The type D points were all located in discharges of 2 m or larger, selecting within this class, the most representative. For the application of the monitoring network, in this work all the 13 discharge stations (D) of the general network were selected, but not all the 21 type R sites were used (only 15 of them). Fig. 4 is a simplified partial view

212

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

Fig. 2. Map of the discharges and monitoring network in the HCLR.

of the monitoring network. Only the discharges and river sites that were sampled in this work are shown, also visualizing the relative position of the rejects with respect to the river stations. In this graphic, the tributaries marked with a flag correspond to small rivers that feed the Lerma (Xonacatlán, S. Catarina, San Lorenzo, Temoaya, Verdiguel and Tejalpa). In the lower part of the figure, the sequential order of all the sampling points (type R and D together) is given, for its use in later graphics as reference. The first two stations of the network (R1 and R2) are located in the lagoon which was in the past the origin of the

river and that nowadays is isolated with a soil barrier to avoid that, in summer times, the direction of flow inverts and part of the discharges reach the lagoon. Just after the lagoon, at the beginning of the river course, are R3 (not used in this work) and R4, which allow to evaluate the water quality before and after the first important discharges, the Tenango Drainage and the Mezapa Stream (D1 and D2), which are part of the discharges monitoring network. The rest of the first half of the HCLR is covered by stations R5, R6 and R7, focused on detecting the impact of the municipal drainages of Chapultepec, San

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

213

installed D-sampling station in the last length since there is no any important discharge in that zone. 3.3. Characteristics of the river waters

Fig. 3. Pareto diagram for discharge sizes.

Lucas Tunca and Ocoyoacac (D3), Metepec and San Mateo (D4), as well as the discharge from a WWTP of an industrial park (D5). The R8 to R12 points are placed between the 7 small rivers. In this zone, six out of the 7 small tributaries, as well as two canals that conduct the effluents of the two macro-plants of Toluca constitute the rest of the monitoring network for the discharges (D6 to D13). The Alzate reservoir, located in the last length of the HCLR, accommodates 6 R-stations (R13 to R18, to conserve the sampling sites intensively used by several researchers in the past). From those, only two (R15 and R16) were included in the monitoring performed in this work. The last stations of the R-network (R20 and R21) were located between the dam curtain and the HCLR limit (9 km down the curtain), in order to detect the changes suffered at the exit of the reservoir and follow their behavior along two more locations. There was no any

Table 1 and Fig. 5 show the spatial and temporal variation of the water characteristics in the river and in the discharges. By focusing on the spatial variability of the R stations parameters (inside the course), it is observed that the quality of the water column is deteriorated since the beginning, in all the seasons, as it is notably put in evidence by the dissolved oxygen profiles (Fig. 5A). The D.O. that reaches relatively high values in the first part of the lagoon (5 and 8 mg/L), all of the sudden relapses to levels between 0 and 1 mg/L before exiting the lagoon, as well as after leaving it. At the end, but only temporally, it was observed that the water is re-oxygenated due to the fall of the liquid from the Alzate dam floodgate (point numbered as 27). No matter the periods, the COD reaches values higher than 150 mg/L from the first sampling points, inside as well as outside the lagoon (Fig. 5B). Throughout the thirty COD measurements performed during the two periods, only 4 presented values lower than 100 mg/L (90, 83, 56 and 64 mg/L in wet season, respectively for the points numbered by #5, 21, 25 and 26, or stations R5, R11, R15 and R16). In the dry season sampling, the COD fluctuated from 190 to 1200 mg/L depending on the sites. Also for the dry season (only), conductivity and total dissolved solids (Fig. 5C), as well as toxicity (Fig. 5D) showed relatively high values: on average, conductivity of 1300 in dry season against 700 μS/cm in wet season, 14 TU (EC50 =7%) in dry season against less than 1 TU (EC50N 100%) in general, during the wet season.

Fig. 4. Simplified partial view of the designed monitoring network. D: discharge stations; R: river stations; Flags: tributaries.

214

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

Table 1 Characteristics of river waters and discharges Parameters

River

Discharges

Dry season

pH (units) Cond (μS/cm) Turb. (NTU) D.O. (mg/L) COD (mg/L) TSS (mg/L) TDS (mg/L) VSS (% of TSS) VDS (% of TDS) Toxicity (TU) #

Wet season

Dry season

Wet season

Mean

Range

Mean

Range

Mean

Range

Mean

Range

7.5 1344 161 1.2 388 164 1978 71% 31% 14.3

7.0–8.5 936–2040 49–598 0–8.0 188–1206 66–770 1522–2558 52–89% 22–41% b1–151+

6.9 684 61 1.3 149 92 401 43% 25% 1.9

6.4–7.2 461–1532 7–207 0–5.0 56–209 17–235 275–600 18–88% 11–40% b1–15

7.2 1079 136 2.1 364 124 2106 74% 31% 5.1

5.5–7.6 144–4860 44–307 0–8.0 21–1312 30–316 1759–3372 44–92% 22–39% b1–29

7.1 698 68 1.6 112 115 460 40% 25% 1.5

6.0–8.6 108–3000 17–209 0–6.4 36–224 36–220 90–1368 15–74% 13–51% b1–170+

Cond: Conductivity; Turb: Turbidity; D.O.: Dissolved Oxygen; COD: Chemical Oxygen Demand; TSS: Total Suspended Solids; TDS, Total Dissolved Solids; VSS: Volatile Suspended Solids; VDS: Volatile Dissolved Solids. # TU: Toxicity Units = 100/EC50d +: singular points (isolated maximum) not included in the averages.

The averages for turbidity and total suspended solids (TSS) in the river were about 170 NTU and 164 mg/L TSS against 61 NTU and 112 mg/L TSS, in low and high waters, respectively. The maximum values were manifested in low waters with 600 NTU of turbidity and

770 mg/L TSS. Oil and grease measured in a limited number of points were about 35 mg/L in average in both campaigns. About the Microtox (Fig. 5D), it is important to observe that the two sites located in the lagoon (number

Fig. 5. Spatial and temporal variability of the characteristics of the river and its discharges.

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

of order 1 and 2 or R1 and R2) do not reveal toxicity (b1 TU) in all the periods, independently that the COD, in particular, was between 170 and 250 mg/L in these sites. The no-detection of toxicity and the existence of fish in some parts to the lagoon lead to the conclusion that this water body has a much better quality than the HCLR. However, the oxygen oversaturation of the lagoon by moment (up to 8 mg/L) during day-time, due to the presence of algae, also means a real risk of oxygen deficit during the night, which may explain fish-death episodes that are occurring in this place. The environmental recovery of the lagoon is feasible, needing only some pollution control measures around the water body. Besides the action already taken of closing the superficial communication between the lagoon and the river, it is necessary to eliminate all the small and large discharges to reduce the COD and dissolved solids, and hence, the increasing proliferation of algae. During the wet season, water is not toxic (b1 TU) in almost all the sampling points of the river water column, except for the third point (R4, 15 TU), apparently affected (just locally) by the first important discharge (D1) that has an exceptionally high toxicity of 170 TU. For the dry season campaign, the trend is the opposite: except in the two points of the lagoon (as discussed previously), all the samples were showing some toxic characteristics (3 to 150 TU). The maximum of 150 TU detected locally in the R11 site (point numbered # 21) is not explicable only by the monitored discharges which show a much more less toxicity. When using the Microtox for river monitoring (aquatic-life protection), currently, intervals of b 1 TU are considered as “non-toxic”, 1.01 to 1.70 as “slightly to moderately toxic”, from 1.8 to 2.6 as “toxic to very toxic” and N 5.3 TU as “extremely toxic” (Bennett and Cuttage, 1992). In the present research, for comparison, the waters from a local municipal wastewater treatment plant were evaluated, finding a toxicity of 9.24 TU in the influent and b 1 TU in the effluent (hormesis). The latter shows that due to its high sensibility, the no-detection of toxicity in the lagoon and other sites means the guarantee of non-distress, whereas the medium values (around 9 TU) may only mean the presence of more or less loaded normal municipal wastewaters. This is in concordance with the results of Gutierrez et al. (2002) who evaluated the responses of activated sludge and Microtox bacterial species to wastewaters. The HCLR does not comply with the requirements for non-restricted irrigation, as regulated by the Mexican Agreement CE-CCA-001/89 that stipulates the Ecologic Criteria used to classify water bodies in the country (DOF, 1989).

215

3.4. Discharges characteristics and their impacts In this section the results of all the discharges in general are included, whereas in the next section an emphasis on tributary rivers shall be made. The standards that regulate the discharges in the Lerma River are established in the Mexican Norm, NOM 001 (DOF, 1997). For the vocation of use in irrigation that the course has, the accepted limits for rejects are: T max of 40 °C, pH of between 5 and 10, and average monthly concentrations smaller to 150 mg/L of TSS and 150 mg/L of BOD5, among other parameters. The information about the discharges characteristics is also summarized in Table 1 and Fig. 5. The maximum temperature (T) detected in the discharges was of 32 °C for the D5 station (effluent from a treatment plant of an industrial park), but there was not any exceeding of the Norm with respect to T. A low pH (5.5) was detected in the discharge from Tenango (D1, point number # 3 in Fig. 5A: locality with an important industrial park). There was not any violation of the pH limits, but the minimum value of 5.5 detected in D1 constitutes a first alert about the nature of this discharge. In high waters, there was an exceeding in the limits for the TSS (graphic not shown) in 5 of the 13 evaluated discharges (D1, D5 and tributaries D7, D10 and D12, with 162 to 200 mg/L TSS). In low waters, D1 and D12 were still above the limits (with 210 and 167 mg/L TSS) and two more discharges were added (D4 and D6 with 316 and 210 mg/L TSS). As observed in-situ, the suspended solids, in the case of the tributary rivers D7, D10 and D12, were the result of an erosion (inorganic matter), rather than pollution. The bulk of the total solids was almost composed by the dissolved fraction (80 to 93% in average). In the dry season, the total and dissolved solids in the discharges (as well as in the river waters) were particularly high (1760 top 3370 mg/L TDS, with an average of 2100 mg/L TDS, Fig. 5C). In the wet season, the dilution substantially lessens the total and dissolved solids concentrations (90 to 1370, with an average of 460 mg/ L TDS). The maximum TDS values were registered in the discharge D5 (point # 10 in Fig. 5C). About the oxygen demand, the standards for discharges are stipulated in BOD5 (150 mg/L as monthly average), while the measurements in this research were done in COD. With a conservative approach (BOD5/COD ratio of 0.4; Metcalf and Eddy, 1991), not any discharge would exceed the 150 mg/L BOD5 limit during the rainy period, whereas in low waters, the D1 discharge (Tenango industrial park, 525 mg/L), D5 (industrial plant effluent, 400 mg/L) and D6 (Toteltepec Stream, where the effluent

216

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

from a municipal plant joins, 212 mg/L) would probably exceed the BOD5 limits. The parameters of COD, dissolved oxygen, conductivity, dissolved solids and toxicity are not included in the current regulation of the discharges. However, they allow appreciating even better the conditions of the rejects and their impact. Generally, in Fig. 5, except for the D.O., every time when a discharge has shown a relatively high value of a parameter, there was an evident response in the river (peaks denoting a local impact of some discharges). In Fig. 5A, it can be seen that, even if some discharges (mainly the tributary rivers) get to the Lerma River with high oxygen levels, they do not succeed in elevating the low O2 levels in the HCLR, due to the high COD. As it was mentioned before, the temporal increment of D.O. in the penultimate station of the river, was only the result of an aeration due to the liquid's fall to the Alzate dam floodgate (point # 27). The two first COD and TSS peaks, that are perceptible in the dry season data, are imputable to D1 and D5 (1300 and 1000 mg/L COD, Fig. 5B, points # 3 and 10). The highest conductivity (Fig. 5C) still takes place in D5 (4800 and 3000 μS/cm depending on the period). In the wet season, the first of the only 3 discharges that show toxicity is, precisely, the D1 (170 TU) with a noticeable impact in R3. The other two discharges with a toxic nature were D5 with 7 TU and mildly D6, with 1.5 TU. On the contrary, in the dry season, almost all the discharges (except D8 and 4 of the 6 tributary rivers evaluated) revealed a toxic nature (between 2 and 29 TU). The maximum was registered in D1 and D5. Among the evaluated rejects, the D1 and D5 discharges outstand as the ones of higher impact, however, in dry season, out of the 4 tributary rivers and the canal of one of the municipal plants (D8), all the large discharges contribute to the deterioration of river Lerma. Apart from the two first points located in the lagoon, all the river sites were affected (3 to 54 TU and a point at 150 TU, Fig. 5D). 3.5. Particular conditions of the tributary rivers and their impact With respect to the tributaries, the BOD5 of the water, estimated through a conservative BOD5/COD ratio of 0.4, was not exceeding the oxygen demand limits for use in irrigation (see Fig. 4 concomitantly with Fig. 5 for identifying the position of the different tributaries). Within the 6 small rivers that were evaluated, the Verdiguel (D11) would present a worse condition with an estimated BOD5 of 90 mg/L in high waters and 120 mg/L in low waters, followed by the river Xonacatlán (D7, with 16 mg/L in high waters and 125 mg/L in low waters). The other tributary rivers, with an estimated BOD5 average of 24 mg/L,

in high waters and 31 mg/L in low waters, are in a much better condition than the Lerma (60 and 155 mg/L). They act as diluents of the waters and contribute to maintain the flow rate in the Lerma River. The quality of the Temoaya River (D13, between 8 and 15 mg/L of BOD5) for the two periods and the Santa Catarina River (D9, with 14 to 17 mg/L) outstand in that matter. In fact, the condition of the tributary rivers, D7 and D11 on one hand (group 1), and D9, D10, D12 and D13 on the other (group 2), is confirmed with their clearly discriminated levels of dissolved oxygen (Fig. 5A, last discharge sites), conductivity (Fig. 5C) and toxicity (Fig. 5D). In particular, none of the 4 tributary rivers of group 2 showed toxicity, contrary to group 1 (D7 and D11, with 2.8 and 8.6 TU, in the dry season). The oxygen was very low in the Verdiguel and Xonacatlan (D11 and D7, group 1: between 0.2 and 0.7 mg/L in the two periods), contrary to the other rivers (group 2) where it was between 1.9 and 6.4 mg/L in the rainy period, against 4 to 8 mg/L in the dry season. The highest O2 levels were obtained in the Temoaya, followed by the Catarina (D13 and D9, included in group 2). If in the wet period none of the tributary rivers showed toxicity, the Verdiguel and Xonacatlan (D11 and D7: group 1), in low waters, were again pointed with certain toxicity (2.8 and 8.6 TU), opposite to the other 4 rivers considered clean (Fig. 5D, within the last 7 discharge points). Temoaya, Tejalpa, San Lorenzo, and Santa Catarina (D13, D12, D10 and D9: group 2) do not reveal any toxicity in the two sampling periods. In summary, 4 of the tributaries are in relatively good conditions whereas the other are more deteriorated. The information on the conditions of the different tributaries and the lagoon should influence the recuperation strategies of the HCLR. 3.6. Similitude between the average characteristics of the river and its discharge From Fig. 5 it can be seen that the characteristics of the discharges, as the river, present a significant change in their magnitude order, between the dry and rainy seasons (except for pH). Other very noticeable general trend in this figure, for several of the represented parameters, is the closeness of the average characteristics of the river water, compared to the discharge waters, in each same season. The information presented in Table 1 gives evidence of a great similitude between the average characteristics of the waters, in the Lerma river versus in the discharges, in the same season (dry season or wet season). It can be observed, for example, that the average conductivities are

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

of 1344 (river) and 1079 μS/cm (discharge) in wet season against 684 and 698 μS/cm in dry season, for the river and discharge waters respectively. Other illustrative pairs are (161, 136) and (61, 68) NTU for turbidity, (388, 364) and (149, 112) mg/L for the COD, (164, 124) and (92, 115) mg/L for TSS. More convincing, the pairs of data relative to the content of the volatile organic matter (in % of the TSS or % of the TDS) present very similar signatures. The VSS average contents in the river vs in the discharge were 71 vs 74% for the dry period, and 43 vs 40% for wet season. With respect to the VDS, the river vs the discharge average contents were 31 vs 31% and 25 vs 25%, in the dry and wet seasons respectively. Fig. 6 represents the parameter averages and their respective 95% confidence limits. It allows realizing a statistical comparison of the means, between discharges and river, in the same season (dry or raining season). The error bars in Fig. 6 show that in general there are no significant statistical differences in the different pairs of

217

data (river vs discharge) at a same season period. The observed similitude was obtained in a context where the measured individual values, which are included in the calculation of the averages, correspond to discharges of varying quality levels. This can be appreciated with the ranges given in Table 1 and in Fig. 5. Moreover, Fig. 7 is a way to quantitatively analyze the change tendencies in each parameter from one season to the next (dry season vs wet season), for the water column on one side and on the other for the discharges. The D function represented is a dilution factor for each parameter, calculated as the relation between the average in the dry season and the average in the wet season. The sequence of the parameters in the figure was arranged in ascending order through the value of D in the river. The parameters that suffer most change from one period to the other are those that have a lower D value, i.e, those that appear at the beginning of the x-axis.

Fig. 6. Statistical comparison of the means from the discharges and river at each same season (RD: river—dry season; DD: Discharges—dry season; RW: river—wet; DW: Discharges—wet).

218

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

charged from the third WWTP (D5, industrial park) does not comply the required quality. From the situation depicted above, it is clear that increasing the number of municipal WWTP will result in an almost immediate positive effect in the river. At the same time, there is a need to change the strategy that intents to treat the wastewaters from different types of industries in a park, by use of a unique centralized WWTP. 4. Conclusions Fig. 7. Dilution factor for the different parameters.

The parameters that presented a major change from one season to the other, for the river, are toxicity, dissolved solids and the total solids (D = 0.13, 0.20 and 0.23, respectively). Opposing, the D.O. and pH are the parameters that suffer the less changes (for the river water D = 1.09 and 0.92, respectively), being both of them very stable, the first with very unfavourable concentration levels, and the second, in a very adequate range. Now it is possible to compare the dilution factor for the river water vs the discharge waters, by the sequence that follow the parameters and by the difference in the magnitudes of D. If the sequence of the parameters is not exactly the same for the discharges compared to the river, for none of them, the order has changed in more than two units. On the other hand, the value of the dilution factors for 7 of the parameters concurs relatively well (average difference of less than 20%). Two of the three parameters that have differences higher than 20% are the D.O. and the toxicity, parameters that not necessarily keep their values between the entrance and the exits from a system, due to possible external inputs, and synergic or antagonist effects. The evidences shown support the finding of a great similarity between the discharges and river average characteristics, establishing that the deterioration of the water bodies is the result of the extension of the sewer network, without the development of an adequate and sufficient treatment infrastructure. The pollution of the river comes mainly from the discharges that conduct the wastewaters (WW) from several municipalities. While the sewer coverage reaches more than 90%, it is known that only about 20% of the wastewaters are treated (CNA, 2005). The two big municipal WWTP comply relatively well their function, but their volumetric capacity is far to cover the needs. Also, the well treated effluent from one of the plants arrives the river as a re-contaminated stream (D6) since some nontreated flows are by-passed to the Lerma through the same canal D6. On the other hand, the water dis-

Along the 60 km of the High Course of Lerma River, 51 discharges with a diameter or width larger than 0.3 m were registered, including 7 small tributaries that flow into the Lerma River. The more visible tubes, canals and tributaries have a diameter or width that vary from 0.3 to 15 m, with an average of 2.8 m, a median of 0.75 m and a mode of 0.5 m. Based on the inventory of the discharges, a monitoring network of 21 stations in the river and 13 in the important discharges (N 2 m) was proposed. The application of the network to monitor the discharges and the Lerma waters quality showed a great similitude between the average characteristics of the inflows and the river itself, in both the wet and dry seasons. The average dilution ratio of the parameters (mean conc. on the wet season/mean on dry season) was quite similar for the river and the discharges. This showed that the state of the river corresponds exactly to the average discharge it receives, without any self-depuration or dilution capacity. In the two evaluated periods, oxygen was found exhausted (b0.5 mg/L) almost all along the river, with COD and TDS average levels of 390 and 1980 mg/L in dry season, against 150 and 400 mg/L in wet season. In the wet season, the river water, in general is not toxic (TU b 1) whereas that in dry season, excluding the lagoon, all the sites revealed some toxicity in the Microtox system (2.9 to 150, with an average of 27 TU). Among the evaluated discharges, two outstood as the ones with the most important impact, being the parameter values measured at these points the more extreme obtained during the campaigns (up to 4860 μS/cm conductivity, 1300 mg/L COD, 170 TU and pH of 5.5). However, in the dry season, out from the 4 tributary rivers and the effluent of one of the municipal plants, all the large discharges contribute to the deterioration of the River Lerma. Four of the six small evaluated tributaries, as well as the lagoon, are in relative good conditions (2 to 8 mg/L D.O., TU b 1), acting as diluents and renewal of the HCLR flow rate. The other two evaluated tributaries showed evidences of deterioration. Moreover, it is important to note that, in general, the results of the Microtox system were coherent

C. Fall et al. / Science of the Total Environment 373 (2007) 208–219

with the diagnosis of good or bad water qualities performed through the physico-chemical parameters, at the different sites. In summary, the situation of the HCLR, as diagnosed in this research, is that the river works as a sewer main collector. Its restoration will necessarily pass by the construction of the required treatment plants, but in the first place, a protection plan for the components that are not very deteriorated, which are the tributary rivers and the lagoon, is of the utmost urgency. The high surface water contamination by untreated wastewaters that is depicted in this research should be taken into account in the Millennium Goals strategies, by promoting treatment plan works simultaneously with the extension of the sewer networks. Acknowledgements This research was supported by the Mexican National Council of Science and Technology (grant # CONACYT 37909T). The authors wish to thank the National Institute of Nuclear Research of Mexico (Dr. Pedro Avila and his team, ININ), for their support during the sampling campaigns. References Ahipathy MV, Puttaiah ET. Ecological characteristics of Vrishabhavathy River in Bangalore (India). Environ Geol 2006;49(8):1217–22. APHA. Standards methods. 17 th ed. Washington D. C., USA: APHAAWWA-WPCF; 1989. Avila-Pérez P, Balcázar M, Zarazúa-Ortega G, Barceló-Quintal I, DíazDelgado C. Heavy metal concentrations in water and bottom sediments of a Mexican reservoir. Sci Total Environ 1999;234(1–3):185–96. Barceló-Quintal ID, Solís CE, González CC, Avila PP, García JA. Determination of cadmium and lead species in the water column of the Jose Antonio Alzate Reservoir, Mexico. Water Environ Res 2000;72(2):132–40. Bennett J, Cuttage J. Review and evaluation of Microtox test for freshwater sediments. Report of the Washington State Department of Ecology, Olympia, Washington; 1992. 32 pp. Clement L, Thas O, Vanrolleghem PA, Ottoy JP. Spatio-temporal statistical models for river monitoring networks. Water Sci Technol 2006;53(1):9–15. CNA. Inventario Nacional de plantas municipales de potabilización y de tratamiento de aguas residuales. Report of the National Water Commission (CNA), Mexico D.F; 2005. 177 pp.

219

Cotman M, Zagorc-Koncan J, Drolc A. Study of impacts of Treated Wastewater to the Krka River (Slovenia). Water Sci Technol 2001;44(6):47–54. De la Fuente JL, Hernández G, Marín LE. Heavy metal concentration in the sediments of the upper Lerma Basin, Mexico. In: Bachuys Publisher, Leiden, the Netherlands, editor. Aquatic Ecosystems of Mexico; 2000. p. 15–27. DOF. Norma CE-CCA-001/89: criterios ecológicos de calidad del agua. Diario Oficial de la Federación, Dic., 2; Mexico D.F., Mex; 1989. DOF. NOM-001-ECOL-1996. Diario Oficial de la Federación, Jan, 6; Mexico D.F., Mex; 1997. Dyer SD, Peng C, McAvoy DC, Fendinger NJ, Masscheleyn P, Castillo LV, et al. The influence of untreated wastewater to aquatic communities in the Balatuin River, the Philippines. Chemosphere 2003;52(1):43–53. Ekhaisei FO, Anyasi CC. Influence of breweries effluent discharge on the microbiological and physico-chemical quality of Ikpoba River (Nigeria). Afr J Biotechnol 2005;4(10):1062–5. Fall C, Cuenca-Mendoza F, Jiménez-Moleón MC. Patrón de contaminación de las aguas residuales procedentes de laboratorios universitarios. Inf Tecnol 2002;13(1):55–60. Gutierrez M, Etxebarria J, de las Fuentes L. Evaluation of wastewater toxicity: comparative study between Microtox (R) and activated sludge oxygen uptake inhibition. Water Res 2002;36(4):919–24. Hinojosa-Peña A. Diseño de una red estratégica de monitoreo para el curso alto del río Lerma, M.Sc. dissertation, Mexico State Autonomous University, Toluca, Mexico, 2006. Karaer F, Kuçukballi A. Monitoring of water quality and assessment of organic pollution load in the Nilufer Stream (Turkey). Environ Monit Assess 2006;114(1–3):391–417. Kolpin DW, Furlong ET, Meyer MT, Thurman EM, Zaugg SD, Barber LB, et al. Pharmaceuticals, hormones and other organic wastewater contaminants in U.S. streams, 1999–2000: a national reconnaissance. Environ Sci Technol 2002;36:1202–11. Metcalf and Eddy Wastewater Engineering. Treatment, disposal, reuse. 3d ed. New York: McGraw-Hill; 1991. 1334 pp. Ning SK, Chang NB. Multi-objective, decision based assessment of a water quality monitoring in a river system. J Environ Monit 2004;4(1):121–6. Roberts PH, Thomas KV. The occurrence of selected pharmaceuticals in wastewater effluent and surface waters of the Lower Tyne catchment. Sci Total Environ 2006;356(1–3):143–53. Rodriguez PL, Pizarro H, Maidana N, Afonso MD, Bonaventura SM. Epixylic algae from a polluted lowland river of Buenos Aires Province (Argentina). Cryptogam Algol 2006;27(1):63–83. Strobl RO, Robillard PD, Shannon RD, Day RL, McDonnell AJ. A water quality monitoring network design methodology for the selection of critical sampling points: part i. Environ Monit Assess 2006;112(1–3):137–58. United Nations. The millennium development goals, 2005 report. New York, USA: United Nations Department of Public Information (DPI/2393); 2005. 45 pp.