Scholarly article on topic 'Assessing seawater intrusion in an arid coastal aquifer under high anthropogenic influence using major constituents, Sr and B isotopes in groundwater'

Assessing seawater intrusion in an arid coastal aquifer under high anthropogenic influence using major constituents, Sr and B isotopes in groundwater Academic research paper on "Earth and related environmental sciences"

CC BY
0
0
Share paper
Keywords
{"La Paz aquifer" / Overexploitation / Salinization / "Cationic exchange pool" / " 87Sr/86Sr" / δ11B}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — J. Mahlknecht, D. Merchán, M. Rosner, A. Meixner, R. Ledesma-Ruiz

Abstract The La Paz aquifer system (Baja California Sur, Mexico) is under severe anthropogenic pressure because of high groundwater abstraction for urban supply (city of La Paz, around 222,000 inhabitants) and irrigated agriculture (1900ha). In consequence, seawater has infiltrated the aquifer, forcing the abandonment of wells with increased salinity. The objective of this study was to assess seawater intrusion, understand the hydrogeochemical processes involved and estimate the contribution of seawater in the wells tested. The aquifer comprises mainly the alluvial filling and marine sediments of a tectonic graben oriented north-south, in contact with the Gulf of California. Groundwater samples were collected in 47 locations and analyzed for major constituents. A subset of 23 samples was analyzed for strontium and boron concentrations and isotopic signatures (87Sr/86Sr and δ11B). Results were interpreted using standard hydrochemical plots along with ad hoc plots including isotopic data. Seawater intrusion was confirmed by several hydrogeochemical indicators, such as the high salinity in areas of intense pumping or the Na+-Ca2+ exchange occurring in sediments that were previously in chemical equilibrium with fresh water. However, seawater contribution was not sufficient to explain the observed concentrations and isotopic signatures of Sr and B. According to the isotopic data, desorption processes triggered by a modification in chemical equilibrium and an increase in ionic strength by seawater intrusion significantly increased Sr and probably B concentrations in groundwater. From a calculation of seawater contribution to the wells, it was estimated that one-third of the sampled abstraction wells were significantly affected by seawater intrusion, reaching concentrations that would limit their use for human supply or even irrigated agriculture. In addition, significant agricultural pollution (nitrates) was detected. Planned management of the aquifer and corrective measures are needed in order to invert the salinization process before it severely affects water resources in the long term.

Academic research paper on topic "Assessing seawater intrusion in an arid coastal aquifer under high anthropogenic influence using major constituents, Sr and B isotopes in groundwater"

ARTICLE IN PRESS

STOTEN-22074; No of Pages 14

Science of the Total Environment xxx (2017) xxx-xxx

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Science of the Total Environment

Assessing seawater intrusion in an arid coastal aquifer under high anthropogenic influence using major constituents, Sr and B isotopes in groundwater

J. Mahlknecht a,*< D. Merchán a, M. Rosner b, A. Meixnerc, R Ledesma-Ruiz a

a Centro del Agua para América Latina y el Caribe, Tecnológico de Monterrey, Mexico b IsoAnalysis UG, Berlin, Germany

c Faculty ofGeosciences and MARUM-Center for Marine Environmental Sciences, University of Bremen, Germany

HIGHLIGHTS

GRAPHICAL ABSTRACT

1 Seawater intrusion has increased salinity in La Paz aquifer. 1 Significant exchange processes affected major cations.

Strontium increased mainly as a consequence of cationic exchange. 1 Boron probably increased as a consequence of desorption processes. 1 One third of the samples showed significant seawater contribution.

ARTICLE INFO

ABSTRACT

Article history:

Received 2 December 2016

Received in revised form 16 February 2017

Accepted 16 February 2017

Available online xxxx

Editor: D. Barcelo

Keywords:

La Paz aquifer

Overexploitation

Salinization

Cationic exchange pool

87Sr/86Sr

The La Paz aquifer system (Baja California Sur, Mexico) is under severe anthropogenic pressure because of high groundwater abstraction for urban supply (city of La Paz, around 222,000 inhabitants) and irrigated agriculture (1900 ha). In consequence, seawater has infiltrated the aquifer, forcing the abandonment of wells with increased salinity. The objective of this study was to assess seawater intrusion, understand the hydrogeochemical processes involved and estimate the contribution of seawater in the wells tested. The aquifer comprises mainly the alluvial filling and marine sediments of a tectonic graben oriented north-south, in contact with the Gulf of California. Groundwater samples were collected in 47 locations and analyzed for major constituents. A subset of 23 samples was analyzed for strontium and boron concentrations and isotopic signatures (87Sr/86Sr and 811B). Results were interpreted using standard hydrochemical plots along with ad hoc plots including isotopic data. Seawater intrusion was confirmed by several hydrogeochemical indicators, such as the high salinity in areas of intense pumping or the Na+-Ca2+ exchange occurring in sediments that were previously in chemical equilibrium with fresh water. However, seawater contribution was not sufficient to explain the observed concentrations and isotopic signatures of Sr and B. According to the isotopic data, desorption processes triggered by a modification in chemical equilibrium and an increase in ionic strength by seawater intrusion significantly increased Sr and probably B concentrations in groundwater. From a calculation of seawater contribution to the wells, it was estimated that one-third of the sampled abstraction wells were significantly affected by seawater intrusion, reaching concentrations that would limit their use for human supply or even irrigated agriculture. In addition, significant agricultural

* Corresponding author. E-mail address: jurgen@itesm.mx (J. Mahlknecht).

http://dx.doi.org/mi 016/j.scitotenv.2017.02.137

0048-9697/© 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.Org/licenses/by/4.0/).

ARTICLE IN PRESS

2 J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

pollution (nitrates) was detected. Planned management of the aquifer and corrective measures are needed in order to invert the salinization process before it severely affects water resources in the long term.

© 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://

creativecommons.org/licenses/by/4.0/).

1. Introduction

Mexico, with 9330 km of coastline, has experienced an important growth in population and agricultural surface in coastal areas over the past 30 years, which has led to the overexploitation of several aquifer systems (SEMARNAT and CONAGUA, 2014). The problems of seawater intrusion are most notable west of Sonora and in the Baja California Peninsula, with minor problems in the Yucatan Peninsula, and in the states of Veracruz, Sinaloa and Nayarit (Cardoso, 1993; Marín, 2002). The coastal aquifer system of La Paz (Baja California Sur) represents the sole water resource of La Paz city and the surrounding towns, an urban area of over 222,000 inhabitants (INEGI, 2010) and around 1900 ha of irrigated agriculture (CONAGUA, 2010a). Following demographic and economic growth, 155 wells have been drilled in the alluvial sediment unit to address an increasing demand for water for urban use, tourism and agriculture over the past decades (CONAGUA, 2010b).

Eventually, a lack of planned water resource management has led to overexploitation of the underlying aquifer with the consequent reversal of the groundwater hydraulic gradient near the coastline and marine intrusion (CONAGUA, 2001, 2010b; Tamez-Meléndez et al., 2016). Moreover, the arid climate and frequent periods of drought adversely affect the water reserves. To meet increasing demand, La Paz city has started searching for new water sources. The local and federal governments have proposed the construction of a desalinization plant which would produce additional 200 L s-1 of freshwater (CONAGUA, 2010b) while controlling seawater intrusion. However, civil organizations have voiced a preference for management of the underlying groundwater resources and/or managed aquifer recharge projects (Pronatura, 2010).

Previous research on this coastal aquifer system has dealt with hydrogeological characterization (CONAGUA, 2001; Cruz-Falcón, 2007; Cruz-Falcón et al., 2010), assessment of marine intrusion using geophysical methods and models of flow and transport (CONAGUA, 2001, 2010a; Escolero and Torres-Onofre, 2007; Monzalvo, 2010), and evaluation of groundwater quality (CONAGUA, 2002, 2010b). One published study (Tamez-Meléndez et al., 2016) evaluated groundwater flow using multivariate statistical analyses, hydrogeochemical indicators, water stable isotope methods (6D and 618O in water molecules) and groundwater dating methods (radiocarbon and CFC's). The study concluded that groundwater in that area was mainly of meteoric origin, with minor regional groundwater flow influence, and that agriculture development had provoked an ever-growing quality problem in groundwaters related to anthropogenic pollutants and seawater intrusion.

In this sense, conventional geochemical methods such as the use of chemical concentrations of major and trace elements or stable isotope ratios may help to elucidate the mixing of end members (e.g., Hernández-Antonio et al., 2015). Nevertheless, the use of these tools can be limited because there are no clear mixing end members, as their signature is altered by surface and near-surface physical, chemical and biological alteration processes, such as evaporation or transpiration (Kendall and Caldwell, 1998).

In this context, the use of strontium and boron isotopes can improve our knowledge of hydrogeochemical processes. The 87Sr/86Sr ratio is not affected by fractionation processes and its value depends on source waters or rocks (e.g., Faure and Mensing, 2005). In consequence, strontium isotopes have been widely used to identify source waters in groundwater studies (e.g., Petelet-Giraud et al., 2003; Bakarietal., 2013; Paces and Wurster, 2014). Similarly, boron stable isotopes signature can be used in groundwater because of its natural occurrence in most waters, and the wide range of expected values possible from potential intense

fractionation processes as a consequence of the high relative mass difference of 10B and 11B (Vengosh and Spivack, 2000; Venturi et al., 2015). Thus, a combination of the classical hydrogeochemical tools and these complementary tools may provide better understanding of the geochemical processes present in aquifers affected by seawater intrusion (e.g., Pennisi et al., 2006; Langman and Ellis, 2010; Meredith et al., 2013; Cary et al., 2015; Petelet-Giraud et al., 2016). The use of multi-isotope studies provides complementary information from different sources, allowing a comprehensive comparison of natural and human-affected processes (e.g., Puig et al., 2016).

Therefore, the present study aimed to further investigate the origin of groundwater salinity in a coastal aquifer affected by seawater intrusion incorporating the use of the radiogenic strontium isotopes ratio (87Sr/86Sr) and stable boron isotopes (611B) for a thoughtful hydrogeochemical evaluation of major constituents. The specific objectives include an assessment of water quality in relation to human uses (mainly human consumption and agriculture), to better understand the controlling geochemical processes and to estimate the seawater contribution to the abstracted water.

2. Study area

2.1. General setting

La Paz aquifer system is located in the southern portion ofthe Peninsula of Baja California, Mexico (Fig. 1A). It is situated within the municipality of La Paz, in the state of Baja California Sur, about 60 km north of the Tropic of Cancer. The aquifer is in direct contact with La Paz Ensenada, which is hydraulically connected to the Sea of Cortez (Gulf of California). The main economic activity in La Paz and its surrounding area is tourism, with a great increase in the number of visitors in recent years, putting the area's water resources under severe pressure.

The climate in the study area is predominantly warm and dry with an average annual temperature and rainfall of 22.5 °C and 263 mm, respectively (CONAGUA, 2010b). Precipitation is not evenly distributed during the year, with almost 75% of precipitation occurring between July and October. Given the high potential evaporation, recharge of this aquifer occurs mainly after rainfalls associated with tropical storms and cyclones at the end of the summer (CONAGUA, 2002).

2.2. Geological background

The geomorphological structure of La Paz valley consists of intermountain plains, table mountains and mountain ranges of up to 900 m above sea level. Structurally, the valley of La Paz corresponds to a graben (Fig. 1B; Hausback and Frizzell, 1984; Cruz-Falcon et al., 2010) in north-south orientation which is limited to the west and east by El Carrizal and La Paz normal faults, respectively (Alvarez et al., 1997; Cruz-Falcon et al., 2010). The presence of structurally controlled wadies and bathymetric data suggest the existence of a series of secondary faults inside the graben (Alvarez et al., 1997; Cruz-Falcon, 2007).

La Paz region is located to the west of the boundary between the Pacific and North American plates, a boundary that extends throughout the Gulf of California. Tectonically, the region forms part of the Gulf Ex-tensional Province (Munguia et al., 2006). The valley of La Paz is the result of distention tectonics which created the Gulf of California, probably during the Pleistocene, and the subsequent accumulation of marine and continental granular sediments in the lower topography (Aranda-Gomez and Perez-Venzor, 1995). The base of the lithological column in the La Paz area is a metamorphic complex which consists of Mesozoic

ARTICLE IN PRESS

J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

Fig. 1. A) Surface lithology of the study site with equipotential lines corresponding to the summer of 2012. B) Hydrostratigraphic cross-section. Adapted fromSGM (1999) and CONAGUA (2001).

ARTICLE IN PRESS

4 J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

shales, phyllites, schists, sillimanite and gneiss (Monzalvo, 2010). This unit is overlain by Cretaceous intrusive rocks (gabbro, granite and granodiorite) in the oriental part of La Paz fault, constituting the Las Cruces and El Novillo mountain ranges (Cruz-Falcón et al., 2010). This group of intrusive rocks is known as Los Cabos block (Fletcher et al., 2000). Towards the western part of the valley, a sequence of outcropping sandstones and conglomerates of the Pleistocene are present. A sequence of Miocene volcanic and sedimentary rocks surrounds the city of La Paz. The filling materials of La Paz valley consist in a Quaternary sequence of marine sediments (alternating lutites and sandstones) and alluvial deposits (conglomerates, sandstones and unconsolidated sediments of sands, silts and clays; Álvarez et al., 1997; Cruz-Falcón et al., 2010, 2013).

to the well owners, in general the screening interval covered the whole well. Privately-installed pumps were used to purge the well (three-well volumes) and collect the samples. The spring sample was collected manually. Field parameters were measured in situ, and recorded after the variables had stabilized. In-situ measurements included electrical conductivity corrected to 25 °C (EC), temperature (T), pH, dissolved oxygen (DO) and alkalinity. A pre-calibrated portable multi-meter (Orion Star A329, Thermo Fisher Scientific, Waltham, MA, USA) was used for EC, T, pH and DO. Alkalinity was determined in the field by titration with 0.02 N H2SO4. Samples were filtered using 0.45 |jm membrane filters, preserved at 4 °C and sent to laboratory for analysis within a week.

3.2. Analytical procedures

23. Hydrogeology

In the valley of La Paz it is possible to identify two aquifers. The first is an intergranular-porosity aquifer (CONAGUA, 1997, 2001, 2002; Cruz-Falcón, 2007; Escolero and Torres-Onofre, 2007). It is an uncon-fined aquifer, but it behaves semiconfined locally where different levels of clays are present. The porous medium is developed in clastic materials mainly of fluvial origin, which are widely distributed in the valley. It presents mainly continental facies, which are made up of sand and gravel sometimes embedded in an argillaceous matrix. These materials reach a maximum thickness of 400 m in the center of the valley (CONAGUA, 2010a). Below the intergranular aquifer, a fractured aquifer system is located within the sequence of igneous and metamorphic rocks, associated with the crystalline complex that forms mainly the eastern part of the valley (CONAGUA, 1997, 2010a). There are no conclusive data regarding the hydraulic parameters of this deep aquifer.

Natural recharge originates mostly from precipitation captured in the mountainous elevations located east and southeast (Las Cruces and El Novillo mountain ranges). Runoff infiltrates when it reaches the flat areas at the bottom of the valley. In addition, rainwater infiltrates through the cracks or fractures of the igneous and weathered metamor-phic materials, providing minor lateral and vertical recharge to the intergranular aquifer (Cruz-Falcón et al., 2011). Recharge also occurs, although to a lesser extent, by direct infiltration of rainfall in the valley (CONAGUA, 2010a). The depth to groundwater level in this aquifer varies from around 5 m near the coastline to > 70 m south of La Paz city (Cruz-Falcón et al., 2010; Monzalvo, 2010), presenting under natural conditions a rather low hydraulic gradient in respect to the sea.

Recharge in the aquifer system has been estimated at 28 million m3 (CONAGUA, 2010b). According to official estimates, current annual extractions average 32 million m3, of which 68% is extracted for urban water supply, 27% for agriculture, 3% for livestock and 2% for industry and other services. However, these data are uncertain because they do not account for illegal extractions. This scenario creates a deficit in long-term water availability. In consequence, over pumping has caused hydraulic gradient inversion (CONAGUA, 2010b), so thatthe groundwater flow occurs mainly in a southwest direction from the coast, allowing seawater to enter the aquifer from the Gulf of California. According to official estimates, in 2004, the wedge intrusion of seawater was located approximately 5.25 km away from the coastline on the east side, and 6.25 km on the western side of the aquifer, with an estimated speed of advance into the valley between 120 and 200 m year-1 (CONAGUA, 2010a). A decrease in the quality of groundwater has been reported by the water authority, and had even led to the closure of some abstraction wells.

3. Methodology

3.1. Sampling procedures

Forty-six production wells and a spring were sampled in August 2013. The well depths varied between 3 and 200 m (Table 1). According

All the chemical analyses were performed in Activation Laboratories Ltd., Ancaster, Ontario, Canada. The concentrations of dissolved cations (Na+, K+, Ca2+, Mg2+), silica (SiO2) and Sr2+ and B were measured by inductive-coupled plasma mass spectrometry (ICP-MS) using a Perkin Elmer SCIEX ELAN 6000 ICP/MS equipment (Waltham, MA, USA). During the trace element determinations, a blank and two water standards were run at the beginning and end of the analyses. Measurement uncertainties for all parameters were evaluated and controlled using regular laboratory duplicates of samples and verifying the precision/calibration of the instruments through regular runs of various primary standard solutions. The international geostandard SRM-1640 (trace elements in natural water - certified by the National Institute of Standards and Technology NIST) was measured during the analyses of groundwater samples to check the accuracy of the ICP-OES and ICP-MS methods. The accuracy of these analyses (defined as the systematic difference between the reference values and the measurements of the geostandard SRM-1640) was lower than 9% for major ions and trace elements. If upper limits of major cations were exceeded, they were measured by inductive-coupled plasma optical emission spectrometry (ICP-OES) using an Agilent Axial ICP Optical Emission Spectrometer. Major anions (CI-, SO|-, NO-) were determined and quantified by ion chromatography in un-acidified samples using a Dionex DX-120 equipment (Sunnyvale, CA, USA). Samples with high Cl- (>75 mg L) and SO4-(>375 mg L) concentrations were diluted in order to avoid oversaturation. The charge balance was cross-checked as follows:

Charge balance(%) = 100

^ meq Cations—^ meq Anions ^ meq Cations + ^ meq Anions

Absolute charge balances averaged 3.7%. One sample (LP17, C.B. = 13%) exceeded the 10% acceptable criteria and was discarded for posterior interpretations.

A subset of 26 samples was selected for isotopic investigations. Strontium isotope ratios (87Sr/86Sr) of water samples were determined by IsoAnalysis UG (Berlin, Germany) using thermal ionization mass spectrometry. Prior to mass spectrometry, strontium was separated from the matrix by ion chromatography with Sr-Spec resin. The raw data were corrected for interfering Rb and mass fractionation (86Sr/88Sr = 0.1194) and normalized to a NIST SRM 98 7 87Sr/86Sr ratio of 0.71025. During the course of the analysis, three aliquots of 1APSO seawater yielded an 87Sr/86Sr ratio of 0.709181 with an external reproducibility of 0.000049 (2 SD). An expanded uncertainty of 0.00007 for 87Sr/86Sr isotope ratios of samples was exemplary calculated for a sea-water standard. Strontium isotopic composition is expressed as the atomic proportion 87Sr/86Sr.

Boron isotopic composition of the groundwater samples was analyzed in the Isotope Geochemistry Laboratory at the MARUM-Center for Marine Environmental Sciences, University of Bremen (Germany). The entire analytical procedure follows the procedure for fluid samples in Kutzschbach et al. (2016). A cation column separation using Bio-Rad AG 50W-X8 resin and mannitol (Bio-Rad Laboratories, Hercules, CA, USA) was performed to isolate boron from matrix elements. Purified

ARTICLE IN PRESS

J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx 5

La Paz aquifer system hydrochemical data collected on August 2013.

Sample Depth T EC pH O2 Na+ K+ Ca2+ Mg2+ Cl- SO2- HCO- NO- Br- Sr2+ 87Sr/86Sr B 811B

Units m °C |jS cm-1 — mg L-1 mmol L-1 |jmol L-1 |jmol L-1 Ratio |jmol L-1 %o

Group 1 LP7 150 31.1 2109 7.1 7.7 3.52 0.13 3.87 2.48 14.70 0.80 1.90 0.50 16.6 20.3 0.705764 8.7 37.1

LP14 60 29.6 3480 7.1 5.6 9.92 0.13 6.56 3.73 26.40 1.42 3.70 0.68 44.7 46.2 0.705888 80.3 54.9

LP15 200 30.6 2868 7.2 6.6 14.22 0.21 3.34 2.09 19.75 1.73 4.01 0.84 23.9 16.5 n.d. 75.0 n.d.

LP16 30 29.9 2552 7.3 7.9 13.01 0.20 2.59 1.68 15.99 1.76 3.94 0.74 23.3 9.9 0.705908 79.4 42.0

LP18 64 31.0 5100 7.0 6.3 17.01 0.23 8.36 7.65 39.49 2.03 5.77 1.99 42.4 33.5 0.706010 80.5 52.0

LP19 64 29.6 5570 6.9 6.4 15.40 0.26 12.48 6.67 47.67 3.02 3.72 0.43 62.7 51.2 n.d. 93.2 n.d.

LP20 60 29.3 3840 7.1 7.2 10.13 0.20 8.66 4.94 31.59 1.54 4.31 0.29 36.8 34.5 n.d. 76.7 n.d.

LP21 130 33.4 2124 7.1 7.8 5.09 0.14 4.97 2.82 15.51 1.11 4.05 0.46 17.5 20.3 n.d. 33.3 n.d.

LP22 15 30.7 7520 7.0 7.5 46.98 0.26 8.88 5.39 63.75 5.10 4.97 0.89 84.9 58.3 0.706114 299.0 46.6

LP23 26 30.0 2603 7.3 8.3 14.92 0.15 5.11 2.50 20.99 1.58 5.11 0.34 23.9 28.3 0.706010 65.0 50.8

LP26 18 29.1 3080 7.2 8.1 17.18 0.10 4.09 2.51 20.54 1.67 7.23 0.80 28.4 22.9 n.d. 56.8 n.d.

LP27 69 28.2 2944 7.2 4.8 19.27 0.08 3.67 2.34 16.64 3.29 4.41 29.21 6.3 23.7 n.d. 99.9 n.d.

LP28 15 29.0 6880 7.0 8.6 33.19 0.23 10.50 7.65 57.26 4.59 6.33 1.21 80.9 76.4 0.705962 219.1 47.6

LP29 20 29.7 3460 7.1 5.6 12.31 0.15 6.51 4.53 26.35 1.43 4.75 0.67 25.5 45.9 n.d. 93.4 n.d.

LP30 22 27.9 5160 6.8 3.9 17.18 0.18 10.53 6.46 44.85 1.49 5.67 0.71 49.6 74.4 0.705723 82.9 41.0

LP31 30 27.1 4770 7.1 5.2 30.14 0.09 4.72 4.53 32.16 2.37 12.88 3.49 29.0 35.2 n.d. 80.3 n.d.

LP32 70 29.0 2589 7.2 5.8 5.83 0.09 5.91 2.30 20.62 0.52 3.25 0.76 37.0 35.5 0.705730 18.2 44.1

LP33 70 29.3 1777 7.4 8.7 5.31 0.08 3.27 1.59 11.23 0.72 3.40 0.45 15.5 23.7 n.d. 27.6 n.d.

LP35 70 31.3 2751 7.2 6.1 4.44 0.08 6.36 3.07 22.37 0.67 2.76 0.58 32.0 29.0 0.705509 21.0 50.3

LP40 28 28.7 8920 7.2 7.1 25.36 0.36 16.42 14.16 83.50 2.53 9.83 0.40 89.6 92.8 n.d. 118.4 n.d.

LP41 50 29.4 6000 7.0 5.7 24.45 0.20 8.53 6.54 51.90 1.83 6.55 0.02 64.8 52.5 0.705913 134.3 38.3

LP43 60 30.3 4250 7.3 7.4 14.14 0.24 7.76 4.32 32.16 3.04 5.27 0.63 36.7 31.0 n.d. 84.8 n.d.

Group 2 LP1 n.a. 28.9 893 6.8 8.3 2.94 0.10 1.27 0.68 4.49 0.39 1.66 0.72 4.8 9.5 0.705692 11.7 41.5

LP2 n.a. 30.1 702 7.1 8.1 2.31 0.10 1.02 0.58 3.02 0.29 1.75 0.63 2.1 7.3 n.d. 8.7 n.d.

LP3 n.a. 29.2 1537 6.9 7.8 4.35 0.14 2.57 1.39 10.86 0.36 1.85 0.35 13.6 16.7 0.705797 11.1 37.2

LP4 76 31.5 1566 7.2 8.6 5.96 0.11 2.06 1.63 9.65 0.46 3.05 0.27 12.6 8.2 0.706034 21.9 33.6

LP8 57 29.9 1142 7.8 8.2 4.21 0.10 1.30 0.95 5.81 0.34 2.10 0.26 7.5 5.2 0.706005 16.8 33.3

LP11 80 25.0 998 7.4 7.9 4.57 0.10 1.03 0.70 4.74 0.38 2.32 0.63 5.4 6.3 0.706171 14.9 36.8

LP12 80 25.0 1150 7.4 6.1 9.70 0.14 0.97 0.43 8.01 0.78 3.25 0.40 10.5 5.1 n.d. 51.6 n.d.

LP13 60 31.8 1155 7.6 7.1 6.13 0.14 1.09 0.42 5.30 0.56 2.50 0.35 6.6 4.8 0.706225 34.2 35.7

LP24 35 31.0 1295 7.5 4.5 8.13 0.07 1.65 0.76 7.45 0.44 4.25 0.28 10.8 10.7 n.d. 29.8 n.d.

LP25 38 32.7 1292 7.4 7.7 6.26 0.11 2.04 0.97 7.50 0.39 4.05 0.15 10.5 13.5 0.706047 22.8 31.4

LP34 75 29.7 1023 7.6 8.4 3.48 0.07 1.60 0.87 5.16 0.44 2.50 0.14 6.3 12.4 n.d. 16.8 n.d.

LP36 35 31.0 1207 7.4 7.1 3.07 0.07 2.27 0.95 6.91 0.30 2.36 0.23 7.6 11.3 n.d. 13.3 n.d.

LP37 60 29.4 1658 7.3 5.9 6.00 0.08 2.64 1.10 10.72 0.69 2.60 0.17 14.1 17.7 0.705638 25.5 43.0

LP38 24 29.5 799 7.8 7.3 2.60 0.05 1.11 0.85 4.63 0.12 1.86 0.11 5.8 6.0 0.705807 10.2 27.8

LP39 n.a. 29.9 683 7.6 7.0 2.54 0.06 0.83 0.63 2.70 0.15 2.18 0.13 3.8 5.3 n.d. 20.6 n.d.

LP42 50 30.2 1182 7.4 7.3 2.58 0.09 1.98 1.63 8.15 0.20 2.06 0.20 11.3 11.8 0.705974 12.8 29.2

Group 3 LP5 201 32.5 642 7.1 7.8 1.77 0.06 1.12 0.65 2.28 0.15 2.15 0.21 3.1 4.5 0.705565 6.5 34.8

LP6 201 32.5 701 7.4 7.7 1.94 0.04 1.17 0.60 3.13 0.11 1.75 0.21 4.6 5.1 0.705397 7.6 33.1

LP9 201 30.5 865 7.4 6.7 3.55 0.07 1.31 0.82 3.33 0.49 2.80 0.10 4.1 7.7 0.705764 13.4 36.0

LP10 156 33.0 561 7.7 7.3 2.92 0.10 0.85 0.47 1.96 0.10 2.54 0.11 2.1 4.2 n.d. 9.5 n.d.

LP44 80 30.2 630 7.2 7.1 2.18 0.04 0.84 0.64 2.53 0.15 1.98 0.58 4.1 4.0 0.705517 9.8 36.9

LP45 50 28.5 610 7.2 7.8 1.60 0.05 0.98 0.89 1.75 0.12 2.75 0.21 2.0 4.3 n.d. 8.8 n.d.

LP46 21 28.1 924 7.1 6.3 3.52 0.06 1.23 1.14 3.02 0.15 3.97 0.17 3.8 6.5 n.d. 17.3 n.d.

LP47 0a 27.4 479 8.3 14.7 1.71 0.11 0.69 0.38 1.54 0.08 1.82 0.13 1.6 3.2 n.d. 8.3 n.d.

Discard

LP17 50 29.5 3320 7.1 6.8 28.62 0.25 4.54 2.62 23.50 1.87 5.39 0.92 24.5 20.3 0.705799 81.7 43.9

n.a.: Not available. n.d.: Not determined. E.C.: Electrical conductivity. a LP47 is a spring, the sample was collected at the surface.

samples and processed standards were dissolved in 2% HNO3 and adjusted to 100 ppb B concentrations. All standards and samples were analyzed repeatedly on a MC-ICP MS Neptune Plus (Thermo Fisher Scientific). The analyses were performed in the standard sample bracketing mode using a 100 ppb SRN NIST 951 reference solution. A 2% HNO3 solution was analyzed before and after each standard and sample. The average of these two measurements represents the analytical baseline and was used for baseline correction. Boron isotopes are expressed in the per-mil notation relatively to untreated boric acid (NIST SRM 951). During the study, the processed SRM NIST 951 showed SnBvalues of-0.04 ± 0.06%» (2 SD, N = 4) and indicates no fraction-ation during the chemical separation. Several reference materials were analyzed in parallel with samples. 6nB of analyzed seawater B1 (39.8 ± 0.1% [2 SD, N = 3]), groundwater B2 (14.7%), groundwater B3 (-21.2 ± 0.4% [2 SD, N = 2]) and the internal laboratory seawater

standard BSW Susu Knolls (39.7 ± 0.4% [2 SD, N = 4]) demonstrate the accuracy of the analytical procedure.

33. Results interpretation

In a previous study using major constituents and water stable isotopes from the same sampling campaign, Tamez-Meléndez et al. (2016) classified the obtained water samples into three groups according to their similitude (cluster aggregates). Group 1 (23 samples, crosses in figures) was sampled closer to the seashore, in areas where higher volumes were abstracted for irrigation and presented the highest salinities in the aquifer (brackish groundwater). Group 2 (16 samples, triangles in figures) corresponds to wells in urban areas and some wells in proximity to the seashore; and the group corresponds to fresh to brackish groundwater. Group 3 (8 samples, circles in figures) corresponds to

ARTICLE IN PRESS

б J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

fresh groundwater and was collected in the southern study area, further from the shore and with no significant anthropogenic influence. We followed this classification in our study. Group 3 was considered as representative of natural groundwater. Mean levels of major ions and trace elements found in samples were within the range typically reported for unpolluted groundwater (Appelo and Postma, 2005) with the exception of Na+ and Cl-, which slightly exceeded the reference value of 2 mmol L-1. However, the location of the study area, between the Pacific Ocean and the Sea of Cortez, justifies a higher concentration of these parameters in relation to marine aerosols deposition. In the presented plots (Results and discussion section), the fresh groundwater composition was used as one end-member. The other end-member was a theoretical mixture of seawater composition (Stumm and Morgan, 1996).

From hydrogeochemical data, the saturation indices (SI) were calculated for selected mineral phases using PHREEQC 3.3.7 (Parkhurst and Appelo, 2013) and its database phreeqc.dat. PHREEQC is a U.S. Geological Survey computer program for simulating chemical reactions and transport processes in natural or polluted water, and allows the calculation of a mineral phase SI, i.e., the thermodynamic feasibility of a given water composition to dissolve or precipitate specific minerals. Although variable with different minerals, in general a SI > 0.2 is considered oversaturation (with a tendency to precipitate), while a SI < — 0.2 indicates undersaturation (with a tendency to dissolve) (Merkel and PlanerFriedrich, 2008). Mineral phases used in the calculations included cal-cite, dolomite, gypsum and halite. In addition, CO2 (g) partial pressures in equilibrium with the analyzed water were computed based on the alkalinity and pH of the samples (Kehew, 2001).

Finally, for each groundwater constituent or computed variable (such as saturation indices), the nonparametric Kruskal-Wallis test (Helsel and Hirsch, 2002) was used to infer whether significant differences existed between Groups 1, 2 and 3. If the null hypothesis (H0 = there is no differences between groups) was rejected, the nonparamet-ric version of Tukey's test (Helsel and Hirsch, 2002) was used to detect differences between specific groups.

4. Results and discussion

4.1. Water quality for urban supply and agriculture

The chemical composition of the collected samples in the La Paz aquifer system is displayed in Table 1. Groundwater presented a wide range of salinity, from fresh water (479 |jS cm-1) to brackish water (8920 |jS cm-1), with a median value of 1658 |jS cm-1. Median EC's were significantly different, with 3470 |jS cm-1, 1153 |jS cm-1 and 636 |jS cm-1 for Groups 1, 2 and 3, respectively. pH ranged from 6.8 to 8.3, although this last value belonged to the spring water collected (LP47). Maximum groundwater pH was 7.8, being neutral to slightly basic. Group 1 had a significantly lower pH than Groups 2 and 3. DO ranged from 3.9 to 14.7 mg L-1, but did not differ significantly between groups. Again, the highest value belonged to spring water, being the maximum value for groundwater of 8.7 mg L-1. The levels of oxygen in the aquifer suggest an aerated system.

Regarding water quality for urban use, 64% and 40% of the samples did not comply with the Mexican drinking water standard (NOM-127-SSA1-1994) for Cl- and Na+ concentrations, respectively, while in the case of NO- (23%) and SO4- (4%) the majority of the samples fell within the legal range. The water quality for irrigation was generally low. Around 87% of the samples had EC values higher than 700 |jS cm-1, presenting thus a slight restriction for irrigation purposes (Ayers and Westcot, 1992). In fact, 28% of samples (all from Group 1) presented severe restrictions for irrigation purposes (EC > 3000 |jS cm-1). Sodium adsorption ratio (SAR) values ranged from 1.17 to 12.43, and were significantly higher in Group 1 (average of 4.91). This indicates a low to medium sodium hazard (e.g., Fetter, 2001) in soils irrigated with these waters. Thus, despite the high salinity, groundwater abstracted from the La Paz aquifer system does not present a high sodium hazard. The

geochemical explanation for this finding is presented in the following sections.

4.2. Hydrogeochemistry and controlling processes

Regarding major cations and anions in solution, an evolution in the hydrochemical facies from low salinity Ca2+-HCO- type to Na+-Cl-with increasing salinity was observed (Fig. 2). In general, samples with low salinity were located further from the coast, whereas samples with higher salinity were located near the coast or in those areas highly affected by the hydraulic gradient inversion (Fig. 1; Tamez-Melendez et al., 2016).

Chloride can be considered a conservative tracer of water saliniza-tion, since it is not significantly affected by sorption processes of biological transformations (e.g., Flury and Prapitz, 1993). In addition, chloride concentrations in seawater (ca. 560 mmol L-1, Stumm and Morgan, 1996) are significantly higher than those expected in fresh unpolluted groundwater (<3 mmol L-1). In this sense, the chloride concentration in the La Paz aquifer system may be related to either chloride influx with recharge water (marine aerosols rich in chloride due to sea proximity), the dissolution of chlorine-bearing mineral in the aquifer matrix, or some degree of mixing with seawater. Although no chlorine-bearing minerals were reported in the reservoir, the prevailing arid climate conditions in the study area make it possible to find some chloride salts in the unsaturated zone.

The molar ratio between Na+ and Cl- was close to 1 in low salinity samples, and generally lower in salinity-enriched samples (Fig. 3A). For most of the samples with moderate to high salinity, the Cl- vs. Na+ crosses clearly plot below a theoretical seawater mixing line, suggesting that some Na+-removing processes take place in the aquifer matrix. Cation exchange between Na+ and Ca2+ (Na+ leaves the solution, through sorption in the mineral phase) has been reported in other aquifers affected by seawater intrusion (Appelo and Postma, 2005). The quantity of Ca2+ in samples with increased salinity also indicates a supplementary source of this alkaline earth element (Fig. 3B). Potassium presented a similar pattern to that of sodium (Fig. 3C) whereas magnesium evolution with higher salinity was closer to the theoretical mixing line behavior (Fig. 3D), i.e., only very salty samples presented more magnesium than would be expected from seawater contribution.

To confirm the importance of cation exchange processes, the difference in concentration Дn between the observed values and those

100 Ca2+ 0 0 ci" 100

Cations Anions

Fig. 2. Piper diagram showing the hydrochemical facies of groundwater in the La Paz aquifer system.

ARTICLE IN PRESS

J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

Fig. 3. Major cations in groundwater of the La Paz aquifer system and their relationship with chloride concentration (Cl ). Sodium (A), calcium (B), potassium (C) and magnesium (D). A theoretical mixing line between fresh groundwater and seawater is presented with 2% increases in seawater proportion with each diamond.

expected if mixing with seawater was the only process was obtained through the following mixing equation (Pennisi et al., 2006):

AHx = %(s)

nX(fresh) + (nCl(s) nCl(fresh))

nX(sea)

nCl(sea)

where nX(s), nX(ugw), nX(sea) refer to the concentration of component X in the sample (s), fresh groundwater (fresh), and seawater (sea). As seen, the contribution of seawater is derived from the conservative ion Cl-(Appelo and Postma, 2005).

Fig. 4 represents the relationship between sodium and calcium differences with a theoretical conservative mixture with seawater. Most observed values fell within the minus one slope line, that is, cationic exchange represented in the following equation:

^CaX2 + Na+^NaX + iCa2+

Regarding major anions, sulfate concentrations displayed a wide range of variation, from 0.08 to 5.10 mmol L-1. There were significant differences between Group 1 and the other two groups (p < 0.01; Fig. 5A). Deviations from the theoretical seawater mixing line can be justified either by an extra supply of sulfate or by sulfur-removing processes. Removing sulfate processes (sulfur reduction) have been reported in other coastal aquifers (e.g., Fidelibus, 2003). However, in the La Paz aquifer system DO concentrations observed suggest that these processes were oflittle importance. In contrast, additional sources ofsulfate are possible. Gypsum dissolution was rejected since there is no evidence of gypsum in the aquifer matrix and the sulfate vs. calcium plot (not shown) indicated an evolution that could not be easily explained by the dissolution of this mineral. However, for most of the collected samples, NO3- presented values clearly over those expected if seawater

intrusion were the only process in action (Fig. 5B). Given the agricultural activities carried out in the lower parts of La Paz valley, nitrogen leaching from agricultural soils would explain the high NO3- concentrations observed. In fact, a subtle relationship was detected between NO3-and SO4-, once removed the amount of sulfate provided in the theoretical mixing with seawater (Fig. 6). A strong correlation between NO-and SO4- is typically detected in agricultural areas with no other significant salinization processes in action (Nakano et al., 2008; Menció et al., 2016) because of the application of fertilizers. In this case, the variability in fertilization management along with salinization due to seawater intrusion justifies the lack of a stronger relationship.

Fig. 4. Relationship between calcium and sodium differences (as calculated by Eq. (2)) from a conservative mixture with seawater.

ARTICLE IN PRESS

J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

A X /

X, X X

Cl- [mmol L-1]

O* h X X

^ 0 X O O 2 0

x* X x X X X —e-^ -X- X -©- X 9-O-

Cl- [mmol L-1]

x Group 1 (brackish)

A Group 2 (fresh/brackish) o Group 3 (fresh)

/X /XX x X X

rV fx X

ci- [mmol L1]

Cl- [mmol L1]

Fig. 5. Anions in the La Paz aquifer system groundwater and their relationship with concentrations of chloride (Cl ): sulfate (A), nitrate (B), bicarbonate (C) and bromide (D). A theoretical mixing line between fresh groundwater and seawater is presented with 2% increases in seawater proportion with each diamond.

Bicarbonate concentrations ranged from 1.7 to 12.9 mmol L-1, being significantly higher in Group 1 of samples (p < 0.01; Fig. 5C). Most of the high-salinty samples plotted over the seawater mixing line, suggesting that some geochemical processes provided HCO-. These processes were related to the lower pH of samples from Group 1, and are discussed in the following section.

Bromide concentrations in the La Paz aquifer system ranged from 0.002 to 0.09 mmol L-1, and correlated significantly with chloride, plotting over the theoretical seawater mixing line (Fig. 5D). Br- concentration in seawater averages 0.8 mmol L-1, approximately 200 times the

x Group 1 (brackish) A Group 2 (fresh/brackish) o Group 3 (fresh)

X x «*

-2-10123 AS042- [mmol L1]

Fig. 6. Relationship between excess sulfate and nitrate (as calculated by Eq. (2)) over the theoretical seawater mixing line.

top values reported for natural fresh waters (below 0.004 mmol L-1; Flury and Prapitz, 1993), although higher values have been reported in coastal areas. In addition, Br- is present in many pesticides and fertilizers, and consequently its leachate can reach groundwater (Flury and Prapitz, 1993). Despite this feasible origin, the strong correlation observed between Br- and Cl- concentrations (Fig. 5D) and the fact that they plotted in the theoretical mixing line between fresh groundwater and seawater suggest that the increase of Br- in the aquifer system is mainly due to seawater intrusion.

Thus, the hydrochemical exploration of the data suggests that seawater has partially displaced fresh water in the influence area of abstraction wells located near the coast. The variability in well depth, abstraction rates and hydrogeological context, along with the existence of other hydrochemical processes, makes it difficult to obtain an overall quantification of the process. Coupled with the seawater intrusion, a significant agricultural pollution is inferred. In addition, the overexploitation of the La Paz aquifer system for irrigation purposes over the past decades may have increased groundwater salinity as a consequence of recirculation of irrigation water, i.e., irrigation returns flows enriched in salts from evapo-concentrates and salts stored in the soil reaching groundwater. Evapo-concentration was detected in the La Paz aquifer system by means of environmental isotopes in a previous work (Tamez-Melendez et al., 2016).

4.3. Speciation-solubility calculations

Speciation-solubility calculation provided the saturation indices (SI) of several mineral phases (calcite, dolomite, gypsum and halite) and the CO2 (g) partial pressure in the water samples from the La Paz aquifer system. Calcite (Fig. 7A) and dolomite (not shown) presented a rather

ARTICLE IN PRESS

J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

similar behavior. For low-salinity samples (below 20 mmol L-1 of Cl-), subsaturation and equilibrium (SI below or around zero) with respect to calcite and dolomite dominated. Thus, low-salinity water was in equilibrium with carbonate minerals present in the aquifer matrix, such as the calcareous cement of conglomerates and sandstones. With increasing salinity, the degree of oversaturation increased, and all samples were above 20 mmol L-1 of chloride in the upper fringe of equilibrium or oversaturated with respect to calcite. This finding may be explained by the increase in Ca2+ due to cation exchange, as outlined in the previous section. In consequence, carbonate mineral precipitation is likely in the La Paz aquifer system, especially in zones near the seashore.

La Paz groundwater was undersaturated with respect to gypsum (Fig. 7B) and halite (not shown). SI values of both mineral phases increase with increasing salinity. Although always clearly undersaturated, gypsum SI reaches values relatively close to the equilibrium with gypsum (SI around zero), due to the increase in both SO|-(seawater mixing and marine aerosols) andCa2+ (released from cation-ic exchange pool). Gypsum dissolution was discarded in the previous section. In contrast, the groundwater was clearly undersaturated with respect to halite, even in the most saline samples (SI halitemax = -4.3).

Finally, partial pressure of CO2 (g) in groundwater ranged from 10-272 to 10-123, although it was 10-323 for the spring sampled, and much closer to atmospheric value of 10-34 (Fig. 7c). CO2 (g) content increased with increasing salinity, supporting the calcite precipitation hypothesis, since CO2 (g) is one of the products of the calcite precipitation chemical reaction (Kehew, 2001; Appelo and Postma, 2005). This increase in dissolved CO2 is the main reason for the lower pH and higher bicarbonate concentrations in samples from Group 1, as discussed in Sections 4.1 and 4.2, respectively.

4.4. Strontium concentration and isotopic signature

Strontium is known to have a hydrogeochemical behavior similar to that of calcium (McNutt, 2000). It occurs in nature as positively charged solutes, and is therefore affected by cationic exchange processes (Bullen and Kendal, 1998; Fidelibus, 2003; Faure and Mensing, 2005). In addition, the 87Sr and 86Sr isotope ratio is not affected by fractionation processes, because of their minor difference in atomic weight (Ingraham et al., 1998). In consequence, 87Sr/86Sr can be used as a tracer of the source of dissolved strontium.

Strontium concentrations in the La Paz aquifer system ranged from 3.2 to 92.8 |jmol L-1, with a median value of 16.5 ^mol L-1. In general, the higher the salinity in a sample, the higher the Sr2+ concentration, as depicted in the significant linear relationship (p < 0.01) between chloride and strontium concentrations (Fig. 8A). The measured 87Sr/86Sr ratios of the 26 available samples of groundwater were within a relatively narrow range between 0.7054 and 0.7062 (median 0.7059, Fig. 8B), with no significant differences between groups.

The obtained 87Sr/86Sr values in this study were congruent with those reported by Rosales-Ramírez (2012), who measured strontium concentrations and isotopic ratios in 19 samples collected in the La Paz aquifer system in February 2011. Sr2+ concentrations between 3.9 and 146.1 |amol L-1, with a median value of 27.7 ^mol L-1, and 87Sr/86Sr ratios between 0.7055 and 0.7066, with a median value of 0.7060, were reported in that study. These values do not differ significantly from those obtained in our study, and therefore suggest similar controlling processes.

The strontium isotope ratios for different rocks in the study area (Schaaf et al., 2000) were mostly below the range of the ratios in water samples (0.7036-0.7039, Table 2). Only intrusive granite reached values closer to those obtained in water samples (0.7050). Rosales-Ramírez (2012) reported an 87Sr/86Sr value of 0.7054 for the Comondú Formation calcareous sinter, the formation constituting the lower member of the intergranular aquifer of La Paz system. In contrast, modern seawater contains 92.4 ^mol L-1 of Sr2 + and 87Sr/86Sr ratios of 0.70918 ± 0.00001 (2 SD) (Stumm and Morgan, 1996; Faure and Mensing, 2005), significantly higher than those found in the system (Fig. 8B). As a result, groundwater from the La Paz aquifer system with high Sr2+ concentrations did not show significant influence of sea-water in its 87Sr/86Sr values. In comparison, a theoretical mixture of fresh groundwater (mean of Group 3) and seawater is presented in Fig. 8B following the formulation given in Faure and Mensing (2005).

Given the exposed information and the patterns observed in concentration and isotopic data, the strontium in solution in the aquifer appears to originate mainly from continental sources, i.e., the weathering of igneous rocks and the dissolution of strontium present in the intergranular aquifer. However, there was no detectable trend in the strontium isotopic ratio when compared with the strontium concentration, and all samples presented similar isotopic composition (Fig. 8B). This suggests that the sources of Sr2+ remain relatively stable. In contrast, the observed increase in Sr2+ concentration was significantly higher than that would be expected if seawater were the only input (Fig. 8A), indicating additional sources of strontium.

According to Bullen and Kendal (1998), an equilibrium between the dissolved strontium and the strontium present in the exchange pool may be achieved given enough time. Under this scenario, the strontium available in the exchange pool has the same isotopic composition as that in the dissolved phase due to negligible fractionation processes in strontium geochemistry. A hypothetical cation exchange pool in equilibrium with fresh groundwater from the La Paz aquifer system (Group 3), when in contact with seawater, would retain Na+ and K+ and release Ca2+, Mg2+ and Sr2+ into the solution (Table 3). This pattern is clearly identified in this study (Section 4.2; Figs. 3 and 8A) and would explain the increase in Sr2+ concentration and its non-variable isotopic signatures with increasing concentrations (Fig. 8B).

Therefore, intrusion of seawater produces a shift in the chemical composition of the exchange pool, including the release of significant

Fig. 7. Saturation indices (SI) of calcite (A) and gypsum (B), and partial pressure of CO2 (C) against salinity (as depicted by chloride concentration). The shaded areas indicate the ranges of uncertainty for the calculated saturation indices; ±0.22 for gypsum (Langmuir and Melchior, 1985), and ±0.35 for calcite (Plummer et al., 1990).

ARTICLE IN PRESS

J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

Fig. 8. A) Relationship between chloride (Cl-) and strontium (Sr2+) concentrations in the La Paz aquifer system. B) Relationship between strontium concentration and s7Sr/s6Sr. The rectangular shaded area represents measured whole rock 87Sr/86Sr in the study area (Schaaf et al., 2000; Rosales-Ramírez, 2012). A theoretical mixing line between fresh groundwater and seawater is presented with 2% increases in seawater proportion with each diamond.

quantities of Sr2+. According to the data, the increase in Sr2+ concentration is mainly due to these exchange reactions, and is clearly higher than the increase provided by seawater (Fig. 8A). This finding was supported by the good relationship found between differences in calcium and strontium from the theoretical mixture with seawater (Fig. 9). Similar patterns in exchangeable strontium and its relatively constant isotopic composition with increasing concentrations have been previously reported for a coastal aquifer in Tuscany, Italy (Pennisi et al., 2006).

An alternative explanation to the Sr2+ observed values is related to the leaching from irrigated soils. Some fertilizers can have significant Sr concentrations, ranging from < 1 to approximately 4500 ppm (Otero et al., 2005). Part of this Sr can reach the saturated zone when applied irrigation water exceeds the soil retention capacity. The 87Sr/86Sr of fertilizers varies widely, from 0.702 to 0.835 (Vitoria et al., 2004). However, given the unknown specific fertilization practices in the study area, it is difficult to further assess the contribution of agricultural soils to groundwater Sr2+.

4.5. Boron concentration and isotopic signature

The Mexican regulations for water quality do not establish a limit value for boron. The World Health Organization recommends a limit of 2.4 mg L-1 (222 |amol L-1) as a guideline, but recognizes that in most sites, natural levels are below 0.5 mg L-1 (WHO, 2011). In contrast, the range between boron deficiency and boron toxicity is very narrow, so boron applications can be extremely toxic to some agricultural

Table 2

Strontium isotopic compositions (87Sr/86Sr) and concentrations of strontium (Sr) and rubidium (Rb) in several rocks sampled in the study area. (Schaaf et al., 2000; Rosales-Ramírez, 2012).

Sample Rock Sr 87Sr/86Sr Reference _[ppm] [-]a_

Comondú formation

CS Calcareous sinter 862 0.705420 ± 08 Rosales-Ramírez, 2012

Deformed granitoids

LCB-1 Granite 750.3 0.705029 ± 30 Schaaf et al., 2000

Undeformed granitoids east of La Paz

LCB-3 Tonalite 450.7 0.703949 ± 31 Schaaf et al., 2000

Mafic to ultramafic intrusives from the El Novillo complex

LCB-5 Qz-diorite 364.6 0.703777 ± 31 Schaaf et al., 2000

LCB-6 Qz-diorite 523.0 0.703603 ± 30 Schaaf et al., 2000

LCB-7 Tonalite 400.1 0.703943 ± 31 Schaaf et al., 2000

LCB-8A Gabbro-norite 271.1 0.703555 ± 38 Schaaf et al., 2000

LCB-8B Hornblendite 232.4 0.703620 ± 41 Schaaf et al., 2000

a Deviation ( ± ) refers to last two significant digits.

crops at concentrations only slightly above optimum levels for others (Gupta et al., 1985). Contaminated irrigation water is one of the main causes of boron toxicity in plants, as the concentration build-up of boron in soils, especially in arid regions with high evapo-transpiration rates (Gupta et al., 1985). As a sample, different crops tolerance to boron toxicity along with other effects of boron (human health, livestock or ecosystems) are available in Moss and Nagpal (2003). Thus, the processes leading to an increase in boron concentration in ground-water are of paramount importance in La Paz aquifer system, given the agricultural use of the abstracted groundwater.

Boron in natural waters is normally present either as boric acid, the B(OH)3 planar trigonal complex or as borate, the B(OH)- tetrahedral complex. The proportion of each species depends on pH, concentration, temperature, salinity, the presence of other cations, and the ionic strength (Vengosh and Spivack, 2000; Hemming and Hönisch, 2007). In most situations, B(OH)3 is enriched in the heavy isotope 11B with respect to B(OH)-, with differences in their 811B value between 19 and 23%o (Vengosh and Spivack, 2000; Faure and Mensing, 2005). In contrast, the tetrahedral B(OH)- is preferentially selected in sorption processes, especially in clay mineral surfaces. In consequence, adsorbed boron is usually enriched in the light isotope, while residual boron in the aqueous phase is enriched in the heavy one.

In the collected samples, boron concentrations ranged from 6.5 to 299.1 |jmol L-1, with a median value of 25.5 ^mol L-1 (Fig. 10A). Only one sample in this study was over the WHO recommendation for boron levels in drinking water. The highest concentrations were found near the coastline, and the lowest in remote recharge areas. In general, the more salinity in a sample, the higher the boron concentration. This finding is supported by the significant correlation between Cl- and B concentrations (p < 0.01), although the correlation coefficient was lower than that obtained for other correlations (e.g., Cl--Ca2+ or Cl--Sr2+). For the samples with isotopic analysis, the 811B varied from + 27.8%» to + 54.8%» (median + 38.3%») (Fig. 10B). Boron concentrations and isotopic signatures in Group 1 of the samples were significantly higher than in the other groups (p < 0.01).

According to Vengosh and Spivack (2000), there are several potential sources of boron in an aquifer system, and each presents a range of isotopic signatures. Those possible in this study case were: (1) boron provided by the dissolution of soluble phases in igneous rocks (811B between - 3 and +3%); (2) boron supplied by marine aerosols in the recharge water (up to 30%); (3) seawater boron introduced with seawater intrusion (427 ^mol L-1 with a stable 811B of 39.5%); (4) boron present in brines ( + 39% and higher); and (5) boron from domestic wastewater or synthetic borate fertilizers (ca. 90 |jmol L-1 and approximately -15 to +10%). In addition, sorption processes can significantly modulate both concentrations and iso-topic signature (e.g., Pennisi et al., 2006).

ARTICLE IN PRESS

J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

Molar concentration (M), relative molar proportion of cations in water (F) and exchanger pool in equilibrium (X, as computed using PHREEQC) of selected cations in fresh groundwater (GW) of La Paz aquifer system and seawater (SW); expected behavior in relation to cation exchange pool in the aquifer.

Cation Mgw Fgw XGW Msw Fsw XSW XGW/XSW Exchanger behavior

mmol L-1 % % mmol L-1 % %

Na+ 2.398 57.08 4.86 480.18 86.35 70.54 0.07 Highly retained

K+ 0.066 1.57 0.71 10.46 1.88 7.68 0.09 Retained

Ca2+ 1.022 24.33 65.64 10.54 1.90 5.29 12.41 Highly released

Mg2+ 0.700 16.66 28.38 54.41 9.78 16.43 1.73 Released

Sr2+ 0.005 0.12 0.41 0.09 0.02 0.06 6.83 Released

Ba 0.010 0.24 n.a. 0.43 0.08 n.a. ca. 3a Releaseda

Total 4.201 100 100 556.11 100 100

n.a.: Not available (no equilibrium constant in PHREEQC database). a Boron in most natural waters is in boric acid (H3BO3) or borate (H4BO-) form. The behavior is estimated from actual water concentrations instead of proportions in the exchange pool.

In samples with low B concentration (Group 3), marine aerosols contributing to recharge can explain the isotopic signature. Igneous rocks were discarded as a potential source because they have 611B values close to zero. Some samples contained concentrations and isoto-pic signatures explained by a minor fraction of wastewater (Group 2, Fig. 10B). These samples were collected in urban areas, and therefore a contribution of wastewater in their chemistry was most likely. In contrast, in high-concentration samples (Group 1), an influence of seawater can be inferred. Nevertheless, the observed boron concentration values were significantly higher than those expected from mixing with seawater (Fig. 10A).

A desorption process from the aquifer matrix may explain the observed results. Desorption of boron from sediments following seawater intrusion has been reported in other coastal aquifers (Bianchini et al., 2005; Pennisi et al., 2006), and was explained by the increase in water ionic strength along with a shift in the chemical equilibrium between aqueous phase and sediment. According to Goldberg et al. (2000), soils with a high cationic exchange capacity exhibit a higher B adsorption capacity. In the La Paz aquifer system, a high cationic exchange capacity is expected from the lithology and supported by the observed cationic exchange patterns for Na+, Ca2+, K+, Mg2+ and Sr2+. However, such a process is expected to produce a decrease in 6nB in the water phase. In the La Paz aquifer system, this pattern was not observed, since the 611B of group 1 samples was higher than both fresh groundwater and seawater (Fig. 10a).

An alternative hypothesis that may explain the observed isotopic signature in this group is related to the precipitation of calcite covered in previous sections. Group 1 samples presented a saturation index with respect to calcite which was significantly higher than that seen in the other groups (Fig. 11), along with higher levels of dissolved CO2 (Fig. 7C), this indicates that some degree of calcite precipitation occurred in these water samples. Carbonate precipitation tends to incorporate the tetrahedral complex B(OH)- to the crystalline structure of

X / x Group 1 (brackish) »Group 2 (fresh/brackish)

x y X 0 Group 3 (fresh)

X X y = 2.50x-1.66

,lfx R2 = 0.8

ACa2t [mmol L1]

calcite. The precipitated carbonates have 811B values ca. 17% lower than those present in solution (Hemming and Hönisch, 2007). In consequence, the preferential removal of this isotope will produce an increase in the 811B in the remaining soluble fraction. However, this hypothesis does not explain the great increase in boron concentration observed in the La Paz aquifer system.

As reported for Sr, fertilizers may also be a significant source of B. Otero et al. (2005) reported concentrations on the order of 100 ppm of B in fertilizers commonly used in Spain. The B isotopic signature of fertilizers depends on the fertilizer origin and varies widely (811B between — 10 and +40%, Tirez et al., 2010). Nevertheless, most of the 811B values reported in Tirez et al. (2010) were below those observed in the La Paz aquifer system groundwater, with the only exception being B from hog manure. Hog's farming is not a common practice in La Paz valley, and observed 611B values were not likely to be related to fertilization practices.

Despite the different sources of B in groundwater discussed above, none of the presented hypotheses for boron dynamics completely explains the obtained data. A combination of these processes, along with others still unidentified probably plays a significant role in boron hydrogeochemistry. In addition, groundwater in the La Paz aquifer system presented 87Sr/86Sr signatures of crustal rocks while its 611B is derived mostly from seawater. Therefore, boron isotopic dynamics were not completely understood in our study, because of the lack of a clear isotopic signature of different sources and the variety of processes affecting it. Further studies are required to clarify this point.

4.6. Seawater intrusion estimation

Regarding human uses, mixing only 2% seawater in a fresh-water aquifer exceeds organoleptic objectives for the upper limit of chloride (water begins to taste salty). If the mix exceeds 4%, the water becomes mostly unusable, and if the mix exceeds 6%, the water can only be used for cooling and flushing purposes. In this context, an estimation of average seawater contribution to collected samples was carried out using Cl- and Br-, two halides known for their conservative behavior in groundwater (Flury and Prapitz, 1993). Assuming that there were no contributions other than fresh groundwater (Group 3) or seawater, and that both anions behave in a conservative way, the seawater-contributing fraction to a particular sample can be computed (Appelo and Postma, 2005):

msample mfresh msea — mfresh

Fig. 9. Correlation between differences in calcium and strontium with a conservative mixture with seawater (as calculated by Eq. (2)).

where m is the molar concentration of the anion tested (Cl- or Br-) in the sample, freshwater (f) or seawater (s). However, some increase in groundwater salinity in the aquifer is expected as a consequence of the recirculation of groundwater (pumping for irrigation purposes and infiltration of irrigation return flows). For example, groundwater salinity increased by a factor of three in a heavily pumped aquifer in southern Portugal following the infiltration of irrigation return flows (Stigter

ARTICLE IN PRESS

J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

350 300 — 250

1150 100 50 0

X x> X w ■

bvX ^SeaVa^ i

20 40 60 CI" [mmol L-1]

80 100

* Sorption-Desorption B

0cfo0OOft00OOOOOO0000000000<?b00000

Sea water mixing line

x Group 1 (brackish) A Group 2 (fresh/brackish) o Group 3 (fresh)

Waste water

0 50 100 150 200 250 300 350 B [nmol L"1]

Fig. 10. A) Relationship between chloride (Cl-) and boron (B) concentrations. B) Relationship between boron concentration and its isotopic composition (8nB) in the La Paz aquifer system. A theoretical mixing line between fresh groundwater and seawater is presented with 2% increases in seawater proportion with each diamond. A mixing line for urban wastewater is also presented. Grey arrows indicate the expected shift after sorption processes (lower concentration, heavier isotopic composition; Vengosh and Spivack, 2000).

et al., 2006). The authors of that study suggested that concentrations may rise up to five times their natural values should the heavy abstraction regimen continue. Thus, in our estimation we consider the freshwater salinity as five times that observed in fresh water from the La Paz aquifer system (Group 3), providing a margin of confidence to state whether a well is affected by seawater intrusion.

Results obtained for both Cl- and Br- anions were in reasonable agreement given the uncertainties of this estimation, with a slope close enough to 1.0 and a regression coefficient of 0.94, as depicted in Fig. 12A.The mean of both estimates is used as a measure of seawater contribution. According to this estimate, 65% of the studied wells (30 samples) presented seawater contributions below 2%, while 35% surpassed this threshold (Fig. 12B). It is notable that 2% of seawater in freshwater implies around 14 mmol L-1 of Cl-, two times higher than the Mexican guideline value for drinking water (250 mg L-1 or 7 mmol L-1). Seven samples contained at least 6% seawater, being severely affected not only for drinking purposes but also for irrigation of most crops (Ayers and Westcot, 1992). These seawater contributions are expected to increase since measures against groundwater overexploitation are lacking in the study area.

5. Conclusion

Groundwater abstraction rates in the La Paz aquifer system over the past decades have been higher than natural recharge values, producing an inversion in the hydraulic gradient and therefore an advancement of

X < X

O £ A X « X

0 A A A

x Group 1 (brackish)

A Group 2 (fresh/brackish)

o Group 3 (fresh)

-0.5 0.0 CalciteSI

Fig. 11. Relationship between calcite saturation index and S"B of groundwater in the La Paz aquifer system. The shaded areas indicate the ranges of uncertainty for the calculated saturation index, i.e., ±0.35 for calcite (Plummer et al., 1990).

the seawater intrusion front. Seawater intrusion has produced an increase in groundwater salinity and a shift in its hydrochemical facies, severely affecting water quality for urban supply and/or irrigation purposes in several wells. Although maximum seawater influences were estimated as over 6% in only seven sampled wells, approximately one-third of studied wells showed over 2% of seawater contribution, enough to limit their use for human supplies. If corrective measures are not taken, this proportion is expected to increase in most of the wells, due to continued groundwater abstraction.

The coupled use of major ion chemistry and both Sr and B isotopic signatures allowed the identification of the geochemical process taking place in the La Paz aquifer. An important cationic exchange process is occurring in the reservoir, with seawater sodium replacing continental calcium (and strontium) in the aquifer clay surfaces. Although with a greater degree of uncertainty, we hypothesize that boron is also being significantly released from clay surfaces to aqueous phase, an issue of relevance given its phytotoxicity. This fact has been rarely observed in other study areas. Moreover, cation exchange processes may impose long-term water quality problem due to reverse exchange in freshening waters (waters enriched in sodium with its related soil-structure problems), unless current abstraction rates would decrease.

In contrast, a considerable agricultural pollution is developing in the aquifer, especially in the lower areas of the valley. Irrigation return flows have resulted in increasing nitrate concentrations in the aquifer, and probably sulfate and other pollutants not measured in this study (such as pesticides), which can further threaten the reliance on water resources in an area were groundwater is the solely supply source for human and agriculture uses.

Since seawater intrusion is a severe problem in coastal aquifers, the adequate assessment of its importance is of capital interest for water resources management. Relatively simple measures, such as electrical conductivity monitoring in pumping wells can serve as a proxy of sea-water intrusion. Nevertheless, given the difficulty of planning and law enforcement in groundwater politics, extreme actions must be taken by the water authorities in coastal areas to find a suitable solution to this environmental crisis.

Acknowledgements

The study was co-funded by Fundación FEMSA and the Water Science and Technology Chair of Tecnológico de Monterrey. Fundación FEMSA had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. J. Mahlknecht and R. Ledesma-Ruiz were supported by funding from European Commission project WATERCLIMA-LAC Coastal Water Management, reference: EuropeAid/135857/DH/ACT/MULTI (RAL&RCA); A. Meixner was

ARTICLE IN PRESS

J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

Fig. 12. A) Agreement in the proportion of seawater estimation with Cl and Br tracers. B) Histogram of seawater proportion in the sampled production wells in the La Paz aquifer system.

supported by funding from the Deutsche Forschungsgemeinschaft (DFG Major Research Instrumentation Programme INST 144/308). The authors acknowledge the contribution of C. Támez-Meléndez, A. Hernández-Antonio and C. Gaona-Zanella, who participated in the sampling campaigns and in the analysis of a previous related paper. The thoughtful review of the manuscript by A. Mora is greatly appreciated. The review by four anonymous experts is appreciated.

References

Álvarez, A.A., Rojas, S.H., Prieto, M.J., 1997. Geología de la bahía de la Paz y áreas adyacentes. In: Ramírez, R.M., Urbán, R.J. (Eds.), La Bahía de La Paz: Investigación y Conservación. UABCS-CICIMAR-SCRIPPS, Mexico, pp. 13-29. Appelo, C.A.J., Postma, D., 2005. Geochemistry, Groundwater and Pollution. second ed.

CRC Press, Amsterdam, The Netherlands (649 pp.). Aranda-Gómez, J.J., Pérez-Venzor, J.A., 1995. Reconocimiento geológico en el límite occidental del bloque de Los Cabos, Baja California Sur, La Paz. Tercera Reunión internacional sobre la geología de la península de Baja California. Resumen. Sociedad Geológica Peninsular, Universidad Autónoma de Baja California Sur, pp. 13-14. Ayers, R.S., Westcot, D.W., 1992. Water quality for agriculture. FAO Irrigation and Drainage Paper 29. Food and Agriculture Organization of the United Nations, Rome, Italy. Bakari, S.S., Aagaard, P., Vogt, R.D., Ruden, F., Johansen, I., Vuai, S.A., 2013. Strontium isotopes as tracers for quantifying mixing of groundwater in the alluvial plain of a coastal watershed, south-eastern Tanzania. J. Geochem. Explor. 130:1-14. http://dx.doi. org/10.1016/j.gexplo.2013.02.008. Bianchini, G., Pennisi, M., Cioni, R., Muti, A., Cerbai, N., Kloppmann, W., 2005. Hydrochemistry of the high boron groundwaters of the Cornia aquifer (Tuscany, Italy). Geothermics 34:297-319. http://dx.doi.org/10.1016/j.geothermics.2005.04. 002.

Bullen, T.D., Kendal, C., 1998. Tracing of weathering reactions and water flow paths: a multi-isotopic approach. In: Kendall, C., McDonnell, J.J. (Eds.), Isotope Tracers in Catchment Hydrology. Elsevier BV, Amsterdam (The Netherlands) (839 pp.). Cardoso, P.R., 1993. Saline water intrusion in Mexico. Transactions on Ecology and the

Environment. 2. WIT, Southampton. Cary, L., Petelet-Giraud, E., Bertrand, G., et al., 2015. Origins and processes of groundwater salinization in the urban coastal aquifers of Recife (Pernambuco, Brazil): a multiisotope approach. Sci. Total Environ. 530-531:411-429. http://dx.doi.org/10.1016/j. scitotenv.2015.05.015. CONAGUA (Comisión Nacional del Agua), 1997. Censo de captaciones de aguas subterráneas y colección de datos geohidrológicos en la zona La Paz-El Carrizal, BCS. Subdirección General Técnica. Gerencia de Aguas Subterráneas. CONAGUA (Comisión Nacional del Agua), 2001. Estudio de caracterización y modelación

de la intrusión marina en el acuífero de La Paz (B.C.S. México D.F.). CONAGUA (Comisión Nacional del Agua), 2002. Determinación de la disponibilidad de agua en el acuífero de La Paz, Estado de Baja California Sur. Dirección General Técnica. Gerencia de Aguas Subterráneas. Subgerencia de Evaluación y Modelación Hidrogeológica.

CONAGUA (Comisión Nacional del Agua), 2010a. Manejo integrado de las aguas subterráneas en el acuífero La Paz, Baja California Sur. Organismo de Cuenca Península de Baja California. Dirección Local Baja California Sur. Subdirección de Asistencia Técnica Operativa (La Paz, Baja California Sur). CONAGUA (Comisión Nacional del Agua), 2010b. Situación actual y posibles escenarios de intrusión salina en el acuífero de La Paz, Baja California Sur, y su aprovechamiento como fuente de desalación para abastecimiento de agua potable. Informe Final, Mexico City.

Cruz-Falcón, A., 2007. Caracterización y diagnóstico del acuífero de La Paz, BCS, mediante estudios geofísicos y geohidrológicos. (PhD thesis). Instituto Politécnico Nacional. Centro Interdisciplinario de Ciencias Marinas, La Paz, Baja California Sur.

Cruz-Falcón, A., Vázquez-González, R., Ramírez-Hernández, J., Salinas-González, F., Nava-Sánchez, E.H., Troyo-Diéguez, E., 2010. Depth estimation to crystalline basement in the valley of La Paz, Baja California Sur, Mexico. Geofis. Int. 49 (4), 213-224.

Cruz-Falcón, A., Vázquez-González, R., Ramírez-Hernández, J., Nava-Sánchez, E.H., Troyo-Diéguez, E., Rivera-Rosas, J., Vega-Mayagoitia, J.E., 2011. Precipitación y recarga en la cuenca de La Paz, BCS. Universidad y Ciencia 27 (3), 251-263.

Cruz-Falcón, A., Ramírez-Hernández, J., Vázquez-González, R., Nava-Sánchez, E.H., Troyo-Diéguez, E., Fraga-Palomino, H.C., 2013. Estimación de la recarga y balance hidrológico del acuífero de La Paz, BCS, México. Universidad y ciencia 29 (1), 87-100.

Escolero, O., Torres-Onofre, S., 2007. Análisis de la intrusión de agua de mar en el acuífero de La Paz (México). Bol. Geol. Min. 118, 637-648.

Faure, G., Mensing, T.M., 2005. Isotopes: Principles and Applications. third ed. John Wiley & Sons, Hoboken, New Jersey (897 pp.).

Fetter Jr., C.W., 2001. Applied hydrogeology. third ed. Prentice-Hall Publishing Co, New York (691 pp.).

Fidelibus, M.D., 2003. Environmental tracing in coastal aquifers: old problems and new solutions. In: López-Geta, J.A., Gómez, J.D., de la Orden, J.A., Ramos, G., Rodríguez, L. (Eds.), Coastal Aquifers Intrusion Technology: Mediterranean Countries. Serie: Hidrogeología y Aguas Subterráneas n° 8. Instituto Geológico y Minero de España, Madrid (Spain) (335 pp.).

Fletcher, J.M., Kohn, B.P., Foster, D.A., Gleadow, A.J.W., 2000. Heterogeneous cooling and exhumation of the Los Cabos block, southern Baja California: evidence from fission-track thermochronology. Geology 28,107-110.

Flury, M., Prapitz, A., 1993. Bromide in the natural environment: occurrence and toxicity. J. Environ. Qual. 22, 747-758.

Goldberg, S., Lesch, S.M., Suarez, D.L., 2000. Predicting boron adsorption by soils using soil chemical parameters in the constant capacitance model. Soil Sci. Am. Soc. J. 64, 1356-1363.

Gupta, U.C., Jame, Y.W., Campbell, C.A., Leyshon, A.J., Nicholaichuk, W., 1985. Boron toxic-ity and deficiency: a review. Can. J. Soil Sci. 65, 381 -409.

Hausback, B.P., FrizzellJr., V.A., 1984. Cenozoic volcanic and tectonic evolution ofBajaCal-ifornia Sur, Mexico. Geology of the Baja California Peninsula, Pacific Section. 39. Soc. of Economic Paleontologists and Mineralogists, Los Angeles, Cal. U.S.A, pp. 219-236.

Helsel, D.R., Hirsch, R.M., 2002. Statistical methods in water resources. Techniques of Water Resources Investigations, Book 4. US Geological Survey, Reston, VA (Chapter A3). (510 pp.).

Hemming, N.G., Hönisch, B., 2007. Boron Isotopes in Marine Carbonate Sediments and the pH of the Ocean. In: Hillaire-Marcel, C., De Vernal, A. (Eds.), Developments in Marine Geology. Proxies in Late Cenozoic Paleoceanography 1. Elsevier BV, Amsterdam, The Netherlands (717-734 pp.).

Hernández-Antonio, A., Mahlknecht, J., Tamez-Meléndez, C., Ramos-Leal, J., Ramírez-Orozco, A., Parra, R., Ornelas-Soto, N., Eastoe, C.J., 2015. Groundwater flow processes and mixing in active volcanic systems: the case of Guadalajara (México). Hydrol. Earth Syst. Sci. 19:3937-3950. http://dx.doi.org/10.5194/hess-19-3937-2015.

INEGI (Instituto Nacional de Estadística y Geografía), 2010. Censo de Población y Vivienda. p. 2010.

Ingraham, N.L., Caldwell, E.A., Verhagen, B.Th., 1998. Arid catchments. In: Kendall, C., McDonnell, J.J. (Eds.), Isotope Tracers in Catchment Hydrology. Elsevier BV, Amsterdan, The Netherlands (435-466 pp.).

Kehew, A.E., 2001. Applied chemiCal Hydrogeology. Prentice Hall, Upper Saddle River, N.J. (368 pp.).

Kendall, C., Caldwell, E.A., 1998. Fundamentals of isotope geochemistry. In: Kendall, C., McDonnell, J.J. (Eds.), Isotope Tracers in Catchment Hydrology. Elsevier BV, Amsterdan, The Netherlands (435-466 pp.).

Kutzschbach, M., Wunder, B., Meixner, A., Wirth, R., Heinrich, W., Franz, G., 2016. Jeremejevite as a precursor phase for the formation of olenitic tourmaline: an experimental and boron isotope study. Eur. J. Mineral. (accepted for publication. DOI: EJM-S-16-00025).

Langman, J.B., Ellis, A.S., 2010. A multi-isotope (sD, 81SO, 87Sr/86Sr, and 8"B) approach for identifying saltwater intrusion and resolving groundwater evolution along the Western Caprock Escarpment of the Southern High Plains, New Mexico. Appl. Geochem. 25:159-174. http://dx.doi.org/10.1016/j.apgeochem.2009.11.004.

ARTICLE IN PRESS

14 J. Mahlknecht et al. / Science of the Total Environment xxx (2017) xxx-xxx

Langmuir, D., Melchior, D., 1985. The geochemistry of Ca, Sr. Ba and Ra sulfates in some deep brines from the Paulo Duro Basin, Texas. Geochim. Cosmochim. Acta 49, 2423-2432.

Marín, L.E., 2002. Perspectives on Mexican groundwater resources. Ground Water 40, 570-571.

McNutt, R.H., 2000. Strontium isotopes. In: Cook, P., Herczeg, A. (Eds.), Environmental Tracers in Groundwater Hydrology. Kluwer Publisher, Boston, Dorderecht, London, pp. 233-260.

Menció, A., Mas-Pla, J., Otero, N., Regas, O., Boy-Roura, M., Puig, R., Bach, J., Doménech, C., Zamorano, M., Brusi, D., Folch, A., 2016. Nitrate pollution of groundwater; all right..., but nothing else? Sci. Total Environ. 539:241-251. http://dx.doi.org/10.1016/j. scitotenv.2015.08.151.

Meredith, K., Moriguti, T., Tomascak, P., Hollins, S., Nakamura, E., 2013. The lithium, boron and strontium isotopic systematics of groundwaters from an arid aquifer system: implications for recharge and weathering processes. Geochim. Cosmochim. Acta 112: 20-31. http://dx.doi.org/10.1016Zj.gca.2013.02.022.

Merkel, B.J., Planer-Friedrich, B., 2008. Grundwasserchemie. Praxisorientierter Leitfaden zur numerischen Modellierung von Beschaffenheit, Kontamination und Sanierung aquatischer Systeme. Springer, Berlin, Heidelberg http://dx.doi.org/10.1007/978-3-540-87469-0.

Monzalvo, M., 2010. Simulación hidrodinámica del acuífero La Paz y su aprovechamiento como fuente de desalación. Tesis de Maestría en Ingeniería. Universidad Nacional Autónoma de México. Programa de Maestría y Doctorado en Ingeniería. Facultad de Ingeniería (196 pp.).

Moss, S.A., Nagpal, N.K., 2003. Ambient water quality guidelines for boron. Water Protection Section. Ministry of Water, Land and Air Protection. ISBN: 0-7726-5048-9 (Technical report, 113 pp.).

Munguía, L., González, M., Mayer, S., Aguirre, A., 2006. Seismicity and state of stress in the La Paz-Los Cabos region, Baja California Sur, Mexico. Bull. Seismol. Soc. Am. 96 (2), 624-636.

Nakano, T., Tayasu, I., Yamada, Y., Hosono, T., Igeta, A., Hyodo, F., Ando, A., Saitoh, Y., Tanaka, T., Wada, E., Yachi, S., 2008. Effect of agriculture on water quality of Lake Biwa tributaries, Japan. Sci. Total Environ. 389:132-148. http://dx.doi.org/10.1016/j. scitotenv.2007.08.042.

Otero, N., Vitoria, L., Soler, A., Canals, A., 2005. Fertiliser characterisation: major, trace and rare earth elements. Appl. Geochem. 20:1473-1488. http://dx.doi.org/10.1016/j. apgeochem.2005.04.002.

Paces, J.B., Wurster, F.C., 2014. Natural uranium and strontium isotope tracers of water sources and surface water-groundwater interactions in arid wetlands - Pahranagat Valley, Nevada, USA. J. Hydrol. 517:213-225. http://dx.doi.org/10.1016/johydrol. 2014.05.011.

Parkhurst, D.L., Appelo, C.A.J., 2013. A Computer Program for Speciation, Batch-reaction, One-dimensional Transport, and Inverse Geochemical Calculations. PHREEQC Interactive. USGS. United States Geological Survey.

Pennisi, M., Bianchini, G., Muti, A., Kloppmann, W., Gonfiantini, R., 2006. Behaviour of boron and strontium isotopes in groundwater-aquifer interactions in the Cornia-Plain (Tuscany, Italy). Appl. Geochem. 21:1169-1183. http://dx.doi.org/10.1013/j. apgeochem.2006.03.001.

Petelet-Giraud, E., Négrel, P., Casanova, J., 2003. Variability of 87Sr/86Sr in water draining granite revealed after a double correction for atmospheric and anthropogenic inputs. Hydrol. Sci. J. 48 (5), 729-742.

Petelet-Giraud, E., Négrel, P., Aunay, B., Ladouche, B., Bailly-Comte, V., Guerrot, C., Flehoc, C., Pezard, P., Lofi, J., Dörfliger, N., 2016. Coastal groundwater salinization: focus on the vertical variability in a multi-layered aquifer through a multi-isotope

fingerprinting (Roussillon Basin, France). Sci. Total Environ. 566-567:398-415. http://dx.doi.org/10.1016/j.scitotenv.2016.05.016.

Plummer, N., Busby, J., Lee, R., Hanshaw, B., 1990. Geochemical modelling of the Madison aquifer in parts of Montana, Wyoming, and South Dakota. Water Resour. Res. 26, 1981-2014.

Pronatura, 2010. Diagnóstico de la Cuenca de La Paz. Reporte Final del Convenio Niparaja - Pronatura Noroeste "Diálogos del Agua". Programa Agua y Humedales, Pronatura Noroeste, Agosto 2010.

Puig, R., Soler, A., Widory, D., Mas-Pla, J., Doménech, C., Otero, N., 2016. Characterizing sources and natural attenuation of nitrate contamination in the Baix Ter aquifer system (NE Spain) using a multi-isotope approach. Sci. Total Environ. 5xx:999. http://dx. doi.org/10.1016/j.scitotenv.2016.11.206.

Rosales-Ramírez, T.T., 2012. Reconocimiento químico e isotópico del acuífero costero de La Paz, B.C.S. Evaluación de una posible intrusión marina. Centro de Investigación Científica y de Educación Superior de Ensenada, Baja California. Universidad Autónoma de Baja California, México (M.Sc. Thesis. 140 pp.).

Schaaf, P., Bohnel, H., Pérez-Venzor, JA, 2000. Pre-Miocene palaeogeography of the Los Cabos Block, Baja California Sur: geochronological and palaeomagnetic constraints. Tectonophysics 318, 53-69.

SEMARNAT (Secretaría de Medio Ambiente y Recursos Naturales), CONAGUA (Comisión Nacional del Agua), 2014s. Estadísticas del agua en México. edición 2014. SEMARNAT and CONAGUA, México, D.F. (242 pp.).

SGM (Servicio Geológico Mexicano, 1999. Carta geológica-minera La Paz, 612-10-11 (Baja California Sur).

Stigter, T.Y., Carvalho Dill, A.M.M., Ribeiro, L., Reis, E., 2006. Impact of the shift from groundwater to surface wáter irrigation on aquifer dynamics and hydrochemistry in a semi-arid region in the south of Portugal. Agric. Water Manag. 85:121-132. http://dx.doi.org/10.1016/j.agwat2006.04.004.

Stumm, W., Morgan, J.J., 1996. Aquatic Chemistry. third ed. Wiley and Sons, New York (1022 pp.).

Tamez-Meléndez, C., Hernández-Antonio, A., Gaona-Zanella, P.C., Ornelas-Soto, N., Mahlknecht, J., 2016. Isotope signatures and hydrochemistry as tools in assessing groundwater occurrence and dynamics in a coastal arid aquifer. Environ. Earth Sci. 75:830 (17 pp.). http://dx.doi.org/10.1007/s12665-016-5617-2.

Tirez, K., Brusten, W., Widory, D., Petelet, E., Bregnot, A., Xue, D., Boeckx, P., Bronders, J., 2010. Boron isotope ratio (d11B) measurements in water framework directive monitoring programs: comparison between double focusing sector field ICP and thermal ionization mass spectrometry. J. Anal. At. Spectrom. 25:964-974. http://dx.doi.org/10. 1039/c001840f.

Vengosh, A., Spivack AJ., 2000. Boron in ground water. In: Cook, P., Herczeg, A. (Eds.), Environmental Tracers in Groundwater Hydrology. Kluwer Publisher, Boston, Dorderecht, London, pp. 479-485.

Venturi, S., Vaselli, O., Tassi, F., Nisi, B., Pennisi, M., Cabassi,J., Bicocchi, G., Rossato, L., 2015. Geochemical and isotopic evidences for a severe anthropogenic boron contamination: a case study from Castelluccio (Arezzo, central Italy). Appl. Geochem. 63: 146-157. http://dx.doi.org/10.1016/j.apgeochem.2015.08.008.

Vitoria, L., Otero, N., Soler, A., Canals, A., 2004. Fertilizer characterization: isotopic data (N, S, O, C, and Sr). Environ. Sci. Technol. 38:3254-3262. http://dx.doi.org/10.1021/ es0348187.

WHO (World Health Organization, 2011. Guidelines for Drinking-water Quality. fourth ed. World Health Organization (564 pp.).